A Hand Book of Media and Communication Research

1111234567891012341115111678920123456789301234567894012345678950111

A Handbook of Media andCommunication Research

‘This Handbook takes on a full range of communication research approaches with an intelli-gence, skill, and agility not found in methodology texts.’

Vincent Mosco, School of Journalism and Communication, Carleton University, Canada

‘A Handbook of Media and Communication Research presents an excellent overview of andintroduction to the concepts and elements within major quantitative and qualitative researchprocesses.’

Marit Bakke, Department of Media Studies, University of Bergen, Norway

‘An authoritative, stimulating and rigorous survey of diverse research traditions in media andcommunications. The emphasis on identifying the potential for convergence across these tradi-tions is both original and welcome.’

Sonia Livingstone, London School of Economics and Political Science

A Handbook of Media and Communication Research presents qualitative as well as quan-titative approaches to the analysis and interpretation of media, covering perspectives fromboth the social sciences and the humanities. Combining practical approaches and theoreti-cal concerns, the Handbook offers a comprehensive review of earlier research and a set ofguidelines for how to think about, plan and carry out studies of media in different socialand cultural contexts.

The Handbook comprises three main elements:

• a historical account of the development of key concepts and approaches• a systematic section covering media production, texts and audience as well as

the wider social, cultural and global contexts of media, and the changing role of computer-mediated communication

• a practical element taking readers through the stages of the research process as carriedout in student projects.

Written by internationally acknowledged specialists in each area and supported throughoutby keywords, up-to-date references and graphic models, the Handbook will be a standardreference work for students and researchers in the field of media, communication and cul-tural studies. Contributors include: Barrie Gunter, Stig Hjarvard, Klaus Bruhn Jensen, PeterLarsen, Amanda Lotz, Graham Murdock, Horace Newcomb, Paddy Scannell, Kim ChristianSchroder, Gaye Tuchman.

Klaus Bruhn Jensen is Professor in the Department of Film and Media Studies, Universityof Copenhagen and Adjunct Professor at the University of Oslo. His previous publicationsinclude A Handbook of Qualitative Methodologies for Mass Communication Research(coeditor, Routledge, 1991) and News of the World: World Cultures Look at TelevisionNews (editor, Routledge, 1998).

A Handbook of Media andCommunication ResearchQualitative and quantitative methodologies

Edited by

Klaus Bruhn Jensen

1111234567891012341115678920123456789301234567894012345678950111 London and New York

•T

aylor & Francis Group

•

RO

UTLEDGE

First published 2002by Routledge11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canadaby Routledge29 West 35th Street, New York, NY 10001

Routledge is an imprint of the Taylor & Francis Group

© 2002 Klaus Bruhn Jensen for selection and editorial; individual chapters to their authors

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Datahas been applied for

ISBN 0–415–22514–0 (hbk)ISBN 0–415–22588–4 (pbk)

This edition published in the Taylor & Francis e-Library, 2002.

ISBN 0-203-465105 Master e-book ISBNISBN 0-203-773349 (Glassbook Format)

List of illustrations viiNotes on contributors ixPreface xi

1 Introduction: the state of convergence in media and communication research 1Klaus Bruhn Jensen

PART 1HISTORY: SOURCES OF MEDIA AND COMMUNICATION RESEARCH

2 The humanities in media and communication research 15Klaus Bruhn Jensen

3 Media, culture and modern times: social science investigations 40Graham Murdock

PART 2SYSTEMATICS: PROCESSES OF MEDIATED COMMUNICATION

Media organizations

4 The production of media fiction 62Horace Newcomb and Amanda Lotz

5 The production of news 78Gaye Tuchman

6 The study of international news 91Stig Hjarvard

1111234567891012341115111678920123456789301234567894012345678950111

Contents

Media texts

7 Discourses of fact 98Kim Christian Schrøder

8 Mediated fiction 117Peter Larsen

Media audiences

9 Media effects: quantitative traditions 138Klaus Bruhn Jensen

10 Media reception: qualitative traditions 156Klaus Bruhn Jensen

Media contexts

11 Contexts, cultures and computers: the cultural contexts of mediated communication 171Klaus Bruhn Jensen

12 History, media and communication 191Paddy Scannell

PART 3PRACTICE: SCIENTIFIC APPROACHES AND SOCIAL APPLICATIONS

13 The quantitative research process 209Barrie Gunter

14 The qualitative research process 235Klaus Bruhn Jensen

15 The complementarity of qualitative and quantitative methodologies in media and communication research 254Klaus Bruhn Jensen

16 The social origins and uses of media and communication research 273Klaus Bruhn Jensen

References 294Index 326

Contentsvi

FIGURES

1.1 Media in the structuration of society 41.2 Time-in culture and time-out culture 51.3 A model of social spheres 71.4 The stages of communication 81.5 The levels of communication 91.6 A model of media, communication and culture 101.7 An anatomy of media and communication research 112.1 A brief chronology of human communication 182.2 The hermeneutic circle 212.3 The process of semiosis 242.4 Two levels of signification 262.5 Four models of meaning 305.1 The news organization in a field of social forces 805.2 Sources and news media as interconnected organizations 896.1 Four approaches to the study of international news 927.1 Dimensions of critical discourse analysis 1077.2 Comparison of different approaches to qualitative discourse analysis 1108.1 The actant model 1278.2 A model of narrative communication 1298.3 Eight levels of narration 1308.4 The visual code of television news 1319.1 The ‘stages’ of mediated communication, as defined by audience research

traditions 1399.2 Examples of differential rates of the diffusion of innovations 1419.3 Dimensions of campaign objectives and effects 1479.4 Chart of values among Danish respondents according to RISC principles (1995) 153

10.1 Milestones of media and communication research 15710.2 Two varieties of reception study 16610.3 Three flows of media use 16911.1 Three media flows in and between countries 17511.2 Three types of interactivity 18511.3 Four modes of communication 18511.4 Horizontal and vertical intertextuality 18711.5 Three levels of articulation 189

1111234567891012341115111678920123456789301234567894012345678950111

Illustrations

III.1 Six prototypical empirical methods 20713.1 Types of validity 21313.2 Advantages and disadvantages of different forms of survey administration 21613.3 The normal distribution curve 23214.1 The role of language in qualitative methodologies 24114.2 Examples of documents relating to media production and reception 24514.3 Three variants of qualitative data analysis 25215.1 Two paradigms of social and cultural research 25515.2 Six levels of empirical research 25815.3 Dimensions of validity and reliability 26715.4 Three domains of reality, incorporating three types of phenomena 27015.5 Empirical microcosms, theoretical macrocosms 27116.1 Types of media research organization 283

PLATES

7.1 BP advertisement 1128.1 The introduction to The Big Sleep 1228.2. Bellour’s (1973) shot-to-shot analysis of a segment from The Big Sleep 124

TABLES

7.1 Representation of protagonists on Review of the Year 1998, BBC1 1998 10113.1 Summary of a regression analysis for variables predicting four measures of

homicide newsworthiness 234

ANALYSIS BOXES

4.1 Field visit to the production set of Any Day Now (1999) 707.1 Content analysis of Review of the Year 1998, BBC1 1998 1017.2 Discourse analysis of advertisement for BP 1129.1 Correspondence analysis of living conditions, lifestyles and media use 153

12.1 Historical research on archival data 20313.1 Surveys and experiments compared: the case of cultivation research 22913.2 Testing for statistical significance: chi-square 23314.1 Discourse analysis of qualitative interview data 24916.1 The signs of science 27516.2 Ten rules for empirical student projects 291

RESOURCE BOXES

1.1 General reference works and journals 312.1 National and international varieties of media studies 19016.1 Histories of media and communication research as a field 28016.2 Key studies and reference works for individual media 288

Illustrationsviii

Barrie Gunter is Professor of JournalismStudies and Director of Research, Departmentof Journalism Studies, University of Sheffield.He is a chartered psychologist who has workedin the media sector as an audience researcherand has written 40 books and around 200 otherpublications on a range of media, marketing,and psychological topics.

Stig Hjarvard, Ph.D., Professor, Department of Film and Media Studies, University ofCopenhagen. Head of the research program‘Global Media Cultures’ (http://www.global.media.ku.dk). Recent books are Internationaletv-nyheder [International TV news] (1995), Tv-nyheder i konkurrence [TV news in compe-tition] (1999), Audiovisual Media in Transition(co-edited with Thomas Tufte, Sekvens, 1998)and News in a Globalized Society (editor,2001).

Klaus Bruhn Jensen is Professor in the Depart-ment of Film and Media Studies, University of Copenhagen, and Adjunct Professor at theUniversity of Oslo. His work has addressedinterdisciplinary theories and methodologieswithin media studies, with a special emphasison reception analysis. Earlier publicationsinclude A Handbook of Qualitative Methodol-ogies for Mass Communication Research (co-editor, 1991) and News of the World: WorldCultures Look at Television News (editor,1998). His current work focuses on computer-mediated communication, and its implicationsfor the field of ‘media’ and ‘communication’research.

Peter Larsen, Professor at the Department ofMedia Studies, University of Bergen, Norway.He is the author of books and articles on semi-otics, rhetoric, music in the visual media,textual analysis, and text theory in connectionwith film, television, and other forms of visualcommunication.

Amanda D. Lotz (Ph.D., University of Texas atAustin) is a Mellon postdoctoral fellow andinstructor in the Program in Film and MediaStudies and Program in American CultureStudies at Washington University in St. Louis.Her research focuses on the representation ofwomen and feminism on television, with atten-tion to emerging feminist theories such as post-feminism and third-wave feminisms, and theU.S. institutional environment of the post-network era.

Graham Murdock graduated from the LondonSchool of Economics and went on to doresearch at the Centre for Mass Communica-tions Research at Leicester University, beforejoining the interdisciplinary department ofSocial Sciences at Loughborough in 1990. Hehas been a visiting professor at the Universitiesof Bergen, Brussels, California, Mexico City,and Stockholm, and his work has been trans-lated into thirteen languages. His most recentbooks include, as co-author, Researching Com-munications (1999), and as co-editor, Tele-vision Across Europe (2000).

Horace Newcomb is Lambdin Kay Distin-guished Professor and Director of the PeabodyAwards Program at the University of Georgia,

1111234567891012341115678920123456789301234567894012345678950111

Contributors

and was, until recently, the F. J. Heyne Cen-tennial Professor in Communication at theUniversity of Texas at Austin. He is co-author,with Robert S. Alley, of The Producer’s Medium (1983). He is also author of TV: TheMost Popular Art (1974), editor of six editionsof Television: The Critical View and editor ofThe Museum of Broadcast CommunicationEncyclopedia of Television (1997).

Paddy Scannell is Professor and Head ofResearch in the School of Communication andCreative Industries at the University of West-minster. He is a founding editor of Media,Culture and Society, author of Radio, Tele-vision and Modern Life (1996), and co-author,with David Cardiff, of A Social History ofBritish Broadcasting, 1922–1939 (1991).

Kim Christian Schrøder is Professor of Com-munication at Roskilde University, Denmark.

His research has dealt with discourse analysisand reception analysis of advertising and tele-vised serial fiction. He has published widely onthe theoretical and methodological aspects ofqualitative audience research. His currentresearch deals with political discourses and themedia in a combined text/audience perspective.

Gaye Tuchman is Professor at the University ofConnecticut, and has specialised in researchwithin the Sociology of Culture, Gender, andTheory. She is the author of many publicationsin these areas, such as Making News (1978),Edging Women Out: Victorian Novelists, Pub-lishers, and Social Change (1989, co-author)and a contribution to the first edition of Hand-book of Qualitative Research (1994). Currentresearch interests include special education andinvisible differences in children.

Contributorsx

This book covers the state of media and com-munication research – its development, presentstatus, and future potential. It is addressed tostudents, researchers and media professionalsseeking an in-depth treatment of the field.

In recent years, media studies have gonethrough a process of convergence betweensocial sciences and humanities, quantitative andqualitative approaches. This book presents thediverse theoretical sources of current mediastudies and provides examples of differentresearch techniques. It also outlines the profileof this academic field, as it relates to the rest ofthe academy and to contemporary society.

To serve as an accessible yet comprehensivehandbook, the volume includes a number offeatures:

• examples of the main types of media analy-sis, including production research, textualanalysis and audience studies;

• reviews and comparisons of the centraltraditions of theory and methodology;

• resources and extensive references for theplanning of empirical research projects;

• keywords and cross-references;• abstracts, as well as figures and tables sum-

marizing the main points of each chapter.

In preparing the volume, I have had the priv-ilege of cooperating with a number of com-petent and generous people. First of all, I amgrateful to the contributors to the volume, whoagreed to join me in the process of develop-ing this reference work. Simultaneously, I havebenefited from many discussions with, andsuggestions from, colleagues at the Universityof Copenhagen and at the University of Oslo.

While drafting the text, I have drawn muchinspiration as a member of two research pro-grams: Global Media Cultures (1999 to 2001)at the University of Copenhagen (http://global.media.ku.dk) and DIWA (1999 to 2003)(Design and use of Interactive Web Applica-tions – http://www.diwa.dk), a joint project of four Danish universities. Svein Østerud hasbeen a continually constructive partner indebates on methodology for more than tenyears. Special thanks are due to Peter Dahlgrenand Søren Kjørup who both took the time toread and offer constructive criticisms on earlierdrafts of several chapters.

My deepest thanks go to Ghita – my wife,my friend, and a real human being.

Klaus Bruhn JensenCopenhagen, April 2001

NOTE

Key concepts and discussions of key terms areindicated by a marginal note beside their firstplace of mention in the text.

The symbol � in the text indicates a crossreference to the preceding text which can befound below.

The symbol � at the foot of a column indi-cates the cross reference linked to its mentionin the above column.

While the publishers have made every effort tocontact copyright holders of material used, theywould be grateful to hear from any they mayhave been unable to locate in order to rectifyany omissions in subsequent printings of thisvolume.

1111234567891012341115678920123456789301234567894012345678950111

Preface

MEDIA, STRUCTURE AND AGENCY

At least since the self-consciously titled ‘Fermentin the Field’ issue of the Journal of Communi-cation (1983), there has been a recognitionwithin media and communication research thatthe diverse theoretical and methodologicalsources of the field, in the social sciences and inthe humanities, hold a significant potential forconsolidation through integration. Toward thisend, one comprehensive conceptual frameworkis available in the work of Giddens (1984), evenif its particular relevance for media remains tobe developed (for assessments, see Bryant andJary 1991; Held and Thompson 1989). Hisstructuration theory is, first and foremost, ameta-theory which seeks to move both empiri-cal and theoretical studies beyond certainentrenched dualisms from more than a centuryof social and cultural research, including sub-jectivist or objectivist, interpretive or causal,hermeneutic or materialist, micro- or macro-approaches to society and culture.

The key to Giddens’ integrative move is thenotion of a ‘duality of structure,’ which defineshuman agency and social structure each as anenabling condition of the other. Human agency,accordingly, is not the manifestation of any freewill, as exercised by individuals or collectiv-ities, nor is social structure a set of external con-straints on their action. Instead, social subjectsand social systems must be seen as continuallyreproducing and, to a degree, reforming eachother, and they interact, not as abstract prin-ciples, but in concrete practices and contexts:‘structure exists . . . only in its instantiations insuch practices and as memory traces orientingthe conduct of knowledgeable human agents’(Giddens 1984: 17). To exemplify, the pressconsists simultaneously of its structural prop-erties – its economic, legal, technological, aswell as cultural-conventional permanence – andof the myriad activities of journalists, advertis-ers, regulators, and audiences who both main-tain and contest these properties. Like othersocial institutions, the press, and the media as

1111234567891012341115678920123456789301234567894012345678950111

IntroductionThe state of convergence in media and communication research

Klaus Bruhn Jensen

1

• a reassessment of the field with reference to the three concepts of media, social structure, and humanagency

• a distinction between media of three degrees: speech, technologically reproduced communication, andcomputer-mediated communication

• a comparison of culture in the narrow sense of aesthetic works and in the broad sense of a whole way oflife

• a definition of modern media as institutions-to-think-with• a presentation of a communication model which integrates traditional transmission and ritual models• outline of the handbook, its elements, and interrelations.

duality ofstructure

such, are not only reinterpreted, but re-enactedon a daily basis.

In order to explain how some measure ofstability emerges from social flux, structurationtheory places special emphasis on the concept of reflexivity. (For additional discussions ofreflexivity and the meaning–action nexus, seeBeck 1999: 109–132; Bourdieu 1977.) Giddensdescribes reflexivity as a general interpretivefaculty that enables humans to ascribe mean-ing to their transactions with others, both in one’s most intimate relations and in encounterswith institutions of political or religious auth-ority. Importantly, this meaning may not bearticulated in any explicit form, neither in dis-course nor even in consciousness. ‘“Reflexivity”should be understood not merely as “self-consciousness” but as the monitored characterof the ongoing flow of social life’ (Giddens1984: 3).

The point is that reflexivity orients peopleand allows them to act, to go on, and that itwould be possible for them, as a rule, to justifytheir actions if they were challenged. In mostcases, however, people will simply go abouttheir business, and will be able to coordinate itwith that of others to a remarkable degree, byrelying on shared, implicit assumptions, whatGiddens terms their practical consciousness.This is in contrast to discursive consciousness,a focused form of intentionality, that can bemobilized in response to one’s own doubts orto alternatives advanced by others. (The thirdelement of Giddens’ model of consciousness isthe unconscious, which is largely comparableto its original Freudian version.) Listening tothe radio, for instance, often serves the practi-cal purpose of monitoring a morning routineuntil it is time for members of the householdto leave for work or school, but a particularnews item about public transport or road con-ditions may shift the listeners’ attention into adiscursive key because this might requireactions out of the ordinary.

The media play a special role, both as meansof reflexivity and as sources of social struc-turation. Giddens recognizes this, in part, byone of his central distinctions between techno-logically mediated and non-mediated socialinteraction or, in his terminology, system inte-

gration and social integration. In contrast tosocial integration, which refers to local, face-to-face interaction, system integration is definedas ‘reciprocity between actors or collectivitiesacross extended time–space, outside conditionsof co-presence’ (Giddens 1984: 377). Oneoutcome of such mediated interaction is a ‘dis-embedding’ of people from their traditionalrelations and environments, and a ‘re-embedd-ing’ into different social formations. Thereference is primarily to the modern era, whichis characterized by the coordination of eco-nomic, political, and cultural activity acrossgreat distances and time differences, whatGiddens calls ‘time–space distanciation,’� in-creasingly on a global and round-the-clockscale (see also Giddens 1990, 1991). Whilesocial integration is performed, above all, byoral communication, system integration hasdepended on shifting technologies and insti-tutions of communication, from handwrittenadministrative and accounting systems tobroadcasting and beyond.

Nevertheless, Giddens has paid surprisinglylittle attention to the media as a condition ofmodernity. ‘Signification,’ including its techno-logically mediated forms, is one of his threedimensions of social systems, the other twobeing domination, namely the exercise of powerthrough political and economic institutions, andlegitimation as exercised typically by legal insti-tutions (Giddens 1984: 29). But the pervasivecommunicative aspects of each of these, and ofpractically any type of social action, have re-mained a blindspot in Giddens’ work (see Jensen1995; Silverstone 1999; Thompson 1995).

A meta-theoretical framework that treatscommunication not as incidental, but as a nec-essary constituent of social life, is relevant notjust to the media field, but to theories of societyand culture as such. In order to move beyondthe lingering dualism of ‘the duality of struc-ture,’ it is helpful to introduce a third categoryof medium, on a par with agency and structure.In a historical and anthropological perspective,media include spoken and body language,scratch notes and government administrations,

Introduction2

reflexivity

practicalconsciousness

discursiveconsciousness

the uncon-scious

� time–space distanciation, disembedding, and re-embedding – Chapter 11, p. 182

broadcasting and the Internet. Their commoncharacteristic is that they serve to orient humanagency as it enacts social structure, partly at thelevel of practical consciousness and everydayroutines.

MEDIA OF THREE DEGREES

The growth of computer-supported communi-cation has recently presented ‘mass’ mediaresearch with a need to reconsider its central

objects of study. Computers can integrate pre-vious media technologies in a single meta-medium (Kay and Goldberg 1999 [1977]: 112);to a degree, computers can also simulate em-bodied, interpersonal communication. To indi-cate the scope of this handbook, it is useful todistinguish three prototypes of media:

1 Media of the first degree. The biologicallybased, socially formed resources that enablehumans to articulate an understanding of

Media of three degrees 3

1111234567891012341115678920123456789301234567894012345678950111

RESOURCE BOX 1.1 GENERAL REFERENCE WORKS AND JOURNALS

The following texts provide general resources and overviews for the field of media and com-munication research. The titles cover different traditions, as indicated, and include qualitative aswell as quantitative methodologies.

Encyclopedia• Barnouw 1989 – a multi-volume comprehensive reference work on most aspects of

communication, including mediated and interpersonal forms• Watson and Hill 1999 – a concise reference work.

Abstracts• Communication abstracts – a listing with abstracts and keywords of current research.

Handbooks and textbooks• McQuail 2000 – a solid introduction to positions in the field, with a relative emphasis on

social-scientific traditions• Berger and Chaffee 1987 – a somewhat dated, but still useful overview summarizing work

defining communication studies as a ‘science’• Jensen and Jankowski 1991 – an overview delineating the contributions of qualitative

research, both social-scientific and humanistic, to the media field• Lindlof 1995 – a reference work emphasizing the interpretive legacy in social science and

its relevance for qualitative empirical studies.

Journals• Journal of Communication – since the mid-1970s a central journal in the field,

accommodating both quantitative and qualitative, administrative and critical work• Communication Theory – a more recent addition to the field, covering interpersonal

communication as well, and with important theoretical contributions to the media field• Critical Studies in Media and Communication and Media, Culture and Society – two

representatives of a primarily critical as well as interpretive strand of media research• Journal of Broadcasting and Electronic Media and Journalism and Mass Communication Quarterly

– two representatives primarily of the quantitative,American mainstream of media research• Screen – one of the journals focusing on film (and television), which also includes

contributions with implications for the wider field• New Media and Society – one of several journals currently shaping the area of computer-

mediated communication.

reality, for a particular purpose, and to engagewith others in communication about it. Thecentral example is verbal language, or speech,but additional ones include song and othermusical expression, dance, drama, painting,and creative arts generally. Such media dependon the presence of the human body, and operatein local time–space, often relying on compara-tively simple, mechanical techniques such asmusical instruments and artistic and writingutensils as constitutive elements. (Handwritingpresents a special case, which has supportedcomplex historical communication systems.However, its comparatively inefficient forms ofreproduction and distribution arguably madethis a transitional cultural form (Meyrowitz1994: 54).)

2 Media of the second degree. The technicallyreproduced or enhanced forms of representa-tion and interaction which support communi-cation across space and time, irrespective of thepresence and number of participants (Benjamin1977 [1936]). Early modern examples includedthe standardized reproduction of religious andpolitical texts through the printing press(Eisenstein 1979). In radio talkshows, conver-sation took on new forms, just as acting styles were adapted to cinema and television.Thus, media technologies have performed a ‘re-embedding,’ both of the media of the firstdegree and of people in relation to distantothers, issues, and arenas.�

3 Media of the third degree. The digitallyprocessed forms of representation and interac-tion which reproduce and recombine previousmedia on a single platform. The central currentexample is the networked personal computer.�This ‘interface’ is likely to change substantiallyas the technologies are adapted further to the human senses, and integrated into bothcommon objects and social arrangements. Incertain respects, humans are media; in certainrespects, media can substitute the social rolesof humans.

Figure 1.1 brings media into a revised con-ceptual table of structuration theory. WhereasGiddens (1984: 374) has linked discursiveconsciousness with verbal expressions only, it remains important to examine how the fullrange of media relate to different levels ofconsciousness and forms of culture. Moreover,particularly in contemporary society, media areamong the central social ‘resources’ in Giddens’terms, just as they are vehicles of many of the‘rules’ that inform social interaction. This hand-book focuses on media of the second degree;reviews media of the third degree as a growingfield of social activity and study; and includesdiscussion and references on media of the firstdegree, as they relate to the technologicalmedia. Each of these media types facilitatessocial structuration in specific ways, and theydo so by participating in the production andcirculation in society of meaning, which accu-mulates as culture.

THE DUALITY OF CULTURE

Like studies of society, research on culture hasbeen divided by dualisms, with two dominantdefinitions criss-crossing the humanities andsocial sciences.� On the one hand, culture has

Introduction4

� history of media and communication – Chapter 12

� computer-mediated communication – Chapter 11, p. 182

Agency

• Discursive consciousness• Practical consciousness• the Unconscious

Structure

• Resources– Allocative (Re: objects, goods, material

phenomena)– Authoritative (Re: persons, actors)

• Rules (Re: meaning and sanctioning of socialconduct)

Media

• Media of the first degree• Media of the second degree• Media of the third degree

Figure 1.1 Media in the structuration of society

� the concept of culture – Chapter 11, p. 172

been conceived as representations of reality –texts and other artefacts – which express someprivileged insight, often in an arena such as amuseum that is separate in space and time fromthe rest of social life. This understanding ofculture as entities of meaning and vehicles oftradition was captured in Matthew Arnold’sdefinition (1869), ‘the best which has beenthought and said in the world.’ On the otherhand, culture has come to be understood as thetotality of human expression, artefacts, andforms of interaction, what Raymond Williams(1975 [1958]: 18) summed up as ‘a whole wayof life.’ The two definitions have been associ-ated with a further set of dualisms, including afocus on either text or context, high or popularculture, a normative-critical or descriptive idealof science, and the qualitative interpretation orquantitative measurement of culture.

A next step beyond dualism is to recognizea duality not only of social structure, but ofculture as well. Culture is both product andprocess, and both aspects enter into socialstructuration at large. To begin, the duality ofculture may be illustrated through conceptsfrom the world of sports: time-out and time-in.In basketball and (American) football, forexample, coaches can call for an interval todiscuss strategy with their teams. While tem-porarily suspending the game, the time-outoccurs within and addresses the total time-in of the game. By analogy, an institutionalizedcultural activity such as media use partly sus-pends other activities, but still takes placewithin the everyday and with reference tofamilies, parliaments, and other well-knowninstitutions. In different respects, news, soapoperas, and talkshows offer a cultural forumfor collective reflexivity (Newcomb and Hirsch1984).

Time-out culture places reality on an explicitagenda, as an object of reflexivity, and providesan occasion for contemplating oneself in asocial or existential perspective, perhaps sug-gesting new avenues for agency. In this regard,mediated communication joins other culturalforms, from religious rituals to fine arts. Time-in culture is continuous with, constitutive of,and orients everyday life, thus regeneratingsocial structure. As such, it supplies the often

implicit premises and procedures of social inter-action. Time-out culture prefigures social ac-tion; time-in culture configures social action.

Figure 1.2 illustrates the embedding of time-out within time-in culture and the permeableboundary between the two. Extending Giddens’(1984) terminology, one might say that variouscultural practices enable their participants tocommute between practical and discursiveconsciousness. Importantly, time-in culture andtime-out culture are not separate activities ordiscourses, but simultaneous and complemen-tary aspects of, for instance, media use. Goingto the cinema can be an occasion to reflect onmoral dilemmas in either marriage or business,not only in a generalized fictional universe, butequally in one’s own life. Compared to reflex-ivity in psychotherapy, by which one may move from discursive and practical conscious-ness toward the unconscious, everybody thusmoves back and forth between practical anddiscursive consciousness many times a day.While this recurring movement at the level ofindividual consciousness resembles the inter-change between time-out and time-in culture ata systemic level, it is important to examine indetail how the two processes are intertwinedunder shifting historical circumstances. In

Duality of culture 5

1111234567891012341115678920123456789301234567894012345678950111

time-outculture

time-inculture

culture asaesthetic

representation

culture associal practice

Time-in

Time-out

Figure 1.2 Time-in culture and time-out culture

social systems, reflexivity is commonly dele-gated to institutions of religion, science, andcommunication. In modern social systems, thepractices of reflexivity are increasingly techno-logically mediated.

INSTITUTIONS-TO-THINK-WITH

The modern media are understood in this handbook, most basically, as technologies thatenable reflexivity on a social scale, as theyproduce and circulate meaning in society.Beyond this meta-theoretical framework, one ofthe most applicable substantive theories of themedia field has been provided in JürgenHabermas’ (1989 [1962]) early historical work,which is compatible with Giddens’ systematicapproach, despite differences of opinion andemphasis (e.g., Giddens 1984: 31).

Habermas’ (1989 [1962]) main conclusionwas that the social system of industrialcapitalist democracies may be described as a setof interconnected, but relatively autonomous‘spheres’ (Figure 1.3). The figure notes, to theright, the role of state agencies in providing astable economic and legal frame for social life.To the left, industrial and other private enter-prise or business unfolds in what is termed thesocial sphere, while the intimate sphere is thedomain of family life. The mediating elementof the system is the public sphere, comprisingthe major political and cultural institutions aswell as the press as Fourth Estate.

Whereas, historically, the public sphere hada proactive function in asserting the economicand political rights of the individual, it can besaid, more generally, to negotiate the terms ofcooperation between social agents and thestate. Most importantly, the public sphere ispremised on the ideal of rational, democraticcommunication about the ends and means ofsocial life. While Habermas emphasized the lib-erating, utopian potential of the public sphere,even while deploring its contemporary decline(see also Sennett 1974), later studies have con-tinued to debate the status of the model as ahistorical, systematic, or normative theory ofcommunication (see Calhoun 1992; Mortensen1977; Negt and Kluge 1993 [1972]).

In this context, the model serves two pur-

poses. First, it locates media on a conceptualmap with the central institutions of contempo-rary society. Although Habermas (1989 [1962])departed from early newspapers and literaryclubs in Europe, the public sphere may be seento include media of the first, second, as well asthird degrees. What is commonly at issue, inboth theoretical and normative approaches tomedia, is the nature of the interrelationsbetween the spheres, especially their relativeautonomy and the forces regulating conflictinginterests. Evidently, the current media are gov-erned as much by an economic logic as by aspirit of democratic dialogue, just as, in earlierperiods, religious institutions and patrons ofthe arts set the conditions for cultural produc-tion. This handbook covers different theoriesof, and empirical findings about, the place ofmedia in relation to the other spheres.

Second, the public sphere model offers anillustrative case of how the duality of structure,and of culture, operates. Rather than being aneutral organizational plan or an instance of‘false consciousness,’ the model refers simulta-neously to a structure of social institutions andto social agents’ imagined relation to theseinstitutions. In imagining this configuration,social agents reproduce, or contest, the institu-tional structure. Like the body (Johnson 1987),society is thus present in the human mind as a predisposition to act in particular ways.Because the public sphere model appears toinform the very organization of daily events, itis likely reproduced as common sense, as hege-mony (Gramsci 1971) – ‘a sense of absolutebecause experienced reality beyond which it isvery difficult for most members of the societyto move, in most areas of their lives’ (Williams1977: 110). Nevertheless, the public sphereinstitutions have introduced a potential fortime-out reflexivity, as performed by indi-viduals as well as collectivities.

To sum up, the modern technological mediaas social institutions are embedded in, butenable reflexivity about, the time-in of every-day life. They are institutions-to-think-with.The terminology derives from Claude Lévi-Strauss (1991 [1962]), who spoke of objects-to-think-with within anthropology (see alsoDouglas 1987). Especially animals, that can be

Introduction6

the publicsphere

objects-to-think-with

eaten, become means of classifying and hencemastering reality. It is not so much that theyare ‘good to eat,’ but that they are ‘good tothink (with)’ (Lévi-Strauss 1991 [1962]: 89). Ina different culture, the same object may meansomething else, or may not be considered goodto think with.

Compared to other objects-to-think-with,media technologies have material and structuralfeatures that designate them as specific culturalresources, the key feature being the ‘program-mability’ also of pre-computer media. Unlikeother artefacts, they serve as flexible, program-mable vehicles of meaning in different culturesand historical periods. (In addition, media mayfunction like other objects-to-think-with when,for example, a painting becomes a nationalsymbol or a film genre is taken as symptomaticof cultural decline.) The most common way tolink concretely the domain of media and com-

munication research to wider cultural andsocial processes has probably been throughmodels (McQuail and Windahl 1993; Meyro-witz 1993), being at once means of theoreticalconceptualization and empirical operationali-zation.

THREE MODELS OF COMMUNICATION

In a 1975 benchmark article, James Carey(1989: 15) pinpointed the two communicationmodels which arguably have been premises ofmost previous media research. (Several side-and substreams are documented in later chap-ters.) On the one hand, communication is thetransmission of entities of meaning from asender to a receiver via some contact. Thismodel has typically informed social-scientificapproaches to the field, and was given its classic formulation in Lasswell’s (1966 [1948])

Three models of communication 7

1111234567891012341115678920123456789301234567894012345678950111

Object

Institution

Object

Institution

Intimate sphere

Religion, sexuality,emotion,friendship, etc.

Family

Social sphere

Private economicactivity, productionand sale/purchaseof commodities,including labor

Private enterprisesand stores

Cultural publicsphere

Preaching, art,literature, music,etc.

Organizations,clubs

Political publicsphere

‘Politics’ and ‘theeconomy’,including socialissues

Parliamentaryorgans,representingpolitical parties,and the press

The (agencies ofthe) stateensure(s) thematerialinfrastructure,overall economicstability, lawenforcement, andregulation ofconflicts byeconomic,coercive, legal,and ideologicalmeans

Private sphere Public sphere

StateSociety

Figure 1.3 A model of social spheresSource: Adapted from Habermas 1989 [1962]; Mortensen 1977

transmissionmodel

questions, ‘Who/Says What/In Which Channel/To Whom/With What Effect?’ A comparableformulation was the ‘mathematical’ theory ofcommunication (Shannon and Weaver 1949),which, though initially addressing the transferof messages in engineering terms as signals,became influential across the field, partlybecause Weaver’s portion of the original textrelied on metaphors that conflated notions ofinformation and meaning (Jensen 1995: 8f.).

On the other hand, Carey (1989) reasserteda ritual model of communication as the collec-tive sharing of meaning and tradition and,hence, as a condition of community. The ritualmodel, with a particular emphasis on texts asthe vehicles of shared meaning, can be tracedin much humanistic research.

In a convergence perspective, both scholarlymainstreams can be seen to share two funda-mental concerns about the stages and levels ofcommunication in society, even if they haveconceptualized and operationalized their con-cerns in distinctive ways. Alhough Lasswell’smodel has been widely criticized, both beforeand after Carey (1989), for compartmentalizingthe stages in a topology of the communicativeprocess, the humanities have joined the socialsciences in recognizing that communication isalso a matter of message transmission in stages.One of the most influential humanistic modelsof communication (Figure 1.4) identified muchthe same elements as Lasswell (1966 [1948]),but conceptualized them as functions which, tovarying degrees, manifest themselves withintexts (Jakobson 1960). As a result, the textitself has been operationalized as the centralobject of analysis, leaving issues of production,reception, and context generally to be inferredfrom textual traces as well as supplementaryevidence – much of which is available, not least,in organizational and audience studies from thesocial sciences.

Jakobson’s (1960) model contributed twospecific points to interdisciplinary research(italicized in Figure 1.4). First, communicationdepends on a code, i.e., a register of signs andsymbols, beyond the physical contact or chan-nel of Lasswell (1966 [1948]). Second, thecontext of communication should, for manyanalytical purposes, be conceived literally as

discourse, as a con-text. Both of these points,in fact, have been taken to heart in much socialscience during its recent ‘cultural turn’ (Rayand Sayer 1999).

In addition, the several levels of communi-cation in society which maintain and transformcommunity, as emphasized in the humanistic-textual tradition, have been acknowledged inthe ‘dominant paradigm’ of communicationscience (Gitlin 1978). A standard referencework such as Berger and Chaffee (1987) con-ceptualizes the field according to different sociallevels and contexts of communication, even ifthe individual acts of communication may tendto be operationalized as transmissions. Thelevels of interpersonal (for an overview see, e.g.,Knapp and Miller 1994) and organizational(for an overview see, e.g., Goldhaber andBarnett 1988; Windahl et al. 1992) communi-cation are complementary to that of ‘mass’communication.

Figure 1.5 illustrates the social levels ofcommunication, each involving stages, agents,and differential forms of feedback, and eachoffering opportunities for both humanistic andsocial-scientific approaches, within and acrosslevels. In addition, the figure reminds both trad-itions that mass-mediated communication isone of several types, and that, at least quanti-tatively, other forms of communication may bemore important to the ongoing structuration ofsociety.

Figure 1.6 proposes a way of integrating theritual and transmission views in order to modelthe specific role of technological media. At thecenter of the model are the media (of the secondand third degrees) operating in the publicsphere. These media are the communicators –transmitters – of ideas and worldviews that feedinto the processes of time-out as well as time-

Introduction8

ritual model

stages ofcommuni-

cation

Context

Message

Contact

Code

Addresser Addressee

Figure 1.4 The stages of communicationSource: Adapted from Jakobson 1960

social levels of communi-cation

interpersonaland organiza-tional com-munication

in culture, offering both a ritual sphere ofreflexivity and practical means of coordinatingeveryday activities. The permeable boundariesindicate that the media themselves are part andparcel of social structuration. (The overlapbetween the media rectangle and the time-outcircle is meant to suggest the distinctive role ofmedia as institutions-to-think-with, not thedegree of this overlap.)�

The foundational question for the field,accordingly, is what difference the media make,not just in terms of their ‘effects’ on audiences,but for the rest of the social structure and forhuman agency, culture, and communication.This handbook takes the distinctive feature ofmedia to be the production and circulation ofmeaning in modern societies, enabling collec-tive reflexivity and coordinated action on anunprecedented scale. This implies that themedia themselves occupy the center of interest

in the field – its identity – in a methodologicalsense. The fact that media are at once busi-nesses, aesthetic forms, and cultural resourcesis of theoretical and empirical interest primar-ily to the extent that these features shape themediated production of meaning. Preciselybecause of the complexity of the media asobjects of analysis, the field must rely on arange of theoretical, disciplinary, as well asinterdisciplinary approaches, taking into con-sideration a large periphery of explanatoryfactors that converge on its center.� The finalsection of this introduction explains how thetheoretical, methodological, and empiricalcomponents of the handbook are prioritizedand interrelated.

In review, Carey’s (1989) account of the twodisciplinary models indirectly recognized theirintegrative potential, when he quoted the prag-matist philosopher, John Dewey: ‘Society existsnot only by transmission, by communication,but it may fairly be said to exist in transmis-sion, in communication’ (p. 13f.). In a recon-structive formulation, society exists both bytransmissions and in rituals. In structurationterms, communicative agency serves to bothreproduce and contest social structure throughtransmission of, as well as ritual participationin, meaning.

Carey’s 1975 definition of communication inpublic life as ‘a symbolic process wherebyreality is produced, maintained, repaired, andtransformed’ (Carey 1989: 23) seemed to anti-cipate Giddens’ (1984) concept of ‘the doublehermeneutic.’ Building on Winch (1963), andwith special reference to how social scienceinterprets an already interpreted reality, Giddens notes that such a double hermeneuticexchanges ideas back and forth between ‘themeaningful social world as constituted by layactors and the metalanguages invented by socialscientists’ (p. 384). Classic examples are theworks of Freud and Marx, which have enteredcontemporary common sense. Opinion pollsand policy studies similarly redirect the actionsof both decision-makers and publics.

The modern media perform a continuous

Three models of communication 9

1111234567891012341115678920123456789301234567894012345678950111

Institutional/organizational

Society-wide

Intergroup/association

Intragroup

Interpersonal

Figure 1.5 The levels of communicationSource: Adapted from McQuail 2000: 7

� models – and other forms of representations inresearch – Chapter 16, p. 275

� the media research institution as field or discipline? –Chapter 16, p. 279

the doublehermeneutic

and publicly accessible hermeneutic of sociallife. Media and communication research cancontribute to a double hermeneutic by describ-ing, interpreting, and explaining what differ-ence the modern media make to differentaudiences and sectors of society.

OUTLINE OF THE HANDBOOK

This handbook is divided into three parts, cov-ering the ‘history’ of media and communicationstudies in different disciplines and theoreticaltraditions; the ‘systematics’ of empirical re-search on different stages and contexts of com-munication; and the ‘practice’ of planning,conducting, and applying research. Figure 1.7lays out this structure with reference to the scopeand focus of each part, noting a key premise of each section; there can be no ‘view fromnowhere’ in either communication (Epstein1973) or science (Nagel 1986). Throughout this volume, methodologies are given specialattention, because they encapsulate the ana-lytical operations of the various traditions,along with their theoretical justifications, thusenabling concrete comparisons of the explana-tory value of different forms of research.

• Part 1 – History traces the main explana-tory concepts and approaches that haveinformed media and communication research.While noting the roots of the field in classicalscholarship, such as rhetoric and poetics, andin early sociology, the two chapters, on thehumanities and the social sciences, focus oncontemporary conceptions of such key notionsas culture and communication, interpretationand interaction. By way of introduction, Part Ialso delimits the media field from related cul-tural forms and areas of inquiry. One premiseof the handbook, as outlined in this introduc-tion, is that a convergence of social-scientificand humanistic perspectives has been takingplace particularly since the 1980s. While thisshift has arguably strengthened the overallquality and relevance of the field, the contrib-utors also weigh the evidence regarding prob-lems of convergence and integration.

• Part 2 – Systematics turns from a diachronicto a synchronic view of different ‘schools’ ofresearch. In a systematic perspective, thecurrent schools represent distinctive concep-tualizations of the stages, levels, and contextsof mediated communication. Whereas certaintraditions are incompatible in their epistemo-logical assumptions or research policies, theycan be seen, in the systematic perspective, tocomplement and sometimes overlap each otherin unrecognized ways. One divisive issue hasbeen the degree to which technological, eco-nomic, and other conditioning factors could besaid to determine either the structure or impactof media. Another premise of the handbook isthat such factors exercise a determination in thefirst instance (Hall 1983), setting the limits ofpossible cultural forms, but not a determina-tion in the final instance of either their socialuses or aesthetics. This premise is in keepingwith the dualities of structure and culture. Thenature and degree of structuration in differentaspects of mediated communication are empir-ical issues.

• Part 3 – Practice deals with various types ofempirical research, and also addresses the socialorigins and applications of media studies. Theorientation is to the future, to the extent that

Introduction10

Time-in

Media

Time-out

Figure 1.6 A model of media, communication and culture

convergenceof social-scientific andhumanisticperspectives

determinationin the firstinstance

researchers and students are constantly re-enacting the field as part of a double hermen-eutic that involves both the research institutionand the rest of society. Special emphasis isplaced on the methodological aspects of design-ing and conducting empirical studies, partic-ularly the different procedures or ‘logics’ ofqualitative and quantitative research – and theinterfaces between them. Corresponding to thefirst premise concerning convergence, the lastpremise of the handbook is that a greatermeasure of unification is both possible anddesirable for the field as such. Crucially, unifi-cation is more likely to succeed if it is devel-oped not in the first instance, through a

standardization of the elementary researchinstruments, but in the final instance, througha comparative and meta-theoretical analysis offindings from different traditions, each with its distinctive explanatory value. Examiningresearch as a social practice in itself, the finalchapter returns to the political and historicalcontexts of media and communication studies,from the classic normative theories of the press,to contemporary media policy, and to instru-mental uses in research and development. Likeother sections of the handbook, Part III in-cludes extensive references and boxed informa-tion with resources for further research.

Outline of the handbook 11

1111234567891012341115678920123456789301234567894012345678950111

Scope

Focus

Premise

History

Theoreticalconcepts andanalyticalprocecures from:• Humanities• Social sciences

Past sources ofideas on media,culture, andcommunication

Convergence

Systematics

Empiricalapproaches to thestages and contextsof communication,as defined bydifferent researchtraditions

Presentoperationalizationsand findings

Determination inthe first instance

Practice

Practices ofconducting,justifying, andapplying research

Future uses ofresearch inscientific and socialinstitutions

Unification in thefinal instance

Figure 1.7 An anatomy of media and communication research

unification in the final

instance

1111234567891012341115678920123456789301234567894012345678950111

History

SOURCES OF MEDIA AND COMMUNICATION RESEARCH

Media and communication research has developed from a long heritage of diverse scientific dis-ciplines to address new conditions of communication in modern society. The two chapters in Part I present and assess the contributions of the humanities and the social sciences – the two‘faculties’ or areas of inquiry which have been the main sources of theoretical concepts andanalytical procedures.

As a preview, the following topics indicate the scope and focus of the field in its present configuration:

• Media and communication. The center of interest in the field are the technological media,but because they are examined as means of communication, they lend themselves to conceptsand forms of analysis which have been derived from both oral and literate forms of commu-nication down through history. And, because the media are studied in their social and culturalcontexts, most of the research questions that have been developed by disciplines such as soci-ology and anthropology have proven their relevance for media studies, as well. ‘Media’ and‘communication,’ in addition, are currently being redefined in the light of computer-mediatedcommunication.

• Humanities and social sciences.The modern humanities can be traced from the early nine-teenth century; the social sciences date from the late nineteenth and early twentieth centuries,when they separated institutionally from other ‘human sciences.’ During this entire period, thetwo areas of inquiry have been in contact, and frequently in dialogue. However, it was in thepost-1945 period that specific components of each were brought together in the emergingfield of media studies.

• Field or discipline? It remains an open question whether these contributions have mergedsufficiently to constitute a traditional ‘discipline,’ as defined by its subject matter and consen-sual methods, as well as by its permanent institutional status. Several chapters in the volumeaddress the question, and Chapter 16 elaborates on the current role of media studies in thewider society.

• Convergence.While leaving the question of disciplinary status open, this volume notes that aconvergence between humanistic and social-scientific approaches has been ongoing in recentdecades.The volume describes different elements and stages of this process, and indicates somepossible lines of development for future research.

IPART

A CLASSICAL AGENDA

Contrary to a widespread notion, the humani-ties are not direct descendants of classical Greek philosophy (Kristeller 1961: 3–23). Intheir recognizably modern form, the humanities date from the early nineteenth century, whenuniversities were taking shape as institutions of research, as initially associated with theHumboldt tradition in Germany (Fallon 1980;Rudy 1984). The understanding of knowledgeas a product of research had been preceded by atleast two alternative conceptions of knowledge,either as self-awareness (summed up in theDelphi oracle’s admonition to ‘Know thyself!’)or as traditional learning, administered andpassed on by a class of learned people (Kjørup1996: 31). While the latter two concepts are stillencountered as subtexts, it is the development ofanalytical procedures and conceptual frame-works for research which has occupied human-istic scholars during the immediate ‘prehistory’of media and communication research.

Much of the agenda for this development,to be sure, was inherited from the classics. Theycontinue to be suggestive, less about what tothink than which issues to think about, oneexample being conceptual problems sharedacross theories of knowledge and theories ofcommunication. This chapter, accordingly, firstretraces part of the classical agenda behindcommunication studies, before outlining themain traditions from the history of ideas thatentered into the modern humanities – rhetoric,hermeneutics, phenomenology, and semiotics.These traditions, in turn, informed the discip-lines – from linguistics and literary studies toart history and film studies – which ultimatelyfed into the field of media and communication.The chapter concludes with an assessment ofhow recent interdisciplinary challenges – post-modernism, feminism, and cognitivism – mayredraw the map also of the media field.

A common denominator for the productionof meaning in mediated communication and the production of knowledge in science is

1111234567891012341115678920123456789301234567894012345678950111

The humanities in media andcommunication researchKlaus Bruhn Jensen

2

• a classical agenda: the heritage of philosophy for communication theory• medium theory: the implications of modern media technologies for classic issues of communication and

culture• four theoretical traditions originating in the humanities: rhetoric, hermeneutics, phenomenology, semiotics• humanistic disciplines feeding into the interdisciplinary field of media studies: art history, literary criticism,

linguistics, film studies• current challenges to the humanities as well as to media and communication research: postmodernism,

feminism, cognitivism.

knowledge asa product of

research

intersubjectivity:� How, by what means, is ashared understanding of certain phenomena inreality possible? Aristotle is sometimes creditedwith the first explicit formulation in Westernphilosophy, in De Interpretatione (c. 320 BCE),of the means by which a person can thus inter-act with other people about reality:

Now spoken sounds are symbols of affec-tions in the soul, and written marks symbolsof spoken sounds. And just as written marksare not the same for all men, neither arespoken sounds. But what these are in thefirst place signs of – affections of the soul –are the same for all; and what these affec-tions are likenesses of – actual things – arealso the same.

(Aristotle 1994: 43)

Things, mental impressions, and communica-tive expressions about each of these – thoseelements would be subscribed to by mosttheories of knowledge as well as of communi-cation. But their interrelations, such as theassumption that the reality and the humanexperience to which communication bears wit-ness, are in all circumstances ‘the same,’ havegenerated some of the most longstanding dis-putes across arts and sciences. Research hasdebated the existence of a unified reality, thesocial and cultural variability of experience,and the very relation between reality, the sub-ject, and signs.

One particular implication of Aristotle’sconception for communication theory was thatsigns serve as evidence of what is at least tem-porarily absent (Clarke 1990: 11) – in space, intime, from other people’s purview, and fromone’s own immediate experience. Differentmedia can represent phenomena that are notaccessible through the senses (including the caseof virtual or thought experiments). And certaindurable media enable people to interact in eachother’s absence. Present signs allow for absentrealities or communicators, or both.

If abstract thinking, self-awareness, andsymbolic communication had provided the

minimal conditions for humanitas, for civilizedsocial interaction (Megarry 1995: 48), writingand later technological media introduced apeculiar set of tools for human activity. Theyradically extended the potential for both socialinteraction and human transformation of thenatural environment. That structural or condi-tioning ‘effect’ of media, for one thing, dwarfsany other effects the media may be shown tohave on either individuals or institutions. Foranother thing, the understanding of such a‘medium effect’ is one of the under-recognizedcontributions of historical and other humanisticresearch to media studies.

MEDIUM THEORY

In an overview and development of ‘mediumtheory,’ Meyrowitz (1994: 50) summed up itscentral question: ‘What are the relatively fixed features of each means of communicatingand how do these features make the mediumphysically, psychologically, and socially differ-ent from other media and from face-to-faceinteraction?’ The last question in particular,suggesting a measure of technological deter-minism,� has given rise to a range of contro-versial answers.

The most well-known proponent of a strongposition regarding the scope and depth ofmedia impact on consciousness and culture,summed up in his dictum, ‘the medium is themessage,’ remains Marshall McLuhan. One ofhis most influential books (McLuhan 1964)examined the media as extensions of the humansenses with fundamental and permanent conse-quences for the awareness of self, others, andhistory. Among his polemical conclusions wasthat writing and, not least, print had con-strained people within the linear logic of a typo-graphic culture. In contrast, electronic media,and above all television, was said to signal arelease of cultural creativity from ‘the Guten-berg galaxy’ (McLuhan 1962). This hope for anew polymorphous epoch is reminiscent of laterutopian visions regarding cyberspace.

Being a literary scholar, McLuhan may havebeen tempted to map changes in the textual and

Humanities in media and communication research16

� intersubjectivity as a condition of communication andof research – Chapter 15, p. 267

humanitas

� technological determinism – Chapter 12, p. 196

‘the mediumis themessage’

thematic structures of media rather directlyonto the social and cultural context. His mainsource of inspiration, Harold A. Innis, anotherCanadian, but a political economist and histo-rian, stated his arguments in a more traditionalscholarly fashion (Innis 1964 [1951], 1972[1950]). Applying principles from the study ofeconomic monopolies to information mono-polies, he identified some of the ways in whichpower can be exercised, and subverted, viamedia. Like other medium theorists, he pointedto the historical significance of media challeng-ing religious authorities, when ‘the medievalChurch’s monopoly over religious information,and thereby over salvation, was broken by theprinting press’ (Meyrowitz 1994: 51). Innissuggested, more grandly, that the structure ofentire cultures and empires has a ‘bias’ towardeither space or time, in the sense that thedominant media may favor either stability overtime or extended territories.� Examples in-clude stone tablets, whose inscriptions last along time but do not travel well, as opposed topapyrus and paper, which will support theadministration of distant provinces, but whichare vulnerable to destruction and to beingappropriated for purposes of social change.

Though less familiar in the media field, otherresearchers working from historical and anthro-pological angles have substantiated the rele-vance of medium theory. Like Innis, they seek toavoid any strong technological determinism,exploring the social and cultural forms in whichtechnology is diffused and adapted. In a histor-ical perspective, Havelock (1963) suggestedhow writing and literacy had paved the way fora novel category of social system, interpretingPlato’s attack on the poets as announcing thepassing of an oral culture. Poets could no longerbe trusted in social matters, such as politics, thewriting of history, and science, even if theirpoetry could still be appreciated as personalopinion or myth. Regarding the next categori-cal transition, Eisenstein (1979) showed how,initially, it was the scribal culture of elites cen-tered in monasteries, not oral or popular cul-ture, that was transformed by the printing press,leading into Renaissance and Reformation.

From an anthropological perspective, Goodyand Watt (1963) questioned relativist positionsregarding communication systems and the cul-tures they will support. One of their points wasto reassert the understanding of literacy as astrategic resource and a necessary condition of, for example, political democracy (see alsoGoody 1987, 2000). The general point wasdocumented empirically in a major study byScribner and Cole (1981), who additionallyshowed that, within one culture, several differ-ent literacies and social uses may be associatedwith different languages (Vai, Arabic, and Eng-lish). Also today, oral story-telling is a socialmovement in different cultural contexts as wellas an active field of research (MacDonald1998).�

It seems an open question whether suchdifferent forms of communication and con-sciousness, as associated with different mediatechnologies, should lead to categorizations orperiodizations involving oral, scribal, print,electronic, and digital ‘cultures.’ Walter Ong(1982), for one, has described the distinctiveexperiential qualities of oral as opposed to lit-erate culture, but has been tempted by his Jesuit frame of reference to privilege the quasi-religious authenticity of the Word. What thework of Ong and others does suggest, however,is the importance of studying the specificity ofdifferent media, as the technologies are given aparticular social shape in historical and culturalcontext. As if to prove their point that ideas areshaped by contemporary media, several of thekey studies cited here were published within the span of a few years (Goody and Watt 1963; Havelock 1963; McLuhan 1964). Thishappened at a time when television was con-tributing to a new agenda for research as well as public opinion about ‘culture’ and ‘themedia.’� In the academy as well as in the restof society, television challenged taste culturesand social lines of division, as later studied byMeyrowitz (1985).

Medium theory has delineated a fertile areaboth for empirical research and for theoreticalintegration, occupying a middle ground

Medium theory 17

1111234567891012341115678920123456789301234567894012345678950111 � the ‘bias’ of communication – Chapter 12, p. 195

cultures: oral,scribal, print,electronic,digital

� history of the media and of communication – Chapter12, p. 194

between the textual focus of the humanities andthe institutional focus of the social sciences.Certainly, it lends itself to deterministic andhyperbolic formulations of the kind that madeMcLuhan a personality also in the media. Still,if applied carefully, medium theory can supporta consolidation of media studies in a middlerange (Merton 1968: 39),� steering empiricalresearch between the Scylla of myopic method-

ologies and the Charybdis of grand theories. Inthe humanities, the main currents of ideas havethemselves been shaped, in part, by their ori-entation toward a specific medium and its usesin social practice. The primary example, bothhistorically and in terms of its continued influ-ence on the study of communication, is the art,practice, and science of rhetoric, which revolvesaround spoken language.�


– c. 25,000– c. 3100– c. 1800– c. 1600– c. 1450– c. 1200– c. 1000– c. 730

1041124114561609

181418391844184618671876188818951895191119201936

1945194719481956195719621963196919711976197619781979198019811981198219841991

cave paintings by prehistoric humanshieroglyphics and cuneiform writingLinear A writingfirst known alphabet (Palestine)Linear B writingChinese ideographic writingPhoenician alphabetphonetic alphabet (Greece)

printing from movable type (China)metal type for printing (Korea)Gutenberg prints from movable metal type and hand press (Germany)regularly published newspaper (Germany)

flatbed cylinder pressphotographytelegraphdouble-cylinder rotary presstypewritertelephonephonograph for public salefilms shown to publicradio transmissiontelevision transmissionscheduled radio broadcastingscheduled television broadcasting

programmable electronic computertransistorlong-playing gramophone recordvideotapesatellite (Sputnik)television transmission via satellitecompact cassette audiotapeARPANETmicroprocessorVHS video cassette recorderteletexttelefax (with international standard)WalkmanCable News Network (CNN)Music Television (MTV)IBM personal computeraudio compact diskApple Macintosh computerWorld Wide Web

Figure 2.1 A brief chronology of human communicationSource: Adapted from Rogers 1986: 25–26; see also Project 1996; Robinson 1995;Winston 1998

� middle range theories – Chapter 3, p. 55� speech as a medium of the first degree – Chapter 1, p. 3

FOUR TRADITIONS IN THE HISTORY OFIDEAS

Rhetoric

The rhetorical tradition is by far the oldest setof ideas informing humanistic research, and itremained centrally influential in both scholar-ship and education under different cultural andinstitutional circumstances from Antiquity intothe nineteenth century (Kennedy 1980). Itslegacy for contemporary research on commu-nication and culture can be summarized brieflywith reference to three sets of concepts. First,the rhetorical tradition refers to five stages inpreparing a speech:

• inventio (the collection and conception ofthe subject matter)

• dispositio (structuring the speech)• elocutio (its linguistic articulation)• memoria (memorizing the resulting configu-

ration of form and content)• actio (performing the speech).

Of these five stages, inventio in particular recog-nizes an intimate relationship between knowingsomething and knowing how to communicateabout it. Certain ways of speaking are appro-priate in a political arena, others in a court-room, others again on a festive occasion; eachcontext has its own purpose and subject matter,which are given material shape in the speech.Dispositio and especially elocutio, next, supplythe concrete means of shaping the speech.Rhetorical figures of speech and various sym-bolic forms have entered into both the study andpractice of much literature and other arts.

Second, in addressing the audience throughactio, a speaker draws on three means of persuasion:

• ethos• logos• pathos.

These focus, respectively, on the character ofthe speaker, the quality of his/her arguments,and the emotions which the speech is designedto evoke in the listeners. Importantly, the threemeans all appear in any act of communication,but in different measures and combinations,

depending on the purpose and, hence, the genre of communication. Whereas the threeaspects of addressing an audience have the mostobvious relevance for such explicitly ‘persua-sive’ genres as advertising and political com-munication, they lend themselves to the studyof most types of mediated communication.

Third, the concept of topos, which classicalrhetoric considered as one part of inventio, isof special interest, because it suggests a figureof thought that suffuses modern humanities aswell. Topos means ‘place,’ and implies thatcommonplaces are, literally, common places ina known or imagined terrain which speakersshare with their audience. This understandingof reality as a text and, conversely, of the textas a spatio-temporal universe that can besearched for traces and clues, has been a per-sistent metaphor in the humanities up until, andincluding, theories of postmodernity and cyber-space. A ‘topical’ form of argument can rely on one or a few concrete examples, rather thana great deal of inductive evidence, as long asthe examples fit into the commonplaces whichmake up the working assumptions of people intheir everyday affairs. Accordingly, Aristotlenoted, rhetoric is the source of a kind of know-ledge which is probable and reasonable. In thisregard, rhetoric is complementary to logic,which addresses other aspects of reality aboutwhich certain or necessary knowledge can beachieved (Clarke 1990: 13).

The close link between communication andknowledge in classical rhetoric was graduallyrelaxed, as witnessed by the development of somany practical manuals of speaking well inpublic. This shift helps to account for thecommon pejorative reference to ‘only rhetoric,’to communication as a superficial form and asmisrepresentation and manipulation. Further-more, rhetorical concepts were redevelopedfrom their oral sources and applied to literateforms of communication, for example, literaryfiction and, later, to ‘texts’ in general. Never-theless, the rhetorical tradition has remained animportant source of ideas regarding the natureof both interpersonal and mediated communi-cation. During the twentieth century, rhetoricenjoyed a renaissance, sometimes under theheading of a ‘new rhetoric’ (Perelman 1979),which, among other things, gave more specific

Four traditions in the history of ideas 19

1111234567891012341115678920123456789301234567894012345678950111

topos

the newrhetoric

attention to the interaction of communicatorswith their concrete audiences. A second inspi-ration for research on the social uses of com-munication came from analytic philosophy andits examination of the structure of informalargument (Toulmin 1958).

Although European scholarship may havebeen especially instrumental in feeding therhetorical tradition into contemporary commu-nication theory, also in this area there has beena vigorous American substream of humanisticresearch (Kennedy 1980). Among the mostimportant influences has been the work ofKenneth Burke, who developed a view of lan-guage as action (Burke 1950), in addition toaccounting for literature as a social, as well asan aesthetic, phenomenon (e.g., Burke 1957).His perspective was subsequently applied to the mass media, for example, in the case ofpolitical communication (e.g., Duncan 1968;Edelman 1971). In the media field, Carey(1989) has been a central figure advocatingsuch a rhetorical as well as historical perspec-tive on the interrelations between modernmedia and earlier cultural forms. As expressedin his ritual model,� also technologically medi-ated ‘communication’ serves to create andmaintain ‘community’ – the common root ofthe two words suggests as much.

Saying something is doing something. Onelesson of rhetoric for current media studies is the understanding of communication as a form of action. Throughout the history of thehumanistic disciplines reviewed below, an im-portant challenge has been to strike a balancebetween this pragmatic-processual conceptionof communication and a focus on the structureof linguistic and other textual vehicles. Atdifferent times, humanistic research has takeneither linguistic� or pragmatic� turns, respec-tively emphasizing communication as structureor as action.

A similar balancing act has been per-formed by various literary as well as linguisticapproaches to the contents or texts of media.

This has been especially important in the mediafield, since media texts are so evidently a partof social life. In comparison, literature andother fine arts have been studied, not least for their inherent structure and value. As aresult, current media research has relied ex-tensively on discourse analysis and otherapproaches to the social uses of language (foran overview see van Dijk 1997; Wetherell et al.2001). A rhetorical-pragmatic concept of com-munication also furnishes one of the connec-tions between humanistic research on mediatexts and social-scientific research on practicesof communication.

The basic rhetorical insight, that languageserves to configure knowledge, is borne out,finally, in the concrete analytical procedures ofmany humanistic media studies. Humanisticresearchers tend to assume that ideas take ontheir final shape only when committed to lan-guage and paper. Just as children do not merelylearn to read, but equally read to learn (Heath1980: 130), so both researchers and otheradults do not merely learn to write – they writeto learn. Since the 1970s, a growing number ofempirical studies have shown how the human-ities as well as other scientific fields articulatethemselves and their findings in distinctiverhetorical forms (e.g., Brodkey 1987; Gilbertand Mulkay 1984; Gross 1990; Hacking 1983;Latour 1987; Nelson et al. 1987; Simon 1989).The prototypical social-scientific journal articletends to assume that factual findings and a sub-sequent interpretive discussion can and shouldbe placed in separate sections, and that the the-oretical categories being examined can be statedsuccinctly from the outset. In contradistinction,humanistic articles or reports will move repeat-edly among and across these several levels ofanalysis, in part to capture the many layers andcontexts of their characteristic objects of analy-sis. These two formats of publication,� in fact,may be seen to mirror the two standard com-munication models: Social scientists ‘transmit’their findings to readers, whereas humanisticscholars seek to join their readers in a form ofrhetorical ‘ritual.’


� ritual model of communication, see Chapter 1, p. 8

� the linguistic turn, p. 26

� the pragmatic turn, p. 38� publication formats of different research traditions –Chapter 16, p. 274

Hermeneutics

While the point of departure for rhetoric wasspeech, particularly concerning matters of factand how to argue about them, hermeneuticsdeveloped out of the practice of reading andunderstanding written texts,� not least narra-tives. Its general purpose has been to clarify thenature and conditions of interpretation, withreference both to the text and to the activity ofthe reader.

The texts at issue originally belonged to reli-gion and law. Over time, the principles and pro-cedures of hermeneutics came to be applied tothe arts as well as to other kinds of texts, andindeed to human experience as such. In a his-torical overview, Ricoeur (1981: ch. 1), himselfa central contributor to modern hermeneuticphilosophy, identified a transition in the earlynineteenth century from a ‘regional’ to a‘general’ hermeneutics which now covered bothsecular and religious texts. The transition wasadvanced particularly in the work of theGerman theologian and philosopher, FriedrichSchleiermacher.

Hermeneutics thus participated in a longmovement within the history of ideas, from areligious cosmology toward a historical, secularunderstanding of human existence.� Themedieval analogy between The Book of Natureand other books as means of insight into ‘thegreat chain of being’ (Lovejoy 1936), in whicheverything has its divinely sanctioned place,was finally breaking down. Texts became a newfocus of attention in themselves, as sources of scientific evidence and of aesthetic contem-plation, rather than being primarily interfaceswith the hereafter. At the inception of themodern humanities, then, the Text was takingcenter stage, together with the individualhuman being. Hermeneutics appeared to offermeans of resolving interpretive conflicts, alsobeyond matters of faith and aesthetics.

The implications of hermeneutics for doingcommunication research are suggested by thekey concept of a ‘hermeneutic circle’ (Figure

2.2). The most basic, ancient insight of hermen-eutics is that the meaning of a textual part canonly be understood in relation to the whole of the text. While perhaps commonsensical, this insight contradicts the equally commonassumption that the meaning of a message isthe sum of its parts, so that communicationmight be studied by breaking it down into itsconstituent elements. Hermeneutics suggeststhat the very process of both reading andanalyzing a text is incremental and creative –readers gradually work out their categories ofunderstanding in order to arrive at a coherentinterpretation.

This dialectic at the level of the individualtext, however, is only the first step in workingout its meanings and implications. Next, thetextual whole must itself be interpreted as partof larger totalities. For example, a novel mayexpress the mindset of its author, who might besaid to articulate the worldview of his epoch orculture. Such an understanding of the text–context nexus as one aspect of the part–wholedialectic dates from the period when a generalhermeneutics, applying to all kinds of texts,was conceived.

A further development of the hermeneutic


1111234567891012341115678920123456789301234567894012345678950111

� writing as transitional form between media of the firstand second degree – Chapter 1, p. 4

�secularization – Chapter 11, p. 174

the greatchain of being

the hermen-eutic circle

Pattern of interpretation

Text

PART

UNDERSTANDING

WHOLE

Dialogue

Sub-interpretation

PRE-UNDERSTANDING

Figure 2.2 The hermeneutic circleSource: Alvesson and Sköldberg 2000: 66.With permission of SagePublications Ltd.

circle which, more specifically, took intoaccount the role of the reader, occurred duringthe twentieth century. The point appears fromanother set of twin concepts – pre-understand-ing and understanding – which was developedespecially in the work of Gadamer (1975[1960]). On the one hand, any understandingrequires a pre-understanding – a ‘prejudice’ inthe neutral sense. On the other hand, under-standing serves to reproduce or contest thereader’s pre-understanding. Throughout theactivity of reading, people perform minimal orsub-interpretations, which may realign theirframes of interpretation. In doing so, they enterinto a ‘dialogue’ with the text and, by exten-sion, with other minds, past as well as present.Most grandly, they partake of culture – the cul-tivation of tradition.

The hermeneutic circle may be understoodas a model of communication, as it evolves notjust in the here and now, but down throughhistory and across cultures. In hermeneutic (andphenomenological) terminology, communica-tion involves a ‘fusion of horizons’ – a meetingand merging of the expectations that commu-nicators bring with them into the exchange.Especially recent contributors to hermeneuticshave pointed to the dangers of fusion and of animmediate empathy with tradition. Referring tothe works of Marx, Nietzsche, and Freud,Ricoeur (1981: 46) identified and developed a ‘hermeneutics of suspicion.’ Its particularpurpose is to discover hidden principles behindwhat other people as well as social institutionssay and do, thus enabling a distinction betweensurface and ‘reality.’ Since the principles atwork may equally be hidden to those personsand institutions, the social task of hermeneuticsmay be seen as a general one of readingbetween the lines of society and feeding itsinterpretations back into appropriate fora. Inparticular, the critical tradition in social science,and in media research, has taken on this taskin studies both of texts and social practices.

The lasting influence of hermeneutics onmedia studies lies at two different levels. First,hermeneutics has been a source of specific inspi-rations, such as theoretical frameworks inreception studies of literature and media (for anoverview see Holub 1984), and methodological

approaches to, for instance, source criticism inhistoriography (for an overview see Alvessonand Sköldberg 2000: ch. 3). Second, thehermeneutic tradition serves as a constantreminder that all social and cultural research isa hermeneutic, and doubly hermeneutic�

(Giddens 1984), activity, interpreting the inter-pretations of others.

Phenomenology

Interrelated with hermeneutics in the history of ideas and as a methodological orientation in some current social research (e.g., Kvale1996), phenomenology emerged as a distinctive school of philosophy around 1900. The date issignificant, because phenomenology may beunderstood as a defensive reaction against thereductionism, in the form of either positivism�

or ‘psychologism,’ which was then seen tothreaten a humanistic understanding of con-sciousness as a lived and interpreted whole. In response, the phenomenological traditioninsisted on the unique qualities and insights ofordinary human experience (for an overviewsee Schmitt 1972). In the social sciences, whereits influence has arguably been at least as strongas in the humanities, phenomenology came toprovide a philosophical legitimation for inter-pretive studies of social life, a minority positionin the social sciences.� In the humanities, phe-nomenology entered into the mainstream, butproposed to redevelop key concepts of thehumanistic heritage, such as interpretation andsubjectivity.

Edmund Husserl, the originator of modernphenomenology, was also the author of itsambiguous motto, Zu den Sachen selbst (to thethings themselves) (Alvesson and Sköldberg2000: 36). These ‘things’ were not materialobjects, but those elements which constitute thecore of human experience and existence, whatHusserl referred to as the ‘lifeworld.’ In orderto gain a better understanding of one’s life-world, one should perform ‘reductions’ of


fusion ofhorizons

hermeneuticsof suspicion

� double hermeneutic – Chapter 1, p. 9

� positivism – Chapter 15, p. 260

� phenomenological social science – see Chapter 3, pp. 42, 56

lifeworld

various types in order to capture its qualitativeessence. Far from breaking up experience intominimal units corresponding to, for instance,sense data, a phenomenological reductioninvolves a ‘bracketing’ (epoché) of experienceas a whole from its circumstances. Husserl’sambition was to reinvent philosophy as ascience in the strict sense. Central to this ambi-tion was an attempt to close the subject–objectdivide, which had been haunting Western philo-sophy at least since Descartes (Rorty 1979).Husserl argued that human consciousness, orintentionality, is always intentionality of some-thing, not a mental category that is foreverseparated from external reality.

In order to explain more concretely howhuman subjects relate to objects in reality,Husserl introduced the concept of a ‘horizon’(as present also in hermeneutics). The conceptrefers to the configuration of a person’s life-world at any given moment, pointing bothforward and backward in time. A horizon com-prises a set of interpretive categories availableto the person from a long biography of social-ization and acculturation. This horizon willchange or be modified over time, as the personenters into new contexts, undertakes new pro-jects, and interacts with other people. From thisgeneral philosophical category, other human-istic research derived a discursive conception ofhorizons, defined as historically and culturallyspecific frameworks of expectations that guidethe interpretation of texts. A ‘misunderstand-ing’ of a text can result from an incompatibil-ity between the horizon implicit in the text andthe reader’s horizon of expectations; a ‘dis-agreement’ about the meaning of a text can bethe product of conflicting interpretive horizons.

Phenomenology has had less of an influenceon media studies than either rhetoric orhermeneutics. This may be, in part, because itsabstract conceptual analyses have seemed lessapplicable to the concrete textual vehicles andsocial practices of communication – phenomen-ology has no evident ‘medium.’ Nevertheless,some film studies in particular have taken theirlead from the phenomenological bracketing ofexperience, and have gone on to bracket filmtexts in order to get at their essential experien-tial qualities that resemble the multimodal life-

world. More recently, Scannell (1996a) has sug-gested, with particular reference to broadcast-ing, that the radical hermeneutics (Ricoeur1981: 45) and phenomenology of Martin Hei-degger may capture the distinctive features ofan increasingly mediatized modern existence.

Semiotics

Of the four humanistic traditions, semioticsexercised the most direct influence in the for-mation of media studies as a field, and thus callsfor some elaboration. Defined most famouslyas a science that studies ‘the life of signs withinsociety’ (Saussure 1959 [1916]: 16), semioticsbecame one of the most influential interdiscip-linary approaches to the study of culture andcommunication generally from the 1960sonward. The tradition addresses all types ofmedia,� and offers methodological procedures,theoretical models, as well as constituents of atheory of science. In its most ambitious formu-lations, semiotics proposed to examine lan-guages, images, psyches, societies, even biologyand cosmology as sign processes (Posner et al.1997–98; Sebeok 1986). More commonly, thesemiotic tradition has contributed analyticalprocedures and methodological frameworks,which have lent a new form of systematicity tohumanistic research on texts.

Semiotics had two founding fathers in thelate nineteenth and early twentieth centuries –the American philosopher and logician CharlesSanders Peirce and the Swiss linguist Ferdinandde Saussure. Their disciplinary backgrounds arekey to their different conceptions of the studyof signs. Recovering an undercurrent in thehistory of ideas, going back to Aristotle, Peircedeveloped a comprehensive philosophy of signswhich he understood as a form of logic thatwould support inquiry into the nature of bothknowledge and being (for key texts see Peirce1992, 1998). In his definition, any sign hasthree aspects:

A sign, or representamen, is somethingwhich stands to somebody for something in


1111234567891012341115678920123456789301234567894012345678950111

epoché

horizon

� semiotics on media of the first, second, as well asthird degrees – Chapter 1, p. 3

sign, object,interpretant

some respect or capacity. It addresses some-body, that is, creates in the mind of thatperson an equivalent sign, or perhaps amore developed sign. That sign which itcreates I call the interpretant of the first sign.The sign stands for something, its object.

(Peirce 1931–58: vol. 2: 228)

Although signs are here said to mediatebetween objects (material and non-material) inreality and concepts in the mind, Peirce rejectedany idealist, nominalist, or skepticist position.Peirce instead attempted to marry a classical,Aristotelian realism with the modern, Kantianinsight that humans necessarily construct theirunderstanding of reality in particular cognitivecategories. Signs, then, are not what we know,but how we come to know what we can justifysaying that we know, in science and in every-day life.

Peirce further suggested that human under-standing is a continuous process of interpreta-tion – semiosis – not a singular internalizationof reality once and for all. In this perspective,examples of semiosis range from the ongoingcoordination of everyday media use and othersocial life, to scientific discovery and aestheticcreativity. Figure 2.3 displays the process ofsemiosis, noting how any given interpretation(interpretant) itself serves as a sign in the nextstage of interpretation. Even though Peirce’soutlook was that of a logician and a naturalscientist, the model may be taken to refer to thecommunicative processes by which cultures aremaintained and societies reproduced.

Saussure (1959 [1916]), in comparison,focused almost entirely on verbal language.Even though it was he who coined the phraseanticipating a general science of signs (of whichlinguistics would be a subdivision), in practiceSaussure and his followers took language astheir model for the study of other sign types aswell. Saussure’s main achievement was todevelop a framework for modern linguisticswhich, to a degree, proved applicable to othersocial and cultural phenomena.

In contrast to the emphasis of earlier phil-ology on the diachronic perspective of how lan-guages change over time, Saussure wanted tostudy language as a system in a synchronic

perspective. Language as an abstract system(langue) could be distinguished, at least for ana-lytical purposes, from its actual uses (parole),and this system could further be analyzed intwo dimensions. Along the syntagmatic axis,letters, words, phrases, and so forth, are theunits that combine to make up meaningfulwholes; each of these units has been chosen asone of several possibilities along a paradigmaticaxis, for example, one verb in preference toanother. The resulting combinatory systemaccounts for the remarkable flexibility of lan-guage as a medium of social interaction.

In addition, the interchange between theparadigmatic and syntagmatic axes gives rise totwo specific means of expression and represen-tation – metaphor and metonymy – which havebeen important in the understanding and analy-sis of media as textual messages. Metaphor maybe defined as a paradigmatic choice which acti-vates an additional frame of reference (incor-porates it within the syntagm of the message as a whole). For example, in the 1980s, the


semiosis

Interpretant 2/Sign 3

Interpretant 1/Sign 2

Sign 1 Object 1

Object 2

Figure 2.3 The process of semiosis

langue andparole

syntagms andparadigms

metaphor andmetonymy

American politician and civil rights activist,Jesse Jackson, became the spokesperson for a‘rainbow coalition’ which sought to bringtogether various ethnic and other disempow-ered groups as a new political force. This termis significantly different from the traditionalreference to the U.S.A. as a ‘melting pot.’Whereas the melting pot eliminates variety, therainbow derives its essential quality and beautyfrom difference and contrast, so that speechesand campaigns can activate this sense for a par-ticular political purpose (just as the melting-potimage serves a political purpose). Metonymy, inturn, is the process by which a single signevokes the full syntagm to which it belongs.When ‘the White House’ was said to commenton the ‘rainbow coalition,’ the reference to thebuilding would evoke both the American presi-dency and the political practices and vestedinterests associated with it (Drotner et al. 1996:195).

An important contribution of Saussure washis account of the arbitrariness of the linguis-tic sign. The sign is said to have two sides: asignified (conceptual content) and a signifier(the acoustic image or physical token associatedwith it). The relation between the two is arbi-trary or conventional, as suggested by thewealth of terms in different languages for thesame phenomenon. Saussure has sometimesbeen taken to suggest that also speakers of thesame language are paradoxically free to choosetheir own meanings, and thus almost destinedto remain divorced from any consensual reality.The point is rather that the linguistic system asa whole and its interrelations are arbitrary inprinciple, but fixed by social convention.

Much later research has extended the prin-ciple of arbitrariness to account for other socialand cultural forms, for example, artworks,myths, and subcultures, all of which may beunderstood as ‘languages’ in the broad sense.The structural anthropology of Claude Lévi-Strauss (1958) was highly influential of suchstudies of culture as more or less arbitrarysystems of signification. A similar perspectivewas developed within structuralist theories ofsocial power (e.g., Althusser 1977 [1965];Foucault 1972) and of the unconscious as a‘language’ (e.g., Lacan 1977). A common argu-

ment in this work has been that the currentorganization of society, and of the humanpsyche, is a historical coincidence, and is opento challenge and change.

To be precise, Saussure referred to thescience of signs, not as semiotics, but as semi-ology. The two terminologies point to the factthat Peirce and Saussure, while contemporary,had no knowledge of each other’s work. Duringits consolidation from the 1960s onwards,semiotics became the agreed term, as symbol-ized by the formation in 1969 of the Inter-national Association for Semiotic Studies. Itwas also during this period that the programfor studying the life of signs in society, unful-filled in Saussure’s own work, began to beimplemented in largely systemic, Saussureanapproaches to culture and society. In recentdecades, Peircean approaches have begun toinform a social semiotics, combining semioticmethodology with other social and communi-cation theory (Hodge and Kress 1988; Jensen1995).

Roland Barthes’ widely influential model oftwo levels of signification (Figure 2.4) suggestedan interface between Saussurean and Peirceansemiotics, and a common agenda of the semi-otic tradition: the analysis of concrete signvehicles – texts and images – as vehicles ofculture, ideology, or ‘myth.’ Barthes (1973[1957]) showed how the combination of asignifier and a signified (expressive form andconceptual content) into one sign (e.g., a mag-azine cover with a picture of a black man in aFrench uniform saluting the flag) can becomethe expressive form of a further, ideologicalcontent (e.g., that French imperialism was nota discriminatory system). The two levels ofmeaning are normally referred to as denotationand connotation, building on Hjelmslev (1963 [1943]). Barthes’ critical point was thatthis two-layered semiotic mechanism serves tonaturalize particular worldviews while silencingothers, and should be deconstructed analyti-cally. (Despite the predominantly critical orien-tation of semiotics, it has also been recruitedfor instrumental purposes, for example, in‘marketing semiotics’ (Umiker-Sebeok 1987).)

It is only a relatively short step to the previous model of semiosis (Figure 2.3). The


1111234567891012341115678920123456789301234567894012345678950111

arbitrarinessof sign

relations

signified(signifié) and

signifier(signifiant)

semiology

denotationand conno-tation

Peircean model refers to a potentially infiniteprocess, but recognizes that the process isalways, in practice, arrested. Sooner or later,individual sign users must arrive at interpreta-tions, and act, in particular social contexts (seefurther Barthes 1970). These questions, con-cerning meaning as process and as product, andconcerning the relationship between interpreta-tion and action, continued to occupy thehumanities as such throughout the twentiethcentury, and became one of the key issues forthe media field.

DISCURSIVE TURNS

In one sense, the humanities have always beenpreoccupied with language – as a means of per-suasion, as testimony of the past, and as amodel for other kinds of communication. Heim(1987) refers to ‘the Logos tradition,’ arisingfrom classical Greek philosophy, as a long-livedworldview that has given priority to verbal lan-guage by assuming a ‘transcendental intimacyof thought, words, and reality’ (p. 42).

The twentieth century, however, witnessedan emphatic turn to the study of language andother signs in a number of disciplines, first andmost explicitly in philosophy, and with a sharedambition of making the procedures for analyz-ing texts, artefacts, and events more explicitand systematic. In philosophy, ‘the linguisticturn’ (Rorty 1967) denotes the effort, particu-larly among Anglo-American analytic philos-ophers since the early 1900s, to examine thestructures and functions of language as a pri-mary way of knowing the structures of reality

as well as the conditions for knowing aboutthem. Much ancient and premodern philosophyhad been ontological (asking ‘What does theworld consist of?’); from the eighteenth century,philosophy came to address the same issues inepistemological terms (‘What can we knowabout the world?’). It was these same issueswhich the twentieth century examined as a lin-guistic matter (‘What do we mean by “know”and “world”?’). The linguistic turn, in effect,signaled a somewhat technical and reductionistconception of language as a modularized inter-face with reality.

The turn to the concrete and formal vehi-cles of knowledge posed problems as well aspotentials. On the one hand, close analyses oflanguage seemed to promise a new degree ofprecision, also in other humanistic research. Onthe other hand, language might end up beingtreated as a formalist universe unto itself, andas a fetish of science. Such an approach wouldbe alien to the traditional cultural, historical,and aesthetic concerns of the humanities. Ontop of this, an overemphasis on verbal languagewould not do justice to the specific qualities ofeither nonverbal fine arts or popular audiovi-sual media. (From a different theoretical per-spective, the analytical distinction between textand image was hotly debated also within semi-otics.)� These dilemmas, which continue toaffect media research today, can be laid out,first, with reference to their primary articula-tion in philosophy and, second, with reference


Myth

Language

III SIGN

III SIGNIFIED

2. Signified1. Signifier

I SIGNIFIER3. Sign

Figure 2.4 Two levels of significationSource: Adapted from Barthes 1973 [1957]: 115.With permission, Farrar, Strauss & Giroux, and Random House.

the linguisticturn

� semiotics of images as well as texts – Chapter 8, p. 120

to the solutions proposed in different human-istic disciplines.

The icon of twentieth-century philosophywas Ludwig Wittgenstein. In Tractatus Logico-Philosophicus (1921), the early Wittgensteinheld that all knowledge must be founded on ele-mentary propositions about minimal features ofreality. Laws of nature as well as other types ofinference and generalization should ultimatelybe reducible to direct observations of rudimen-tary phenomena. The ideal of both philosophyand science, accordingly, is to establish a cor-respondence between the structures of realityand the linguistic, logical, and other discursivestructures expressing our understanding of it.In this way, Wittgenstein prepared argumentsthat were extended and applied also to thestudy of culture and society by logical positiv-ism� of the 1920s to 1930s.

The later Wittgenstein himself rejected thisformal and structural view of language. Instead,he came to define language as a complex set ofdiscursive activities, so-called language gamesthat are inseparable from the life form, or prac-tice, which they serve to constitute. His dictum,that ‘meaning is use’ (Wittgenstein 1953: 20e),summed up the beginning of a larger shift from language as a medium of representation, tolanguage as a means of action. It is no longer theform of language in itself, but its social uses which are the center of interest in muchresearch. In the social sciences, a related shift hasbeen noticeable – what is sometimes referred toas a ‘cultural turn’ (Ray and Sayer 1999) – whichhas entailed a renewed emphasis on the inter-pretive dimensions of social life, compared to itsstructural and institutional aspects.

It may be argued that these various discur-sive turns, promoting a precise focus on lan-guage and culture, have enabled the humanitiesto become ‘harder’ in methodological terms. Incertain respects, the social sciences have grown‘softer.’ Indeed, concepts of ‘discourse’ – thesocial uses of signs – have entered broadly intothe convergence of the two scientific cultures.At the same time, several varieties of a discur-sive turn may be traced, also within humanis-tic disciplines.

FOUR DISCIPLINES IN HUMANISTICRESEARCH

Art history

In many respects, the modern media have taken over the social functions of both religiousand secular fine arts – visual arts (painting,sculpture, architecture), music, and literature(including theater). The arts had taken shape asseparate domains of social and aesthetic prac-tice in the eighteenth century, as described byHabermas (1989 [1962]) under the heading of a cultural public sphere (see Figure 1.3).Defined programmatically by Kant in hisCritique of Judgment (1790), the arts could be seen to establish a world apart, in which (initially only) the privileged classes wereinvited to engage in a disinterested contempla-tion of the beautiful, the true, and the good.Artworks could and should be appreciated fortheir immanent logic. In this perspective, artand culture have no ulterior purpose.

The classical understanding of the arts waschallenged both from within and from without(for key texts see Harrison and Wood 1992).Internally, Romantic as well as avant-gardemovements questioned the autonomous statusof the artwork. Externally, the invention anddiffusion of technological media increasinglyundermined any understanding of art andculture as a realm apart, and one devoid of con-flict. Quite evidently, the media were embeddedin social institutions of power and in the dailylives of all classes. For one thing, culture wasnow manifestly a business – a source of capital.For another thing, the consumption of culturewas a means of positioning oneself in society,being a symbol and a source of distinction(Bourdieu 1984 [1979]).

Museums for the general public emergedalongside the press and other media as an insti-tution of record, but also one of diversion (e.g.,Bennett 1995). All the while, traditional formsof stage entertainment remained central carri-ers of the cultural heritage (e.g., Levine 1988).Nevertheless, the modern media became a cat-egorically new form of cultural infrastructure – a continuous and conspicuous presence in the everyday, as well as an arena of socialstruggle. Over time, this common cultural infra-

Four disciplines in humanistic research 27

1111234567891012341115678920123456789301234567894012345678950111 � logical positivism – Chapter 15, p. 260

fine arts:visual arts,music,literature

distinction

museums

structure served to blur many of the traditionaldistinctions of high and low cultural sectors.From a research perspective, this meant thatany cultural expression or artefact could bestudied for its inherent meaning – as ‘ texts.’ Itfurther meant that theories and approachesfrom the one sector might be transferred toaccount for the cultural practices of the other.Both art history and other disciplines have per-formed a very fruitful application of ‘high’theory to ‘low’ culture (MacCabe 1986).

One obvious point of contact has beenbetween painting (and other visual arts) andphotography (and other media images). (Sculp-ture and architecture have recently attainednew relevance for studying the virtual worldsof digital media (see Rush 1999).) First, ana-lytical models and methods, as derived, in part,from the psychology of perception and appliedto different historical artforms (e.g., Arnheim1974; Gombrich 1960), have provided some of the necessary tools for examining form,perspective, color, as well as iconographic con-ventions in film and television. Semiotics andstructuralism were especially influential, instudies of both art and visual media, in apply-ing linguistic models to visual phenomena.Building on Erwin Panofsky’s iconology,Barthes (1984 [1964b]) contributed one of themost cited accounts of how image and text canrelate to each other in different genres andmedia. Either the text ‘anchors’ or delimits themeanings of the image, or the reader’s attentionis allowed to wander back and forth betweenimage and text as in a ‘relay.’�

Second, media studies took up and re-emphasized what has remained a more mar-ginal interest within art history, namely therelationship between the arts and their socialcontext. Art history itself has tended to repro-duce elements both from a classic understand-ing of the artwork as autonomous, and frommodern aestheticism, studying art for art’s sake. Confronting this tradition from a mater-ialist and psychoanalytic position, Hauser(1951), for example, began to rewrite thehistory of Western art, including new artformssuch as cinema. Much media research went

on to examine in detail the ways in which theselection of a particular visual subject matter aswell as its composition may carry historicallyand socially specific messages, thus readingimages as signs of their times. In an early exem-plary study, Berger (1972) identified the manyanalogies between traditional oil paintings ofthe wealthy and powerful, and advertisingwhich depicts role models that consumers areinvited to identify with.

One implication of textual studies in thisvein has been that visual media also carry aformal ‘message,’ in the sense that they frameand reinforce one particular perspective on theworld. Another implication is that imagesaddress and attract viewers, who not onlyspend time and money, but ‘invest’ their sub-jectivity and identity in the discursive universeon offer. The general notion that the mediaserve to incorporate the human subject into adominant view of society by textual means (andthat subjects as well as social institutions canbe understood as ‘texts’) has been central tomuch humanistic research since the 1960s (foran overview see Coward and Ellis 1977).

An issue which the ongoing cross-fertili-zation between elite art and popular culture has raised for future research is when the timemay be ripe for an inclusive history of modernculture that bridges the high–low divide. Both media-oriented overviews (Pelfrey 1985;Walker 1994) and traditional art histories(Janson 1991) have, to varying degrees, recog-nized the interconnections – from the ‘ready-mades’ or everyday objects elevated as art bythe Dada movement, through pop art, to instal-lations and computer art. Presumably, the twofields of research must derive some form ofintegrative framework from related fields suchas the history of ideas (e.g., Hughes 1991; Kern1983; Peters 1999). The issue goes far beyondthe question of whether the present is charac-terized, for instance, by a specifically visualculture (Evans and Hall 1999). Rather, the verycategory of ‘art,’ along with the distinctionbetween ‘text’ and ‘context,’ has been compli-cated by about 100 years of technological andinstitutional shifts in the field of culture as awhole.


texts

� anchorage and relay – Chapter 7, p. 111

Literary criticism

Of the fine arts, literature, including theater, has often been ascribed a special role in mod-ern culture, because it enables individuals toimagine radically different realities. Havingimagined and reflected upon these, people willreturn to their everyday reality and practicalconcerns with a new awareness. This process of reflexivity is conducted within a range oftraditional themes, forms, and genres, whicheducated readers are expected to master as partof their ‘cultural literacy.’ While authors mightbe expected to serve, to a degree, as the keepersof cultural tradition, a distinctive ambition ofmost modern high culture has been to tran-scend and constantly reinvent tradition. In theRomantic epoch, a new degree of autonomywas granted to the individual author. Thegenius which found expression in literaturecould be understood, in the words of the Britishpoet William Wordsworth, as the ‘spontaneousoverflow of powerful feelings’ (Abrams et al.1962: 103). In modernism, authors came to beexpected specifically to provide striking newinsights for their readers, not least by formal,linguistic means. As a social institution, modernliterature was called upon to ‘make it new,’ topresent reality to its readers in new formats(Berman 1982; Huyssen 1986).

Literary research only gradually emerged asan area of specialized study, having beenincluded as an element within historical andphilological research in general. Moreover,literary research was, for a long time, foundedon biographical studies of major authors andon historical approaches to the artworks and their place in a genealogy of styles, forms, andthematics (for an overview see Eagleton 1983;Wimsatt and Brooks 1957). The common term‘literary criticism’ further suggests the affinityof this research to an appreciation of the‘canon’ and a continuous evaluation of new art-works in this perspective. It was the rise ofseveral formalist approaches during the twenti-eth century which promoted a more distant-iated, disinterested study of literature and a sustained comparison with popular artforms.

The first internationally influential formalist‘school’ was the New Criticism. From the

1940s, a group of scholars in the U.S.A. andthe U.K. advocated studying literature as anobjective, self-contained structure of textualparadoxes and ambivalences. Any interest inauthorial intention or affective impact waslabeled intentional and affective fallacies (seeWimsatt 1954). Although this position tendedto isolate literature from its social circum-stances, it helped to professionalize the analy-sis of literature and to gain academic legitimacyfor the discipline. By performing close readingsof ‘the texts themselves’ and bracketing readers,writers, as well as the literary institution, theNew Criticism participated in the turn to animmanent analysis of texts.

Russian formalism, as developed fromWorld War I and into the 1930s (Erlich 1955),prepared the main ingredients of structuralism,which gradually became the dominant positionin literary theory from the 1960s. It was theformalists who noted that literature and otherarts do not merely make reality new; rathertheir distinctive capacity is to defamiliarizereality, to make it ‘strange.’ An important ambi-tion of structuralism was to account for suchgeneral features of literature, and of all textsand media.

Behind structuralism lies a ‘generative’ con-ception of language and meaning. The key ideais that any message amounts to a variation ona structural matrix of constants. The term ‘gen-erative’ is sometimes associated with the trans-formational-generative grammar of Chomsky(1965), who described a ‘deep’ structure thatproduces the variable ‘surface’ structures ofconcrete sentences. However, the scope of theidea is much wider. Particular narratives alsoamount to variations on a few components, and a generative model is implicit, for instance,in the social-scientific concept of social roles,which in principle can be filled by anyone (e.g.,Merton 1968). The most famous account of themodularity of narratives was the analysis byPropp (1958 [1928]) of the basic constituentsin the Russian folktale and their recombinationin any given tale. The model has been applied,for example, by Eco (1987 [1965]) in an analy-sis of the James Bond stories.�


1111234567891012341115678920123456789301234567894012345678950111

historical–biographicalapproaches

the NewCriticism

Russianformalism

defamiliar-ization

structuralism

generativemodel oflanguage

� narratology – Chapter 8, p. 123

In Figure 2.5, the generative model is treatedas a model of meaning generally, and is relatedto three other models in a typology. Thetypology arises from the combination of twodimensions, comprising the constituents andthe combinatorial structures of meaning. Inaddition, the model refers to meaning in termsof units as well as events to allow for empiri-cal research on both the systemic and theprocessual aspect of meaning production.

It may be argued that the main model ofcurrent humanistic media studies is generative,whereas the mainstream of social-scientificmedia research is stochastic. However, bothdeterministic and indeterministic conceptionsof meaning easily creep into actual research,even while few researchers may align them-selves explicitly with such models. For example,deterministic models continue to underlieassumptions about the direct impact of mediaon audiences (e.g., Postman 1985). In addition,postmodernist studies in particular tend toimply that both the textual content of mediaand its interpretation amount to free-floating,indeterministic processes (e.g., Fish 1979).

The concept of ‘intertextuality’ is anothermain contribution of structuralism to mediaand communication research.� The many

interconnections across individual texts, series,and genres in the media suggest the particularimportance of thinking of the units of contentnot as singular works, but as instances of‘textuality.’ The idea was coined, again, byRussian linguists and literary scholars duringthe interwar years (Bakhtin 1981; Volosinov1973 [1929]), and has been extended andreworked to apply to mediated communication.The concept has opened a terrain for a morefocused analysis both of such classic issues asthe relationship between ‘text’ and ‘context,’and of current phenomena such as ‘hypertex-tuality’ in computer-mediated communication.

Linguistics

To recapitulate, modern linguistics was one ofthe main accomplishments of the semiotic trad-ition. It performed a shift from a diachronicphilology that studied language in the broadcontext of history and culture, to a synchroniclinguistics which examines language in systemicterms.

At first, much linguistics consisted in a rel-atively formal study of three main dimensions,namely grammar, semantics, and phonetics,often with reference to single, abstracted sen-tences. From the 1970s, however, there was anoticeable shift in the direction of also study-


Predefinedstructureconfiguringfinalunits/events

Deterministic

Predefined inventory of final units/events

Stochastic Indeterministic

+

Generative

–

+

–

Figure 2.5 Four models of meaning

� intertextuality – Chapter 11, p. 186

fromdiachronicphilology tosynchronic,systemiclinguistics

ing the social uses of language in context. In linguistic terminology, the shift involved arenewed emphasis on a fourth dimension oflanguage study: pragmatics (for an overview seeCoulthard 1985). In the wider perspective ofsocial and cultural research, linguistics thusprovided the basic ingredients of discourseanalysis,� as widely applied in interdisciplinaryresearch. It was in this shape that linguistic con-cepts and procedures exercised most of theirinfluence on the media field.

During the early stages in the developmentof humanistic media research, literary theoryhad been relatively more influential. This islikely explained, in part, by the initial under-standing of media as sources of mass commu-nication. The media could be seen, liketraditional cultural forms, as one-way carriersof narratives, rather than as means of inter-action comparable to conversation. (In thebroader field of ‘communication,’ includinginterpersonal and organizational communica-tion, linguistics has been a steady source ofideas (e.g., Knapp and Miller 1994).) Literarymodels might also seem more relevant fordescribing the relationship between media textsand ideological or psychological issues, com-pared to the more narrowly focused linguisticmodels.

In the past few decades, linguistics hasgained somewhat of an equal standing with lit-erary criticism in textual media studies. Thismay again be explained, in part, by develop-ments in the media themselves. Ordinary con-versation and informal modes of address havebecome pervasive in media, and have directedthe attention of research toward spoken lan-guage as a constitutive feature of most medi-ated communication. Not least talkshows (e.g.,Livingstone and Lunt 1994), but also othergenres such as sportscasting (Kuiper 1996),have been studied. More generally, a rediscov-ery of the importance of interpersonal com-munication both in and around media hastaken place (e.g., Gumpert and Cathcart 1986;Scannell 1991), and reception studies havenoted the embedding of media in other every-day communication. Most recently, computer-

mediated communication has been studied forits hybrid genres, mixing interpersonal withmass communication.

The increased use of linguistic approachesin media studies has been facilitated by theredefinition of linguistics, incorporating morelevels and contexts of language use among itsobjects of analysis. In addition to being a meansof analyzing media texts, discourse analysis isalso an instrument and a means of documenta-tion in research. Media and communicationresearchers rely extensively on language forconceptualization, operationalization, data col-lection, data analysis, and reporting conclu-sions. Qualitative methodologies especiallyincreasingly employ some form of languageanalysis for both discovery and documenta-tion.�

Film studies

As in the case of art history and literary criti-cism, the social status as well as the nomen-clature of ‘film’ or ‘cinema’ are contested. Tobegin, it may be perplexing that film studies aretreated here as a source, rather than a variant,of media research. The fact that film was anearly, and a specific type of medium – the firstnon-alphabetic form of communication with amass audience – is only part of the explanationfor the separate development of film studies.More important is the fact that academicresearch on film from the outset defined itprimarily as an artform. It might have beendefined from the outset as a medium of com-munication which, like books and other media,comprises high and low forms of expressionand production.

Growing out of literary studies in severalnational contexts, film studies have remainedcomparatively segregated from other mediastudies. This is witnessed in journals and con-ferences, and in separate film and media depart-ments in universities. One consequence hasbeen that the humanities account for the main-stream of film studies, compared to the social-scientific mainstream of international media


1111234567891012341115678920123456789301234567894012345678950111 � pragmatics and discourse analysis – Chapter 7, p. 104

� linguistic analysis of research discourses – Chapter14, p. 248

film asartform

research. Film scholarship remains character-ized by its aesthetic research questions, its‘textual’ analyses, and its grand theory (for anoverview see Andrew 1976) (for key texts seeMast et al. 1999).

Before the 1960s, film theory is normallydivided into two traditions. The constructivistor formalist tradition emphasized the extent towhich films literally produce and present onepossible world for the spectators. The selectionand recombination of elements, first, within thesingle shot and second, through the montagejuxtaposing these elements, is a construction ofreality (Berger and Luckmann 1966). The actof construction, moreover, is said to involve thespectators, who are invited to rediscover notonly that world, but themselves in the processof viewing.

Despite different conceptions of film as ‘lan-guage’ or ‘art,’ and different interpretations ofits psychological or political implications, keyfigures such as Sergei Eisenstein, Béla Balázs,and Rudolf Arnheim all centered on the pro-ductive moment of montage. For Eisenstein, afilmmaker as well as theorist, the constructivistposition further entailed a special responsibil-ity on the part of the filmmaker as well as a‘politics of film’:

the filmmaker must lead [the spectator to aconfrontation with the theme] with his eyesopen, exposing to the spectator his means,his mechanism, not merely because this styleis preferable to the illusionary realism whichis the hallmark of Hollywood but becausethe film derives its energy from the con-scious mental leaps of the spectator.

(Andrew 1976: 63)

Such an understanding of communication as aform of ‘cognition through disruption’ may betraced across different media and philosophiesof art. Different articulations are found in themodernist aesthetics of ‘making it new,’ in the‘defamiliarization’ of Russian formalism and in Bertolt Brecht’s dramaturgical strategy ofVerfremdung, as well as in recent poststruc-turalist notions of poetic language.�

Realism was the second main tradition ofclassic film theory and, indeed, of much film-making. Siegfried Kracauer, who together withAndré Bazin (Bazin 1967–71) defined filmicrealism, identified two tendencies in all filmproduction until 1960, namely realism and for-malism (Kracauer 1960). Whereas formalismseemed to characterize early and silent films,realism, with its longshots and longer-lastingshots, as well as its moving camera and un-edited scenes, appeared to be dominant in manylater works. As suggested by the subtitle ofKracauer’s 1960 volume, ‘the redemption ofphysical reality,’ realism in the shape of a pho-tographic offprint of reality may be the natureof film, manifesting itself after the medium hasmatured.

Up to this point, most film studies are bestdescribed as ‘criticism,’ developing theorythrough aesthetic and conceptual analyses ofsingular works, frequently in an essayistic form.From the 1960s onwards, film semiotics devel-oped a much more elaborate and systematic setof analytical procedures and conceptual frame-works. Reacting against widespread and loosetalk about the ‘language’ of film, the centralfigure of film semiotics, Christian Metz (1974[1968]), set out to dismantle this mystifyingmetaphor. Metz concluded that although film,like television and other visual communica-tion, shares certain categories with other signsystems, film is not a language system (langue)in Saussure’s sense, consisting of minimal unitswith distinctive features that are recombined inspeech (parole). Instead, Metz went on to applysemiotic concepts to selected aspects of the filmmedium, particularly the different types of syn-tagms (complex sequences of meaning) thatstructure a given film.�

In some of his later work, Metz (1982)linked up with the psychoanalytic current infilm theory, which became central to the disci-pline from the 1970s, as represented by theinfluential journal Screen. Taking its lead fromJacques Lacan’s reinterpretation of Freud (e.g.,Lacan 1977), a basic assumption of thisresearch was that film is in a special position


the formalisttradition

montage

� poststructuralism – p. 33

the realisttradition

film semiotics

� semiotic analysis of film syntagms – Chapter 8, p. 121

to reactivate infantile forms of experience andidentification in the spectator because of itsapparent immediacy and sensuousness. Anearly text by Mulvey (1992 [1975]) elaboratedon the gendered nature of the spectator’s gaze,particularly the male scopophilia (pleasure oflooking), which seemed to be anticipated by thenarrative structure and mode of address in theclassical Hollywood film.

Within the psychosemiotic current as well as in the great majority of other film studies,the main approach has been structuralist andother close analyses of film ‘texts.’ On thisbasis, researchers have made inferences, some-times quite far-reaching, about the receptionand impact of cinematic works, and about theirpolitical or psychoanalytic implications. Filmhistory, by and large, still consists of the sumof scholarly interpretations, subject to ongoingrevision, of the main genres and their master-pieces.

In comparison, research on film productionand reception, or on institutional frameworks,has remained a minority interest. Nevertheless,the relatively few studies of this kind have indi-cated the importance of placing film in its socialand cultural context. Among the contributionshave been Gans’ (1957) study of how film-makers’ views of their audience affect theirfilms; and Bordwell et al.’s (1985) comprehen-sive analysis of the organizational as well astextual structures in Hollywood cinema. Rec-ently, cognitivism has questioned the theoreti-cal and methodological assumptions of muchprevious film research, being one of the currentchallengers to the humanistic mainstream.�

INTERDISCIPLINARY CHALLENGERS

Postmodernism

In order to assess the implications of one of themost controversial ideas in research as well aspublic debate in the past two decades, it is nec-essary to distinguish three different phenomenathat are sometimes conflated in the term ‘post-modernism’ (for key texts see Foster 1985).Originally, the term was introduced into liter-

ary and cultural research to refer to an anti-modernist style in various arts. While the stylemay be found in the literary experiments of, for instance, Richard Brautigan and ThomasPynchon, it has been associated particularlywith architecture. In an influential volume,Robert Venturi and his co-authors suggestedthat architects should take their inspirationfrom popular culture, as epitomized by LasVegas, in order to construct more imaginativebuildings and environments around people’slives (Venturi et al. 1972). Precisely the ques-tioning of traditional distinctions between highand low culture has been a position taken upby much postmodernist thinking.

Second, postmodernist styles have beentaken by some theorists as one symptom of an entire epoch of ‘postmodernity.’ In more orless radical versions, researchers have claimedthat modernity – specifically the Enlightenmentproject of achieving general social progressthrough rationalist science and politics – hasended, or is ending. Such projects have beenlabeled as ‘grand’ narratives, which could andshould, according to Lyotard (1984 [1979]), be replaced by ‘little’ narratives. These latternarratives amount to open-ended and democ-ratically accessible language games which,ideally, enable an unending dialogue betweenequal participants in communication. Beyondthe level of narrative and discourse, Jameson(1991) has gone on to argue that postmodernculture has a concrete liberatory potential, andthat postmodernism as such may be interpretedas ‘the cultural logic of late capitalism.’ Themost radical view of postmodernity has beenexpressed by Baudrillard (1988), who suggeststhat the categorical distinction between dis-course and reality has broken down, and whocelebrates the resulting ‘hyperreality.’

Third, if postmodernism is a general currentof ideas, ‘poststructuralism’ may be understoodas its theoretical branch. (Another word,‘deconstructionism,’ is associated with U.S.variants of the position, and is perhaps becom-ing the common term, but it derives from thesame seminal texts, such as Derrida (1976[1967]).) The analytical strategy is to exposeinternal contradictions in texts and to under-mine their apparent intentions. The theoretical

Interdisciplinary challengers 33

1111234567891012341115678920123456789301234567894012345678950111

psycho-semiotics:

the gaze

� cognitivism, see p. 36

postmodern-ism as an

aesthetic style

postmodernityas an epoch

poststructur-alism as atheoreticalposition

deconstruc-tionism

premise is that no textual meaning is stable, noris any genuine human insight into oneself orothers a possibility.

Compared to some earlier philosophy andtheory, the poststructuralist agenda is anemphatic skepticism and relativism. The aim isnot merely to show that knowledge is uncer-tain, or to argue that it should be obtained bydifferent means – knowledge as traditionallyunderstood is said to be literally impossible. Yetmuch poststructuralism tends to conceive ofitself as a critical and progressive position,implying alternative ideals. One argument isthat any fixation of meaning and knowledgemay be seen as a way of exercising definitionalpower, even of committing violence against theworldviews of others. In many ways, then,poststructuralism reiterates the ideology cri-tique of the 1960s–1970s, which sought tounmask texts as the carriers of bourgeois ideol-ogy (e.g., Barthes 1973 [1957]; Negt and Kluge1993 [1972]; Williamson 1978).

The social and psychological consequencesof texts, and the consequent need to ‘decon-struct’ them, have been emphasized withinpoststructuralism. Studies of discursive struc-tures as means of oppression and suppressionhave taken their inspiration from the works of,among others, Foucault (1972) and Lacan(1977). Despite this orientation toward actionand change, however, the practice of poststruc-turalist research has remained as centrallyfocused on the interpretation of texts as that ofstructuralism.

In media and communication research, post-structuralist studies have especially examinedfilm, television, and other visual texts (for anoverview see Brooker and Brooker 1997;Morse 1998). An underlying assumption is thatimages are relatively ‘open,’ and thus in prin-ciple allow viewers to take up shifting subjectpositions, but are stifled in current mediaforms. The psychosemiotic current above hasexplored how different visual genres may besaid to fixate their audience in particular spec-tator positions. In addition, poststructuralismhas lent itself to the study of emerging genressuch as the music video (e.g., Kinder 1984) andof other media formats which are under com-mercial pressures to innovate; for example,

advertising or contemporary television (e.g.,Caldwell 1995). Most recently, computer mediaand virtual realities have presented themselvesas natural candidates for examining a pheno-menon such as decentered identities (e.g., Stone1991).

It should be added that references to ‘post-modernism’ in media studies sometimes do notexpress a particular theoretical or aestheticcommitment. Instead, they may articulate ageneral and unexamined sense of living inuncertain, ‘postmodern times.’ Particularlywith reference to the assumption that thepresent period is an epoch of postmodernity,Harvey (1989), among others, has suggestedthat ‘postmodern’ culture is more appropriatelythought of as a set of styles and practices whichtypify modernity at a particular historical junc-ture. These styles and practices – for example,an aestheticization of politics and a growingmediatization of everyday life – are perhapsbest seen as the outcome of an intensified butcontinuous process of modernization. At leastin this regard, no categorical shift may haveoccurred between the ‘control society’ (Beniger1986) of the interwar period and the ‘networksociety’ (Castells 1996) of the new millennium.

Feminism

The study of gender has been central to anthro-pology and other social sciences from the outsetabout a century ago (Reinharz 1992), but it hasbeen given a more explicitly critical inflectionin recent decades, when feminism came to thefore also in the humanities (for key texts seeMarks and de Courtivron 1981; Nicholson1997).

Feminist media studies have been part of the‘second wave’ of political as well as theoreticalactivity centering on women’s liberation sincethe 1960s (the ‘first’ wave being associated withphases of securing political and other formalrights). Here, a notion of ‘difference femin-ism’ (Nicholson 1997: 3) has been influential. An important premise of difference feminism is that gender is not only a characteristic ofindividuals (a separate ‘variable’), but a consti-tutive factor in all social interactions andinstitutions, including the research institution


ideologycritique

gender:variable orconstituent?

itself. Feminism has served as one reminder thatscience is shaped, in part, by its historical andideological circumstances, and by the lives sci-entists live outside the academy. With referenceto the ‘double hermeneutic’ of science (Giddens1984), one may say that interpretive framesalso feed from the historial and social contextinto the very activity of research from the outsetof a study, not just vice versa at its conclusion.

Three aspects of feminist studies should benoted for their contribution to media and com-munication research (for an overview see alsoPress 2000; van Zoonen 1992). The threeaspects correspond approximately to phases inthe history of feminism, and to positions on ascale ranging from liberal to radical concep-tions of gender and research.

The earliest version of feminist mediastudies focused on ‘images of women,’ particu-larly stereotypical representations and theirlikely effect in socializing both female and male audiences (e.g., Tuchman et al. 1978).Such studies pre-date difference feminism, inthe sense that they tended to assume equalitybetween men and women as a neutral standard,and as an ideal that might be enforced byformal criteria. This premise is reminiscent ofliberal feminism, whose activist struggles forequal political and economic rights go back,depending on the national context, to the earlytwentieth century and before.

In other humanistic disciplines such as liter-ary studies, one concern of feminists has beento make visible, or to recover, female authorsfrom the past, on the assumption that theywould provide alternative images of women(see Moi 1985: 50–69). In media studies, itappears that the industrial organization ofimage production tends to cancel other differ-ences between media professionals, includinggender. Hence, more women employees maynot result in a more balanced representation ofgender (van Zoonen 1992: 43–65). A furtherargument may be (and has been) made that thisstate of affairs is due to an inherent ‘masculine’logic of industrial production, also in themedia.

It is such a wider, cultural conception ofgender which has informed the second formu-lation of (difference) feminism. Here, gender is

conceived as a pervasive condition of all indi-vidual and social being, affecting the think-ing and actions of both women and men. The notion of ‘difference’ also, to a degree,acknowledges that the category of ‘woman’ isnot unified, but is traversed by, for example,class and ethnicity (Press 2000: 29). Referringto the phenomenological tradition, one mightsay that the purpose of a gender-specific analy-sis of, for instance, media use, becomes, not toisolate the gender element but to bracket it.�The aim is to capture its essential qualities andto understand how specific everyday uses ofmedia arise from and feed back into genderedcircumstances of social life.

By shifting the emphasis from formalpolitical and economic institutions to culturalpractices, feminism has joined cultural studiesin the premise that culture is a site of socialstruggle in its own right. Nevertheless, certainconceptions of gender are likely to exercisehegemony (Gramsci 1971), because they are sopervasive and ingrained in cultural artefactsand practices. It is such hegemony that hasespecially been targeted for critical analysis byfeminist media research. Studies range fromexaminations of the gendered gaze in visualcommunication (Mulvey 1992 [1975]) andideology critiques of popular fiction (see studiesin McRobbie 1991), to more recent attempts atlegitimating the secret or ambivalent pleasuresof popular culture aimed at women (e.g.,Hobson 1982).

The third variety of feminism assumes aradical difference between women and men, notonly regarding their cultural categories, but inepistemological terms of what men and womencan and cannot know. Perhaps certain forms ofknowledge are available only to women. Anadditional argument is sometimes made thatthis unique potential for insight originates inthe specific psychology, or physiology, of thefemale body. With the ambiguous title, ‘ThisSex Which Is Not One,’ Irigaray (1997 [1977])has suggested several points – that there is nosimple relationship between biological sex andcultural gender; that no man or woman has ahomogeneous or unified gender; and that the


1111234567891012341115678920123456789301234567894012345678950111

images ofwomen

differencefeminism � phenomenological bracketing – p. 23

dual shape of the female genitals may beretraced in female forms of culture and dis-course. Female culture might be said to privi-lege process and dialogue, inclusion and care.The main difficulty of assessing the merits ofsuch arguments, however, is their shifting,metaphorical premises. The implication quicklybecomes an untenable essentialism that con-flates biology and epistemology.

The dangers of essentialism for researchhave also been noted within feminism. Still, thenotion that there may be separate or radicallydifferent orders of experience – a kind ofgendered ‘private languages’ – has continued to hold an attraction for feminist research. One justification for exploring this form oftheorizing has been the history of oppressionwhich women have suffered, arguably in largemeasure because of a dominant ‘masculine’epistemology. Some theorists, such as GayatriSpivak, have advocated a ‘strategic essential-ism’ (Nicholson 1997: 318) as a necessary andjustified position from which research maycounter oppression.

Essentialist feminism has had relatively littleinfluence on media and communication re-search as such. It has, however, affected thefield indirectly through the theory of science. Inrecent decades, this has been an interdisci-plinary battleground regarding the inter-connections between procedures of ‘science’and social ‘interests.’ Advocates of ‘standpointepistemology,’� for instance, have argued thatit is possible, indeed necessary, to conductresearch from particular gendered and politicalstandpoints.

Cognitivism

The latest challenge to textual and other human-istic research comes from cognitivism. Over thepast decade or so, this tradition has proposed amajor reconceptualization of humanistic, social,and cultural studies. Cognitivism derives itstheoretical inspiration from ‘hard’ sciences –neuropsychology, cognitive psychology, med-ical science, and computer science. From the 1950s onwards, these disciplines joined forces in

the interdisciplinary field of cognitive science, in part around notions of artificial intelligence(for an overview see Gardner 1985). Recentdevelopments in digital technologies have added to its perceived relevance for studies ofcommunication and culture.

In media studies, cognitivism represents a‘third’ academic culture, in addition to the twocultures of social sciences and humanitieswhich, for the past few decades, have been con-verging.� Snow (1964) had identified a deepintellectual gap within the postwar academybetween ‘soft’ arts and ‘hard’ sciences. This gapreproduced itself in the media field in the earlystand-off between hard social science and soft,interpretive humanities. Cognitivism reactual-izes issues from those debates, most basicallyabout nature and nurture, as well as structureand agency, as explanatory factors in social andcultural research. If feminism has remindedmedia research, sometimes in metaphoricalterms, that humans and their ways of commu-nicating are gendered, cognitivism has insistedthat both women and men are literally com-municating animals.

Evolutionary and other biological perspec-tives on human communication, including itscontinuities with (other) animal communi-cation, has long been a niche interest in thewider field of media studies (for an overviewsee Capella 1996). Studies of visual arts havealso been influenced by cognitive science (Solso1994). It was, however, film scholars who,during the 1980s, reintroduced cognitive theoryinto research on technologically mediated com-munication (for an overview see Zillmann andVorderer 2000). A particular influential volumewas David Bordwell’s Narration in the FictionFilm (Bordwell 1985a), which sought to inte-grate an information-processing perspectivewith structuralist concepts of texts and theirinterpretation. Bordwell redeveloped aspects ofliterary theory, specifically Russian formalism,to account in detail for the interchange betweenthe discursive structures of film and the viewer’scognitive activity.

The position of this approach in relation toearlier film theory was summed up in the title


essentialism:biological and

epistemo-logical

� standpoint epistemology – Chapter 16, p. 286

cognitivescience

� two cultures of research – Chapter 16, p. 279

naturalevolution

of a stock-taking volume, co-edited by Bordwellwith Noël Carroll and entitled Post-Theory(Bordwell and Carroll 1996). The Introductionnotes that the title was not intended to signal‘the end of film theory,’ but to question Theory,defined as ‘that aggregate of doctrines derivedfrom Lacanian psychoanalysis, Structuralistsemiotics, Post-Structuralist literary theory, andvariants of Althusserian Marxism’ (p. xiii),which had hardened into an orthodoxy of‘grand theory’ in film studies.� The stated aimof the editors, and of their contributors, was topromote pluralism in the choice of both theo-retical frameworks and analytical procedures.Moreover, the articles advocated a return tomiddle-range and empirical work informed by concrete research questions, including eco-nomic and historical studies. In effect, however,the volume confronted cognitivism with thedominant psychoanalytic theories of film. The Introduction does recognize that the essaysin Post-Theory converge on cognitivism, butargues that this stance allows for several artic-ulations – by polemical contrast to a monolithicpsychoanalysis (p. xvi).

Despite the natural-scientific inspirations ofcognitivism, the center of interest has remainedsquarely on the film text and its appeals to theviewer. It is somewhat ironic that an approachwhich gains much of its legitimacy from anemphatically experimental tradition of scienceshould still be practiced by solitary scholarsscrutinizing texts. The ‘findings’ – or readings– within cognitive media studies are mostly theoutcome of a methodological business-as-usual,relying extensively on hermeneutic and narra-tological procedures of analysis. This is espe-cially controversial since cognitivism has givenparticular attention to audiences, and has pro-posed far-reaching inferences regarding thenature of media experience. In this respect,cognitive studies resemble semiotic textualstudies, as performed before the advent ofempirical reception studies.�

This analytical practice may be due to therecency of cognitive media theory. It is also

explained, in part, by the focus of cognitivescience on universal, rather than the historic-ally and culturally variable aspects of humancognition. Yet, for future media studies, an inte-gration of cognitivism with the two other aca-demic cultures of humanities and social sciencesseems a necessary next step (see, e.g., Buckland2000).

It should be mentioned that some cognitivemedia research has offered experimental as wellas other non-textual evidence to substantiate itsconclusions (e.g., Messaris 1994; Reeves andNass 1996; Zillmann and Bryant 1994). Thecomputer – originally the key metaphor for ageneral cognitive science about the human mind – has been promoting concrete researchinterest in more diverse forms of evidence. For one thing, the distributed forms of computer-mediated communication and interaction callfor different analytical strategies from thematching of singular films with abstract cogni-tive schemata or frames.� For another, as atool of analysis and documentation, the com-puter facilitates inquiries into the process ofcommunication that are at once large scale andin depth.

Following an early cognitive conception of film viewing as puzzle-solving and otherinformation processing, recent studies haveaddressed the emotional aspect of media recep-tion. Grodal (1997), for example, has outlineda typology of genres in visual communicationwith reference to their different emotional andbodily affordances. Finally, while most of therecent interest in cognitivism has been directedtoward the audiovisual media and their visual specificity, this literature remains to beintegrated with earlier media research, forexample, on audiences’ recall of information,as presented also in print media.�

It is inviting to speculate on the timing ofcognitivism in the media field. It is likely that the abstract, contemplative, and occasion-ally idiosyncratic qualities of psychoanalyticTheory (Bordwell and Carroll 1996) contrib-uted to swinging the pendulum in an opposite


1111234567891012341115678920123456789301234567894012345678950111

� film studies – p. 33

� empirical reception studies – Chapter 10

� schemata and frames – Chapter 9, p. 149

� studies of media recall – Chapter 9, p. 144

direction. At the same time, it seems plausiblethat a focused interest in the biological, evolu-tionary aspects of culture and communicationwas delayed by discomfort at their essentialist,and potentially racist, implications – and by arelated critical agenda in much early mediaresearch. If humans are assumed to have thecapacity to emancipate, resocialize, and recul-ture themselves in radical ways, in part throughthe intervention of critical studies, then theunchanging conditions of communication willmove out of focus in research.

In conclusion, cognitivism has restatedclassic issues in the history of ideas (for anoverview see Lakoff and Johnson 1999). Thebiological and technological conditions ofculture and communication, and their differenthistorical forms and social variations, areresearch questions which are shared across thehumanities and social sciences. In examiningthe human body as a medium of thinking,representing, and acting, cognitivism links upwith parts of phenomenology which have sug-gested that bodies are not something humans‘have,’ but something we ‘are’ (Merleau-Ponty1962 [1945]: 174). However, cognitivist con-ceptions of technological media still appearpreliminary. Johnson (1987) concluded thathumans carry ‘the body in the mind,’ in theshape of metaphors and other basic-level con-cepts emerging from our bodily orientation intime and space. It remains equally necessary formedia research to explore the ways in whichwe carry ‘society and history in the mind,’ aswe interact with media and other humans.

A PRAGMATIC TURN

To sum up: the humanities have performedseveral discursive turns over the past two cen-turies. They are currently taking another,pragmatic turn as part of their engagement withboth social and natural sciences. The omni-presence of mediated communication in sociallife may have contributed to a renewed under-standing that language and texts are not merely,or even primarily, means of representation and sources of evidence about the past. Texts,genres, and media are cultural resources whichare at once material and discursive.

As suggested by the trajectory of twentieth-century philosophy, there has been a noticeableinterdisciplinary shift from an understanding oflanguage (and media) as a formal structure, toan emphasis on its social uses. The emphasis onthe pragmatic or performative aspect of com-munication was given a name in speech–acttheory (Austin 1962; Searle 1969). As an epis-temological position, speech–act theory entaileda recognition of the intimate relation betweenlanguage, experience, and action. It also drovehome a realization of the limitations of formalsystems of representation in accounting for thestructure of human experience. As a researchprogram, the pragmatic turn is at present beingimplemented in empirical as well as theoreticalresearch across and between the humanities and social sciences. In this process, media andcommunication research occupies a strategicposition.

In the long perspective of the humanities,the pragmatic turn may be summarized withreference to three characteristics, all of whichinvolve the media:

1 Performative concept of communication.Retrieving aspects of classical rhetoric, andintegrating elements of modern social theory,much current humanistic research examines thespecific ways in which individuals and institu-tions co-construct society and culture. Themodern technological media are significantsources of the frames of understanding thatenter into such construction.

2 General category of ‘ texts.’ The humanitieshave gradually come to include more audio-visual media and more everyday phenomenaamong its objects of analysis. Although thewhole world is not a text, it may be studied assuch.

3 Secular notion of culture. Related to thegeneral conception of texts, culture is now com-monly understood in the humanities as anyvehicle or practice of meaning. Secularization,in this sense, goes beyond the denial of anydivine origin and purpose of the world, andinvolves a questioning of any absolute stan-dards in social meaning production.


speech–acttheory

The most clearly delineated representativeof the pragmatic turn in media and communi-cation research has been the cultural studiestradition (for an overview see During 1999;Grossberg et al. 1992; Hartley and Pearson2000; Striphas 1998). Although researchers indifferent national contexts have, predictably,cultivated different approaches,� the Britishvariety of cultural studies, as associated withthe Birmingham Centre for ContemporaryCultural Studies, has been particularly influen-tial internationally (e.g., Hall et al. 1980). Since the 1970s, it has established itself as anidentifiable ‘school’ alongside, for example,political economy (e.g., Murdock and Golding1977) and cultivation research (e.g., Gerbnerand Gross 1976). In certain respects, culturalstudies represents a hybrid of humanistic andsocial-scientific perspectives. Within the mediafield as a whole, however, it has especially con-tributed a humanistic inflection of key concepts.

In a theoretical overview, Stuart Hall(1986), himself a seminal figure in the school(Morley and Chen 1996), argued that two par-adigms came together in cultural studies (sug-gesting that ‘paradigms,’� as introduced byKuhn (1970), may not be as incommensurableas some later commentators have made them).On the one hand, ‘structuralism,’� as notedabove, tends to stress the determination of con-

sciousness and communication by economic,political, and other social frameworks. Halltraced this emphasis particularly to Althusser(1977 [1965]) and Lévi-Strauss (1958), in addi-tion to Marx.

On the other hand, ‘culturalism’ insists onculture as a relatively autonomous practice andas a site on which some of the most importantsocial struggles of the modern period have beenconducted. The inspiration for cultural studiesin this regard came, not least, from U.K. histo-rians and literary scholars (Hoggart 1957;Thompson 1968 [1963]; Williams 1962, 1975[1958]). This is one explanation for the specialBritish connection in cultural studies. For Halland others, the two paradigms were joined byGramsci (1971), whose concept of hegemony�

– a dominant range of worldviews that re-inforce the social status quo – seemed to accom-modate both cultural autonomy and relativedetermination.

The traditions and literatures that arereviewed in the chapters of this volume, in dif-ferent ways, bear witness to influences andissues associated with the pragmatic turn. Thepragmatic turn also manifests itself in theongoing convergence of humanities and socialsciences in the media field. The contribution ofthe social sciences to media and communicationresearch is the subject of Chapter 3.

A pragmatic turn 39

1111234567891012341115678920123456789301234567894012345678950111

culturalstudies

� national research traditions – Chapter 11, p. 190

�paradigms – Chapter 15, p. 255

�structuralism – p. 29 � hegemony – Chapter 1, p. 6

culturalism

LINEAGES OF THE PRESENT

Many of the social institutions and patterns of everyday life with which we are now sofamiliar, assumed their present forms in thefour decades between 1880 and 1920. Theirdevelopment was inextricably tied up with thegrowth of modern media. The arrival of wire-less telegraphy in 1895, and of automaticswitchboards in telephone exchanges in 1892,allowed larger volumes of information and conversation to be transmitted instantly, overincreasingly greater distances. The ability toreproduce photographs in newspapers andmagazines for the first time (1880) transformedthe popular press. And the launch of the gramo-phone, in 1887, and the arrival of cinema, in1895, laid the basis for novel kinds of enter-tainment and experience.�

This complex of new communicationsmedia played a central role in constructing the

contemporary social order in four main ways.Firstly, they allowed both the large businessenterprises that were coming to dominate theeconomy, and the new forms of state and gov-ernment which were emerging in the politicalarena, to manage their proliferating activitiesmore effectively. Modern business worked withever more complex chains of supply, produc-tion, and distribution. Modern nation-stateswere assuming greater responsibility for socialwelfare in areas like pensions and education.With the age of total war ushered in by WorldWar I, they faced the problem of managingmilitary operations spread over a huge geo-graphical area. Even in peacetime, the opera-tions of many large corporations and westernnation-states were global in scope. Empires,whether territorial or economic, posed formi-dable problems of command, coordination, andcontrol.�

Second, modern communications werecentral to the ways in which governments and

Media, culture and modern timesSocial science investigations

Graham Murdock

3

• a description of media as central constituents of the modern period, and of the social sciences studying it• a recovery of the different types of research, analysis, and activism which characterized early social science,

and which fed into later media studies• a review of work examining the relations between media and democracy, social space, and public culture• a reassessment of precursors of current social science, e.g., regarding media effects and reception• a summary of the transition from structural functionalism to political economy and cultural studies in

contemporary media studies.

� chronological table of media and communication –Chapter 2, p. 18 � the control society – Chapter 16, p. 275

business corporations secured public support.Mature industrial capitalism was based on themass production of standardised goods, typi-fied by Henry Ford’s Model T motor car (firstintroduced in 1909). To maximise profits in this new system, mass production had to bematched by mass consumption. Consequently,the development of modern assembly line manufacture and of mass offices using the newtypewriters (introduced in 1873) was accom-panied by the rise of department stores andmail order catalogues that were designed toengineer consumer desire and to translatewishes and wants into purchases.

At the centre of this selling system stood thenew advertising agencies and public relationsfirms dedicated to promoting their clients’brands, massaging their company images, andcombating criticisms and attacks. Modern gov-ernments too were in the business of selling.With the development of representative democ-racy and the extension of the vote, politicalparties who wished to form a government hadto compete aggressively for public support andtry to ensure that ‘public opinion’ reflected their priorities and constructions of events. Thepublic media, particularly the popular pressand later commercial broadcasting, rapidlybecame the main arenas where these competi-tions for sales and votes were played out. Aswe shall see below, many commentators sawthe private ownership of key news media, andthe ever-present possibility of self-interestedinterventions by owners and advertisers, as afundamental problem for democracies based onthe ideal of open debate in the pursuit of thepublic interest.

Third, these concerns were part of a widerdebate about the state of public culture. Moreand more of the language and imagery throughwhich people understood and interpreted theworld was manufactured by professionalsworking in media industries that were dedi-cated mainly to making a profit. This situationgenerated fierce debates over whether art andthought which challenged dominant assump-tions could survive in the factory-like conditionsof the modern newspaper office, film studio, orpopular song-writing system. Another questionwas whether manufactured meanings could ever

be an ‘authentic’ expression of popular feelingsand opinions.

Fourth, questions about the central tenden-cies of mediated public culture led easily on to concerns about their impact on ‘ordinary’people. How did popular media influencebeliefs and behaviour? How were they chang-ing everyday patterns of activity and socia-bility? How did people manage the relationsbetween mediated meanings and meaningsgrounded in everyday experience?

When these issues first emerged, theyattracted the attention of a variety of analystsand commentators – from cultural critics toinvestigative journalists and social activists. Buta number of the most important contributionscame from people who thought of themselvesin one way or another as social scientists.�There are good reasons why we should rereadtheir pioneering interventions. For one thing,the issues they grappled with remain central to contemporary debates, and many of theirdefinitions of core problems remain highlypertinent to present-day debate. For another,their attempts to develop systematic techniquesof investigation have essential lessons to teachus about the potentials and limits of differentmethods and how they might be productivelycombined. To ignore this legacy is to condemnmedia research to perpetually rediscovering old methods, having forgotten their previoushistory. We develop intellectually by climbingon other people’s shoulders so that we can seefurther and more clearly. We have no choice,but, as a field of research, we should at leastacknowledge our debt to them and call themby name.

RESEARCHERS, REPORTERS, ANALYSTSAND ACTIVISTS

The social sciences were relatively late additionsto the intellectual field. Their emergence as theuniversity-based disciplines we now know coin-cided exactly with the development of contem-porary social and media systems. Up until then,thinking about individual consciousness, socialaction and the good society had been over-

Researchers, reporters, analysts and activists 41

1111234567891012341115678920123456789301234567894012345678950111

advertisingand public

relations

� types of theory and analysis – Chapter 16, p. 274

whelmingly concentrated in departments ofphilosophy, with history and law making minorcontributions. Consequently, as John Watson,one of the founders of modern psychology, putit, establishing the social sciences meant break-ing philosophy’s ‘stranglehold’ over socialinquiry (1924: xi).

This assault on intellectual monopoly beganin earnest in the late 1880s. In 1887, ÉmileDurkheim was appointed to the first namedprofessorial chair of sociology in France, atBordeaux. In 1890, Alfred Marshall publishedhis Principles of Economics, deliberately reject-ing the established tradition of political eco-nomy with its strong links to moral philosophyand its concern with the relations betweeneconomic activity and ‘the good society.’Instead, he argued for a new discipline thatwould analyse ‘the economy’ as a boundedsphere of action. In 1892, William James(brother of the novelist Henry James) invitedthe German scholar, Hugo Münsterberg, toHarvard to take charge of the pioneering psy-chology laboratory he had established there.And in 1898, two English anthropologists,Haddon and Rivers, set off for the TorresStraits, intent on replacing the impressions ofmissionaries and adventurers with a study oftribal life based on firsthand observation.

Underpinning many of these initiatives was the idea that the social sciences were con-tinuous with the natural sciences and shouldadopt the same basic aims and techniques ofinquiry.� That conception privileged system-atic experimentation and testing, objectiveobservation purged of all personal values, andthe translation of ‘results’ into statistics wher-ever possible. The aim was to identify the stablepatterns governing social life and to formulatethe connections binding them together as rela-tions of cause and effect. The result was know-ledge that could be used to manage existinginstitutions more effectively.

This positivist conception of ‘science’ wasstrongly opposed from the outset by those whoinsisted that the study of social life remained aninterpretive activity dedicated to understandingwhat life looks and feels like to other people.

One of the founding texts for this approach wasAlfred Schütz’s 1932 book, The Phenomen-ology of the Social World, in which he arguedthat social-scientific inquiry must always beginwith ‘the already constituted meanings of activeparticipants in the social world’ (Schütz 1956:10).� This argument – that social ‘reality’ isbeing continually built and rebuilt by ‘ordinary’people as they struggle to make sense of theirsituation and devise strategies for action – hasproved massively influential. It underpins thebroad tradition of work that is now often called‘constructionism’ (a term deliberately chosenfor its strong associations with building work)as well as much recent work on ‘active audi-ences.’� By highlighting the range of skillsrequired by everyday social practices andexploring the logics underpinning unfamiliarbeliefs, early contributors to this traditionhoped to deconstruct prevailing stereotypes anddemonstrate the creativity with which ‘ordi-nary’ people negotiated the circumstances inwhich they found themselves.

The impetus to illuminate the connectionsbetween biography and history was shared bycritical researchers aiming to expose the gapsbetween official promises and everyday reali-ties, ideals, and performance. They rejectedwhat they saw as the phenomenologists’ ‘radi-cally individualistic and subjective’ approach tothe social (Outhwaite 1975: 92), insisting thatpeople’s lives and worldviews were shaped infundamental ways by structural forces whichthey may misunderstand, or not even be awareof. This argument laid the basis for what hascome to be called the critical realist tradition ofsocial analysis (see Sayer 2000).� Its support-ers saw themselves less as celebrants of thecomplexities of lived reality and more as criticsof the processes which had produced theseexperiences and meanings and of the prevailingpatterns of power that blocked alternatives.

These three positions entailed different rolesfor the professional social analyst:

Social science investigations42

� natural and social sciences – Chapter 15, p. 255

construc-tionism

� phenomenology and social science – Chapter 2, p. 22

� active audiences – Chapter 10

� critical realism – Chapter 15, p. 268

• management• interpretation• critique.�

All these roles found a niche within the emerg-ing social science disciplines, but their practicewas not monopolised by university depart-ments. Important contributions were made bywriters working outside the academy, and thelines separating the academy from the widerworlds of letters and political action were fluidrather than fixed.

Upton Sinclair, who produced one of thebest early critiques of press performance(Sinclair 1920), had already achieved earlyfame with his novel, The Jungle (1906), anexposé of conditions in the Chicago slaughter-houses and meat-packing plants, and later wenton to run (unsuccessfully) for the governorshipof California on a radical ticket. Robert Parkhad been a journalist as well as the secretary to a well-known black activist before taking aposition in the Sociology Department at theUniversity of Chicago, where he played acentral role in developing the tradition of urbanethnography, including studies of the media ineveryday life, for which the department is bestknown. The influential cultural analyst, Sieg-fried Kracauer, moved from the department ofsociology at Berlin to become a journalist, andlater a cultural critic, for the leading Germandaily newspaper, the Frankfurter Zeitung.

There was traffic, too, across the borderbetween universities and business. When JohnWatson was dismissed from his chair of psy-chology at Johns Hopkins University for moralimpropriety, he went to work for the country’sbiggest advertising agency, J. WalterThompson, where he eventually rose to becomeVice-President. Similarly, when Ernest Dichter,who had worked with Paul Lazarsfeld, one ofthe leading academic figures in early communi-cations research, in Vienna, moved to theU.S.A., he set up his own commercial researchconsultancy where he developed his ideas onthe motivations behind consumer choices andbecame one of the central figures in postwarmarket research (e.g., Dichter 1947).

The traffic across institutional borders had several important consequences. First, itmeant that research agendas were often shapedby issues originally raised in political debate, and that leading academics frequently wrote for major magazines and newspapers as well as for professional journals. Many thought ofthemselves primarily as public intellectuals�

contributing to national and internationaldebate, and only secondarily as researcherspursuing academic careers.

Second, because media and communicationswere new areas of inquiry without an estab-lished tradition of methodology, researcherswere more open to improvisation. Sometimesthe choice of methods was the result of purechance, as in the case of the pioneering Chicagosociologist, W. I. Thomas (see Janowitz 1966:xxiv). One morning, while walking down aback alley in the Polish community on the WestSide, he had to sidestep quickly to avoid somegarbage just thrown out of an upstairs window.It contained several packets of letters arrangedin chronological order. Since he read Polish, hestarted to read one of the bundles and decidedto use letters as the basic data for his monu-mental study of migrant experience, later pub-lished as The Polish Peasant in Europe andAmerica (Thomas and Znaniecki 1927).

Early analysts were also more willing towork with a range of methods. True, somefavoured rigorous experimentation, samplesurveys, and structured questionnaires, whileothers preferred to work with depth interview-ing, personal documents, and ethnographicobservation. However, the separation betweenquantitative and qualitative methods was oftenless rigid than it later became. Paul Lazarsfeld,for example, is remembered today as one of the pioneers of statistical methods, and is oftendismissed as a crude empiricist. In fact, he was a pioneer of multi-method research. Hischoices were always dictated by the issue to beaddressed, and there was no question of onemethod being suitable for all questions(Morrison 1998: 140).

A third consequence of this movementacross intellectual check-points is that the story

Researchers, reporters, analysts and activists 43

1111234567891012341115678920123456789301234567894012345678950111 � three knowledge interests – Chapter 16, p. 279

the ChicagoSchool ofSociology

� intelligentsia – Chapter 16, p. 278

of the social-scientific contribution to mediaand communication research is as much ahistory of commercial research and criticalinquiry as of academic projects. Lazarsfeld,whose early work was often funded from com-mercial sources, argued that because of the wayhe formulated problems he always ‘got resultswhich were interesting for the theoreticalpsychologists and worth the money to the busi-ness man’ (1934: 71). His critics disagreed,insisting that intellectual inquiry must remainindependent of pressure from both governmentand business, and must define its own terms ofinquiry.

In reply, Lazarsfeld could point to PersonalInfluence (Katz and Lazarsfeld 1955), whichidentified the ways people drew on the opinionsand recommendations of friends and peoplethey admired when making everyday choices.The research was funded by two commercialsponsors, the magazine publisher Mcfaddenand the Roper polling organisation, and exam-ined people’s choice of consumer goods andfashions as well as their political choices.However, Lazarsfeld and his co-author, ElihuKatz, used the opportunity to refine the ‘two-step flow’ model of communications which hademerged (serendipitously) in Lazarsfeld et al.’s(1944) earlier work on an election campaign.The model immediately attracted enormousinterest and was widely used in later academicstudies of media audiences.� Ironically, how-ever, new ideas and methods that were pion-eered within the academy were often moreenthusiastically adopted by commercial re-searchers than by academics. As we shall seebelow, focus groups are a case in point.

Retrieving the history of the social-scientificwork on media and reassessing the contempo-rary relevance of key studies is a mammothtask, and will have to wait for another occa-sion (see Murdock and Pickering forthcoming).Here, I simply want to indicate how some ofthe most interesting early contributors set abouttackling central issues which we still face, andI will suggest how their pioneering work illu-minates contemporary debates. Important con-tributions have come from across the range of

social science disciplines – from sociology,anthropology, and psychology, to economics,political economy, and political science (for anoverview and references see Alexander et al.1997; Boudon et al. 1997). It was primarilypolitical scientists and other concernedobservers who raised the central question of therole of media in modern democracies.

MEDIA AND DEMOCRACY

Liberal democratic models of mass democracypresupposed universal access to certain basicinformational and cultural resources:

• Citizens required comprehensive and dis-interested information on developments thataffected their personal and political choices,particularly where changes were being pro-moted by agencies – whether governmental orbusiness – with significant power over theirlives.

• Citizens were entitled to see their own expe-riences, opinions, and aspirations given a fairrepresentation in the major media of publicculture.

• Citizens had a right to participation in opendebates over the relative merits of competingexplanations of prevailing conditions and rivalproposals for change.�

Surveying the state of public media at theturn of the twentieth century, particularly thepopular press which was the main forum forinformation and debate, many commentatorsdetected a substantial gap between democraticideals and press performance. As America’sleading political philosopher of the time, JohnDewey, noted in his influential book of 1926,The Public and its Problems, the popular mediahad signally failed to convert a people into apublic fully involved in the political process.Their failure to provide adequate communica-tive resources for citizenship had, he lamented,left ‘the public . . . formless . . . seizing andholding its shadow rather than its substance,’ a


� two-step flow model – Chapter 10, p. 159

social sciencedisciplines

politicalscience

� information, representation, and participation asdemocratic resources – Chapter 16, p. 276

phantom, glimpsed occasionally only to disap-pear again (Dewey 1926: 141). Dewey’s criti-cism focused on the eclipse of political newsand debate by coverage of ‘crime, accident,family rows, personal clashes and conflicts’ (p. 180). As we shall see below, his alarm atthis ‘sensationalist’ turn was widely shared.

More radical critics, however, offeredanother explanation of the press’s democraticfailure, focusing on journalism’s role in pro-viding what Upton Sinclair in his savagely crit-ical book, The Brass Check (1920), called the‘day-to-day, between-elections propaganda’which allows the ‘Empire of Business’ to main-tain its effective ‘control over political democ-racy’ and to ensure that its agendas andpriorities prevail (p. 222). In Sinclair’s view, thiseclipse of the public interest by private interestshad two origins. First, newspapers were pri-vately owned and therefore open to abuses ofpower by proprietors pursuing their opinionsor those of their allies and associates. Second,they were funded primarily by advertising,leaving them open to pressure from companieswanting to promote their corporate interests.

‘What,’ asked another prominent critic, WillIrwin, ‘does the advertiser ask as a bonus inreturn for his business favour ? . . . suppressionof . . . disasters injurious to his business’ (Irwin1969 [1911]: 15). He supports his argument bypointing to the substantial publicity given to the relatively trivial sums which some trainconductors had ‘fiddled’ from their employer,compared with the lack of coverage given to acase of wrongful arrest for shoplifting broughtagainst a department store by a woman whohad almost died of shock when apprehended.He gleefully notes that in both cases the com-panies involved were major advertisers. Theargument that money talks and also buys silencehad been made even more forcefully by EdwardRoss, a sociology professor at the University ofChicago. In one of America’s leading journals of opinion, The Atlantic Monthly, a year before,he wrote a piece provocatively entitled, ‘TheSuppression of Important News’ (1910). Hismajor concrete examples concerned the presscoverage of labour relations and strikes, whichcould be considered the touchstone of a corpo-ration’s treatment of its employees.

Other commentators at the time analysedthe news management activities of government,focusing particularly on the mismanagement ofinformation flows in times of national crisis.This second major ‘filter’ of news (alongside thefilters operating within the media themselves)was starkly illuminated by the prominent polit-ical scientist, Walter Lippmann, and his associ-ate, Charles Merz, in their case study of theNew York Times’s coverage of the RussianRevolution, entitled ‘A Test of the News’ andpublished in 1920. Their detailed textual analysis revealed that, at every stage of theRevolution and the ensuing Civil War (in which the U.S.A. had intervened to support the Communists’ main opponents), some ofAmerica’s leading journalists had been seri-ously misled by ‘the official purveyors of infor-mation’ offering opinions dictated by vestedpolitical interests rather than ‘trustworthynews’ (Lippmann and Merz 1920: 41). As aresult, ‘a great people in a supreme crisis couldnot secure the minimum necessary informationon a supremely important event’ (p. 3).

Still other critics of press performancefocused on ways in which the press amplifiedincidents of social deviance and helped provokea public and, subsequently, an official reactionwhich was out of proportion to the scale of theoriginal threat and may well have added to theproblem rather than alleviating it. In 1922, forexample, Roscoe Pound and his colleaguesundertook a detailed quantitative study ofcrime reporting in Cleveland newspapers forthe month of January 1919, using column inchcounts.� They found that whereas, in the firsthalf of the month, the total amount of spacegiven over to crime was 925 inches, in thesecond half it leapt to 6642 inches. This was inspite of the fact that the number of crimesreported to the police had only increased from345 to 363. They concluded that although thecity’s much publicised ‘crime wave’ was largelyfictitious and manufactured by the press, thecoverage had very real consequences for theadministration of criminal justice. Because thepublic believed they were in the middle of acrime epidemic, they demanded an immediate

Media and democracy 45

1111234567891012341115678920123456789301234567894012345678950111 � quantitative content analysis – Chapter 13, p. 219

response from the police and the city authori-ties. These agencies, wishing to retain publicsupport, complied, caring ‘more to satisfypopular demands than to be observant of thetried processes of law’ (Pound and Frankfurter1922: 546). The result was a greatly increasedlikelihood of miscarriages of justice and sen-tences more severe than the offences warranted.

This vigorous tradition of news critique inthe context of political democracy, althoughlargely unacknowledged now, paved the wayfor four major currents of work that have beenvery influential in media research in recentyears:

1 Work within the political economy of com-munications which explores the corporatecontrol of public communications. It is typifiedby Robert McChesney’s recent volume, RichMedia, Poor Democracy (1999).

2 An account of the various filters blockingand directing news flows on behalf of thepowerful has been outlined in the influential‘propaganda model’ of news production whichwas developed by Edward Herman and NoamChomsky in their book Manufacturing Consent(1988). The model is pursued in detail in theirsuccessive case studies of reporting in theAmerican press in times of war and crisis.

3 A substantial body of work has analysedsystematic biases in news reporting of socialconflict, including labour disputes. The Glas-gow Media Group’s Bad News study (1976)has played a particularly prominent role.

4 Analyses of moral panics and media ampli-fication generally have played a major role instudies of public responses to youth subcultures(see Thompson 1998) and to environmentaland other risks (see Douglas 1997; Renn 1991).

MEDIA, NETWORKS AND EMPIRES

As commentators recognised at the time, therapid growth of the telegraph in the 1830s and1840s was a truly revolutionary developmentin human communications. By separating com-munication from transportation for the firsttime in history and translating messages of all

kinds into patterned electrical pulses – travel-ling over wires, through undersea cables and,later, through the atmosphere – it dismantledthe material barriers that are presented bywashed-out roads and storms at sea. The effectwas to create new communicative spaces re-leased from the confines of geographical place.These spaces inhabited a no-man’s landbetween the point of transmission and the pointof arrival.�

The importance of this development wasforcefully put by Charles Cooley in a long essayon ‘The Theory of Transportation,’ published inthe official journal of the American EconomicAssociation in 1894. He had been a senior civilservant in the transport division, and he went onto become a significant figure in early Americansociology. At the time of writing, however, hewas working as a political economy teacher atthe University of Michigan. He argued that any ‘study of communication from the point of view of place relations . . . cannot penetratemore than skin-deep into the social meaning ofcommunication’ because ‘in communication,place relations, as such, are of diminishingimportance, and since the introduction of thetelegraph it may almost be said that there are no place relations’ (Cooley 1894: 293). Or, as a poem of 1872, written in honour of SamuelMorse, whose code of dots and dashes hadbecome the global language of telegraphy, put it: ‘And science proclaimed, from shore toshore, / That Time and Space ruled man nomore’ (quoted in Standage 1998: 23).

At the time, this vivid perception that com-municative spaces were displacing geographicalplaces as the key organising nodes of socialactivity was not pursued particularly vigorouslyby social analysts. Most saw ‘society’ as moreor less synonymous with the nation-state.� TheCanadian political economist, Harold Innis,was a notable exception. His early work, The Fur Trade in Canada (1930), had explainedthe choices facing Canada with reference to its place within the centre–periphery systemswhich characterised relations between empires


� space and social organization – Chapter 11, p. 182

� the nation-state – Chapter 11, p. 173

and their colonies more generally. His laterwork, Empire and Communications (1950) andThe Bias of Communication (1951) focusedmore concertedly on the relations betweencommunications and empire, arguing that allcommunication systems exhibited a ‘bias’ thatfavoured either control over time or controlover space. By facilitating coordination overincreasingly greater distances, the developmentof portable and mobile media had facilitatedboth the extension of modern imperial systemsand the corporate concentration of control overkey communications facilities.� Innis’s densewriting style and the fact that much of hisempirical material was historical discouragedmany social scientists from pursuing his ideas.But the perception that communications wascentral to contemporary forms of imperialismwas gathering momentum elsewhere.

By the mid-1920s, it was clear that Ameri-can popular culture was rapidly becoming aglobal lingua franca, led by Hollywood’sascendancy in the international film market. Asone observer put it, ‘Language varies, mannersvary, money varies, even railway gauges vary.The one universal unit in the world to-day isthat slender ribbon [of film] which can carryhocus-pocus, growing pains and dreams’ (Merz1926: 165). Countries at the other end of thiscultural chain were rapidly coming to seeAmerica’s cultural ascendancy as a new form ofimperialism, based on the annexation not ofterritory but of imagination. Commentators inBritain were particularly alarmed, viewing thelanguage they shared with America as a Trojanhorse. In 1937, the radical art historian, F. D. Klingender, collaborated with the activist,Stuart Legg, to produce The Money Behind theScreen, detailing the parlous state of Britain’snational film industry and the extent of Ameri-can dominance. It was one of the first studiesof its kind and set an important precedent forfuture work on the organisation of the mediaindustries. However, because it focused mainlyon the economic and financial aspects of film-making and distribution, it had little to sayabout the social and cultural impact of im-ported movies. It was left to an American,

Charles Merz (Lippmann’s collaborator on the‘Test of the News’ study), to put these issues onthe table.

Merz saw the sale of movies in overseas markets as the heavy artillery in a trade warwhich was designed to make the American wayof life the model for popular aspirations acrossthe globe. This assault was organised around the ideology of consumerism – the promise thatconsumer goods can deliver both happiness andfreedom – as dramatised in the lives of the starsand the lush settings of popular films. It wasexactly this incorporation of film into the culture of selling that had so angered HugoMünsterberg when he set out to defendAmerican movies from their critics and todemonstrate their potential to become a new artform, in his influential book, The Photoplay,the original title of his 1916 work (Münsterberg1970).� In his view, whenever spectators wereencouraged to think of themselves as consumers,any possibility of true art was annihilated. ‘Theinterior decoration of the rooms is not exhibitedas a display for a department store . . . A good[film] is not an advertisement for the newestfashions’ (Munsterberg 1970: 81). Writing lessthan a decade later, Merz was in no doubt thatthe battering ram of seductive consumer stylewas breaking down resistance to ‘the AmericanWay’ around the world. He observed that

Automobiles manufactured here are orderedabroad after screen shadows have beenobserved to ride in them . . . rich Peruviansbuy piano-players; orders come to GrandRapids from Japanese who have admiredmission armchairs in the films . . . Yorkshiremanufacturers of boots and clothing havebeen obliged to alter their plants because thenear east now wishes to dress like RudolphValentino.

(Merz 1926: 162)

Later writers have returned to these economicand imaginative flows by way of the notion of ‘cultural imperialism,’� as outlined in its

Media, networks and empires 47

1111234567891012341115678920123456789301234567894012345678950111 � ‘medium theory’ and history – Chapter 2, p. 16

consumerism

� film studies – Chapter 2, p. 31

� cultural imperialism – Chapter 11, p. 177

most forceful form in Herbert Schiller’s book,Mass Communications and American Empire(1969).

The tangled connection between modernisa-tion� and Americanisation was also a majortheme in studies of the waves of economic andpolitical migrants who had moved from Europeto the U.S.A. in the two decades on either sideof 1900. The shifting relations between placeand space were at the heart of the immigrantexperience. They were simultaneously living inthe United States and being members of globaldiasporic cultures� which linked their newhomes to their original homelands throughmultiple ties of kinship, memory, travel, andeconomic remittance.

As one of the major destinations for inwardmigration to the U.S.A., Chicago offered a richlocale for investigating these issues, and anumber of sociologists took up the challenge.In these studies, the role of mediated commu-nications in sustaining old allegiances and iden-tities, as well as in building new ones, emergesas a major theme. The letters collected inThomas and Znaniecki’s (1927) study of thePolish migrants contain multiple references, forexample, to the importance of family photo-graphs in cementing memories and markingchange. Robert Park, with his background injournalism, undertook a major study of TheImmigrant Press and its Control (1922). Oneof the Chicago department’s graduate students,Paul Cressey, went on to make young people’smovement through the imaginative landscapesof American consumerism, as offered in themovies, one of the major themes in his unpub-lished (1936) ethnography of juvenile delin-quency in the Italian quarter of New York (seeJowett et al. 1996).

In a situation where migration and dias-pora are increasingly characteristic of people’s everyday experience around the world (themajor focus, then and now, of anthropologicalresearch) and where American brands andcultural productions have achieved global cur-rency, these pioneering attempts to explore how

place and space intersect to organise mem-ories, identities, and desires are more pertinent than ever. Similarly, as Manuel Castells hasargued in his influential analysis of trans-national networks (1996), the role of electroniccommunications in shifting the relations be-tween space and place is central to under-standing the structural organisation of socialand economic life in the era of globalisation.�Other commentators, like John Urry, go further,arguing that, faced with these challenges, soci-ology needs to focus not on ‘societies,’ but onglobal mobilities and flows – of images, infor-mation, peoples, and goods (Urry 2000: 2–3).

This new agenda places issues of communi-cations at the heart of social inquiry. The pointcan be oversold, however. People still have tolive somewhere, and many aspects of their livesare still structured in fundamental ways byinstitutions and processes operating at localand national level. As a consequence, nationaland local media remain crucial resources forpeople’s understanding of themselves and theirsituation.

TRAVELS CLOSER TO HOME

Although the impact of the shifting relationsbetween physical locations and communicativespaces was felt most acutely by migrants, it wasalso an integral feature of the way everyday lifewas being restructured everywhere, even in themost conservative-looking towns. In 1921,Robert Lynd and his wife, Helen, set off for themodest-sized city of Muncie in Indiana to assessthe effects of social and cultural change oneveryday life. Muncie was the archetypal smalltown, middle-of-the-road politically and in themid-West geographically, qualities the Lyndscaptured perfectly in their fictitious name,Middletown. The study was funded by a reli-gious foundation which was particularly inter-ested in shifts in religious behaviour, but fromthe outset, the Lynds decided to look at life inMuncie in the round.

The researchers were immediately struck byhow comprehensively and rapidly a town sixtymiles away from the nearest sizeable city and


� modernization – Chapter 11, p. 193

� diaspora – Chapter 11, p. 180

anthropology

� globalisation – Chapter 11, p. 181

half a day’s train ride from the nearest large citywas being transformed by change. They wereparticularly impressed by the way that ‘increas-ingly frequent and strong culture waves sweepover us from without, drenching us with thematerial and non-material habits of othercentres’ (Lynd and Lynd 1929: 5) They saw this process as centrally driven by the increas-ing diffusion of two technologies – cars andradios – the first offering physical release fromthe confines of the town, the second imagina-tive mobility. They pictured the radio ‘rollingback the horizons of Middletown for the bankclerk or the mechanic sitting at home andlistening to a Philharmonic concert . . . or toPresident Coolidge bidding his father goodnight on the eve of election’ (p. 269). And theywere adamant that it was impossible to study‘Middletown as a self-contained, self-startingcommunity . . . when one watches these space-binding leisure-time inventions imported fromwithout . . . reshaping the city’ (p. 271).

The fact that Muncie had a population ofonly 35,000 made it possible for the Lynds tostudy it relatively intensively, using a range ofmethods from documentary research to per-sonal observations and interviews. Largertowns and cities, however, posed formidablelogistical problems for this intensive style of‘community study’ (as it came to be called).Consequently, urban ethnographies of everydaymedia activity have tended to focus on partic-ular neighbourhoods or locales within a city.�

The Lynds’ central focus on the role of com-munication technologies in restructuring every-day life gave added impetus to longstandingconcerns about the range of cultural and infor-mational resources which the popular mediaoffered for meaning-making.

MEDIA AND PUBLIC CULTURE

Looking at the popular press – the major forumfor information and debate before the launchof radio broadcasting in the 1920s – earliercommentators had seen the discussion of polit-ical affairs and public policy being steadilyedged out by crime, scandal, and human

interest stories. They lamented this movementtowards the appeals and styles of entertainmentand theatre because it seemed a move awayfrom the rational public debate which democ-racy required, and to mark an eclipse of activecitizenship by spectatorship.

In the ensuing debate on ‘sensationalism’ (asthis new journalistic style came to be called),many contributors were content to rely on illus-trative examples selected for their rhetoricaleffect. Within the emerging social sciences,however, there was a desire to calibrate thisshift more precisely. Addressing the first meet-ing of the German Sociological Association in 1910, the country’s leading sociologist, Max Weber, urged his audience to pick up‘scissors and compass to measure the quantita-tive changes of newspaper contents during thelast generation.’ Interestingly, however, he wascareful to emphasise that these initial countsshould be seen as the first stage of a widerresearch programme that ‘will proceed to qual-itative [analyses]’ (quoted in Hardt 1979:181–182).

In the USA, however, Weber’s programmewas already well underway. In 1900, forexample, Delos Wilcox had published an exten-sive study of the differences between ‘serious’and ‘yellow’ (or ‘tabloid’) newspapers in one ofAmerica’s most prestigious scholarly publica-tions, the Annals of the Academy of Politicaland Social Science. His painstaking ‘column bycolumn’ count of the space occupied by variouscategories of content demonstrated that criticsof the popular press were right (Wilcox 1900:65). ‘Yellow’ newspapers carried more news ofcrime and vice, whereas ‘serious’ (or ‘conserv-ative’ papers as he calls them) carried more ofthe political and business news as well as lettersand exchanges, as required by classical democ-ratic models of the press as a public sphere(Habermas 1989 [1962]). This argument is alsoencountered today in discussions over the‘tabloidisation’� of contemporary news andaccusations of ‘dumbing down’ (see, e.g.,Sparks and Tulloch 2000). Quantitative contentanalysis continues to play a central role in fur-nishing empirical evidence for these debates.

Media and public culture 49

1111234567891012341115678920123456789301234567894012345678950111 � ethnography – Chapter 10, p. 164

sensationalism

� tabloidisation – Chapter 7, p. 107

By no means all commentators saw thepopular press as the enemy of democraticideals, however. In a seminal paper on thehuman interest story, the Chicago sociologist,Helen Hughes, argued strongly that by incor-porating the stories of ordinary people into thenews, the popular press was rebuilding theempathy for strangers which modern life haderoded, and which a democracy that is basedon equal respect and attention to plural voicesrequired. In her view, ‘moral speculations arenot evoked by news of court procedure; theytake form on the reading of an intimate storythat shows what the impact of law and con-vention means as a private experience’ (Hughes1937: 81). Again, analysts have recently re-turned to these relations between personal testi-mony and democratic sensibilities in detailedstudies of both tabloid news (e.g., Langer 1998)and a new wave of confessional television talk-shows (e.g., Murdock 2000).

While some early social scientists weremapping popular representations using contentanalysis, others were addressing the second partof Weber’s programme, namely the develop-ment of detailed case studies� of particulartexts. Critical analysts were particularly inter-ested in how the themes and styles of popularmedia constructed the world in ways that sup-ported prevailing relations of power. SiegfriedKracauer’s (1927) essay, ‘The Little ShopgirlsGo to the Movies,’ is a particularly goodexample of this tradition of ideology critique�

in action. Movies, he argues, ‘are babbling arude secret, without really wanting to. In theendless sequence of films, a limited number oftypical themes recur again and again [revealing]the sum of the society’s ideologies, whose spellis broken by . . . interpretation’ (Kracauer1995: 295). This characterisation of the criticas a guerrilla fighter in the war of signs, reveal-ing the power behind the pleasures, has beencentral to much subsequent work in textualanalysis. However, proponents of this model ofstrong media have too often assumed that audi-ences are vulnerable or, at the very least, not

properly equipped to see through the disguisesof power without the help of a professionalanalyst.�

Kracauer emphatically sees the ‘little shop-girls’ watching the films that he deconstructs as open to suggestion. After sitting absorbed ina film that reveals the soft heart concealedwithin the iron chest of a Berlin businessman,for example, he imagines them learning toforget their exploitative and unequal workingconditions and coming ‘to understand that theirbrilliant boss is made of gold on the inside aswell’ (p. 300). Interestingly, Roland Barthes em-ployed almost exactly the same metaphor thirtyyears later in his (1957) book, Mythologies. Heargued that it is only once ‘a typist earningtwenty pound a month recognizes herself in thebig wedding of the bourgoisie’ that ideologycan be said to be fully effective (Barthes 1973:154).

Barthes’s semiotic approach to critiquingpopular media went on to exert an enormousinfluence on later work in media and culturalstudies. However, it was also subjected to fierce criticism from commentators who argued that it did scant justice to the complexities ofwomen’s relations to the media they consumed.This argument was later pursued in a range ofdetailed empirical work which fed into a moregeneral celebration of the ‘active audience.’Once again, however, far from marking a deci-sive break with the past, as some commenta-tors have claimed, this general perspective haslong roots in social science inquiry.

In 1913, Emilie Altenloh, a graduate studentat the University of Heidelberg, published thedoctoral thesis in sociology which she had con-ducted under the supervision of Alfred Weber(Max Weber’s brother). Entitled The Sociologyof Cinema, it was based on replies to over athousand questionnaires given out to cinema-goers, supplemented by personal observationsand interviews. In contrast to Kracauer, she sees the women at the centre of her study notas dupes of prevailing ideology, but as self-reflexive actors in the new social and imagina-tive spaces opened up by the cinema and the


� case study – Chapter 14, p. 239

� ideology critique – Chapter 2, p. 34� audience as semiotic guerilla fighters – Chapter 10, p. 167

modern city. Their choices are crucially deter-mined by the dynamics and pressures of amature capitalist economy (see Hansen 1990).However, this sociological approach to audi-ences as actively negotiating mediated meaningsand experiences, but embedded in wider socialcontexts that structured their responses incomplex ways, had an uphill struggle. It wasfaced with the growing enthusiasm for the stim-ulus–response models being developed withinindividual psychology.

QUESTIONS OF INFLUENCE AND EFFECTS

One of the leading proponents of this psycho-logical model was John Watson. His ‘behav-iourist’ approach saw all human action, from‘jumping at a sound’ to ‘writing books’ asresponses to environmental stimuli (1930: 6).He further insisted that only records of behav-iour that was observed directly in the labora-tory counted as relevant data in social inquirieswishing to claim the mantle of ‘science.’

As he conceded, however, ‘the social prob-lems which psychology sometimes has to study’could ‘probably never be brought under labo-ratory control’ (Watson 1924: 28). He haddiscovered this for himself when he set out in the summer of 1919 to evaluate the impact on the sexual behaviour of young people of agovernment-sponsored film which warned thatpromiscuity could lead to venereal disease.After a monumental research effort involvingover a thousand questionnaires, nearly a hun-dred personal interviews, and a number ofobservations of film screenings, he reluctantlyconcluded that his evidence suggested that‘there is no indication that behaviour is modi-fied significantly’ (Lashley and Watson 1922:216). A powerful stimulus, expressly designedto persuade, appeared to have had no dis-cernible effect that could be detected by hismeasuring instruments. Unfortunately, this cau-tionary tale did not dissuade subsequent psy-chological researchers from pursuing the HolyGrail of direct media ‘effects,’ both in and outof the laboratory.

Had later researchers looked closely atWatson’s own description of his research, theywould have come across a brief account of an

incident that pointed to an alternative strategyof inquiry. Watson was disturbed to find thatfar from instilling the fear and disgust that theproducers had hoped for, the film frequentlyprovoked ribald and risqué comments. Afterone screening, members of the research teamdecided to talk ‘with a number of young men,loafers about the hotel lobby, and the like.’They told the researchers that when ‘boys andgirls who had seen the picture talked about itafterward . . . flippancy and innuendo prevailedin their talk’ (Lashley and Watson 1922: 204).Watson saw this ‘talk’ simply as an interest-ing supplement to his main data. For other researchers, however, listening attentively tocasual talk and loafing around hotel lobbiesand other places where people socialisedspontaneously provided the basis for an entirelydifferent approach to investigating the role ofmedia in everyday life.

LIVING WITH MEDIA: ACTIVE AUDIENCES

When Paul Cressey set out to study the placeof movies in the lives of young men living inthe predominantly Italian area of East Harlemin New York in the early 1930s, he declared hiswish to ‘avoid the “social vacuum fallacy” soprevalent in much . . . psychological research’and to ‘see the motion picture’ in ‘relationshipto all the other forces and influences which bearin upon the delinquent boy in these areas’ andin the motion picture theatre (quoted in Jowettet al. 1996: 160). Cressey had already under-taken an acclaimed ethnography of a dance-hall (later published as Taxi Hall Dance) for his Master’s degree in sociology at Chicago. He decided to approach this new study as an ethnography of an urban neighbourhood,drawing on systematic observation and open-ended interviews to construct a thick descrip-tion� of the cinema’s role in the social andimaginative lives of the boys he was asked tostudy.

Cressey produced a path-breaking manu-script which suggested fertile links between themovies’ equation of success and happiness withmoney and consumption, his subjects’ sense of

Living with media: active audiences 51

1111234567891012341115678920123456789301234567894012345678950111

psychology

behaviourism:stimulus and

response

� thick description – Chapter 14, p. 242

themselves, and the careers actually open tothem. Crime, he argued, was not a simple‘effect’ of crime films, but was produced in the cracks between consumer aspirations andblocked social opportunities. ‘For boys whohave been restricted on all sides by poverty, the appeal of [the] expensive apartments, costlyautomobiles, and “flashy” clothes’ enjoyed bythe gangsters in the films they watched is ‘initself an invitation towards that type of activ-ity and aspiration’ (ibid.: 209).

Unfortunately, the study was never pub-lished, and the unfinished draft lay unread inthe archives and was rediscovered only recently.However, if we look at other enthographies ofyoung people undertaken in the same period,such as Frederick Thrasher’s The Gang (1927),we see the same drive to anchor accounts ofdeviance in everyday experiences and to presentyoung people as actively creating their ownmeanings. Some of these meanings (as Thrasherargued) might form the basis for a distinctivesubculture with its own rituals and emblems ofidentity.

Again, these early studies of youth subcul-tures, together with the preference for ethno-graphic modes of inquiry and the insistence onplacing everyday media activity in its widercontexts, have all been enormously influentialwithin academic studies of audiences and con-sumption over the past two decades or so. Thishas been the case particularly within culturalstudies.� Prompted by the early ethnographicwork of the Birmingham Centre, researchershave returned to the tradition of inquiry pio-neered by Cressey, Thrasher, and others. Nowas then, however, there have been relatively fewfull ethnographies based on contact and obser-vation sustained over several years. For practi-cal reasons, most researchers have had to settlefor qualitative studies compressed into muchshorter time periods, using open-ended inter-views, focus groups, or personal documents toaccess grounded interpretations and everydayexperiences. Given Lazarsfeld’s reputation as anarch number-cruncher, it is ironic that three ofthe most important efforts to develop qualita-tive methods and apply them to the study of

audience activity should have come either fromresearchers working on projects he helped toset up, or from his close collaborators.

The value of conducting detailed interviewswith respondents had been amply demonstratedin a study of public reaction to Orson Welles’s1938 radio dramatisation of H. G. Wells’s TheWar of the Worlds. Welles, who already had areputation for breaking established genericrules, decided to present this story of a Martianinvasion of Earth in a form that was as closeas possible to the conventions of radio newsreporting. Some listeners believed it was anewscast and panicked. Although the leader ofthe research project, Hadley Cantril, was a psy-chologist, he saw very clearly that personaldifferences could not explain the complexvariations in responses to the programme, asdescribed by the 138 people his research teaminterviewed. Instead, he sought to account fordifferences in terms of inequalities in theamounts and types of social and cultural capi-tal� people possessed, such as how familiarthey were with contemporary artistic conven-tions, and whether they had someone theytrusted whom they could ask for advice.

Read in this way, Cantril’s explanatoryframework is closer to the account that theFrench sociologist, Pierre Bourdieu, gives of thesocial basis of cultural taste in his highly influ-ential book Distinction (published in English in1984). To characterise it as simply anotherexercise in ‘selective influence,’ as one widelyread American textbook on media researchdoes (Lowery and DeFleur 1983: 83), is toignore how thoroughly the study emphasisesthe ‘social’ in social psychology.

Cantril’s Invasion From Mars, published in1940, was the first book produced by the RadioResearch Project, directed by Paul Lazarsfeld.The following year, Herta Herzog, one ofCantril’s co-authors and a member of Lazars-feld’s project team, published a path-breakingstudy of women’s relations with daytime radiosoap operas. The study was based primarily onpersonal interviews with a hundred womenliving in the Greater New York area, initiallyopen-ended, but later using a prepared ques-


� cultural studies – p. 57 � social and cultural capital – Chapter 9, p. 154

tionnaire. This work is often presented as anearly example of ‘uses and gratifications’research.� Herzog certainly uses these words,but again, to see it simply in these terms is tomiss its more radical implications. Like EmilieAltenloh almost thirty years before, whatemerges if we read the data carefully is anaccount of women’s strategies for coming toterms with the expectations, inequalities, andburied resentments of domesticity, and of howthe imaginary worlds of the soaps intersect withthe mundane realities of everyday life. As such,it stands as an early precursor of the very influ-ential qualitative work on women as audiencesfor romantic fiction and television soaps thatfollowed the emergence of feminist scholarshipin the 1970s.�

Another of Lazarsfeld’s intellectual col-leagues, Robert Merton, also used personalinterviews to good effect in a study of newsmagazine readership in a small town (Merton1949). After preliminary work, he was in-trigued to find that influential people in thecommunity approached the magazine with verydifferent purposes. ‘Local’ influentials, who hadgrown up in the town and owed their positionto their local social connections, showed very

little interest in stories dealing with national of international affairs. In comparison, ‘cos-mopolitan’ influentials, who had come fromelsewhere and owed their reputation to nation-ally recognised qualifications, tended to gravi-tate to these items first.

Interestingly, it was exactly this last groupthat Edward Ross had had in mind when in1910 he called for the establishment of a news-paper endowed by public subscription topromote serious news and debate. He admittedthat it was unlikely to attract a general reader-ship, but he hoped that ‘it would inform theteachers, preachers, lecturers and public men,who speak to the people eye to eye’ (Ross 1910:311). This describes exactly the ‘two-step flow’model of communications that Merton andLazarsfeld developed in their work. However,to see Merton’s work on ‘locals’ and ‘cos-

mopolitans’ simply in these terms is once againto miss its full implications. Merton’s emphasison variations in cultural and social capital, and on differences in the strategies of distinc-tion and advantage pursued by different socialsegments, prefigures Bourdieu’s work. In thisregard, Merton may be seen as one of the pion-eers in establishing a distinctively sociologicalapproach to differences in media consumptionand cultural tastes.

Merton also played a major role in pio-neering a method that has been central to recent work on active audiences: focus groupresearch.� He developed the technique whichhe called ‘the focused interview’ during hisinvolvement in wartime studies of militarymotivation and morale. Just after World WarII, he published a paper outlining its uses andrationales in one of the major professional jour-nals of sociology (Merton and Kendall 1946).At the time, however, the growing ascendancyof quantitative methods tended to discourageacademic researchers from pursuing its possi-bilities. It was taken up within market research(see Morrison 1998), but it has only recentlyre-emerged as a major technique in social-scientific studies of audiences.

Sometimes, however, direct interviewingproved impossible because of shortage of time,limited resources, or difficulties of access. Inthese instances, researchers wishing to workwith qualitative materials fell back on the‘frozen speech’ embodied in personal docu-ments. We have already noted how W.I.Thomas built his study of Polish migrants toAmerica around collections of existing familyletters. Where there were no existing docu-ments, researchers set out to create them.�Herbert Blumer, another leading figure in theChicago Sociology Department, collected ‘anumber of autobiographical accounts dealingwith motion-picture experiences’ (Blumer andHauser 1933: 20). He then used these as thebasis for a study eventually published asMotion Pictures and Youth. As with Cressey’sstudy (which formed part of the same Payne

Living with media: active audiences 53

1111234567891012341115678920123456789301234567894012345678950111

� uses and gratifications research – Chapter 9, p. 142

� qualitative studies of women as audiences – Chapter10, p. 163

local andcosmopolitansocial types

� focus group research – Chapter 14, p. 241

� autobiographies and other documents – Chapter 14,p. 243

Fund research programme),� Blumer’s materi-als, with their strong emphasis on situatedexperience, led him to emphasise the impor-tance of ‘social milieu’ and ‘social background’in forming film preferences and organisingresponses (ibid.: 202).

At the same time, the scope of Blumer’sstudy was limited by the project’s focus on thepossible links between the cinema and juvenilecrime. A more open account of the relationsbetween cinema-going, everyday life, and aspi-rations was offered by J.P. Mayer in his 1948study, British Cinemas and Their Audiences.Mayer had hit on the bright idea of collectingcinema autobiographies by placing an adver-tisement in Picturegoer, one of the major mag-azines for film fans, asking people to write tohim. Ien Ang was later to use the same tech-nique to collect Dutch women’s accounts oftheir experiences of watching the Americantelevision soap, Dallas, for her very influentialstudy, Watching Dallas (Ang 1985).

Mayer freely acknowledges his debt toBlumer’s pioneering efforts, but uses his ownmaterial to explore a range of themes nottackled in Blumer’s work. These include thecentral role of films, particularly American filmsand their stars, in offering models for personalstyle and consumption. As one of the womenwho wrote to Mayer explained,

When I was 17 I saw a star (I forget her name)about whom the boy I was with said: ‘She hasthe most lovely feet and her shoes are alwaysbeautiful.’ I had nice feet and made a vowthat the same should be said of me. I don’tknow whether it ever was, but I alwaysbought the nicest shoes and stockings I couldafford and shoes are still my pet luxury.

(quoted in Mayer 1948: 25)

Mayer claimed in the Introduction to thestudy that, despite the self-selecting nature ofhis sample, his ‘anonymous contributors speakfor twenty million or more’ addressing us, thereaders, directly in their own words, ratherthan ‘through the mouth of the “superior”intellectual who by chance or choice went to a

better school, to university’ (p. 11). This cele-bration of the authenticity of everyday talk andits ability to offer the analyst direct, unmedi-ated access to ‘real’ experiences and feelings is a blindspot comparable to the wholesale re-jection of such evidence by some quantitativeresearchers. This inclination to romanticise‘ordinary’ lives runs through much later work on everyday life and media consumption within cultural studies and qualitative sociol-ogy like a goodwill message embedded in astick of seaside rock. Its impetus is generousand democratic in spirit, but by playing downthe researcher’s responsibility to reinterpretpeople’s own accounts and to tease out thehidden threads that bind biographies to histo-ries, strategies to structures, it blunts the criti-cal edge of social investigation.

REACTIONS, RUPTURES ANDREDISCOVERIES

The account offered here has deliberately con-centrated on work conducted in the first half ofthe twentieth century in order to show thatcurrent research in media and communica-tions continually draws on traditions of socialscience inquiry which are longer, richer, andmore varied than many contemporary writersimagine. However, there is no doubt that themajority of research in the field has been con-ducted since 1950.

Along with other specialisms, media andcommunication research benefited enormouslyfrom the rapid growth in social scientificresearch both inside and outside the universitiesin the postwar years.� But it was also pro-foundly shaped by the political climate createdby the onset of the Cold War between the USAand the Soviet Union. This ideological conflictdominated the intellectual landscape. In theUSA, the obsessive hunt for communist ‘subver-sives’ in cultural and intellectual life, spear-headed by Senator McCarthy’s investigationsinto ‘un-American’ activities, had a profoundlychilling effect on scholarship, and it comprehen-sively discouraged the pursuit of critical inquiry.


� Payne Fund studies – Chapter 10, p. 158� development of media research institution – Chapter16, p. 278

The dominant model of the social orderduring that period, structural functionalism,fitted this climate of caution perfectly. The prin-cipal architect of the position was the Americansociologist, Talcott Parsons, whose key book,The Social System, published in 1951 (in theearly years of the Cold War), saw every socialinstitution as having a particular role or ‘func-tion’ in maintaining social stability and cement-ing consensus. Media and communicationssystems were assigned a central role in thisgluing-together process.

This model of society as a smoothly func-tioning, self-correcting organism was itselfhighly functional politically. It presentedpostwar America as a society that had success-fully transcended the class conflicts that theSoviets (following Marx) saw as the majordriving force of historical change. It was also auseful intellectual weapon in the fierce global-wide struggle for the hearts and minds of citi-zens in the former colonial territories that hadachieved political independence after the war.It allowed the USA to present the ‘Americanway’ of doing things as a self-evidently supe-rior path to ‘modernisation,’ both materiallyand morally, one that would deliver social inte-gration as well as economic improvement. Inline with this view, degrees of ‘modernisation’came to be measured in terms of how closely acountry approximated to the USA in terms ofthe relative distribution of selected consumergoods, including communications goods.

Although the long shadow cast by the ideo-logical and military struggle between theworld’s two great superpowers silenced ordeflected many sceptical voices within theAmerican academy, there were exceptions, mostnotably C. Wright Mills. Mills had studiedsociology at the University of Wisconsin underEdward Ross (whose blistering attack on thepress we looked at earlier). He shared hismentor’s radical populist perception that thedemocratic rights of ‘ordinary’ people werecontinually blocked by entrenched centres ofpower intent on retaining their privileges andextending their influence. This led him to arguethat, far from being agents of organic unity andvoluntary consensus as Parsons supposed,mediated communications were central to the

play of power. We cannot, he argued, ‘merelyassume that some set of values, or legitima-tions, must prevail lest a social structure come apart’ (Mills 1970: 46). On the contrary,if there is

a unified symbol sphere, one monopolisedby certain master symbols, [it] is more likelyto be the result of a monopoly of channelsof communications, and of forceful taboo-ing of certain countersymbols, than theresult of any harmonious institutional basis.It is more likely to be imposed than to grow.

(Gerth and Mills 1954: 297)

Mills went on to develop this point fiveyears later, in his most influential book, ThePower Elite (1959). In it, he presented America,not as the ideal democracy of official Cold War rhetoric, but as ‘a naked and arbitrarypower,’ controlled by the interlocking interestsof industry, the military, and government, inwhich ‘the second-rate mind is in command ofthe ponderously spoken platitude’ and ‘its menof decision enforce their often crackpot ideason world reality’ (Mills 1959: 360–361). Meas-ured against the scale of this perceived threatto democratic ideals, it is not surprising thatMills found that most work by his colleaguesin the social sciences fell some way short of thechallenge.

In his manifesto for engaged inquiry, TheSociological Imagination (1959), he argued thatthe social sciences are distinguished from otherforms of commentary precisely by their ‘capac-ity to range from the most impersonal andremote transformations to the most intimatefeatures of the human self – and to see the rela-tions between the two’ (Mills 1970: 14). Thisrequired all aspects of social life to be placedsquarely in their full historical and structuralcontext. Their failure to do this was his princi-pal complaint against Merton and Lazarsfeld,his colleagues at Columbia, where he wasworking at the time.

If Parsons’ bland, empty categories repre-sented the betrayal of ‘grand theory,’ Mills saw Lazarsfeld and Merton’s concentration on ‘theories of the middle range’ (workingconcepts and models like the ‘two-step’ flow

Reactions, ruptures, and rediscoveries 55

1111234567891012341115678920123456789301234567894012345678950111

structuralfunctionalism

thesociologicalimagination

theories ofthe middlerange

model of communications) as prime examplesof the derelictions of what he called ‘abstractedempiricism.’ How, he asked, could Lazarsfeld,in his well-known study of the 1940 electioncampaign in Erie County, Ohio, The People’sChoice (1944), focus so enthusiastically onvoting behaviour and still make ‘no referenceto the party machinery for “getting out thevote,” or indeed to any political institutions’(Mills 1970: 63)?

As noted above, though, this characterisa-tion of Lazarsfeld as an essentially conservativefigure does him less than justice. In a speech hemade soon after arriving in America fromVienna (where he had been active in the social-ist movement), Lazarsfeld described himself as‘a Marxist on leave.’ Mills clearly felt that thisleave had become permanent. However, a careful reading of Lazarsfeld’s writings revealsstrong continuities with his radical youth. In an essay written jointly with Merton, publishedin the same year as The People’s Choice, he had presented a strong critique of corporate control over communications, arguing that‘increasingly, the chief power groups, amongwhich organised business occupies the mostspectacular place [seem] to have reduced directexploitation, and turned to a subtler type ofexploitation, achieved largely by disseminatingpropaganda through the mass media of commu-nication’ (Lazarsfeld and Merton 1960: 493).

Mills died in 1962, before one of the othermajor challenges to structural functionalism,the revival of phenomenology, had got fullyinto its stride. Had he lived to comment, hisobservations would have been highly critical,since the leaders of this movement signallyfailed to take account of ‘the historical struc-tures in which the milieux of everyday life areorganised,’ or to relate biographies to historiesas he had advocated (Mills 1970: 175). Theattempt to build on the phenomenologicalapproach to everyday life, as pioneered byAlfred Schütz, was led by Harold Garfinkel,who called his work ethnomethodology be-cause it was based on closely observed accountsof the ‘methods’ that people (‘ethnos’) use ineveryday encounters. In many ways, the re-search collected in his best-known book,Studies in Ethnomethodology (1967), may be

read as an attempt to radicalise Parsons byinterrogating the tacit agreements that hismodel of social consensus depends on, butwhich he had taken for granted (see Sharrockand Anderson 1986: ch. 3). By exposing theprovisionality of the rules and schemas gov-erning personal interaction, and detailing howthey were continually recreated and reaffirmedin everyday social encounters, Garfinkel pre-sents ‘ordinary’ people as the true architects ofsocial order and social change. In this concep-tion, ‘social structure cannot refer to anythingmore than members’ everyday sense [of it] sinceit has no identity which is independent of thatsense’ (Filmer et al. 1972: 54).

This radically reductionist account of socialstructure, with its exclusive focus on the micro-politics of everyday life, could not offer a con-certed challenge to structural functionalism,because it refused to examine the structural sed-imentation of institutional power. Such chal-lenges were beginning to gather momentumelsewhere, however. When the political consen-sus was finally broken open in the 1960s – bythe Civil Rights movement, the opposition tothe American war in Vietnam, and the inter-national student movements – critical traditionsof inquiry began to gain increasing currencyand support. Two intellectual movements, inparticular, have had a major impact on con-temporary work in media and communications:the revival of critical political economy and thedevelopment of cultural studies.

Although some of the key figures in thereturn to political economy, such as DallasSmythe and Herbert Schiller, had trained aseconomists, their political allegiances werealways to the older tradition of inquiry with itsdeep roots in longstanding debates about therelations between economic organisation, cul-tural life and the common good. From the mid-1960s onwards, writers like HerbertSchiller returned to this central focus on therelations between the production and cir-culation of material goods and the constitu-tion of the good life and the good society.Borrowing from Marx and from earlier radicalcommentators like Upton Sinclair, they devel-oped a powerful critique of the role of theAmerican media and communications indus-


abstractedempiricism

ethno-methodology

criticalpoliticaleconomy

tries in supporting prevailing inequalities ofpower and benefit, both at home and overseas.

At the same time, the emerging field of cul-tural studies� was exploring how the generaldynamics of capitalist cultural productionworked themselves out in styles and repre-sentations which were carried by everydayartefacts, and in the strategies devised by audi-ences and consumers in constructing their ownmeanings and uses. From the outset, culturalstudies was a point of intersection between twointellectual traditions – ideology critique andethnographies of everyday cultural practice.Consequently, within the corpus of literature ithas produced, accounts of powerful media havecontinually rubbed up against celebrations ofaudience refusal and resistance. Analyses of theunequal distribution of vernacular and radicaldiscourses and meaning systems (as in DavidMorley’s influential work on The ‘Nationwide’Audience (1980)) attempted to mediatebetween the two traditions.

ON NOT REINVENTING THE WHEEL

For many analysts, including myself, illuminat-ing the exercise of power and structural con-straints and exploring the possibilities forchange remain the central aims of a criticalsocial-scientific approach to media and com-munication. As I have tried to indicate, in pur-suing this task, we have a rich stock of conceptsand methods to draw upon. Their originatorsare not distant figures to be consigned to dustyback rooms in the museum of ideas. Theyremain our contemporaries. We still confrontthe central questions they grappled with, andtheir search for answers still has much to teachus. We are part of a continuing conversationabout the structure and meaning of moderntimes and the ways they are changing. Theystand at our shoulder, advising, carping, urgingus on. To refuse their invitation to debate is tocondemn ourselves to regularly reinventing thewheel.

On not reinventing the wheel 57

1111234567891012341115678920123456789301234567894012345678950111 � cultural studies – also Chapter 2, p. 39

1111234567891012341115678920123456789301234567894012345678950111

Systematics

PROCESSES OF MEDIATED COMMUNICATION

An implicit model of communicationThe bulk of previous media studies are premised on a basic and sometimes implicit model ofcommunication which centers attention on senders, messages and recipients. That is in spite of the fact that most current research recognizes problems in thus segmenting the process ofcommunication, and in divorcing mediated communication from its social and cultural contexts.

The chapters of Part II follow this model of communication as a structuring device that facil-itates a review of earlier empirical as well as theoretical contributions. Each chapter, in variousways, also considers how evidence about, and conceptions of, one stage of communication reflecton the process as a whole. In particular, Chapter 9 returns to the issue in an analysis of the severalstages of media uses and effects. Furthermore, Chapters 11 and 12 examine research on mediaas they enter different cultural and historical contexts.

The different traditions of research which are covered in the following chapters are often char-acterized by a focus, for instance, on particular types of media organizations, or on specific genres.Similarly, the traditions are distinguished, in part, by their reliance on particular methodologies.

Media organizationsThe three chapters on media organizations address both fiction and fact, entertainment and news,which have given rise to different forms of production. The presentations also draw attention todifferences between media types, and between the national and international levels of media organ-ization. Finally, the chapters include both reassessment and critique of classic studies, and a con-crete illustration of field research on media production.

• Ficton production (Chapter 4). The chapter reviews the several interrelated levels ofdetermination which shape the final media product, including the international economy,technological developments, and professional work routines.The importance of each level isexemplified with reference to a case study of production for cable television.

• News production (Chapter 5). An analysis of news studies over the last five decadesidentifies three main traditions, and goes on to suggest their compatibility, in the light alsoof changes within the media themselves.As such, the chapter addresses the various levelsof determination in media production (presented in Chapter 4) from the perspective ofdifferent research traditions.

IIPART

Introduction60

• International news (Chapter 6).The area of news, and of political communication generally,has particular implications for international relations, which have been an importantconcern in media studies.The last chapter on media production reviews studies of the flowof news in the world as a fourth tradition of news research, alongside the three traditionsidentified in Chapter 5.

Media textsThe messages of mediated communication have been studied as ‘texts’ and as ‘contents’ by dif-ferent theoretical and methodological traditions. Fiction has primarily lent itself to qualitativeapproaches, as derived from the study of literature and other arts, whereas factual genres havebeen examined by a mixture of quantitative and qualitative approaches. In addition, the moderntechnological media have posed research questions concerning the specificity of audiovisual ‘texts,’and concerning the interrelations between media texts in ‘intertextual’ networks.

• Media fact (Chapter 7). The first chapter addresses both quantitative content analysis andqualitative discourse analysis of media texts.The chapter reviews developments withincontent analysis in recent decades, and compares different varieties of discourse studies.One of the example analyses takes up still images and their interrelations with verbal texts.

• Media fiction (Chapter 8). Following a brief account of quantitative studies of fiction, thesecond chapter gives special attention to semiotic, structuralist, and narratologicalapproaches to media texts. Referring to the classic feature movie, The Big Sleep, as itsrecurring example, the chapter shows how various models of analysis may be applied todifferent media, including moving images and sound.

Media audiencesFrom the beginnings of the field, audiences have been a central object of analysis, and have beenstudied especially by quantitative social science through experimental and survey methodologies.More recently, research with a textual and qualitative orientation has begun to examine audiencesempirically. Indeed, ‘reception’ and ‘effects’ have been a key area of convergence in recent decades.

• Media effects (Chapter 9). After a brief history of the notion of ‘effects,’ the chaptersummarizes the multiple traditions of inquiry which have developed since the 1930s.Theirvarious contributions are systematized with reference to stages of communication andinfluence, and to short-term and long-term effects.

• Media reception (Chapter 10). This chapter begins by conferring the recognized ‘milestones’of the effects tradition with important contributions to reception studies. Next, the chapterreviews varieties of reception analysis, and considers the potentials and problems of‘ethnography’ as currently practiced in media studies.

Media contextsThe last two chapters under the heading of systematics take up the interconnections of mediawith their social, historical and cultural contexts, thus emphasizing the communication aspect ofmedia and communication research.The purpose is to present relevant adjoining fields of research,

Processes of mediated communication 61

1111234567891012341115678920123456789301234567894012345678950111

some of which have already been incorporated into the media field, while others may hold anuntapped potential. Specifically, computer-mediated communication is redefining the distinctionsbetween media, text, and context, and calls, not least, for theory development in further research.

• Space – culture (Chapter 11). On the one hand, media are embedded in social and culturalspaces – from the local to the global level.The chapter gives special attention to researchon the cultural formations that are associated with geographical areas, but also considers,for example, the subcultural formations which the media help to maintain. In addition, thechapter provides an overview of research so far on computer media.

• Time – history (Chapter 12). On the other hand, the technological media are embedded in,and emerge from, the long history of human communication. Perhaps surprisingly, ‘themedia’ is a recent notion, dating from the 1960s.The chapter covers the history both of‘communication’ and of particular ‘media,’ and presents concepts and methodologies forunderstanding the modern media in historical perspective.

Determination in the first instanceEach chapter examines a number of factors that affect processes of mediated communication,both within the media and in their social contexts.A shared premise of the chapters is that eachindividual factor – whether technological, economic, political, or cultural – may exercise a deter-mination in the first instance (Hall 1983: 58), but not a determination in the final instance. Neitherthe concrete products nor the actual practices of communication are the outcome of any simplecausality. Different traditions of research have identified, and tried to explain, empirical variationsof mediated communication. Here, the traditions are brought together and compared as part ofa systematics of media and communication studies.

Since the mid-1980s, the most intensivelyresearched topics in media studies have beenaudiences, their reception and uses of mediatedtexts. Following the preponderance of priorresearch that made claims about the meaningsof such texts, their ideological significance andtheir social effectivities, the emphasis on audi-ences seems both predictable and warranted.One result of this focus, however, is that duringthe same period far less work has focused onproduction processes. This also follows, in part,from the presumption that mass-mediatedfiction remains, in many cases, a ‘factory prod-uct,’ standardized content emerging fromroutinized production processes. We suggest,however, that the situation is more complicatedthan this. Variations within routines, the pointsof tension engaged in contests between stan-dardization and differentiation, are of equalsignificance and equally instructive in exploringthe significance of media fictions. It is necessary,then, to provide a more thorough and detailedanalysis of production practices than the usualgeneralizations about ‘media factories’ allow.

One approach is outlined in David Bord-well’s concept of ‘historical poetics’ (Bordwell1989). Under that category, he suggests that it

is necessary to explore options open to media-makers at given points in time and in specificsocial contexts, attending to industrial, eco-nomic, and regulatory factors. While Bordwelltends to focus on those options generating thestandardized elements found in much mass-mediated material, we believe it is possible todiscover important manipulations of produc-tion processes that indicate the ‘relative auton-omy’ of individuals, groups, even organizationswithin media industries. What this suggestsmost importantly is that any study of the production of media fiction must recognizemultiple types of influence. Factors rangingfrom policy formation to the application of newtechnologies may affect the production of anyparticular media fiction. While research takingaccount of multiple causal elements usuallyprovides stronger explanations, it often remainsthe case that specific instances of productionstudies privilege particular aspects over others.In some cases, this results in reductive asser-tions about causation or influence, although inmost instances it simply means that one factoris taken as dominant, overdetermining allothers. These factors and the relations amongthem may be described as ‘levels of analysis.’

The production of media fictionHorace Newcomb and Amanda Lotz4

• an outline of the main levels of analysis in production research – from political economy to professionalroutines

• a description of the diverse sources of evidence and of relevant methods of analysis• a case study: the production of a cable television series• a review of findings from the case study in relation to the levels of analysis.

relativeautonomy

We emphasize that this term should not implya universally effective hierarchy of influence ordetermination, but it is the case that such a cat-egorization of influence works from moregeneral to more specific sites and applications.The most effective production research willseek to identify the interdependence of theinfluences, while demonstrating their variationsin different cases, settings, and systems.

LEVELS OF ANALYSIS

National and international political economyand policy

The production of particular media artefactswithin specific industrial systems obviouslytakes place within more general contexts.Among the most influential works at this levelare those exploring differences between com-mercial broadcasting or film industries andmedia reliant on various forms of state support(e.g., Blumler 1992; Katz and Wedell 1977;Schiller 1969). Although analyses of productionpractices generally acknowledge such differ-ences, broad assumptions rather than detailedanalysis commonly guide the study of the rela-tions between policy and production. This is so,in part, because descriptive and source materialrelated to media policy often focuses on gener-alizations rather than cases, while detailed casestudies take the constraints of politicaleconomy for granted. Overlooked in bothmodels is the fact that individual productionsare enabled as well as constrained by generalconditions. Varied responses to those condi-tions illuminate the complexity of the largerstructures, reminding us that while media pro-duction is indeed a modern, factory product,the differences among the products are astelling as their similarities. Nevertheless, anumber of policy works provide useful contex-tual information for production research.Among them, Alexander et al. (1998), Hoskinset al. (1997), and Moran (1996) offer extensiveoverviews of contemporary media industrypolicies that can be applied analytically.

Some studies do, in fact, bring together themacro-levels of policy and economic structurewith analyses of cases. They include explo-

rations of the ways in which media productsare affected by social problems, such as ‘cen-sorship’ or ‘violence.’ Doherty (1999) andGardner (1987) provide examples of the firsttopic, showing how particular American filmswere produced before and during periods ofheavy social control. Cowan (1979) focuses on engagements with policy by individuals(e.g., Norman Lear) and institutions (e.g., TheWriters Guild of America) with regard to sexand violence on television, and shows how pro-duction strategies were affected by congres-sional actions mandating a ‘family hour’ forcommercial television.

As these publications suggest, a majorapproach to studying the relationship betweenpolicy and production has been historical.�Boddy (1990), for example, explores relationsamong television executives, the U.S. Congress,and television critics in the 1950s. He carefullyestablishes how, in the struggle among thesegroups, industry executives managed to securetheir economic interests through legislative andjudicial decision-making. The outcome of thesebattles led to major industry developments,such as the shift from ‘live’ television produc-tion in New York to filmed programming fromHollywood, resulting in fundamental changesin aesthetics, altered production practices, andultimately the distinctive place of televisionfiction in U.S. culture.

Methodologically, all these works havedepended on the analysis of archival data.�While public policy records are usually freelyavailable, corporate papers have sometimesbeen deposited in reference archives, makingaccess for researchers relatively simple. In othercases, such materials may be proprietary andaccess severely restricted. These records areessential for production research, because thedocuments contain evidence both of conflictingpoints of view and of concrete decision-makingrelated to particular media artefacts.

Levels of analysis 63

1111234567891012341115678920123456789301234567894012345678950111

� historical studies of media – Chapter 12

� archival data – Chapter 12, p. 203

commercialand state-supported

media

Specific industrial contexts

Historical approaches have also been promi-nent in research examining the institutionalconfiguration of media industries, but hereanalysis is focused more precisely on specificindustrial practices. Among the strongest exam-ples is The Classical Hollywood Cinema(Bordwell et al. 1985), which examined thedevelopment of the Hollywood film industryand the resulting reliance on a particular nar-rative style, as indicated in the title.� The studytracks the establishment of regularized indus-trial strategies, consequent divisions of labor,instrumental applications of new technologies,and other features of the Hollywood film fac-tories. It suggests that, after a period of exper-imentation, the U.S. film industry narrowedinto certain industrial operations that weredeveloped in the service of particular narrativeconventions. The analysis also demonstrateshow this general pattern of regularization wasrealized in particular films. The central argu-ment points to the reduction of possible narra-tive strategies and, ultimately, of the styles,genres, and meanings within an industry thatwas increasingly successful on its own terms.The ideological result was the establishment ofa particular cultural meaning of ‘cinema’ to theexclusion or marginalization of alternativeforms.

Here again, researchers rely heavily on pri-mary historical records – contracts, inter-officememoranda, extant interviews, handbooks,production manuals, instructional pamphlets,variously revised scripts, and story conferencememoranda recording decision-making pro-cesses. These are explored in order to describe,analyze, and contextualize the actual productionpractices involved in film- and television-making. Because sustained archival research isneeded to uncover more evidence, and becausethe study of media fiction, in particular, is ofrelatively recent development, new historiescontinue to refine our knowledge of much-needed background and circumstances. Hilmes(1997), for example, provided new informationregarding the shifting arrangements among

media industries, including the radio industry,which have altered the cultural definition of‘film’ and ‘television’ in the U.S.A.

A contemporary example illustrates howthis level of analysis takes research beyondmore generalized descriptions. Montgomery(1989) focused on the different ways in whichinterest groups engage television networks inorder to gain a more favorable representation.Using interview and ethnographic methods, aswell as analysis of records and contracts, shealso examined individual television texts toshow how these groups variously succeeded orfailed in their attempts to alter detailed tele-vision production practices. A similar case is in progress, and invites study, as we write. Inthe U.S.A., the National Association for theAdvancement of Colored People has recentlysecured contractual arrangements with majortelevision networks that will lead to theemployment of more people of color in televi-sion’s executive and creative communities(Jensen et al. 2000). In this case, as in casesstudied by Montgomery, interest groups haveused strategies, such as boycotts or threats ofboycotts, which recognize the role played bypolitical economy and government policies inmedia production.

Particular organizations: studios, productioncompanies, networks

Studies of institutional relations often rely onthe next more specific level of analysis, explor-ing the connection between an organizationand the industrial configuration in which itoperates. Textual analysis of individual worksor collections of films and television programsis much more prominent at this level, frequentlywith an emphasis on genre and format asindicative of an organizational ‘style.’ An out-standing example of this type of analysis isSchatz (1988), who focused on regularized andsystemic aspects of the film industry, usingsources and approaches similar to other histo-ries of cinema. In using archival sources, how-ever, his primary method was the case study,and by focusing on several studio organiza-tions, the production of particular films, andthe roles of powerful individual studio heads,

Production of media fiction64

� Hollywood cinema – also Chapter 8, pp. 119, 121

interestgroups

he placed greater emphasis on human agencyand documented important variations. Thefindings, again, indicate greater diversity within‘the studio system’ than is sometimes assumedin studies of the general institutional arrange-ments.

For television, a primary example is Feueret al. (1984), a study of the MTM productioncompany, which argues for the existence of a‘signature style’ associated with a number of itsproductions. The identification of that styleenabled the authors to describe variationswithin the general structures of both genre (thesituation/domestic/workplace comedy) and theU.S. television industry as such. Cunningham(1988) provided a similar example of housestyle from Australian television. Such studiesmust rely heavily on company histories andproduction case studies in order to support the textual analyses that identify particularelements as a distinctive ‘style.’ In addition,they may make use of interviews and observa-tions when access to individuals and ongoingproductions can be arranged.

Individual productions

A yet more specific level of analysis focuses onindividual artefacts – films or television pro-grams or series. Here, for example, we wouldinclude works exploring ‘the making of’ par-ticular films and television programs such asCarringer’s The Making of Citizen Kane(1996). Often, such works are designed to bemore popular in appeal, providing behind-the-scenes information for fans or interestedobservers. Their popularity, however, does notnecessarily diminish their usefulness for morecomplex research, and they may be cited as evi-dence in any of the other types of analysisdescribed above. And, when such cases areexamined within a more generalized theoreticalframework, they can result in productionstudies of great analytical power. Indeed, schol-arly works often offer similar information,making them informative for general readers as well as for researchers. A primary examplefor television is Gitlin (1983), who exploredvarious fiction productions in order to showboth variation and similarity within the process

and in the resulting product. Relying on inter-view and observation methods, combined withclose analysis of both production techniquesand narrative strategies, Gitlin used his cases tosupport more far-reaching inferences regardingthe role of television in American culture.

One of the most significant recent examplesof work focused on an individual television pro-duction is Julie D’Acci’s Defining Women: TheCase of Cagney and Lacey (1994). In this book, D’Acci traces the development of the pro-gram, explaining the roles of individual writer-producers, actors, studio executives, networkheads, programmers, publicity teams, and otherparticipants. She also examines the responses ofcritics, viewers, and organized interest groups,showing how their commentary contributed tokeeping the series on air, in addition to contin-uing the debate on television portrayals ofwomen. By combining this wealth of back-ground material with her own detailed textualanalysis, D’Acci presents one of the most com-plete pictures of the production of a fictionaltelevision program to date.

Individual agents

Closely related to case studies of projects arethose focusing on the ‘makers’ of fictional con-tent, on their enactment as well as manipulationof all these structural factors. Many works atthis level, among them most studies dependenton ‘auteurist’ theories of creative control, grantextraordinary freedom to individuals and their‘genius’ (e.g., Bogdanovich 1967; Sarris 1968).Equally as significant, however, are contribu-tions critiquing such notions. One of the mostinfluential studies of the television industry, forexample, is Cantor (1971). Using surveys andinterviews in which producers remain anony-mous, her work highlights the systemic con-straints on ‘true’ notions of creativity as it mighthave been exercised by producers. In Cantor’sanalysis, the fundamental structure of Americanmedia industries – rooted in capitalism, sup-ported by advertising, organized as oligopoly,and structured as factory labor – prevents theircreative potential from being realized.

Other approaches have worked from adifferent assumption: that personnel involved

Levels of analysis 65

1111234567891012341115678920123456789301234567894012345678950111

the studiosystem of film

house style intelevision –

also Ellis(1982)

the making of

social roles inproductionunits

auteur theory

in creating media fiction are in fact aware, tovarying degrees, of the constraints and oppor-tunities implied in the ‘levels’ of influence, asreviewed here, which affect their work. Thefinal emergence of any fiction film or televisionprogram is seen as the result of intensely col-laborative processes – something frequentlyacknowledged within the industries as well asin theory. Thus, the work of individuals isviewed as tightly woven into such collabora-tion, which, further, is embedded in the moregeneral levels of influence. In another study ofAmerican television producers, Newcomb and Alley (1983) emphasized the potential forcreative manipulation of the very same systemicconstraints cited by Cantor. Their analysis alsodepended on interviews with producers self-reporting their decision-making processes, andwas amplified with textual analyses designed tocheck those reports.

Primary data for studying the work of indi-viduals need not always be gathered by inter-view or survey, however. For example, ourlibrary lists eighty-five works related to AlfredHitchcock, and while many, perhaps most, ofthese are textual analyses of aspects of Hitch-cock’s films, a number provide original com-mentary on the production process. Gottlieb(1995) collected, in the director’s own words,explanations, theories, and accounts regardingthe production of ‘his’ works. In a related cross-reference, Behlmer (1981), by gathering David O. Selznick’s memos relating to numerousproductions, offered another, the producer’s,perspective on some of Hitchcock, the director’s,projects. A complete analysis of the specific pro-jects on which they collaborated, and of theirrespective individual contributions, may bedeveloped only in the production contexts em-bedding their sometimes conflicted relationship.

To sum up, these examples support thegeneral recommendation of this essay: toaccount fully for the production of mediafiction, it is necessary, at some point and in somemeasure, to acknowledge the extraordinaryrange of levels of influence, from the broadeststructural arrangements to the most particularcreative or administrative decisions made. It isthe interdependence of these factors which,above all, defines media production practices.

SOURCES AND METHODS

In order to develop an accordingly complexstudy of media production, it is necessary toapply a wide range of analytical approaches toan equal range of sources. Research on currentfiction production may usefully begin bydescribing the historical development of thecontemporary situation. As already indicated,newer histories continue to provide substantialadditional detail, hence elaborating a moreprecise understanding of how media industriescame to their current status. A project wouldnext require a description of the general regu-latory and economic context at the level of themedia system (commercial, public service,mixed, etc.). In addition to legislative sourcesconcerning such industrial formations, macro-level information is usually available in national statistical abstracts. These lattersources provide details of import and export,viewer ratings and other statistics for television,gross numbers of completed productions forvarious media industries, distribution andattendance figures for films, the contribution ofspecific industries to the Gross Domestic Prod-uct, and so on. The significance of this infor-mation is often best recognized in comparativestudies, as in Sinclair et al. (1996), where shiftsin national policies and support systems arelinked to changes in production practices.Comparative data may also be found in publi-cations of organizations such as the EuropeanAudiovisual Observatory or the EuropeanBroadcasting Union.

For the study of individual corporations orproduction companies, some limited inform-ation may often be found in public corporaterecords, annual reports, and similar documents.More general information such as size of com-panies, principal officers, location and address,and recent projects is provided in sources such as the annually published InternationalTelevision and Video Almanac and the Inter-national Motion Picture Almanac. It is far moredifficult to obtain access to current corporaterecords concerning specific projects or corpo-rate strategies. As alternatives, or complements,both original interviews and trade press report-age are valuable on this topic, but researchers


creativity andsystemicconstraints

nationalstatistics

corporaterecords

must be aware that much of the informationmay be designed for public relations purposes.

Even more difficult to obtain is informationrelated to the costs and other financial arrange-ments of particular productions. Still, general-ized budgets are widely acknowledged. It is wellknown, for example, that the average cost of aone-hour episode of a U.S. fictional televisionseries was between $1.5 and $2 million in1999. With this factor in mind, it is possible toassess the significance of the decision by theNBC network to pay $13 million per episodefor its top program, ER. Similarly, high filmbudgets, such as those reported for Titanic, orlow expenditures, such as those for The BlairWitch Project, are discussed in the trade andgeneral press as directly related to aestheticchoices, creative decision-making, and resultingworks.

But financial matters are only part of thecomplex negotiations leading to the productionof fiction. It is more difficult still to account forthe exercise of power that is involved in bring-ing a film to the screen or a television programto distribution, because the process involvescomplicated interactions involving many com-plex organizations. It is perhaps for this reasonthat historical production studies, with somebenefit of hindsight, have been among the most informative in recent years. Works suchas Schatz (1988) are based on archival recordswhich, somewhat surprisingly, maintain de-tailed accounts of some of the most compli-cated, acrimonious, and revealing exchanges inthe production of particular films. The accountof struggles within the creative process areextremely instructive, as long as one remembersthat each case is likely to work variations onstandard industry practices.

Comparable contemporary ‘behind thescenes’ information, while often among themost important sources in these matters, maybe the least available. Most of it must be gath-ered from the ‘trade press,’ newspaper andmagazine publications focused on the mediaindustries. In the U.S.A., the daily and weeklyeditions of Variety provide extensive coverageof the financial arrangements within the film,television, cable, news media, music, and legit-imate (stage) theater industries. Moreover, these

papers present detailed information about indi-vidual productions, publish running records ofbox-office receipts, provide extensive coverageof countries other than the U.S.A., and fre-quently offer interpretive analysis of industrychanges. Many other trade journals are equallyuseful, for instance, Broadcasting and Cable,Advertising Age, and Electronic Media.

Ultimately, in order to develop a broadunderstanding of any fiction production, it isnecessary to supplement such backgroundinformation with field research. This entailsobservation of production practices and inter-viewing of the personnel involved, often sup-plemented by published interviews and otherlibrary sources. Perhaps surprisingly, it is oftenrather easy for academic researchers to gainaccess to media production sites, where creativepersonnel working on a project are likely to berather open regarding the choices they make,though less likely to provide details related toindividual power struggles.

Thus, preparations addressing the variouscontextual levels should precede the analysis ofa specific set of creative practices. The morethorough the preparation, the more precise andefficient the observations and interviews con-ducted in the case study. Demonstrating fullpreparation also makes it more likely thataccess to the production site will be granted byself-conscious professionals. To illustrate thesepoints, and to elaborate the various levels ofanalysis in production research, the followingsection explores a specific research project.

ANY DAY NOW: A CASE STUDY

Preparing the project

The case study� was designed as part of dis-sertation research, conducted by Amanda Lotzfor a project focusing on the representation offeminist discourses� in contemporary U.S. television. The larger study examines three tele-vision series on three different networks insome detail – Ally McBeal (FOX), Sex in the

Any day now: a case study 67

1111234567891012341115678920123456789301234567894012345678950111

productionbudgets

trade press

field research

� case study – Chapter 14, p. 239

� feminism – Chapter 2, p. 34

City (HBO), and Any Day Now, an originalproduction for the Lifetime cable network. On-site production research was conductedonly for this last series, and is supplementedwith detailed institutional, textual, and audi-ence analyses. The last program was selected,in part, because Lifetime promoted itself as‘television for women,’ but also because of therichness of its textual representations. Theseries is, in addition, distinctive by being pro-duced originally for cable distribution, ratherthan purchased from a studio following broad-cast on network television, which was thecommon practice in earlier periods of cabletelevision programming, and still more com-mon than the production of original programs.

Any Day Now is an hour-long dramaticseries focusing on the friendship between twowomen living in Birmingham, Alabama: Rene,an African-American, financially successful,single lawyer, and Mary Elizabeth, a white,working-class housewife. Set both in the presentand in the 1960s childhood of the two women,the series dramatizes the different forces thatmake their friendship challenging. The seriesuses an innovative narrative structure, alternat-ing representations of the present with those of the past, including scenes in which the twofriends are played by child actors. This linkingof past and present carries over into theportrayal of long-term family relationships(relatives appear in both time frames), and ofsocial issues: the primary narrative consistentlyfocuses on issues of race in American society,specifically in the deep South. Nancy Miller,Gary Randall, and Sheldon Pinchuk are theexecutive producers of the series, which is shoton set and on location in Santa Clarita,California, forty minutes north of Los Angeles.

Box 4.1 describes the steps involved inarranging and carrying out interviews with keypersonnel as well as observations of the serieswhile it was in production in the summer of1999. Not described in this chapter are addi-tional preparations which included develop-ment of a thorough working knowledge of theproduction history of the program, textualanalysis of previously aired episodes, andexamination of published critical responses tothe show.

As an hour-long dramatic series, Any DayNow is produced on a schedule that has become standardized in U.S. television produc-tion. Following extensive writing and pre-production, each episode is filmed in about eightdays, and the actual production schedule for anepisode is not available until a few days beforeshooting for the episode begins. (Post-produc-tion, including editing, dubbing, and prepara-tion of the sound-track, occurs in a brief periodfollowing production, and was not observed for this project. But since decisions made duringthese processes may be significant for the finalproduct, researchers may attempt to observeand discuss these aspects of the productionprocess as well.) A production schedule dictateswhether the crew is filming on the standing setor on location each day of the ‘episode shoot.’Because it was necessary, as is commonly thecase, to arrange the research visit in advance,the hope was to make arrangements so that theproduction would be filming primarily on setduring the observation period in Los Angeles.Filming on set would provide more opportuni-ties for interviews with key creative personneland for observing various production activitiesthat might not occur on location. Chances weregood that this would be the case, since thebudget for this series allowed only two days ofshooting on location for each episode, with thebulk of filming done on standing interior sets.

It is important to add that the accessafforded in this instance may be atypical of tele-vision production in general. Any Day Nowwas beginning its second season on LifetimeTelevision. Although in no danger of cancella-tion, the executive producers were on a cam-paign to increase the visibility of the series, andwere focusing especially on the program’s atten-tion to racism and the ethnic diversity of thecast. Because of the series’ history – NancyMiller, one of the executive producers, spenteight years trying to get the series produced,and personally viewed it as a vehicle for socialchange rather than financial profit – our expres-sion of interest in the series, and the limitedexposure the research could provide, wereimportant to the producers. In addition, be-cause it did not draw an audience as large as anetwork series, our request was less common


pre-production

post-production

than might be the case for a network programdeigned ‘significant’ by some academicresearchers and journalists for whom cable tele-vision programs, with their smaller audiences,remain less important.

Preparing the field visit

Before making extensive plans for this casestudy, or for making the overall project designexclusively dependent on the single case, pre-liminary telephone calls were made to the Any Day Now production office to discuss thepossibility of a visit. The calls were made a fewmonths before the intended visit, allowing theproject to be reconfigured if the producers werenot open to a visit to their production site. (On the importance of this type of preliminaryplanning, see ‘A counter-example,’ below.) Thefirst contact was made with Executive ProducerMiller’s assistant in April 1999. The assistantresponded positively to the research interests,and suggested that the visit be made later in thesummer, when the writers returned from theirhiatus, the break between periods of produc-tion. At this point, it was clear that Lotz hadestablished a personal relationship with aperson who was aware of the research interestin the series, and who had tentatively approveda visit.

Before granting permission� for the visit,the executive producers requested informationregarding the purpose of the research, thegeneral topics of study, lists of individuals whomight be involved in interviews, and details ofan official university affiliation. Such requestsare common and natural, and researchersshould be prepared to respond in detail. Theinformation was provided in a succinct two-page letter which described the research ingeneral terms, and explained the significance ofthis series for the larger project.

The letter also outlined the aspects of pro-duction that Lotz wished to observe. Theseincluded being present at writers’ meetings,observing production in progress, and inter-viewing writers, producers, and actors. At this

stage, no interviews were firmly scheduled. It isoften difficult to pre-plan which aspects of theprocess will be most beneficial for the largerproject, or which appointments might have tobe canceled later. Because the production ofmedia fiction is affected by anything from anactor’s illness to bad weather, researchers mustbe prepared to respond quickly to changes inschedules, and to seize occasions for gatheringinformation in unexpected and unplannedways.

A counter-example

Although, as noted above, access to televisionproductions can often be readily obtained, thisis not always the case, as suggested by a lesssuccessful endeavor within the same project. Anattempt was made to arrange an interview withDavid E. Kelley, the writer and ExecutiveProducer of Ally McBeal. As a ‘top’ series inU.S. network television at that time, this pro-gram received much more popular and criticalattention, and has become something of atouchstone in contemporary discussions offeminism.

A letter was sent to Kelley’s offices request-ing an interview during the Los Angeles visitarranged for Any Day Now. The letter alsorequested a telephone interview if a visit wasnot possible. The letter was sent via expressmail to make it stand out from the volumes ofcorrespondence received at Kelley’s productioncompany, and a telephone message was leftwith Kelley’s assistant. In that conversation,Lotz described the project, and the assistantwas informed of the forthcoming letter. A fewweeks later, Lotz received a message from theassistant explaining that Mr Kelley was toobusy for an interview. While the popularity ofKelley’s series may have made an interviewimpossible, it was still feasible to gather a rangeof interviews from trade publications, popularmagazines, and other media which would bevaluable in the larger project. This information,however, would not allow firsthand knowledgeof the negotiating process that is part of anyproduction. Given Kelley’s extremely unusualposition in the television industry and on thisseries (he has written or co-written every


1111234567891012341115678920123456789301234567894012345678950111

� permission of access – research ethics, Chapter 16, p. 289


ANALYSIS BOX 4.1 FIELD VISIT TO THE PRODUCTION SET OF ANY DAY NOW(1999)

Once the agreements had been confirmed, the visit was scheduled, and over the next few monthsthe practical arrangements were completed through several telephone calls. The last call beforethe visit secured logistical information such as directions to the studio and where to park. Becausemany production sites have security guards at entrance gates, it is important to know such details,as well as whose name to give the guard upon arrival.

Other information on logistics made it clear that the daily schedule was often quite unpre-dictable.Writers tend to work fairly stable eight-hour days from about 10 a.m. to 6 p.m., but indi-vidual writers are also variously involved in general production meetings, depending on whethertheir script is currently in production, or if they are writing an upcoming episode. A productioncrew works twelve-hour days, with each morning’s call-time dependent on when work was finishedthe previous night. Typically, shooting concludes between 9 and 10 p.m., and production resumesthe following day around noon, although many members of the crew are on hand and preparingearlier. Securing this information early on made some aspects of production at least partially predictable, but, as we indicate below, hardly controllable.

The first entry on the set of Any Day Now came at mid-afternoon on a Tuesday, following arrival in Los Angeles that morning, securing a rental car, and driving to the studios north of thecity. Production had started for the day, and the writers were at lunch.The assistant who had beenthe main contact provided a tour of the site, including the studios and the offices. She introducedLotz around the office and to the production staff as ‘Amanda from the University of Texas who is doing her dissertation on representations of ethnicity and feminism on television. So she’slooking at Any Day Now’. Once people were aware of the general purpose of the visit, they were eager to answer questions and discuss the show.

Three and a half days were spent visiting the series, a duration largely dictated by the timeavailable to Lotz for this portion of her project. Despite the relatively brief length of the visit, itwas sufficient for the purpose of the inquiry, and, because such visits are granted out of profes-sional courtesy, it is unlikely that a longer stay could be arranged except under exceptional cir-cumstances. By the end of the second day, the components of the production company and thegeneral rhythm of production were clear. The final day and a half was helpful in securing inter-views – which were repeatedly rescheduled, but which also, unexpectedly, made it possible toobserve a promotions meeting. Many events and meetings were scheduled, canceled, and resched-uled in just these few days, making it difficult to plan ahead for every event, and, because a limitedamount of time had been allotted for the studio visit, much of what actually took place had tobe left to chance. However, even if the visit was thus completed without securing some import-ant interviews, a relationship had been established, assuring that additional or follow-up interviewscould easily be arranged at a later date.

On the set of Any Day Now, Lotz was largely free to do as she wished. If the writers weremeeting, she generally sat with them, and when they were out, or working independently, shevisited the set. Here, the observations were, in part, determined by the concrete productionprocess in film and television, which is very slow going, indeed quite tedious. Production of AnyDay Now followed standard television industry practices and protocols. For each scene, the actorscome on the set first, rehearse their lines and the ‘blocking,’ the process in which they learn their‘marks,’ their positions and movements during the scene. Next, the actors move out, and the‘second team’ comes on: doubles for the actors take their places while the production crew setsthe lighting, camera, and audio equipment. The full process may take as much as an hour for ascene of less than five minutes. Once the stage is set, the actors return and perform the scene,which is repeated until the director is satisfied with the ‘take.’ Being present on the set affordedparticular opportunities to observe the dynamics among the writer, director, and actors.

logistics

theproduction

crew

theproduction

set

the director

cont.


1111234567891012341115678920123456789301234567894012345678950111

For this particular project, however, observing the writing process and the roles of the exec-utive producers was more important. In their meetings, the central decisions regarding the seriesconcept, the contribution of individual episodes to that concept, and the general social and politicalgoals of the executive producers became increasingly clear. During the visit, the writers wereworking on scripts for the final episodes of the season, planned to air about six episodes afterthe one in production. The writers used Nancy Miller’s office, a comfortable space lined with overstuffed couches and chairs and decorated with memorabilia from Any Day Now and Miller’sother series. Here, it was possible to observe meetings on each step of the writing process, whichfollowed a well-known and relatively routine procedure.

The process began with outlining script ideas, and proceeded to the presentation of ideas toExecutive Producer Miller, followed by Miller’s discussion of the ideas with the Lifetime executivesassigned to work with the show. After securing approval from Miller and Lifetime, the writerswould continue to develop the story, and the individual writer assigned to an episode would spend a few days writing alone. In the meetings, Lotz was able to watch the group dynamic ofdeveloping and polishing scripts that were in the later draft stages. In other meetings, writersbrainstormed ideas for many other possible episodes and discussed the future trajectory of theseries with Miller.

One fortunate aspect of this visit was that it occurred during ‘pilot season,’ when the companywas in the process of presenting (‘pitching’) ideas for new series to various networks.The environ-ment was constantly chaotic, and it was actually being present within this activity that allowed thebest understanding of the overall production process. A significant amount of research time wasspent merely sitting in the production company office observing the assistants to each of the exec-utive producers. Lotz developed a relationship with the assistants, and gained a great deal of infor-mation about the series through talking with them.Their tasks provided additional insight into theways in which the series was being developed.They also agreed to maintain contact after the visit,making themselves available for inquiries about developments, ratings information, and for address-ing questions that would inevitably arise during the analysis and writing related to the series.

During the visit, it also became clear that there was much to be learned by looking around,listening, and asking simple questions. For example, on the wall in the writers’ office was a list ofcriteria – a reminder to the writers – of the vision the executive producers were aiming for ineach episode. Similarly, by being present in the office during the daily telephone calls between thestudio and Lifetime, it was possible to develop an understanding of the intricate relationshipbetween the producers and the network airing the series.

Because of developments in production, interviews with the executive producers were repeat-edly postponed. In some ways this was beneficial, for as the week progressed, other personnelanswered some questions, while new questions arose. It was necessary to be flexible, but alsopersistent, in order to get some of the important interviews, for example, finally being granted ‘afew minutes’ during a smoke break – that ended almost an hour later.

Although only some interviews were taped, the recorder was also valuable for reviewing eachday’s events during the hour-long commute back to the city. While it would probably have beenpossible to tape the entire writers’ meetings, these were often long, rambling discussions aboutcharacters and current events which went on for hours, and which might have been inhibited by recording. Instead, it was possible to create notes about specific discussions and to gatherinformation important to the larger study.

Staying close to Miller during the week led to attendance at some meetings discussing topicsnot previously defined in the production research literature. For example, a ‘tone meeting’ washeld as part of planning the production of the next episode. Here, the writer, director, and firstassistant director met with Miller, and went through the script to make sure all the participantsagreed on how the episode should be acted and shot in terms of tone and attitude. While the

writers andthe writingprocess

networkexecutives

pilot season

assistants toexecutiveproducers

note-taking vs.tape-recording

cont.

episode to date), the lack of an interview as wellas of observational sources is unfortunate, butnot decisive for the total project.

THE LEVELS OF ANALYSIS REVISITED

We begin this analysis by emphasizing that thecase is not intended as the only model of qual-itative production research. Moreover, it ismost emphatically not an ethnographic study.�That term implies much more extensive in-volvement, more detailed encounters withinformants, and closer engagement with theentire ‘culture’ of the production process (e.g.,Alvarado and Buscombe 1978; Elliott 1972).The present field visit was intended as obser-vational and informational. It provided know-ledge that could not be obtained in other ways,but the primary purpose was to gather materialwith which to supplement textual analysis andlibrary research. It was therefore more open-ended, more opportunistic than a systematicethnographic project would be.

This sort of project, often more available toresearchers, thus reiterates the main pointregarding research on the production of fictionalentertainment. The practices observed and the information obtained in this project – or in any other – must be examined from severalperspectives, from macro- to micro-contexts, allof which affect the final form of the televisionprogram.

At the general level of political economy,�the television series Any Day Now is producedwithin the U.S. commercial television system.Its key purpose is to attract audiences and makethem available to advertisers. Producer-writers,actors, and the entire production staff and crew are well aware of this goal, and scripts,production practices, and post-production deci-sions are made in ways which acknowledge that the fictional narrative will be broken intosections, or acts, between which commercialadvertisements will be shown. Creative person-nel may or may not know what specific prod-ucts will be advertised, but they are aware ofthe general range of items likely to be includedin a program targeted to particular demo-graphic groups. In the U.S.A., and increasinglythroughout the world, both producers andother creative personnel are regularly informedof their audience. On the one hand, the fictionsthey create must serve as relatively comfortablesettings for those products. On the other hand,the fictions serve as forms of expression forstories the producers wish to tell for personal,ideological reasons. In addition, each fiction ispresented on a network with its own strategicgoals within the changing context of U.S.broadcasting.

The specific industrial context of U.S. tele-vision has been fundamentally reconfigured inrecent years by changes in both technology and


term ‘tone meeting’ may be specific to this production company, it is likely that others engage insimilar activities, but may not have been observed in the process. In another instance, by sheerchance, a promotions meeting was held during the visit, and proved an excellent opportunity forgaining information on competing visions of the show held by various participants in the produc-tion process. The meeting included representatives from two promotions companies, one hiredby Lifetime Television, one hired by Miller and Randall (two of the three executive producers), co-star Annie Potts’s publicist (by telephone), and (also by telephone) a representative from SpellingEntertainment, the parent production company.

Much of the information gathered during this field visit confirmed earlier conceptions and ideasrelated to the program, and to the topics of feminism and racism embedded within it, but otherinformation amplified and refined those ideas, and provided details that would not otherwise havebeen available.The observations and interviews further enabled later stages of the analysis to drawon the multiple perspectives of those involved in the creative and production process.

� ethnography – Chapter 10, p. 164� political economy: audience as product – Chapter 9,p. 143

policy. Any current exploration of media pro-duction is therefore undertaken at one of themost fluid moments in media history. As aresult, identifying the general commercial goalof television programming, as we did above,tells us even less than in the past about acomplex process. While the commercial goalstill implies particular economic as well as ide-ological configurations of organizations andcontents, the industrial relations that are estab-lished to accomplish those goals have changedin a dramatic fashion. The recent (2000) mergerof an internet service provider, America Online(AOL), and a media production and distribu-tion company, Time-Warner, illustrates not onlythe merging of massive capital accumulations,but also the media’s increasing dependence onboth vertical and horizontal integration (inorder to control all stages of production anddistribution, or to expand one’s reach withinone stage). AOL, the distribution site, can nowdepend on Time-Warner, the content provider(and, with its cable holdings, also a distributor)for extensive influence throughout multiplemedia industries and multiple audiences.

Fundamental to these changes has been thediversification of distribution outlets, which hasmodified existing production practices in theU.S.A. and around the world. Cable television,along with home video, has placed pressure onboth commercial and public service networksin different national contexts, partly intensifiedby satellite transmission and digital streamingfor computer delivery. As a result, both tele-vision and film industries are now required toproduce more, more cheaply, and to vary theappearance of their products to attract audi-ences who can choose from a variety of contentand playback options.

Describing this context, Caldwell (1995)argues that ‘television’ has shifted into an ageof ‘televisuality,’ in which ‘style’ has become theprimary content of the medium. Television pro-ducers work within the conditions of ‘post-network’ television, in which every programmust strive for some sort of distinction thatmight catch viewers, made ‘mobile’ with theirremote control device. Curtin (1996, 1999)makes a similar point when defining an age of‘neo-network’ television in which a distinctive

‘edginess’ is the norm for the medium, ratherthan any generalized, consensual narratives. Itis within this context that cable networks andtheir programming strategies take on primarysignificance.

The particular organization, Lifetime Tele-vision, serves as an example of the descriptionsoffered by Caldwell and Curtin. It is certainlymore carefully ‘targeted,’ more precisely de-fined than traditional broadcast networks. It isa twenty-four-hour cable network, included in the ‘basic,’ non-subscription offering of cable systems. Launched in February 1984, itis jointly owned by the Walt Disney Co. (50percent) and Hearst Communications (50percent). The network reaches approximately78 million potential viewers, placing it eleventhamong basic cable offerings, but it is viewed inan average 1,319,000 homes in prime time,ranking it sixth in viewership among cablechannels. A more useful description of viewer-ship emerges when we look at the response toa special two-hour movie version of Any DayNow which aired in October 1999. Thatscreening received a 2.3 rating in the overalltelevision viewing population (2.3 percent of allU.S. homes with television, namely about 2.3million homes). Within its cable ‘universe’(around 78 million homes with access toLifetime), the program received a 3.1 rating, or2.25 million homes. Revenues for 1998, com-bining advertising and license fees, totaled $523million, almost double that of the average basiccable channel, and were projected at $550million for 2000 (McAvoy 2000). Expendituresfor programming reached $196.4 million, morethan twice the average among cable channels(McAdams 1999). Included in this figure arefunds for Any Day Now, Lifetime OriginalMovies, and other, non-fiction programming.

In 1994, Lifetime began promoting itself as‘Television for Women.’ In 1998, the networkaveraged 208,000 women viewers in the 18 to34 age group during prime time. In the 18 to 49 age group, Lifetime’s ‘target’ demographicgroup, the network averaged 523,000 womenviewers, and in the 25 to 54 age group, theaverage was 594,000 viewers (McAdams 1999).It is within this context that Any Day Now isproduced.

The levels of analysis revisited 73

1111234567891012341115678920123456789301234567894012345678950111

integration:vertical and

horizontal

moredistribution

technologies

ratings andrevenues

At this individual production, the network’sgoals and successes are among the conditionsenabling Miller to produce her show. Previouslyconsidered too specific, and perhaps too con-troversial in its treatment of race, the series wasrepeatedly rejected by the major U.S. televisionnetworks that still attempt to reach the largestand most widely defined audience in demo-graphic terms, and plan much of their program-ming accordingly. For Lifetime’s more preciselytargeted audience, Any Day Now was an‘appropriate vehicle’ – which is not, of course,the same as a ‘safe bet.’ The show could havefailed to attract audiences, in which case Life-time would have canceled it. But Lifetime’sexpectations of ‘acceptable numbers’ are sub-stantially different from those of conventionalnetwork television, and the show is considereda success.

This is not to say that the program wouldnever have been developed for another distrib-ution system. Indeed, recently, many majornetwork programs have been far more nar-rowly targeted than previously, precisely be-cause cable television has been successful withthis type of edgier, potentially more controver-sial programming. Broadcast networks mustcompete for audiences which are increasinglyaccustomed to such choice, which, in turn,means that those networks are willing to acceptsmaller audiences. The numbers required tokeep a television program on air have droppedsignificantly in the past ten years. Put anotherway, Any Day Now would probably have abetter chance of being programmed on a majornetwork today than when it was first boughtby Lifetime. These changes have benefitted pro-ducers, such as Miller, who not only have moreoutlets for their work, but can present materialmore clearly defined by their own vision andpersonal goals.

The network maintains strong ongoinginvolvement with all aspects of the productionthrough making suggestions and decisions. Likemost television networks, Lifetime assignsexecutives to be responsible for ‘Current Pro-gramming.’ The ‘Current’ office and staff areparalleled by another organizational divisionresponsible for the ‘Development’ of new pro-gramming. Executives from these offices

express their views in the form of ‘notes,’ orsuggestions, presented to producers, writers,and sometimes actors.

As is typical of series–network relations,Executive Producer Miller cleared script ideaswith Lifetime, first when the writers originatedgeneral plot ideas, and again once episodeoutlines were firmly in place. Miller expressedsatisfaction with the network’s willingness toair more controversial themes and more explicitdialogue than a broadcast network mightaccept. Still, the network exerted its influencein other ways. One writer commented that the network often expressed concern that thecontent was not ‘dramatic’ enough, and soughtfor the writers to incorporate more melodrama.The network’s influence was also evident whena writer suggested a story about aging. Millerdismissed this idea before it was developed,because the network did not want to acknowl-edge the fact that the characters would be intheir forties, perhaps near their fifties, if a realtimeline were applied to their fictional world –Lifetime attempts to appeal to a younger demo-graphic segment.

As with scripts, casting decisions are care-fully discussed with the network. Notes fromthe network are, of course, always backed withthe authority of financial support, which couldbe withdrawn. Such executive commentaryfrom positions of considerable power is oftenresented by the creative personnel, and maylead into a negotiating game, in which somesuggestions are implemented while others areignored. Sometimes, however, the notes areappreciated and important adjustments madeto the program. In either case, the more suc-cessful a program becomes, the more it winsautonomy for the producers, and a higherdegree of creative freedom.

Other participants in the creative processare also given varying degrees of responsibilityand involvement. It is not coincidental, forexample, that a representative of one of the twoprincipal actors was present at the Any DayNow promotion meeting. The role of AnniePotts’s publicist was to put the views of hisclient into play as part of the advertisingprocess. Star performers can be exceptionallypowerful in determining the shape of a pro-


networkpolicies

stars

duction. Although this tends to be even morethe case in film than in television, the concernsof ‘talent,’ like those of all powerful partici-pants, enter into an interactive process. Whilethe contributions of stars and their representa-tives could be considered at the next, individuallevel of analysis, we treat them here as expres-sions of roles within the production contextand of particular types of organizational power.

To analyze the roles of individual agents infiction production, it is necessary to place allthe previous ‘levels’ in relation to the finalproduct. We emphasize the role of the execu-tive producer, who has primary authority in theday-to-day, ‘on the ground’ creative process.While Executive Producer Nancy Miller is fully aware of the commercial role of televisionin American culture, of the network’s goals and its power, and of the influence of the starperformers, she is equally aware of her ownauthority. Indeed, one of her key creative activ-ities is negotiating the concerns of all thoseinvolved parties while maintaining her owngoals for Any Day Now, which she has createdto express her views and those of her collab-orators.

Miller and co-creator Deborah Joy Levinesold the original conception for Any Day Nowto CBS in 1990. This early version of the series,about the interracial friendship of two girls in1960s Birmingham, was designed as a half-hour production combining comedy and dramain what is sometimes referred to as a ‘dramedy.’Miller based the series on a childhood rela-tionship of her own, but changed the ethnicityof one character to make it more interesting.The series was to be produced by the Orionproduction company, then headed by GaryRandall. Days before the series was to go intoproduction, Randall stopped it, feeling the time was not right for a series about children.Next, Miller spent eight years trying to find an outlet, and became partners with Randallbefore she was contacted by Lifetime. Lifetimewas preparing to launch their first drama, andwanted Miller’s series if she could add a con-temporary dimension to the show by depictingthe girls as adults. Miller and Randall reshapedthe series, focusing the genre more precisely andadding the adult characters.

During the field visit, Miller expressed heraim of having the discussion of racial politicscarry over from the series to become ‘water-cooler conversation’ for the country. She grewup in Oklahoma, often visiting relatives in thedeep South, and feels she understands south-erners, but not the bigotry she saw daily in theshape of separate water fountains and otherpractices excluding blacks. Still, Miller said sheunderstands how people can be racist yet havea good heart, a contradiction she explores inthe series. She also feels the country’s struggleto understand ethnic difference and racism is apart of U.S. cultural history largely ignored instorytelling, and another gap she seeks to fill.�

In addition, Miller recognized a dearth ofmulti-dimensional female characters on tele-vision, or even the presence of many women’sstories. She recounted that, after writing forpolice shows, she was sick of writing aboutfemale rape victims and wives who functionedonly to service their husbands. Miller said sheseeks to disprove conventional industry wisdomthat if women are ‘not twelve and weigh eightpounds and have 38D breasts, they don’t wantto see you.’ Instead, she argued, ‘if you makepeople laugh and you’re telling compellingstories,’ you will draw an audience.

Miller designed this fictional world for U.S.,commercially sponsored television, knowing itslimitations and restrictions, but recognizingcable television as a less constrained environ-ment in which executives would be willing tosupport her somewhat riskier project. BecauseAny Day Now has been successful in cable tele-vision terms, as noted above, she has been ableto determine many of the decisions regardingcontent, direction, and ‘tone.’ Her power tocontinue creative control over the programresults from its success, and from her ability todraw on the related commitments of thenetwork and others who hold degrees of powerto shape ‘her’ show. But she must still work tomaintain that trust, with all these factors inmind, by making dozens of decisions, small andlarge, on a daily basis.

So, too, must those on her staff who occupypositions of creative involvement. Although


1111234567891012341115678920123456789301234567894012345678950111

the executiveproducer

� ethnicity – Chapter 11, p. 179

they do not hold the same level of responsi-bility for the overall success and vision of theprogram, many of them hope to achieve justthat authority at some point in their careers.For some of them, working as staff writers, or holding the title of producer, is part of thetraining process leading toward a time whenthey will create programs of their own. Just as Miller negotiates with the executives ofLifetime, then, other producer-writers and staffmembers negotiate with her. These efforts maybe described as a balance between short-termnarrative stakes and long-term narrative inten-tions. In the observed meetings of writers, dis-cussion focused on how the particular narrativegoals of an episode would suit the larger goalsof the series. And while the Executive Producermaintains primary control over the long-termnarrative, she is also immediately concernedwith each episode, even with a particular sceneor a single line of dialogue. In fact, executiveproducers may, and often do, rewrite the effortsof other writers in order to maintain focus. Thefact that writers for Any Day Now specificallynoted that Miller does little rewriting of theirmaterial indicates a clarity regarding narrativegoals, characterizations, and overall content.This is borne out by the list of ‘tone’ sugges-tions, posted in Miller’s office, which werealways visible to other writers as a materialindication of the collective creative process.

Following completion of the writing processand the production of an episode, Miller, herproduction staff, and numerous technical per-sonnel will work to add music, special sound,visual, and other enhancements, all within verytight time constraints. In this final phase, aswell, the negotiation rarely ceases, and if theprogram remains successful the process willbegin once again, when production for the nextseason goes into operation during the follow-ing July.

CONCLUSION

We conclude by emphasizing that whereas thetypes of social interaction observed in this studyof Any Day Now, and common to most mediafiction production, have been repeatedly char-acterized as ‘struggles’ and addressed in terms

of ‘power relations,’ they may also be definedin terms of ‘collaboration.’ Disputes and dis-agreements, sometimes severe, are inevitable,but it would be wrong to suggest that unequalpower relations always reflect fundamentallyopposed perspectives, or that ‘winners’ exercisepower in order to obliterate the ideas and con-tributions of ‘losers.’

This is not to say that the outcome shouldbe studied, much less explained, in terms ofindividual agency, at least if individuals areunderstood only as ‘autonomous,’ ‘coherent,’‘free’ subjects. Individuals often express andenact internalized institutional goals as well asrestraints, rather than any uniquely personalperspectives, as part of an ongoing structura-tion of media and society (Giddens 1984). Ofcourse, collaborations occur within acceptedideological ranges – from single lines andscenes, to episodes, the series as a whole, andeven the media system in which it appears. Suchconstraints and restrictions are recognized,worked within, battled against, and acknow-ledged by the various agents as forces thatshape, but do not totally determine the finalproduction. While the collaborative processdoes work to shut out some perspectives, theemerging fictions can themselves be critical ofsystemic aspects of the social structure. Cer-tainly, this is the case with Any Day Now andits critique of persistent racism in the U.S.A.Our point has been to illustrate the complexityof fiction production as a socially interactiveprocess, carried out by individuals, within insti-tutional, organizational, technological, andother structural contexts.

The discussion of the roles of individualsalso serves as a reminder that the analyticalprocess should move back ‘up’ the levels ofanalysis. For instance, we have noted the tran-sitional role of writers in television, wherewriters move into the producer role and assumecreative control of a series. In the film industry,creative control is primarily assumed bydirectors, while writers are relegated to a lowerstatus and involvement. They are not undercontract, but make professional arrangements,through agents and lawyers, to work on indi-vidual projects. And, unlike television pro-ducers, who are most often directly involved in


filmproduction vs.televisionproduction

the creative process, the role of the film pro-ducer is generally focused on arranging financefor specific productions. In the context of themore general economic arrangements for fictionproduction, still other differences emerge. Filmfinancing is based on income from the box-office and home rentals, both nationally and internationally. Television financing comesfrom either advertising or nationally regulatedlicense fees and other support, or a combina-tion of the two. These differences in political-economic contexts, which will tend to shapeinstitutional structures, organizational prac-tices, and the consequent roles of individuals,should be taken into account in concrete pro-duction studies.

Partly for these reasons, many of the mostsignificant current questions within productionresearch involve corporate mergers, technolog-ical interconnections, and the cultural andother social implications of new industrial con-figurations. When a book may become a moviethat becomes a television series that becomes atheme park ride that becomes a video game thatbecomes a line of toys, production researchersfind themselves involved with new sets ofissues.� The research process may start withany one of these media products, studying the distinctive work processes at a given pro-duction site. But it should ultimately address

not only the goals of creative individuals, but also the configuration of media organizationswith particular industrial strategies which areembedded in, and responding to, large-scalepolitical-economic conditions.

The current state of television in the U.S.A.,and increasingly elsewhere, as defined in partby technological developments leading to more,and more differentiated, distribution outlets,may favor those who, like Nancy Miller, workto place their visions on screens, even if viewedby comparatively smaller numbers. Within thecommercial political economy, the creativeprocess of producing media fiction remainscomplex, dense, and variously inflected bythose involved. Particularly at a time whenmedia systems throughout the world are in aprocess of vital change, resulting from eco-nomic, regulatory, and technological develop-ments, it is important that comparative studiesof production processes be undertaken, includ-ing different national and regional contexts.Analyzing these processes in more detail willcomplement the studies of audiences in relationto media products, and will help explain theequally complicated responses recorded inreception studies. In this way, our understand-ing of the social roles of mediated fiction willbe enriched and more precisely understood.


1111234567891012341115678920123456789301234567894012345678950111

� integrated production and intertextuality – alsoChapter 11, p. 186

Contemporary approaches to research on theproduction of news can be read as a reactionto those functionalist models of the mass mediawhich were dominant in the mid-twentiethcentury, particularly in the U.S.A., but also inother societies that were then embracing aspectsof a dominant American culture (Lemert 1993).Some current research has drawn its theoriesfrom political economy, others from symbolicinteractionism – two traditions focusing respec-tively on the entire social system and on theinteractions among and between individualsand institutions, what sociologists refer to asmacro- and micro-theories.� Still others haveexamined news as a cultural form with par-ticular ideological implications. Despite theircommon rejection of structural functional-ism,� however, these traditions have yet to beintegrated in a satisfactory fashion. There issome irony in this, because the functionalmodels at least claimed to have integratedmicro- and macro-analysis. This chapter takesstock of news production studies with reference

to ongoing changes within the media systemitself. These changes may once again forceresearch to reconsider its premises and proce-dures (see also Cottle 2000), and to exploreways of integrating the several traditions in thearea in future empirical practice.

Following a characterization of the earlyfunctionalist models, this chapter reviews someof the main contributions to news productionresearch since the 1950s. A key point of laterstudies has been that journalists and othernewsworkers do not merely select and combineinformation; rather they can be seen to literally‘produce’ the news in the context of organ-izational and other social frameworks. Threetraditions have emerged which, in distinctiveways, study the social origins and consequencesof news. With shifting technological and eco-nomic frameworks, moreover, have come a dif-ferent set of circumstances for the productionand circulation of news in society. Currentchanges in the social system of news are exem-plified next, with special reference to the U.S.A.The final section revisits the three research tra-ditions and reassesses their explanatory valuewith regard to these changes, exploring to whatdegree they may ultimately be ‘complementary.’

The production of newsGaye Tuchman5

• a description of early functionalist production research• a comparison of three separate traditions in subsequent studies of news production• a review of recent developments within U.S. news media• an assessment of the ability of the three research traditions in combination to account for current

developments of news media.

� macro- and micro-theories of production – Chapter6, p. 91

� structural functionalism – Chapter 3, p. 55

FROM GATEKEEPING TO NEWSPRODUCTION

Media sociology in the mid-twentieth century

One common feature of various functionalistmodels during the 1950s was that theydescribed media as components of a largersocial system in which media organizations andtheir audiences were equal and parallel entities.Drawing their inspiration as students of RobertK. Merton (e.g., Merton 1968), Riley and Riley(1965) proposed one of the best known andinfluential of these (see also DeFleur 1975;Wright 1959). Applying Merton’s middle-rangeapproach,� Riley and Riley implied that masscommunications are a balanced set of socialstructures which contribute to the stability ofthe social system. On the one hand, employees,managers, and other ‘members’ of media insti-tutions respond to the interests of such refer-ence groups as co-workers and competitors. Onthe other hand, individuals receive messagesfrom media institutions, and respond in turn,according to their orientation toward their own reference groups, who represent a parallelstructure within the social system. In a well-functioning system, the reactions and subse-quent actions of individuals in relation to mediaamount to a (rudimentary) feedback, whichprompts the media to either reiterate or modifytheir output, so as to ensure the continued oper-ation and stability of the social system as such.

To be sure, not all sociologists agreed with the full model, as suggested by alterna-tive graphic depictions (e.g., DeFleur 1975;McQuail 1976). Rather than drawing parallelboxes, other models placed the media above therecipients of messages, suggesting a hierarchi-cal system of communication and social inter-action. The media thus seemed to be morepowerful, in both institutional and discursiveterms, than those ordinary social agents who read newspapers, go to the movies, listento the radio, or watch television. Nevertheless,the underlying metaphor of communication re-mained that of the transmission of an inform-

ation flow, which might be used selectively oreven cut off, but which relayed full-fledged messages about a pre-existing reality to mediaaudiences.�

Like much social science, early research onnews had often been qualitative, exploring theplace of this strategic genre in political andother social life. Max Weber (1958a [1918])had found that the press was a set of politicalorganizations rather than being neutral relay-ers of information or scandal, and that jour-nalists were ‘professional politicians.’ Robert E. Park (1922) had examined the role of news-papers in building social cohesion in theheterogeneous American nation. By the 1950s,however, news studies were joining the emerg-ing social-scientific orthodoxy of quantitativemethodologies coupled with functionalisttheories. Although some research combinedqualitative and quantitative approaches (e.g.,Janowitz 1967), most studies involved contentanalysis� or the quantitative examination ofthe editorial and other journalistic decisions of individuals termed ‘gatekeepers’� (White1950). When, for instance, participant obser-vation was employed, it was either placed in afunctionalist framework (Breed 1955), or it waslargely dismissed as ‘unscientific,’ only to berecovered in later news studies (Lang and Lang1953).

An alternative, preliminary model is dis-played in Figure 5.1, focusing on the news org-anization amidst a field of often conflictingsocial forces (see also Curran 2000). Theseforces comprise at least four types:

1 Economic agents. Owners and investorswill, in the nature of the matter, aim for newspolicies and procedures that ensure a profit, in the short or long term. In doing so, theyrespond to both audience and advertiser inter-ests. At the same time, each news organizationmust orient itself toward its competitors, whichoperate in the same or a related force field.

From gatekeeping to news production 79

1111234567891012341115678920123456789301234567894012345678950111

� social theories of the middle range – Chapter 3, p. 55

� transmission model of communication – Chapter 1, p. 7

� quantitative content analysis – Chapter 13, p. 219

� gatekeeping studies – Chapter 6, p. 91

2 Political agents. Political, legal, and otherregulatory entities establish the general frame-work in which news organizations operate. Atleast in the European setting, moreover, publicfunding of broadcasting represents an alterna-tive or a substantial supplement to advertising.

3 Source agents. Sources, on the one hand,represent necessary lines of information feedinginto news, and hence are cultivated to ensurecontinuous coverage of key social sectors. Onthe other hand, both official political sourcesand ad hoc interest organizations amount topressure groups seeking a voice in the news.

4 Audience agents. In addition to being amarket of consumers whose interests any newsmedium must continually try to fathom, in awider sense the ‘audiences’ of a news organi-zation comprise a heterogeneous set of con-stituencies, from other professional journalistsevaluating it as peers, to the general public whowill label and rank news media in relation toeach other.

In responding to and accommodating suchexternal forces, a news organization devisespractices and procedures internally for accom-plishing its daily news work. Because theseforces have so many diverse origins and exer-cise their impact on the news at differentmoments, current news research has come toemploy a range of methodologies.

Making news

During the late 1960s and throughout the1970s, many media sociologists discarded thefunctionalist legacy. Despite their differentsources of theoretical inspiration and theirmore or less explicit ambitions of opposing the powers that be in news and politics, thealternative positions all emphasized a notion of ‘production’ – news is made, not found.Further, it is not the attitude or ‘bias’ of indi-vidual journalists, but their social and organi-zational context which primarily determineshow news is made. This common denominator

Production of news80

News organization C

Pressure groups

Sources

Investors

Owners

News organization BNews organization A

Regulatory agencies

Advertisers

Other social institutions

Audiences

Figure 5.1 The news organization in a field of social forcesSource: Adapted from McQuail and Windahl 1993: 61

was signaled in the titles of a whole list of in-fluential studies on ‘making,’ ‘creating,’ ‘manu-facturing,’ and ‘deciding’ what is news. (Afurther shift toward research on audiences ‘mak-ing sense’ of news began during the 1980s (e.g.,Jensen 1986).) While some comparative pro-duction studies included different regions of theworld (e.g., Golding and Elliott 1979), most ofthis work centered on the American and Britishmedia (e.g., Altheide 1976; Fishman 1980; Gans1979; Schlesinger 1978; Tuchman 1978).

The notion of production, however, wasgiven several inflections. Whereas some criticaland Marxist approaches proposed to subsumenews and media as such under general political-economic categories, examining their specificcontribution to the production and reproduc-tion of the social status quo, other work reliedon a phenomenological framework to explorehow (news) media necessarily contribute toproducing and maintaining society as a sharedsymbolic environment. In addition, textualstudies analyzed the ways in which news textsproduce particular perspectives on social real-ity, while blurring or obscuring others. In anoverview, Schudson (1991) thus identified threetraditions in recent research on the social pro-duction of news, superseding functionalism:

1 Political economy. Building upon classicMarxist ideas, the political economy traditionhas argued that both functionalist and othermodels of mediated communication tend toignore issues of class and power generally (e.g.,Golding and Middleton 1982; Herman andChomsky 1988; Murdock 1982; Murdock andGolding 1977). One relatively separate strand,drawing on Gramsci (1971) and French neo-Marxism (e.g., Althusser 1977 [1965]), has rec-ognized the semi-autonomous status of media,but has suggested that the media neverthelessexercise hegemony by limiting both the specificagendas of the political process and the culturaluniverses made available through media repre-sentations (e.g., Sallach 1974).

2 Phenomenology and ethnomethodology. Asecond group of studies has emphasized the meaningful nature of social life, and haveexamined the role of news media in constitut-

ing and disseminating meaning (e.g., Molotchand Lester 1974; Tuchman 1978). While theinspirations range from traditional pheno-menology� to symbolic interactionism andethnomethodology,� a common premise of thistradition is that individual journalists, singlenews outlets, as well as the media institution as such, collectively accomplish the social con-struction of reality (Berger and Luckmann1966). (For recent studies see, e.g., Chomsky1999; Sumpter 2000.)

3 Textual studies. The third type of newsresearch has argued for a closer reading of newsas narratives that are replete with symbols ofthe society which shaped them, and which they,in turn, reshape and reassert.� Whereas suchstudies typically do not examine the productionprocess in an empirical sense, news texts evi-dently bear witness to this process; constitutethe media-audience nexus in the further dis-semination of news; and can be seen as neces-sary objects of analysis also in productionresearch.

The different empirical foci of the threetraditions are explained, in part, by their theo-retical assumptions. Similarly, their character-istic purposes of inquiry have entailed differentmethodological choices. Political economy hasaccounted for the economic, legislative, and, toa degree, technological conditions of mediatednews, and for inequalities and inadequacies inthe resulting infrastructure, often by relying on quantitative analyses of available statisticalevidence as well as of original datasets. Socialconstructionism has uncovered the routin-ized, yet finely nuanced practices of daily news work, primarily through observational andinterview techniques. In addition, textual stud-ies have typologized and interpreted the for-mats of news, their historical roots, and currenttransformations, especially in qualitative dis-course studies and, to a more limited extent, inquantitative content analyses.

From gatekeeping to news production 81

1111234567891012341115678920123456789301234567894012345678950111

� phenomenology – Chapter 2, p. 22

� symbolic interactionism and ethnomethodology –Chapter 3, p. 56

� news as texts – Chapter 7

‘All three approaches . . . have greatlyadvanced our understanding of the media byfocusing on the specific institutions and thespecific processes in those institutions respon-sible for creating the cultural product we call news’ (Schudson 1991: 150). However, bothSchudson and others sometimes appear toassume that these traditions have not beeninformed by one another. In fact, they have,more often than not, defined themselves throughcompetition or conflict, first with the function-alist heritage and second with each other (e.g.,Ferguson and Golding 1997). Schudson (1991),moreover, chastised all three traditions for being‘indifferent to comparative as well as histori-cal studies’ (p. 156). Partially in response, thischapter begins to show how, jointly, the threeapproaches may help us to understand ongoingchanges in news as a social institution and as apolitical genre over the past three decades.

The next section accordingly reviews recentchanges in the American media system, asrevealed by content analysis, participant obser-vation, as well as structural evidence. The U.S.media serve as an example of more generalissues, in media research as well as in mediapolitics, partly because of their internationaldissemination, partly because of their status asrole models also in professional education andnews organizations around the world. Further-more, the process of ‘conglomeration,’ so evi-dent in U.S. media, is also advancing in othercountries (Croteau and Hoynes 2001: 72–145).Writing well over twenty years ago about theBritish communications industry, Murdock andGolding (1977: 31) noted ‘the overall concen-tration of ownership,’ ‘the network of inter-locking shareholdings,’ and the resulting ‘highdegree of connectedness and communality ofinterests between the various sectors of indus-trial and financial capital.’ All these factorsframe the more localized processes throughwhich media professionals produce both newsand entertainment on a daily basis. Indeed, themain trend of media conglomeration at presentmay not be Americanization (Tunstall 1977),but transnationalization.�

These transformations are explained, to alarge degree, by the simple but far from trivialfact that news is a product – manufactured,sold, and consumed daily. To paraphrase acomment that Max Weber made at the end ofthe Protestant Ethic (1904), however much par-ticipants in an institution, or the institutionitself, may claim that the ‘maw’ of capitalismis irrelevant to its existence, no news mediumor news audience is untouched by the economicconditions of their communication (Weber1958b: 181–182). For empirical media studies,one central question is how changing economicconditions affect both the process of news pro-duction and the final news product.

WHAT’S NEW – IN THE NEWS MEDIA

In hindsight, it is easy to forget a general devel-opment in news media since the 1970s: the dra-matic growth in the total number of sourcesthat consumers have available and combine intheir daily use. To briefly list the newcomers intheir order of appearance, they included cabletelevision, additional television networks, cablenews networks, the Internet, and linkagesbetween cable news stations and the Internet.This general expansion has resulted in in-creased competition, not only within butbetween media.

A daily newspaper and a national nightlynetwork news show compete with other similarproducts, but also with all media distributingnews. On a local level, in southern Connecticut,for example, the New Haven Register vies forcirculation with the New York Times. On anational level, what used to be the three majorAmerican networks in the mid-1970s heldabout 90 percent of the total television audi-ence at any given time, whereas the four largestnetworks currently attain about 50 percent ofthe total (Croteau and Hoynes 2001: 124). Inthis context, network news has faced a crisiswhich is comparable to that experienced bynewspapers and other print media after thecoming of electronic media, as the classic net-works have been challenged by new networksin general, and by cable news networks inparticular (Croteau and Hoynes 2001: 46–68;Hyuhn 2000).


� transnational concentration and conglomeration –Chapter 4, p. 73

competitionbetween moretypes of newsmedia

In order to unravel this increasingly inte-grated news system, it is helpful to identifysome specific challenges experienced by twoclassic news media – newspapers and television– and their attempted solutions in terms of bothproduct development and new productionprocesses.

Newspapers

Once upon a time in the U.S.A., afternoonnewspapers both presented fresh stories andupdated the morning newspaper (Emery andEmery 1988). Because of their competition par-ticularly with television from the 1950s, after-noon editions typically devoted more space toanalysis. When that practice, too, failed tomaintain their circulation, most afternoondailies folded (Kaniss 1991: 35). To survive,newspapers sometimes entered into a ‘jointoperating agreement,’ in which they sharedadvertising and production costs but wereresponsible for their own reporting and edit-orial content, as currently done by the DenverPost and Rocky Mountain News. But even suchcooperative arrangements could not preventfailure for some afternoon newspapers. Thus,in 1998 the afternoon Nashville Banner foldedafter having a ‘joint operating agreement’ withthe Nashville Tennessean for about twentyyears.

Alongside competition, a second reason forthe demise of the afternoon daily in most U.S.cities has been the changed nature of metro-politan areas. As more workers moved to thesuburbs, their need for a daily paper becameless self-evident. For one thing, more peoplecommuted to work in the city by car than bypublic transport, which would have facilitatedreading during the ride. For another thing,many suburbanites no longer identified with thecity (Kaniss 1991).

By the late 1960s, many morning news-papers had succumbed to the competition fromradio and television (Emery and Emery 1988;Kaniss 1991). The newspapers’ response wasmulti-pronged. One strategy for maintaining,or even broadening, circulation was to invokethe central position of the city in all of metro-politan life as well as to present specific activi-

ties as matters of concern to suburban resi-dents. Such ‘metropolitanism’ entailed coverageof entertainment, but also of the urban politi-cal process, which affected the living conditionsof suburban readers both within and outsidethe city itself. An additional, highly practicalmotivation for newspapers to promote metro-politanism was that the city as a political centeris much easier to cover than its many satellites.During the 1980s, such newspapers as the NewYork Times and the Los Angeles Times tried tomake the best of both urban worlds when theybegan to issue special editions for selected loca-tions, featuring regional news, but also repre-senting the city as a pivot in the lives of peoplein each suburb or region.

A second set of strategies in the fight fornewspaper circulation in the U.S. went beyondspecial content, and affected the format of theindividual article (Benhurst and Mutz 1997).Whereas some newspapers now carried shorterarticles, and others aimed for longer ones, bothpatterns may be related to the presentation ofnews on television. As is frequently pointed out,also in self-reflective articles in other news-papers, the format of USA Today (launched in 1984) mimics television: short articles, eye-catching typography (‘blips’), and ‘factoids’(precise but decontextualized items of informa-tion) that reproduce features of national televi-sion news. In addition to the attempt to createa product for readers throughout the country(the U.S. press has traditionally been entirelylocal and regional), USA Today especially seeksto appeal to those who have neither the incli-nation nor the time to seek out more in-depthcoverage.

In an opposite strategy, other newspapershave reacted against television, and have foundalternative ideals for their news formats in themagazine press. By presenting fewer and longerstories, this group of papers announces an insis-tence that print media are capable of moredepth and nuance than the ‘headline service’ oftelevision. Indeed, some research has found that American newspapers today tend to offermore analysis than they did thirty years ago (Benhurst and Mutz 1997). In coveringelections, for instance, newspapers devote rela-tively more space to the issues, and less to the

What’s new – in the news media 83

1111234567891012341115678920123456789301234567894012345678950111

afternoonnewspapers

joint operatingagreements

morningnewspapers

suburbs

metro-politanism

article formats

candidates’ ‘horse race.’ Compared to thespread of ‘sound bite’ news in television duringthe same period (Hallin 1992), this newspaperstrategy is one more indication of how eachnews medium has sought to achieve ‘productdifferentiation’ within an increasingly inte-grated news system.

Within this last group of newspapers, a thirdstrategy of differentiation has been in evidence,which involves a redefinition of the professionalstatus of news workers. In contrast to ‘editors’papers’ or ‘reporters’ papers,’ some presses havecome to be termed ‘writers’ papers’ (Manzella1997). The implication is that managementspermit their writers greater latitude or crea-tivity in developing stories outside of the mostfamiliar protocols or conventions of newswriting (see also Becker 1982). For example,one reporter for the Hartford Courant wrote amulti-week series of articles in 1997 about hisexperiences as he paddled a canoe from thesource of the Connecticut River through theLong Island Sound to Manhattan Island. Aprize-winning series, these stories were notpremised on getting and reporting ‘the facts,’but on telling a real story. Another change ofprotocol at such newspapers has been that theyno longer require stories appearing on the front page to be structured in a classic invertedpyramid, proceeding from the basic elements of who, what, when, and where, to how and, perhaps, why, and to other elaboration,exemplification, and background (Manzella1997).While these journalistic formats recall the‘new journalism’ originating in the 1960s aspart of a political as well as aesthetic counter-culture (Wolfe 1973), current developments areprobably best seen in market terms – in thecontext of such new media as the Internet andall-news cable networks that deliver updated,basic information in a speedier fashion than anynewspaper could.

Television

If the newspaper was the main vehicle of amodern political public, in the U.S. (Schiller1981; Schudson 1978) as in Europe (Habermas1989 [1962]), television took over centralaspects of that role in the post-1945 period,

especially at the national or network level. It isalso at this level that the most important trans-formations have occurred, as increased com-petition from new networks, such as FOX, andnews networks, such as CNN, has spelled adecrease in profits for the three classic networks(ABC, CBS, NBC) (Croteau and Hoynes 2001).Whereas in the 1950s network executives werecharmed to realize that the nightly news couldmake a profit (e.g., Epstein 1973), today’s exec-utives are beset by smaller audience shares andby the demand from the media conglomerateswhich now own these once powerful networksthat they make a profit in themselves (althoughthe audience for any one news show is com-paratively small). Of course, a smaller audienceshare means a smaller budget also for newsoperations. As a result, the individual networknews divisions have decreased their number of correspondents both within the U.S.A. andaround the world (Remnick 1998). At the sametime, the news divisions as a group havebecome more explicitly interdependent, evensharing information, as done, for instance, forexit polls in the 2000 U.S. presidential election,and buying materials from competitors, espe-cially if they can be rebroadcast in a differentformat.

Admittedly, network news reporters werealways interdependent, covering the same storyfor different companies, and routinely chat-ting with one another, which could result ingroup and sometimes ‘pack’ journalism (Crouse1973). Reporters received the same news re-leases, interviewed the same people, and usedthe same techniques to transform film andsound into stories; but through what reportersstill like to call ‘enterprise,’ they might put dif-ferent spins on their stories. Today, both themore limited number of correspondents and thestandardization of the available image andsound, and of press conferences and interviews,have decreased the possibility of differentiationat the level of the individual story.

And yet news channels must differentiatethemselves, their ‘brand,’ to audiences. Onestrategy has been to develop a distinctive mixof stories. Hyuhn (2000), for instance, foundgreater diversity in the news stories offered onbroadcast networks than on cable stations.


productdifferentiation

writers’papers

invertedpyramid

newjournalism

moretelevisionnetworks

newscasts asbrands

Another and perhaps more prominent strategyhas been to profile the anchorperson in at leasttwo different ways. First, the well-documentedprevalence of shorter sound bites (Hallin 1992)has meant that news anchors (and reporters)are on screen for a larger proportion of theprogram. On the one hand, the identification oftheir statements as ‘analysis’ represents a stepaway from a previous practice of introducingand summing up politicians and other news-makers, and thus insisting that the ‘facts’ speakfor themselves. On the other hand, the newsteam’s increased time-on-screen brings them tocenter stage. Some audience members haveprobably always been motivated to watch, forinstance, ABC nightly news because they pre-ferred Peter Jennings to Dan Rather of CBS orTom Brokaw of NBC, but the greater promi-nence of news-readers and, to a lesser extent,of reporters, might suggest that the staff orteam, not the news, is the product.

A second, related development has placedthe anchor more centrally not only in the newsmessage itself, but in its ‘mode of address’ toviewers.� From the advent of television news-casts in the 1950s, anchors had sat behind adesk and read the news, switching their glancefrom hard copy on the desk to telepromptersto the audience, a tradition which the new-comer, CNN, still maintains. However, in 2000,the nightly news shows of CBS and NBCopened with shots of standing anchorpersons,seemingly active and ready to enter the viewer’shome, and symbolically different from theclassic anchormen sitting authoritatively behinda desk and waiting for the camera to slowlypan toward them. (In February 2001, DanRather of CBS was back behind his desk,whereas Tom Brokaw of NBC continued tostand.) While still wearing stylish suits and tiesand speaking in their convincing announcers’voices, these anchorpersons have come to seemless like a doctor or lawyer pronouncing judg-ment, and more like a messenger reaching outto the viewer.

The presentation of news-readers on CNNand local television news appears to support theperception of news teams which the viewer is

invited to join, symbolically, in dialogue. Here,news-readers sit behind long desks and inter-act continuously, a practice introduced by U.S.local news stations in the 1960s. At CNN,news-readers are frequently paired by gender(one man, one woman), and joined at the deskby on-staff commentators or outside experts.On local news, anchors share their desk with asports reporter or even a weatherperson, whothen moves to stand before maps and diagramsin order to present the forecast.

In sum, while these modified forms of repre-sentation and modes of address in televisionnews, in one sense, merely bring out certainpotentials of the audiovisual medium, they arerecruited for specific economic and culturalpurposes in a particular historical context.Communicative immediacy, involving newsworkers and audiences in what Thompson(1995: 82) has called ‘mediated quasi-inter-action’ may be a unique selling proposition(Schudson 1984: 50) for television at a timewhen many more print and electronic outletshave become available; when audiences havelost part of their faith in the immediacy of newsimages as representations of reality; and whencomputer media do not yet offer the sensoryrichness of television. It is such a complex inter-change of economic and cultural factors whichempirical production research is called upon toaddress in future studies, and it is this com-plexity which may be engaged by combiningseveral complementary research traditions.

THREE TRADITIONS OF NEWSPRODUCTION RESEARCH IN REVIEW

The three main traditions in news productionresearch, delineated by Schudson (1991), mightinitially seem to lead in different directions. Thepolitical economy position would, in one senseor another, return research to Marx’s dictum inThe German Ideology (1845–46) that the ideasof the dominant class are the ideas of an age.Hence, it would encourage examination of howthe technological media participate in a widerideological ‘mediation’ – the transformation ofideas into forms of social domination. Theclassic issues have been re-emphasized at themomentous and still contested transition to a

Three traditions of news production research 85

1111234567891012341115678920123456789301234567894012345678950111

sound bites

� personalized address to viewers – Chapter 7, p. 106

global ‘network society’ (Castells 1996), whoseseminal ideas and slogans, again, may servepowerful minority interests. Next, a focus onthe social organization and production of newsmight raise questions, for instance, of hownews workers’ use of the Internet affects thesorts of information they obtain and report asnews (e.g., Borden and Harvey 1998; Garrison2000). In addition, within textual mediastudies, historical and comparative research onhow myths of individualism and of the ‘team’coexist in the wider culture, suggests itself as ameans of accounting for the forms of represen-tation and the modes of address noted abovein current television news, but also for the novelforms of community arising within computer-mediated communication (e.g., Jones 1998).

To indicate the extent to which the three traditions address some of the same researchquestions, I start from within the tradition thatI know best, the social construction of news, orwhat Schudson (1991) called the social organi-zation of news, and consider three U.S. studies(Fishman 1980; Gitlin 1980; Tuchman 1978).To begin, all three books take up both the socialorganization of newswork and what Murdockand Golding (1977) had identified as theproblem of ‘mediation’ – how the ideas of adominant class become the ideas of an epoch.Of the three books, Gitlin’s work on media cov-erage of the U.S. student movement relied theleast on formal participant observation or atheoretical framework of phenomenology. Hisempirical data were derived partly from hisown records of participation in Students for aDemocratic Society (SDS), partly from inter-views with news workers and friends who had been involved in the news stories beinganalyzed. Nonetheless, one basis for his largerargument that journalists and editors ‘putreality together’ was concepts from symbolicinteractionism, such as Goffman’s (1974)notion of a frame� which brings some aspectsof reality into view, while excluding others. Oneof the most interesting portions of Gitlin’s(1980) data concerns those cases where per-sonnel at news headquarters in New York Citytold their team how not to cover a story, and

where, in addition, these instructions had achilling effect on other network teams coveringother stories which they perceived as similar. Ina next step, then, organizational and policydeliberations could be read off at the level oftextual structures as well, even if this was notthe empirical focus of Gitlin’s study.

Compared to such a clear-cut interventionof powerful interests into concrete news deci-sions, the books by Tuchman (1978) andFishman (1980) also included examples of theroutine imposition of what might be called ‘an ideological imperative’ on the selection and structure of news. Tuchman’s study of alocal television station described, among otherthings, how the business office of the stationintervened to obtain newspaper coverage ofstories that involved its advertisers; how theorder of stories in the late evening newscast wasrearranged to please advertisers (an airlinesponsoring the news had an agreement that anair crash would not be the first story when itwas the lead sponsor); and how womenreporters complained that they were orderedspecifically to cover certain stories because theyinvolved women. (The wall between news andadvertising departments has since been break-ing down (Croteau and Hoynes 2001: 163–181; Underwood 1993).) Fishman showed, forexample, how reporters tended to accept andreproduce the presumed distinction betweenpolicy-making and policy implementation with-in municipal government, and hence how newsstories would hide conflicts, compromises, andprocesses of undue influence in local politics.What both Fishman and Tuchman shared with Gitlin, nevertheless, is the assumption thata detailed empirical accounting of the dailyroutines and practices of news production isrequired in order to explain how social infra-structures, institutions, and their interests aretranslated into concrete news texts.

In addition to their different empirical foci, the three studies also relied, in part, on sep-arate theoretical frameworks regarding thedetermination of news by its social contexts,and regarding the agency of journalists as asource of indeterminacy. Gitlin (1980), for one,argued that, in structural terms, the ideas andactions of news workers articulate and serve the


‘mediation’:dominant

classes andideas

� frames – Chapter 9, p. 149

practices androutines ofnewsproduction

interests of a dominant class to which theythemselves belong. However, rather thanaccepting an understanding of ideology as asimple and, perhaps, intentional misrepresenta-tion, the volume resonated with both of theauthor’s classic articles on media sociology,hegemony, and ideology (Gitlin 1978, 1979),suggesting that news systematically preferscertain worldviews.

Fishman (1980) and Tuchman (1978), intheir turn, elaborated the concept of ideologyfrom Smith (1974), and argued that newsmedia are ideological in the sense that newswork ‘is an interested procedure’� and ‘ameans not to know.’ Their point was not the(real) lack of analysis in many news stories, nordid these authors suggest that the ultimatelyclass-based ‘interested procedure’ was the resultof owners or managers interfering with newsprocedures. Instead, like Smith, they proposedto see all methods of knowing, including newsgathering, as ‘embedded’ in a particular socialand historical context (Garfinkel 1967; Giddens1984). As concluded also by Gitlin (1980),news workers do not intend to frame the newsso as to support either private economic or stateinterests, but, because of their institutionalposition and organizational practices, they arelikely to serve those interests, and to reproducetheir structural and power bases.

Any ‘ideological’ impact takes effectthrough concrete news stories – as focused bythe textual studies tradition, but recognized inproduction research traditions as well. Admit-tedly, news texts vary a great deal, for example,in their precise cognitive and narrative frames,which are used to tell particular stories, and which, further, are capable of multipleinterpretations, some of which may oppose adominant ideological message (Gamson 1992).However, in most mainstream media, suchframes as well as other formal features of newsare the product of protocols or conventionswhich transmit at least two additional, implicitmessages. First, news still tends to present itselfto audiences in a few standard formats that canbe taken as evidence of professionalism in

general and objectivity in particular, a notionthat remains central to the self-understandingof news workers, and has been referred to byHallin (1986) as one of the myths of the pro-fession. Second, the selection and combinationof information in familiar news formats nor-mally take place with reference to a relativelysmall set of institutional agendas,� which tendto be reinforced as reporters interact with theirinstitutional sources. To exemplify, the U.S.media only came to question the Vietnam Warwhen their elite sources no longer had oneagenda, but became willing to disclose theirintensifying disagreements with one another asto whether the U.S.A. could win the war (Hallin1986).

When, from to time, the protocols of writingthe news change, it is relevant for research toask how this discursive change relates to pos-sible changes in professional routines and in thepolitical economy of news. The section aboveon changing conventions of newspaper jour-nalism noted the spread of ‘magazine-like’ leadparagraphs in news articles. In political-economic terms, such paragraphs may be inter-preted as a response to competition, specificallythe ability of the Internet and of all-news cablenetworks to deliver information in a muchspeedier fashion than newspapers. In terms oforganizational routines and professional ideals,the actual shift is presumably the result of bothmanagement deliberations and contestationsamong professionals about the relatively pres-tigious ‘product’ which they co-sign. Productdevelopment may occur simultaneously withinseveral organizations, as news workers migrate,and as journalistic associations and publica-tions debate the identity of the genre and of theprofession. As a textual form, the ‘magazinestructure’ seems to announce to readers: ‘Wemay not be the first to bring you the news, butyou will learn it in a form which allows foridentification and interaction with the people inthe news.’ Significantly, the core definition ofnews as the presentation of facts remains un-touched by this joint commercial, professional,and discursive strategy.

Three traditions of news production research 87

1111234567891012341115678920123456789301234567894012345678950111

� interested procedures – knowledge interests, Chapter16, p. 279

objectivity

� agenda-setting – Chapter 9, p. 145

productdevelopment

The analysis of news texts thus provides asupplementary approach to the study of newsin its social and cultural contexts. Like newsevents, societies and cultures become meaning-ful and interpretable when they are consideredin their textual or symbolic forms. As noted by Schudson (1991: 151), quoting MarshallSahlins (1985), ‘an event is not just a happen-ing in the world; it is a relation between acertain happening and a given symbolic sys-tem.’ In this perspective, what Gans (1979)uncovered in his analysis of the basic valuesheld by journalists in television network newsand major news magazines were the elementsof a symbolic system, for example, an idealiza-tion of small-town life, which resonates withmuch of American culture. From a receptionperspective, Gamson (1992) similarly identifiedspecific frames of understanding news, involv-ing elements of, for instance, justice and iden-tity, and referring to common symbolic dimen-sions of technology and nationalism thatAmericans associate with political issues.

Indeed, it may be argued that a concern not only with political life, but with culturebroadly, permeated U.S. news research from theoutset. When Park (1922) examined how theimmigrant press helped newly arrived Ameri-cans to assimilate, he necessarily addressedsome of the main symbols of cultural conso-nance and participation in public life: how todress, how to eat in a mannerly fashion, howto fill out a job application. Equally, whenHughes (1940) identified the criteria of a suit-able human interest story, she necessarily un-covered symbols of what it meant to be aprivate individual in the American context.Although both Park and Hughes merelyclaimed to be analyzing news, they simultane-ously provided a description of the culture inwhich the news was embedded and refracted.�Also in other cultural contexts, variations in thenews form, and in its social and cultural uses,have been subject to structural, organizational,as well as textual study (for an overview seeJensen 1998).

CONCLUSION

The parallel lives of political economy, socialphenomenology, and textual studies in newsresearch since the 1960s may have been due tothe shared ambition of these three traditions of staying out of, and moving against, thefunctionalist mainstream of a field which, at the time, was being institutionalized, not onlyin the U.S.A., but in much of the world.�According to the theoretical and politicalagendas of the period, integration might entailcooptation and neutralization. In the currentlyconsolidated as well as diversified field, anargument may be made for reconsidering thecomplementarity of the three traditions as wellas their interfaces with any remaining main-stream. This chapter has reviewed the trad-itions, and has exemplified some of their theoretical and empirical points of contact.

A common premise of the three traditions isthat news cannot be accounted for in terms ofeither liberal-leftist ‘bias’ or establishment ‘pro-paganda.’ Instead, it is necessary to examineempirically the several moments of its ‘produc-tion’ – its political-economic preconditions, its organizational enactment, and its textualarticulation. Whereas some research, like somepublic debate, still invokes metaphors of ‘selec-tivity filters’ to explain media ‘propaganda’(Herman and Chomsky 1988), the politicaleconomy tradition from the outset has arguedagainst ‘presenting the mass media as a simplerelay system for the direct tranmission of a rul-ing ideology to subordinate groups’ (Murdockand Golding 1977: 34). The same researchershave continued to formulate research questionsconcerning ‘the ways in which the representa-tions present in media products are related to the material realities of their production’(Golding and Murdock 1991: 22). Textual stud-ies of news, in their turn, have recognized thediverse institutional purposes of media dis-courses, as evidenced in their modes of addressas well as their concrete linguistic selections.�

Whichever configuration of the research traditions will be the outcome in future


� acculturation and socialization via media – Chapter9, p. 151

� media research institution – Chapter 16, p. 279

� language of news – Chapter 7

research, they must address classic issues ofpolitical philosophy and of the normative theories of media,� including:

• the relative autonomy of individual newsworkers, particular news organizations, and themedia institution as such within the socialsystem;

• the differences, similarities, and intercon-nections (via the same market) between com-mercial and public service media, as theyproduce and circulate meaning in society.

Figure 5.1 indicated the range of socialagents that exercise an influence on news pro-duction, whereas the different research trad-itions provide methodologies and theoreticalframeworks for studying the nature of theseagents’ interchange. One important issue forfurther studies is the relationship between ‘pull’and ‘push’ in the news market – between jour-nalists pulling together information from vari-ous sources, and sources pushing their agendasand frames of understanding into the media.

‘Push’ activities are increasingly important, asplanned media coverage has become integral to the conduct of both business and politics(e.g., Glenn et al. 1997; Manheim 1998; Shoe-maker and Reese 1991; Windahl et al. 1992).

Figure 5.2 summarizes key elements of thepush–pull nexus. Whereas ‘source organiza-tions’ – typically political bodies and corp-orations – have different aims and procedures from news organizations, they equally pro-cess ‘occurrences’ at the interface between theorganization and its social context, whichcertain strategically placed (source) persons canpresent to reporters as ‘events,’ hoping thatthese will be further processed within newsorganizations and disseminated to a widerpublic as news. At the juncture between sourcesand reporters, ‘source media’ (press releases,government reports, telephone interviews, etc.)provide important and still under-researchedraw material for what ends up as news texts.At the same juncture, it also remains importantto ask to what degree journalists in differentmedia and cultural contexts perceive their role,

Conclusion 89

1111234567891012341115678920123456789301234567894012345678950111 � normative theories – Chapter 16, p. 276

pull and pushin newsmarkets

Source organization

Sourcepersons

Sourceevent

News organization

Reportersand editors

Reportedevent

Sourcemedia

Occurrence News occurrence

Figure 5.2 Sources and news media as interconnectedorganizationsSource: Adapted from Ericson et al. 1987: 41; see also McQuail and Windahl1993: 178

journalisticself-conceptions

for example, as information dissemination,interpretation and investigation, or an adver-sarial relation to institutions of power (Patter-son 1998; Weaver and Wilhoit 1996; Wu et al.1996), despite the overdetermining influence of organizational practices and professionalprotocols.

The areas of agreement between the threetraditions of news research cannot now besummarized in one elegant model, such as thatof Riley and Riley (1965). Instead, it will be important to take another, meta-theoreticallook at the traditions, asking where and howthey may complement one another in further

empirical studies. The constitution of news, likethe constitution of society itself, is perhaps bestdescribed as a complex and continuous struc-turation (Giddens 1984), involving infrastruc-tural, organizational, as well as discursivecomponents. News is both a permanent socialstructure and a means of social reflexivity andcontestation; a product as well as a productiveprocess. For now, the realization that seeminglydisparate research traditions represent com-plementary perspectives on the same object ofanalysis will be a significant step forward instudies of news production.


FOUR APPROACHES TO NEWS RESEARCH

When providing an overview of the field of inter-national news research, it is useful to specify the level of analysis (i.e., micro or macro), andthe kind of perspective applied to the newsprocess (i.e., how the social phenomenon ofnews is conceptualized). Two perspectives on thenews process have dominated research on inter-national news: those of selection and construc-tion. Research using a selection perspectivetraditionally focuses on the news ‘event.’ Eventsin the world are the independent variablesdetermining the structure of foreign news. Here,social institutions (the press, the journalisticprofession, the news market, etc.) are said toplay a secondary, intermediate role as a ‘selector’or gatekeeper; they perform such functions asselection, rejection, re-editing, etc.

The construction perspective takes theopposite view: news is seen as a social artefact.Social conventions and practices, specific val-ues, and the allocation of material resourceswork together to produce a particular outcome:the news. Here, social institutions constitute the primary variable; the form, content, andvolume of foreign news are all products of thesocial practices of news institutions. In Figure

6.1, these two perspectives are paired with themicro–macro distinction, thus producing amatrix of four approaches to the analysis ofinternational news.

1 In the first approach – micro-level and selec-tion perspective – we find the oldest tradition:gatekeeper analysis. This began with theseminal study by White (1950), and was furtherdeveloped by McNelly (1959) to include thewhole chain of gatekeepers, from the initialreporting of the event in a foreign location, tothe editor of a local newspaper in another partof the world. Even though, during the past twodecades, severe criticism of the gatekeeperapproach has been raised, it has not completelylost its place in the field (Shoemaker 1991). Theconcept of gatekeeping intuitively correspondsto certain basic empirical observations at themicro level: foreign news is brought into thenewsroom, some of it is selected, rewritten, andso on.

2 In the second approach – macro-level andselection perspective – we find the news flowanalyses. These studies have so far dominatedresearch on international news; examples are, among many others, UNESCO (1953),

1111234567891012341115678920123456789301234567894012345678950111

The study of international newsStig Hjarvard

6

• a distinction between four types of research on news• a review of previous studies of international news flow• a critique of Galtung and Ruge’s (1965) classic study• a research agenda on news and globalization.

selection vs.construction

of news

gatekeepingresearch

Nnaemeka and Richstad (1980), Stevenson andShaw (1984), UNESCO (1985), and Varis andJokelin (1976). Although news flow analysisentails detailed counting of individual newsitems, it has been conducted at the macro level.The general idea has been that the more dataone collects and summarizes, the more generalpropositions one can make as to the structureof international news.

3 In the third approach – micro-level and con-struction perspective – we find empirical studiesof news content, often comparative case studiesfocusing, for instance, on the different U.S. TVnetworks’ coverage of events in the Third World(see, e.g., Adams 1982). We also find a fewdetailed studies of news agency organizations(see, e.g., Boyd-Barrett 1980; Boyd-Barrett andThussu, 1992; Fenby 1986; Höhne 1984), andstudies of foreign correspondents (see, e.g.,Batscha 1975). In both cases, the underlyingassumption is that social institutions heavilyinfluence the form, content, and volume offoreign news. However, the analysis is, in prac-tice, concerned with micro-level phenomena:the content of individual news stories or the par-ticulars of an individual news agency.

4 In the fourth approach – macro-level andconstruction perspective – we generally findmore theoretical contributions, which are con-cerned with international news in the contextof broader questions of media or cultural impe-rialism, or of the political economy of themedia (see, e.g., Galtung 1971; Mosco 1996;Mowlana 1997; Schiller 1976; Tunstall 1977;Tunstall and Machin 1999).

TV NEWS FLOW STUDIES

In this chapter, I will focus on studies of theinternational flow of television news.� What is

surprising here is that the bulk of research thathas been carried out within the news flow trad-ition fails to examine, in any real detail, theactual flow or exchange of TV news betweennews organizations (news agencies, broadcast-ers, etc.) and across countries. In fact, newsflow research does not take the ‘flow’ metaphorvery seriously. Most studies consist only ofcomparative analyses of the output of nationalbroadcast news. As a consequence, the actualprocess of news flow is left unexplained. Thismay be likened to trying to understand the eco-logical system of the sea by counting thenumber of fish one can catch from the shore.

A few studies have tried to analyze thecontent and volume of television news itemsthat are actually exchanged between actors atthe wholesale level, and between wholesale andretail levels. In the case of Europe, there is ahandful of studies (Golding and Elliot 1979;Gurevitch 1992; Hjarvard 1995, 1999; Melnik1981; Ruby 1965; Varis and Jokelin 1976). Inthe case of the so-called Third World conti-nents, there are only one or two studies percontinent (see, e.g., Turkistani (1988), whichcovers news flow in the Arab region). Takentogether, these few studies also reveal that few– if any – general conclusions may be drawnabout the character (volume, direction, content,etc.) of TV news flow at the global level. Thestudies seem to indicate that stable or uniformpatterns of news flow may instead be identifiedat a lower level, namely as a function of theregion, the historical period, and the specificactors involved (kind of wholesale agencies,kind of broadcasters, etc.). For instance, theidea that international TV news flow could becharacterized as ‘one-way traffic,’ as suggestedby Nordenstreng and Varis (1974) and by

Study of international news92

� three other traditions of news research – Chapter 5

wholesale andretail newsexchange

Micro-levelanalysis

Macro-levelanalysis

Selection perspective

Gatekeeper analysis

News flow analysis

Construction perspective

Ideology critique of foreign news contentAnalysis of news media organizations

Media imperialismPolitical economy of news media

Figure 6.1 Fourapproaches to the studyof international news

UNESCO (1989: 145), receives support in onlysome of these studies, and is actually contra-dicted by more recent findings.

There are at least two major problems withhow news flow data have been interpreted inearlier research. First, key concepts are fre-quently changed when analyses move from thestage of empirical analysis to that of interpre-tation and theoretical explanation – a shiftcommonly occurs from a selection to a con-struction perspective. News flow patterns areoften explained using concepts from theories ofmedia and cultural imperialism or from depen-dency theory.� But little work has been doneto mediate between the construction perspec-tive of these theoretical explanations and theselection perspective implied by the empiricalflow studies. There are very few attempts toexplain how macro-social structures actuallydetermine or influence news flow patterns. For instance, how does media imperialism workat the level of social actors (e.g., in TV newsagency newsrooms) – actors whose combinedactions at the micro level result in the observ-able news flow patterns? What is missing is a theoretical framework that could create acommon ground accommodating both thegeneral social structure and the social actorsinvolved in micro-social actions.

Second, data interpretation has been guidedby a theoretical and/or political position whereinternational news flow is considered to bestructured by a rather limited set of factors.Both the realm of foreign news and the socialworld have often been assumed to be muchsimpler and more homogeneous than is actuallythe case. In particular, the connection betweenthe operations of the news media and broadersocial forces (such as cultural imperialism, eco-nomic dependency, etc.) has been portrayed asa rather simple one. Studies of news flow pat-terns have seldom dealt with the more specific,and relatively autonomous, features of mediainstitutions: editorial practices, journalistic edu-cation, technology, conventions of expression,and so on. Foreign news is both a discursive-symbolic and a social-economic practice, andboth dimensions must be taken into account in

research. Thus, it is necessary to bridge the gap between social-scientific and humanistictraditions in order to provide a comprehensiveanalysis also of international news.

GALTUNG AND RUGE REVISITED

Some of the problems characterizing the studyof international news in the 1960s, 1970s andinto the 1980s may be illuminated through acritical discussion of the most important theo-retical contribution to the field during thisperiod: Johan Galtung and Mari HolmboeRuge’s article ‘The Structure of Foreign News’(1965). The article has been required readingat many schools of journalism and mediadepartments. Galtung and Ruge’s hypothesesand methodology also became paradigmatic for much of the research on international news.Two decades after its publication, the Finnishmedia scholar, Kaarle Nordenstreng, noted thatmuch of the research and critical advocacy ofa New World Information and CommunicationOrder (NWICO)� could be traced to Galtungand Ruge: ‘Most of the criticism, analyticalconceptualization, and guidelines for changevoiced over the past few years were articulatedin the 1960s by Johan Galtung and MariHolmboe Ruge’ (Nordenstreng 1984: 29).

The key propositions of Galtung and Rugeare stated as hypotheses involving twelvefactors (such as lack of ambiguity, reference toelite nations, etc.), which together are said todetermine the structure of foreign news. Theytest the importance of these factors through a quantitative content analysis� of four Nor-wegian newspapers with different political lean-ings. Concretely, the analysis deals with thecoverage of three international crises: Congo in1960, Cuba in 1960, and Cyprus in 1964. Theyfind partial confirmation of their hypothesesand of the relevance of the twelve factors, butalso state that there is much empirical testingto be done, and that ‘much remains to be donein terms of refinement of the hypothesis’(Galtung and Ruge 1965: 70). Later attemptsat verifying the various hypotheses have been

Galtung and Ruge revisited 93

1111234567891012341115678920123456789301234567894012345678950111 � cultural imperialism – Chapter 11, p. 177

� NWICO – Chapter 11, p. 178

� quantitative content analysis – Chapter 13, p. 219

mixed. A major review of the research litera-ture in English caused Hur to conclude that:‘the generality of the proposed hypothesesremains questionable because one hypothesishas been supported in some studies and rejectedin others’ (Hur 1984: 367). Large Germanresearch projects, Schulz (1976) and Wilke(1984), reached similarly mixed conclusions.

In spite of the uncertainty surrounding theempirical validity� of both hypotheses andfactors, they have continued to inform thetheoretical and methodological framework ofresearch on international news. The theoreticalstanding of the position was further strength-ened by the publication of Johan Galtung’sinfluential article, ‘A Structural Theory ofImperialism’ (1971), in which he incorporatedthe field of international news into what helabeled ‘communication and cultural imperial-ism.’ However, Galtung and Ruge’s theoreticaland methodological framework contains aseries of problems that make it untenable as astarting point for empirical analysis.

The first problem concerns their conceptionof the news production process. The proposednews ‘factors’ constitute an abstract synthesisof a whole series of different social structuresand processes which enter into the productionand circulation of foreign news. These includethe correspondent network, editorial structureand policy, the different, distinctive news media(TV, radio, newspapers, etc.), audience charac-teristics, and so on.

The research literature, furthermore, revealsdiverging interpretations of the proposed newsfactors. In some cases they are considered asthe imaginary list of news criteria according towhich individual gatekeepers make theirchoices (e.g., Ahern 1984). In other cases theyare considered as psychological determinants ofthe news process as a whole (e.g., Rosengren1974). Others again see the factors as an ex-pression of macro-social determinants (e.g.,Nnaemeka and Richstad 1980; Pollock andGuidette 1980). In other words, Galtung andRuge’s theoretical construction of the empiricalobject was inadequate, since it is not at all clearwhat kind of social process in the real world is

being uncovered by the analytical use of thenews factors. It may be processes at the micro,meso, or macro level, and it may be social phen-omena as different as economic, journalistic,political, ideological, or psychological aspectsof the news production process.

Moreover, the proposed methodology isflawed. A heuristic model – a ‘radio’ metaphor– informs the development of hypotheses andfactors. According to Galtung and Ruge, thenews factors are ‘nothing but common-senseperception psychology translated into radio-scanning and event-scanning activities’ (Galt-ung and Ruge 1965: 66). It follows from thisthat the factors should be understood as selec-tion criteria regarding the characteristics ofevents. However, the method employed is con-cerned neither with the selection process norwith the characteristics of the events. Instead,a quantitative analysis of the final news productis used to test the hypotheses.

Rosengren (1970, 1974) drew attention tothe fact that a valid test of the hypotheses must include a comparison of extra-media data(the universe of events from which the news isselected) and intra-media data (the selectednews items). Rosengren admits that, in practice,it may be difficult to demarcate such an extra-media dataset. However, the problems involvedin the construction of an extra-media data setshould not justify the application of a ‘faultyErsatz methodology’ (Rosengren 1974: 146). If the factors address the social and symbolictransformation of events into news, hypothesesabout them cannot be confirmed solely throughan analysis of the resulting news. For instance,it is impossible to know whether a predomi-nance of elite persons as actors in the news isdue to a bias in the selection and presenta-tion of news, or to the fact that elite personsactually play a greater role in the eventsreported.

A third problem is the notion that it wouldbe possible, once and for all, to isolate a defi-nite number of factors that determine the wayforeign news is structured in the (Western)world. Galtung’s and Ruge’s ambition was topropose some very general hypotheses aboutthe characteristics of (foreign) news in general,or as Johan Galtung himself expressed it: ‘It is


extra-mediaand intra-media data

� validity – Chapter 15, p. 266

a theory concerning what kinds of things arereferred to as news. “The theory of news”’(Galtung 1973). As such, their hypotheses aswell as the news factors are based on theassumption that, in news from and about theWestern world, there is a high degree of uni-formity across countries, types of media, polit-ical and cultural traditions, and so on.

Consequently, according to Galtung andRuge, the perspective for further research oninternational news was to test and refine suchfactors and hypotheses, which would allowscholars to predict with increased accuracy thelikely content and, hence, the fundamentalstructure of foreign news. Their ambition wasnot to look for specificities or differences. Thelogic justifying their own analysis of four differ-ent newspapers was that the more newspapersthe hypotheses turned out to be valid for, themore support the hypotheses would receive.

Their own empirical analysis can, in fact,validate a quite different – indeed contradictory– relationship, namely that the editorial policyof the specific newspapers in question mayaccount for a high degree of variation in thenews. A report from 1962 (published only in Norwegian) by Johan Galtung and MariHolmboe Ruge used the same empirical mater-ials as the 1965 article (not the 1964 Cypruscrisis, for obvious reasons). It presented a con-clusion which is rather surprising, compared tothat of the 1965 article: ‘What we have foundis simply that even if newspapers receive exactlythe same agency material via the teleprinter, thechanges made by the editorial staff will never-theless be enough to produce a completelydifferent type of news product’ (Galtung andRuge 1962: 74; my translation).

It is, of course, perfectly legitimate to usethe same empirical data to test different hypo-theses, but in this case the hypotheses contra-dict each other. If the hypotheses from the 1965 articles are unable to capture differenceswhich, according to the 1962 report, are promi-nent in the news, this is perhaps because the1965 hypotheses operate at a level of abstrac-tion where one is able, for example, to separatenews from non-news. However, such hypo-theses and their constituent factors are unsuitedfor guiding research that aims at understanding

the differences, variations, and historical devel-opments in international news.

GLOBALIZATION – A NEW AGENDA

The research paradigm established by Galtungand Ruge (1965) was influenced by the ratherstable structure of international society at thetime: the East–West division between commu-nist and capitalist states, and the North–Southdivision between the affluent and poor coun-tries. The critical debates about inequalities ofnews flows, uneven communication resources,and the dominance of Western actors – fol-lowed by the call for a New World Informationand Communication Order – were stronglyarticulated within the Cold War framework ofinternational relations.�

Since the mid-1980s, this structure has grad-ually been eroded, which has given rise to a newresearch agenda that cuts across previous dis-tinctions between micro and macro levels, andbetween selection versus construction perspec-tives. The collapse of communism in the SovietUnion and Eastern Europe, the demise of theapartheid regime in South Africa, and the GulfWar were important events which signaled boththe emergence of new power structures in inter-national politics, and the rise of news media to a new prominence. Increasingly, the newsmedia were important not only for their report-ing of these events, but seemed to play a role in the very course of events. It became clearthat news media were not only part of a dis-tinct political geography (typically within thenational political framework), but that theyalso play a role in creating and rearticulatingboundaries of the spaces in which social com-munities are organized (Morley and Robins1995).

The Gulf War has been the subject of a greatnumber of studies, which, taken as a whole,reflect a more diverse methodological palettethan earlier news research. Some studies of GulfWar coverage focused on the media as tools ofpolitical communication (Denton 1993); otherstudies dealt with the audience reception of war news coverage (Morrison 1992); and still

Globalization – a new agenda 95

1111234567891012341115678920123456789301234567894012345678950111 � the (end of the) Cold War – Chapter 11, p. 178

the Gulf War

others used different forms of textual analysisto uncover differences and similarities in thecoverage of the war in various media andnations (Nohrstedt and Ottosen 2000). Othermajor regional events, such as the Balkan wars, were also subjected to detailed analysis(Thompson 1999).

The new conditions in Europe were exam-ined in studies of changes in the institutionaland market structures of the wholesale newsbusiness. Boyd-Barrett and Rantanen (1999)analyzed, among other things, developments innew Eastern European states, and Hjarvard(1995) examined the impact of deregulation�

on the Eurovision News Exchange and com-mercial TV news agencies. The Europeanizationof news and journalism was examined in lightof the increasingly important role of theEuropean Union (Hjarvard 1993; Meinhof andRichardson 1999; Slaatta 1999). The emergenceof transnational TV news channels was dealtwith in several studies; in particular, CNNInternational was analyzed with regard to bothorganization and content (Flourney and Stewart1997; Volkmer 1999). The previously quitedominant tradition of news flow analysis wasalso renewed in the 1990s, in particular throughthe ‘Cooperative Study of Foreign News andInternational News in the 1990s’ (Sreberny andStevenson in preparation). The study of inter-national news in the 1990s was also influencedby new trends in reception analysis. Receptionstudies began – albeit slowly – to gain ground,and a few comparative studies of the receptionof foreign news were published (Biltereyst 2001;Cohen et al. 1996; Jensen 1998).

During the 1990s, the different contribu-tions to the field seemed to share several char-acteristics. Compared to research in the 1960s,1970s, and 1980s, there was now less of a focuson discerning grand macro-social relationships(such as the ‘one-way flow’ hypothesis). At thesame time, the simple counting of news items atthe micro level was considered insufficient andin need of supplementation or combination withother methods. Increasingly, research on inter-national news found its focus at a middle-range

level.� Working somewhere between the macroand micro levels seemed to increase the likeli-hood of making valid generalizations. Stablepatterns of interaction could be discerned at a regional level (e.g., Western Europe), within aspecific medium or sector (e.g., foreign news on television or the news agencies), or at a par-ticular stage in the communication sequence(e.g., the reception of foreign news). Research inthe 1990s also tried to combine methodologiesfrom the humanities and the social sciences.Ginneken’s (1998) introduction to the field, for one, clearly demonstrated the advantage of using a broader methodological palette (e.g.,organizational studies and discourse analysis) in order to understand global news.

Similarly, recent research has abandoned the dichotomy of construction vs. selection. Instead, social interaction is interpreted as adual relationship involving both action andstructure (Giddens 1984).� International newsmay be understood neither as the mere prod-uct of willful gatekeepers selecting from worldevents with a heavy hand, nor as the soleoutcome of overarching social structures, such as media and cultural imperialism. Social actors in the field of foreign news are know-ledgeable agents, whose actions at times maymake a specific difference. Simultaneously,however, they are informed by, and to someextent subordinated to, the social structuresthat govern the journalistic profession, theparticular media organization, the political andcultural framework of the country in question,and so on.

Research on international news in the 1990shas demonstrated that both actors and struc-tures are subject to considerable change. Aseries of new actors have entered the field,among them dedicated news channels withregional and global reach. Newspapers, newsbroadcasters, and Internet news services havebeen taken over by large media conglomerates,and this organizational and economic concen-tration has created new conditions for the pro-duction and distribution of foreign news.�


CNNInternational

� duality of structure – Chapter 1, p. 1

� conglomeration – in fiction production, Chapter 4, p. 73

dedicated,monogenericnews channels

� deregulation – Chapter 16, p. 278

� theories of the middle range – Chapter 3, p. 55

Another very important structural change is therise of the Internet as a news medium.�Internet news services may entail a further blur-ring of the distinction between wholesale andretail news providers, and may bring newactors with no journalistic background into thebusiness (Paterson 2001). Finally, the verynotions of ‘international news’ or ‘foreignnews’ are no longer self-evident in a globalizedworld with more open boundaries and overlapsbetween the local, regional, national, andglobal levels. For instance, in some countries inthe European Union, news about the EU is no

longer considered foreign news, but a kind ofquasi-national news.

The agenda of globalization constitutes achallenge for future research. An importanttask will be to analyze how globalization bothchanges the content and structure of inter-national news, and influences the political andother social processes that depend on thispublic communication. A parallel task will beto understand how globalization itself is fur-thered by changes in the field of internationalnews media.

Globalization – a new agenda 97

1111234567891012341115678920123456789301234567894012345678950111 � the Internet – Chapter 11, p. 188

TEXTS TO REMEMBER: INTRODUCTION

Every time a year comes to an end, we may usethe chronological threshold to think back onevents of the past year and their significance forourselves and for the many social formations towhich we happen to belong. Within the privatecircles of life, our personal memory is perhapssupported by letters and diaries, photographsor home video recordings. The further we moveinto the history of our local community orregion, via ethnic and national collectivities,toward events on a global scale that haveaffected us despite their geographical distance,the more we have to rely on mediated memo-ries, as recorded and disseminated by others, tounderstand what happened.� This chapterdeals with the ways in which the media select,combine, and present events in the real worldin verbal and visual form, thus constructingversions of reality which shape the meaningsand values that inform our attitudes and behav-iors, primarily as citizens and consumers.

At the end of 1998, the BBC broadcast itsannual Review of the Year program, offering

a selective, collective memory of a year whoseevents included the Bill Clinton–MonicaLewinsky case, the Football World Cup inFrance, British Northern Ireland Minister MoMowlam’s exploits at the negotiating table aswell as her trials as a cancer patient, SouthAfrican President Nelson Mandela’s eightiethbirthday, pop singer George Michael beingarrested for lewd conduct in a public lavatory,Titanic – the most expensive movie ever made(so far), and many other events from thedomestic and international scene. In just overan hour, this retrospective provided one partic-ular answer to the question, ‘What happenedin 1998?’, at the same time implicitly labelingnumerous other events as not worth remem-bering. In addition to omissions and absences,however, textual analysis should be curiousabout the ‘lenses,’ both literal and metaphori-cal, through which events are to be viewed.

This is how anchorwoman Sue Lawley intro-duces the program as she is walking around ahuge, dark television editing room with dozensof tiny screens behind her:

Welcome to Review of the Year 1998, theyear Britain went digital, or a tiny bit of it

Discourses of factKim Christian Schrøder7

• an overview of the development of quantitative content analysis• an example of content analysis in practice• a comparison of three forms of qualitative discourse analysis• a review of studies of visual media content• an example of the analysis of still images and texts in advertising.

� memory: immediate and mediated – Chapter 2, p. 16

did anyway. Very soon, we’re told, we’ll allbe able to watch hundreds, possibly thou-sands of channels, so this year we thoughtwe’d embrace the future as we review thepast. Tonight on Review of the Year welaunch a few channels ourselves. Their con-tent, of course: the events of the past yearstreamed and themed as they might be forthe digital age. The Sports Channel domi-nated by World Cup 1998, The MedicalChannel including a brief history of Viagra,The Women’s Channel featuring MoMowlam, Gerri Spice and more. But webegin with The Men’s Channel, not for men,you understand, but about them, their tri-umphs and their follies: some did well,others behaved badly. . . .

The visual environment, first of all, locatesviewers in a multi-channel, digital universe. Thescreens in the editing room function as a‘metonymic’ visual sign, being ‘part of’ andthereby signifying the ‘whole of’ the digitalage.� In the verbal part of the message, viewersmay respond to the jocular style of the anchor-woman; to the rhetorically balanced phrase,‘embrace the future/review the past’; to therhyming neologism ‘themed and streamed’; andto the definite article in the phrase ‘the eventsof the past year,’ which presupposes that theevents chosen by the program are the eventsworth remembering.

A recurring structural feature is Sue Lawleyaligning herself with viewers through pro-nouns.� The ‘inclusive we’ of ‘we’re told, we’ll all’ differs from the ‘exclusive we’ of ‘wethought’ and ‘we begin’, as she claims mem-bership of the production team. All in all, theanchor relies on the viewers’ complicity in pre-tending, for the sake of this program’s format,that Britain did ‘go digital’ in 1998. In fact, therest of its content is structured in accordancewith the narrative framework of an imaginedmulti-channel universe, including several chan-nels not mentioned in the opening, such as TheDisaster Channel, The Crime Channel, and

The Goodbye Channel, the last commemorat-ing people who died in 1998. Lawley’s refer-ence to some who ‘behaved badly’ prefiguresthe first item on The Men’s Channel:

1998 was the year in which one man foundhimself starring in his own real-life soapopera. Presenting Bill Clinton, the mostpowerful man in the world, and a woman-izer. Monica Lewinsky, the woman whoselips could destroy a president.

The musical, visual, and verbal sequence, com-prising a familiar fanfare, an aerial shot of amassive white building with a helicopter hov-ering in the air, and a split screen with threedifferent pictures of President Bill Clinton, is asustained pastiche of the internationally suc-cessful 1980s soap opera Dallas. Having intro-duced the main characters and the plot line, thesoap-style title sequence carries into an imita-tion of ‘Last week on Dallas’ that brings theviewer up to date on the Clinton–Lewinskyreal-life soap. Beyond this selection (paradig-matically) and combination (syntagmatically)of elements,� as with other media, televisionbrings to the configuration a repertoire ofgenres and styles that are themselves a mediacreation. This self-referentiality� is particu-larly evident here, because the generic lens for describing political events is taken from fiction, and may question not only conven-tional divisions between news and soaps, theserious and the trivial, but between fact andfiction, and between the public and privatespheres of social life.

Whereas this brief, informal analysis has notpursued political or moral implications, it is thepublic’s dependency on media representationsto keep track of events, and years, in order tomake politically or morally informed decisionswhich has often led media researchers to focusattention on the precise manner in which pic-tures of social reality are constructed for us by the media through topical structures andexpressive forms.

Texts to remember: introduction 99

1111234567891012341115678920123456789301234567894012345678950111

� metonymy – Chapter 2, p. 24

� pronouns – p. 106

� paradigms and syntagms – Chapter 2, p. 24

� self-referentiality – intertextuality, Chapter 11, p. 186

While complaints about ‘bias’ in the mediaare legion across the spectrum of public debate,academic research has usually been driven by a‘critical’ interest� in exposing the differentialnews treatment of social groups in conformitywith entrenched hierarchies of power and priv-ilege, for example, in the strategic area of indus-trial conflict (Glasgow Media Group 1976;Hall et al. 1976). Similarly, studies of advertis-ing have taken on stereotypical gender roles(e.g., Andrén et al. 1978; Goffman 1979[1976]; Millum 1975), and the promotion ofconsumption as a solution to complex social or personal problems (e.g., Leiss et al. 1986;Vestergaard and Schrøder 1985; Williamson1978). A third body of research has examinedthe realities implied in media fiction – particu-larly television series – from the 1960s (e.g.,DeFleur 1964), via U.S. commission reports ontelevision as a social issue (Comstock et al.1971), to cultivation research.� Over severaldecades, cultivation studies have identified a discrepancy between crime statistics and theamounts and kinds of violence depicted ontelevision, as well as the possible impact of TVviolence especially on heavy viewers and itspotential political support for conservative law-and-order policies (Gerbner and Gross 1976;for critiques see Hirsch 1980; Newcomb 1978).

The critical orientation is commonly foundon both sides of the quantitative–qualitativedivide, even though the readiness to accept aconception of discursive representations as a‘construction’ has varied in the development of textual and content studies. In a first phase,a number of studies tended to assume that themedia could be studied as a phenomenon sep-arate from the rest of society, asking how ‘faith-fully’ media represented social reality.� Thishas been the rationale of most quantitativecontent studies, but also of ‘critical linguistics,’reviewed below. A second phase has empha-sized the mutual constitution of media andmodern societies, so that it may not be feas-ible, methodologically or epistemologically, to

compare ‘media realities’ with any independentindicators. ‘Critical discourse analysis’ as wellas ‘discursive psychology’, but also certainapproaches to content measurement, increas-ingly take as their premise that media contentis not so much a secondary reflection as an arte-fact and a practice in which society is bothreproduced and contested.

QUANTITATIVE CONTENT ANALYSIS

Questions that audiences – and researchers –ask

Having watched a program like Review of theYear 1998, the audience, being instinctivelyanalytical, or semiotic, creatures, is left withcertain general impressions of what the pro-gram was ‘about,’ and how this ‘content’ islikely to affect both viewers like themselves andthe larger society represented. Predictably, therewill be critical comments about women beingdepicted as secondary in political and economicmatters, about a supposedly serious news re-view spending too much time on the trifles ofthe entertainment world, or about Third Worldevents being all but ignored in an increasinglyglobal society. In interpersonal everyday dis-cussions, people will next try to substantiatetheir claims by referring to specific programelements, but because lay discussion is rarelysustained beyond impressions or speculations,it is difficult to determine who is right and whois wrong. One task of media and communica-tion research, in the garb of ‘content analysis,’is to address such issues, supposedly on behalfof a democratic society as an input to, andinfluence on, political and public debates abouthow its citizens and their concerns shouldcollectively be portrayed by the media.

In order to interpret the implications of find-ings regarding both gender and other cat-egories, content analysts need a standard ofcomparison. If one subscribes to the first-phasefocus on media as separate from other socialstructuration, one may refer to ‘reality itself,’as documented typically with official statisticsregarding, for instance, the proportion ofwomen and men in parliaments or occupationalroles. If, instead, one relies on the second-phase

Discourses of fact100

� critical knowledge interest – Chapter 16, p. 281

� cultivation research – Chapter 9, p. 150

� selection vs. construction of news – Chapter 6, p. 79

Quantitative content analysis 101

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 7.1 CONTENT ANALYSIS OF REVIEW OF THE YEAR 1998,BBC1 1998

Table 7.1 displays a basic quantitative content analysis� of the 1998 Review of the Year. Amongseveral potential categories, the analysis singled out the human individual, specifically characteris-tics of the program’s personalities. It was decided to consider only the protagonists, defined asindividuals at the center of attention by virtue of their being both mentioned verbally by name(in spoken or written form) and appearing visually on screen. Thus, Rupert Murdoch, mentionedin the voice-over but not shown, and Leonardo DiCaprio, shown but not named, were excludedfrom the counts.This criterion is controversial, and led to the inclusion of perhaps less-prominentindividuals like Russian Premier Primakov who had been the premier for only a few months.Thepoint is that such criteria must be justified theoretically through the operationalization of ‘protagonist,’ and then applied consistently across all cases. A particular theoretical motivation for analyzing individual protagonists is that they are the vehicles of particular social and culturalcharacteristics which together delineate a particular possible world. Three types of characteris-tics were studied for present purposes:

1 geographical origin (perhaps suggesting a view of the world from a particular perspective);2 gender (considering claims about male dominance in the world of television);3 social arenas of activity (exploring the prominence of four different sectors with which

protagonists may commonly be associated: politics, showbiz (subdivided into sports andentertainment/media), science, and everyday life).

It should be noted that categorization and subsequent quantification (lasting approximately fourhours for this seventy-five-minute program, including calculation of percentages with a pocket

cont.

Table 7.1 Representation of protagonists on Review of the Year 1998

Men % Women % Total %

Geographical belongingBritish 39 56 15 65 54 58American 8 11 4 17 12 13European 8 11 2 9 10 11Rest of the world 14 20 1 4 15 16Unidentifiable 1 1 0 0 1 1

Total characters 70 100 23 100 93 100

Social arenaPolitics 38 54 7 30 45 48Showbiz 22 31 12 52 34 37

Sports 16 23 4 17 20 22Entertainment 6 9 8 35 14 15

Science 2 3 0 0 2 2Everyday life 8 11 4 17 12 13

Total characters 70 100 23 100 93 100

Gender 70 75 23 25 93 100

� content analysis, sampling, coding, and statistical procedures – Chapter 13, p. 219

premise regarding the mutual constitution ofsocieties and their symbolic forms, an internalstandard suggests itself, such as previous‘reviews of the year’ going back five, ten, ortwenty years, perhaps conferring, as well, withother genres.� Additional sources of compari-son arise from the distinction between public-service and commercial channels, and betweenprint media and this broadcast medium.

It remains to emphasize that the analysis inBox 7.1 provides an illustration of, and reflec-

tion on, basic principles and applications ofcontent analysis, not a full-fledged method-ology of representative sampling, dual coding,and hypothesis testing. A main advantage ofquantitative content studies is that they canserve to confirm or disconfirm intuitive impres-sions by performing a systematic description of a large set of media discourses throughnumbers that express the frequency and promi-nence of particular textual properties. A draw-back, or trade-off, is the inevitable reduction of complexity that follows from the decontex-tualization of meaningful elements. While qual-itative researchers have sometimes been bluntin their critique of what appeared to be a


calculator) is by no means an automatic process. Almost every instance requires the discrimin-ating ability of the analyst before assigning, for example, Princess Diana to the entertainmentcategory, Monica Lewinsky to politics, former astronaut and U.S. senator John Glenn as a con-tributor to science (rather than a politician or media celebrity), Boris Yeltsin (and Russia) to Europe.In one interesting case, the Israeli singer Dana International was categorized as female (before asex change,‘she’ was a man) and European (appearing because she won the European Song Contest).Any discussion about findings and their validity must begin by thus explicitly noting the criteria,procedures, and decisions of the analysis as conditions of intersubjectivity, disagreement, andresearch dialogue.

The findings regarding geographical origin, indicating three British out of every five protagonists,are hardly surprising in a national program, whereas the minor role of European protagonists mightraise eyebrows among EU observers at an insular British outlook. (A further breakdown wouldreveal that four of the ten Europeans are sports stars, four are Russian presidents or premiers,leaving German Chancellor Helmuth Kohl and Serbian President Slobodan Milosevic as the solerepresentatives of non-Russian Europe.) In comparison, the rest-of-the-world figures might seemhigh, but includes protagonists from Africa (three), Asia (six), the Middle East (three), and LatinAmerica (three). Moreover, a preliminary, qualitative assessment suggests that most of the eventsdealt with in this rag-bag category are economic or political crises and natural disasters.

Regarding social arenas, citizens might welcome the fact that almost half the protagonists belongto the realm of politics, and might worry about the fact that showbiz accounts for no less thanone-third (mainly sports personalities) of the program’s main characters, while science, site of bothrevolutionary discoveries and ethical concerns, reaches a low of 2 percent. Again, a qualitative look behind the numbers suggests that politics is seen repeatedly through a lens of ridicule, in thecase of the American presidency, parliamentary struggles in the Russian Duma, and the sexual pol-itics of male British Members of Parliament. In addition, the apparently astonishing fact that peoplefrom everyday life are almost as prominent as entertainment personalities is placed in perspec-tive when it turns out that no less than eleven of the twelve instances represent either victimsor perpetrators of sensational crime.

Finally, regarding gender, across the board three-quarters of the protagonists in this particularprogram are men, a quarter women. The variations, however, suggest that women are relativelymore prominent in the British than in the rest-of-the-world category, less so in politics and science,more so in showbiz, and even outnumbering men in entertainment.

� extra-media or intra-media standards of evaluation –Chapter 6, p. 94

scientistic attempt ‘to eliminate, as far aspossible, any human element or bias’ from theresearch process (Fiske 1990: 135), contentanalysts have also engaged in a critical redev-elopment of their premises and procedures.

Revising categories

So far, this chapter has used the cumbersometerm ‘quantitative content analysis,’ but the lit-erature normally refers simply to ‘contentanalysis,’ while the qualitative study of mediatexts is variously called ‘textual analysis’ or,more recently, ‘discourse analysis.’ In a sense,quantitative social sciences have monopolizedthe generic term, as suggested by classic titleslike Content Analysis in Communication Re-search (Berelson 1952), Content Analysis: AnIntroduction to its Methodology (Krippendorff1980), or Content Analysis for the SocialSciences and Humanities (Holsti 1969). In theearly decades of the twentieth century, quanti-tative newspaper analysis measured the columninches devoted by newspapers to particularsubject matters, seeking to reveal ‘the truthabout newspapers’ and possibly criticize jour-nalistic practice (Street (1909), quoted inKrippendorff 1980: 14). Other studies usedcontent-analytical procedures not just to learnabout media, but to monitor topics in presscontent in continuous surveys so as ‘to estab-lish a system of bookkeeping of the “socialweather” comparable in principle to the statis-tics of the U.S. Weather Bureau’ (Krippendorff1980: 14), and comparable, in certain respects,to agenda-setting research half a century later.�A third ancestor of modern content analysiswere the propaganda studies� of enemy media during World War II, when Americanresearchers monitored ‘domestic enemy broad-casts to understand and predict events withinNazi Germany and its allies and to estimate theeffects of military actions on war mood’(Krippendorff 1980: 16).

Around 1950, following developments insurvey and other analytical techniques since

the 1930s, content analysis was approachingmaturity. In his famous and henceforth author-itative definition, Bernard Berelson (1952) syn-thesized earlier methodological reflections onthe approach: ‘Content analysis is a researchtechnique for the objective, systematic, andquantitative description of the manifest contentof communication’ (p. 18). The fundamentaldistinction in the analytical process, accordingto Berelson, is between ‘knowledge’ and ‘inter-pretation’ of media content (‘inferences’ toBerelson). While he recognized that the endgoal of research is to make interpretations ofthe media, and of their relationship with eitherthe intentions of senders or the consequencesfor the attitudes and behaviors of recipients,Berelson stipulated that such interpretationmust not be mixed into the analytical processproper. The distinction appears from the state-ment that ‘knowledge of the content can legit-imately support inferences about non-contentevents’ (p. 18), and is clarified in a discussionof the difference between qualitative and quan-titative approaches:

In ‘qualitative’ analysis the interpretations(i.e. inferences about intent or effect) aremore often made as part of the analyticalprocess whereas in quantitative analysis theinterpretations are more often likely tofollow the analytic procedure. . . . The ten-dency of the qualitative analyst is to makehis interpretations as he goes through thematerial – whenever a piece of material cueshim in some way. The tendency of the‘quantitative’ analyst is to base his interpre-tation upon the total completed analysis inorder to see particular pieces of content inperspective.

(pp. 122–123)

The quantitative analyst, then, can hope toavoid ‘interpreting’ his data only if he concernshimself entirely with ‘manifest,’ or denotative,meanings and excludes latent, or connotative,�meanings, since manifest meanings are definedas those which everybody (both senders andrecipients of messages) will spontaneously agree

Quantitative content analysis 103

1111234567891012341115678920123456789301234567894012345678950111

� agenda-setting research – Chapter 9, p. 145

� propaganda studies – Chapter 9, p. 147 � denotation and connotation – Chapter 2, p. 25

on. Berelson does acknowledge that, in a strictsense, ‘manifest content’ is merely that end ofa continuum of meaning where ‘understandingis simple and direct’ (p. 20), or where meaningsare more likely to be shared than at the other,radically polysemic� extreme. In hermeneuticterminology,� Berelson still assumed that onecan, after all, examine understanding withouta pre-understanding or horizon of expectations.

What amounted to a suppression in researchpractice of the role of interpretation in anyhuman activity, has been at the heart of con-troversies in the theory of science since Berelson(see the early critique of Berelson by Kracauer(1953: 693). More recent definitions, also with-in content analysis, have involved a markeddeparture from Berelson’s model. Thirty yearson, Krippendorff (1980) advanced an alterna-tive and influential definition: ‘Content analy-sis is a research technique for making replicableand valid inferences from data to their context’(p. 21). In developing his definition, Krippen-dorff was explicitly critical of Berelson’s posi-tion, which he found unnecessarily restrictive.Krippendorff did agree that content analysismust be replicable, and therefore systematic,but he saw no reason why content analysismust be quantitative, and also dismissed theexclusion of latent meanings from the re-searcher’s legitimate horizon of interests.Although his book remained focused on themechanics and accreditation of content quan-tification, his conceptualization of interpreta-tion differed fundamentally from Berelson’s,despite relatively subtle differences of wording.Crucially, Krippendorff identified inferentialprocesses in all stages of the research process,‘from data to their context.’ In the establish-ment of analytical categories, the ascription oftextual units to these categories, as well as thecorrelation of findings with theoretical concep-tions of society and culture, the content analystis inevitably an interpreter. In Krippendorff’s(1980) words, the analyst is a reader of themeanings of a text, someone who is not merelyengaged in ‘“extracting” content from the data

as if it was objectively “contained” in them. . . . meanings are always relative to a commu-nicator’ (p. 22).

With a cautious interpretation, one mightsuggest that this definition of ‘content analysis’implies a constructivist reconceptualization ofquantitative measurement. As such, Krippen-dorff’s (1980) intervention represents one spec-ification of the sometimes rather abstractacknowledgment that quantitative and qualita-tive approaches, also to media texts, are com-plementary in that they produce differentanalytical versions of reality, as an input topublic debate about social reality and its medi-ation. In the qualitative literature on mediatexts, both the constructionist and criticalpoints of departure have been in evidence fromthe beginning.

QUALITATIVE DISCOURSE ANALYSIS

Critical linguistics

In the conclusion to a now classic qualitativeanalysis of a newspaper report about racialconflict in Southern Africa, Tony Trew, one ofthe founders of ‘critical linguistics,’ stated thatthe text ‘is so far from the truth that only apowerful grip on the press and information andthe diligence of the media in resolving the floodof anomalies which they report are adequate topreserve the pretence that the press is truthful’(Trew 1979: 106). The quotation might give theimpression that critical linguistics simplisticallychecks whether a verbal account of politicaland other events is ‘true’ or not. While this isnot the case, analyses in the ‘critical linguistic’tradition will often lapse into direct contradic-tion of the media’s picture of reality, fueled bya strong political and educational commitment.

Critical linguistics developed in the 1970s asan influential early school of discourse analy-sis. The approach was able to document a closerelationship between the linguistic details ofmedia texts and the production of ideologyand, by implication, to substantiate that mediaideology contributes to the reproduction of asocial order founded on inequality and oppres-sion. Its epistemology was a somewhat heavy-handed linguistic constructivism according to


� polysemy – Chapter 10, p. 167

� hermeneutics – Chapter 2, p. 21

which words construct a mental grid. Wordsare conventional, but constitutive of the realitythey designate, so that, for instance, news-papers inevitably construct the social states andevents they describe.

Critical linguistics went on to claim that notjust lexical, but also syntactic choices havesemantic and hence ideological implications, asexemplified below. Relying on Saussure’s notionof the arbitrary sign,� the point is not that lin-guistic choices can be made by journalists atwill, but instead that these choices articulateand enact social power in practice. Since theformation of public opinion in capitalist soci-eties is controlled by those with economic andorganizational power over the mass media, who will see to it that representations do notjeopardize their interests and privileges, publiccommunication has a pervasive ideologicalthrust that renders current social arrangementsnatural, even inevitable, while discreditingalternatives as contrary to common sense. Assummed up by Fowler (1985), the task of crit-ical linguists is to expose such ideology in thebest interest of the majority: ‘Of course frommy point of view this has to be regarded as anundesirable situation and one that a critical lin-guist ought to expose’ (p. 67).

Dissolving seemingly innocent linguistic fea-tures, critical textual analyses can debunk‘warped versions of reality’ (p. 68). Building on Halliday’s functional grammar and socialsemiotic (Halliday 1973, 1978), numerouspublications from the late 1970s and into the1980s (e.g., Fowler et al. 1979; Hodge andKress 1988) identified linguistic features that‘will probably repay close examination,’according to Fowler (1985: 68). He offered thefollowing ‘checklist’ (with examples in paren-theses referring to the opening extract ofReview of the Year 1998):

1 Lexical processes arise from the particularvocabulary, including metaphors, which a mediatext mobilizes concerning a given area ofexperience, such as scientific vocabulary in a cosmetics ad, or management jargon in politi-cal coverage (the ‘launch’ of digital channels

in the BBC program refers simultaneously toactual, future channels and to the program’sown format).

2 Transitivity, a relation obtaining betweenverbs and nouns, adds up to a construction andconfiguration of the participants and processesin a text (the review not only ‘launches’ newchannels, but it ‘embraces’ the future, andanticipates what, very soon, ‘we’ll all be ableto watch’).

3 Syntactic transformations, particularly thoselabeled ‘passivization’ and ‘nominalization,’can be considered ideologically problematicbecause they may obscure agency – who didwhat to whom – or change the participants’relative prominence (‘we’re told’ that a digitalcornucopia will soon be available, but not whotells us, or with what justification).

4 Modality, in the form especially of modalverbs and adverbs, serves to express speakers’and other senders’ assessments of, and attitudestowards, the events rendered by the sentencesconstituting the text (the reference to ‘possibly’thousands of channels is partly a qualificationof the prediction itself, but also partly anexpression of skepticism toward everything‘we’re told’).

5 Speech acts� and turn-taking refer to thefact that speech is a coordinated form of action,in which speakers take turns in complex ways.One insight of discourse analysis generally isthat there is no one-to-one relation betweensentence types and communicative functions.Instead, analysts should ask which speech act isaccomplished by the utterance of a sentence,and how both the single act and the structure ofturn-taking may build positions of power intothe communicative context (e.g., a classroom ora medical consultation). (While the news reviewis not dialogic in the sense of talk or discussionprograms, already the opening ‘Welcome,’ itselfa conventional act of greeting, invites the viewerto attend to the ‘turn’ of this program).

6 Implicature, comprising a variety of com-municative features, is perhaps best explained

Qualitative discourse analysis 105

1111234567891012341115678920123456789301234567894012345678950111 � arbitrariness of signs – Chapter 2, p. 25 � speech acts – Chapter 2, p. 38

as the ability of most language users to ‘readbetween the lines,’ to infer what is ‘really’meant. It is this ability which is made system-atic and procedural in (critical) linguistics,particularly in attempts to recover ‘latent’meanings through a ‘hermeneutics of suspi-cion’� (the reference to The Men’s Channel asbeing not for, but about, men implies the exist-ence of monogeneric pornographic channels,and relies on viewers’ knowledge of this to pokegentle fun at men’s ‘follies’).

7 Address and personal reference, finally,allow linguistic analysis to consider howstylistic choices anticipate an audience with aparticular social or educational status; hownaming conventions affect the degree of for-mality; and how personal pronouns assert aparticular ‘interpersonal’ relationship with anaudience (in addition to the inclusive and exclusive we, ‘you understand’ is an explicitappeal to both female and male viewers toaccept the terminological joke about The Men’sChannel).

Although this checklist has found its way into later discourse studies of media, it is notalways evident that its elements represent pro-ductive analytical categories. Transitivity, forexample, requires a very time-consumingscrutiny of textual detail to establish an ideo-logical pattern (e.g., Fowler 1985), which mightbe predictable from a more cursory examina-tion. Furthermore, it is doubtful whether claimsabout the ‘mystifying’ effects of syntactic trans-formations are warranted, being based onassumptions about the ‘non-recoverability’ ofthe deleted linguistic component (Trew 1979).It seems equally probable that average media-users will have no difficulty inferring who didwhat to whom, despite passive sentences ornominalizations, on the basis of their back-ground knowledge of the world, their familiar-ity with a current news agenda, and theirgeneral communicative competence. Interest-ingly, an empirical study of readers’ receptionof the newspaper articles previously analyzedby Trew found that the readers’ views of the

events reported in the articles were determinedless by their ‘syntactically constructed’ ideologythan by the readers’ social identities and life his-tories (Sigman and Fry 1985).

In sum, critical linguistics made an import-ant departure, both beyond content analyses of manifest meaning, and beyond formal and poetic traditions of textual analysis, butremained closely focused on linguistic indica-tors. Other forms of discourse analysis havegone some way toward placing such indicatorsin their communicative and social contexts.

Critical discourse analysis

Although intellectually rooted in critical lin-guistics, critical discourse analysis (CDA), asdeveloped mainly by Norman Fairclough sincethe late 1980s, represents a significant theoret-ical as well as methodological contribution tothe interdisciplinary study of media discourse(Fairclough 1992, 1995). Its theoretical frame-work, relating textual features, first, to the con-crete social situations in which texts areproduced and consumed, and second, to socialprocesses at large, is usually displayed as amodel of three embedded boxes (Figure 7.1).‘Texts’ stand at the core of the model, and areexplored largely through the categories of crit-ical linguistics. The second dimension of analy-sis concerns ‘discourse practices,’ for instance,the processes through which specific mediatexts are produced in media organizations andconsumed, or ‘decoded,’ by audiences in thecontext of their everyday lives. These discoursepractices are understood as mediators betweencertain delimited texts and much wider ‘socio-cultural practices,’ which constitute the thirddimension of analysis. At this third, macro-social level, the discursive phenomena broughtto light at the first two levels are adduced inclaims about, and interpretations of, the pre-vailing ‘order of discourse’ at a given historicaltime. Thus, CDA has sought to join a linguis-tic, analytical approach to discourses, in theplural, with a critical, theoretical conception of one dominant discourse, in the singular,following Foucault (1972).

In addition to its sources in critical linguis-tics, and in critical social theory, the model of


� hermeneutics of suspicion – Chapter 2, p. 22

twodefinitions ofdiscourse – asconcrete usesof languageand othersigns, and as adominantworldview

CDA is inscribed in current debates about‘structure’ and ‘agency’ (e.g., Beck 1994; Gid-dens 1984, 1994), as suggested by its con-structivist position, specifically the dual role ofmedia as constituted by, as well as constituting,the social formation. The media clearly operatewithin a set of economic and political frame-works, representing the prevailing social order,but its diversity as well as its internal repre-sentational struggles, as articulated by themedia, serve also to poise the media uneasilybetween convention and innovation, reproduc-tion and change. In order to account for suchsocial and discursive processes, ‘intertextuality’has been imported as a key analytical conceptinto CDA.� Concretely, any text is indebted to innumerable source texts, and may itselfbecome the source of an infinite number offuture texts. In the case of news and otherfactual coverage, CDA analyses will examinevarious media genres (news reports, interviews,commentaries, etc.) as ‘discourse types,’ which,in their turn, draw on different political dis-course types (government reports, parliamen-

tary speeches, press conferences, etc.) that growout of related, but different discourse practices.

The intertextual perspective also attains relevance at the level of the ‘orders of dis-course,’ where conventional political discoursemay be intermingled with the discourses ofscience and technology, of grass-roots politics,or of the everyday. It is at this macro-sociallevel that, ultimately, the ‘meaning’ of dis-courses and practices may be evaluated. Oneexample is what Fairclough terms the ‘conver-sationalization’� of public discourse in themedia: the spread of informal speech and colloquial expressions. On the one hand, thismight trivialize complex social issues and relations; on the other hand, by making keyproblems accessible it might be a source of cultural democratization: ‘The communica-tive style of broadcasting lies at the intersectionof . . . democratizing, legitimizing and market-ing pressures, and its ambivalence follows from that’ (Fairclough 1995: 149). Similarly,Fairclough has referred to a colonization of


1111234567891012341115678920123456789301234567894012345678950111

TEXT

DISCURSIVE PRACTICE(production, distribution, consumption)

SOCIAL PRACTICE

Figure 7.1 Dimensions of critical discourse analysisSource: Fairclough 1995: 59

� intertextuality – Chapter 11, p. 186� conversationalization and tabloidization – Chapter 3,p. 49

public speech by commercial discourses as‘marketization,’ while he has been less aware ofthe complementary process of the ‘politiciza-tion’ of market discourses, as ecological andethical considerations gain attention in the business world.

The main limitation of critical discourseanalysis is that no empirical attention is given tothe middle range of the discourse practices.Fairclough has deliberately excluded this aspectfrom his own analyses, stating that ‘my empha-sis will be upon linguistic analysis of texts . . . I am not concerned . . . with direct analysis of production or consumption of texts’ (Fair-clough 1995: 62). Certainly so far, few re-searchers overall have undertaken holistic,empirical studies of media and their circulationof discourses between senders and recipients,some exceptions being Swales and Rogers’(1995) study of corporate mission statementsand Deacon et al. (1999) on the news circuit. As a research program, however, Fairclough’srejection of empirically studying audiences and other social agents of discourse tends tolimit the applicability and explanatory value of the approach.

Conversation analysis and discursivepsychology

In his important overview and redevelopmentof discourse analysis, Potter (1996) discusses anextract from a news report about a bombattack in Cape Town in which ‘the attack was believed to be the work of the UPLAR’(National Public Radio, December 12, 1993;emphasis added). Potter uses the sentence toillustrate a main difference between criticaldiscourse analysis and his own ‘discursivepsychology.’ Critical discourse analysis (andlinguistics) would take issue with the ‘agentdeletion’ of the sentence, the missing ‘believer,’which might covertly support the powers thatbe against a political liberation movementwithout adequate evidence. Without denyingthe potential ideological implication, Potteradopts an interaction perspective, asking whatpragmatic function the agent deletion mayfulfill for a journalist working under certaininstitutional pressures and professional norms,

which make it inappropriate for himself to bethe origin of any such belief. Further, Potterhypothesizes that other available sources, suchas the police, were not perceived as credible incontext. Potter’s (1996) conclusion is that theagent-less construction may serve as ‘a specificdesign solution to a range of fact construc-tion and warranting problems’ (p. 157) whichface a news report intent on establishing itsobjectivity and accountability. At the end of the day the journalist needed sources, did nothave any credible ones, and resorted to makingthem discursively present and absent at thesame time.

Among the heterogeneous ancestry of dis-cursive psychology, the most formative influ-ence has come from the tradition of ethno-methodology� and conversation analysis,particularly when it comes to its analyticalprocedures. For four decades, the aim of con-versation analysis has been to explore thesituational micro-mechanics of verbal inter-action. Many studies have probed, for example,speakers’ turn-taking through so-called adja-cency pairs, the role of silences and interrup-tions in the flow of interaction, the managementof topic development and topic change, andseveral other structuring features (for overviewssee Have 1999; Nofsinger 1991).

Like conversation analysis, discursive psy-chology takes its point of departure in the sit-uational contexts of language use. Its purpose,however, is to go on to account for the ways in which the communicative micro-mechanicsenter into everyday as well as institutionalprocesses of social life, for instance, in the caseof nationalism or racism (Billig 1995; Wetherelland Potter 1992). In thoroughly ‘discursifying’categories like attitudes and memory, discursivepsychology thus departs both from critical dis-course analysis, with its distinctions betweendiscursive and other social processes, and frommuch traditional psychology and social psy-chology (Edwards and Potter 1992; Potter andWetherell 1987).

Discursive psychology is particularly con-cerned with speakers’ fact construction andtheir own positioning in co-constructing social


turn-taking

� ethnomethodology – Chapter 3, p. 56

reality; that is, their attempts to establish theirown ‘versions’ or accounts of social events astrue and factual, and to undermine the factual-ity and truth of their interlocutors’ versions.Familiar examples are television news inter-views and debate programs, but the strategiesmay be found across many genres of talk andtext. When staking a claim to their ownversion, communicators will draw on ‘interpret-ive repertoires,’� which amount to frameworksof understanding that are reminiscent ofFairclough’s orders of discourse, even if thenotion of repertoires conceives discourse as apractical resource, rather than an end-product:

By interpretative repertoires we meanbroadly discernible clusters of terms, des-criptions and figures of speech often assem-bled around metaphors or vivid images. Inmore structuralist language we can talk ofthese things as systems of signification and asthe building blocks used for manufacturingversions of actions, self and social structuresin talk.

(Potter and Wetherell 1996: 89)

In an interview-based study of discourse and racism in New Zealand, the authors foundthat white New Zealanders would draw on two different interpretive repertoires regarding the position of Maori culture in contemporarysociety, which they called culture-as-heritageand culture-as-therapy. The former repertoireframes culture, almost in biological terms, as an endangered species, something to be pre-served and treasured; the latter conceivesculture in psychological terms as a need, par-ticularly for young Maoris, who ‘need to redis-cover their cultural “roots” to become “whole”again’ (p. 90). An important analytical point isthat the repertoires are not mutually exclusive,but may coexist also in a given individual’s dis-course about race, serving different rhetoricalpurposes in different situational circumstances.Most importantly, interpretive repertoires areneither personal nor socially specific ‘attitudes’as in much social psychology and surveyresearch, but the concrete linguistic manifesta-

tion and exercise of social practices of discrim-ination, and thus of social power.

Although the mass media have not beencentral to discursive psychology – a recentintroduction does not even mention its applic-ability to media studies (Potter and Wetherell1996) – it seems clear that the approach hasmuch to offer, theoretically as well as analytic-ally, especially as the electronic media havebecome increasingly dominated by formats thatborrow from or replicate everyday verbal inter-action (as also suggested by Fairclough’s ‘con-versationalization’), and as verbal exchangethrough computer media serves to establishnew forms of ‘virtual’ community. Researchersin the tradition suggest as much in their choiceof empirical materials. For example, Potter’s(1996) analyses of the situational constructionof facticity are full of single examples frommedia discourses, as he demonstrates howspeakers in news programs take great painsverbally to demonstrate that they do not ‘havean axe to grind’; to confess voluntarily tohaving a stake in some matter in order to createan impression of honesty and trustworthiness;to claim that ‘the facts show’ something to bethe case; or to bolster their own credibility by adducing testimony from sources with anuncertain identity (e.g., quoting ‘communitysources’ for information about gang warfare).

So far, however, perhaps the most system-atic and directly applicable work on mediadiscourses has come from scholars who seethemselves as conversation analysts, rather thandiscursive psychologists (or critical discourseanalysts) (e.g., Drew and Heritage 1992;Greatbatch 1998; Heritage and Greatbatch1991; see also Scannell 1991). One growingbody of work has analyzed the inherently inter-active genre of the news interview. In suchstudies, attention is focused not on the possibleideological meanings of a sequence of utter-ances, but on the social as well as the discursivedynamics of the exchange – on communicationas a form of action that is embedded in severalcontexts at once. These contexts of interactioncan be specified, in part, through comparisonof turn-taking patterns within media genreswith those of ordinary conversation, but also by comparing the conditions of discourse


1111234567891012341115678920123456789301234567894012345678950111 � interpretive repertoires – Chapter 10, p. 167

production in, for example, public service andother professional news organizations. Notingthat news interviews deviate systematicallyfrom ordinary conversation in replacing thelatter’s question–answer–receipt pattern with aquestion–answer–question sequence, Heritage(1985) explains this difference with reference to‘the overhearing audience’ of the interview. Byavoiding the evaluation inherent in the ‘receipt’turn of normal conversational sequences, theinterviewer declines the role of being the (only)recipient of the answer, while maintaininghis/her role as the elicitor of ever more answers,in both cases on the part of the overhearingaudience.

An additional consequence of ‘the overhear-ing audience’ for the interactional dynamics ofthe news interview are the frequent so-called‘formulating’ questions (e.g., ‘So you’re sug-gesting that . . . ’), by which an interviewer maymake explicit the potentially controversialimplications of an answer while appearing tomerely rephrase what the interviewee had justsaid. Within the institutional framework of pro-fessional journalism, then, discursive formula-tions can serve to both signal and reaffirm theimpartiality and balance of news in controver-sial matters. By contrast, the overhearing audi-ence is offered an entirely different discursiveposition in the case of a U.S. television phone-in program, analyzed by Crow (1986), in whicha sexologist host gives advice about sexualproblems, a genre which toes the line betweenprivate confession and public talk, explicitlydesigned for an overhearing (and overseeing)audience. Montgomery (1986) throws light onbroadcast monologues as he demonstrates howradio DJ talk, in contrast to the monologue of

radio news to abstracted third-person citizens,operates along the more intimate axis of first-and second-person pronouns. The DJ constructsa particular kind of ‘imagined community’(Anderson 1991) with his listeners in a simu-lated dialogue, in which he does not display anysign of awkwardness when his directive speechacts, such as greetings and questions, are notresponded to by anybody.

Figure 7.2 summarizes the relative emphasesof the three main qualitative approaches tomedia texts. (The parentheses indicate that theelement is recognized in principle, but notdeveloped in analytical practice.) A challengefor future research is the integration and rede-velopment of these several dimensions, boththeoretically and analytically, in interdisciplin-ary studies of the media–society nexus. In thatprocess, theories and analyses of visual mediaremain a special challenge.

VISUAL MEDIA DISCOURSES

In all the above approaches, including quanti-tative content analysis, the visual aspect ofnews and other factual genres in newspapers,magazines, television, and computer media is atbest given secondary attention, in spite of theconventional wisdom of research that the mediaenvironment is increasingly dominated by stilland moving pictures. While one early exceptionin the case of individual TV news stories wasFiske and Hartley (1978: ch. 6), it is next toimpossible to locate full-scale visual analyses ofentire news bulletins. Content analysis may berelatively ill suited for the analysis of visualcommunication beyond the identification of,for instance, types of protagonists or formal


theoverhearing

audience

Socioculturalpractices

Discourse practices

Textual analysis

Critical linguistics

(+)

–

++

Critical discourseanalysis

+

(+)

++

Discursivepsychology

(+)

++

+

Figure 7.2 Comparison of different approaches to qualitative discourse analysis

features, given its categorical conception ofmeaning, as mentioned, vis-à-vis the frequentlycontext-dependent and continuous features ofvisuals.

Qualitative discourse analysis has greaterpromise in this regard, and has been givenseveral different formulations, usually withrecourse to one of two theoretical sources:Roland Barthes and Charles Sanders Peirce (fora detailed analysis of news photographs, seeFiske 1990: 104ff.; Hall 1973a; Deacon et al.(1999: chs 9–10) carry the analysis from stillphotographs to moving images; see alsoHansen et al. 1998: chs 6–8).

Barthes (1964b) suggested that the denota-tive and connotative ‘levels of meaning’ applyalso to the analysis of visual signs. In a photo-graph, for instance, we may distinguish be-tween the denotative level, which carries theinnocent, factual meanings available to anyobserver irrespective of cultural background,and the connotative level, which carries thevisual meanings that a specific culture assignsto the denotative message.

In his 1964 article, moreover, Barthes theo-rizes the possible relationships between differ-ent sign systems within the same message,specifically ‘the linking of text and image,’asking ‘what are the functions of the linguisticmessage with regard to the (twofold) iconicmessage?’ (Barthes 1964: 38). Barthes discussestwo such functions: anchorage and relay, whichmay coexist in a message and which share thefunction of ‘fix[ing] the floating chain of signi-fieds’ of the inherently polysemic image (p. 39).

Anchorage, which may appear in the formof a title, a caption, or an accompanying pressarticle, may be subdivided into identifyinganchorage, where ‘the text helps to identifypurely and simply the elements of the scene’ asa ‘denoted description of the image’ (p. 39), andinterpreting anchorage, which guides the inter-pretation of the connoted meaning away fromtoo personal associations or ideologically unde-sirable meanings. In a formulation that is rem-iniscent of contemporary critical linguistics,Barthes claimed that the anchoring text has ‘arepressive value’ which is able to ‘remote-control [the reader] towards a meaning chosenin advance’ (p. 40).

Relay, which may appear via comic strip bal-loons or film dialogue, is established throughsnatches of text that may be perceived as anutterance spoken by a character in the image.Here, text and image stand in a ‘complementaryrelationship’ to each other, the text ‘setting out . . . meanings that are not to be found in the image itself,’ and the joint meaning of image and text being ‘realized at a higher level,that of the story, the anecdote, the diegesis’ (p. 41).

While analytically useful, this conceptual-ization of the text–image relationship cannevertheless be criticized, echoing a famouschiasmus from the history of communicationresearch, for being mainly interested in whatthe text does to the picture, to the relativeneglect of what the picture does to the text –an issue taken up in the analysis in Box 7.2.

In the case of Peircean semiotics, it is par-ticularly one set of analytical concepts that hasbeen selectively applied to arrive at a typologyof three possible relations between (visual) signsand objects in reality (Peirce 1985). First, a‘symbol’ is a sign whose relation with its objectis a matter of convention, with verbal language,such as the words ‘oil company’ in the BPadvertisement examined in Box 7.2, as the pro-totypical example. Second, an ‘icon’ is relatedto its referent through similarity; a photographentering into a news article or an advertisement(e.g., of the speakers and listeners in the BP ad)is an iconic sign resembling certain real-worldphenomena. Third, an ‘index’ signifies its objectthrough some existential or physical connectionwith it. While classical examples include smokeas an index of fire, in media and other culturalforms an index may operate at a more inferen-tial level, so that the people appearing in theBP ad may be understood as indices of the agegroup, gender, or ethnicity, or the personnelprofile of the oil company, which they appearto represent. In this specific respect, indexicalityis comparable to metonymy in the Saussureantradition.�

The typology is immediately complicated bythe fact that, for example, an (iconic) photo-graph is also an index of a segment of reality

Visual media discourses 111

1111234567891012341115678920123456789301234567894012345678950111

anchorage andrelay

symbol, icon,index

� metonymy – Chapter 2, p. 24


ANALYSIS BOX 7.2 DISCOURSE ANALYSIS OF ADVERTISEMENT FOR BP (THE GREEN MAGAZINE, DECEMBER 1991, INSIDE FRONT COVER)

cont.Plate 7.1 BP advertisement


1111234567891012341115678920123456789301234567894012345678950111

The BP advertisement, ‘Oil companies tend to invite criticism. At BP, we actively encourage it,’appeared in The Green Magazine, a publication whose readership has a greater-than-average inter-est in the environment.

In the reading situation, as the readers turn the front page of the magazine, they encounter apage which presents itself at first glance as a polysemic visual and verbal puzzle.The graphic layoutof the page with its distribution of iconic and symbolic signs, and the BP logo in the bottom right-hand corner, identifies it as an advertisement. The headline functions as a caption of the picture,and as an attention-catching device – many readers read no further! It provides no ‘identifyinganchorage’ for the composite picture, merely an aid to interpretation (‘interpreting anchorage’),assisting the readers’ inferential processes about who the people in the picture are and what theyare doing.

As the caption headline associates ‘oil companies’ with ‘BP’ and a pronominal ‘we’ who ‘encour-age criticism,’ the inferential process of interpretation may perceive the iconic representation ofthe eleven individuals around the periphery similarly as an indexical representation of BP employ-ees. They also indexically represent several social backgrounds: men and women, age groupsbetween the twenties and fifties, and different ethnic groups (one black man, one Asian woman,the remainder white).Their common gesture of placing an open palm behind an ear is an iconic-symbolic sign conventionally used to indicate a listening attitude.The headline may therefore alsobe seen as a collective utterance by the eleven employees on behalf of the company, i.e., text as‘relay.’

The headline’s choice of the nominalization ‘criticism’ begs the question of the agency behindthe criticizing activity, which may, however, be inferred, from the situational context now estab-lished, to be environmental activists, thus filling in the identity of the central picture element.Thespeaker depicted in the middle wears a suit but no tie, which may connote that he belongs neitherto the establishment nor to a counterculture – he is a decent person, someone to be taken seri-ously, his visible teeth perhaps an index of a certain insistence. A first reading thus suggests thatnot only are the diverse employees of BP responsive to criticism, so that there might be no needto shout, but the company is responsive precisely because it is made up of a cross-section of thesociety it serves.

In addition to the paradigmatic choices mentioned above, the syntagmatic ensemble of the adalso springs from the choice to use black and white only, as opposed to color. This is just oneof a number of ‘absences’ in the semiotic composition of this ad as opposed to consumer adver-tising, which distance this ad from any glamorous impression or visual ‘hype,’ instead suggestingsoberness.

The language of the copy is straightforward, fairly colloquial, with no difficult or specialistterminology. It uses declarative sentence structures only, apart from a few cases of independentsentence fragments, one example being the adverb ‘loudly,’ punctuating the rhetorical authority ofthe critics and establishing an intratextual link of recognition to the pictorial representation of themegaphone.The absence of imperatives and interrogatives, compared with consumer ads, comes torepresent the absence of any direct attempt to actively involve the reader (Myers 1994: ch. 4). Againthis adds to the impression of soberness, of the company simply putting its case forward, leaving itup to the reader to pass the verdict of relevance or irrelevance, acceptance or rejection.

This feature of the sentence structure is supplemented by a complete absence of the second-person pronoun ‘you.’ The ad thus, without any sense of intrusion, merely invites the reader towitness an account of a relationship between the company (‘we’) and different representatives ofthird parties, i.e., community and environmental groups (‘they’).The ‘we’ used is an ‘exclusive we’(because it excludes the readers) through most of the text; however, it slides imperceptibly intoan ‘inclusive we’ in the last sentence of the copy, which is repeated in the slogan ‘For all our

cont.

imprinting itself on film. The complicationarises, in part, from Peirce’s epistemology,which emphasized that signs are not entities,but relations or functions mediating knowledgeof objects in reality. One interpretation is tounderstand the three sign relations as proper-ties of any sign, whose combination helps toaccount for the cognitive status and experien-tial qualities of a given sign (Johansen 1993:95). To exemplify, a news photo of the WhiteHouse can be, at one and the same time, aniconic representation of a particular buildinglocated in Washington, D.C.; an indexical rep-resentation of the government of the U.S.A.,housing and standing for its chief executiveofficer; and a symbolic representation, depend-ing on the context, of values conventionallyassociated with the U.S.A. or its president, fromglobal capitalism to liberal democracy.

The semiotic characterization of visual signs

may be taken one step further to suggest whyvisual communication is intuitively ascribed ahigh measure of naturalism, potentially withimportant ideological implications. On the onehand, the visual sign appears to provide somenatural, iconic connection with reality, as rein-forced by common knowledge of how photo-graphy, film, and video is produced. On theother hand, it follows from the indexical, infer-ential aspect of the visual sign that the issue ofwhat, exactly, is being ‘indicated’ is inherentlycontroversial. First, one part of a scene or eventbeing covered is selected and reproduced by a(news) photographer, depending on both prac-tical circumstances and aesthetic priorities.Next, a sequence of editors will make addi-tional choices regarding both editing of thevisuals themselves and of their placement in a larger discursive whole, which serves as acomplex index for specific scenes, events, or


tomorrows.’ This may be seen as an attempt to involve the reader in BP’s objectives by instillinga logic which may run something like this: ‘If BP is doing all of this for our (shared) tomorrows,BP is worthy of my respect.’

The text has two instances of word play. The headline plays on two meanings of the phrase‘invite criticism,’ one to attract unwanted criticism, the other to welcome criticism, to even helpcreate it. The effect may be one of surprise at the fact that BP faces myths about ruthless oilcompanies head-on, and appreciation of the copy-writer’s ingenuity. The other instance of wordplay concerns BP’s ‘patronizing’ the environmental groups, so that they may all the better supplyadvice to BP. Again BP faces a common myth about multinational companies, i.e., that they ‘showinappropriate superiority,’ then turns the myth around by exploiting the other meaning of ‘patron-ize’: ‘to enable someone to develop their full potential.’

These instances of word play may also be seen as a kind of absence vis-à-vis consumer ads.What is absent in the BP ad’s word plays is the frivolousness of the typical advertising pun, thecheap joke impression. In the BP ad, word play is used as an intelligent, even sober, way of antic-ipating the reader’s objection to the claims of harmony between the oil company and environ-mental activists.

This unflawed impression of cooperation and dialogue is made possible, in part, by the absenceof the past in the ad’s universe. There is no mention of ‘yesterday,’ the time when oil companiesdid ruthlessly exploit the environment and when there was antagonism between oil companiesand environmentalists.The absence of a lexicalized ‘yesterday’ is supported by the absence of pasttense verbs.

The verbal and visual signs of the BP ad thus appear to create a universe of meaning in whichall conflicts between oil companies and environmental activists have been harmonized away, in aspirit of sympathy and mutuality. Whether this universe, as mapped here through a mixture ofdescriptive discovery and hypothetical interpretation, corresponds to the respective meaning uni-verses of the corporate communicators and the magazine readers is for the empirical study ofencoding and decoding processes to discover (for the reception of corporate advertising, seeSchrøder 1997; for a holistic analysis of the news process, see Deacon et al. 1999).

issues. Further steps in the indexical chain mayinvolve a widening context of interpretation,from the rest of the newspaper and its genreconventions, to the contemporary social set-ting and readers’ horizon of expectations.�All these factors contradict any assumption ofnaturalism, and instead call for intertextual,reception, and other forms of analysis to estab-lish the status of a particular visual index in thelarger communicative process.

The two sources of semiotic concepts –Barthes and Peirce – have been drawn on,liberally and in combination, by numerousanalysts, and have had considerable heuristicvalue for illustrative analyses of how mediaimages communicate (Hall 1973a; Williamson1978). At the same time, it should be acknow-ledged that there have been few systematic, large-scale, or cumulative studies of the specificity of visual media. One complaint is that analysesmay seem commonsensical, perhaps because of the saturation of current society by news and advertising genres, as well as the critical attitude toward them that has been promoted bya double hermeneutic (Giddens 1984) operatingin both education and public debate.

Other, more specific complaints target keytheoretical premises of visual analysis so far.Despite Barthes’ (1970) redevelopment of thedistinction between denotation and connota-tion, as well as other critiques of the implicitthreshold between ‘natural’ and culturally invested meaning (Eco 1976), studies have inpractice often performed a heavy-handed oper-ationalization of the dyad, just as icon, index,and symbol have been approached as separatesign types, rather than complementary functionsof one sign complex. Equally, anchorage andrelay would seem far from exhaustive as catego-rizations of image–text relations in differentgenres and media. In the further development of visual studies, the under-recognized area ofcomics (e.g., McCloud 1994; Sabin 1993) andthe hypertextual formats of computer media(e.g., Bolter and Grusin 1999), in addition tofilm theory and art history, can be sources oftheoretical concepts and analytical procedures.Recently, two distinct programs for such re-

development have been outlined, drawingrespectively on semiotics and cognitive theory.

With a background in critical linguistics,Gunther Kress and Theo van Leeuwen (1996)have proposed to go beyond a mere extensionof discourse analysis from linguistic to visualsigns. They emphasize that the modern media,from school textbooks to the World Wide Web,increasingly feature texts which are ‘multi-modal,’ comprising a range of representationaland communicative forms within the limits of one text, such as, in addition to verballanguage, sound, graphics, and images (includ-ing gestures and proxemics). To these may be added the tactile, olfactory, and gustatorymodes of communicating which may seemrather peripheral, but which may become morewidespread in computer-simulated form. Allthese means of human communication are saidto require an inclusive ‘discourse semiotics’ onone theoretical platform:

Discourse analysis has, on the whole,focused on the linguistically realized text. Inthe multi-modal approach the attempt is tounderstand all the representational modeswhich are in play in the text, in the samedegree of detail and with the same method-ological precision as discourse analysis isable to do with linguistic text.

(Kress et al. 1997: 258)

The theoretical platform of their discoursesemiotics strongly resembles Fairclough’s (1992)three-dimensional model, whereas their analyt-ical practice is inspired eclectically by linguistics,visual semiotics, film theory, art criticism, as well as media research on advertising in par-ticular. In its present form, however, it is notevident that this multi-modal approach repre-sents a genuine, or completed, innovation ofeither discourse or visual media analysis. Irre-spective of the authors’ protestations to thecontrary, their indebtedness to Halliday’s (1978)social semiotics implies a linguistic bias towardother modes of representation. Moreover, likesome earlier applications of the denotation–connotation distinction, this preliminary inte-gration of diverse ‘visual’ and other modes into a discourse semiotics may appear common-


1111234567891012341115678920123456789301234567894012345678950111 � horizon of expectations – Chapter 10, p. 167

comics

multi-modalcommun-ication

sensical, a limitation which is exacerbated byoccasionally simplistic assumptions regardingthe resulting discursive transfer of ideology toaudiences.

The other program for visual media analysisdeviates more radically from linguistic dis-course analysis and semiotics. Cognitive mediatheory� takes as its premise that visual stimulifrom the media activate the same mentalprocesses as other perception and cognition.Contrary to much semiotics, cognitivism findsthat images have no explicit syntax of minimalelements, communicating instead through icon-icity and indexicality. In one of the relatively few cognitivist analyses of still media images sofar (see also Solso 1994), Messaris (1997:182–203) suggested a more general typology of‘propositions’ or principles in visual commun-ication, involving causality, contrast, analogy,and generalization. In this regard, cognitiveimage analysis joins attempts in other commu-nication theory at integrating the analysis of‘texts’ with an understanding of its place inhuman cognition and action.

CONCLUSION

This chapter has outlined the main approachesto the analysis of media texts, recognizing the division into quantitative and qualitativestudies, as well as identifying variations withineach tradition, some of which suggest sharedconcerns across the traditions regarding boththe definition of discursive or content ‘entities’and the development of context-sensitive ana-lytical procedures. An important, if implicitambition of many discourse and content studieshas been to infer from the properties of mediatexts to their probable ‘effects’ on or, in moremodest but also more complex terms, theirrelative contribution to social and culturalprocesses at large, with special reference topublic opinion or the reproduction of ideology.Despite the ambition of thus informing eitherpolicy or public debate on media reception (or media institutions), content studies haveconformed to the traditional focus of mediaresearch on distinct ‘moments’ of the commu-

nicative process, and have, most commonly, notrelated the ‘textual moment’ methodologicallyto the moments of encoding and decoding,production and use. This is in spite of the factthat ‘discourses’ equally enter into, and areanalytical points of access to, both media pro-duction and reception.�

Since neither the intentions of communica-tors nor the meanings created by audiences can be ‘read off’ media by means of discourseor content analysis, a more complete under-standing of the production and circulation ofmeaning in society requires a differentiatedanalytical attention to media texts as vehicleswhich are necessary but not sufficient con-ditions of communication; to their specificmoment in the communicative process; and totheir interrelations with other moments. Thisrequirement has been underscored by the devel-opment of ‘interactive’ media, in which usersnavigate an enhanced intertextual and inter-medial supply of meaning – selecting, combin-ing, and interpreting what is, to a degree, theircustomized text and one which, in both workand leisure, blends into the context of use. Thechallenge for discourse and content studiesunder changing technological and historical cir-cumstances remains one of clarifying its partic-ular domain of analysis and explanatory value,in particular to distinguish the descriptive dis-covery of a text’s verbal and visual characteris-tics from the hypothetical interpretation of itspotential meanings for different audience-users,and from studies of its actualized meanings in social processes of production and recep-tion. Both in analytical descriptions and in the generation of hypothetical interpretations, the relative relevance of the available, qualita-tive as well as quantitative approaches must be assessed with reference to the purpose ofanalysis and the particular object of analysis.Irrespective of their theoretical and method-ological origins, all these approaches are facedwith socially constructed textual objects, andare themselves socially constructing whateverinsights they provide.


� cognitivism – Chapter 2, p. 36� discourses as general elements in research – Chapter15, p. 251

descriptivediscovery

hypotheticalinterpretation

INTRODUCTION

The word ‘text’ is derived from the Latin verbtexo, meaning to weave or twine together, andto construct or build something. The Latinnoun textum means woven cloth or fabric, butwas also used to denote speech and writing. Awritten or spoken text, then, is like a wovencloth – a construction of meaning made out ofwords.

Since the 1960s, ‘text’� has been used as a general term covering diverse phenomena such as music, still images, films, and so on, inaddition to written and spoken language. Theunderlying argument is that all these means ofexpression are semantic constructions – ‘fabrics’of signs – and most of them are meant to be‘read’ in linear, temporal sequences. Such con-structions may be described and analyzed byanalogy to a verbal text, drawing on concepts,models, and procedures that were developedwithin disciplines and fields like linguistics,semiotics, and literary studies. Analyses of‘texts’ in this broad sense form a central part

of media studies. Whether such analyses are car-ried out as studies of single texts or importanttextual genres, or as part of projects studyingmedia reception and media institutions, thecentral concern is with questions of meaningand interpretation. The point of departure is thetexts themselves – their structure and theircontent.

Like any other form of analysis, textualanalysis examines a given object – a text or agroup of texts – as closely and as systematicallyas possible in order to answer specific researchquestions. These questions can lead to twobasic types of textual analysis: one focused ongeneralities, the other on particulars. The firstdescribes recurrent, typical features in order toestablish textual models or prototypes. Thesecond examines the texts in question as iso-lated occurrences with reference to their speci-ficities.� Obviously, there are both transitionalvariants and logical connections between thetwo types. In practice, generalities are alwaysestablished through the study of particulars,

1111234567891012341115678920123456789301234567894012345678950111

Mediated fictionPeter Larsen8

• a description of quantitative studies of representation in media fiction• formal studies in the semiotic tradition, including shot-to-shot analyses of film• narrative studies and models, applied across media types• a presentation of genre analysis as a middle range, connecting text and context• a discussion of sound as an under-researched aspects of media ‘texts’• The Big Sleep (1946) as the main analytical example.

� texts as a general category – Chapter 2, p. 28� generalities and particulars – theory of science,Chapter 15, p. 256

and the analysis of particulars presupposessome knowledge about generalities. Import-antly, qualitative and quantitative methodolo-gies both lend themselves, in specific ways, tothe study of generalities as well as particulars.

This chapter presents a number of content-analytical and text-analytical traditions inmedia studies. Many of the relevant proceduresand principles are not specific to particularmedia or genres. However, this chapter focusesmainly on fiction in film and television, whichaccounts for the bulk of mediated fiction, andwhich has also been the point of departure forthe most significant developments in research.Throughout the chapter, examples fromHoward Hawks’s film The Big Sleep (1946) willbe used to illustrate the various theoretical posi-tions and analytical procedures.

QUANTITATIVE CONTENT ANALYSIS

Content studies of media representations

Some questions about generalities can beanswered in relatively precise terms by mea-suring or counting certain textual features.Since the 1940s, such quantitative analyseshave played an important role in several formsof media research. In the first general accountof this type of analysis, Bernard Berelson(1952) used the term ‘content analysis,’ defin-ing it as a research technique ‘for the objective,systematic, and quantitative description of themanifest content of communication’ (p. 18).�

As this definition suggests, content analysisis basically descriptive. Certain well-definedtextual elements or characteristics are measuredin various ways; for example, the amount ofnewspaper coverage of certain issues, the topicsof television news programs, the occurrence ofparticular symbols in fictional genres, and soon. Such quantitative descriptions of the fea-tures of a group of texts are then typically usedas a basis for more general inferences about themeaning of these texts and their implicationsregarding various social phenomena.

One early example of this approach was Leo

Lowenthal’s study of ‘mass idols’ from 1941.In a quantitative analysis of biographical art-icles in popular U.S. magazines, he showed thatwhile political leaders, businessmen, and scien-tists were portrayed in 46 percent of the casesat the beginning of the twentieth century, thispercentage had dropped to 25 by 1940, and theinterest was now focused, instead, on filmactors, entertainers, and athletes. This shift hadbegun in the 1920s, and was accompanied byan additional shift of focus from the public lifeand achievements of the people in question, totheir love life, their ways of dressing, and otherprivate matters. Lowenthal interpreted thischange of emphasis from ‘idols of production’to ‘idols of consumption,’ and from the publicto the private sphere, as a textual expression ofgeneral changes in social values during theperiod.

Since the early days of content analysis,then, this quantitative methodology has beenused in a variety of research contexts, often in combination with other methodologies. Themain varieties have included (for a more de-tailed overview see Wimmer and Dominick1994):

• studies of patterns and trends in mediarepresentations as such;

• studies of the relationship between textualrepresentations and ‘the real world’;

• studies concerning media effects.

In some studies, the overall content of a specific media genre (e.g., television news) hasbeen analyzed by the Glasgow UniversityMedia Group (1976 and 1980). Other studieshave undertaken analyses of how the mediahandle specific events, such as political demon-strations (e.g., Halloran et al. 1970) and wars(e.g., Morrison and Tumber (1988) on theFalklands conflict, and Morrison (1992) on the Gulf War).

The depiction of violence, particularly intelevision fiction, has been a favorite topic. In such studies, researchers commonly start bydefining how ‘violence’ should be understood,and in which specific discursive contexts itshould be measured. Next, coders watch sam-ples of television programs and count the

Mediated fiction118

� content analysis – Chapter 13, p. 219; exampleanalysis, Chapter 7, p. 101

incidents which match the definition. Thelongest running and most well known of thesestudies was started in the late 1960s by GeorgeGerbner. In the initial research design, hedefined violence as ‘the overt expression ofphysical force (with or without a weapon)against self or other, compelling action againstone’s will on pain of being hurt or killed, oractually hurting or killing’ (Gerbner 1972: 31).From this definition, Gerbner further estab-lished a series of guidelines regarding the typesof events to be included in studies with referenceto their place in the fictional universe and theirrelation to the overall plot (see Gerbner andGross 1976; Gerbner et al. 1977). Continuedduring the 1990s by Gerbner and his team, this series of studies is an example of researchprojects in which content analysis is used tocompare media representations with actualoccurrences in society. These studies haverepeatedly shown that the world of U.S. prime-time television fiction is a far more violent placethan the real world, and that the portrayal ofcertain social groups as frequent victims ofcrime is inconsistent with social statistics. Thesame set of studies has also addressed mediaeffects. Comparing content analyses with surveydata, Gerbner and his team have argued thattelevision fiction ‘cultivates’ or shapes people’sperceptions of social reality, so that heavyviewers tend to have misconceptions of the roleof violence in everyday life.�

A quantitative analysis of classical Hollywoodcinema

Many content analyses thus form part of largerprojects concerning media effects or the rela-tionship between media representations andreal life. Other quantitative analyses, however,have focused on intratextual research ques-tions. A well-known example of the use ofcontent analysis for describing and interpretingparticular textual genres is a study by DavidBordwell (1985b) of the style of classicalHollywood cinema.

The point of departure for that study was alist of the approximately fifteen thousand

feature films produced in the U.S.A. in theperiod 1915 to 1960. Bordwell and his co-authors used a table of random numbers toselect 841 films from this list. Of these, 100could be located in various collections andarchives. As the authors emphasize, this wasnot, strictly speaking, a random sample, sinceeach film did not have an equal chance of beingselected, because not every film on the originallist had survived. The point, however, is thatthe actual choices of films ‘were not biased bypersonal preferences or conceptions of influen-tial or masterful films’ (Bordwell 1985b: 388).In fact, four-fifths of the resulting sample werefairly obscure productions.

The authors studied each film in the sampleon a viewing machine, recording stylistic detailsof each shot and summarizing actions scene byscene. On the basis of these descriptions, theyconstructed a model of the ‘typical film’ of the period, with special attention to style andnarration, and went on to test this generaliza-tion by analyzing almost two hundred otherHollywood films from the same period.Although many of the films in this secondsample were chosen precisely for their qualityor historical influence, the analyses confirmedthe general model (Bordwell 1985b: 10). Thestudy stands as an impressive account of typicalaspects of the classical Hollywood film in termsof features such as story construction, narra-tional strategies, the construction of time andspace, etc. Moreover, this description of invari-ables and stabilities provides an importantframework for further research on the equallyobvious stylistic changes during the classicalHollywood period.

QUALITATIVE TEXTUAL ANALYSIS ANDSEMIOTICS

While quantitative content analyses are descrip-tive in nature, aiming at ‘the manifest contentof communication’ (Berelson 1952: 18), textualanalyses in the tradition of literary criticism andart history are usually interpretive, aiming atwhat is sometimes termed latent meaning. Thebasic questions, consequently, are ‘qualitative’:What does the text really mean, and how areits meanings organized? Still, analyses may be

Qualitative textual analysis and semiotics 119

1111234567891012341115678920123456789301234567894012345678950111 � cultivation research – Chapter 9, p. 150

manifest vs.latent meaning

concerned with either particulars or general-ities. While focusing on features that are char-acteristic of a single work or a small group ofworks, studies also examine how the givenworks differ from all other works in structuraland thematic terms.

The sources of this analytical practiceinclude the attentive ‘close reading’ practiced bythe Anglo-American school of New Criticism�

and its French counterpart, the explication dutexte. In both of these traditions, it is the sin-gular or particular analytical object, its partsand its whole, which is at the center of inter-est. The underlying assumption, and the veryreason for giving particular attention to theworks in question, is that they are thought tobe significant carriers of cultural values andinsights, or that they provide important andvaluable aesthetic experiences. This type oftextual analysis has been transferred success-fully to film studies – an area of media studiesthat deals with textual objects which are, insome ways, comparable to the unique worksanalyzed by literary critics and art historians(see, e.g., Robin Wood on Hitchcock (1965)and on Hawks (1968)). However, the basic aimof most other media research has been to studyprototypes, regularities, repeated patterns, andfeatures which are shared by masses of texts.Accordingly, textual analyses of media haveusually taken their inspiration from humanistictraditions which are oriented toward the studyof generalities, particularly structural linguisticsand general semiotics.�

Applied to media studies, Ferdinand deSaussure’s (1916) linguistic dichotomy oflangue and parole� has proven to be an effec-tive conceptual tool: concrete media texts maybe regarded as instances of parole, as generatedfrom one or more langues. Even though thevarious instances of parole are carried by dif-ferent types of signifiers (images, written orspoken language, combinations of these, etc.);are transmitted through different media; andare produced according to specific rules, it may

be argued that all sign production is based oncertain general principles. Whether the analyt-ical intention is to reconstruct the languebehind a group of texts, or to study a singletext as parole, knowledge of such generaltextual principles is useful. The underlyinglangue is usually understood as a ‘code’� thatconsists of signs, on the one hand, and of seriesof syntactic rules, on the other. Meaning pro-duction is thus thought of as a process by whichsigns are selected from latent groups of possi-bilities (paradigms) and combined according tothe relevant syntactic rules into strings of text(syntagms).�

Saussure (1916) defined the sign as amaterial object – a signifier – evoking a certainmental representation – a signified.� Corres-ponding to these two aspects, there are twomain types of semiotic media analysis: a formaland material study of signs in the media, andan analysis of media contents as representationswith an inherent meaning. However, since thetwo aspects of the sign are interdependent,there can be no absolute distinction betweenthe two types of textual analysis.

FORMAL ANALYSIS

Formal analyses (of signifiers) highlight thematerial specificity of the medium in question.What are its particular properties, and how dothese properties translate into communicativepossibilities? In the case of visual media, keytheoretical questions have been: What is animage? Are images signs? (Barthes 1964b; Eco1968, 1976; Sonesson 1989). How do photo-graphic images differ from other types of signs?(Barthes 1980). Concrete analyses in the areahave dealt with questions concerning pictorialrepresentation, composition, and style, in somecases relying on concepts and analytical proce-dures from the study of traditional arts (for abrief overview see Bryson 1991).�

Mediated fiction120

� New Criticism – Chapter 2, p. 29

� semiotics – Chapter 2, p, 23

� langue and parole – Chapter 2, p. 24

� code – Chapter 1, p. 8

� paradigms and syntagms – Chapter 2, p. 24

� signifier and signified – Chapter 2, p. 25

� art history – Chapter 2, p. 27

Moving images: syntagm analyses

The ways in which moving images are pro-duced and structured raise special theoreticaland analytical problems.� One of the firstresearchers to address these problems from asemiotic perspective was the French linguist,Christian Metz. In an influential study of howcinematic signifiers are organized, he showedthat the majority of films are based on seventypes of ‘syntagms,’ or seven basic series ofshots (Metz 1968). Two of the most centraltypes are exemplified in Plate 8.1.

Four narrative syntagms indicate temporalrelations between narrative events:

1 The scene is the most widely used of these– a series of shots presenting an event in con-tinuous time and space.

2 In the alternating syntagm, there is cross-cutting between several narrative spaces (forexample, in a ‘chase’) as well as indications oftemporal simultaneity between the shots.

3 In sequences, the shots indicate discontinu-ous time: the ordinary sequence is an ellipticalconstruction in which unimportant events andother details are left out.

4 The episodic sequence, in comparison, orga-nizes the shots so that the omissions suggest acompressed chronological development.

In addition to the four narrative syntagms thereare three a-chronological ones:

1 The descriptive syntagm is a series of shotssuggesting spatial co-presence of people orobjects.

2 The bracket syntagm depicts typical aspectsof a phenomenon or a concept (‘poverty,’‘morning in the city,’ etc.).

3 The parallel syntagm organizes two series ofcontrasting motifs (‘rich and poor,’ ‘town andcountry,’ etc.).

Metz’ intention was to describe cinema as ageneral langue, or at least one of its main com-

ponent structures. His framework has alsoproven useful, however, for the analysis of indi-vidual films. Knowledge of the basic syntagmsystem can serve as an ‘attention-focusingdevice’ for the analyst (Stam et al. 1992: 48),and is also helpful when one aims at definingfundamental formal characteristics of specificgenres or films (see, e.g., Ellis (1975) on Ealingcomedies; Flitterman-Lewis (1983) on soapoperas; Heath (1975) on Orson Welles’s Touchof Evil).

A brief example from The Big Sleep illus-trates this point. The story concerns old GeneralSternwood, who hires the private detectiveMarlowe to solve a mystery. The case involvesthe General’s youngest daughter, Carmen, whois being blackmailed. As Marlowe is working onthe case, additional mysteries are introduced: aformer employee of the General has disap-peared; the blackmailer is found murdered; apowerful gangster plays an obscure role in all ofthese events. Along the way, Marlowe also fallsin love with the General’s oldest daughter,Vivian. At the end of the film, Marlowe solvesthe mysteries, and the gangster, who is the mainculprit, is shot by his own men.

During the first few minutes of the film, wesee Marlowe visiting the General. First, hemeets Carmen in the hall of the Sternwoodmansion; he then meets the general himself ina winter garden; and finally he meets Vivian inher room. A butler leads him from room toroom and finally back to the entrance. Fromthese encounters, Marlowe – and we, the audi-ence – learn a lot about the Sternwoods and thecase. Next follows a shot of a sign saying‘Hollywood Public Library,’ a close-up ofbooks and documents, and a series of shots ofMarlowe sitting in the library, reading andtaking notes.

On the level of signifieds or content, theintroduction provides Marlowe (and the audi-ence) with necessary background information.On the level of signifiers or form, this infor-mation is organized in three distinct ‘scenes’(Marlowe’s three encounters in the Sternwoodmansion, each of them presented in spatio-tem-poral continuity). When he starts working onthe case, a summarizing ‘ordinary sequence’follows (Marlowe is seen in the library study-

Formal analysis 121

1111234567891012341115678920123456789301234567894012345678950111 � analysis of still images – Chapter 7, p. 110

seven types ofsyntagms

ing, but unimportant details are skipped). (SeePlate 8.1.)

Most of the rest of the film is constructedalong these lines. The narrative is presentedmainly in the form of scenes and ordinarysequences, in addition to the occasional alter-nating syntagm. This is a familiar pattern, thepattern of classical Hollywood cinema. Mostfilms from this period use a very limited partof the general syntagm system, as described byMetz. Indeed, Bordwell (1985a: 158), analyz-ing classical Hollywood films from a cognitivistperspective,� arrived at a similar conclusion.The majority of films in this period are basedon only two syntagms or ‘segments’ as he callsthem – the ‘scene’ and the ‘montage sequence’– corresponding roughly to Metz’s ‘scene’ and‘sequence.’

The correspondence between a semiotic anda cognitivist description of the classicalHollywood film is no coincidence. When Metzworked out his system, he followed the generalsemiotic and linguistic principle of describinghis materials according to the distinctions thata ‘native speaker’ – in his case, an experiencedfilm spectator – would make. Each of hissyntagms are defined with reference to thecharacteristic spatio-temporal effect which anymember of a cinema audience would immedi-ately recognize. There is an evident connectionbetween this approach and cognitivist studies ofthe various ‘schemata’ (organized clusters ofknowledge) which people apply in their every-day life. In both cases, the intention is todescribe and systematize what ‘everybody’ intu-itively ‘knows’ about the world (for a discussionof the relation between semiotics and cognitivescience in film studies see Buckland 2000).

Mediated fiction122

Plate 8.1 The introduction to The Big Sleep: Marlowe’s encounters with Carmen Sternwood, GeneralSternwood,Vivian Sternwood, and his visit to the public library

� cognitivism – Chapter 2, p. 36

Compared to other types of films, classicalHollywood films constitute a class of their ownin terms of both formal and narrative strategy.The so-called montage films that were producedin the USSR during the 1920s are obviouscounter-examples to this norm. They areovertly ‘rhetorical,’ constantly presenting polit-ical arguments by visual means. While, forinstance, the ‘parallel syntagm’ is very rare inclassical Hollywood cinema, it is a centraldevice in these Soviet films, where it was usedto suggest all kinds of rhetorical analogies orcontrasts (Bordwell 1985a: 239).

Shot-to-shot analyses

While Metz’s concerns were theoretical, hisaccount of the basic syntagm system providesa useful, if rather broad framework for theanalysis of particular films. Raymond Bellour is an example of a film semiotician whose works are not theoretical in a strict sense, butprimarily concerned with practical, analyticalquestions and, as such, more closely related totraditional literary analysis. During the 1970s,Bellour presented a series of ‘close readings’ ofwell-known Hollywood films (see the collectionin Bellour 1979). By means of meticulous shot-to-shot analyses, primarily of formal featuresand often of very short fragments, he describedthe films in question, established their indi-vidual cinematic systems, as it were, and usedthese characterizations as a basis for conclu-sions about formal aspects of the classicalHollywood system in general.

One example of this procedure is his analy-sis of a short segment from The Big Sleep(Bellour 1973). Consisting of twelve shots, thesegment, inconspicuous, almost trivial, marks atransition between two action-packed, narra-tive climaxes. After a shoot-out at a gas stationin the country, Marlowe and Vivian drive awayin a car; on the road toward the final shoot-out, they declare their love for each other. (Thetwelve shots are rendered in Plate 8.2.)

Bellour describes the twelve shots withregard to framing, camera movement, cameraangles, absence or presence of characters inshots, absence or presence of speech duringshots, and relative length of shots. His results

may be summarized as follows. Althoughnothing much happens, the number of shots isrelatively high – a fact which Bellour explainsas a strategy of discontinuity, of introducingvariations into the filmic space within the giventime frame. At the same time, these variationsappear against a background of similarities.After the location of the scene is established bya shot showing the car driving along seen fromthe outside, there follows:

• a medium-close shot of Marlowe and Vivianside by side;

• a close-up of Marlowe;• a close-up of Vivian;• the remainder of the scene is made up of

almost identical medium-close shots andclose-ups of Marlowe and Vivian.

This repetition of formal qualities produces asense of constraint around the actual variationin the actors’ dialogue and comportment,making the variations stand out as differenceswhich ensure the forward movement and con-tinuity of the narrative. Bellour finds this for-mal arrangement to be characteristic not onlyof The Big Sleep, but of classical Hollywoodcinema as such.

NARRATIVE ANALYSIS

Narrative semiotics or narratology – the studyof basic narrative patterns and procedures – hasbeen one of the most fertile interdisciplinaryfields of study during recent decades, and hasproven particularly effective and valuable foranalytical purposes. By way of introduction, itshould be noted that narrative is both a mentalstructure and a specific type of text. As a mental‘tool,’ it functions as a fundamental interpre-tive frame: humans make narratives in order toorganize their experiences and to make theirworld intelligible (on the relations between nar-rative thought and narrative discourse seeBranigan 1992; Bruner 1986, 1994; Labov andWaletzky 1967).

Narrative texts appear in all kinds of dis-cursive forms and in all kinds of media (Barthes1966). As a textual type, however, the narra-tive is defined solely by its content: a narrative

Narrative analysis 123

1111234567891012341115678920123456789301234567894012345678950111

montage films

shot: con-tinuous visual

sequence,filmed by

one camera,without cuts,

wipes,dissolves, or

other editing

narratology

narrative:textual as wellas mentalstructure

Mediated fiction124

is a representation of events in time and space.These events are organized in series of causesand effects, and viewed in relation to humanprojects which they either further or impede(for detailed definitions of narrative see Bal1985 [1977]; Branigan 1992; Bremond 1966;Rimmon-Kenan 1983; Todorov 1971).

Narratives as sequences of events

The causal organization of events leads to char-acteristic narrative patterns. In 1928, theRussian formalist Vladimir Propp showed thatvirtually all narratives within a large selectionof folk-tales were based on identical events and,further, that these events were presented inidentical order (Propp 1958). Some of theserecurrences could be explained as the result ofgenre conventions, but later studies have con-firmed that there is a more general and sys-tematic regularity at work in the ordering ofnarrative events. At a basic level, not only these

Russian folk-tales but all narratives are seriesof variations on a simple pattern, consisting ofan initial state which is transformed by adynamic event into a new state (Bremond 1966;Todorov 1971).

No matter what a narrative is about, it willalways be constructed from such modules or‘elementary sequences.’ The same characteristicpattern may also be observed at the narrativemacro level. A narrative will usually start withthe presentation of a situation, a setting, theprincipal characters, the general state of affairs,and so on. This initial situation is then gradu-ally transformed by a series of events andactions, until a new situation has been estab-lished and the story ends. According to story-comprehension psychologists, this macropattern is also the mental schema that peoplemobilize when they recognize and try to struc-ture unfamiliar narratives (see Bower and Cirilo1985; Gulich and Quasthoff 1985; Van Dijkand Kintsch 1983; also Branigan 1992).


1111234567891012341115678920123456789301234567894012345678950111

Plate 8.2 Bellour’s (1973) shot-to-shot analysis of a segment from The Big Sleep

elementarysequences

At the macro level, the narrative sequencefrom beginning to end usually involves a move-ment between central thematic positions or‘values’ – from an initial, problematic, andunstable situation, through successive actions,to a final, stable, and acceptable situation. Inanalyses of concrete narratives, it is thereforehelpful to begin by describing how the string ofelementary sequences is organized at the macrolevel. This procedure provides not only a roughoutline, but a way of gaining important insightsinto the underlying value system.

The what and how of narratives

Any narrative may be viewed from two differ-ent points of view. On the one hand, there isthe series of events; on the other hand, there isthe actual text by means of which these eventsare represented. These two aspects may bedefined in terms of the general sign model: thewhat of the narrative (the series of events) isthe narrative signified, while the how (theactual text) is the signifier. This distinction isusually referred to in narratology by the termsstory and discourse (Genette 1980 [1972]; seealso Chatman 1989).

The Russian formalists� employed a similardistinction, but emphasized the audience’sactive participation in the construction of nar-rative coherence. David Bordwell has reintro-duced the formalists’ term syuzhet to cover theseries of events which is presented explicitly inthe narrative text, while fabula refers to thespectator’s construction of connections betweenthese events. No narrative discourse presentsthe full story – there will always be indeterm-inacies and vaguenesses. To understand a nar-rative is precisely to interpret the availableinformation (syuzhet), and to construct textualcoherence (fabula) (Bordwell 1985a).

The construction of the fabula by the audi-ence, then, is dependent on the syuzhet, but thespecific syuzhet, in its turn, is also only oneamong many possible implementations of thefabula. Moreover, the particular way in whichcrucial narrative information from the fabula isorganized and presented in the syuzhet has

important consequences for the audience’sexpectations and comprehension. The fabula–syuzhet relationship may even be a decisiveelement in the definition of a particular narra-tive genre. Mystery and other suspense storiesare well-known examples of this. Their narra-tive strategy is to generate suspense by break-ing the chronological order of events, so thatcrucial fabula information about the originalcause of the mystery is presented very late inthe syuzhet.

Just as one fabula may give rise to a multi-tude of syuzhets, it may also be implemented inmany discursive forms and media types.Consequently, the fabula–syuzhet distinction isvery useful for analyzing adaptations of narra-tives across different media, as illustrated byThe Big Sleep. The film is based on RaymondChandler’s debut novel of the same name from1939. A textual analysis of fabula and syuzhetshows that most of the novel’s events (fabula)as well as their presentation (syuzhet) have beentransferred to the film relatively faithfully bythe scriptwriters. Nevertheless, the film differsfrom the novel in several respects because oftwo types of changes. First, several minorevents and minor characters have simply beencut during the adaptation process. This was,and is, standard procedure in adaptations:‘unnecessary’ details and narrative intricaciesare pruned away for the sake of comprehens-ibility, but also to make long stories fit the rel-atively short feature film format (on adaptationpractices see Lothe 2000; Naremore 2000).

Second, and more interestingly, the constel-lation of the central characters has beenchanged rather dramatically. In the novel, ascheming Vivian tries to seduce Marlowe inorder to obstruct his detection work, while hefor his part firmly refuses her advances; in thefilm, she saves his life, helps him solve themystery, and they end up being lovers. Givinga traditional happy ending to the originalmelancholy story of a lonely, disillusioneddetective is quite a radical rewrite, which callsfor contextual explanations over and above the textual, narrative economy. Film histo-rians usually suggest that Hawks’s previousmovie, To Have And Have Not (1944), hadestablished Humphrey Bogart and Lauren

Mediated fiction126

story anddiscourse

syuzhet andfabula

� Russian formalism – Chapter 2, p. 29

adaptationsacross media

textual andcontextualevidence

Bacall as Hollywood’s new romantic couple,and that the rewriting of Chandler’s story wasdone in order to support this constellation(Clark 1983).

Characters and actants

Narrative series of events, then, are structuredaccording to certain basic patterns. Similar reg-ularities apply to the individual characters oragents who are involved in and affected by theevents. On the face of it, there is a multitudeof characters, each of them endowed with spe-cific, individual traits and qualities. However,these characters have rather limited spheres ofaction. As demonstrated by A.J. Greimas, thereare six ‘basic positions’ that characters canassume in relation to the central project of thenarrative. The multitude of narrative characterscan thus be viewed as surface manifestations ofsix underlying narrative roles, the so-called‘actants’ (Greimas 1966). The actant model ispresented in Figure 8.1.

The moment a narrative project is stated,two actants are established: a Subject whodesires an Object. In addition, all narrativeprojects have to do with communication in the sense of ‘transport’: the Object has to be‘moved’ between two positions. Therefore, the statement of a project also establishes twofurther actants: the potential Sender who ‘has’the Object, and the potential Receiver who‘lacks’ it. The transport of the Object is usuallycomplicated by a conflict between competingprojects within the narrative universe. In suchcases, the Subject is faced with an Opponentwho will try to prevent the transport. Further-more, there will often be a Helper, i.e., anactant who supports the project and works tofacilitate the transport of the Object.

The actant model is a simple and effectiveanalytical tool. Its strength is that it accountsfor all actions and characters from the samepoint of view. Focusing on how characters arepositioned in relation to the core project of thenarrative, the model makes it possible toprovide a description of the most basic relationsand conflicts in a narrative.

In The Big Sleep, old General Sternwoodhires Marlowe to solve the mystery. The basic

project for Marlowe (Subject) is to obtain infor-mation (Object) for his client, the General(Receiver). But who has the necessary informa-tion, i.e., who is the Sender? Trying to find theanswer to this question, Marlowe meets thetwo daughters, assorted gangsters, and severalother people who all try to prevent him fromlooking deeper into the matter. All these char-acters fill the Opponent position. Late in thestory, Vivian decides to side with Marlowe. Shebecomes his Helper, and it turns out that she isalso a Sender, one of the many people who havetried to hide important information from him.After a series of dramatic events, Marlowe ulti-mately solves the mystery.

As this summary analysis shows, there is noone-to-one relation between characters andactants. For one thing, one actant position maybe filled by many different characters. In thiscase, both the Opponent and the Sender posi-tions are filled by large groups of characters.For another thing, the same character can haveseveral functions in the narrative, and maytherefore fill several actant positions. MostOpponents in The Big Sleep are also Senders –quite a common constellation. Another com-mon constellation is identity between the char-acters filling the Subject and Receiver actants.In many narratives, the ‘hero’ undertakes thenarrative project for his or her own sake. Thereare, however, also many examples of Subjectsworking selflessly for others, or – as in detec-tive narratives like The Big Sleep – a Receiverhiring somebody else to do the Subject’s job.

A final observation is that the actant modelmay be used in two different ways. It is mostcommonly applied to the fabula as a whole –after the fact, so to speak. However, as theexample of Vivian suggests, characters can –and often do – change actant position in the


1111234567891012341115678920123456789301234567894012345678950111

agents andactants

Sender Object

Helper

Receiver

OpponentSubject

Figure 8.1 The actant model

course of a narrative. Thus, the model may also usefully be applied to shorter narrativesequences as a means of getting an overview of‘local’ or individual projects and their configu-ration within the unfolding presentation of thesyuzhet.

The canonical story

A textual presentation of a coherent series ofevents centered around a human project – thisis a broad definition aimed at covering all nar-ratives. Some narratives, however, are obvi-ously more well suited for a given purpose than others. Story-comprehension researchershave argued that the main narrative forms of Western culture represent a special variantof the basic pattern.� This ‘canonical story’ isthe foundation, for example, of the classicalHollywood film, which, according to DavidBordwell,

presents psychologically defined individualswho struggle to solve a clear-cut problem orto attain specific goals. In the course of thisstruggle, the characters enter into conflictwith others or with external circumstances.The story ends with a decisive victory ordefeat, a resolution of the problem and aclear achievement or nonachievement of thegoals. The principal causal agency is thusthe character, a discriminated individual en-dowed with a consistent batch of evidenttraits, qualities, and behaviors.

(Bordwell 1985a: 157)

Compared to the general definition of nar-rative, this description of the canonical story isfar more detailed. All narratives involve humanprojects, but in the canonical story, the seriesof events appears totally controlled by individ-uals. The resulting clarity is the most generalfeature of this mode of narration. The centralcharacters are clearly defined in terms of theirpsychological make-up; their problems andgoals are clear; the story ends with a clear res-olution, and so on.

The canonical story is even predefined at thelevel of actants. The main narrative project andthe Object is clearly stated from the outset, andthe other actant positions are filled with indi-viduals. There is always a conflict in this nar-rative universe, i.e., conflicting projects andhence a clearly defined Opponent. In addition,there are also parallel projects. The canonicalstory commonly presents two storylines withthe same Subject. One part of the story is‘private’ and deals with the (usually male)Subject’s project of getting the Object-womanhe loves; the other part is more ‘public,’ in thatit deals with ‘another sphere – work, war, amission or quest, other personal relationships’(Bordwell 1985a: 157). These two projects areusually interwoven – ‘textually’ – in such a waythat the success of the ‘public’ project is a pre-condition for the completion of the ‘personal’project.

The Big Sleep meets most of the require-ments of the canonical story. The narrative ischaracter-driven, the project is clear, there isconflict, and the story ends with a decisivevictory for Marlowe. The film is also a goodexample of parallel projects. During Marlowe’squest for information, he and Vivian fall inlove, and since she is deeply involved in the casehe is working on, he is faced with two inter-twined problems in the last portion of the film.Not only must he solve the General’s mystery,but he must also save Vivian from gangstersand from the police.

Narration and narrators

Some narratives are told face-to-face with anactual narrator addressing an actual audience‘live.’ In technologically mediated narratives,narrators and their audiences are separated intime and space. And yet a novel, a film, and an episode in a television series all seem to‘address’ their audiences, and give the impres-sion of being ‘told’ by someone. Most studiesof such textual signs of ‘narration’ assume thatthe structure of the actual communicativeencounter is repeated, as it were, within thetextual object itself. This view was originallyput forward by Jakobson (1960), who arguedthat most texts contain elements whose prime

Mediated fiction128

local andglobal

narrativeprojects

� Western and non-Western narratives – Chapter 11, p. 179

parallelprojects:private andpublic

function is to refer to the ‘sender,’ the ‘receiver,’and to the ‘message’ or the textual objectitself.�

In an influential study of narration in litera-ture and film, Chatman (1989) adopted thisview, and went on to suggest some additionalconceptual distinctions (Figure 8.2). In the em-pirical communication situation, a Real authorpresents a Narrative text to a Real reader; withinthe Narrative text, there are traces of an Impliedauthor and an Implied reader.

The term ‘implied author’ was coined byWayne Booth to indicate ‘the picture a readergets’ of an author by reading a novel (Booth1961: 70–71). Chatman expanded the term todenote the ‘principle . . . that stacked the cardsin this particular way, had these things happento these characters, in these words or images. . . . It instructs us silently, through the design ofthe whole, with all the voices, by all the meansit has chosen to let us learn’ (Chatman 1989:148). The implied author is shorthand for thenorms and values that a text suggests to itsreader by the way in which it is organized.Similarly, the term ‘implied reader’ refers to ‘theaudience presupposed by the narrative itself’ (p. 150). Chatman’s distinction between the‘real reader’ and the ‘implied reader’ furthercorresponds to a distinction in reception analy-sis: the empirical readers, or the actual mediaaudience, as opposed to the reader-in-the-text,i.e., the various roles or positions which a textoffers its reader by means of its organization(see Eco 1987; Iser 1972, 1976).�

The implied author and reader are struc-tures present in any text. In addition, some

texts explicitly mark themselves as being toldby ‘someone,’ a narrator, usually a characterspeaking in the first person, an ‘I.’ The narra-tor may be a character acting inside the narra-tive universe, or an observer reporting from anexternal position. Sometimes, the narrator hasa counterpart in a narratee, when the narratorexplicitly addresses ‘someone,’ a ‘you,’ a ‘DearReader,’ etc.

Chatman’s simple narration model (Figure8.2) has been criticized by film scholars. This ispartly because it is so obviously based on liter-ature and written language, but also because itimplies the traditional communication sequenceof a message being transported from a clearlydelimited sender to a receiver. Terms like‘author’ and ‘narrator’ inevitably suggest a typeof individual artistic production which is veryrare in modern cultural industries. Arguingagainst Chatman, David Bordwell has sug-gested that one should avoid thinking of nar-ration as a process driven by narrators, asmodeled on the ‘author’ of a novel. Narration

is better understood as the organization ofa set of cues for the construction of a story.This presupposes a perceiver, but not anysender, of a message. This scheme allows forthe possibility that the narrational processmay sometimes mimic the communicationsituation more or less fully. A text’s narra-tion may emit cues that suggest a narrator,or a ‘narratee,’ or it may not.

(Bordwell 1985a: 62)

In fact, Bordwell’s view of narration is notradically different from that of Chatman. Bothagree that identifiable narrators and narrateesare special cases, and both are primarily con-cerned with narration, which both regard as a


1111234567891012341115678920123456789301234567894012345678950111

� three models of communication – Chapter 1, p. 8

� empirical reception analysis – Chapter 10

implied authorand reader

Narrative text

Realauthor

Realreader

Impliedauthor

Impliedreader(Narrator) (Narratee)

Figure 8.2 A model of narrative communication

narrator andnarratee

special type of textual organization (seeChatman 1990; for an overview of the discus-sion, see Lothe 2000). However, for an analy-sis of the means of the narration in actual texts,one needs a more differentiated description ofthe area than the ones suggested by eitherBordwell or Chatman.

Levels of narration

In a helpful discussion of the possible ways of providing narrative information in a film,Edward Branigan argues that there are eightlevels of narration, all of which derive from thetraditional communication structure – ‘some-one tells something to someone’ (Branigan1992: 86f.). Figure 8.3 is a simplified versionof his overview.

The point of departure for Branigan, as forChatman, is a historical author and a histori-cal artefact, a text, presented to a historicalaudience. Inside this text, the communicationstructure is repeated on seven levels:

1 The extra-fictional narrator appears in thetext, but is not part of its fiction, and is there-fore capable of referring to it precisely asfiction. The title sequence of a film is a typicalexample of information provided by this typeof narrator (‘Warner Bros. Pictures Inc. | JackL. Warner, executive producer | Presents |Humphrey Bogart | Lauren Bacall | ‘The BigSleep’ | A Howard Hawks Production . . .’). Theextra-fictional narrator may be marked explic-itly in the text, as in this title sequence. He/shemay also work like Chatman’s ‘implied author,’i.e., as an implicit presence suggested to theaudience by the arrangement of certain ele-ments in the fiction. Early in The Big Sleep,there is an example of an extra-fictional narra-tor ‘stacking the cards’ in order to control thereading of the narrative. Marlowe is waitingoutside the house of the blackmailer, when hehears gunshots and a scream. As he runstoward the entrance door, someone is runningaway on the other side of the house. The audi-ence actually sees the person in the followingshot, but this is cropped in such a way that itshows only the running feet. Clearly, this is notMarlowe’s vision of the scene. Being on the

wrong side of the house, he hears only thesound of footsteps. The shot of the feet mustbe attributed, instead, to ‘someone’ who, froma position outside the fiction, is creating sus-pense. This is accomplished by organizing theaudience view of events in such a way that weare allowed to see the running person, but aredenied full knowledge of his identity.

2 A non-diegetic narrator is present inside thetext as well as inside the fiction, but is outsidethe ‘diegesis,’ or the story world. This type ofnarrator observes the story world, but does notact in it. Well-known cinematic examples arethe titles that help spectators orient themselvesin the story world, ranging from simple state-ments of time or place (‘San Francisco’) to more elaborate presentations of necessaryinformation, such as the one opening JohnHouston’s The Maltese Falcon (1941): ‘In 1539the Knights Templar of Malta paid tribute toCharles V of Spain by sending him a GoldenFalcon encrusted from beak to claw with rarestjewels. But pirates seized the galley carrying thispriceless token, and the fate of the MalteseFalcon remains a mystery to this day.’

3 The diegetic narrator is inside the storyworld and can address events in which he orshe has participated. Narrators of this typemanifest themselves whenever characters in thestory world tell each other about their experi-

Mediated fiction130

Historical author

Extra-fictional narrator

Nondiegetic narrator

Diegetic narrator

Character

External focalization

Internal focalization, surface

Internal focalization, depth

: Text

: Fiction

: Story World

: Events

: Action

: Speech

: Perception

: Thought

Figure 8.3 Eight levels of narrationSource: Adapted from Branigan 1992

diegesis

ences. In addition, entire narratives may be told by diegetic narrators. A novel like JosephConrad’s Heart of Darkness (1899, revised1902) is an example. Here, the complex story of a fateful journey to Africa is told inretrospect in the first person by one of thecharacters. When Francis Ford Coppola madeApocalypse Now (1979), a film adaptation ofConrad’s novel, he chose the cinematic coun-terpart to diegetic narration by letting the maincharacter report and comment on narrativeevents on the sound-track in voice-over (seeLothe 2000; for a discussion of voice-over nar-ration, see Kozloff 1988).

The term ‘narrator’ suggests the presence ofa ‘voice.’ Beyond this, characters within thestory world may provide us with informationabout this world in other ways than by simply‘telling’ it:

• Characters enable us to learn about them aswe observe their actions, what is also knownas ‘character narration’.

• External focalization is one way of learningabout the story world by sharing a charac-ter’s experiences. Here, the character is pre-sented from the outside, but we understandthat we are in his or her world – that it isthis particular character who is the sourceof our knowledge.

• Internal focalizations allow us to see thestory world more literally from a character’sperspective. In the first type, surface focal-ization, we share a character’s subjectiveexperience, for example, by seeing what thecharacter sees by means of a point-of-view-shot.

• In a deep focalization, finally, we – the audi-ence – share a character’s thoughts, dreams,hallucinations, etc.

Narrative analysis is always centrally con-cerned with how a given narrative applies theseseveral modes of narration. A narrative derivesits distinctive character not only from its pre-sentation of various events and characters tothe audience, but also from the specific ways inwhich information about these events and char-acters is organized, and how the audience isaddressed. Some narratives belong to genreswhich adhere to relatively fixed narrationalsystems. A brief account of how Western tele-vision news is presented illustrates this point.The example also serves to emphasize that, astextual form, the narrative cuts across trad-itional boundaries between fact and fiction.

A typical news program is made up of vari-ous stories from real life. These stories are toldprimarily by anchorpersons and journalists inthe studio, and by journalists reporting from the field. In Branigan’s (1992) terminology, theanchorpersons and journalists in the studio are non-diegetic narrators who observe the story world and talk about it from the outside.Journalists in the field, on the scene where the events are taking or have taken place, arediegetic narrators. Reporting from inside thestory world, these journalists may interview people who are directly involved in events.Whentalking about their experiences, these peopleequally become diegetic narrators, as do peoplewho are interviewed by journalists in the studio.

The relations between these narrators areusually indicated by a simple visual code(Larsen 1974). The most important elements ofthis code are, first, the presence or absence ofeye contact with the viewers, and second, thedistinction between being present in the studioor in the field (Figure 8.4).

The anchorpersons and journalists in thestudio as well as the journalists in the field areallowed to address the audience directly, thus


111123456789101234111567892011123456789301234567894012345678950111

voice-overnarration

characternarration

point-of-viewshot

fictional aswell as factualnarratives

Eye contact

No eye contact

Studio

Anchorpersons, journalists

Interviewees

Field

Journalists

Interviewees

Figure 8.4 The visual code of television news

simulating face-to-face communication. In con-trast, the interviewees both in the studio and inthe field have to address the interviewer. Thesedistinctions – direct/non-direct address; studio/field – thus form a simple system which helpsthe audience to distinguish between the varioustypes of narrators, and to understand their rel-ative authority within the narrative hierarchy.The most important feature of each narrativesequence is decided by the eye contact, whichshows, in a very concrete and visually explicitway, who has the right to address the viewers,and to interpret the world for them.

In a mainstream feature film, the relationsbetween the various levels of narration are con-siderably more complex, and not marked asclearly as in a news program. The characters ina film address each other – they talk and act‘inwards’ into the story world, seeminglyunaware of the audience outside the text. Incertain cases, the narration becomes rather con-spicuous – for example, when the audience isallowed to share a character’s perceptions andthoughts by means of internal focalization – butmostly the story seems to tell itself in an almostimperceptible manner. This general impressionof ‘invisible narration’ in the mainstream filmis the result of a subtle narrational system thatis based primarily on non-diegetic narration viathe camera and various forms of external focal-ization (on editing see, e.g., Bordwell andThompson 1997).

GENRE ANALYSIS

Genre is the French translation of the Latingenus, a word originally referring to pheno-mena which resemble each other due to theircommon source or other close, mutual rela-tions. Genus means origin, but also family,biological genus, grammatical gender, and,more generally, kind, sort, or class. Its Frenchcounterpart is similarly used in reference togroups of texts or works of art that are con-nected by a number of common features. Theunderlying assumption is that a unique text isquite a rare phenomenon. Most texts belong tolarger classes of texts, and these classes, in turn,are defined by the features shared by the indi-vidual texts.

Classical writings on textual genres dealtwith a relatively manageable field. The classeswere few and defined by a limited number oftextual features, as witnessed by the traditionalpartitioning of the field into drama, epic, andpoetry. Contemporary theorists of genre whodeal with media texts face a more complex sit-uation. Throughout the media system, and par-ticularly in the area of film and televisionfiction, there is a multitude of genre categories,and a constant proliferation of new genres andsubgenres (on film genres see Altman 1999;Grant 1986, 1995; Kaminsky and Mahan1985; Neale 1980, 2000; on television genressee Newcomb 1974; Rose 1985).

The media industry itself uses genres as con-venient labels in production planning and in themarketing of new products. From the audi-ence’s point of view, references to genre func-tion as appetizers that suggest the type ofinterpretive frame which should be applied toa particular text. Genre references, then, createcertain expectations which are based on theaudience’s prior experience with similar mediaproducts. As such, a genre may be regarded as a kind of ‘contract’ between the media industry and its audience. More specifically, itamounts to an agreement between sender andreceiver about some basic features of a certaintype of textual product which is designed toperform a particular cultural task.

In sum, a genre is a system of textual con-ventions, or – to use semiotic terminology – akind of latent langue governing the productionof individual instances of parole. Because of thecontractual nature of the relation betweensender and receiver, it can also be regarded asa kind of mental tool, provided by the mediaand feeding into cultural processes. Mediaaudiences use such tools to interpret the worldand to address certain recurring socioculturalissues within familiar formats.

Practical and theoretical genres

In some cases, a ‘genre’ refers to a group oftexts which is classified differently by the mediaindustry and by portions of the audience. Theprime example is the term film noir, which wasintroduced by French film critics after World

Mediated fiction132

narrativehierarchy

invisiblenarration

cameranarration

drama, epic,poetry

genre ascontractbetweensender andreceiver

War II to refer to a group of Hollywood filmsthat the industry was marketing as crime storiesor melodramas (Borde and Chaumeton 1955).By using the term film noir, the film critics sug-gested that the films in question were somehowunited across the established industrial genrecategories (the history of the film noir conceptis discussed in Cook 1999).

Film noir has been called a ‘theoretical’genre as opposed to the ‘practical’ genresemployed by the producers and consumers ofmedia texts. The distinction calls attention tothe fact that any given genre term usuallycovers two different phenomena. On the onehand, a genre is part of the practical, almostunconscious knowledge� about textual fea-tures and classes which members of the audi-ence draw upon in their everyday lives. On theother hand, it is also a theoretical and analyti-cal object, or rather it becomes such an objectthe moment people start discussing whichtextual features define a particular group oftexts, or how this group differs from othergroups (see Todorov 1978). Film noir is one ofthe relatively rare examples of a classificationwhich started out as a theoretical genre, wasaccepted by the industry and the audience, andended up being a practical genre.�

The theoretical definition of genres is noeasy task. The history of genre theory showsthat it is, in fact, quite impossible to classifytexts on the basis of simple, well-defined fea-tures, or to make genres form a clear-cut,unambiguous system. One solution is to rely onthe concept of ‘family resemblances,’ as definedby Wittgenstein (1953). While it is quite rarefor a few common features to constitute anentire genre, a group of texts may neverthelessbe closely related because of the partial andincremental overlaps between its various sub-groups – just as individual children within afamily may have features in common with boththeir father and their mother without resem-bling each other. Genres are, indeed, definedwith reference to a common pool of textual fea-tures, but these features may not all be shared

by all members of the group. In this view, agenre amounts to a series of transitions or dis-placements between texts, an aspect whichmakes the concept particularly useful for thedescription and analysis of the historical devel-opment of textual forms. The individual genremay be regarded as a mobile textual field inwhich new subcategories or branches are con-stantly being developed.

When texts are used recurrently to performmore or less similar cultural and social func-tions, prototypical forms develop almost auto-matically. And, as these recurrent discursivefeatures are institutionalized and codified, indi-vidual texts will eventually be produced andreceived in accordance with the codified norms.According to Todorov (1978), a genre is pre-cisely such a codification of discursive charac-teristics, i.e., a group of texts that serve acommon purpose, and which therefore havecertain discursive qualities in common. Still, asimplied by the concept of family resemblances,these characteristics may be associated withmany different textual levels and aspects.

Textual characteristics of genres

The genre label suggests an interpretive frameto the audience as well as to the analyst. Toanalyze a particular film as a western or as afilm noir is to regard it as an example of ageneral type. It is also to identify a series oftextual features which this particular film hasin common with other films, and which may berelevant for the analysis. Another reference toThe Big Sleep illustrates how, and further sug-gests the basic textual aspects at work in genredefinitions (on these aspects see furtherTodorov 1978).

In reference works as well as in popularmovie guides, The Big Sleep is usually labeleda mystery movie (but also sometimes a thriller,a gangster film, or a crime film (see, e.g., Cook1999; Halliwell 1999). The genre label pointsout that this particular film belongs to a largergroup of films which, despite other differences,have a common core: they present a mystery,there is a crime which is to be solved – usually,but not always, by a private investigator or apolice officer. This kind of genre definition

Genre analysis 133

1111234567891012341115678920123456789301234567894012345678950111

film noir

� practical consciousness – Chapter 1, p. 2

� practical and other ‘theory’ – Chapter 16, p. 274

familyresemblances

focuses on ‘content,’ or what linguists mightcall ‘semantic’ aspects of the texts in question.Semantic definitions focus on signifieds (Whatis the text about? What kind of ‘themes’ or‘motifs’ does it deal with?, etc.). In addition,they may raise questions of referentiality (Is the textual universe presented as fictional orfactual? What kind of sociohistorical contextdoes the text refer to?, etc.).

The Big Sleep has the semantic features nec-essary for being labeled a mystery, but it is alsoa mystery movie. The ‘movie’ part of the labelrefers to what Todorov (1978) calls the ‘verbal’aspect of the text, its ‘material’ form. First ofall, there are the basic material conditions ofcommunication: What kind of discursive sub-stance are we dealing with? In the case of TheBig Sleep, the mystery is not presented in any simple, homogeneous discourse (as, forexample, in Chandler’s original detectivenovel), but in a complex sequence of movingimages, dialogue, written texts, music, and soon. Second, and more importantly, the materialaspect concerns questions of ‘style’ or ‘mise-en-scène’ – the way in which events are staged forthe camera in terms of setting, lighting, cos-tumes, and character behavior (see Bordwelland Thompson 1997). To say that a Hollywoodfilm from the 1940s is a mystery movie almostautomatically suggests a fixed stylistic pattern:images in black and white, usually an urbansetting, a certain type of lighting with a prefer-ence for night scenes, rainy streets and dramaticshadows, and so on.

Finally, there is the ‘syntactic’ aspect – theactual structuration of the text, its sequentialorganization or composition. In the case ofmystery movies like The Big Sleep, the centralcharacteristics follow from the relation betweensyuzhet and fabula, as already described: fabulaevents are presented in reverse chronologicalorder in the syuzhet; the film presents murdersand mysteries, but withholds information abouttheir causes until the very last moment.

To sum up, analyzing The Big Sleep as amystery movie implies that special attentionwill be paid to the narrative construction(semantically: What type of mystery is pre-sented here? syntactically: How is it solved?What kind of ideological values are involved?,

etc.) as well as to its material aspect, or the styl-istic presentation of the narrative. It remains toemphasize again, however, that genres are notfixed classes with precise borderlines. The BigSleep is an example that one film may very wellbelong to several genres at once. It shares deci-sive textual features with a number of othermystery films, but it is also one of those filmswhich can be, and usually is, labeled film noir.If one were to analyze it as a film noir onewould cover much of the same ground, but theanalysis would focus more on the ‘dark’ aspectsof the story, the violence, the cynical and disil-lusioned detective, the dangerous femme fatale,and so on – aspects which all play an import-ant role in the definition of the film noir.

The configuration of semantic, material, andsyntactical aspects, noted above, highlights anumber of textual features. Moreover, sincegenres are established and institutionalizedaccording to the functions of texts in fixed,recurring situations, they also always have a‘pragmatic’ aspect. What are the demands of the situation on the discursive character-istics of a given genre? What is the intentiongoverning the production of the texts? What istheir purpose? Even though this fourth aspectof a definition of genre primarily concerns thesocial context, the pragmatic functions which atext has to fulfill within a given context obvi-ously affect the details of its discursive con-struction. In this regard, one particularlyimportant issue is the way in which the textaddresses its audience. As indicated by the briefanalysis of television news, this is one textualgenre in which a central pragmatic question –the interpretive authority of different narrators– is solved by means of a fixed narrationalsystem.

HETEROGENEOUS TEXTS

A film is not only a sequence of moving images,but an organized mixture of images, words,texts, music, and noises. Most media texts arelike this: montages, or heterogeneous construc-tions that are characterized by a constant dis-placement and circulation of meaning. In filmand television, in newspapers and magazines,information is doubled or tripled, presented

Mediated fiction134

semanticdefinition of

genre

materialdefinition of genre

mise-en-scène

syntacticdefinition of

genre

pragmaticdefinition ofgenre

mode ofaddress

through several channels, discourses, and sensesat once.

This complexity raises a series of theoreticaland analytical issues which are often neglectedin media studies. Like much earlier film theory,Christian Metz’s semiotic works in the 1960swere primarily concerned with sequences ofshots. The focus of many current film studies is primarily on the fabula – the narrative asconstructed by a spectator after the fact.Journalism studies frequently describe news-paper stories and even television news as merelyverbal phenomena.

Similarities and differences

Although the need for studies of the hetero-geneity of the modern media was acknowledgedby the semioticians of the 1960s, the problemproved difficult to handle within a theoreticalframework which takes the homogeneousverbal text as its model. In his authoritativeintroduction to semiotics, Barthes (1964a)simply stated that semioticians, like linguists,must work primarily with homogeneous mat-erials. Even if he saw the need for descriptionsof, for example, films or fashion magazines as complex structural totalities, he wanted topostpone the analysis of such heterogeneousmaterials and to concentrate on studies of indi-vidual, homogeneous substances.

Nevertheless, in his influential essay on ‘therhetoric of the image,’ Barthes did suggest somebasic relations between two such substances,namely image and text. According to Barthes,a text can function as an ‘anchorage’ of visualsignifieds, either by identifying the objects rep-resented by the image (e.g., a caption for a newsphoto), or by suggesting and authorizing howthe image should be interpreted (e.g., the titleof a painting). The ‘anchorage’ relation is basedon redundancy: the text repeats or explains theinformation given by the image. In contrast, theother basic relation, ‘relay,’ operates on differ-ences. In this relation, text and image are com-plementary, and the unity of the message isrealized at a higher level. Barthes’s ownexample was the relation between dialogue andimages in a film. Here, text and image carry different signifieds; the text adds meanings to

the narrative which are not found in thesequence of images (Barthes 1964b).�

With this simple outline, particularly theconcept of relay, Barthes hinted at ways of ana-lyzing not only text–image relationships, butthe total mix of substances which are charac-teristic of films and television programs. Invarious versions, this notion of complementar-ity has been a key issue since the 1980s instudies trying to account for the heterogeneityof media materials.

Studies of sound

Most of these discussions have focused on thestatus of sound and its functions in film. In1980, in a seminal issue of Yale French Studieson cinema and sound, Rick Altman (1980) crit-icized traditional film theory as well as contem-porary film semiotics for their focus on thecinematic image and their correspondingneglect of sound. Here and in later writings, heargued that sound is just as important as – andin certain respects perhaps even more importantthan – images for the total cinematic experience(e.g., Altman 1992). From a similar position,French sound theorist Michel Chion has arguedin a series of books that most film theory is char-acterized by ‘vococentrism.’ In so far as filmsound is acknowledged at all, it is usually thehuman voice of the dialogue, or the voice-over, which is privileged analytically in favor of,for instance, music or noise. Chion presents detailed analyses of the importance of sound infilm and television with particular emphasis onthe ‘point-of-hearing,’ i.e., the effect which isobtained by the positioning of the sound sourcein relation to the point of view of the camera(Chion 1999 [1982], 1985, 1988, 1998).

In their justified criticism of the dominantfocus on visuality in previous film and mediastudies, Altman and Chion sometimes placethemselves at the opposite extreme, claimingthe superiority of sound in relation to images.Most later works in this area emphasize thecomplementarity of sound and image. This, forexample, is the position of Sarah Kozloff in her

Heterogeneous texts 135

1111234567891012341115678920123456789301234567894012345678950111

� anchorage and relay, with example analysis – Chapter7, p. 111

point-of-hearing

studies of the various ways in which voice-overand dialogue are used in film (Kozloff 1988,2000). It is also the dominant position atpresent in studies of music and its functions in film.

The role of music

Traditional film studies tended to distinguishbetween two types of music, one supporting theimage, the other working contrapuntal to theimage (e.g., Eisler and Adorno 1947). Arguingthat music should be described more directlyfor its relations with the narrative and itsevents, Chion (1985, 1994 [1990], 1995) sug-gests three basic categories:

1 empathetic music, which supports andexpresses the emotions of the characters;

2 a-empathetic music, which is independent ofthe actions on the screen;

3 didactic music, which is employed as a dis-tanciating, often ironical commentary onthe action.

In concrete studies of music in film and tele-vision, a more differentiated analytical frame-work is needed. As in narrative studies andgenre studies, classical Hollywood cinema hasalso been the central case in studies of filmmusic. One example is Claudia Gorbman(1987), who takes the practice of classicalHollywood as the basis for a general descriptionof narrative film music. Drawing on analyses of,among others, Bernard Herrmann’s score forHangover Square (John Brahm, 1944) and MaxSteiner’s score for Mildred Pierce (MichaelCurtiz, 1945), she argues that most film musicperforms one of four basic functions:

1 a signifier of emotion, i.e., the music setsspecific moods and emphasizes particularemotions suggested by the images;

2 formal and rhythmic continuity betweenshots and in transitions between scenes;

3 narrative cues or interpretations of narrativeevents;

4 formal and narrative unity may be con-structed by means of the repetition andvariation of musical material and instru-mentation.

Most of these functions are examples ofwhat Barthes called anchorage: the music sup-ports the meaning of the images (for a furtherdiscussion of musical anchorage see Larsen2000). There are, however, also examples offilm music functioning as relay, i.e., workingcomplementary to the images and to the dia-logue, and adding qualities of its own to thenarrative. This aspect has been underlined byKathryn Kalinak (1992), who, in a studysimilar to that by Gorbman, analyzes MaxSteiner’s score for The Informer (John Ford,1935) and David Raksin’s score for Laura(Otto Preminger, 1944) as examples of theclassic Hollywood style. She emphasizes thecentral composers’ reworking of the musicalpractices of late Romanticism, in particulartheir use of leitmotifs – short musical themesfunctioning as kinds of signposts announcingthe arrival of characters, or referring to priorevents or central narrative issues. In an analy-sis of John Williams’s score for The EmpireStrikes Back (Irvin Kershner, 1980), Kalinakfurther shows that this classic Hollywood prac-tice has been continued in later productions.

Sound in television

The turn toward sound in film studies duringthe 1980s was, if not prompted, at least rein-forced by the remarkable technological devel-opments of the period. The introduction of theDolby noise reduction system in 1977, the suc-cessive generations of Dolby stereo and sur-round sound, the new multitrack recordingsystems, and so on greatly improved the qualityof cinema sound, and also led to a series ofexperiments with sound by both avant-gardeand mainstream directors. Although televisionhas gone through almost identical technologi-cal developments during the same period,resulting in a similarly increased emphasis onsound communication in the medium as awhole as well as in individual programs andgenres, there has been no corresponding focuson sound in television studies.

The role of music is sometimes discussed instudies of major television genres (see, e.g., onmusic in soap operas, Gripsrud 1995: 183f.).The establishment of international promotional

Mediated fiction136

leitmotifs

channels like MTV during the 1980s promptedsome studies of the music–image relationship inmusic videos (e.g., Kaplan 1988; Larsen 1989).In recent years, there have been additionalstudies of music in relation to multimediapresentations. One example is musicologistNicholas Cook’s (1998) analyses of the ways inwhich music, words, moving pictures, anddance work together in television commercialsand music videos.

On the whole, however, there have beenremarkably few contributions in this area.

Regarding television sound in general, the onlymajor work is a short article from the mid-1980s in which Rick Altman discussed thefunction of sound in relation to the domesticviewing situation (Altman 1986). Obviously,this is an area of media studies in which furtherresearch is greatly needed, not only because ofthe centrality of sound in the media, butbecause such research may contribute to abetter understanding of sound as ‘text,’ in itsown right and in conjunction with other mediatexts.

Heterogeneous texts 137

1111234567891012341115678920123456789301234567894012345678950111

If any one issue can be said to have motivatedmedia studies, it is the question of ‘effects.’From the perspective of policy-makers and thegeneral public, the field has been expected tosupply evidence of what the media may do topeople and to society. From within the acad-emic perspective, the field has justified itself byexamining what specific difference the modernmedia make, compared to other cultural formsand social institutions. The question of effectshas largely been stated in terms of the relativelyshort-term cognitive and behavioral impacts ofdifferent media and their contents on massaudiences, which have been studied by quanti-tative social-scientific methodologies. Especiallyin this area, it is still appropriate to speak of adominant paradigm or model (Gitlin 1978;Webster and Phalen 1997), even if the quanti-tative mainstream is quite differentiated andcurrently in dialogue with a qualitative sub-stream (Chapter 10).

This chapter reviews the main varieties ofaudience studies with reference to their foci onparticular stages of the communicative process.While contemporary research commonly dis-

tanciates itself from a chain-like model of theprocess, it is the case that each tradition isdefined, in part, by its orientation toward a par-ticular moment of the interchange betweenmedia and audiences, whether in the short orthe long term. In addition to being an intu-itively helpful configuration of the empiricalfield, this implicit model holds the potentialboth for specifying the scope of each tradition,and for indicating areas of contact betweenthem. To anticipate one argument of the chap-ter, each tradition may be said to identify asocial ‘context’ of interaction between mediaand audiences. For example, an ‘early’ contextof interaction is examined in terms of nationaland international television ratings and othermeasures of media exposure; a ‘late’ contexthas been studied with reference to the ques-tion of whether media contribute over time toclosing or deepening so-called knowledge gaps in society (Tichenor et al. 1970). In eachof these two examples, an interchange occursbetween audience and medium which has impli-cations for audience members’ actions in othersocial contexts: the act of television viewing is

Media effectsQuantitative traditions

Klaus Bruhn Jensen

9

• a brief history of research on media effects• a review of the main traditions of audience research, focusing on different stages of the process of

mediated communication• a presentation of additional studies on how media serve to socialize individuals and to institutionalize

society• an example of the study of lifestyles as they relate to media use.

‘effects’ –traditionally

defined ascognitive and

behavioralimpacts on

individuals inthe short

term

the dominantparadigm

a way of giving priority to one form of culturalactivity over others; the command of certainforms of public knowledge is one necessarycondition for political participation.

Figure 9.1 lays out the main stages asdefined by different research traditions. Beforeexamining these several moments of mediatedcommunication in society, this chapter brieflynotes the controversial history of effectsresearch. The following sections describe andexemplify each research tradition, with specialreference to its theoretical assumptions and pre-ferred methodologies. A more detailed accountis given of research on lifestyles, which hasrecently attracted attention in media studies,and which is one of the links between the quan-titative mainstream and the qualitative sub-stream.

HISTORIES OF EFFECTS RESEARCH

Because of the sustained social and scientificinterest in effects, this area has more of an iden-tifiable, if contested history than most othersubspecialties. As summarized by McQuail(2000: 417–22), four phases may be singled out:

• Phase 1: All-powerful media. From about1900, when an identifiable research specializa-tion was emerging (e.g., Bryce 1966 [1900]), to1940, the media were widely thought to be able

to shape both opinion and behavior throughpropaganda (e.g., Lasswell 1938). However,‘this view was based not on scientific investi-gation but on observation of the enormouspopularity of the press and of the new mediaof film and radio’ (McQuail 2000: 417).

• Phase 2: Theory of powerful media put tothe test. From the beginnings of academicmedia research in the 1930s, studies began tosuggest that at least no direct link could beestablished between media stimulus and audi-ence response. Klapper (1960) provided aninfluential digest of findings which seemed tosupport this position.

• Phase 3: Powerful media rediscovered. Inpart through a reassessment of the same evi-dence (see Chaffee and Hochheimer 1985;Delia 1987), the 1960s witnessed a return tohypotheses about media power. Such hypothe-ses were also fueled partly by the arrival of tele-vision, partly by research on various cognitiveand structural impacts, over and above effectson individuals’ attitudes, and partly by therediscovery of critical social theory.

• Phase 4: Negotiated media influence. Sincearound 1980, a turn to understanding media asnecessarily constructing meanings, with impli-cations for social structure as well as individ-ual agency, has been in evidence. While partly

Histories of effects research 139

1111234567891012341115678920123456789301234567894012345678950111

Gratificationssought

Diffusion ofinnovations

Contexts ofmedia use

Media insocial action

Gratificationsobtained

ConsumptionRecall

Framing

Socialization

Cultivation

Lifestyles

Agenda-setting

Campaigns

Institutionalization

Decodings

MEDIA SOCIETY

Knowledgegaps

Figure 9.1 The ‘stages’ of mediated communication, as defined by audience research traditionsNote: The qualitative traditions, indicated below the line, are reviewed in Chapter 10, this volume

associated with qualitative reception studies,the shift can also be traced in several of thequantitative traditions reviewed below.

As the periodization suggests, the develop-ment of research, and the definition of ‘effects,’has been intimately related to contemporaryconcerns, including two world wars and the riseof television as a vehicle of popular culture. Inthe process, researchers themselves have alsoheld changing notions of their own social andintellectual role. Moreover, it should be notedthat both this periodization and the definitionof ‘milestone’ studies (Lowery and DeFleur1995) enter into a received history of the field,which is as informative of current ideas aboutthe mission of media effects research as it is ofpast findings (see Figure 10.1).

One of the abiding issues has been the sociallevel at which effects should be conceptual-ized and studied. Overwhelmingly, previousresearch has focused on the individual user’slevel, despite exceptions such as studies on thestructural dependency of different social groupson media (Ball-Rokeach 1985). This chapter,while retracing the individualist mainstream,returns whenever relevant to technologies, insti-tutions, and genres as preparatory conditionsof mediated communication and its impact.�In this respect, the empirical domain of effectsresearch shades into the domains of both‘medium theory’ (Chapter 2) and of research onthe historical and cultural contexts of commu-nication (Chapters 11–12).

MOMENTS OF IMPACT: FROM DIFFUSIONTO CULTIVATION

Diffusion of innovations

One necessary condition of mediated commu-nication on any social scale is the availabilityand accessibility of media as cultural resourcesfor a significant portion of the population. Onthe one hand, the material availability of a tech-nology presupposes its invention, development,distribution, and relative affordability in the

historical context. On the other hand, its dis-cursive accessibility requires both that its poten-tial social uses are perceived as relevant orattractive, and that audiences have the corres-ponding cognitive and cultural competences.This complex social dynamic, in certain res-pects reiterating the development and spread ofliteracy, has been studied in the media fieldunder the heading of ‘diffusion of innovations.’

It is important to note that this research tradition has been quite inclusive in studyingnot only the diffusion of new media technolo-gies, but perhaps more typically the diffusionof information via media. An important back-ground was postwar attempts by Westernnations at modernizing health, education, andother social sectors in developing countries(e.g., Lerner 1958). The central theoreticalmodel was advanced by Rogers (1962), whodescribed a rather linear and hierarchical pro-cess of information diffusion, concerning ideasas well as products. In this process, the mediawere found to be important especially in the first stage of information and awareness(followed by stages of persuasion, decision oradoption, and confirmation). Importantly, per-sonal contacts and concrete experience could beshown to replace technological media as themain vehicles of diffusion in later stages.� Inaddition, studies of news diffusion suggestedthat the higher the proportion of a populationthat knows about some event or issue, thehigher the proportion that learned about thesefrom an interpersonal source (Greenberg 1964).

Later work on development communica-tion� as well as on the introduction of newmedia has moved away from a paternalist andlogistical conception of cultural change, andtoward a recognition of cultural specificity as acondition of how technologies may be imple-mented (Rogers 1986, 2000). A wider implica-tion for communication theory has been thattechnological media are more than engineeringsolutions, and always come with a ‘message’whose interpretation will depend on historicaland social circumstances. With this revised

Media effects140

� structures and conditions as sources of ‘effects’ –Chapter 2, p. 16

materialavailability

discursiveaccessibility

� interpersonal communication – Chapter 1, p. 8

� development communication – Chapter 11, p. 179

agenda, diffusion research has remained rele-vant for studies on how media are adopted invarious cultural settings, and on which seg-ments of a community or society are likely tobe ‘early’ or ‘late’ adopters (see Figure 9.2). Foreach new medium, a recurring concern has been the extent to which media can reinforce,rather than counteract, social inequalities. Thismay be either because the ‘ideal’ S-curve (froma slow and unequal start, via an accelerateddiffusion, to a plateau where the medium ismaterially available to practically everybody)does not occur, or because some audiences withlimited cognitive or cultural competences findthe media less accessible. Both conditions mayresult in ‘knowledge gaps,’ as addressed below.

The most common approach to diffusionstudies has been survey methodologies, which

make it possible to compare and contrast thebasic availability and accessibility of eithermedia or information in different social groupsand contexts. National statistical abstracts aswell as continuous market studies are amongthe sources of evidence, which also lendthemselves to ‘secondary analyses’ that extractadditional patterns of diffusion and use fromexisting datasets. Perhaps partly for thismethodological reason, diffusion research hasgiven less attention to the interpretive aspect ofsuch social processes, or what has been calledsymbolic diffusion (Jensen 1993a). Similarly,questions of material availability have beengiven priority over discursive accessibility, in thesense that generic and other cultural conven-tions as well as audiences’ pre-understandings –their ‘literacy’ in a broad sense (Messaris 1994)

Moments of impact: from diffusion to cultivation 141

1111234567891012341115678920123456789301234567894012345678950111

100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

0%

Ado

ptio

n of

tele

visi

on b

y U

.S. h

ouse

hold

s

1948 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972

Highest quarterof U.S. households

in income

Lowest quarterof U.S. households

in income

Adoption of black andwhite television byall U.S. households

Adoption of colortelevision by

all U.S. households

58%

42%

32%

24%

7%

GAP

97%

3%

6%

12%

18%

2%1%

23%

27%

37%

49%

32%

18%

91%

68%

59%

50%

43%

33%

25%

17%

9%

5%3%

Figure 9.2 Examples of differential rates of the diffusion of innovations, in percent over timeSource: Rogers 1986: 171. Reprinted with permission of The Free Press, a division of Simon & Schuster, Inc., © 1986 The Free Press.

symbolicdiffusion

– remain under-researched, yet important con-ditions of how new forms of communication are diffused.

Gratifications sought

Audiences’ orientation toward media has beenthe main focus of the second tradition, ‘usesand gratifications’ research, so much so that itsposition is frequently summarized by asking,not what media do to audiences, but what audi-ences do with media (e.g., Halloran 1970).Following pioneering studies in the 1940s(Berelson 1949; Herzog 1944), the traditionwas rediscovered and redeveloped from the1960s, partly in response to the contemporaryconclusion concerning ‘limited effects.’ Theresulting research program articulated anambition of relating the impact of media toboth psychological needs and social conditions,including the alternatives to media use. Thatprogram may be read as a restatement of thechain of communication, referring to ‘(1) thesocial and psychological origins of (2) needswhich generate (3) expectations of (4) the massmedia or other sources which lead to (5)differential exposure (or engaging in otheractivities), resulting in (6) need gratifica-tion and (7) other consequences’ (Katz et al.1974: 20).

Uses and gratifications research becamewidely influential, at least into the 1980s(Rosengren et al. 1985), on several counts. Forone thing, studies began to outline generaltypologies of the media–audience nexus. Theconcept of ‘functions,’ familiar in sociologyfrom Durkheim to Parsons, lent itself toaccounts of the relationship between humanneeds, media provisions, and the gratificationsgenerated at their interface. A number of studiesconverged on three main functions of media,namely information-seeking, diversion, and themaintenance of personal identity (see Blumler1979; McQuail et al. 1972). In addition,research addressed the complicated, yet crucialquestion of whether and how the public per-ceives different media and genres as providers ofdistinctive gratifications (see Katz et al. 1973).In certain respects, such studies are comparableto opinion polls, conducted regularly in many

countries, on public perceptions of variousmedia.

The gratifications tradition soon came in fortheoretical criticism from at least two differentquarters. On the one hand, critical theoristsargued that the approach was compromised by its functionalist premises, specifically thepresumption that audience uses of media mightbe examined in relation to a ‘common interest’and an actual or emerging social consensus(e.g., Elliott 1974). Despite references in theliterature to ‘dysfunctions’ of media (Lazarsfeldand Merton 1960 [1948]; Wright 1959), thesetended to be conceived as deviations from anorm rather than indicators of any structuralconflict. On the other hand, scholars of a cul-turalist bent noted that the entire processualand participatory aspect of communicationseemed to fall outside this framework (e.g.,Carey and Kreiling 1974). In addition, empiri-cal research questioned the status and explana-tory value of the key notion of gratifications.Lichtenstein and Rosenfeld (1983), for in-stance, found that heavy and light viewers, aswell as fans and non-fans, tended to agree ontheir gratifications, and concluded that audi-ences at large may reproduce a generallyaccepted image of both content formats andspecific media.

Later work has responded to some of thesecriticisms, and has elaborated the uses and grat-ifications tradition in several respects. Oneresponse has been attempts to integrate gratifi-cations into more general models of communi-cation and media choice (McQuail 2000: 393).Another response has been to explore linksbetween personality and media gratifications(Conway and Rubin 1991). Others again havedifferentiated theoretically between studies ofaudience activity before, during, and aftermedia exposure, which may substantiate ratherdifferent conclusions about the nature of mediause, and which may require different method-ologies (Levy and Windahl 1985). However,despite the occasional experimental or qualita-tive study, the tradition has relied primarily onsurvey techniques to examine the audienceexperience of media as a whole. Thus, whilegratifications research has documented and sys-tematized public perceptions of various media

Media effects142

whataudiences do

with media

functions ofmedia use

publicperceptions of media and genres

and genres in general categories, partly in com-parison with other cultural forms, it has provenless applicable to studying specific audienceresponses or contextual uses of media and theircontents. (One attempt in this last regard istaken up below in the section that distinguishesbetween gratifications sought and obtained.)

Consumption

The bread and butter of media research as amodern institution of social surveillance andcontrol (Beniger 1986) have always been mea-surements which answer questions of whoattends to which media, and to what portionof their contents, when, and for how long. Inaddition to the commercial value of such mea-surements, they document the moment of‘exposure,’ or the concrete interaction betweena medium and an audience, without whichthere would be no communication, interpreta-tion, or effects. While the term ‘consumption’might suggest either an applied, marketingapproach or a critique of the economic func-tions of media, here it refers inclusively to thefact that audiences do spend time and, directlyor indirectly, money on media.

The time and money spent, by varioussociodemographic groups, on different media,genres, and texts, are of interest as necessarybaseline information, even while both the cat-egories of measurement and the social uses offindings will be subject to epistemological aswell as political controversy. For example,Smythe (1977) has argued that the primaryproduct of media organizations is not contents,but audiences – media deliver audience atten-tion to advertisers, or they construct an appar-ently attractive role of ‘being an audience,’ forwhich people are willing to pay money.

Just as the conceptualization of ‘audiences’has remained controversial, the definition oftheir ‘attention’ to media has proven a stum-bling-block to research, in very practical terms.For one thing, audience attention to media istypically discontinuous, either because events inthe immediate context disrupt attention, orbecause the media themselves fail to hold theaudience’s attention, for instance, through acommercial break or a full newspaper article.

For another thing, media use, such as televisionviewing, is often a secondary activity in rela-tion to some primary activity in the household(e.g., Robinson and Converse 1972). Indeed,these activities may be so thoroughly integratedthat it is difficult to distinguish and rank them,for both researchers and respondents.

Partly because of such methodological diffi-culties, audience research has gone throughseveral ‘generations’ of techniques (for anoverview see Gunter 2000). In addition to theincreasing reliance on computers, not just indata analysis, but also in the process of col-lecting and organizing data (e.g., ComputerAssisted Telephone Interviewing (CATI)), themovement has been from in-person techniquessuch as diaries (and interviews), toward semi-automatic techniques (or a combination oftechniques), particularly for television. Fromthe mid-1980s, ‘peoplemeters’� have been usedto go beyond measures of whether the televi-sion set is turned on (and to which channel itis tuned), and have relied on respondents topush a button to identify their presence as‘viewers’ and, in some systems, to indicate theirlevel of appreciation of the program. On asmaller scale, studies have videotaped audiencesin front of the set to assess their presence andattention (see Bechtel et al. 1972; Borzekowskiand Robinson 1999). Other studies haveemployed eye-tracking techniques to determinewhat readers look at, and in what sequence(e.g., Thorson 1994) – techniques which arealso applied to computer media. Also, the useof remote control devices has been examinedby electronic monitoring (Kaye and Sapolsky1997). For all the techniques mentioned, how-ever, there are trade-offs that depend on thepurpose of study, between, for example, thecomprehensiveness and potential intrusivenessof peoplemeter methodologies.

Because of the focus on individuals’ expo-sure to media, particularly in commercialresearch, less emphasis has been placed on ques-tions of ‘where’ and especially ‘with whom’media are used. Whereas qualitative studieshave highlighted the social contexts of mediause (Chapter 10), some quantitative research


1111234567891012341115678920123456789301234567894012345678950111

exposure

audience asproduct

secondarymedia use

� TV- and peoplemeters – Chapter 13, p. 228

viewing theviewers

eye-tracking

has also addressed related questions. For exam-ple, studies of television have traced its impacton family life, and concluded already in the1970s that ‘multiple sets disrupt the establishedpattern of family interaction’ (Comstock 1978:158). Another characteristic of quantitativeaudience research more generally is that com-paratively little empirical attention has beengiven to the ‘what’ of communication – the cul-tural and discursive forms which give rise to theadoption, gratifications, and effects of media.While this helps to explain a frequent com-plaint, also within media industries, that audi-ence measurements fall short of being audienceresearch (Wober 1981: 410), still, exposure isthe sine qua non of communication and oneinevitable focus of media research.

Gratifications obtained

As noted above gratifications research has beencriticized for a functionalist notion of commu-nication, but also on its own terms, because ‘ithas not provided much successful prediction or causal explanation of media choice and use’ (McQuail 2000: 389). That may be due,first, to the contextual, even accidental natureof much media use, which is difficult to accountfor in terms of motivations and plans. Second,the comparative failure of gratifications re-search has entailed a questioning of the pre-ferred survey methodologies of the tradition,which are especially appropriate for examiningconscious, unambiguous, and relatively uncon-troversial aspects of media use.

Nevertheless, in order to capture the evidentgratifications of media – pleasurable experi-ences and valued information – gratificationsresearchers have proposed to distinguish expec-tations of what will be provided (gratificationssought – GS) from the resulting satisfactions(gratifications obtained – GO). Importantly, GS is, in part, the outcome of past experienceswith particular media (genres, texts), and GO amounts to a feedback concerning whichgratifications to seek in the future, and from which media. This processual perspective ongratifications has been summed up in an ‘expectancy value’ model (Palmgreen andRayburn 1985).

In one study, Palmgreen et al. (1980)solicited responses to two sets of statements,presented in telephone interviews immediatelyafter each other. GS was operationalized as, ‘Iwatch television news to keep up with currentissues and events.’ GO, next, was restated withreference to the program that the respondentwould ordinarily watch, thus ‘CBS News helpsme to keep up with current issues and events’(p. 171). Whatever the correlations between thetwo sets of measures, however, the empiricaldesign does not warrant conclusions about the viewer’s satisfactions or experiences in anyspecific sense. The GO and GS statements werepresented in the same interview session, inequally abstract formulations, and, most im-portant, there was no concrete point of refer-ence in an actual program or context ofviewing. Instead, the findings may support apublic-opinion profile of the particular newsprogram or organization in question. In addi-tion, such conceptual and methodological diffi-culties have made the gratifications traditionless influential since the 1980s.

An alternative approach – the ‘experiencesampling method’ – has been applied by Kubeyand Csikszentmihalyi (1990) to television. Thebasic idea is that respondents are contactedwith a paging device at random times, and areasked each time to complete self-reports onwhat they are doing, and how they feel. Thesedata may then be used to explore, for instance,the relationship between types of media use,simultaneous activities, and the respondent’smoods and other mental states. While theapproach requires considerable resources, itoffers one means of securing immediate feed-back from media-users about their experiencesin relatively naturalistic circumstances.

Recall

Parallel to respondents’ statements about theirconscious motivations for, and benefits from,media use, a favorite measure of audienceresearch has been recall, i.e., respondents’ability to reproduce items of information within a relatively short time span after mediaexposure. This aspect of media use has beenexamined especially through experiments and

Media effects144

consumptionof what,

with whom

experiencesamplingmethod

surveys (for overviews see Gunter 2000:215–225; Robinson and Levy 1986).

Basic recall is of interest as a preparatorycondition of other forms of impact, fromfactual learning and education to informedpolitical participation. Studies have normallyfound that recall is positively correlated withthe recipients’ level of prior knowledge (andhence their social status), in addition to the per-ceived concrete relevance of the information.Furthermore, recall studies have offered anopportunity to probe the specificity and ‘effi-ciency’ of different media, including multi-modal communication, not least the visualcomponent of television news. Research in this area tends to suggest that print and audio-visual media do not in themselves constitute ahierarchy of more or less efficient vehicles ofcommunication, but rather that narrative andmessage structure in general, and the meaning-ful integration of verbal and visual elements inparticular, will determine recall and potentiallearning by audiences.

Within the recall stage of communicationand impact, it is possible to identify at least twodifferent strands of research. First, some studieshave departed from the premise that mediahave a significant potential as sources ofenlightenment for the general public. Not sur-prisingly, much of this work has been associ-ated with European public-service traditions ofbroadcasting, and was, to a degree, conceivedas a means of product development in thepublic interest (e.g., Findahl 1985; Gunter1987b). Partly for this reason, studies wouldcommonly analyze programs closely in con-junction with the audience response, thus antic-ipating and complementing qualitativereception studies.

The second body of research has beenmotivated by commercial aims of establishingwhether consumers recall or recognize particu-lar commodities or brands. As such, it entersinto the wider applied field of marketingresearch (e.g., Shimp and Gresham 1983;Thorson 1990). From a more general perspec-tive of theory development, this research hasreiterated some key difficulties in determininghow different stages of communication andimpact are interrelated. In an influential elabo-

ration of the basic stimulus–response model,McGuire (1973) indicated six stages of asequence: presentation, attention, comprehen-sion, yielding, retention, and overt behavior.�In fact, this seemingly logical order is not nec-essarily borne out by empirical research (seeWindahl et al. 1992). For example, consumersmay have no explicit recall of a product, andyet may buy it on sight.

Once again, the preconscious and practical(Giddens 1984) aspects of much media-relatedbehavior present a central methodological chal-lenge, and one which calls for further researchon the nature of the ‘stages’ reviewed in thischapter. The difficulties of ascertaining effectshave repeatedly led to the postulation of addi-tional stages, competing theories, or a reinter-pretation of findings in alternative frameworks.In one early assessment of the finding thatpeople seem to want news, as indicated by their habitual consumption, yet recall little ornothing, Nordenstreng (1972) concluded that‘the main thing retained from the news is thatnothing special has happened’ (p. 390). Recalland other measures of media impact might thusbe reinterpreted, in part, with reference tovarious institutionalized, social uses of media,as audiences deliberate on public issues andmake personal decisions.

Agenda-setting

The research tradition which emphasizes thatthe media set an agenda for public debate anddecision-making, especially within politics, hasfocused not on discrete units of information, astransmitted and recalled, but on ‘issues’ – thoseconfigurations of information that are associ-ated with the activities of particular social insti-tutions, such as parties and parliaments. At thesame time, the agenda-setting tradition hashelped to shift the focus of much audienceresearch. The general move has been away from what used to be the foundational ques-tion at least into the 1960s, namely how themedia may shape and change, for example,people’s attitudes or voting – and toward more general issues of how people process


1111234567891012341115678920123456789301234567894012345678950111

priorknowledge

print andaudiovisual

media

� ’stages’ of stimulus and response – Chapter 3, p. 51

information as part, for instance, of their polit-ical participation (for an overview see Krausand Davis 1976).

This shift is summed up in the conclusionthat the media do not tell people what to think,but what to think about. In American research,this formulation is normally credited to Cohen(1963), but, before him, Trenaman andMcQuail (1961: 178) concluded, ‘The evidencestrongly suggests that people think about whatthey are told but at no level do they think what they are told.’ In addition, an early classicsuch as Lazarsfeld et al. (1944) had noted thatmedia, among other things, serve to structurepolitical issues. It is also worth noting that such earlier suggestions in the literature weregiven new attention from the 1960s, followingthe conclusion that the media had little or noeffect. In this regard, agenda-setting joined grat-ifications research in the attempt to differenti-ate the concept of effects, asking what peopledo with media and what, consequently, theythink about.

Agenda-setting was given its name and itsfoundational work by McCombs and Shaw(1972). In brief, their study in a small NorthCarolina community during the 1968 presi-dential campaign found that although differ-ent news media would represent political lifedifferently, there was a significant correlationbetween issues that were defined as importantby the media and by the voters: ‘voters tend toshare the media’s composite definition of whatis important’ (p. 184).�

Subsequent research has debated and re-examined the approach, both methodologicallyand theoretically. One methodological concernis that both the media and audience sides of theequation are often examined in the aggregate,through surveys and content analysis, but withlittle empirical attention to, for example, vari-ations in the public’s media use or in the fociof news coverage over time. Accordingly, oneambitious program of ‘parallel content analy-sis’ has been outlined by Neuman (1989). Itspurpose would be to examine the ‘content’ ofboth media and public opinion, their thematic

foci and formal articulation, including varia-tions over time. In addition, Lang and Lang(1981) called for more empirical attention notonly to agenda-setting on the audience side, butto ‘agenda-building’ in the interplay betweenmedia and various political agents.

An alternative theoretical explanation forthe correlations identified in agenda-settingresearch follows if one reverses the causality,i.e., if one assumes that issues and prioritiesflow from the public to the media and/or topoliticians in an ideal democratic system. Morerecent work has sought to address the com-plexity of the entire process by distinguishingthree kinds of agendas – those of the media, ofthe public, and of policy (Dearing and Rogers1996). The majority of studies appear toconfirm the impact of media on the publicagenda, whereas the evidence concerning thepolicy agenda is more uncertain (p. 87).

While some reviews conclude that theagenda-setting hypothesis is ‘a plausible butunproven idea’ (McQuail 2000: 456), it remainsa fertile area for further studies. At the theoret-ical level, it has affinities with Newcomb andHirsch’s (1984) proposal that media be studiednot as distribution systems, but as a ‘culturalforum’ to which matters of public interest may be brought, articulated, and contested.Research topics include the short-term manage-ment of agendas and sources by different inter-est groups, and the more long-term adaptationof institutional agendas and logics to shiftingsocial circumstances. Dearing and Rogers(1996: 98) also note that the approach lendsitself to research on other genres than news andpolitical communication.

Despite its distinctive terminology, agenda-setting research can thus be seen to convergewith other work on the role of media in thesocial construction of reality (Berger andLuckmann 1966), and on the media’s ‘framing’of reality, as addressed below. This convergencehas been supported by experimental researchon how the media may perform a ‘priming’ ofthe audience-public as to which issues are thedecisive ones in assessing political candidates(Iyengar and Kinder 1987). A similar directionfor research is indicated by recent surveystudies of so-called second-level agenda-setting,

Media effects146

what to thinkabout

� hypothetico-deductive methodology of agenda-settingresearch – Chapter 15, p. 262

parallelcontentanalysis

three agendas:media, public,policy

media ascultural forum

priming

second-levelagenda-setting

which affects the audience perception of vari-ous detailed attributes of issues, candidates,and so forth (McCombs et al. 2000).

Campaigns

Given the embedding of modern media incomplex and distributed social systems, it ishardly surprising that they should be employed,and studied, as planned means of communica-tion and coordination. In this capacity, mediaare vehicles of ‘campaigns’ to inform, persuade,or mobilize a mass public. Especially in view ofthe market-driven nature of most contemporaryeconomies, it is equally unsurprising that mostcommunication campaigns are of a commercialnature, and have been theorized accord-ingly (for an overview see Windahl et al. 1992)– following early influential work on media asmeans of propaganda during wartime (Hov-

land et al. 1953). Recently, ethical and socialaccountability have become more importantingredients of how corporations and otherorganizations present themselves in public.Campaign studies thus have more general im-plications for the understanding of social com-munication processes. As conducted in severalmedia and over time, campaigns may also beunderstood in discursive terms as instances of‘intertextuality.’

In their overview, Rogers and Storey (1987:821) noted four common features of cam-paigns: ‘(1) a campaign intends to generatespecific outcomes or effects (2) in a relativelylarge number of individuals, (3) usually with-in a specified period of time and (4) through an organized set of communication activities.’From this definition, the authors go on to elaborate several dimensions of a campaign.Figure 9.3 summarizes three main dimensions,


1111234567891012341115678920123456789301234567894012345678950111

propaganda

ethical andsocialaccountability

LEVEL OF OBJECTIVES

Receiver

Sender

LOCUS OF BENEFIT

To inform

Individual Society

To mobilize(promote or prevent behavior change)

LOCUS OF CHANGE(Level of analysis)

Figure 9.3 Dimensions of campaign objectives and effectsSource: Rogers and Storey 1987: 823.With permission of Sage Publications Ltd.

each of which are continua, not dichotomouscategories.

First, the objectives of a campaign may bemore or less ambitious, from informing or per-suading, to mobilizing the audience to act in aparticular way. Importantly, the aim may beeither to promote or to prevent a change ofbehavior, in the latter case, for example, main-taining the audience’s preference for a commer-cial brand. Media planners need to differentiate,then, between communications which are de-signed to introduce, maintain, or reposition aproduct or an idea.

Second, the campaign may be designed toeffect changes at the individual, organizational,institutional, or social levels. Although the con-crete recipients of information will be individ-uals, the communication effort may be focusedon whole segments of a population, such assmokers or car owners (by tobacco and car manufacturers, environmental interest groups,or health authorities). Partly for this reason,communication campaigns – and research aboutthem – also need to take into account the placeof interpersonal communication among theseindividuals’ sources of information. The dimen-sion of change is also where the empirical focusof a campaign study is specified.

Third, campaigns may be distinguished interms of their primary beneficiary. The proto-typical campaign may be the one illustrated inFigure 9.3, with an individual receiver and an institutional sender, who is likely to benefit.It is important, however, to note varieties also of this prototype. While commercial adver-tising is on behalf of the advertiser, the commu-nication is also, arguably, of some informationaland economic value to consumers. And, whereasa health campaign is designed to prevent, forinstance, employees or the general public fromsuffering as patients, at the systemic level it islikely to save money for corporations andhealthcare systems.

Because of their diversity and complexity,campaigns have been studied by means ofvarious methodologies – surveys, experiments,focus groups, and so on. These approaches caneach tap different moments of impact, whichdo not always constitute a logical sequence butwhich may complement each other in assessing

a particular campaign. An additional consider-ation, for media planners as well as researchers,is what might happen in the absence of a cam-paign – a question which normally cannot be addressed in any clear-cut, experimentaldesign, but calls for indirect evidence. Advertis-ing professionals and researchers tend to recog-nize that it is ‘almost always impossible toestimate the impact of advertising on salesvolume’ (cited in Schudson 1984: 17). Eventhough it can be difficult, then, to document themarginal value of both advertising and public-service campaigns, it still may be necessary foran advertiser, interest group, or governmentagency to maintain a presence in the social fieldof attention that is made up by the media.

Campaign research shares its focus on theseveral media and moments of communicationwith studies of intertextuality,� even if theirtheories and methodologies differ substantially.A point of convergence is the sort of ‘plannedintertextuality’ that is characteristic of muchcurrent media marketing, from the merchan-dising of fictional characters across severalmedia, artefacts, and theme parks, to the redis-covery and recirculation of figures such as theBatman (Pearson and Uricchio 1990). Oneexemplary intertextual study, which alsoaddressed the economic aspects of the JamesBond figure since the 1950s, is Bennett andWoollacott (1987).

Knowledge gaps

Compared to diffusion research, which, to adegree, has examined inequalities in the avail-ability of different media, research on ‘know-ledge gaps’ has stressed the extent to whichsuch inequalities may be cumulated, over timeand in conjunction with other forms of socialinequality. Furthermore, this tradition hasemphasized that the point may not be thematerial availability of the technologies, buttheir discursive accessibility – their perceivedrelevance as resources for civic life and dailyliving.� An implicit premise of much public

Media effects148

promoting orpreventing

change


� availability vs. accessibility – see p. 140

debate on the media, and of many media prac-titioners, is that information and communica-tion technologies make information andaesthetic experience much more accessible thanat any earlier point in history. In response tothat premise, the knowledge-gap hypothesisasserts that the flow of information in societyvia the media tends to widen, rather than close,the gaps between the information-rich andinformation-poor (Tichenor et al. 1970).

Like other research on cumulative and long-term effects, knowledge-gap surveys tend toproduce suggestive rather than conclusive evi-dence. This is due, in part, to pre-existinginequalities in the distribution of informationaland cultural resources between different socialsegments, of which the media are only onesource. However, by examining public know-ledge of specific issues and events whichreceived extensive media coverage during a particular period, the original group ofresearchers concluded that better-educatedpeople were better able to the assimilate theinformation (Tichenor et al. 1970; Donohue et al. 1987; for experimental evidence seeGrabe et al. 2000).

The same researchers did note that themedia may work to close gaps under conditionsof social conflict, when public informationbecomes more urgent, relevant, and hencelearnable (Donohue et al. 1975). Other studieshave gone on to counter the general hypothe-sis, suggesting that audiences’ levels of interestmay be more important for their recall thantheir level of prior knowledge (Genova andGreenberg 1979). It has also been argued thateach medium works in specific ways, so that,for instance, television, which reaches a higherproportion of the population than print media,and whose formats do not privilege the welleducated, might work as a leveler of knowledge(Neuman 1976). Of course, any further assess-ment of such knowledge-leveling would have totake into account the types of television systemsand presses being compared.

In view of previous studies, also of recall ofmedia content in the shorter term, the contri-bution of the media to knowledge gaps insociety remains a plausible hypothesis. Morerecent work has suggested links between know-

ledge gaps and other aspects of audiences’ cog-nitive orientations, such as a disaffection withpolitcs (Fredin et al. 1994). Overviews of theliterature conclude that media appear not to beclosing or narrowing gaps (Gaziano 1997;Viswanath and Finnegan 1996). Returning tothe widespread assumption that media canserve as vehicles of democracy, at least in theformal sense of making information universallyavailable, one can note that knowledge gaps doseem to persist in diverse areas of economic,political, and cultural activity. To be sure, suchconclusions are themselves premised on an‘educated’ definition of information and know-ledge, which may depreciate other forms ofcognition and competence. Nevertheless, ex-plicit and operational knowledge regardingpolitics and economy must be considered cen-tral social resources, as suggested most recentlyby debates addressing a ‘digital divide’ in aninformation or network society.

Framing

The concept of ‘frame’ suggests that an item ofinformation, whether arising from media, fromother sources, or from direct perception, isinvested with meaning only once it is placed ina context of other information. In framingcertain items of information, we simultaneouslybracket these from potentially endless massesof information. Compared to studies of recalland knowledge gaps, the focus is shifted awayfrom information as discrete entities, andtoward the complex and socially specific world-views that audiences hold. Compared to‘agendas,’ which may be understood as a tem-porary set of priorities, ‘frames’ amount tomore permanent dispositions of a social ormental variety. As such, frames are of specialinterest for understanding the media–societynexus, but correspondingly difficult to opera-tionalize in empirical research.

Perhaps for this reason, studies of framinghave rather diverse sources, and have producedsome strange bedfellows. In an overview ofprevious research, Scheufele (1999) noted thecontributions of both experimental social psy-chology, particularly Heider’s (1958) classicattribution theory, and of qualitative micro-


1111234567891012341115678920123456789301234567894012345678950111

levels ofinterest andknowledge

digital divide

social andmental frames

sociology in the form of Goffman’s (1974)frame analysis. In certain respects, work relyingon concepts of framing has begun to bridgeboth such theories and the qualitative–quanti-tative divide. In one exemplary study thatemployed depth interviewing, surveys, contentanalysis, as well as experiments, Neuman et al.(1992) showed how audiences would rely onframes which were derived largely from per-sonal experience in order to make sense of thenews – interpretive frames that differed fromthose offered by the news media. This is con-sistent with qualitative reception studies whichhave found that audiences employ ‘super-themes’ in order to establish meaningful linksbetween the worlds of news and of everydaylife (Jensen 1988, 1998).

One further potential of the concept offrames (and of agenda-setting) is that it enablesa more detailed comparison of media pro-duction and reception. On the one hand, jour-nalists and other media professionals areconstantly engaged in framing content, and inanticipating how they may be interpreted (e.g.,Ettema and Whitney 1994; Gans 1957). On theother hand, media audiences necessarily employinterpretive frames that are generated andreformulated over time, with reference to mediaas well as other sources (Gamson 1992; Graber1984). Presumably, audiences in their turn are also aware that both news professionals and other sources have a stake in the dissemi-nation of particular perspectives, which can be resisted. In this regard, framing research has theoretical affinities with Hall’s (1973b)seminal work on processes of encoding anddecoding, including oppositional frames ofinterpretation.�

Scheufele (1999) makes the additional pointthat agenda-setting and framing may representdifferent, but interrelated phenomenal aspectsof media impact. ‘While the process of issueselection or agenda-setting by mass media . . .needs to be a conscious one, framing is basedon subtle nuances in wording and syntax andtherefore . . . most likely [has] unintentionaleffects, or at least effects that are hard topredict and control by journalists’ (p. 19).

Extending this line of argument, it is natural toconclude that processes of agenda-setting andof framing will be interdependent, and thatsuch more or less conscious and plannedaspects of the audience response will requirecomplementary methodologies. In a com-parative perspective, framing has introducedanother important differentiation in the notionof effects, since ‘framing influences how audi-ences think about issues’ (p. 19), just as agenda-setting influences what issues to think about.

Cultivation theory

This section begins to enter ‘an area wherethere is much theory and speculation but littlefirm evidence of confirmed relationships be-tween the mass media and matters of values,beliefs, opinions and social attitudes’ (McQuail2000: 459). Whereas both findings and theo-retical models from the research traditionsreviewed so far have certainly been contested,the possibility of empirically studying the rela-tively more short-term effects of media is rarelychallenged. When it comes to the contributionsof media to long-term and unplanned socialchange (or stability), it has proven considerablymore difficult to conceptualize and opera-tionalize ‘effects.’

The difficulties are of at least two differentkinds. On the one hand, media evidently con-tribute to the socialization of individuals – butin exceedingly complex ways. On the otherhand, media enter into an interchange also withall other social institutions, with current events,and with the cultural practices of the public atlarge. The final sections of this chapter reviewissues regarding the study of ‘socialization’ and‘institutionalization’ via the media. First, thissection examines one of the most extensivelystudied theories of how media may cultivatepeople’s worldviews – an approach which is sit-uated in the borderlands between individualand systemic effects of media. It is the cradle-to-grave socialization of everybody (Americans)through media (television) which has been thefocus of cultivation research.

The key assumption of the cultivationhypothesis is that television has taken on sucha central place in modern American culture that

Media effects150

super-themes

productionand reception

as framing

� oppositional frames – Chapter 10, p. 162

how to thinkabout issues

cultivationhypothesis

it constitutes a ‘symbolic environment’ in andof itself. As such, it competes with, distorts,and, to a degree, substitutes the reality of per-sonal experience, as informed by other mediaand interpersonal sources as well (e.g., Gerbnerand Gross 1976). The methodology designed totest the hypothesis has been a combination of survey and content analysis, comparing therepresentation of social reality on television(not least violence, crime, and other ‘risks’)with viewers’ awareness of and attitudestowards such concerns. The conclusion of theoriginal group of researchers, with GeorgeGerbner as the seminal figure, is that such a cul-tivation of knowledge and attitudes does occur.Consequently, ‘heavy’ viewers are more likelyto produce ‘the television answer’ to questions,for example, about risks to them personally, but also about issues such as ethnicity, gender,or poverty. While debated, the approach hascontinued to attract interest, and to recruitresearchers to study diverse aspects of cultiva-tion, but documenting rather mixed evidence(Shanahan and Morgan 1999).

In certain respects comparable to agenda-setting research, cultivation research goesbeyond any simple stimulus–response model,both by noting the cumulative nature of theimpact, and by studying it within a carefullyarticulated theoretical and methodologicalframework. Moreover, the cultivation studieswere originally conceived as part of a holistic‘cultural indicators’ research program. The pur-pose was, from a critical position, to documenthow the media, being the central modern meansof cultural expression, maintain and reinforcethe social status quo. The research program alsoincluded analyses of television as a social insti-tution, although most efforts have gone into thecombined audience-content studies.

Criticism, especially of the more specific cul-tivation hypothesis, has been severe, and hastaken two forms. First, cultivation researchershave been attacked on their own turf througha re-analysis of the original datasets. In brief,critics have argued that, if sufficiently stringentstatistical controls are introduced, the cultiva-tion hypothesis cannot be upheld (Hirsch 1980,1981). In addition, humanistic researchers havequestioned the entire theoretical rationale

behind the datasets, arguing that the approachneglects television as ‘texts’ and their variableinterpretation by viewers (Newcomb 1978).Second, a number of studies in countries otherthan the U.S.A. have not found support for thehypothesis.� Beyond surveys, some of thisresearch has included a component of experi-mental methodologies, which came to beapplied to cultivation studies during the 1990s(Shrum 1996).

A third issue is whether the focus of culti-vation research on television has been mis-guided, theoretically and historically. Not onlyare the media institutionally and discursivelyinterconnected, as suggested by, for instance,campaign and intertextual studies, but ‘the ageof television’ is probably coming to an end, atleast in the present technological and institu-tional definition of ‘television.’ If the theory ofaggregate cultivation stands and falls with onemedium, it is not sufficiently general to guidemedia research. If, in addition, one takes intoaccount the ongoing extension and redefinitionof the field of media and communicationresearch, it may be an appropriate occasion toreconsider the status and explanatory value ofcategories such as ‘cultivation,’ ‘socialization,’and ‘effects.’

SOCIALIZATION BY MEDIA

One response to the difficulty of ascertainingthe long-term effects of media on humans andtheir societies in empirical studies has been areturn, for at least a decade, to more ‘grand’theories of culture and society. This reorienta-tion has entailed a return, as well, to the disci-plinary sources of the interdisciplinary field ofmedia studies. For example, history has cometo the fore in the shape of ‘medium theory’(Meyrowitz 1994), and sociology has been thesource of theories of modernity, as increasinglyapplied to media (e.g., Thompson 1995). Acommon aim has been to relocate media on amore inclusive and, hence, more illuminatingconceptual map, as witnessed also by the con-vergence of different research traditions.

Socialization by media 151

1111234567891012341115678920123456789301234567894012345678950111

culturalindicators

� cultivation research in non-U.S. contexts – seeChapter 13, p. 229

Grand theory, however, does not precludeempirical testing. An example is research on‘lifestyle’ and, more broadly, the ‘life forms’ ofdifferent social classes. Here, the seminal workof Bourdieu (1977; 1984 [1979]) on ‘habitus,’‘cultural capital,’ and their role in differentsocial fields of practice has been especially influ-ential. In addition, however, lifestyle researchhas become a somewhat surprising meeting-place of critical social theorists and marketingprofessionals. The two groups are broughttogether by an ambition of understanding cul-tural forms, values, and styles, not just as indi-vidual expressions but as the manifestation ofembodied, persistent, yet contested social struc-tures. This perspective follows, in part, fromthe ‘rediscovery’ of agency and reflexivity incurrent social theory, including Bourdieu andGiddens, following the conflicted coexistence ofstructural functionalism and Marxism in thepost-1945 period. During this same period,marketing practitioners have been challenged torefine their tools of prediction and planning, soas to gain a marginal advantage in the market.The means to this end, increasingly, has becomea better understanding of how consumers makesense of products, and of themselves.

Box 9.1 presents an example of lifestyleresearch as it relates to media issues. Figure 9.4further illustrates the methodology of the studyin question and its way of displaying findings,both of which are characteristic of lifestylestudies.

Research on living conditions, lifestyles, andmedia uses offer one indication that any inclu-sive terminology referring to ‘effects’ in generalmay be a misnomer. Instead, ‘socialization’ viathe technological media is an inescapable con-dition of modernity – just as the medievalchurch and early modern technologies (e.g., theprinting press) were constitutive elements of thesocial order of their epochs.

An issue currently facing media and com-munication research is how to specify the ele-ments and processes of socialization, includingthe distinction between primary and secondarysocialization. Most basically perhaps, sec-ondary socialization, as associated with othersocial institutions than the family (schools,interest organizations, media), has arguably

taken on greater importance overall, and hasbecome more formative of the individual overtime. This appears to be the case, both in quan-titative terms of the time spent on formal edu-cation as well as mediated leisure, and inqualitative terms of where decisive social stan-dards and knowledges come from (e.g., Beck etal. 1994). Moreover, the line between primaryand secondary socialization has becomeblurred, if it was ever clear. The fact that theprimary socialization of children and youth, astraditionally performed in the family, is increas-ingly interdependent with media, has been sug-gested in a wide variety of earlier research –from the Payne Fund studies of the the 1930s(Jowett et al. 1996), via Himmelweit et al.(1958) and Schramm et al. (1961), toLivingstone and Bovill (2001), and Rosengrenand Windahl (1989). For centuries, print mediahave been agents of socialization that originatein the public sphere, and take effect in theprivate sphere (e.g., Drotner 1988), whereas the electronic media have come to redefine the very meaning of ‘private’ and ‘public’ (e.g.,Meyrowitz 1985).

At the level of individual socialization, then,the modern media are bound to inform howpeople think and act, even if the evidence willnormally be indirect or circumstantial. At thelevel of social institutions as well, the media areone of the inevitable structural conditionsunder which social interaction takes place.

INSTITUTIONALIZATION BY MEDIA

In view of the comparative recency of ‘themedia,’ research on their structural or systemicinterdependence with other social institutionsmay be premature, and might be left for futurehistorical studies. However, the field of mediaand communication research has had an ambi-tion both of explaining current institutionalarrangements and of projecting the future ofmedia. This has been due, in part, to the strat-egic placement of the media sector from the per-spective of commercial and political interests.

While each of the traditions reviewed in thischapter have produced some evidence for infer-

Media effects152

lifestyle andlife forms

primary andsecondary

socialization

� Payne Fund studies – Chapter 10, p. 158

Institutionalization by media 153

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 9.1 CORRESPONDENCE ANALYSIS OF LIVING CONDITIONS,LIFESTYLES AND MEDIA USE

A common method of relating people’s material living conditions with their interpretation of thoseconditions, as informed partly by the media, is, first, to ask them a complex set of survey ques-tions regarding fundamental values (in addition to sociodemographic information). Next, both thevalue statements and other items of information are analyzed and compared through a multivari-ate technique of ‘correspondence analysis’ (Greenacre 1993).� In essence, the analysis is able toestablish the degree to which different answers are correlated, and the pattern of these correla-tions, as they apply to both single respondents and entire samples. The analysis thus results in aset of interrelated values, which may be understood in terms of a ‘worldview’ that affects people’spolitical and cultural orientations and actions. Because the approach normally relies on largesamples, lifestyle research can produce robust findings.And, because it is longitudinal, tracking shiftsand trends� in the public’s tastes and preferences, it might be said to offer an approximation ofthe Zeitgeist – the cultural universe of an age.

In technical terms, the different values constitute points in a multidimensional mathematicalspace, but are normally represented in a two-dimensional model, as in Figure 9.4. The two

cont.

� correspondence analysis and multivariate statistics – Chapter 13, p. 232

� trend studies – Chapter 13, p. 218

Innovation

Tradition

Technology

Pragmatism Idealism

Challenge

Target fixation

Internationalization

MarketReason Unreflective

Work over leisureInformal

Flexibility

Unselfishness

Heterarchy

The environmentbefore industry

Alternative life

DominanceCommunity sense

The new family

Individualism

Personal expression

Blurring of sexesIntraception

Fitness and healthinvestment

Open citizenship

Spiritualism

NetworkingDelight in natural beauty

CellulizationExploring mental frontiers

Active protection ofthe environment

EmotionalexperienceFear of technology

The family

RootsFear of violence

Aimlessness

Conformity

Economic security

Privatization

SelfishnessStatus

Hypernatural

Taste for risks

Materialism

Fact fixationUncertainty and complexity

Feeling of political efficiency

Self-respect

Self-instrumentalism

Personal appearance

Figure 9.4 Chart of values among Danish respondents according to Research Institute on SocialChange principles (1995)Source: Schrøder 1999: 57. Reprinted with permission of Sage Publications, Ltd.

ences about the long-term significance of media,four additional types of research have focusedmore specifically on the long term. In each case,a late ‘stage’ of the interaction between mediaand audiences in a social context is taken up inorder to examine and typologize changes in howpeople go about their business, in public or inprivate, as a result of the presence of media.

• Natural experiments on the introduction ofmedia. Particularly in the case of television(which developed after the formation of mediaresearch as a field), the first introduction of themedium into a community or culture has pro-vided opportunities to assess any changes in,for example, time use and the role of differentmedia as reference points, for instance, for playor conversation. Because of the explicit before-after situation, such issues have lent themselvesto study in natural experiments.�

• Public events. A distinctive role of modernmedia is to present events and issues to thepublic for some form of collective considera-tion. In doing so, they necessarily shape thecontent and form of the available coverage, butthey may also influence the course and finaloutcome of events in other social institutionsand sectors. Whereas agenda-setting researchhas focused on the part played by media inarticulating political issues, studies have alsoexamined how other types of social activityunfold and are resolved, in part, through mediaintervention. In contrast to the common lamentthat media create ‘pseudo-events’ (Boorstin1961), several important qualitative studieshave shown how media both frame and enablepublic participation in major social events –such as national ceremonies, political scandals,

Media effects154

dimensions of this model correspond to those two underlying principles which are found, in thegiven analysis, to be the most important in accounting for the particular pattern. Figure 9.4 exem-plifies a configuration of values with reference to a study of everyday life and political participa-tion in Denmark, conducted in 1994 to 1995 (Schrøder 1999).The two dimensions (which recurin comparable studies) indicate different degrees of orientation toward a ‘modern’ (innovative) or a ‘traditional’ life form, and toward an idealist or a pragmatic view of life. Furthermore, thesepatterns are more or less analogous to a differential distribution in society of what Bourdieu (1984[1979]) terms economic and cultural capital (typically operationalized as money and education).

Specific configurations suggest different conceptions of self and society, and will vary acrosscultural contexts as well as historical time. At issue along these dimensions are classic questionsof economy, politics, technology, and ethics. In the present context, the social field of media usemay be broken down along similar dimensions, so that media preferences correlate with funda-mental values as well as sociodemographics. One example from the European context arises fromthe different types of broadcasters. While the original public service monopoly in the Danishstudy has its stronghold in the upper, ‘northern’ part of the chart, especially its ‘eastern’ segment,a preference for the commercial broadcasters in the past ten to fifteen years is associated withthe values clustered in the lower, ‘southern’ part, especially its ‘western’ segment.

It should be noted that details of both the theory and operationalization of ‘lifestyle’ vary.Compared to Bourdieu’s (1984 [1979]: 128–129) social-systemic chart of values, other studies havetaken a more individualist conception of lifestyles (e.g., Inglehart 1990; Inglehart et al. 1996), orhave adapted the concept for marketing purposes. (For a philosophical elaboration of the cate-gory of ‘taste,’ see Douglas 1996.) In addition to the U.S.A., the general approach has been employedin several European countries, in part as an instrument of planning television programming in theface-off between commercial and public service broadcasters. However, comparable studies havealso suggested that the explanatory principles may hold, in a modified form, in a cross-culturalperspective for studying both modern and modernizing cultures (e.g., Schwartz and Bilsky 1990).

economic andcultural capital

taste

cont.

� natural experiments – Chapter 13, p. 226

and international sports – by making them into ‘media events’ (e.g., Dayan and Katz 1992;G. E. Lang and Lang 1981; K. Lang and Lang1953).

• Institutional practices. Media occupy aspecial position among other modern institu-tions because they constitute a common arena– a cultural forum (Newcomb and Hirsch1984) – in which the standards, priorities, and instruments of other institutions may benegotiated and coordinated. Again, this is espe-cially evident in political life, which is subjectto a specific public accountability, and which,further, is increasingly carried on in the media.(In comparison, private economic enterprise is,to a large extent, exempt from detailed publiccontrol; and high forms of culture are still com-monly conducted and covered in separatearenas or media.) Accordingly, much researchhas examined how the media may be changingthe practices of political democracy, and hasworried, not least, that the actual participationand influence of the audience-public is beingdiminished (for an overview see Blumler andGurevitch 1995). In addition to structural and normative-theoretical analyses,� researchin this area has often relied on either contentor discourse analysis to identify the shorteningof political dialogue into ‘sound bites’ (Hallin1992) or historical shifts in how prime minis-ters communicate via television (Seymour-Ure1989).

• Cultural formations. In the most generalsense, the technological media can contributeto making certain cultural practices and world-views dominant. On the one hand, a criticalstrand of research has, for example, made thegeneral argument that a ‘media logic’ (Altheide

and Snow 1979) has been invading other socialcontexts. Other critical studies have employedcontent studies to document systematic omis-sions in news coverage, which might fragmentpublic understanding and legitimate the socialstatus quo (Group 1976, 1980). On the otherhand, more recent work has argued for thedemocratizing potential of popular culture, ascarried by the media. As is suggested, forinstance, by Hobson’s (1982) study of televi-sion fans demanding ‘their’ soap back on thescreen, the media may encourage audiences toinsist on their cultural preferences and plea-sures. In addition, ‘medium theory’� is amongcurrent attempts in research to capture thespecificity of modern culture. Much furtherresearch is needed, not only on each ‘stage’ ofcommunicative impact, but perhaps even moreon the nature of their interrelations.

In conclusion, this chapter has reviewed themain traditions of quantitative audience re-search, with special reference to their theoreti-cal premises and methodological procedures.The review has suggested, first, the comple-mentarity of the several stages and levels inves-tigated in previous studies. Second, it hasindicated that quantitative audience researchhas been particularly successful in accountingfor the ‘early’ stages of mediated communica-tion, such as the diffusion and consumption ofeach medium, and for ‘mid-term’ effects, suchas recall and agenda-setting. Qualitative recep-tion studies, as reviewed in the next chapter,have proven able to supplement quantitativeresearch by focusing on other stages of the com-municative process, including the everyday con-texts of media use and some of the long-termcultural implications of media.

Institutionalization by media 155

1111234567891012341115678920123456789301234567894012345678950111

media events

cultural forum

� normative media theories – Chapter 16, p. 276 � medium theory – Chapter 2, p. 16

MILESTONES REVISITED

One way of encapsulating the field of study hasbeen through systematics and models, such asFigure 9.1, which identified some main stagesof the communicative process and the tradi-tions of audience research associated with each.Another means of taking stock has been toidentify historical ‘milestone’ studies which,arguably, defined appropriate ways of concep-tualizing and examining the various stages ofcommunication empirically. By way of intro-duction, and of joining qualitative and quanti-tative traditions, this chapter revisits some ofthe milestones of audience studies, as received– and sometimes overlooked – in later research.

Figure 10.1 lays out fourteen milestones, asdefined by Lowery and DeFleur (1995) throughthree editions of their widely circulated text-book, and adds a number of candidates fromoutside the dominant paradigm (Webster andPhalen 1997). While any such listing is essentiallycontestable, the aim is to consider a broaderrange of defining contributions to the field inorder to explain, and promote, the process ofconvergence between research traditions.

In the case of the established ‘Milestones I,’it may be especially apparent that several con-tributions were shaped by their historical andsocial context – from war to political or publicdemand for evidence on media effects.� At thesame time, it is appropriate to note also thatthe primarily European, qualitative, and criti-cal perspectives under ‘Milestones II’ bearwitness to the impact both of war and of thecultural revolution of the 1960s on scientificideas and practice. In both cases, these cir-cumstances do not disqualify either findings orapproaches, but serve as a necessary context forassessing the specific explanatory value of theentries and for considering their commensur-ability in future theory development.

To begin a review of the milestones, a fewcaveats are in order. First, the listing considersonly publications post-1900. As indicated inChapter 2, before the earliest origins of ‘media’research, most thinking on communication and its effects had been conducted in rhetoricaland aesthetic traditions of inquiry,� from

Media receptionQualitative traditions

Klaus Bruhn Jensen

10

• a presentation of the established ‘milestones’ of quantitative audience research together with seminalqualitative contributions

• a review of the main forms of reception analysis• an assessment of recent work in ‘media ethnography’• a discussion of the place of media discourses and other relevant forms of evidence in further reception

studies.

� social origins of media research – Chapter 16

� rhetorical and aesthetic effects – Chapter 2

Aristotle’s rhetoric, via Kantian aesthetics andnineteenth-century hermeneutics, to semiotics –all of which continue to inform contemporarycommunication theory. Second, the revised setof milestones remains limited by the fact thatso much of the field’s activities – as currentlyinstitutionalized internationally through jour-nals, conferences, and translation – originatesin either the U.S.A. or Europe. Third, a partic-ular difficulty in selecting additional candidatesfor ‘milestones’ is that much previous research,while bearing on reception and effects, has notalways examined audiences empirically. Thedifficult dividing line between ‘text’ and ‘audi-ence’ is a product both of the rudimentary the-orization of content that has characterizedquantitative research traditions, and of thequite recent turn to empirical audience studiesamong humanistic scholars.

The two groups of ‘milestone’ studies arenoted here for the record, and as a point ofdeparture for theory development and empiri-cal collaboration. Having largely developed inseparate spheres of research, much of theirpotential for convergence simply remains to be

assessed. In addition to the affinities betweenthe Payne Fund and Mass Observation researchprograms, the diffusion of innovations, forexample, and medium theory, as anticipated byWalter Benjamin, would seem to share anumber of research questions regarding thesocial consequences of new technologicalresources for communication. Further, agenda-setting studies overlap conceptually with muchqualitative research on the socially and cultur-ally specific frames of decoding media texts(e.g., Gamson 1992; Liebes and Katz 1990).Cantril et al.’s (1940) contribution, employinga liberal mix of qualitative and quantitativemethodologies, was an early indication of whatmay be gained if research programs as well asindividual scholars reconsider their theoreticaland methodological predilections.

The rest of this chapter reviews the contri-bution of qualitative research traditions. Whilebeing late starters in terms of empirical studies,these traditions have produced new insight intoat least three different moments of the processof mediated communication. Reviewed in thefollowing section, the three foci of reception

Milestones revisited 157

1111234567891012341115678920123456789301234567894012345678950111

20001900

Payne Fund(1933–35)

Diffusion ofInnovations(1943)

Project Revereon leaflets(1952–54)

TV andBehavior(1982)

Violence andthe Media(1969)

Television in theLives of ourChildren (1961)

RadioGratifications(1944)

MassObservation(1937ff.)

The Work of Art in the Ageof Technical Reproduction(1936) Constituents of a

Theory of the Media(1970)

The People’sChoice(1944)

Communicationand Persuasion(1953)

TheInvasionfrom Mars(1940)

Film experimentson Americansoldiers (1949) Personal Influence

(1955)

Myth Today(1957)

Language andCinema (1971)

The ‘Nationwide’Audience (1980)

TV: Technology andCultural Form (1974)

TV Culture(1987)

Agenda-setting (1972)

Surgeon General(1971)

Reading theRomance(1984)

The SocialUses of TV(1980)

The Exportof Meaning(1990)

Figure 10.1 Milestones of media and communication research

Media reception158

cont.

MILESTONES I

• Payne Fund.This first substantial mediaresearch program in the U.S.A. examinedthe effects of movies on children andyouth through a variety of content, survey,experimental, and other methodologies,arising from and feeding intocontemporary concerns and debates (foran overview see Jowett et al. 1996).

• The Invasion from Mars. Cantril et al.’s(1940) multimethod study of how theAmerican public responded to OrsonWelles’s radio dramatization of War of theWorlds suggested, among other things, howto combine qualitative and quantitativemethodologies for a concrete researchpurpose. (For a critical reassessment, seeRosengren et al. 1978).

• Diffusion of Innovations. Following 1940sstudies on the adoption of agriculturaltechniques, diffusion research expanded toaddress other kinds of innovations andtheir place in processes of social change,including media and their dissemination ofinformation (for an overview see Rogers1962).

• The People’s Choice. In certain respects theinaugural work of U.S. communicationresearch, this study focused on the place of media in political democracy, relying onpanels and other state-of-the-art surveymethodology (Lazarsfeld et al. 1944).Atleast in its reception history, the study hasoften been taken to suggest that mediareinforce rather than change people’spositions, and that, overall, media servedemocracy (for a classic critique see Gitlin1978).

• Radio Gratifications. Among thecontributions of the first generation of U.S.researchers were some of the seminalideas which were later developed as uses-and-gratifications research. Apart from itsinherent quality, the work by Herzog(1944) was unusual in considering the‘feminine’ genre of daytime radio serials.

MILESTONES II

• Mass Observation.The mass observationstudies in the U.K., from the 1930sonward, documented various aspects ofsocial life, including cinema-going and othermedia use (for an overview see Richardsand Sheridan 1987). Comparable in certainrespects to the American Payne Fundstudies, the studies represent an earlyproposal for ways of examining the placeof media in an increasingly urbanized andmodernized everyday, providing valuableinsights into processes of mediaconsumption and reception.

• The Work of Art in the Age of TechnicalReproduction.This title (sometimestranslated as ‘mechanical’ reproduction)covers Walter Benjamin’s (1977 [1936])early attempt to capture what wasdistinctively new about the cultural formsbeing disseminated by the technologicalmedia.Taking ‘medium theory’ (seeChapter 2) to the level of concrete textualanalysis, this work identified the loss of‘aura’ in media, compared to traditionalartworks, as a key factor which may besaid to overdetermine the interpretationsand effects of specific media items. Likeother critical social theory of the 1930s to1940s (e.g.,Adorno and Horkheimer 1977[1944]), Benjamin’s work was, in part, anintellectual response to the conditions offascism – a response that was different inkind from the U.S. milestones participatingin the war effort, but which engagedrelated social circumstances.

• Myth Today. Following the rebuilding ofpostwar Europe, new societies (andresearch institutions) were taking shape, inwhich the definition of ‘culture’ was againat issue, not only in debates over‘Americanization’ (see Webster 1988), butwith reference to popular culture as such.One of the undoubted, early influences onresearch in this area was Barthes’s (1973[1957]) work on the modern ‘mythologies’disseminated in large part by the media.

Milestones revisited 159

1111234567891012341115678920123456789301234567894012345678950111 cont.

cont. MILESTONES I

• Film experiments on American soldiers. Aspart of U.S. involvement in World War II,a series of experimental studies wereconducted of Frank Capra’s Why We Fightfilms, asking to what extent they might notonly provide information, but also shapeattitudes (Hovland et al. 1949).

• Communication and Persuasion. Departingfrom this and other earlier experimentalresearch, and from theories of the selectiveinfluence of media on individuals, the YaleProgram of Research on Communicationand Attitude Change proposed togeneralize the perspective, but were ableto document especially short-term changes(Hovland et al. 1953).

• Personal Influence. Elaborating, in part, the(serendipitous) finding of Lazarsfeld et al.(1944), that technological media work,not least, by being mediated further ininterpersonal communication by opinionleaders, Katz and Lazarsfeld (1955) helpedto establish the two-step flow ofcommunication as a generally influentialmodel.

• Project Revere. At the height of the ColdWar, and with the Korean War still beingfought, this project (1952–1954) wasfunded by the U.S. military to explore theuse of airborne leaflets to replace, orcircumvent, other channels (for anoverview see DeFleur and Larsen 1987).

• Television in the Lives of our Children.Thisfirst major study of TV’s effects on childrenin the U.S.A. was characterized partly bythe diversity of issues (and methodologies)included, such as social differences, partlyby a renewed uses-and-gratificationsperspective (Schramm et al. 1961). A close(and earlier) European cousin, notrecognized among the ‘milestones’ ofLowery and DeFleur (1995) (but dulynoted in the preface to Schramm et al.),is Himmelweit et al. (1958).

• Mass Media and Violence. Growing out ofU.S. government and public concern over

MILESTONES II

Particularly his essay, ‘Myth Today,’introducing the distinction betweendenotation and connotation into analysesof how meanings are offered by media,and taken by audiences, has shaped theunderstanding among the first generationof European media researchers of howmedia come to take effect on bothindividuals and societies.

• Constituents of a Theory of the Media.Enzensberger’s (1972 [1970]) articlereturned to Bertolt Brecht’s pointregarding radio in the 1930s, that theequipment for receiving could as well beused for sending messages and participatingin a society-wide dialogue, thus suggestingthe critical potential of media (as well asthe generally constitutive role of media insocial systems). In addition to the period-specific ideological message, Enzensberger’stheoretical message anticipated, to adegree, later work on the ‘active’ audienceand on computer media.

• Language and Cinema.The title of one ofthe film theorist, Christian Metz’s (1974[1968]), main publications suggests the‘milestone’ nature of theoretical workduring the 1960s to 1970s on the specificverbal and visual vehicles of communica-tion, and on their mode of address to theaudience.While other contributors tosemiotics and discourse analysis could havebeen singled out (see Chapter 2), the keyquestion of whether cinema and othernon-alphabetic communication might bestudied as ‘languages’ (and if not, thenhow), has been given perhaps its mostelaborate articulation in Metz and otherEuropean semiotics and structuralism.

• TV:Technology and Cultural Form.Williams’s(1974) definition of broadcasting (radio and TV) as a ‘flow,’ rather thanautonomous works, identified a key aspectof much mediated communication, and anticipated certain features of computer media (see also Ellis 1982).

studies so far have been the everyday contextsof media use; the decodings or textual inter-pretations of media content; and audience usesof this content as a resource in other social con-texts. The following section discusses recentproposals for a more inclusive research positionof ‘media ethnography.’ Finally, the chapterreturns to the importance of ‘texts’ or ‘dis-courses’ for the study of reception, and of medi-ated communication as such, and describes the

relevance of various discourses and other typesof evidence for future reception studies.

MOMENTS OF INTERPRETATION

Contexts of media use

As previewed in Figure 9.1, one focus of qual-itative audience studies has been the context of use, particularly in the household. Whereas

Media reception160

cont. MILESTONES I

social unrest during the 1960s, this report(Lange et al. 1969) summarized earlierresearch, and presented new evidence andconclusions, particularly on the impact ofviolence in entertainment programs,relying on approaches which have beenextended in cultivation research (seeChapter 9).

Surgeon General.Traditionally referred to as the‘Surgeon General’s Report on Televisionand Social Behavior,’ the relevant work(Comstock et al. 1971) is, in fact, a largecollection of studies which informed such a report. In one sense following up on theprevious ‘milestone’ on a grand scale, theissues and findings of this publication were,again, oriented toward specific policyissues.

Agenda-setting. Since the 1970s, agenda-settingresearch has been influential in restatingone key question of audience research asthat of setting (political) agendas, ratherthan changing attitudes and votingpreferences (McCombs and Shaw 1972).

Television and Behavior. Explicitly following upon the ‘Surgeon General’s Report,’ this last‘milestone’ took stock of research tenyears later, but included a wider range ofissues than televised violence (Pearl et al.1982).While no new research wascommissioned, the integrative reviews ofthe massive research output since 1971provided an authoritative overview, withspecial reference to television.

MILESTONES II

• The ‘Nationwide’ Audience. Drawing on bothdiscourse analysis and gratificationsresearch, Morley’s (1980) was the firstmajor publication to bring together social-scientific and humanistic perspectives onthe audience in a qualitative empiricaldesign regarding television news reception,and is reviewed below, together with thefollowing entries.

• The Social Uses of Television. In the sameyear, Lull (1980) went beyond the individualfocus of gratifications research to studythe social uses of television in the family,also relying on qualitative methodology andre-emphasizing the explanatory value ofmicrosociological traditions.

• Reading the Romance. At the samejuncture, Radway (1984) provided anintegrated study of the institutions, texts,and audiences of the print romance genre,of which especially the audience portionand its fieldwork established a model forfurther research.

• Television Culture. As part of a textbook ontelevision studies, Fiske (1987) presentedan operational approach to intertextualitythat went beyond immanent analyses ofeither the narrative structures or modes ofaddress in media texts (see Chapter 11).

• The Export of Meaning. Finally, thisqualitative study of the experience of theDallas television series among differentcultural and ethnic groups (Liebes and Katz1990) was one of the first attempts toconduct comparative reception research(see also Jensen 1998).

media use inthe household

predecessors included, for instance, the Britishmass observation studies (Richards and Sheri-dan 1987) and innovative techniques for video-taping television viewers (Bechtel et al. 1972),the seminal recent study was by Lull (1980).Referring to the social-scientific undercurrentsof symbolic interactionism, ethnomethodology,and other ‘microsociology,’� and relying onprolonged observation of television viewing inAmerican households, the study developed atypology of the ‘social’ uses of television. Onthe one hand, it went beyond the individualfocus of uses and gratifications research; on the other, it restricted the definition of ‘social’uses to the private family setting, and did notaddress the uses of media for political, eco-nomic, or cultural purposes in the publicdomain. The typology noted, first, the ‘struc-tural’ uses of television in generating an en-vironment or atmosphere, and in regulating, forinstance, bedtime. Second, the study docu-mented several ‘relational’ or interactional usesof television, for example, for facilitating oravoiding contact between parents and children,or between spouses.

A number of later studies have examinedhow media are integrated with and, to a degree,modify other everyday activities (for overviewssee Alasuutari 1999; Moores 1993). Despiteresearch considering media use in public places(Lemish 1982), in quasi-public settings such asa prison (Lindlof 1987), or in peer groups(Buckingham 1993), the focus has remained onhouseholds and the dynamics of family life,with some reference to other social contexts(e.g., Gauntlett and Hill 1999). This is ex-plained, in part, by the simultaneous focus ofreception studies on the television medium,which has lent itself as a convenient referencepoint in exploring the family as a natural socialand cultural unit.

Radio, while generally under-researched(presumably in part because its heyday camebefore the institutionalization of mediaresearch, which coincided with the rise of tele-vision), has received less attention, but has beenstudied, for instance, in a historical perspective(e.g., Moores 1988; Scannell 1988). Similarly,

print media, from newspapers and magazinesto books, which are eminently transportableand adaptable to different social circumstances,apparently have rarely been studied on site.Although this may be due to the seemingly soli-tary experience of reading, print media have arange of social uses, from seeking to ward offcontact, and for individual readers to establisha separate social sphere in which they can be‘absent’ though present (see Radway 1984).The film medium, again, has been studied moreas an artform than as a social practice, althoughrecent work has begun to cover the decodingof films as well as the social acts of going tomovies, and sharing the experience with others(e.g., Barker and Brooks 1998; Gomery 1992;Stacey 1994; Stokes and Maltby 1999a,b). Thereception literature also includes studies of howseveral media enter into the everyday lives ofindividuals and families (e.g., Barnhurst andWartella 1998; Jordan 1992).

As with work by Lull (1988b, 1991), somestudies have employed participating observa-tion to produce a fine-grained analysis of mediain a particular locale (e.g., Gillespie 1995).However, in a number of cases, the mainmethod of data collection has, in fact, beenqualitative interviewing, whether individuallyor in groups (e.g., Gray 1992; Hobson 1980;Morley 1986). While relevant, interviewmethodologies depend on the respondents’introspection, retrospection, and verbal recol-lection of their actions, which necessarilyreproduce events from a current perspective.Particularly in research on the concrete contextsof media use, a weighing of the strengths ofobservation and interviewing, depending on thepurpose of study, is thus of the essence.

Observational studies, in their turn, haverarely been able to capture audience members’response to particular narratives and othercontent elements. Indeed, Lull (1980) did not explicitly relate the different social uses oftelevision, and their relative prominence, todifferent genres. One tendency of receptionstudies so far, then, has been for observationalstudies to detail the everyday lives of peoplearound media, and for interview studies toprobe their interpretations of specific mediadiscourses.

Moments of interpretation 161

1111234567891012341115678920123456789301234567894012345678950111

structural andrelational uses

of television

� microsociology – Chapter 3, p. 56

media use inpublic places

It should be added that computer mediapose new challenges as well as opportunities forresearch on the contexts of media use. Over andabove the various physical locales in which newmedia are used – households, cybercafés, peersettings – the computer tends to blur the dis-tinction between medium and context� (Jensen2000a). Compared to, for instance, televi-sion, where social interaction occurs around,and with reference to, the medium as a distincttechnological and cultural artefact, social inter-action is itself constitutive of many genres ofcomputer media, from chat to online gaming.If the interaction ends, the medium simplybecomes inoperative. In this regard, interactiv-ity is not merely an offer, but a requirement anda condition of computer-mediated communica-tion. This condition entails a different concep-tualization and operationalization of ‘media’and ‘contexts,’ and may lead research to ex-plore, in more concrete terms, the performativeaspects of media use.

Decodings

Research on ‘decodings’ – audience interpreta-tions of specific media discourses – accounts forthe bulk of previous qualitative receptionstudies (for an overview see Livingstone 1998).Much of the tradition originates from a cultur-alist and, in part, literary interest in howmeaning is produced and shared.� This inter-est led to a dual strategy of including the actualmedia ‘works’ in the empirical research designs,but also of insisting that cultural artefacts canhave several meanings, and may have no onecore meaning.

On the one hand, then, decoding studieshave retained a number of concepts and ana-lytical procedures from literary and other cul-tural theory. On the other hand, these empiricalaudience-cum-content studies� have recuper-ated and redeveloped those parts of the literarytradition which have underscored the rela-

tive indeterminacy of meaning, particularly thework of reception aesthetics (for an overviewsee Eco 1987; Holub 1984). Other importantsources of inspiration have been critical socialtheory noting the volatility of the dominant ide-ology, as presented also in media (Parkin 1971).Uses and gratifications studies served, simulta-neously, as a main object of criticism becauseof their preference for functionalism and quan-tification as a source of ideas.

This somewhat eclectic mixture of social-scientific and humanistic notions was broughttogether in the cultural studies tradition.�Taking his lead from the work of Hall (1973b)on the less than perfect correspondence be-tween media practitioners’ encoding of textsand the audience decodings, Morley (1980)presented the seminal study on socially specificforms of decoding news. Employing a focus-group methodology, Morley documented arange of decodings of what was assumed to bethe (ideologically) ‘preferred’ meaning of thenews discourse – from an accepting or ‘domi-nant’ reading, via a ‘negotiated,’ to an ‘oppo-sitional’ reading. Across the focus groups in thestudy, these readings appeared correlated withthe social positions and organizational involve-ment of the participants. For example, the com-bination of low social status with shop-floorunion involvement seemed to produce some of the most explicitly oppositional readings (p. 141). In this regard, audiences themselvescould be said to perform a ‘hermeneutics of sus-picion.’�

In retrospect, it was perhaps the general per-spective, of linking the social-systemic and thediscursive-interpretive attributes of the respon-dents, which most influenced later work in thearea. Morley (1981) himself was among thefirst to criticize the somewhat undifferentiatedconception of decoding as the (degree of) repro-duction of a dominant ideology. He also calledfor the study of, for example, basic compre-hension and derived pleasure as aspects of newsreception (see also Lewis 1983, 1985). Laterstudies have noted several varieties of opposi-

Media reception162

� context as locale and as virtual space – Chapter 11,p. 189

� culturalist or ritual model of communication –Chapter 1, p. 8

� audience-cum-content studies – p. 167

� cultural studies – Chapter 2, p. 39

� hermeneutics of suspicion – Chapter 2, p. 22

dominant,negotiated,andoppositionalreadings of atextually‘preferred’meaning

tional decodings, and have uncovered the diffi-culties that even comparatively sophisticatedaudiences may have in deconstructing theimplied message of a media text (Hacker et al.1991).

In keeping with its critical-political agenda,Morley’s (1980) study paid special attention tothe class status of the different respondentgroups. Later studies have examined genderand class on a more or less equal footing, andhave sometimes explored the interrelationbetween these factors in concrete media expe-rience (e.g., Press 1991; Schlesinger et al. 1992).In comparison, ethnicity and cultural differ-ence, for a time, received relatively less atten-tion as backgrounds to decoding and mediause, but figure more prominently in recentstudies (e.g., Duke 2000; J. Lewis 1991;Parameswaran 1999). Age has rarely been atthe center of reception and audience studies (orof theory development in other media research)(but see, e.g., Jensen 1990; Press 1991; Tulloch1989). This is in spite of the familiar concernespecially with the potentially adverse effects ofmedia on children and youth, which hasinformed, and helped to fund, many audiencestudies.

Television, again, has been the preferredmedium of decoding studies. The centrality ofthe medium in the field as such may have beenreinforced in research on decoding by twodevelopments in the social context of research.For one thing, deregulation, which made more,especially commercial, formats available, par-ticularly in the European setting, raised ques-tions concerning the international ‘cultivation’of local cultures and audiences, which might beaddressed by in-depth audience studies. Foranother thing, almost simultaneously, researchseemed to rediscover ‘pleasure’ as a legitimatetopic of research – with television seeminglyoffering the bulk of pleasurable experience forthe public at large. (At the same time, researchon television audiences – whether qualitative orquantitative – has been limited, in part, by itsreliance primarily on verbal methodologies tostudy both pleasure and knowledge arisingfrom an audiovisual medium.)

While the news genre has remained centralin reception studies (for an overview see Jensen

1998), it was joined, not least, by melodrama,including daytime as well as prime-time soaps,as a favorite area of analysis. The cross-culturalwork of Liebes and Katz (1990) was one influ-ential contribution, and was complemented byseveral other theoretical and methodologicalperspectives (e.g., Brown 1994; Livingstone1990; Seiter et al. 1989). Despite their focus ontelevision, previous reception studies have alsotaken up a variety of other media and genres,for example:

• books (e.g., Parameswaran 1999; Radway1984)

• films (e.g., Cooper 1999; Stacey 1994)• magazines (e.g., Hermes 1995; Lutz and

Collins 1993)• advertising (e.g., Mick and Buhl 1992;

Schrøder 1997).

Media in social contexts of action

The third moment or stage of reception – theembedding and use of media in other socialcontexts of action – has produced a smaller andmore heterogeneous body of research. This maybe due partly to the difficulty of operational-izing a vast empirical domain, partly to a wide-spread tendency to conceptualize media asmeans of representation, rather than as re-sources for action.

At issue in this third group of receptionstudies is how, concretely, media are integratedinto both everyday life and various institution-alized practices. In this regard, reception studies have a significant overlap with the issuesaddressed as ‘agenda-setting’ and ‘framing,’�and, to a degree, ‘lifestyles.’ Moreover, qualita-tive methodologies enable reception studies toaddress the stages of ‘socialization’ and ‘insti-tutionalization’ in ways which complement themore common survey approaches in thoseareas.� When it comes to documenting thecontextual processes of interpreting, adapting,and integrating media into social life, a quali-tative, case-based approach may be preferable

Moments of interpretation 163

1111234567891012341115678920123456789301234567894012345678950111

class, gender,ethnicity, age

television,news,melodrama:main areas ofreceptionanalysis

media asrepresentationor resource

� agenda-setting and framing – Chapter 9

� socialization and institutionalization – Chapter 9

to a variable-based approach in this ‘late’ stageof media impact.�

Three examples may indicate the nature ofthis third form of reception research:

1 In her study of print romances, Radway(1984) showed not only that readers produce avariety of textual interpretations, but also thatthey conceive of romances as a resource forliving. Romances could be seen to provide indi-rect advice on married life, and to allow theirwomen readers to insist on their ‘own’ time forreading. Compared to decoding studies, thefocus here is shifted away from the text, evenits interpretation, toward the act of reading andits place in a wider field of social action.

2 In a study of American television news, Iexamined news stories from the viewers’ per-spective, less as accounts than as resources forboth political participation and daily living(Jensen 1986). The findings suggested thatviewers may approach the news genre with adivided, ambiguous consciousness. While argu-ing for the dutiful importance of the daily actof news-viewing, the respondents bore witnessto the limited instrumental value of the infor-mation for voting or any other practicalpurpose.

3 Studies of fan cultures have shown how the‘foundational’ texts of fandom (a feature film,the music of a band) become the object of – aresource for – a range of cultural practices (seeJenkins 1992; L. Lewis 1991). In addition togiving rise to active and creative decodings,partly within groups of readers, the texts areprocessed and redeveloped by fans in their ownwriting, music, or audiovisual production.

What unites these previous studies is theiremphasis on the contextual and socially inter-ested nature of all (media) interpretation.Media reception is embedded in a historical andcultural context, as well as in the immediatecontext of media use; audiences’ interpreta-tions, equally, are prefigured by their ‘interests,’their frames, lifestyles, and trajectories of

socialization. Whereas most reception (andother media) studies have left behind anyKantian notion of the disinterested contempla-tion of reality through aesthetic artefacts,� thetext is often given theoretical priority. One con-tribution of qualitative reception studies hasbeen to elucidate how media are recruited byaudiences as resources to inform and enhancetheir actions in a variety of social contexts.Media use is a specific, reflexive form of socialaction (Jensen 1995).

MEDIA ‘ETHNOGRAPHY’?

A recent development, designed in part to re-emphasize the status of media use as a form ofsocial action, has taken place under the headingof ‘ethnography,’ particularly since the mid-1980s. Most generally, the widespread invoca-tion of ethnography may be understood as oneway of coming to terms with the convergencebetween social-scientific and humanistic inspi-rations – a development that has been especiallynoticeable in audience studies. It may be nec-essary to carve out a specific research position,or niche, which would recognize that media useis simultaneously a discursive and a performa-tive phenomenon. It may also be necessary stillto oppose this position, as a camp, to the dom-inant paradigm (Webster and Phalen 1997). Infact, an earlier edition of Lowery and DeFleur(1995) had credited the coming of a ‘meaningparadigm,’ which, curiously, was remarginal-ized in the third edition of the ‘milestones’ (fordiscussion see Jensen 2000b).

Radway (1988: 363), early on, pinpointedthe underlying problem, and suggested a curefor both textual reification and the decontex-tualized study of audiences:

Audiences . . . are set in relation to a singleset of isolated texts which qualify already ascategorically distinct objects. No matterhow extensive the effort to dissolve theboundaries of the textual object or the audi-ence, most recent studies of reception,including my own, continue to begin withthe ‘factual’ existence of a particular kind of

Media reception164

the act ofreading

� aesthetic contemplation – Chapter 2, p. 27� case-based and variable-based research – Chapter 15,p. 256

text which is understood to be received bysome set of individuals. Such studies per-petuate, then, the notion of a circuit neatlybounded and therefore identifiable, locat-able, and open to observation.

Her less reifying alternative was classic ethno-graphic fieldwork about all the cultural as wellas other social practices of an entire commu-nity, covering work and leisure:

a collaborative project that would beginwithin the already defined boundaries of apolitically constituted municipality andattempt to map there the complex, collec-tive production of ‘popular culture’ acrossthe terrain of everyday life . . . a project thatwould take as its object of study the rangeof practices engaged in by individuals withina single heterogeneous community as theyelaborate their own form of popular culturethrough the realms of leisure and then artic-ulate those practices to others engaged induring their working lives.

(Radway 1988: 368)

Ang went several steps further in assumingthat an ethnographic methodology represents aradically different epistemology from bothqualitative reception studies and quantitativeforms of audience research. Indeed, she dis-credited both positions, since, to Ang, the sci-entific act of categorization is, in itself, an actof violence committed against the public andthe everyday:

From this [ethnographic] perspective, ‘tele-vision audience’ is a nonsensical category,for there is only the dispersed, indefinitelyproliferating chain of situations in whichtelevision audiencehood is practised andexperienced – together making up thediffuse and fragmentary social world ofactual audiences.

(Ang 1991: 164)

It seems difficult to specify how then, concretely,empirical studies should proceed. Nevertheless,as a theoretical point, this line of argumentbecame influential. Also Radway’s (1988) less

radical position implied a considerable compli-cation of the task at hand – of the relevantobjects of analysis, of the collaborative researchactivity, and of the eventual interpretation of‘findings,’ if that term is appropriate.

The record suggests that media studies haverarely delivered the sort of projects outlined by Radway (1988). Instead, there has been atendency, in some publications, to describeassorted qualitative methodologies, includinginterviewing, as the equivalent of ‘ethno-graphy.’ This, in turn, has provoked severe criticism, particularly of the standards of datacollection, analysis, and reporting, summed up in the conclusion that ‘“ethnography” hasbecome an abused buzz-word in our field’ (Lull1988a: 242).

In response, some researchers (e.g., Drotner1994, 1996) have attempted to reclaim the‘ethnographic’ terminology, not only by ack-nowledging the technical requirements of theapproach (e.g., Schensul and LeCompte 1999),but by suggesting that it has its own method-ological, and perhaps epistemological, speci-ficity. Accordingly, ‘ethnography’ might implya focus, not on media, but on the social prac-tices and gendered identities which media helpto constitute across contexts. Still, it remainsunclear how this approach would differ fromthe qualitative methodologies and construc-tivist epistemologies of most other receptionstudies. More seriously, perhaps, ‘media ethno-graphy’ appears to be a contradiction in terms– it assumes that an inclusive and holistic studyof a social group or context could, neverthe-less, legitimately predefine media use as itsempirical focus. If, in addition, ‘media ethno-graphy’ would propose to cover all moments ofthe communicative process, not just audiences,it would become a misnomer for a wholevariety of multimethod approaches� to thepractice of mediated communication as such,already in the literature (e.g., Aron 1998; Gayet al. 1997).

One problem with media studies with an‘ethnographic’ ambition is that they tend toconfuse the inevitably limited field of empirical

Media ‘ethnography’? 165

1111234567891012341115678920123456789301234567894012345678950111 � multimethod research – Chapter 15, p. 272

inquiry with the more inclusive domain oftheoretical interest. Hence, it becomes hard to justify what not to study – a dilemma thatmay lead either to ‘data death’ or unwarrantedclaims about doing ‘ethnography.’

A more pertinent alternative is to describereception analysis and so-called ‘ethnographic’approaches as complementary varieties ofmultimethod, as well as multidisciplinary, ap-proaches to the audience experience and use ofmedia. Figure 10.2 indicates two such proto-typical varieties. Both varieties take the mediaas their methodological focus, but emphasizeeither the distinctive uses of a medium or genreby several audience groups in different contexts,or the uses of several media and genres by onespecific social group across contexts, who thusdefine their culture and lifestyle, in part, withreference to media. In both cases, the receptionand social uses of media are, to varying degrees,analyzed and interpreted with reference to thewider historical and social context embeddingboth media and audiences. The methodologicalorientation of media research must be towardthe specific difference media make in society,even while its theoretical frame of reference isthe entire social system which shapes, and is

shaped by, media. In methodological terms,meaning flows from media to society; in theo-retical terms, it flows from society to media.

The upshot of debates since the 1980s on‘ethnography’ is that anthropology still hasmuch to offer to the media field in the way oftheory and methodology. The abiding relevanceof the anthropological tradition has been indi-cated, as well, by studies of media audiencesoriginating outside the media field as such (e.g., Dickey 1993; Lutz and Collins 1993).However, the comprehensive, classic techniqueof ethnography, as applied to research in bothpremodern and modern settings, seems lesssuited for the specific purposes of media andcommunication research. Other recent contri-butions to media and cultural research havesuggested additional, alternative ways of con-ceptualizing audiences with reference to theirexercise of agency in different social and his-torical contexts (e.g., Abercrombie and Long-hurst 1998; Butsch 2000; Lembo 2000).

MEDIA DISCOURSES AS RECEIVED

A recurring issue – in ‘ethnographic’ studies, inreception analysis, but also in quantitative

Media reception166

Audience 1

Medium/genre 1

Medium/genre 2

Medium/genre 3

METHODOLOGICAL ORIENTATION

THEORETICAL ORIENTATION

Medium/genre Audience 2 Audience

Audience 3

Figure 10.2 Two varieties of reception study

research traditions – has been the status of thetext in audience studies. A specific aim of qual-itative audience research has been to develop a conceptual understanding of the media–audience nexus that would not privilege one orthe other side. From the outset, reception analy-sis was motivated by a concern that, in human-istic textual studies, the media text was beingreified, or at least privileged as a source ofevidence. In response, some researchers haveworried that reception studies would them-selves lose sight altogether of the texts whichare, in fact, received and interpreted (e.g.,Brunsdon 1989). In this regard, receptionstudies might repeat the mistakes of some quan-titative audience research that has neglected the specific cultural forms in which ‘content’ iscommunicated. Such worries were reinforcedby influential theoretical contributions whichasserted that (media) texts are ‘empty,’ waitingonly to be filled with meaning by audi-ences acting as ‘interpretive communities’ (Fish1979). Other interventions seemed to over-emphasize, in semiotic terms, the ‘polysemy’ ofmedia texts – the variability and indeterminacyof the meanings which audiences mightattribute to them (Fiske 1987). Eco (1976: 150)even suggested the possibility of ‘semioticguerilla warfare,’ as performed by media audi-ences as well as analysts.

The challenge for reception analysis hasbeen, and remains, to perform audience-cum-content analysis – comparative studies of mediadiscourses and their interpretation and uses byspecific audiences. This section presents someof the key concepts which have been developedfor this purpose, before describing the severaltypes of ‘texts’ which are relevant for empiricalreception studies.

The audience in the text

An important legacy of literary theory forempirical reception analysis has been the notionof the reader, or ‘audience in the text.’ Whileresembling earlier notions of ‘point of view’ inliterary genres, the argument is that texts liter-ally anticipate and ‘inscribe’ readers into theirstructures. The idea was developed into ana-lytical tools within several research traditions –

by 1960s structuralism and semiotics on howtexts ‘enunciate’ their message to readers,� andperhaps most influentially by German receptionaesthetics. Jauss (1982) advanced the conceptof a ‘horizon of expectations’� within which areader encounters a text; Iser (1976) identifiedthe structural ‘blanks’ of a text as an invitationfor the reader to complete it, and later outlineda more inclusive ‘literary anthropology’ (Iser1989). Eco (1987), further, contributed a dis-tinction between texts which, respectively, are‘open’ or ‘closed’ in terms of their range ofpotential decodings. Eco also distinguishedbetween ‘interpreting’ a text along the lines sug-gested by its structural attributes, and ‘using’ atext for the reader’s more or less idiosyncraticpurposes – perhaps reading it against the grain.

Whereas literary reception studies haveremained largely preoccupied with the textsthemselves, a number of its concepts have thusproven applicable to the study of media audi-ences. Here, the concept of ‘interpretive com-munities’ has been particularly influential. Ahelpful redevelopment for empirical purposeshas been Potter and Wetherell’s (1987) moreoperational notion of ‘interpretive repertoires.’Rather than assuming that individuals ‘belong’to delimited communities, the latter notion sug-gests that they rely on a range of repertoires,depending on their own background, the textat hand, as well as the particular context ofmedia use. Accordingly, data analysis can tracethe semantic relations, metaphors, and otherinterpretive procedures that different respon-dents employ (e.g., Jensen 1990). As such,‘interpretive repertoires’ avoid the implicationeither that ‘interpretive communities’ aremonolithic formations, or that they are consti-tuted entirely ad hoc, for each text and reading,or by analogy to fan groups. One promisingperspective for audience research, and formedia studies generally, is that such interpretivecategories may be mapped onto sociodemo-graphic categories, thus enabling research toexplore both discursively and socially specificaspects of mediated communication.

Media discourses as received 167

1111234567891012341115678920123456789301234567894012345678950111

polysemy

receptionaesthetics

textual blanks

open and closed texts

interpretationsor uses of texts

� enunciation and semiotic analysis – Chapter 8, p. 128

� horizon of expectations – Chapter 2, p. 22

interpretivecommunitiesandrepertoires

interpretiveand socio-demographiccategories

The texts in front of the audience

Compared to literary theory which, by tradi-tion, has centered on the singular artwork,studies of the technological media must addressa broader variety of content structures andformats. Moreover, some of these formatsemerge from the activity of the audience – mostnotably in ‘interactive’ media, but also, forexample, in the case of radio and television flowwhose elements are selected and combined byaudiences. While Chapters 7 and 8 cover theanalysis of various types of media contents, itis appropriate to review the full range of mediadiscourses in the present context of receptionstudies.

Decoding studies in particular, as notedabove, have often examined television pro-grams and other delimited media texts. In addi-tion, one may identify the following discursivetypes:

• Discursive elements. Although receptionanalysis has most commonly involved entirenarratives or news stories, some studies havetraced the presence of particular themes orimages in the media and their reproduction byaudiences (e.g., Philo 1990). In this regard,reception research was preceded, and influ-enced, by studies of recall and comprehension�

(e.g., Findahl 1985).

• Singular texts. Media content is perhapsbest understood generally as ‘discourse’� – thesocial uses of language and other signs. Still,individual texts may call for a sustained recep-tion analysis, because of their presumed impacton the public (e.g., Cantril et al. 1940), or, forinstance, because of their prolonged presence as part of ‘international’ media culture (e.g.,Gripsrud 1995).

• Genres. A central aim of qualitative recep-tion studies so far has been to explore the audi-ence response to generic attributes acrossindividual media texts. The purpose is todescribe in depth what may be certain generalinterpretive procedures that audiences apply to

major genres. The ambition is found in muchdecoding research, and has been developedexplicitly in studies of media as social resources(e.g., Jensen 1986; Radway 1984). Anotherreason for focusing on genres is that their modeof address, in addition to their form andcontent, anticipates particular uses of media insocial contexts.

• Media. Compared to their ‘contents,’ otheraspects of media – as means of structuring dailyroutines, or as physical artefacts – have receivedrelatively less attention. Recently, such aspectshave been addressed, for example, in work onthe place of media within everyday life (for anoverview see Silverstone 1994).

• Flow. Like genre, flow (Williams 1974) intelevision, radio, and some other media, pre-sents itself as a strategically important level ofanalysis. Flow captures a distinctive feature ofan important portion of all mediated commu-nication. Yet studies so far have been compar-atively few in qualitative research. One reasonpresumably is the difficulty of operationalizingthe audience experience across several contentelements, across channels, and over time.Jensen (1994) outlined an approach to docu-menting the concrete movements of viewersacross the several channel flows of a televisionuniverse, what could be described as a superflow. This record lent itself to an analysis of theresulting viewer flow with reference, forexample, to (narrative) structure and recurringthemes. A model of the several types of flowinvolved in media use is given in Figure 10.3.

• Hypertexts. The discursive structures ofcomputer media are comparable, in certainrespects, to the flow of broadcasting. A chal-lenge for current reception studies is to movebeyond the manifest structures and narrativesof virtual environments, to the experiences anduses by their ‘audiences.’

• Media environments. At the most generallevel of analysis, the media constitute a culturalenvironment of sorts. Its complexity may seemforbidding, but its interrelations remain acentral topic for further research. One optionis studies of the ‘life history’ of a news item(Deacon et al. 1999) (or fictional event, or

Media reception168

� recall studies – Chapter 9, p. 144

� two definitions of discourse – Chapter 7, p. 106

resonant metaphor) with reference to the con-texts in which it is interpreted and appropri-ated by audiences. So far, intertextual studieshave gone some way toward defining issues andrelevant approaches (Bennett and Woollacott1987; Gay et al. 1997; Pearson and Uricchio1990).� At the same time, the blurring of thedivisions between ‘text’ and ‘context’ in virtualenvironments is adding to the complexity andchallenge.

THE MEDIA–AUDIENCE–SOCIETY NEXUS

The trajectory of reception analysis may besummarized with reference to the concreteempirical materials being examined within different varieties of audience research. Whatdistinguishes reception studies, perhaps most ofall, in comparison with the majority of thequantitative milestones, is their focus simulta-neously on media discourses and ‘audience dis-courses,’ i.e., the statements and actions ofaudiences which serve as evidence about theirexperiences and uses of media. The conceptu-alization and operationalization of these vari-ous discourses call for specification in furtherqualitative empirical research.

In an overview of different varieties of com-munication theory, Meyrowitz (1993) distin-guished between conceptions of media as eitherconduits, languages, or environments. The dualstatus of media as channels and texts – withintransmission and ritual models of communi-cation – is increasingly recognized as part of the convergence between social-scientific andhumanistic communication studies. A betterunderstanding of media, in addition, as a cul-tural environment is one of the next challengesalso for reception studies. In meeting this chal-lenge, research will have to consider at leastthree types of analytical objects:

1 This chapter has referred to the contents ofmedia as discourses – vehicles of potentialmeaning which are actualized and enacted byaudiences in various everyday circumstances.The interpretive actualization of media contenthas also been approached in reception studiesas discourses, as analyzed through variousqualitative methodologies.

2 Media are material artefacts – they have avery concrete presence in the space, time, andinterpersonal relations of everyday life. Chap-ter 1 noted that the modern media may bedefined as institutions-to-think-with. However,in their local and distributed settings, they arealso objects-to-think-with (Lévi-Strauss 1991[1962]). Relatively little attention has beengiven so far to this aspect of reception, eventhough some studies have examined the place oftelevision sets in the family context in differentcultural settings (Lull 1988b) and the repre-sentation of television as a new element of socialspace during its introduction (Spigel 1992). TheWalkman (Gay et al. 1997) and mobile com-puting devices also suggest themselves as objectsof research in this perspective.

3 Media, finally, serve to constitute practices.Not only are media received and applied with-in particular social and cultural environments,but they help to constitute those environ-ments, as also suggested by ‘medium theory’(Chapter 2). The media enter into specific formsof interaction, association, and community – invariable contexts from the local to the globallevel. At this point, media research blends intoother forms of social and cultural research.

The media–audience–society nexus 169

1111234567891012341115678920123456789301111234567894012345678950111 � intertextuality and hypertext – Chapter 11, p. 187

Channelflows

Audience flow

SUPERFLOW

Figure 10.3 Three flows of media use

In conclusion, audiences, broadly speaking,constitute a strategic nexus between media and their social contexts. It is for this reason,in part, that audience research has been acentral ingredient of media and communica-tion studies, both its quantitative and, more

recently, its qualitative variants. Other types of research have gone on to address the widercultural and historical contexts of mediatedcommunication. These research traditions arethe topics of the next two chapters.

Media reception170

FROM MEDIA, TO COMMUNICATION, TOCULTURES

Contextualization and recontextualization

An important characteristic of ‘media’ researcharound 2000 has been a renewed emphasis onthe practices of ‘communication.’ Informationand communication technologies support mul-tiple forms of interaction at many levels of thesocial structure. Neither theoretically norempirically can the media be understood as sep-arate entities – add-ons or plug-ins. Instead, thetechnological media are necessary infrastruc-tural resources in maintaining and structuringmodern societies (Giddens 1984; Thompson1995).

The moments of mutual constitution, con-crete interchange, and structuration may beexamined as ‘contexts.’ These contexts extendfar beyond the stages of impact of media onparticular audiences.

This chapter reviews selected research trad-itions which examine the geographical, tech-nological, institutional, and discursive contextsof media – linkages between media and the restof the social structure (even while not all of

these traditions may subscribe to the meta-theoretical perspective of ‘contexts’). The logicof the presentation proceeds from subculturesand communities, via the nation-state as anessentially modern context that has dependedupon print and broadcast technologies, toissues of transnationalism and postcolonialism.The final section examines the (still unresolved)status of those social ‘contexts’ which are estab-lished in and of computer media networks.While frequently at odds with each other intheir theoretical and political assumptions, eachtradition has been important in addressingquestions of media impact and determination,not at the level of individual response or social-ization, but at the institutional and culturallevels of social structuration.

From the classic question regarding the rel-ative determination of social life by either eco-nomic, political, or cultural factors, follows thequestion of how the media, in particular, affectand respond to their various social contexts.The distinctive, reflexive function of media maybe summed up, first, as ‘contextualization.’ Themedia relate issues and events in the world toeach other, and allow audiences to contemplate

1111234567891012341115678920123456789301234567894012345678950111

Contexts, cultures, and computersThe cultural contexts of mediated communication

Klaus Bruhn Jensen

11

• a definition of key concepts relating media to the rest of society: context and culture• a review of studies concerning different levels of the cultural contexts of media: national, local,

intercultural, and global• an examination of research on subcultures, cultural imperialism, and postcolonialism• a special section on computer media and virtual cultures.

contexts as nexusbetween

media,cultures, and

societies

and, perhaps, act upon such relations – withreference to a specific level or domain of socialreality. Second, the media can serve to ‘recon-textualize’ those events and issues, indicating arange of alternative perspectives and interpre-tations.� With reference to science and politics,Rorty (1991: 94) has suggested that these insti-tutions perform recontextualizations whichrange ‘from instinctive revision of intentionsthrough routine calculation toward revolution-ary science and politics.’ Although Rorty’s posi-tion tends to go overboard in its postmodernisttextualization of social issues (for a critique, seeJensen 1995: 180–185), the concept of recon-textualization is helpful if understood as a con-crete activity which both media practitionersand audiences perform. By contextualizing andrecontextualizing the world, the media partici-pate centrally in maintaining society, and theydo so across a variety of settings where cultureis constructed and contested.

Culture and cultures

In order to anticipate some familiar problems indefining ‘culture,’ a few conceptual distinctionsare in order. Kroeber and Kluckhohn (1952), bythe 1950s, noted as many as 164 definitionsof culture. Similar problems affect commu-nication studies, where Anderson (1996) hasrecently identified 249 definitions or theories of communication.

An inclusive, modern definition of culturedates from the Romantic age, specifically thework of Johann Gottfried Herder (Ideas for aPhilosophy of the History of Mankind, 1784).Culture is that which simultaneously unites and differentiates human beings (for an over-view see Williams 1983). On the one hand, all human beings share culture – the ability toexperience, reproduce, and communicate mean-ing. On the other hand, humans can be distin-guished according to the specific meanings with which they align themselves – the culturalformations to which they ‘belong’ or ‘owe’ theiridentity and solidarity. While cultures, in theplural, have traditionally been coupled with

particular social groups in a geographical loca-tion, cultures have increasingly come to extendacross space and time with the advent ofwriting, print, and later communication tech-nologies. It is this extension, not least, whichrequires a careful account of the several culturalcontexts in which media are embedded.

An additional complication for media stud-ies arises from the distinction between culture as product and culture as process (e.g., Jensen1986: ch. 5). Culture has been understood, onthe one hand, as specific ‘products’ of humancreativity, as ‘containers’ of culture, and often asprivileged means of insight for social elites. Withindustrialization came a wider range of culturalproducts which addressed a mass public, as well.On the other hand, culture has been examinedas a ‘process’ which pervades all of mental andsocial life. This process was traditionally under-stood literally as the cultivation of the humanspirit either in a secular or a religious sense.Increasingly, reference has also been made to themundane and lived reality of the everyday.

This last reorientation has itself been asso-ciated with the industrialization of cultureduring the modern period. Industrialization hasentailed a recognition – for commercial as well as ideal purposes – of popular, informal,and emergent cultural practices alongside thelong-established and institutionalized arts. Inacademia, the reorientation has entailed a rela-tivization, in part, of both classical learning andWestern civilization. The relativization has beenperformed over the past century most notablyby anthropology and, more recently, by re-search fields on (mass) communication and(popular) culture. The distinction betweenculture as product and as process has never-theless been persistent, and has challengedresearch to conceptualize and operationalizecommunication in ways that capture both sidesof the cultural coin of exchange.

One implication for media and communica-tion research has been a call, increasingly, tostudy not only cultures – in the plural – butcultural contexts of multiple kinds. These con-texts lend themselves to empirical studies asmoments of interchange between media andother social and cultural circumstances, broadlyspeaking. Accordingly, in the following reviews,

Cultural contexts and computers172

� (re)contextualization and mediated reflexivity – time-in and time-out culture, Chapter 1, p. 5

164 defini-tions of

culture; 249definitions of

communi-cation

culture unites and

differentiatesall humans

culture asproduct andprocess

both media production and reception are takenup as sites of the production of culture. Keyissues in research as well as public debate – con-cerning cultural identity or authenticity, andconcerning economic and political power –have frequently revolved around the produc-tion–reception nexus. What is the relative deter-mining influence of, for example, transnationaleconomic infrastructures, national political reg-ulation, and local cultural formations in theexperience and use of mediated communica-tion? For each of the contexts below, a broadworking definition of culture is adopted: cul-ture has been studied as reproduced, cumu-lated, and sedimented structures of meaning.The empirical questions have centered on thespecific contributions the media make to cul-tural production and social interaction.

NATIONS AS MODERN CULTURES

The rise of the technological media coincideshistorically with the rise of the nation-state. Farfrom one being a singular determinant of theother, both developments entered into a third,long-term shift toward the modernization�

of social relations in most areas of humanactivity. Whereas the scope and current statusof modernization remain debated, its mainingredients are fairly well established (for anoverview see Thompson 1995):

• industrialization and capitalization of thematerial economy, along with a growing divi-sion and rationalization of labor, leading intovariable phases of market competition, incor-poration, imperialism, and conglomeration;

• democratization and bureaucratization ofthe institutions and practices of politicalrepresentation and government;

• secularization of the cultural forms ofexpression, including the securing of nichesfor non- or anti-religious reflexivity, and therecognition of popular alternatives or com-plements to the fine arts.

From the outset, the modern nation-statewas a contested entity. It developed, moreover,

in conjunction with other nations as part of anew world economic system, centered in West-ern Europe. Wallerstein (1974: 348) has madethe point that, in contrast to previous suchsystems – for example, in the Middle East orChina – this international system of capitalistproduction and trade did not grow into anempire in any traditional sense. The modernworld system, taking shape from the sixteenthcentury to c. 1900, developed as an economic,not a political entity. Hence, it allowed for, andsupported, not only a global division of labor,but the formation of political structures at thelevel of the nation-state (see also Wallerstein1980, 1989).

In the domain of culture, nation-buildingwas supported by arguments and narrativesconcerning the legitimacy of such arrange-ments, as widely disseminated by print media.The process of dissemination, in turn, was facil-itated by the fact that these cultural artefactscould be distributed as commodities. Morepeople – developing into a mass public – movedabove the level of subsistence, and became ableand willing to pay for mediated experiences.

Anderson (1991) has offered the helpfulnotion of ‘imagined communities’ to character-ize nations not only in terms of geography and constitutions, but as cultural and discur-sive formations. In one sense, all communities,from face-to-face to virtual interaction, are, toa degree, imagined (Jensen 2001). Anderson(1991), however, described in further detail themany forms of mediated communication whichcame to serve as means of symbolic control andrepair within nations – from newspapers andnovels, to museums, maps, and the census.Moreover, he noted the role of print technolo-gies in stabilizing and enforcing national lan-guages, over and above the sacred languages of scripture, and as means of collective self-definition.

The media did not only help to replacereligious frames of reference with national ones,and to question the status of ‘divine’ rulers. Themedia also became unprecedented resources for redefining the future – for oneself and forsociety. If historical time and its subjects wereno longer subordinated to a divine cosmology,then the future presented itself as an open field

Nations as modern cultures 173

1111234567891012341115678920123456789301234567894012345678950111 � modernization – Chapter 3, p. 48

culture ascommodities

imaginedcommunities

of action, both for individuals and for socialinterest groups. The future had to be made. Itis this inherent openness which some theoristshave identified as the essence of modernity(Berman 1982; Giddens 1991; Huyssen 1986),recognizing, as well, that openness is a sourceof existential uncertainty and perhaps ambiva-lence. It is this same openness which mediahave been called upon as institutions to inter-pret, with the nation as the new center of animagined universe.

The history of the modern nation-state is, in large measure, a history of expansionism and warfare, as supported, in part, by mediapropaganda.� Also internally, the definition of what is a nation, and what should be thepolitical and economic rights of citizens, hasbeen subject to continuous controversy, carriedon partly through media. In the Europeansetting, Habermas (1989 [1962]) described the development of the classic public sphere.�The eighteenth-century press was instrumentalin defining both citizenship and nationhood.(Habermas, interestingly, also noted the com-plementary role of interpersonal communica-tion in clubs and coffee-houses.) Regarding theearly history of the press, moreover, studiesfrom both the U.S.A. and the U.K. have sug-gested that left-radical papers lost out in theeconomic competition and political conflictsthat accompanied the establishment of the press(Curran and Seaton 1997; Schiller 1981). Theseconditions, arguably, resulted in less diversenational media systems also in a longer andcurrent perspective.

Since the early 1990s, the formation of arenewed public sphere has been in progress, fol-lowing the transition of Eastern Europeannations, and of nations within the former SovietUnion, to various liberal press systems. At thishistorical juncture, media, legislators, as well asaudiences face many more technologies and arange of policy choices regarding public serviceand market-driven organizations (see, e.g.,Kelly and Shepherd 1998; Sparks 1997). Oneconsequence of the transition has been a height-ened awareness, also in research, regarding sub-

merged national and cultural sentiments (whichsocial science had largely failed to predict asforces behind 1989 and its aftermath (K.B.Jensen 1999: 427)). Another consequence hasbeen publications which aim to document andrecuperate earlier periods of these cultures, forexample, of the Baltic nations before Sovietsupremacy (Høyer et al. 1993).

During the twentieth century, broadcastingbecame a central reference point in the con-struction and maintenance of national culture.Radio, and later television, have depicted espe-cially the official life of the nation. Broadcastschedules have also served, more instrumen-tally, to maintain cycles and rhythms of sociallife, from the twenty-four-hour cycle of thefamily to the seasonal ceremonies of the nation(sometimes understood as a ‘family’) (Scannell1988; Scannell and Cardiff 1991).

In developing countries, nation-building hasbeen a special concern for policy-makers as wellas media practitioners. The diffusion of com-munication technologies� has commonly beenperceived as a necessary condition for materialprogress and, more generally, modern lifeforms. A prototypical example was the 1975 to1976 SITE experiment in India, where com-munity television sets were introduced into2330 largely rural villages. The purpose was totransmit especially educational and informa-tional programs on agriculture, health, andfamily planning (for an overview see Journal ofCommunication, 29(4), 1979: 89–144). Beyondthe concrete goals of enlightenment, an import-ant structural effect of such projects has beento promote particular understandings ofnationhood and citizenship.

It bears repeating that nation-states havedeveloped in conjunction, as part of a trans-national system. While later sections detailresearch traditions that have focused ontransnational issues, it is useful here to distin-guish the several flows of communication thatreach audiences in their national context.Indeed, by far the majority of all mediated com-munication is still transmitted by nationalsenders to national audiences. Figure 11.1 indi-cates three flows:


� propaganda studies – Chapter 10, p. 159

� public-sphere model – Chapter 1, p. 7 � diffusion of innovations – Chapter 9, p. 140

1 National flow, which refers to foreign orimported (over and above home-produced)contents that are adapted by distributors to thetastes and cultural background of the nationalmarket. Importantly, such adaptation includesnot merely translation, but the recontextualiza-tion or ‘domestication’ of news items (Cohen et al. 1996), and the redevelopment of formatsfor fiction series and other entertainment (e.g.,O’Donnell 1999; Silj 1988). In this respect,national media, from early newspapers andpopular literature onwards, have always beentransnational.

2 Bilateral flow, which covers direct, regular,and ‘accidental’ transmission to audiences in a

neighboring and often culturally affiliatedcountry.

3 Multilateral flow, which refers to explicitlyinternational forms of communication, such as CNN and MTV, as well as many Internetservices.

SUBCULTURES AND LOCALITIES

Below the national level of social organization,media studies have specifically addressed twotypes of ‘communities of interest’ – local orregional forms of culture and communication,and subcultures. In both cases, the implicitpoint of reference has tended to be the nation-

Subcultures and localities 175

1111234567891012341115678920123456789301234567894012345678950111

COUNTRY 1 COUNTRY 2

Production

Distribution National

Reception

Bilateral

Multilateral

COUNTRY 3

Figure 11.1 Three media flows in and between countriesSource: Sepstrup 1989; see also McQuail and Windahl 1993: 225

state. The nation could be defined as a (metro-politan) center with an under-recognized peri-phery of self-reliant cultural formations. Thiscenter could also be understood as a historicalform of social organization that has grown outof touch with its constituencies in terms of age,class, gender, and ethnicity. The nation mighteven be considered an obstacle to the full cul-tural realization of such local communities andlived constituencies.

Research interest in communities and sub-cultures dates back at least to early sociology.Tönnies (1974 [1887]) presented a distinctionwhich has remained influential:

• Gemeinschaft – homogeneous, communitar-ian, and non-contractually governed socialformations;

• Gesellschaft – heterogeneous, individualist,and contractual arrangements.

The two terms do not refer to distinct types ofsociety, but to forms of social organizationwhich, to varying degrees, characterize partic-ular historical cases. In addition, the Chicagoschool sociological tradition� has been animportant source of ideas for research on ‘com-munities’ since the early twentieth century.

The two prototypes also represent imaginedforms of social organization. Since the Roman-tic age, modern societies and cultures have,from time to time, entertained the notion ofreturning to, or reintroducing elements of,Gemeinschaft in the modern context. Thepotentials and implications of local agency havecontinued to occupy social theory. In mediastudies, various community media, especiallylocal radio and television, attracted renewedattention from the 1960s. This was due partlyto new technological possibilities of produc-tion and distribution. However, research alsoseemed to be attracted by the democratizingand locally empowering potential of such com-munication (for an overview see Jankowski etal. 1992). In some cases, studies sought todevelop this potential further through actionresearch.�

Studies of local and community media haveprimarily served to document what remains a significant level of social communication. Incomparison, research on subcultures has beenthe more theoretically innovative tradition. Subcultural studies developed particularly fromthe 1970s, and examined primarily youthgroups which had manifested themselves,through their cultural practices, from the 1950s(for an overview see Gelder and Thornton1997). As the term suggests, these cultural for-mations could be seen as somehow subordi-nated to a national culture of sorts. In response,subcultures might express and gratify particularcultural preferences and lifestyles alongside thenational provision. Alternatively, subculturesmight, in a stronger sense, represent full-fledged social and cultural universes. As such,they would be separate from, and largely inopposition to, a perceived mainstream nationalculture, in part with reference to the politicaland economic interests underpinning the main-stream.

It is especially this last conception whichsubcultural research has defended and ex-plored. A central object of analyis has been sub-cultural universes that are simultaneously localand transnational, being practiced in severaldifferent geographical locations and generatinga sense of community and solidarity acrossnational borders. Concretely, variants of rock-’n’roll have been a center-piece, not only of fangroups, but of international social movementsof some scope, from anti-Vietnam War pro-tests to Live Aid concerts. The actual culturalartefacts have generally come out of Americanor multinational media industries, but thecommon perception of members of these sub-cultures (and of many researchers) has beenthat the media are thus redefined and reappro-priated for alternative purposes.

A particularly influential approach to sub-cultures has been associated with ‘the Birming-ham School’ (Hall et al. 1980).� Beyond itsfruitful combination of humanistic and social-scientific inspirations, the cultural studies trad-ition has been premised on the definition ofculture as a constitutive ingredient of social


center andperiphery

� Chicago School sociology – Chapter 3, p. 43

� action research – Chapter 16, p. 287 � Birmingham cultural studies – Chapter 2, p. 39

conflict and change. Rather than being sec-ondary to, or derived from, economic or political matters, subcultural practices are anexample of culture becoming a site of socialstruggle in its own right. To cite the title of aninfluential volume, the members of subculturesexercise ‘resistance through rituals’ (Hall andJefferson 1975).

Certainly, resistant cultural practices can bemore or less self-consciously and proactivelysocial in their orientation. At least in somerespects, membership of a subculture may not becategorically different from fandom. Critics suchas Gitlin (1997) have described cultural studiesresearchers more broadly as disappointed revo-lutionaries who find it ‘reassuring to detect“resistance” saturating the pores of everydaylife’ (p. 30). Furthermore, subcultural researchhas not been without its romantic and self-right-eous streaks, re-emphasizing the authenticityand legitimacy of specific subcultures. Conse-quently, there has been a relative neglect of theirposition within the wider social structure. Asnoted by Middleton (1990: 161) with referenceto music-oriented youth subcultures, studieshave given less attention to ‘the subcultures’relationships with their parent cultures, with thedominant culture and with other youth cultures.’Nevertheless, subcultural studies have been suc-cessful in conceptualizing and interpretingculture as a multiply contextual practice, thusextending Williams’s (1975 [1958]) agenda forstudying culture as ‘a whole way of life.’

A special focus of subcultural research, not surprisingly, has been popular music. Post-war subcultures emerged as part of anti-authoritarian turns within the family as well as in educational, cultural, and other socialinstitutions. The generalized revolt was sym-bolized and articulated in innovative rhythmsand lyrics. One difficulty, however, has been toaccommodate the analysis of music as – music.Middleton (1990), for one, argues that subcul-tural theory has tended to rely excessively andabstractly on the identification of homologies –structural similarities between the organiza-tion of, for instance, musical materials and the social interactions around the music, as inreferences to rock ’n’ roll as ‘screw and smashmusic’ (p. 158).

In fairness, music is one of the most under-researched aspects of mediated communicationacross the entire field. The resulting gap intheory and evidence is presumably due, at leastin part, to the neglect of popular music by trad-itional disciplinary musicology. Compared to,for instance, literature departments, popularculture only very slowly seems to be makinginroads into most departments of music.However, it is incumbent on the media field todevelop more comprehensive theories of com-munication which incorporate the differentvehicles and modalities of communication –verbal, visual, auditive – also in view of themassive presence of music in the current mediaenvironment as a whole.�

BETWEEN CULTURES

Cultural imperialism

Given the role of culture, often in the guise ofreligion, throughout centuries of warfare andinternational conflict, it is hardly surprisingthat the modern media have been recruited forsimilar ends. Media research has been alert tothe potential consequences, and to the possiblerole of research in documenting and counter-acting abuses. To reformulate the military his-torian, Karl von Clausewitz, who concludedthat war is a continuation of politics by differ-ent means, many communication studies have been motivated by a concern that themedia might continue both politics and war bycultural means. Specifically, the media might serve to subordinate non-Western cultures in anextension of colonialism, which was officiallybeing dismantled after 1945, and particularlyfrom the 1960s.

Much work in this area of the media fieldmay be summed up as a critique of ‘culturalimperialism.’ While the nuances of the researchand debate are many (for an early overview see Tunstall 1977), the main thesis can be stated briefly. The Western media industriesserve as the agents of a continued imperialism,as exercised most noticeably by the U.S.A. overdeveloping nations, and made possible by the

Between cultures 177

1111234567891012341115678920123456789301234567894012345678950111

analysisthrough

homology

� research on media and music – Chapter 8, p. 135

principles of an international capitalist eco-nomy. For one thing, the relevant technologiesand professional competences have not beendiffused sufficiently to allow for indigenousmedia production in most world settings. Foranother, comparatively cheap news and fictionwhich have already made a profit in theirprimary, Western markets can be marketedmassively and cheaply to secondary markets inthe developing world.

Two aspects of the historical context of cul-tural imperialism research should be noted.First, the general setting was a divided worldin which two superpowers, the U.S.A. and theU.S.S.R., were engaged in a ‘Cold War.’ Againstthis background, the ‘Third World’ could beconsidered a residual to be enlisted, also bycultural means, against the opposing side. Theseveral different philosophies of the press which representatives of all three worlds in-voked were thus clearly informed by additionaland conflicting conceptions of both culture andimperialism, markets and governments. Thedifferent positions emerged most pointedly inthe debates and reports which grew out ofattempts, sponsored by the United Nations andconducted within UNESCO, to outline a NewWorld Information and Communication Order(NWICO) (see MacBride 1980).�

A second background to assessing the cul-tural imperialism thesis is the ambiguousresponse which American and other popularmedia had evoked in various national settings,including Europe (see Webster 1988). At times,advocates of the thesis seemed to resent thischallenge to national cultural establishments, orthey resisted modernization as such, as sym-bolized by contemporary American popularculture. In the European setting, much researchwas conducted which questioned the actual cul-tural diversity especially of television program-ming (for overviews see Richardson andMeinhof 1999; Wieten et al. 2000).

Being part of the revitalization of criticalsocial theory from the 1960s, research on cul-tural imperialism derived much of its theoreti-cal impetus from the ‘political economy’�

tradition which was then taking shape in mediaresearch (e.g., Murdock and Golding 1977).Whereas Schiller (1969) became an inspirationto a generation of critical U.S. researchers, acharacteristic feature of studies in this area wasthe alliances fostered among researchers onseveral continents, including both Latin andNorth America (see Roach 1997). A classicstudy of the ideological implications of DonaldDuck comics (Dorfman and Mattelart 1975[1971]) is representative of one variety ofresearch which performed ‘ideology critique’�of media texts. Perhaps more important wasresearch which documented the technological,knowledge, and other gaps affecting inter-national communication and cooperation (e.g.,Nordenstreng and Schiller 1979), and the spec-ific studies of imbalances in the internationalflow of news that informed the NWICOdebates (e.g., Sreberny et al. 1985).�

Recent publications have come to questionsome earlier positions regarding cultural im-perialism (for an overview see Golding andHarris 1997). Again, this has happened along-side changes in the world orders of politics andeconomy, including the dissolution of the SovietUnion. Media research has largely rejected themore determinist assumptions of the tradition,and has identified, for example, the substantialrole of regional media organizations (e.g.,Boyd-Barrett and Thussu 1992). In addition,studies have documented how local culturesand communities can both accommodate andcounteract transnational media. Besides recep-tion studies (e.g., Biltereyst 1991; Liebes andKatz 1990), textual and production studieshave also found that ‘local’ contents (albeitsometimes in ‘Americanized’ formats) may havethe greatest audience appeal (e.g., O’Donnell1999; Silj 1988). Faced with more recentresearch, the cultural imperialism tradition hasbeen hampered by the generality of its argu-ment and by its (paradoxically) limited researchinterest in the specific cultures at issue.


the ThirdWorld in the

Cold War

� NWICO – Chapter 16, p. 277

� political economy – Chapter 3, p. 56

� ideology critique – Chapter 2, p. 34

� news flow – Chapter 6, p. 92

Intercultural communication

Several other approaches to differences betweencultures and their ways of communicating aresometimes gathered under a heading of ‘inter-cultural communication’ (for overviews seeMartin and Nakayama 1999; Park 1998).Whereas much of this work emphasizes inter-personal or organizational communication, thissection briefly notes some key points applyingto mediated communication.

The term ‘intercultural communication’ isnormally associated with the ‘codes’ of com-munication, in the broadest sense, as concep-tualized within interpersonal communicationstudies (Gudykunst and Kim 1997; Jandt2001). A common premise is that these codesmay inhibit understanding between people ofdifferent national or ethnic origins. Nonverbalbehavior is an important topic of analysis, as isthe question of how different communicativebehaviors may aid in ‘uncertainty reduction.’Invoking educational and other proactivemeans of avoiding or managing conflict, thetradition operates with such ideals as ‘the inter-cultural person as a model for human develop-ment’ (Gudykunst and Kim 1992: 253). Somereference works in the area do cover techno-logical media (Asante and Gudykunst 1989),but these contributions rarely have a distinctivetheoretical focus, or they attempt to extrapo-late from the interpersonal to the macro-sociallevel of communication.

The discipline of anthropology,� tradition-ally the home of comparative studies of cul-tures, has been less of a direct influence on themedia field than, for instance, sociology or psy-chology. It is worth noting that, despite thepotential of the media field in this respect, com-prehensive comparative studies of media andtheir uses in cultural context are still rela-tively few (see Blumler et al. 1992). Instead, theimpact of anthropology has been more piece-meal, for example, methodologically in the case of debates on ‘ethnography’� as well asin specific theoretical inspirations (e.g., Dayanand Katz 1992).

Perhaps partly for these reasons, intercul-tural and cross-cultural issues have sometimesbeen examined in an ad hoc or ‘applied’ fash-ion. The classic example is the (largely failed)attempt in ‘development communication’ toemploy information and communication tech-nologies to advance economic and political life in the developing world (Lerner 1958;Schramm 1964; for an assessment and dis-cussion see Schramm and Lerner 1976). Inaddition, the cultural imperialism tradition,operating from a different point of the politicalspectrum, could be said to represent an ad hoceffort, since little theory development hasaddressed the ‘cultures’ at stake.

Common to the conceptions of culture andcommunication considered in this volume isthat they grow out of Western models of scienceand scholarship, for historical and institutionalreasons. However, a number of publicationshave explored the potential links of communi-cation theory with other ontologies, includingthe world’s religions (see Christians and Traber1997; Dissanayake 1988b; Kincaid 1987). Insome cases, interventions have been thinlyveiled attempts at promoting and applying reli-gious axioms to contemporary policy issues(Mowlana 1993; see also Mowlana and Wilson1990). In other cases, publications have sug-gested relevant conceptual family resemblancesbetween East and West with a view to theorydevelopment and meta-theory. If, as noted byDissanayake (1988a: 1), ‘a preoccupation withmetatheory is a clear sign that a given disciplinehas attained a certain level of maturity,’ moreintercultural approaches to theorizing ‘culture’and ‘communication’ may contribute to matu-rity and consolidation in the future.

Postcolonial theory

A recent approach to intercultural relations ispostcolonial theory, which has re-emphasizedthe implications of the colonial past for presentsocial interaction and cultural forms. The issueshave been addressed from a variety of perspec-tives by intellectuals in the former colonies aswell as by literary and historical researchers(for an overview see Ashcroft et al. 1995). The most distinctive position, however, has its

Between cultures 179

1111234567891012341115678920123456789301234567894012345678950111

� anthropology – Chapter 3, p. 48

� ethnography – Chapter 10, p. 164

developmentcommuni-cation

non-Westernontologies and(meta)theories

theoretical background in poststructuralist the-ories of discourse.� Compared, for example, toa historical definition of postcolonialism, whichmight focus on economic and political mech-anisms of oppression, a discursive definitionshifts the attention toward narratives andworldviews as cultural means of oppression. In particular, postcolonial theory has beeninformed by a Foucauldian notion of discursiveformations which privilege certain worldviewsover others (Foucault 1972). The position hasbeen influenced, in addition, by the psychoan-alytic theory of Lacan (1977). He has con-tributed the notion that language (and othercultural forms) creates a ‘speaking position,’from which a dominant group differentiatesitself from – excommunicates – their ‘others.’

The classic statement on cultural ‘othering,’which tends to label other cultural formationsas inferior, is Saïd’s (1978) work on ‘oriental-ism.’ From a wide variety of sources and genres,his volume documents a deep-seated tension inhow Western ‘authors,’ in a broad sense, haveexpressed their understanding of non-Westerncultures. These ‘others’ are, on the one hand,appealing, and, on the other, repelling. Thisambivalent experience, next, leads to the perception that the ‘others’ must be mastered,and may more or less legitimately be controlledfor ulterior purposes. While the mechanism of ‘othering’ becomes more controversial andurgent in periods and in settings where severalcultural formations are brought together bymigration, disaster, or conflict, it may be takenas pervasive, and as an obstacle to interculturalcommunication.

More recently, influential contributions tothe area have shifted the emphasis furthertoward a discursive conception of oppressionand power generally (Bhabha 1994; Spivak1988). Despite nuances regarding the extent towhich these authors recognize a potential foralternative viewpoints, resistance, or liberationunder present sociocultural circumstances,Bhabha and Spivak have both concentratedtheir analytical efforts on developing abstractand rather stipulative theories of discourse. Acentral question has been how – through what

discursive forms – an autonomous and gen-uinely postcolonial subject might finally articu-late itself. As a result, these authors could besaid to neglect both the material contexts ofdiscourse and the variable interpretations of discourse by actual, embodied human beings.These factors would seem to be necessary con-ditions in any explanation of how colonialismis reproduced and perpetuated. The discursivefocus has, in turn, evoked complaints fromother researchers in the area, who suggest thatpostcolonial theory may end up as an intro-verted ‘academic glass-bead game’ (Slemon1995: 52).

In the context of media studies, postcolonialtheory might serve as a corrective to the infra-structural and institutional focus of researchtraditions ranging from development communi-cation to cultural imperialism. So far, however,its influence has been relatively limited, and hasmanifested itself particularly in the culturalstudies tradition (e.g., Morley 2000). A keyconcept whose relevance for the media field assuch is becoming apparent is that of ‘diaspora’– the dispersion and diversification of previ-ously more homogeneous cultures acrosslocales. Another question suggesting itself tothe field as a whole is how the term ‘post-colonial’ should be applied: to the relativelyshort period since the postwar independence of many developing nations, to the longerperiod since the original colonization, orbroadly to social and cultural affairs in a worldaffected by diverse and only partially recog-nized forms of colonialism (Ashcroft et al.1995: xv)? This question may take on new relevance at a time when digital means of com-munication are again redefining the terms ofinternational interaction.

AMONG CULTURES

The lines of division between intercultural,transnational, and global communications canbe difficult to draw, both in practice and con-ceptually. Nevertheless, the 1990s witnessed amarked, renewed research interest in the globallevel of mediated communication. This hap-pened partly in response to the wider diffusionof computer and satellite technologies. Indeed,


� poststructuralism – Chapter 2, p. 33

the other

orientalism

diaspora

these developments within communicationswere key to a commonly perceived trend to-ward the globalization of economy and politicsgenerally. Thus, changes in the social contextof media and communication have once againoffered the field new opportunities in address-ing contemporary issues.

To some degree, studies of media and glob-alization have extended earlier work on thetechnologies and institutions of modernity (foroverviews see Morley and Robins 1995;Thompson 1995). In addition, recent work onglobalization has taken some of its inspirationfrom inclusive theoretical frameworks regard-ing either culture (e.g., Appadurai 1996; Han-nerz 1996) or economy and society (e.g.,Castells 1996; Harvey 1989; Held et al. 2000).The difficulties of grasping the nature and scopeof ongoing social transformations may accountfor the somewhat ‘grand’ nature of some ofthese theories. Symbolized by the networkedcomputer, with its potential for ‘interactivity’across space and time, and its radically dis-persed ‘virtual’ cultures, current developmentshave, in particular, made the area of computer-mediated communication a strategic subspec-ialty of research, examined in the next section.

Despite technological and economic change,it is important to bear in mind that ‘globaliza-tion’ is not unique to the present age. Oneoutcome of modernization was the economic‘world system’ (Wallerstein 1974) that devel-oped from the sixteenth century. This systemnever took the shape of any full-fledged politi-cal structure, is comprised of independentnation-states, and is likely to remain that wayfor the foreseeable future. This is in spite of thefact that a variety of supranational treaties andentities have emerged over time, and may becurrently on the rise, as exemplified by theEuropean Union.

In present circumstances, the media con-tinue to occupy a dual position. Media are:

• commodities in a transnational economy• national and local political institutions.

This duality helps to identify some of the con-flicts which enter into processes of globalization– conflicts previously addressed in research

under headings of cultural imperialism andpostcolonialism. It is in the field of culture thatmany of these conflicts are articulated andaddressed, and it is this cultural aspect of glob-alization which has brought media to the foreof current research on globalization (for over-views see Robertson 1992; Tomlinson 1999).

In order to describe the place of media inglobalization, it is helpful to distinguish be-tween the world as a context of action and asa frame of reference (Tomlinson 1999: 11). Asa context of action – including business invest-ments, transborder political mobilization, andtourism – the material world has undoubtedlybecome more interconnected in recent decades.In the media sector, this has perhaps been especially clear on the production side throughthe formation of large conglomerates in aprocess of intensified concentration.� Thenews and popular music industries are amongthe foremost examples of this development(even though cultural industries, including thepress and cinema, have long been large inter-national enterprises) (for an overview seeSreberny 2000).

One countervailing tendency in a compli-cated field is that, in terms of distribution andconsumption, local media systems may beholding their own. For example, the relativeproportion of home-produced content (and itsconsumption) in the culturally central televisionmedium appears to have grown at least in someregions of the world. This may be attributed to‘the enormous expansion of television produc-tion and transmission outside the United Statessince the 1970s’ (McQuail 2000: 232).

It is, however, as a frame of reference thatglobalization applies most directly to media.For most people most of the time, the rest ofthe world is an imagined entity, occasionally animagined community (Anderson 1991) theyencounter through media. What I do to otherpeople and institutions elsewhere in eithermaterial or discursive terms – and what they doto me – only affects us in the most indirect andcumulative of ways. This condition does notrender the world less real, nor does it make the

Among cultures 181

1111234567891012341115678920123456789301234567894012345678950111

globalization

the modernworld system

the world as context of action

� concentration, integration, conglomeration – Chapter4, p. 73

the world asframe ofreference

individual less of an agent – it is a general con-dition of social interaction under modernity,and it is this condition that is accentuated aswell as modified by globalization.

The modern condition has been character-ized by, among others, Giddens (1990) as ‘time-space distanciation.’ The term refers to the factthat media and other types of modern institu-tions allow for larger and more complex socialentities. Interaction in the literal sense of ‘actingtogether’ within a shared frame of referencedoes not equal being present in the same spaceand time. One result of interaction at a distanceis a ‘disembedding’ (Giddens) or ‘deterritorial-ization’ (Tomlinson 1999: 106) of much ofactual social life from co-presence in familiararenas. Presumably, such interaction results ina measure of disorientation for the individual.A central means of reorienting oneself – ofbeing ‘re-embedded’ in various contexts ofaction – is the technological media. From theindividual’s perspective, then, the media areambivalent resources under conditions ofmodernity and globalization. The media helpsupport a social system that is not only complexbut opaque; they also provide means of makingthat system more transparent.

It is through such a multidimensional con-ception of globalization that references to ‘thelocal in the global,’ and vice versa, can be spec-ified and justified. For media conglomerates, the entire globe is a context of action. However,for purposes of product development, con-glomerates also need to consider the globe asan imagined entity. In order to attract audi-ences, who are always localized, producers anddistributors of media commodities must takeinto account the several frames of reference –national, regional, transnational – which enterinto ‘local’ cultures. One lesson from mono-generic transnational television channels, suchas MTV and CNN, appears to be that, whenfaced with national competition, they can beforced to regionalize or customize their product(Roe and De Meyer 2000; see also Volkmer1999). In theories of globalization, this pro-cess has been referred to as ‘glocalization’(Robertson 1995).

In sum, it is important to distinguish be-tween several sectors or levels of globalization:

• economic – which has been ongoing for cen-turies;

• political – which remains limited, at least ininstitutional terms;

• cultural – which has been intensifying inrecent decades.

It is these several aspects of globalization whichcome together in the case of media. Further,they have different implications for differentmoments of the process of mediated communi-cation:

• Media production. The impact of trans-national conglomeration in different media andgenres, while significant, is tempered by theconcurrent expansion of national and local pro-duction and transmission.

• Media discourses. The export of meaning(Liebes and Katz 1990) from the U.S.A. andother global media centers may be reinforcedby national adaptations of similar formats. Buthomogenization is counteracted by the specialand continued popularity of local product inthe local language.

• Media audiences. The reception of ‘foreign’media and genres further serves to mediate theircultural impact. The modes of reception havebeen shaped, in part, by the cultural practicesand conventions of local contexts. A case inpoint is sports (e.g., Coakley 2000; Whannel1992) – a ritualized form of play and inter-national combat which nevertheless remainsanchored in the nation-state. Internationalsports is an ingrained part of national sched-ules and seasons. It also lends itself both tolocal, collective participation on site and to vic-arious, mediated forms of experience.

COMPUTER-MEDIATED CULTURES

Building on the seminal theories of Alan Turingabout the computer as a universal machine, andabout its social uses from visionaries such asVannevar Bush, Ted Nelson, and Alan Kay (for key texts see Mayer 1999), the computerbecame recognizable as a medium particularlyduring the 1980s. This followed the diffusion


time–spacedistanciation,disembed-ding, and re-embedding

glocalization

sports

of personal computers, as suggested by adver-tising introducing it to the general public(Jensen 1993a). It was not until the 1990s,however, that the computer seemed to be rec-ognized, not as a mass medium, but as a newcategory of medium for a mass public. Thishappened following the introduction of theWorld Wide Web and of many more networkconnections in the industrialized world.

Studies of the computer as a medium arecurrently being conceptualized in relation toother forms of communication and culture.Research is simultaneously being institutional-ized in relation to a variety of disciplines andfields across the social, natural, and humansciences. The purpose of this section is tooutline the place of the computer on the currentmap of the field of media and communicationresearch, and to review some of the main con-tributions to conceptualizing and studying thecomputer as a medium (for overviews seeBenedikt 1991; Jones 1998; Porter 1997).

Methodological preface

Introductory textbooks in the emerging sub-specialty of computer media studies rightly note the importance of technological integra-tion or convergence between previously sepa-rate media (Slevin 2000; van Dijk 1999). Thisfeature, however, facilitates social divergence:media devices enable more diverse, and moredispersed, forms of interaction.

The dual process of convergence and diver-gence poses a methodological challenge for re-search. For example, ‘audience’ or, better, ‘user’studies must gain access to and document more,and more differentiated, contexts of media use.The methodological complexity is only under-scored by phenomena such as ‘mobile’ and‘ubiquitous’ computing. With a growing minia-turization of input–output devices and an in-creased transmission capacity, the mobility ofmedia and the time–space distanciation of com-munication (Giddens 1990) are accentuated.Further, with the embedding of computers intomore everyday objects, the definition of com-puter ‘media’ is brought more radically intoquestion.

Standard reference works are in the making

(Jankowski and van Selm in press; Jones 1999;Lievrouw and Livingstone 2002; Mann andStewart 2000), transferring theories, empiricaldesigns, and methods to digital settings and, in part, devising new ones. In overview, at least three different sets of issues and concernswhich are specific to computer media may beidentified:

1 Online data collection. Computer networksoffer an accessible, economical, and efficientinstrument of data collection (for example,online surveys (Witmer et al. 1999) or focusgroups (Mann and Stewart 2000)). However,their potential is limited in several respects.First, a comparatively large percentage of mostpopulations of interest will not be connected to a network. Second, there are difficulties ofverifying the nature of a sample contacted in ‘cyberspace.’ So far, then, online data col-lection is most appropriate for studying onlineinteraction or online communities as such. Inaddition, traditional methodologies such asobservation and interviewing take on a newmeaning in a news group or chat room, andblend into content or discourse analysis (of logfiles) (e.g., Baym 2000; Downey 1998; Hine2000; Miller and Slater 2000).

2 Validity of findings. A particular aspect ofinteraction via the Internet is that people can,and do, present and explore multiple identitiesor personae (e.g., Turkle 1995). This clearlylimits the explanatory value both of the basicinformation provided and of any inferencesabout the respondents. Furthermore, comparedto census or postal information, computer‘addresses’ may be less precise in identifying thesource of a response. Online findings also tendto focus attention on the mediated interactionin itself, rather than on its social contexts, asfocused by recent media studies.

3 Research ethics. Whatever evidence becomesavailable online raises a number of ethicalissues. This is, in part, because the status ofcomputer-mediated communication as a socialcontext – on a scale from public to private – isunresolved. As professional guidelines andsome form of a cultural consensus are still beingnegotiated, research projects need to consider

Computer-mediated cultures 183

1111234567891012341115678920123456789301234567894012345678950111

technologicalconvergence

socialdivergence

ubiquitouscomputing

multiplepersonae

log files

carefully issues of anonymity, confidentiality,and ‘informed consent.’

Interactivities

A key characteristic of computer media is theextent to which they are becoming integratedwith, and constitutive of, other social contexts.The dividing lines between ‘medium,’ ‘text,’and ‘context’ pose a host of theoretical andanalytical questions. The relationship betweenmedia as means of representing reality, and asresources for intervening in reality, has thusbeen re-emphasized as a research question bydevelopments in digital technology.

While the term ‘interactivity’ has beenwidely employed to refer to the defining char-acteristic of computer media, it has been usedin a variety of senses (for an overview see J. F.Jensen 1999). Indeed, it may be argued that theterm refers to several distinct phenomena, andshould be studied in the plural.

Figure 11.2 lays out three types of interac-tivity with reference to key concepts of mediaand communication theory, as introduced inChapter 1.

• Interactivity 1 refers to the medium-userrelation of communication. In computer sci-ence, this relation has been examined, forinstance, by the influential human–computerinteraction (HCI) tradition. Its research issuesinclude both technological design and ergo-nomics (Preece 1994). Compared to print aswell as electronic media, computer users antici-pate and redirect later stages of the interactivesequence. This affords greater flexibility anddynamism, with computer games as the proto-typical example, but also in, for instance, inter-active fictions (e.g., Murray 1997). Most con-cretely, this type of interactivity may be definedas selectivity – selections by the user are re-quired for the interaction to proceed at all. As acommunicative sequence, moreover, the inter-change may be described as a structured turn-taking,� in certain respects corresponding to

the turns of a conversation (Sacks et al. 1974).Compared to, for instance, film directors andtheir crews, system developers and program-mers play a specific role in facilitating turn-taking – over and above content providers – andare, in one sense, ‘communicating’ with users.Users, in turn, may customize their interface.Together, these various communicative optionsraise to a new degree the question of who – howmany, in what capacity – is the ‘author’� ofmedia discourses.

• Interactivity 2 refers to a broader relationbetween media and the rest of the social struc-ture. The research questions are familiar to themedia field – media have been said to functionas a watchdog, a Fourth Estate (Cater 1959), apublic sphere (Habermas 1989 [1962]), or aninstitution-to-think-with (Chapter 1). At issueare classic questions of how computer-mediatedcommunication shapes, and is shaped by, thepolitical and economic infrastructures of sociallife. A helpful attempt at modeling this link hasbeen the typology of Bordewijk and Kaam(1986) (Figure 11.3). The conceptual matrixdistinguishes between central and distributedcontrol over an available information base, andbetween central or distributed control over theretrieval of particular items at a particular time,resulting in four ideal types of communication.The more familar ones are ‘conversation,’ as in face-to-face or real-time online dialogue (e.g., chat or conferencing), and ‘allocution,’which corresponds to traditional mass commu-nication. Next, ‘consultation’ is a variant espe-cially associated with computers, for instance,in the use of databases or online gaming.Finally, ‘registration’ refers to the more or less automated documentation of users’ trajec-tories on the Internet and other systems. This‘communication’ raises a host of issues regard-ing data security, confidentiality, as well as com-mercial (ab)uses of information. One import-ant implication of the model is that all four (as well as other) communicative forms can be brought together on a single platform,making the computer a ‘metamedium’ (Kay andGoldberg 1999 [1977]).


� research ethics – Chapter 16, p. 289

� turn-taking – Chapter 7, p. 108

textualrepresenta-

tion andcontextual

action

� authorship – media production, Chapter 4, p. 65

computer asmetamedium

• Interactivity 3 refers to the individual–society nexus, as mediated by computer media.Social agents participate continuously andlocally in the structuration of society, throughreference to a medium. The potential for local-ized action via computer media has been highlighted by some utopian visions regardingcitizen involvement in society (e.g., Negroponte1995) (the dystopian counterparts have notedthe capacity of computers for surveillance ofcitizens by state and other social agencies). Inmore general theoretical terms, computer mediausers perform a range of actions – the (re)pro-duction of physical objects, of personal rela-tionships, organizations, communities, and ofentire societies. To exemplify, the material pro-

duction of commodities increasingly involvescomputer-aided design and manufacture (CAD/CAM). A second example is organizations, such as banks, insurance companies, and grass-roots movements, which cohere and operate, inpart, through computer media (see Hauben and Hauben 1997; Rheingold 1994). If com-munities and societies can be understood assystems of communication, computer-mediatedcommunication is one preparatory conditionfor new social formations. Finally, politicaldemocracy is one specific example of a radicallydistributed form of social interaction. Withnews services on the Internet and elsewhere asan increasingly strategic center-piece, media-users participate in, and reproduce, political


1111234567891012341115678920123456789301234567894012345678950111

INTERACTIVITY 3

INTERACTIVITY 1

INTERACTIVITY 2

Agency

Medium

Structure

Figure 11.2 Three types of interactivitySource: Jensen 2000a

CONTROL OF TIMEAND ITEMS SELECTED

Central

Distributed

CONTROL OF INFORMATION BASE

Central

allocution

consultation

Distributed

registration

conversation

Figure 11.3 Four modes of communicationSource: Bordewijk and Kaam 1986; see also McQuail and Windahl 1993: 209

societies assystems ofcommuni-cation

institutions. Like other institutions, politics hasno unified appearance or locale. This does notmake that political participation via (digital)media less than real, or necessarily inferior toconversation. Mediated communication is anecessary condition of political democracy inmodernity.

Intertextualities

The spread of digital and distributed forms of communication has given the theoreticalconcept of intertextuality renewed importance,also for empirical studies of media. To clarifythe present relevance of the concept, its geneal-ogy must be briefly retraced.

The seminal contributions were made byMikhail Bakhtin and the circle around him in the early decades of the twentieth century (see Bakhtin 1981; also Bell and Gardiner1998). Bakhtin’s most basic concept, ‘dialogism’(which was translated by Kristeva (1984) as‘intertextuality’) underscores ‘the necessaryrelation of any utterance to other utterances’(Stam et al. 1992: 203). ‘Utterances’ are hereunderstood in a broad semiotic� sense as vehi-cles of meaning, including conversations, plays,novels, newspapers, or hypertext systems. Theconstituent signs of each of these communica-tions may be said to acquire their meaning inrelation to other signs, past as well as present.Given this abstract definition, the reference to a‘necessary’ relation between utterances may bepuzzling. In addition to the trivial observationthat a set of utterances is related in time andspace, and by the communicators, the point isthat they enter into a particular configuration,and that this structure will follow certain social and cultural patterns. A culture mighteven be considered an (almost infinitelycomplex) instance of intertextuality.

In contrast to some later appropriations of hisidea, Bakhtin (1981) recognized the social deter-mination of texts and communications. Themore difficult question is how society makes itspresence felt in language. A preliminary andalmost poetic answer was offered by the Bakhtincircle, in a central statement attributed to

Volosinov (1985: 53): ‘The immediate social sit-uation and the broader social milieu whollydetermine – and determine from within, so tospeak – the structure of an utterance.’ The phrase‘determine from within’ suggests the theoreticalambition in the Bakhtin circle of transcending aclassic Marxist position that would see the ideo-logical or cultural superstructure as more or lessdirectly determined by, and secondary to, thematerial basis of society. The distinction between‘immediate social situation’ and ‘the broadersocial milieu,’ moreover, pointed to the severalmoments of social determination. It is such aweaker, processual notion of determination, orhegemony,� as developed also by other socialtheorists such as Gramsci (1971), which hasinformed much work about intertextuality.

The methodological question, however, ofhow to examine intertextual structures and thesocial processes shaping them has tended todivide researchers. The majority of intertextualstudies, as exemplified by Genette (1997 [1982]),have centered on texts as relatively self-containedentities. This research has made little or no ref-erence to complementary sources of evidenceabout literary and media institutions, or aboutthe audiences who respond to intertextuality.

An alternative approach was outlined byFiske (1987), who distinguished different typesof intertextuality according to their social func-tions. ‘Horizontal’ intertextuality concerns thetransfer and accumulation of particular mean-ings over historical time, as preserved inmetaphors, themes, characters, and genres.‘Vertical’ intertextuality operates during a moredelimited time period, but across several mediaand social contexts. To clarify this synchronicperspective, Fiske referred to three types oftexts: ‘Primary’ texts are comparable to tradi-tional artworks, being vehicles of central mean-ings or insights. (In a historical or horizontalperspective, the primary texts are the center ofattention.) If the primary text, for instance, isa feature movie, the ‘secondary’ texts willconsist of studio publicity, reviews, and criti-cism. The ‘tertiary’ texts are contributed byaudiences in conversations and other interac-tion around media. Together, the two axes of


� semiotics – Chapter 2, p. 23 � hegemony – Chapter 2, p. 39

horizontal and verticalintertextuality

primary,secondary,and tertiarytexts

intertextuality may be understood as a modelof how textual meaning is produced and circu-lated in society (Figure 11.4).

It should be added that Fiske’s (1987) ter-minology still privileges media texts as theprimary sources of meaning, apparently in con-tradiction of the logic of his argument (seeJensen 1995: 61–62). Surely, media are not themain sources of meaning in social life. Instead,as suggested in Figure 10.2, meaning ratherflows from society to the media. This is in spiteof the fact that the methodologies of mediaresearch must be designed to account for thespecific difference media make in society.

A natural candidate for intertextual analysisin previous research has been (radio and) tele-vision flow� (for an overview see Jensen 1994).The several channels in the television universe

of most countries allow viewers to select andcombine discursive elements into an intertextualconfiguration. Flow thus anticipates aspects ofcomputer-mediated communication.

‘Hypertext’ and ‘hypermedia’ serve to linkmessage elements in computer-mediated com-munication (see Aarseth 1997; Bolter 1991;Landow 1997). ‘Texts’ (or parts of texts) canbe linked, for example, for easy access or forpurposes of indexing. ‘Media’ can be con-nected, for instance, in the form of verbalexplanations of video sequences in ‘multimedia’discourses. In each case, the ‘hyperlink’ may beunderstood as a form of instrumental or oper-ationalized intertextuality – a means of makingexplicit, retrievable, modifiable, and communi-cable what might have remained a more or lessrandom association. This discursive structurehas been one practical key to the development,for example, of the World Wide Web.


1111234567891012341115678920123456789301234567894012345678950111

Relatedgenres,figures,themes

Tertiary texts:conversationabout media

Relatedfigures oractors

Earliertexts in thesame genre

Earliertexts in thesame series

HORIZONTAL INTERTEXTUALITY

VERTICAL INTERTEXTUALITY

Primary text

Secondary texts:references withinmedia

Figure 11.4 Horizontal and vertical intertextuality

� flow – Chapter 10, p. 168

hypertext andhypermedia

For theory development and furtherresearch, it may prove helpful to conceptualizethe computer as a multiply semiotic machine,which processes and communicates informationthrough several links and levels of ‘code’ (e.g.,Finnemann 1999). A generative model of mean-ing,� which focuses on the different levels atwhich ‘information’ and ‘meaning’ are coded(and which was originally inspired by comput-ing), suggests itself for this purpose. Media andcommunication theory still has to arrive at abasic typology of computer media discourses,noting how particular types of hardware andsoftware allow for different forms of navigationand interaction.

Toward this end, one can return to the basicvehicles or units of meaning – verbal, visual,and other signs. What distinguishes verbal lan-guage from images, above all, is ‘double artic-ulation’ (Martinet 1964 [1960]). A verbalstatement and a photograph both have a firstlevel of articulation, in the sense that they eachcommunicate a meaning, but they do so in dis-tinctive ways. Only verbal language has asecond level of articulation, i.e., a fixed reper-toire of minimal units – sounds or letters. Theformal constituents of an image cannot begraded or parsed in any similar way.

Hyperstructures, being a key organizing prin-ciple of digital media, may be understood as athird level of articulation. Both the meanings ofordinary delimited media messages (first articu-lation) and their constituents (second articula-tion) have normally been ascertained ‘internally,’with reference to an immanent structure of dis-course. Compared to such a structure, a hyper-link may be said to articulate an ‘external’meaning, joining one text (verbal, visual, other)with another and suggesting a whole that is notinherent in the parts. This discursive whole, to besure, may still be studied in its own right or withreference to decodings and uses. The point is thathyperlinks and other features of digital mediamay be different in kind from most other mediatexts, including the links between image and textwhich have been at the center of visual commu-nication studies since Barthes (1984 [1964b]).Hyperstructures give rise to different types of

interpretation as well as intervention, becausethe entire system of communication is both moreflexible in itself and modifiable by users.

Figure 11.5 illustrates the three levels ofarticulation. The solid vertical lines indicate the connection between first- and second-levelarticulation in texts; the horizontal lines indi-cate connections between text and image at thefirst level of articulation; and the dotted linessuggest links across the third level of articula-tion. Here, texts and their verbal elements –letters, words, sentences, sections, etc. – as wellas images (or fields within them) can serve as‘nodes.’ When joined at the third level, theselinks produce meanings of a different order andcomplexity. It is the nature of these links andmeanings that further research must accountfor – a task that is complicated, but also madepotentially rewarding, by the dynamic natureof the full discursive system.

Intermedialities

With computer-mediated communication, re-search on ‘media’ has come full circle in tworespects. First, the individual media are increas-ingly interrelated – technologically, in market-ing terms, and as cultural forms (for anoverview see Bolter and Grusin 1999). As notedin Chapter 1, it is helpful to conceptualize thecomputer as a third degree of medium,� whichincorporates several previous technologicalmedia, and which re-emphasizes features other-wise associated with face-to-face interaction.Compared to some earlier notions of an ‘information society’ (e.g., Nora and Minc1980; Porat 1977), which pointed to a growinginformation sector and its consequences forculture and society, it may be the integral‘informatization’ of work, leisure, and mostmajor social institutions which will distinguisha ‘network society’ (Castells 1996).

At present, research is beginning to explorethe several ‘intermedialities’ and social formsthrough which computers take effect. Compu-ters enter into the social infrastructure; theyalso provide imagined partners in interaction,from avatars and intelligent agents to cus-


� generative model of meaning – Chapter 2, p. 30

doublearticulation

theinformationsociety

� media of the third degree – Chapter 1, p. 4

the networksociety

tomized communication systems. For the timebeing, research in this strategic and much pub-licized area must proceed amidst hyperbolicprojections. Utopian proclamations note thedemocratic and intercultural potential of theInternet, while dystopian statements anticipatethe demise of humans and the rise of cyborgs(cybernetic organisms). Such visions shouldcome as no surprise after two centuries ofambivalently embracing modern technologies,from Shelley’s Frankenstein (1818), via 1950sartificial intelligence, to the cyberspace of Gib-son’s Neuromancer (1984).

Besides the convergence of previously sepa-rate media, a second implication of computer-mediated communication is sometimes taken tobe a questioning of any absolute dividing line,not only between ‘media’ and ‘context,’ but alsobetween ‘reality’ and ‘virtuality.’ In fact, farfrom the collapsing material and virtual realms,computer media may simply offer one moreprosaic resource for human cultural activity (seefurther Jensen 2001). Referring to Anderson’s(1991) notion of national (and other) culturesas ‘imagined communities,’� one may arguethat, in premodern as well as modern settings,

media have been instruments for imaginingother people, projects, and solutions – for get-ting jobs done. Certainly, virtuality can andmust be examined as a matter of degree inphenomenal terms, since different modes ofcommunication give rise to different experi-ences of sociality. However, it is an importantpremise for research that the differentiation and distanciation of communicative interactionvia computers do not make either the inter-action or the social context less real. By imag-ining other people, in face-to-face as well astechnologically mediated communication, webecome virtual partners in real social activity.

To sum up, several traditions of research,each with a more or less explicit model of com-munication, have informed the attempt so farto conceptualize computer media:

• the computer as yet another channel ofsocial communication, for example, in the inter-disciplinary tradition of computer-mediatedcommunication (CMC) (for an overview, seeJones 1998);


1111234567891012341115678920123456789301234567894012345678950111

2nd level

TextImage

Image

1st level

3rd level

1st level

2nd level

Text

Figure 11.5 Three levels of articulation

cyborgs

cyberspace

virtual reality

� imagined communities – p. 173

• the computer as an augmentation of thehuman intellect, not least as a tool for reorgan-izing work and education, as represented byhuman–computer interaction (HCI), but datingback to early conceptions of computing as com-munication (e.g., Licklider and Taylor 1999[1968]);

• the computer as a substitution of humansaltogether, as evidenced in artificial-intelligenceresearch (for an overview see Gardner 1985),and as radicalized in current notions of thecyborg (e.g., Haraway 1997);

• the computer as one resource in the consti-tution of contemporary societies, mediatingbetween agency and structure – a position elab-orated in Chapter 1.

In conclusion, computer media have con-tributed new issues to the agenda of media and communication research, and have high-lighted other issues, such as the global level of communications and the relationship be-tween representation and (inter-)action. At the same time, research is well advised to keep in mind the specific local and national circumstances in which media continue to be produced and received. A lesson of mediastudies has been that the social consequences of media depend, to an important degree, on the concrete contexts and cultures in whichthey are introduced and applied. This lesson is brought home by historical studies of media, which are the topic of the next chapter.


RESOURCE BOX 11.1 NATIONAL AND INTERNATIONAL VARIETIES OF MEDIA STUDIES

Different national and cultural contexts, not surprisingly, have developed different traditions of research.Whereas the volume focuses on the English-speaking cultures that have been central in institutionalizing themedia field internationally in recent decades, this box indicates additional references to other literatures. Infact, much of this literature is available in English translation or in excerpts, as reviewed in these titles.Moreover, the section in this chapter entitled ‘Between cultures’ references contributions which have soughtto overcome East–West as well as North–South divides, also within media and communication research.

Russia• Kelly and Shepherd 1998 – a broad, non-sectarian collection addressing Russian cultural research,

including the Soviet period and the 1991 transition, with many references.

Japan• Ito and Tanaka 1992 – an overview of institutions of media research and education in Japan.

Latin America and Spain• Martin-Barbero 1993 – an original contribution to social and cultural theory, as well as an

overview of Latin research traditions;• Graham and Labanyi 1995 – a broad collection covering developments in both Latin America and

Spain.

Germany• Burns 1995 – a broad collection, including historical overviews of the place of Germany in

international and intellectual history;• Schulz 1997 – a review focusing on political communication, and on the 1990s, but with coverage

of the development of the field.

France• Reseaux – a journal presenting translations in English of selected research;• Forbes and Kelly 1995 – a broad collection, including historical overviews of the place of France

in international and intellectual history;• Cayrol and Mercier 1998 – a review focusing on political communication, but with extensive

contextualization and coverage of the development of the field.

INTRODUCTION

In this chapter, I review developments in his-torical studies of communication and media. Inorder to contextualize these developments it isnecessary to consider, historically, the formationof history as an academic discipline. Through-out the chapter, I attend to theoretical andmethodological issues concerning historio-graphy (the practice of writing history) in orderto illuminate some of the more specific problemsin writing histories of media and communica-tion. What is not dealt with in this chapter, since to do so would turn a chapter into a book, arequestions concerning the nature of history itself. The given facticity of history as an acad-emic field of inquiry is taken as the startingpoint, but it is obvious that history itself – whatever it may be – is not something inventedby historians; that is, the emergence of historyas an academic domain is itself a reflection ofthe historical process. It is precisely a ques-tion for historians as to why it should be – at aparticular time and in particular places – that a new kind of historical consciousness should emerge, one of whose manifestations is acade-mic historiography.

A historical sensibility comes fully intobeing in nineteenth-century Europe and NorthAmerica, and is undoubtedly connected withthe unprecedented speeding up of technologi-cal innovation, economic, political, and othersocial change – in short, the phenomenon ofmodernity.� Transformations in communica-tion and the continuing development of ‘new’media (radio in the 1920s was as much a matterof wonder for the contemporary world as the Internet is today) have contributed to theheightened sense of social change – the sharpawareness of the difference between past andpresent, then and now – which is the mark ofhistorical consciousness. Recent social theory(e.g., Giddens 1990; Harvey 1989) has alsoshown a renewed interest in time and space,and especially in the consequences of theirshrinkage to the point of disappearing as obstacles to be overcome in the circulation ofinformation and commodities. At this point intime, however, the Internet has no history tospeak of, but only a future. The work of his-torians – unlike that of social theorists – beginsonly as the past emerges as distinct from the

1111234567891012341115678920123456789301234567894012345678950111

History, media and communicationPaddy Scannell12

• an overview of the history of history as a field of research• a discussion of ‘media history’ as an inclusive phenomenon• a presentation of histories of oral and written communication• a review of histories of different technological media, including the press and broadcasting• an example of archival research on historical data.

historiography

� modernity – Chapter 11, p. 173

present. That this has begun to happen inrespect of modern media and communicationtechnologies is what gives the possibility of thesubstantive concerns of this chapter.

THE HISTORY OF HISTORY

The professionalization of history began in the nineteenth century. A chair in history at the University of Oxford was first establishedin 1848. In his inaugural lecture, the firstRegius Professor of History, H. H. Vaughan,defined the job of the historian as the ‘disclo-sure of the critical changes in the conditions ofa society’ (Stone 1987: 4). From the late nine-teenth century, history developed into an inde-pendent academic discipline. Separate historydepartments were established with Ph.D. pro-grams for the training of future historians.Professional associations were created, taught programs at the undergraduate and graduate levels were established, and a curriculum (an‘agenda’) for the study of history as a distinc-tive academic subject was put in place withinthe emerging fields of the humanities and socialsciences (Stone 1987: 5).

There were two aspects to this development:first, a definition of the right way of ‘doing’history (the professionalization of methods),and, second, a specification of the subject mat-ter of historical inquiry (what history is prop-erly ‘about’). Both aspects came to the fore inGermany, where there was pioneering concernwith the establishment of history as an aca-demic discipline on a ‘scientific’ basis. Whatthis meant, above all, was a renewed concernabout the ‘objectivity’ of historical researchwhose scientific truth claims were underpinnedby rigorous research methods linked to tempo-ral distance on the part of the historian fromthe field of inquiry, and careful scholarly inter-pretation and evaluation of the evidence. Thedoyen of German ‘historicism’ (see Bambach1995) was Leopold von Ranke, whose name isparticularly associated with a revolution insources and methods – a shift away from earlierhistories or ‘chronicles’ to the use of the offi-cial records of governments. Historians nowbegan to work regularly in archives, and theyelaborated an increasingly sophisticated set of

techniques for assessing the reliability of docu-ments. They claimed that their own historieswere therefore more objective and more ‘scien-tific’ than those of their predecessors (Burke1992: 7).

The primary subject matter of these inquirieswas the nation-state� and, in particular, its administrative and constitutional develop-ment, as linked to its diplomatic and militaryrelations with other nation-states. Nationalrecord offices were opened up to academic his-torians, and the basic political documents relat-ing to these matters were indexed and madepublicly available for research free of charge.Thus, by the beginning of the twentieth century,the academic profession of history was in place. Its agenda consisted of narrative histories– firmly supported by archival research intoprimary documentary source material – whosesubject matter was the activities of those politi-cal and military elites who created, defined, andmaintained the modern nation-state. Historiansno longer wrote for an educated general public(Stone 1987: 3–7). They wrote for each other orfor their students, and talked among themselvesas a self-legitimating subset of a new historicalgenus – homo academicus (Bourdieu 1988).

Not unnaturally, these foundational devel-opments, set in place between the 1870s andthe 1930s, were subject to critique in the course of the twentieth century. On the onehand, a set of methodological objections wasraised against the privileging of ‘narrative’ asthe definitive mode of historiography (historywriting). On the other hand, the narrowly elitist base of the subject matter of history asthe actions of ‘great men’ came under increas-ing critical scrutiny. The issue of history as nar-rative is fundamental to it as an academicdiscipline, since it raises crucial questions con-cerning the nature and status of historicalevents and of historical agents. In short, theproblems are: What is history about, who arehistorical agents, and what is historiography(the writing of history)?

Political history, history of events, and nar-rative history – these are almost interchangeableterms, as Paul Ricoeur points out in his review

History, media, communication192

history as anacademicdiscipline

Germanhistoricism

and scientificobjectivity

� the nation-state – Chapter 10, p. 173

narrativehistory

of ‘the eclipse of narrative’ (Ricoeur 1983:95–120). They were treated as such by theFrench historian, Fernand Braudel, a leadingmember of the Annales School, so called afterthe journal of that name founded in 1929 byMarcel Bloch and Lucien Febvre, who, alongwith Braudel, were the leading critics of tradi-tional history and vigorous opponents of whatthey called histoire evenementielle (Burke 1991:7–8). They argued that it was naive to think ofhistorical events as being ‘made’ by historicalactors, because what determines or structuresevents has no level of explanation, in traditionalhistorical narratives, beyond the deeds of sig-nificant individuals themselves. Their struc-turalist� critique of history-as-events (or the‘great man’ view of history) draws attention to all those shaping factors, ignored in narra-tive histories, that determine the character andbehaviour of historical actors and the scope oftheir actions. The objectivity claimed by politi-cal histories is at the expense of underlying eco-nomic, social, political, cultural, and religiousfactors which all, in different ways, combine topredetermine the form and content of historicalevents. Moreover, and this was the crucial pointof Braudel and others, the nature of historicaltime remains under-theorized in narrative histo-ries, since they presuppose, without any criticalreflection on the matter, first, that the time ofevents is somehow self-evident, and second, thatit is essentially the same as the time(s) of history.

If history is the sum of all that has ever hap-pened in the past, then a basic methodologicaland theoretical issue for historians is how theyestablish manageable time spans for their activ-ities. As Braudel points out, ‘All historical workis concerned with the breaking down of timepast, choosing among its chronological realitiesaccording to more or less conscious preferencesand exclusions.’ He is highly critical of ‘tradi-tional’ history’s concern with the short timespan and ‘the headlong, dramatic, breathlessrush of its narratives’ which are tied to ‘theindividual and the event’ (Braudel 1980: 27–28). He points to new kinds of economic and social history whose concern is not withparticular individuals, moments, or events, but

with the larger economic and social forces thatstructure their terrain, that shape the groundupon which events are enacted. Such macro(large-scale) histories necessarily take a morelong-term view of the processes of historicalchange. They are concerned with, for instance,the mapping of the cyclical rise and fall ofprices, which calls for complex, large-scaledata-gathering of quantitative informationabout prices, wages, money, rent, output percapita, capital investment, overseas trade, andother key economic variables (Stone 1987:13–14) compared to the qualitative method-ologies most prevalent in historical research.�The focus of macro histories is not on individ-uals and their actions, but on underlying fac-tors that mould the processes of social changethrough time. These ‘conjunctural’ historiesattempt to analyse the ways in which a rangeof economic variables combine in a particularhistorical conjuncture to precipitate crises thatplay themselves out at the level of events andactions. The intermediate time span of such his-tories is measured in decades; ten, twenty, fiftyyears; but beyond this lies the long-term – thelongue durée – which stretches over centuries.It is a kind of ‘motionless history’ which triesto establish the most fundamental constraintsupon the shape and scope of human actions:the impact of climate and geography, forexample, upon the historical life of a regionsuch as the Mediterranean, the subject ofBraudel’s most famous work (Braudel 1972,1974).

The critique of narrative history as privileg-ing elites came to the fore in the 1950s and1960s. New kinds of social history were devel-oped which insisted, as the historical novel had done (Lukacs 1989 [1916]), on the signif-icance of the apparently insignificant – ordinarymen and women, and the character of everydaylife (Burke 1991). These ‘histories frombelow’� (Sharpe 1991) – as opposed to historyfrom above (political history) – rediscovered

The history of history 193

1111234567891012341115678920123456789301234567894012345678950111

the AnnalesSchool

� structuralism – Chapter 2, p. 29

macro andlong-termhistory

the longuedurée

� quantitative methodologies – Chapter 13. See also thediffusion of innovations, Chapter 9, p. 140, and thespread of literacy (Vincent 2000)

� histories from below and oral history – Chapter 14,p. 241

such ‘hidden’ or ‘invisible’ histories as those ofthe English working class (Thompson 1968), orof women (Rowbottom 1974; Scott 1991), ofthe family (Stone 1979), or the body (Porter1991). Thus, in the second half of the twenti-eth century, academic history began to broadenout and diversify. It was in this context ofexpanding areas of research and inquiry thathistorical attention began to focus on, amongother things, the media.

MEDIA HISTORY?

What would a history of the media be about?We must first determine what the word refersto. Note that it is a plural noun: a ‘medium’,the media. This plurality is usually taken toinclude the press and broadcasting (radio andtelevision), and cinema too, perhaps. One use-ful way of reconstructing how terms gathercertain meanings is to look at their historicalusage. When did they begin to be used in a par-ticular way, and why? What we can say of ‘themedia’ is that the sense in which it is commonlyused today – and by now it is a commonsenseterm and part of everyone’s ordinary usage – isno older than the early 1960s. Earlier usage ofthe word ‘medium’ was associated with the late nineteenth-century vogue for spiritualismwhich, as John D. Peters (1999) has cogentlyshown, was linked with the impact of new com-munication technologies: the telegraph and tele-phone. The idea of a ‘media society’ shows upfor the first time in the 1960s, and coincideswith the rapid rise of television as the dominant‘medium’ of daily life in Europe. In everydayusage, the term now recognizes that, in ourkind of society, the press and broadcasting are central social institutions and a taken-for-granted part of everyone’s ordinary, everydaylife, but the ‘media society’ as we know it cameinto being only in the second half of the twen-tieth century.

To put it like this invites the question, ‘Whatwas there before that?’ The Norwegian hist-orian, Hans Fredrik Dahl (1994), has pointedout that before the ‘media society’ there was‘mass society’, and terms such as ‘mass culture’and ‘mass communication’ referred to the im-pact and effect of new technologies of commu-

nication on early twentieth-century society.Thus, in the 1930s, we find a whole series ofdebates (in Britain, Europe and the USA) aboutthe effect on social values, the beliefs and tastesof new kinds of ‘mass’ journalism, Hollywoodcinema, commercial radio, and an emerging‘popular’ music industry. The issue Dahl raisesis as follows: first, do we try and study histor-ically the development of ‘the media’, or do wego back and try and trace the development ofparticular ‘mediums’? Second and relatedly,what is the difference between studying ‘themedia’ and studying ‘communication’?

In this chapter, I will treat this difference asessentially one of historical scale. Communi-cation is as old as mankind, insofar as languageis constitutive of human being. Modern mediaof communication – daily newspapers, radioand television services – stand in a long tem-poral continuum that reaches back throughwritten systems of communication (often takenas the point of entry into history) to spokenlanguage and the beginnings of human sociallife. Thus histories of communication areanterior to, and distinct from, media histories.Moreover, ‘the media’ is a synthetic descriptorand as such misleading when used in respect ofsomething called ‘media history’, for the differ-ent mediums that make up ‘the media’ havethus far only been studied historically as dis-crete phenomena, and not as the relationaltotality which the term itself implies. Accord-ingly, I begin with histories of communicationand proceed thence to more recent histories ofdifferent mediums which make up the knownand understood ‘media society’, familiar assuch to all of us today. I will then briefly reviewsome issues that arise when trying to ‘do’ his-torical work on aspects of ‘the media’.

THE HISTORY OF COMMUNICATION

The fundamental, universal form of communi-cation in all human societies is language, or,more precisely, spoken language, or speech.�We know that systems of writing, as devices forrepresenting spoken language, developed muchlater – on current evidence around 7000 years


‘the media’from the

1960s

writing asbeginning ofhistory

� speech – Chapter 2, p. 19

ago. There are, of course, different systems ofwriting, but today two systems prevail: alpha-betic systems – the Roman alphabet is mas-sively dominant, but there are others (Greek,Cyrillic) – and ideographic systems (Chinese,Japanese). We can take these two modes ofcommunication – speech and writing – as inti-mately connected, but there is no doubt thatspoken language is ontologically (as well as his-torically) prior. That is, speech is constitutive ofour (human) way of being: some would say itis the species-defining characteristic. It is thevery essence of our human, social life. Allhuman societies have a spoken language. Farfewer have developed systems of writing. Thusa primary question is: If both are systems ofcommunication, what are the differencesbetween them? What is it that we can andcannot do with speech as our only availablemeans of communication? What does writingdo for us (and what do we do with it)? Andwhat is it that later means of communication(print, radio, television) do for us, and what dowe do with them? Such questions have a posi-tive and negative side to them: What does anyparticular means (medium) of communicationenable us to do, and what are the limits to whatwe can do with it?

To put this slightly differently, a fundamen-tal constraint on the scope of human actions isimposed by the temporal and spatial charac-teristics of our available resources, includingour resources for communicating with eachother. The question, first clearly put by theCanadian economic historian, Harold Innis, is:What are the implications of the availablemeans of communication for the character of aparticular human society (Innis 1972 [1950])?What is a society like that has only speech asits available means of communication? Innis’sanswer was that such societies are necessarilyspatially (geographically) small in scale. Theymay be nomadic or pastoral (they may moveabout or remain in one place) but, since theirshared spoken language is what makes themwhat they are and holds them together, the sizeof their community is restricted by its face-to-face character. The group’s social memory – itsknowledge in the present of ‘how and what andwhen’ to do things, as derived from past prac-

tices – is orally transmitted in a spoken trad-ition that is passed on from one generation tothe next.

Systems of writing, Innis argued, developedas means of coordinating and controllinghuman activities across extended time andspace. Writing is a system of record: a way ofputting things down, so that information maybe transmitted through space over great dis-tances and be preserved through time as arecord of what was said and done. Innis drewspecial attention to the different materials usedfor writing, and the ways they affect the scope,character, and purposes of the messages theyrecord. Messages carved with a chisel on heavy,durable materials – such as slate, granite ormarble – have a monumental character thatendures through deep, or slow, time. Messageswritten on lightweight materials such aspapyrus or paper (invented in China and avail-able in Europe, via the Moors, from the thir-teenth century) are portable and easily carriedover great distances.

More generally, Innis argued that differentmedia, using different materials, had differentconsequences for the human, social control oftime and space: he called this ‘the bias of com-munication’ (Innis 1964 [1951]). He noted theimportance of written technologies for theestablishment and maintenance of empires – thecreation of power blocs spread over great dis-tances and preserved through many genera-tions. He observed that the complex task ofadministering, coordinating, and policing a vastimperium, such as the Roman empire, requiredwritten systems both as means of recordingdecisions, laws and so on, and of communicat-ing them throughout the empire. Writing wasone essential underpinning of the whole colos-sal enterprise. Innis’s analysis of written systemsof communication shows clearly that writing,in its primary functions, has always been linkedto political and economic power. Whereverwriting has established itself, it has immediatelyproduced a fundamental distinction betweenthe lettered and the unlettered, the educatedand the uneducated. To have the skills ofreading and writing is a passport to individualself-advancement. Literacy gives rise to edu-cated elites (in the past they were called priests;

The history of communication 195

1111234567891012341115678920123456789301234567894012345678950111

alphabetic andideographicsystems ofwriting

time and space as

aspects ofcommunication

systems

writingmaterials

the bias of com-munication

writing asrecord andcommuni-cation

literacy and socialdifference

today they are called academics) that gravitateto the centres of power.

Innis’s major work focused on the characterof communication in ancient societies. It wasMarshall McLuhan who picked up on Innis’sideas and applied them to modern societies.McLuhan, following Innis, made a fundamen-tal distinction between ‘oral’ and ‘written’ cul-tures. He focused particularly on the impact ofthe new technology of printing in the late fif-teenth century on the older ‘manuscript culture’of the Middle Ages. McLuhan argued thatEurope’s take-off into the modern world waspowerfully supported by the new culture of thebook, which rapidly spread through Europe inthe sixteenth century, creating new ‘knowledgecommunities’ based on print (McLuhan 1962).He went on, in Understanding Media (1964),to compare print cultures with the newer elec-tronic cultures of modern media, especiallyradio and television, arguing that they had theeffect of rescuing oral forms of communicationfrom the dominance of printed and writtenforms. McLuhan was one of the first real ana-lysts of modern media as communication tech-nologies. It is in his writings that ‘the media’are first examined as interrelated technologiesthat come together to create media societies. Itis from McLuhan that we learned to think of‘the media’.

Perhaps his most striking and importantperception of the impact of electronic mediawas that it could no longer be thought of asparticular to this or that society. What was newabout electronic media was their global impact:they created, for the first time in history, thepossibility of point-to-point, instant, live com-munication between any two points on theglobe – ‘the global village,’� in his famousphrase (McLuhan 1964). Electronic media havepowerful transforming effects on the characterof social time and space. In effect, they help tocreate and sustain the ‘world-historical’ char-acter of modern life (Giddens 1990). It isincreasingly difficult not to know what is goingon elsewhere in the world today. Time andspace have drastically shrunk. Each one of us

has the world in our living room. Global tele-vision is a very real phenomenon when millionsand millions of people, scattered all over theworld, have simultaneous access to an eventsuch as the Olympics (Dayan and Katz 1992).Television today is intimately linked to thecharacter of national and international politics,business and war. Through television, culturalnarratives, images, songs and jokes circulatearound the world. When a programme likeDallas is watched in over a hundred differentcountries it is clear that electronic media con-tribute to the formation of a global culture – aculture of consumption, many have argued,underpinned by global capitalism – which hasserious implications for national and local cul-tures. A key aspect of the ‘gap’ between Northand South, the so-called developed and devel-oping regions of the world, is a communica-tions divide – the unequal flow of informationand entertainment between first and thirdworlds.

Innis and McLuhan provide a fascinatingway of identifying the impact and effect ofmedia of communication on human societies byfocusing on them as technologies that extendthe scope and scale of human, social activity.Their emphasis on the ‘bias of communication’implicit in different media technologies drawsattention to their effect on the social organiza-tion of time and space. Their approach has an intrinsically historical thrust and, indeed,conceives of history in terms of time periodsdominated or characterized by different media:from small-scale oral societies, to empiresunderpinned by writing, to the early modernEuropean culture of the book, to today’s elec-tronic global culture. It is valuable to empha-size that the media are not simply ‘neutral’carriers of information, but have a determinateeffect on the character and scope of what theyconvey.

We should, however, note some of the lim-itations of this approach and, in particular, thecharge of ‘technological determinism’ which isoften made against it. As Raymond Williams(1974: 13) succinctly put it:

New technologies are discovered, by anessentially internal process of research and


� the global village and globalization – Chapter 11, p. 180

technologicaldeterminism

development, which then sets the conditionsfor social change and progress. Progress, inparticular, is the history of these inventions,which ‘created the modern world.’ Theeffects of these technologies, whether director indirect, foreseen or unforeseen, are as itwere, the rest of history. The steam engine,the automobile, television, the atom bomb,have made modern man and the moderncondition.

The kind of argument advanced by Innis andMcLuhan has the effect of making technologiesof communication into instruments that appear,somehow, outside and beyond the control oftheir makers. It is another subtle kind of his-toricism: a ‘grand narrative’ of progress inwhich the hero is technology. The media shapehuman societies in their own likeness. The effectof such an argument is to remove the very stuff of history – the actions of human beings,politics, war, culture, entertainment – from theanalysis. Do we use media or do they use us?For McLuhan, as he famously put it, ‘themedium is the message.’ Its content is irrelevant.But what is missing in this kind of approach isany attention to how media are caught up in thehistories of human societies, and are used invarious ways and for various purposes.

Nevertheless, these pioneering historicalstudies of human communication, as deter-mined by media technologies which shape andconstrain human action, have become increas-ingly influential after a period in the 1970swhen McLuhan fell out of favour, particularlywith the Left, whose theories were then infashion (Winthrop-Young and Wutz 1999:xi–xvi). Anthropologists took up the basic dis-tinction between ‘oral’ societies and those withwritten or print-based cultures (Goody 1968;Ong 1982). McLuhan’s thesis about the transi-tion from manuscript to print cultures inEurope was explored by Elizabeth Eisenstein(1979). More recent work, such as FriedrichKittler’s historical study, first published in1986, of Gramophone, Film, Typewriter tookup the basic insights of Innis and McLuhan,and reworked them in a postmodern, anti-humanist reading of the ways in which ‘mediadetermine our situation’ (Kittler 1999: xxxix).

Kittler’s work, in turn, has been an import-ant source for John Durham Peters’s remark-able historical study of the idea of communica-tion (Peters 1999). There is a history not just of technologies of communication, but also of how we think about communication (ourchanging understanding of what it means). Theconnections between the two are explored byPeters in a highly original way. Kittler’s workshowed how new communication technologies– the gramophone, for instance – change theways in which we think of ourselves. Today weare all familiar with the idea that human brainsare rather like computers. This seems a naturalidea to us. A hundred years ago, it seemednatural to think of the brain as rather likeEdison’s phonograph (Kittler 1999: 38–45), anidea that to us now seems bizarre.

Peters, following Kittler, shows in fascinat-ing detail some of the ways in which new com-munication technologies have shaped how wethink about ourselves and others. He traces theidea of communication as occurring not onlybetween human beings, but also with the dead,with angels, with aliens, with animals, and with machines. If this sounds odd, consider the huge investment, since the 1950s, in ‘scientific’efforts to communicate with ‘the rest of theuniverse’ on the assumption that somewhere –out there in infinite time–space – there are intel-ligent alien beings to receive the messages thatwe are beaming out into the void. Peters’s bril-liant description and analysis of the SETIproject (the Search for Extra-Terrestrial Intelli-gence), an international scientific programmethat began in the USA after World War II andwhich still continues, is indicative of a themethat is central to his book, namely that anincreasing awareness of communication gener-ates increasing anxiety about it (Peters 1999:246–261).

It is only in the modern era, with the hugeproliferation of technologies of communica-tion, that communication has become an issuefor us. Today we are highly aware of the frag-ility of ‘successful’ communication, and of theendless possibilities of communicative failure or breakdown both as technical problems andas social problems of our relations and inter-actions with others. To think about how we

The history of communication 197

1111234567891012341115678920123456789301234567894012345678950111

the medium isthe message

history of the idea of communi-cation

think about communication as such in relationto new means of communicating is the distinc-tive contribution Peters makes to the historicalstudy of communication.

MEDIA HISTORIES

The historical study of communication tends,as the brief review above indicates, to exhibitthe following characteristics. Especially in Innisand McLuhan, a synoptic approach is takenwhich operates within large, epochal time spans, which embraces a ‘total’ or global viewof history transcending national boundaries,and which deploys an analysis of the historicalprocess whereby the emphasis is not on humanbeings as historical agents (i.e. as ‘makers’ ofhistory) but on external (technological) factorswhich determine the scale and scope of humanagency. In all these ways, the history of com-munication is part of the ‘new history’ thatreacted against the ways in which the academicdiscipline and field of study was originallyestablished. Histories of communication mobil-ise a deterministic (or structuralist) approach totheir field of study and are strongly anti-humanist, notably so in Kittler’s (1999) work,which throughout writes mockingly of ‘so-called Man’.

Histories of media tend, by comparison, tobe old-fashioned, humanist narratives whichfocus on the histories of particular media insti-tutions or of particular historic ‘moments’. Theyeschew synthetic, synoptic accounts. These dif-ferences should not, of course, be thought of interms of the preferences of individual historians.It is rather that these two different historical‘topics’ – communication and media – call for different methods and accounts. Written histories are, of necessity, about the past, buthow much time must pass before histories can begin to be written? The history of print goes back centuries. The history of television goes back fifty years. The history of the Inter-net goes back ten years, if that. In order forHistory to appear, there must be some distancebetween the present time (in which history iswritten) and the past time (which is to bewritten about). In other words, it may not yet be possible to write histories of ‘the

media’, since this presupposes a synthesis sorecent that its historical lineaments are scarcely visible at this time. What we find instead, asDahl (1994) pointed out, are institutional his-tories of particular media: newspapers, radioand television.

But even this is problematic. What are thegrounds for taking newspapers, radio and tele-vision as ‘the media’? Different taxonomiesshow up in media and communication depart-ments in European and North American uni-versities, where radio, film and television areoften studied together along with ‘journalism’.The common tendency to bracket film and tele-vision together because they are thought of as‘visual media’, and the tendency to neglect thestudy of radio (often called ‘the Cinderella’ ofmedia studies) are the results of another ‘hiddenhistory’, namely, the history of the developmentof the academic field of study itself.� This too,like its object domain of inquiry, is as yet sorecent as to be barely visible.

The legitimacy of any academic field ofstudy takes time to become established. At thebeginning of the twentieth century, the study ofmodern literatures did not exist as an academicdiscipline, and, within a university culture that privileged the study of ancient literatures(Greek and Roman), the study of things like themodern novel seemed trivial and irrelevant(Doyle 1989; Mathieson 1975). The first under-graduate degree in media studies in the UK wasestablished at the then Polytechnic of CentralLondon (now University of Westminster) in1975. Time has yet to work its alchemy on theacademic study of the media, transforming itfrom its present lowly status (in Britain at least)into a respectable field of study such as modernliterature. Meanwhile, it should be noted thatthe emergence and definition of the field hasbeen fitful and uneven, and has often been theoutcome of internal institutional reorganizationand the rationalization of existing resources.Film studies developed earlier than mediastudies, and usually within modern literaturedepartments, from which they sometimes brokeaway. Film histories, however, continued to


institutionalhistories ofsingle media

� history of media research institutions – Chapter 16, p. 278

privilege the cinematic texts.� As film studiesdeveloped, it extended its interests to what itregarded as cognate fields of inquiry, namelyphotography and television. There is no neces-sary correspondence between the structure ofthe media in ‘the real world’ and the ways inwhich ‘the media’ are studied in universities.Historically and ‘as a matter of fact’, radio andtelevision are very closely linked in terms oftheir technological, economic, political andinstitutional formation. Yet they are seldomstudied or researched by academics as related toeach other under the rubric of ‘broadcasting’.It is thus necessary to argue the case for treat-ing ‘the media’ as, in essence, the interconnec-tions between newspapers (the press), radio andtelevision (broadcasting). I have made this caseelsewhere, arguing that the shared structuralcharacteristic of all three is ‘dailiness’; a specifictemporality that is articulated and given sub-stance in the contemporary world by the inter-play of daily newspaper, radio and televisionservices (Scannell 1988, 1996c, 2000).

PRESS HISTORIES

The time spans of the development of news-papers, radio and television are quite different.Newspapers as we now know them in NorthAmerica and Europe can be traced back to theseventeenth century. The technology of wirelesstelephony was worked out in the late nine-teenth century, and its widespread social appli-cation as radio begins after World War I. Thetechnology of television dates from the earlytwentieth century, and the history of its socialusage begins properly after World War II (foran overview see Winston 1998).� Not surpris-ingly, historical work on these developmentshas a different time span in each case.

In the UK, histories of newspapers go back at least to the mid-nineteenth century, andthere is a large body of literature on particularnewspapers (a five-volume history of The Timesbeing the outstanding example (Anon 1935–1958)), journalists and newspaper owners,

nearly all of which was written by people in thebusiness, rather than by professional historians.As such they are largely anecdotal, concernedwith significant individuals (there is a subgenreof journalism biographies about outstandingnewspaper editors and outrageous newspaperproprietors) and the politics of their day. How-ever, as James Curran has argued in an influ-ential synoptic account of British press history(Curran and Seaton 1997), a distinctive con-sensus emerges as to the meaning and signifi-cance of its historical formation. It is yetanother variant of the ‘Whig interpretation ofBritish history’ (Butterworth 1965), in whichthe emergence of a ‘free’ political press is seenas a victory over government censorship andcontrol (Curran and Seaton 1997: 5–113). Thekey moment is the mid-nineteenth century,when newspapers were freed from punitiveforms of taxation imposed by government toprevent ‘unlicensed’ radical newspapers fromcriticizing those in power and fomenting civilunrest. In this account, ‘the market’ appears asthe hero; the guarantor of civil liberties and theprotector of the political independence of news-papers which no longer function as the mouth-piece of various political factions. In Curran’sview, the most authoritative academic historyof the British press (Koss 1981, 1984) simplyextends this ‘myth’ through to the present(Curran and Seaton 1997: 28–31). Curran,however, has his own ‘alternative’ account ofthis history, in which the hero is the ultimatelysuppressed radical press of the early nineteenthcentury, and the villain is the market, which, heargues, has destroyed the possibility of a prop-erly political press. Thus, his account reversesthe terms of the Whig teleology, but retains itspremises. The terms in which press history isthought in Curran’s account remain within thetraditional concerns of political historiography.The past is interrogated and criticized in thelight of a set of unexamined political concernsin the present.

The most influential history of the US pressstarts from rather different premises. It is asocial, rather than a political history, and beganas a Ph.D. dissertation in the Sociology Depart-ment of Harvard University. Its author, MichaelSchudson, was interested in the history of ideals

Press histories 199

1111234567891012341115678920123456789301234567894012345678950111

� film studies – Chapter 2, p. 31

� chronological table of media and communication –Chapter 2, p. 18

dailiness

British press

Whig inter-pretation ofhistory

and the sociology of values. He combined bothin a historical study of the American press,which began as ‘a case study in the history ofprofessions and in the genesis of professionalideology’. As such, its main concern was withthe relationship between the institutionalizationof modern journalism and general currents ineconomic, political, social and cultural life(Schudson 1978: 10–11). The issue that bringsinto focus all these concerns is the question of‘objectivity’ as an ideology or norm for prac-tising journalists:� it is the history of the riseand fall of this idea that is traced in Discover-ing the News. Schudson’s account, though achronological narrative, avoids the pitfalls of a teleology. ‘Objectivity’ is neither the triumphof value-free journalism, nor the hidden, self-deceiving bias of a dominant, market-drivenpress. What is shown, rather, is that ‘objectiv-ity’ as a social value is intrinsically historical.The early American press had no concept of‘objectivity’, which arose only as the press wasindustrialized and journalism became fully pro-fessional in the twentieth century. The separa-tion of ‘facts’ and ‘opinion’, information andcomment was the complex outcome of the com-mercial pressures of mass production and theextension of democracy. Objectivity became thelodestone of professional journalism in theearly nineteenth century, not out of a naiveempiricism and faith in ‘facts’, but rather thereverse. It was a bulwark, Schudson argues,against scepticism and anxieties about the cor-rosive relativism, the arbitrariness of values,and the subjectivity of belief in a world whereone man’s opinion was as good as another’s and neither had any special claim to truth.Journalistic objectivity was one response to ‘thehollow silence of modernity’ (Schudson 1978:158). Yet even as it was articulated as a pro-fessional value in the 1930s, the criticism of ‘themyth’ of objectivity was enunciated, and thiscriticism became routine in the adversarial Leftculture of the 1960s (from which Curranattacks the dominant interpretation of Britishpress history).

� journalistic objectivity – Chapter 5, p. 87

HISTORIES OF BROADCASTING

Radio and television are products of the twen-tieth century, and their historical significance aswell as their histories have only recently begunto emerge. The oldest, and longest running,history of broadcasting is by the British hist-orian, Asa Briggs, who began working on theinstitutional history of the British BroadcastingCorporation (BBC) in the 1950s. The fifthvolume in this massive history was published in1995 (at which point Briggs laid down his pen), and a sixth volume (to be written by JeanSeaton) has just been commissioned. WhenBriggs began, he was operating, as he noted, inuncharted waters. There were no general his-tories of broadcasting in any country to serveas models. The three-volume history of Ameri-can broadcasting by Erik Barnouw (1966–1970) had not yet appeared. Neither commu-nication nor media studies had yet appeared asacademic fields of study.

Briggs therefore found himself in the posi-tion of ‘inventing’ broadcasting history, and hisreflections on the methodological and otherproblems he faced are thus of particular inter-est. He wanted to write a ‘total history’ thatcovered all aspects of the activities of the BBC.This was a truly monumental undertakingsince, by the time he got up to World War II,the BBC’s activities were on a global scale; in1942 it was broadcasting in the short-waveradio band to all parts of the world. There werereal difficulties of narrative organization andcoherence as his history moved forward slowlyon all fronts: home broadcasting, overseasbroadcasting, radio, and, later, television. Therewere difficulties of writing an institutionalhistory, which might result in a ‘history fromabove’ (reproducing the management view ofthe BBC and underplaying the attitudes ofinsubordinate dissidents in the ranks). Therewas the danger of not being able to see thewood for the trees: it was hard to find roomfor a historical account of the history of output– the programmes themselves – and theirimpact and effect on contemporary society.There was, finally, the issue of perspective.Briggs was anxious to avoid ‘reading back intothe past current fashions of description and


BBC:institutionalhistory

explanation.’ He wanted his history to be,above all else, definitive: ‘to be as accurate andas well backed up (in footnotes) by all avail-able evidence as possible.’ He elected to quoteat length from original sources – even thoughthis was not then fashionable among historians– partly to capture lost modes of expression andmood, but also to enable readers who mightdraw different conclusions from his own assess-ments to have the necessary amount of pub-lished evidence before them (Briggs 1980: 8).

Briggs’s work is, above all, a historian’shistory, which allows subsequent researchers topursue in greater detail the study of, say, pro-gramme making without having to reconstructyet again the house that Briggs built. The socialhistory of the beginnings of broadcasting in theUK that I wrote with David Cardiff (Scannelland Cardiff 1991) could not have been writtenwithout the prior existence of a definitive ‘total’history of the BBC. Historians of public servicebroadcasting in other European countries havealso acknowledged their debt to Briggs’s history(e.g., Dahl 1976).

It is a truism that histories are as good asthe archives on which they depend. Historiansof British broadcasting are lucky to have sucha remarkable resource as the BBC WrittenArchives at Caversham (near Reading), whoseholdings currently cover all aspects of the BBC’sactivities from the earliest days through to theend of the 1970s. Historians of the beginningsof broadcasting in other countries have had amore difficult task, since radio almost every-where began on a local, commercial basis, anddata about the very first radio stations have allbut vanished. The perishable character of thespoken word shows up in stark contrast withthe permanent trace of writing. Leslie Johnson,historian of early Australian broadcasting,notes what a difficult and disheartening task itwas to gather information about the radio sta-tions of the 1920s and 1930s (Johnson 1988:viii–ix). Susan Smulyan, in the absence of datafrom the stations themselves, draws heavily onearly popular radio magazines for her historyof the commercialization of American radio inthe 1920s (Smulyan 1994). C Cecile Meadelfaced similar difficulties when writing thehistory of the beginnings of French radio. When

she did find data on some of the many radiostations of the period, they were invariablyscanty, poorly kept and unclassified (Meadel1994: 16–17).

In spite of such difficulties, in NorthAmerica and in many European countries thereare now in place histories of the beginnings anddevelopment of national broadcasting systems,starting with radio and continuing with televi-sion. These institutional histories have beencomplemented by more detailed ‘genre’ studiesof particular aspects of broadcast output; news,drama, and documentary, for instance, havingreceived considerable attention. Alongside theseare social histories of audiences and of the placeof radio and television in everyday life (e.g.,Lacey 1994; Moores 1988; O’Sullivan 1991;van Zoonen and Wieten 1994).

‘DOING’ HISTORY

The past in its fullness is beyond recall. Onlyits traces remain today, and the task of piecing together the fragments, interpreting and reworking them into coherent narrativeaccounts and other formats, is, of necessity, acumulative and collaborative process. There isno one ‘correct’ way of doing this. Historicalwork is partly determined by the available data(or lack of it), and partly by the predispositionsand attitudes brought to bear upon the past byhistorians in their present. The literature refer-enced above is characterized by the diversity ofits approaches and attitudes to history, butthose attitudes are only partly determined bythe concerns and prejudices of the historian.They are also shaped by the nature of the topicand the extent to which it is, or is not, estab-lished as an object of historical inquiry. In thisrespect, the problems of ‘doing’ broadcastinghistory are instructive.

Why should it be that in the first placebroadcasting histories tend to be accounts ofparticular institutions? We may note that thefirst ‘period’ of broadcasting (a historiographi-cal construct, of course) is the history of radioand its institutionalization within the context ofparticular nation-states. For although radiobegan, almost everywhere, on a small-scale localbasis, technical, distributional and economic

‘Doing’ history 201

1111234567891012341115678920123456789301234567894012345678950111

BBC: socialhistory

historicalarchives

histories ofgenres

pressures rapidly led to the consolidation of the production and distribution of programmesinto centralized, large-scale, national institu-tions, funded either by advertising revenues orby some form of hypothecated tax (a ‘licencefee’). The history of broadcasting thus beginswith accounts of that process. When that is inplace, other historians can engage with whatmight be called ‘second-generation’ history.That is how I think of my own historical workon British broadcasting. The existence ofBriggs’s history made it possible to go back intothe archives and begin a closer, more attentivestudy of the complex business of making pro-grammes, across the major areas of broadcastoutput. How were the practices of gatheringand broadcasting ‘news’ developed? How didbroadcasters discover how to produce enter-taining programmes for radio? How was music(what music? whose music?) developed as amajor area of radio output? Such questions canonly be addressed in detail if accounts of theinstitutional framework within which theseactivities were situated are already in place. Itthen becomes a complex business of gatheringdata, interpreting it and putting it into a coher-ent narrative frame.

It requires patience and persistence to siftthrough a holding as massive as the BBCWritten Archives. Such work should ideally bedone in a concentrated, continuous way. It isno use doing it occasionally over an extendedperiod. You have to immerse yourself in anarchive as I did, several days a week continu-ously for over a year, in order to feel your wayaround the holdings (so that you get a sense ofwhere to look for a ‘missing link’), but aboveall in order to begin, in time, to see how thebits and pieces fit together. It is like doing agigantic jigsaw puzzle, except that a lot of thepieces are inevitably missing, and so you willhave to fill in the blanks by circumstantial evi-dence where necessary. Archival work, then,shares several features with other qualitativeresearch, including an extended stay in a localeas well as iterative procedures of sampling andanalysis.�

The most obviously ‘lost’ pieces in thepuzzle, for historians of early radio, are thebroadcast programmes themselves. The BBCbegan broadcasting in 1923, but there are noextant recordings of any programmes until theearly 1930s, and only a tiny fraction of outputbefore World War II was recorded and pre-served. So how do you deal with this? There is, of course, a complete record of what wasbroadcast, day by day, in the programmeschedules of the Radio Times, which started uponly months after the BBC got going in 1923.Before the war, all programmes were scripted,and in many cases the transcripts of the pro-gramme as broadcast (with last-minute emen-dations and corrections pencilled in) were kept.For example, a very good picture of develop-ments in radio documentary (or ‘features’ asthey were called at the time in the BBC) can bebuilt up from recordings and transcripts of pro-grammes, and from a number of very usefulproduction files which reveal (in the best cases)the whole process of the making of a particu-lar programme from the initial idea through totransmission (Scannell 1986).

However, to take a notorious example, thereare no transcripts of prewar news broadcasts(with the exception of the General Strike in1926, and transmissions at the height of theMunich Crisis in 1938). The practice of keepingall bulletins began only with the outbreak ofwar. We simply do not know what news bul-letins sounded like nor what, in detail, thestructure and content of any bulletin was likebefore the war. Accounts of the beginnings ofbroadcast news are thus, in this strict sense, cir-cumstantial. There is, nevertheless, a wealth ofsecondary source material about news fromwhich a reasonably accurate account of itsdevelopment may be built up. The files of theBBC’s various management, policy-making and departmental committees were invariablyuseful. The magnificent press-cuttings collec-tion was invaluable. Contemporary biographiesby politicians and broadcasters were occasion-ally illuminating. Hansard, the verbatim recordof the day-to-day proceedings of Parliament,was indispensable whenever there was apolitical row about news.


� qualitative empirical methodologies – Chapter 14

recordings ofprogrammes

schedules andtranscripts

organizationaldocuments,press cuttings,and bio-graphies

‘Doing’ history 203

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 12.1 HISTORICAL RESEARCH ON ARCHIVAL DATA

On top of questions concerning the availability of data, the nature of the topic and its status inprevious research, one needs to be alert to connections between pieces of data one comes acrossin different contexts and on apparently different issues. In one of the many news files I examined(Scannell and Cardiff 1991), I came across an exchange of letters between the Home Office (theBritish ministry of interior affairs) and the BBC.The Home Office wrote to John Reith (DirectorGeneral of the BBC) on 31 October 1932, querying the inclusion in the previous night’s bulletinof a protest from the Metropolitan Police Federation about proposed cuts in police pay. Wherehad the BBC got this story, it wanted to know? Reith immediately replied that it came from theusual agency sources (the Press Association) and added that he had given instructions that ‘nostuff from the agencies re cuts in pay, protest etc is to be broadcast . . . only HMG’s [His Majesty’sGovernment] official statements in future.’ He further remarked that the same bulletin had includeda message telephoned through to the News section from the Commissioner of Police, asking thepublic not to go out to watch demonstrations (Scannell and Cardiff 1991: 46–47).

It was difficult to make much sense of this at first, in the absence of any records of news bul-letins, but I noticed the date and recalled that around this time unemployment had peaked at over3 million. I then remembered accounts of the National Unemployed Workers’ Movement which Ihad read as background to the BBC’s handling of the issue of unemployment (Hannington 1977;Kingsford 1982).When I checked out Wal Hannington’s book, all became clear. A massive protestmarch had arrived in London a day or so earlier and ‘subversive’ leaflets had been distributed inadvance, addressed to the police force: ‘Policemen! Defeat your own pay cuts by supportingTuesday’s demonstration against the economies!’ (Kingsford 1982: 156). On the day of the demon-stration – the day the Home Office queried the news item – there had been a huge gathering inTrafalgar Square and an impassioned speech from Hannington appealing to the police not to usetheir batons against the marchers and the large supporting crowd. But violence broke out, andtwo days later Hannington was arrested and charged with attempting to cause disaffection amongmembers of the Metropolitan Police, contrary to the Public Order Act of 1919.

This example well illustrates the delicate hermeneutic� task that the historian faces whenconfronting primary data. As always, the date of the document, if available, is a crucial clue. Butthe text remains an enigma without detailed knowledge of what was happening in the country atprecisely that time. Only when that has been discovered does the task of historical interpretationproperly begin.What exactly does this nugget of information tell us? Does it show something ofthe subservience of the broadcasters (of Reith) to the government of the day? Perhaps. But howshould that be understood? And here, the attitude of the historian comes into the equation. Forit would be easy to construct some kind of conspiracy theory of hidden complicity between thebroadcasters and their political masters in the management of news in ways that served the inter-ests of those in power, thereby showing the role of the BBC in maintaining ‘the dominant ideol-ogy’ or something similar. Broadcasters themselves, however, were well aware of such dilemmas.I vividly remember the moment when I came across a memorandum, written days after the MunichCrisis, by the ex-Head of News, in which he declared that ‘we [the BBC] have taken part in con-spiracy of silence’, and went on to say that the BBC had failed the British public by not giving itthe necessary information as to the causes of the crisis or its implications – the inevitability ofwar (Scannell and Cardiff 1991: 88–89). A simple ‘reading’ of the developing relations betweenbroadcasting and politics, starting with the General Strike of 1926 and going through to the out-break of war, could plausibly develop a ‘Bad News’ thesis (Glasgow Media Group 1976) about theBBC’s suppression and distortion of what was happening at the time. But this could be achieved

cont.� hermeneutics – Chapter 2, p. 21

The study of any aspect of broadcasting,and of other media, is impossible withoutknowledge of the wider political, social and cul-tural contexts within which they are situated.Again and again in our study of the productionprocess within the BBC, David Cardiff and Ihad to reconstruct other histories: of entertain-ment in the prewar period, or of the state ofmusical culture at that time. There were nostandard social or cultural histories to guide uson such matters, and so we had to reconstructthem ourselves from contemporary bio-graphies, non-academic accounts and frommagazines. It is essential to have a feel for thewider contexts within which broadcasting operates in order to write informed historicalaccounts of the activities of the broadcasters. Ido not wish to be prescriptive as to how this isachieved. History is not a social science, and ittends to resist the impositions of rigid method-ologies. The precise method of any particularhistorical investigation will be shaped andinformed by the nature of the topic and the

available resources, but a willingness to readwidely and to explore supplementary sources isvital. To this, I would add a sense of historicalcontinuities and change over longer time spansthan the actual period of study. The historianmust look at what came before and after the topic which engages her or him. There is atendency today still to write of the BBC and themeaning of public service broadcasting� as ifit were set in place, once and for all, by its firstDirector-General, John Reith. But the meaningof public service has undergone significant andimportant changes in the past eighty years, andnecessarily so, since it operates today in condi-tions very different from the 1920s when itbegan (see Scannell 1990, 1996b).

CONCLUSION

Histories of communication and of the mediaconverge in the twentieth century, as is appar-


only by ignoring all the evidence of attempts by the broadcasters to resist the manipulations ofgovernment, most notably over two of the most sensitive political issues of the 1930s: unem-ployment and foreign affairs. We have documented in detail the resistances of programme- andpolicy-makers in both cases (see Scannell and Cardiff (1991: 64–68) on the 1934 Talks series onunemployment called Time to Spare; and pp. 74–78 on a 1935 series, The Citizen and His Government,which was suppressed by the government because it contained communist and fascist speakers).

The past must be understood in its own terms, in the first instance, before rushing to judge-ment with the wisdom of hindsight. In this regard, distant periods pose some of the same analyt-ical challenges as other cultures.� Interpretations of the beginnings of broadcasting should besensitive to the problems encountered by the broadcasters as they became involved in politics.It should be remembered that the formal conditions of full representative democracy were notestablished in the UK until 1919 by The Representation of the People Act.The meaning of democ-racy is something that gets worked out and discovered through the actual processes of history.This learning process has been at the heart of the relationship between broadcasting and politicsfrom the beginning through to the present. A hermeneutics sensitive to this unending processwould note, as we tried to, how the broadcasters gradually came to understand their own com-plicity with the political powers in the ‘management’ of news and information in the interwarperiod (Scannell and Cardiff 1991: 23–133). It would point to the basic problems for broadcast-ers of discovering how to ‘do’ informing, news-gathering and news-telling. There was no NewsDepartment in the BBC until 1934, and the practices of contemporary broadcast journalism withwhich we are all familiar today go back no further than the 1950s in the UK and the USA, andthe 1960s and later in most other countries.

cont.

� cross-cultural and comparative research – Chapter 11, p. 179

� public service broadcasting – Chapter 16, p. 277

ent in the works by Peters (1999) and Kittler(1999) already cited, and in Kern’s (1983) fas-cinating study of ‘the culture of time and space’at the turn of the last century, and Winston’s(1998) synoptic history of media technology.The early ‘mainframe’ institutional histories ofthe press and broadcasting have been supple-mented by all kinds of histories on particularaspects of the press and broadcasting in the pasttwenty-five years. The establishment of acad-emic journals devoted to historical research� –Media History and The Historical Journal ofFilm, Radio, and Television – also indicateshow well established this hybrid discipline, splitacross history and media and communicationstudies, has become.

What of the future? There is, first, a press-ing need for historical work on all aspects ofmedia and communication elsewhere in theworld other than Europe and North America.If we take, for instance, the continent of Africa,only recently emerged from the epoch of whitecolonial domination, it is a difficult matter, inrespect both of principle and method, to beginthe necessary business of retrieving the historyof a continent so long robbed of its own iden-tity. In the postcolonial era,� along with every-thing else, there is the delicate question,forcefully posed by Ngugi Wa Thiong’o, con-cerning the decolonization of mind. For whomdo African intellectuals write, in what language,and from what perspectives (Ngugi 1986)? Theacademic disciplines, whose normative status islargely taken for granted in the West, them-selves pose tricky matters for African aca-demics. ‘History’ as written and taught in theeducational system of South Africa at everylevel in the apartheid era is, post 1994, in needof, and receiving, a thorough rethinking andrewriting. The press and broadcasting (radio isan important medium in many African coun-tries) are a double-edged inheritance from thedeparted imperial powers. In Africa, they sitalongside oral traditions for both poetry and

song which may still be, for the majority ofAfricans, the dominant means through whichstories circulate and opinions are formed. Theinterpretation of African culture in its ownterms – without depending wholly on the intel-lectual and theoretical frames of the West – isa matter of concern in the emerging field ofAfrican media and cultural studies. There iscurrently a lively debate about the significanceof ubuntu for the interpretation and under-standing of African society and culture ingeneral and the media in particular (Blanken-berg 1999; Kamwangamulu 1999).

The profoundly ruptured history of com-munication in Africa – a history barely begun,because there was no time, before now, that itcould be written – will be marked by these pres-sures and tensions. Meanwhile, what remainsto be tackled in Europe and North America aresynthetic media histories of one kind oranother. A comparative review of national his-tories of the press and broadcasting wouldreveal, on the one hand, the cultural specificityand difference, country by country, in thesedevelopments, and, at the same time, theircommon patterns both in form and content. If it is a basic theoretical and methodologicalquestion for historians as to whose historiesthey are writing, one major concern is how tomove beyond the histories of particular nation-states. Comparative media history does not yetexist, but it could begin to reveal the sharedstructural characteristics of national mediasystems and the unities and differences of mediaoutput in different countries and in differentparts of the world. Such work might thenconfirm or disprove the claims of ‘mediumtheorists’ (mostly followers of McLuhan; seeMeyrowitz 1994) as to the allegedly inherent,structural characteristics of written and elec-tronic forms of communication, and therebybegin to reconcile those histories that privilegeform with those that privilege content.

Conclusion 205

1111234567891012341115678920123456789301234567894012345678950111 � journals and reference works – Chapter 1, p. 3 � postcolonial theory – Chapter 11, p. 179

ubuntu:‘humanness’(literally:‘I only existbecause of my ties to others’),emphasizingcommunitylife, collectiveresponsibility,and com-munication

comparativemedia history

1111234567891012341115678920123456789301234567894012345678950111

Practice

SCIENTIFIC APPROACHES AND SOCIAL IMPLICATIONS

Doing media researchMedia and communication studies have developed a variety of analytical procedures and designs, drawing on thesocial sciences as well as the humanities. In overview, empirical studies have relied on six different prototypicalmethods, as laid out in Figure III.1.

In addition to the qualitative-quantitative distinction, these methods are characterized by their forms of datacollection and the resulting types of evidence. Each basic method, as employed singly or in combination withother methods, raises a number of methodological, theoretical, epistemological, as well as political questions.

The following chapters present key concepts and criteria for the practice of empirical research.Two chaptersdescribe and exemplify how various qualitative and quantitative methodologies may be applied to different typesof research questions. One chapter explores and illustrates the complementarity of the two types of approaches:

• Quantitative studies are covered both in terms of their basic categories and operations, and with referenceto the different relevances of surveys, experiments, and content analyses;

• Qualitative studies are presented, similarly, in terms of the systematics of interviewing, observation, andtextual research, including issues of data collection and data analysis;

• The complementarity of qualitative and quantitative research is examined with reference both to concreteexamples and to classic problems in the philosophy of science.

IIIPART

Speech/verbal language

Action/behavior

Texts/documents

Qualitative

interviewing

observation

discourse analysis

Quantitative

survey

experiment

content analysis

Figure III.1 Six prototypical empirical methods

cont.

Unification in the final instanceThe chapters take as their premise that different methodological approaches can and should be combined soas to supplement each other. It is this premise, as borne out by a number of examples and arguments, whichis summed up in the heading ‘Unification in the final instance.’ While much of this volume has already tracedthe convergence of previously distinct research traditions, the following chapters present and review a varietyof practical resources available for present and future media studies.

Media research as social practiceLike mediated communication, media studies are themselves an integral part of the societies and cultures beingstudied.The final chapter returns to the social origins and contexts of research, and to the interests guiding dif-ferent types of projects in different sectors of society. Many media students and researchers wish to make adifference through their work. One lesson from the field so far is that media studies necessarily make a differ-ence.The question is how, in whose interest, and through which empirical and theoretical approaches?

cont.

INTRODUCTION

The field of media and communication researchis characterised by quite a variety of differentresearch perspectives. That fact stems from thehybrid nature of this field of empirical inquiry,in which investigative approaches have beenderived from longer established academic disci-plines in the social sciences.� Anthropology,economics, geography, history, linguistics,political science, psychology, sociology have allcontributed theories and methodologies forstudying the structure, organisation, content,uses and impact of media. While media schol-ars have, accordingly, debated the merits andshortcomings of different theories and method-ologies within limited spheres of inquiry(Neuman 1994; Wimmer and Dominick 1994),perhaps the most significant debate within aca-demic circles (though not the highest profile onein the public sphere) has centred on a disputebetween different philosophies of social scienceabout the research perspective that offers themost sensitive and meaningful insights into the

role and influence of the media in society. A‘positivist’ or hypothetico-deductive school ofthought has been lined up against critical andinterpretive perspectives.� These differentsocial-scientific perspectives vary in terms of theperceived objectives of research, the way socialreality and human beings are conceived, therole of theory-driven empirical inquiry and thekind of evidence to which most weight is given(see Neuman 1994). Hypothetico-deductiveapproaches to media inquiry are concernedwith the setting up, proving or disproving ofhypotheses, and the eventual establishment of theoretical explanations of events or causallaws which explain relationships between indi-viduals’ activities in, and experiences of media,and their knowledge, beliefs, opinions andbehaviour. These phenomena are usually oper-ationally defined in quantitative terms to facil-itate measurement of the strengths of causallinks or degrees of association between them.

It is not the purpose of this chapter to elab-orate upon the distinctions between these

1111234567891012341115678920123456789301234567894012345678950111

The quantitative research processBarrie Gunter13

• an overview of basic concepts within quantitative research, including hypothesis testing and sampling• reviews of survey research, quantitative content analyses and experimental studies of media, with analytical

examples• a comparison of surveys and experiments, their strengths and weaknesses• a presentation of quantitative data analysis, including examples of statistical procedures.

� positivism and other philosophy of science – Chapter 15� social-scientific sources of media studies – Chapter 3

philosophies of social science, nor to discusstheir relative strengths and weaknesses. Thischapter is concerned with the nature of quanti-tative research, examines the basic concepts ofthis kind of empirical inquiry, and discusses theprincipal forms it takes. The latter will bedivided in terms of whether their main aim is toinvestigate associative links (surveys) or causallinks (experiments) between variables. A furtherdistinction will be made between methodologiesconcerned with the study of either media audi-ences or media content – the two foci of quan-titative media studies. Since quantitative studiesof media content largely involve a form of sur-veying of media output, this methodology willbe discussed under the heading of surveyresearch. It should perhaps be noted at thispoint, however, that the systematic analysis ofmedia output may form part of an experimen-tal investigation, too.

The chapter will also examine some of thebasic principles of quantitative data analysisand presentation. In quantitative research, mea-surement is conducted through numbers. Thequality of the research is crucially affected bythe effectiveness of data processing, analysisand interpretation. The discussion of quantita-tive research methods inevitably examines notjust their inherent strengths and weaknesses,but also when their application is appropriateand inappropriate. To be complete, this reviewalso considers the evolution of quantitativeresearch methodologies in relation to theoreti-cal models of media analysis, and attempts toaddress any weaknesses of this link.

BASIC CONCEPTS IN QUANTITATIVERESEARCH

The basic concepts that characterise quantita-tive research methodologies concern relevantmodes of measurement and procedures toanalyse the relationships between such mea-surements. A central notion is the variable,which is linked to additional fundamental ele-ments of quantitative research such as conceptsand constructs.

A concept represents an abstract idea thatembodies the nature of observable phenomena,or an interpretation of why such phenomena

occur. For example, individuals may be differ-entiated in terms of their use of media – heavyusers may be distinguished from light users.This media usage, in turn, may be linked to pat-terns of social behaviour, and may even be usedto explain different behavioural patterns. In thiscontext, media usage becomes an explanatoryconcept. In a different context, the concept of‘cultivation’ may be used to describe distinctivepatterns of perceptions or beliefs that individ-uals who are heavy media-users hold, andwhich distinguish them from the perceptionsand beliefs held by light media-users (Gerbneret al. 1986; Wober and Gunter 1988).

A construct comprises a combination ofconcepts. This term is often used to describe adefining characteristic of individuals that isassociated with their personality type. Forexample, the personality construct of ‘sensa-tion-seeking’ is used to distinguish betweenpeople who seek varying levels of optimal stim-ulation from their environment (Zuckerman1994). High sensation-seekers generally needhigher levels of environmental stimulation thanlow sensation-seekers. High sensation-seekers,for instance, may be characterised by such con-cepts as sociability, tolerance for strong stimu-lation, risk-taking and a sense of adventure.Constructs have a dimensional quality, so thatindividuals may be classified as high or low onthe personality dimension of sensation-seeking.

A variable is an empirical representation ofa concept or construct. Whereas concepts andconstructs have an abstract quality, variablesprovide operational measures that can be quan-tified and manipulated by researchers. Forexample, gender, age and socioeconomic classare variables. Personality characteristics such asaggressiveness, locus of control, extraversionand neuroticism can all be treated as variables.The amount of newspaper reading or televisionviewing also exemplify variables. Variables maybe further defined and differentiated in terms of their constituent attributes. Attributes arevalues or categories into which variables can bedivided. In the case of gender, there are two cat-egories – male and female. In the case of age,individuals can be differentiated by age groupor actual age. Television viewers can be differ-entiated by amount of viewing into values such

The quantitative research process210

concepts

constructs

variables withattributes

as light, medium and heavy viewers on the basisof self-reported or independently observedviewing behaviour.

Types of variables

Variables can be defined further in terms oftheir relationship with each other. A funda-mental distinction is made between indepen-dent and dependent variables. The independentvariable is a variable which can be manipulatedby the researcher. It is also known as the‘causal’ variable, in that it is a concept or con-struct that is believed to produce some mea-surable response or outcome. The independentvariable is also referred to as the predictor vari-able in some studies. The dependent variable isthe measure of the response or outcome. It istherefore also known as the ‘effect’ or ‘cri-terion’ variable. A principal objective of quan-titative research is to establish the closeness of the relationship between independent anddependent variables (Neuman 1994; Wimmerand Dominick 1994). Ultimately, researcherswish to provide evidence that a particular inde-pendent variable has a causal relationship witha particular dependent variable (Bailey 1994).

To illustrate this idea, a media researchermay be interested in demonstrating that tele-vision news can improve the audience’s know-ledge of a political event or issue.� Onetheorist might argue that this outcome is depen-dent upon the way political news is presentedin a news bulletin. Another might argue that itdepends upon the frequency with which indi-viduals are exposed to such stories (Gunter1987b). In the first case, a study might belaunched in which the position of a politicalnews story in a bulletin is varied. For someviewers, the story is presented at the beginningof the bulletin, while for others it occurs at theend or in the middle. A serial-position hypoth-esis would lead one to predict that presentationin the middle of the bulletin would result inpoorer post-viewing recall of the story byviewers (Gunter 1979; Tannenbaum 1954). Inthis case, the independent variable is the posi-tioning of the story and the dependent variable

is the measurement of story content recall byviewers after the bulletin has been presented.

In the second case, a sample of viewersmight be presented with a questionnaire thatasks them to report how often they havewatched television news over the past week ormonth (Gunter 1985b). Similar questions mightbe asked about exposure to radio news andnewspapers. The same respondents would thenbe asked a series of knowledge questions abouta political news story. A separate analysis mightbe carried out of the frequency with which thatstory was covered by the news media. The lattermeasure could be combined with the measureof self-reported news media exposure to definethe independent variable of relevant politicalnews exposure. Respondents with heavy andlight news exposure could then be comparedregarding their scores on the measures ofpolitical story knowledge, which would be thedependent variable here.

Hypothesis testing

Quantitative research is primarily concernedwith demonstrating cause–effect relationships,and any research project begins by setting up ahypothesis. A hypothesis is a proposition to betested, or a tentative statement of a relationshipbetween two variables.� While hypothesistesting is not unique to quantitative research, itis one of its fundamental elements, and almosta required aspect of quantitative academicresearch. (Qualitative research, under a ‘posi-tivist’ school of social science, may eschewhypothesis testing if its aim is purely explora-tory, in which case it is concerned more withthe discovery of potential hypotheses for futuretesting.)

Hypotheses make prognostications aboutthe links between variables (Bailey 1994). Theypropose that under one set of conditions, if an independent variable is manipulated in acertain way (as in experimental studies) or isassumed to have a certain strength (as in surveyresearch), it may be expected to exert a mea-surable impact on a designated dependent vari-able. The researcher then sets out to discover if

Basic concepts in quantitative research 211

1111234567891012341115678920123456789301234567894012345678950111

independentand

dependentvariables

� recall studies – Chapter 9, p. 144 � hypothetico-deductive research – Chapter 15, p. 262

that prediction holds true. The essence of quan-titative scientific enquiry is to prove or disprovehypotheses, and the outcome is seen as a con-tribution to the growth of knowledge. Scientistsrarely restrict themselves to testing single hypo-theses, however. Through repeated hypothesistesting and the development and verification ofnew hypotheses, a body of understanding isdeveloped that essentially comprises a series of acceptable and accepted explanations for arange of dependent variables.

Reliability and validity

A core aim of social scientific inquiry, particu-larly within the ‘positivist’ tradition, is the estab-lishment and demonstration of the reliabilityand validity of research findings. Reliability con-cerns the dependability and consistency of therelationship between two variables or in thescore obtained on a single variable at more thanone point in time. Validity indicates whether ameasure properly captures the meaning of theconcept or construct it represents. If heavy view-ing of violent programmes on television isbelieved to cause aggressive behaviour in view-ers, then for this belief to be accepted as true,two conditions must be met. First, repeated evi-dence must emerge that exposure to violence ontelevision is followed by increased aggressive-ness. Second, the measures of exposure to tele-vision violence and of aggressive behaviour mustaccurately represent those behaviours.

Reliability can be established by carryingout repeated tests of phenomena and relation-ships between phenomena, by repeating suchtests among different groups of people with thesame results, and by having several researchersrun the same test (Siegel and Hodge 1968).Where a particular concept or construct – forexample, a personality dimension such as extra-version or sensation-seeking – is measured by aseries of items, it should be possible to select atrandom any 50 per cent subset of the total setof items and find that they differentiate betweenindividuals in the same way as any other 50 percent subset of those items. This is known as split-half reliability testing. Alternatively, aresearcher could construct two different ques-tionnaires designed to measure the same

concept and administer both to the same groupof respondents. This is called multiple forms(Goode and Hatt 1952) or alternate forms(Sellitz et al. 1976) reliability.

Validity can be much more difficult to estab-lish with certainty. ‘Measurement validity is thedegree of fit between a construct and indicatorsof it’ (Neuman 1994: 130). Partly because ofthis complexity, validity is assessed in severalways (Figure 13.1):

• Face validity offers a basic level of judge-ment that a measured variable really measuresthe phenomenon it represents. A test of proof-reading ability might ask individuals to readand correct errors in a passage of text. The testwould provide a clear behavioural representa-tion of the kind of ability being measured.

• Predictive validity assesses the ability of ameasure to predict a future event that is logi-cally connected to a concept or construct. If atest of extraversion distinguishes between howsociable and outgoing people are, then highscorers should be observed to initiate more con-versations and speak to more people at a partyfull of strangers than would low scorers.

• Concurrent validity means that a measure is associated with another indicator that hasalready been shown to be valid. Thus, a newmeasure of sociability would be expected toexhibit a high and significant correlation withscores on an established extraversion scale.

• Construct validity is a more complex methodof establishing a measure’s validity. Since a con-struct usually comprises a collection of conceptsand indicators, construct validity requires that a new measurement is shown to be related to a variety of other established and previously verified measures. In the develop-ment of a questionnaire measure of aggression,for example, researchers may ask whether it differentiates among individuals who might beexpected to exhibit different aggression levelson the basis of other established aggressionmeasures, or whether the new questionnairemeasure distinguishes among individuals whomother judges (e.g., friends, peers, parents, teach-ers) have distinguished in the same way, or among whom one would expect such differ-


research ascumulation of

knowledge

reliabilitythrough

repetition ofmeasurements

and tests

split-halfreliability

testing

multiple oralternateformsreliability

ences to occur on theoretical grounds (e.g.,gender differences in aggressiveness) (Milavskyet al. 1982).

A distinction may also be made betweeninternal validity and external validity. Internalvalidity means that the design of a researchproject is free from theoretical or methodolog-ical error, and is a term mostly used in experi-mental studies. An experiment’s results could beinvalidated if it emerged that its measurementsfailed to capture the phenomena they purportedto measure. External validity is also usedprimarily in the context of experiments. Thisconcept addresses whether the results can begeneralised to other situations or groups ofpeople. Low external validity means that theresults are unique to the specific experimentalsetting in which they were obtained, but areunlikely to occur anywhere else.

Levels of measurement

There are four principal types of measurementin quantitative research. The type of measure-ment reflects the kind of concept or constructit represents. Some measurements make fairlysuperficial or crude distinctions between en-tities, while others operate at a higher or morerefined level. An initial distinction may be madebetween continuous and discrete variables.Discrete variables can be measured at the nom-inal or ordinal levels, while continuous vari-ables can be measured at the interval or ratiolevels. Continuous variables have a number ofmeasurable values which can be located alonga mathematical continuum, and measurementis along a scale that rises in increments. Hence,the measurement of time can be made in termsof seconds or minutes; the measurement of dis-

tance can be made in terms of various units oflength. Discrete variables, instead, make dis-tinctions according to relatively fixed attributes.Objects fall into one category or another.Examples of this form of measurement includegender (male or female), marital status (single,married, divorced, widowed) and religion(Protestant, Catholic, Muslim, Hindu, etc.).

• The nominal level is the weakest form ofmeasurement. Numbers can be used here onlyto signify categories of objects. For instance,voters can be classified in terms of the specificpolitical party they voted for. Any ‘object’which is thus placed within a particular cate-gory is deemed to be equivalent to any other,and there is no indication of the degree towhich an object belongs to a category.

• At the next level of ordinal measurement,objects of analysis are ranked along a dimen-sion, such as smaller to greater, or lower tohigher. In the case of socioeconomic class, forexample, people may be ranked as belonging todifferent classes, with some classes deemed tobe higher than others. Middle class is thereforehigher than working class, but there is no indi-cation given of how much distance lies betweenone class and another.

• At the interval level, entities are measuredalong a dimension that has equal intervals. One commonly used example is temperature.The temperature scale rises in degrees with each degree mathematically equal to any otherdegree. The weakness of the interval scale is thatit lacks a true zero or a condition of nothing-ness. An intelligence scale is another example ofan interval scale. Lacking a true zero, however,it is not possible to say that a person with an IQof 100 is twice as intelligent as one with an IQ of 50.

Basic concepts in quantitative research 213

1111234567891012341115678920123456789301234567894012345678950111

Judgement-based

Face validity

Criterion-based

Predictive validity

Concurrent validity

Theory-based

Construct validity

Figure 13.1 Types of validitySource: Wimmer and Dominick 1994: 99

internal andexternalvalidity

continuousand discrete

variables

• The ratio scale is the most powerful form ofmeasurement. This has all the properties of aninterval scale and the existence of a true zero.Time, distance and speed are examples of ratiolevel scales. An object moving at ten miles anhour is moving exactly twice as fast as onemoving at five miles an hour.

These basic concepts are applied in quanti-tative research, in particular, to measure mediaaudiences, media content and cause–effect rela-tionships between media and audiences.Quantitative methods have demonstrated eitherassociations between media and audiences ordirect, causal connections between them, and inthe remaining sections of this chapter, thesemethods are examined in turn. In each case, theprincipal designs are reviewed first, beforeattention shifts to illustrations of how they havebeen applied by media researchers.

SEARCHING FOR MEDIA–EFFECTASSOCIATIONS: SURVEY RESEARCH

Surveys are a major form of quantitativeresearch that does not involve any manipula-tion of participants or their circumstances inadvance. Surveys collect data after the fact.Because they obtain information from respon-dents about their knowledge, beliefs, attitudes,values and behaviours on a post hoc basis,surveys cannot test cause–effect relationshipsdirectly. Surveys instead explore relationshipsor degrees of association between variables.Thus, surveys are entirely dependent upon self-report information supplied by respondents,whereas experiments can complement ques-tionnaire responses with direct observations byresearchers. In the media and communicationcontext, surveys have been conducted with boththe general public (i.e., media audiences) andspecialised groups (e.g., media producers). (Forfurther information about best practice in con-ducting surveys, see Babbie 1990; Fink 1995a,1995b; Oppenheim 1992.)

The original form of the modern survey, his-torically, was the census (Converse 1987; Moserand Kalton 1971). A census compiles informa-tion about the characteristics of an entire popu-lation. Early censuses were conducted to assess

property ownership for taxation purposes, butthey also provided a means to establish the avail-ability of young men for military service and, indemocratic societies, assisted in the division ofpopulations and territories into constituencieselecting their representatives in government.Surveys, in their turn, were developed to docu-ment poverty following industrialisation andurbanisation in the nineteenth century.

While censuses attempt to obtain data fromeveryone in a population, surveys use samplingtechniques to select subsets of a population foranalysis. With a population numbering manymillions of people, it is usually not feasible toquestion everyone, so that smaller and moremanageable numbers must be selected for datacollection. The most important objective hereis to ensure that the achieved population subset,or ‘sample’, represents the population as awhole. During the first half of the twentiethcentury, survey research benefited from ad-vances in scientific sampling procedures andquestionnaire design techniques.

A classic study illustrating the early adoptionof survey research regarding media influenceswas made by Lazarsfeld et al. (1944). This studyconducted a survey of American voters to try tounderstand, among other things, more aboutthe role of the media (radio and newspapers) inelection campaigns. Repeat interviewing ofrespondents was carried out across the durationof a presidential election campaign in order toassess their exposure to campaign material, aswell as their opinions about candidates andawareness of policies. Respondents’ social cate-gory memberships (sex, age, residence, eco-nomic status and education) emerged asimportant variables that influenced their degreeof exposure to mass communication and theirpolitical candidate preferences. Further researchby Katz and Lazarsfeld (1955) utilised surveymethodology to establish the importance ofinformal social networks for public opinionformation. A specific communication process,labelled the ‘two-step flow of communication’,posited that the media have an indirect effectupon public opinion, which operates through‘opinion leaders’.�


relations ofassociation

and after thefact

self-reportinformation

the census

populationand sample

� two-step flow – Chapter 10, p. 159

Surveys may be differentiated in terms oftheir purpose; their administration, includingsampling; and their time span.

Purpose of a survey

Surveys may be broadly divided into descrip-tive and analytical exercises:

• A descriptive survey simply attempts to doc-ument current conditions or states of affairs.Public opinion polls, for instance, can provideinformation about people’s present attitudes on a specified topic. Historically, descriptivesurveys can be traced back to the censuses,whose purpose was to define general charac-teristics of entire populations.

• Analytical surveys also collect descriptivedata, but attempt to go on to examine rela-tionships among variables in order to testresearch hypotheses. Accordingly, a survey mayassess the impact of an advertising campaign onpublic awareness of a brand and changes in themarket share of a product. Such explanatorysurveys have also played a prominent part inresearch into the social effects of the media(e.g., the impact of media violence).

Forms of administration

Surveys collect data through either question-naires or interview schedules (see Figure 13.2).Respondents may complete questionnaires bythemselves, or answer questions that are put to them by an interviewer. Self-completionquestionnaires are often posted to respondentswho complete this ‘instrument’ at home in their own time and then mail it back to theresearcher. Such questionnaires must be self-explanatory, because respondents are not guided through the data collection procedureby a researcher in person. Questionnaires may also be administered to groups of mul-tiple respondents simultaneously in a theatre or classroom situation. On such occasions,researchers can be on hand to assist respon-dents with questionnaire completion.

In addition, respondents may be interviewedorally, and here they are led through the ques-

tions by an interviewer. When this happens,respondents rarely see the complete question-naire, whether the interviews are conducted bytelephone or face-to-face. The latter form ofadministration may take place in respondents’own homes, in the researcher’s office or in thestreet. Telephone interviews have the advantagethat they can accomplish the data collectionvery quickly, and that they are relatively cheap.Interviews by telephone can also be conductedwith respondents who, for geographicalreasons, may be difficult to reach at home.However, such interviews must be kept fairlyshort, and cannot use questions where respon-dents need to be shown something visually.Personal, face-to-face interviews representperhaps the most efficient form of surveyadministration. This is partly because longerinterviews of up to an hour are possible in thehome, although interviews in streets or shop-ping malls may be even shorter than telephoneinterviews. In addition to dealing with anymeaning difficulties in questions, interviewerscan use visual prompts, and with computer-assisted techniques, they can complete bothdata collection and analysis rapidly.

The issue of sampling

It is essential that the individuals in a surveyshould be representative of the total populationfrom which they are drawn, if researchers wish to generalize their findings to the popula-tion as a whole. A key aspect of quantitativeresearch, therefore, is sampling. Samples maybe constructed either on a probability or non-probability basis. A probability sample isselected according to mathematical guidelineswhere-by the chance for the selection of eachunit is known; a non-probability sample doesnot follow such guidelines. The advantage ofprobability sampling is that it allows research-ers to calculate the amount of sampling errorin a study. This means that researchers candetermine the degree to which a sample is dif-ferent from the population as a whole in termsof specific characteristics, when the distributionof those characteristics for the general popula-tion is already known.

Searching for media–effect associations: survey research 215

1111234567891012341115678920123456789301234567894012345678950111

questionnaires

telephoneinterviews

face-to-faceinterviews

Non-probability samplingThis type of sampling is often used in mediaresearch. People are selected for study on thegrounds that they are available, convenient toaccess and prepared to participate. Conveni-ence samples may comprise college studentsenrolled in a researcher’s own courses, orpeople intercepted in the street. Volunteersamples may be obtained by advertising forparticipants on notice-boards or in newspapers.In each of these cases, the researcher has littlecontrol over who comes forward to take partin the study. Consequently, such samples arelikely to be biased in their demography andpsychological characteristics as compared tothe population in general.

More systematic forms of non-probabilitysampling are, however, available. While thesestill do not meet the mathematical requirementsof probability sampling, they may neverthelessdeliver more robust samples:

• A purposive sample – often used in adver-tising research – is taken when respondents areselected according to a specific criterion, suchas their purchase of a particular product.

• A quota sample is another selection pro-cedure whereby participants are chosen tomatch a pre-determined percentage distributionfor the general population. If, for example, thedistribution of males and females is 49 per centand 51 per cent in the total population, respon-


Mail questionnaires

Advantages

Cheap to runCan reach wide geographical areaRespondents complete questionnaire at own paceOffer anonymityAvoid interviewer bias

Telephone interviews

Advantages

Relatively cheap to runGenerate higher response rate than mail questionnairesResearcher can control order in which questions are answeredCan provide rapid data collection and processing if computer-assisted techniques are used

Face-to-face interviews

Advantages

Have the best response ratePermit the longest interviewsVisual prompts can be usedInterviewers control the way questions are answeredInterviewers can probe for more detailed responses

Disadvantages

Questionnaires not always returnedMay suffer delays in responsesResponses higher for some social categories than for othersNo control over how respondents complete questionnairesNo one available to clarify questions if parts of questionnaire are not understood

Disadvantages

Interviews must be kept shortCan only reach respondents with telephonesNo visual prompts possibleOpen-ended questions are difficult to use

Disadvantages

Very expensive to runInterviewers may have problems reaching certain locations (e.g., remote areas, unsafe areas)Interviewer bias can be a problem

Figure 13.2 Advantages and disadvantages of different forms of survey administration

convenienceand volunteer

samples

dents are selected for a survey until this distri-bution is matched in the sample.

(A further technique is haphazard sampling,whereby participants are selected on the basisof appearance or mere practical convenience.This approach relies on subjective judgementsby researchers, rather than any clearly devisedsystem of selection. In some cases, researchersmay attempt to recruit large numbers in orderto compensate for the lack of a selection sys-tem, on the mistaken assumption that biggermeans better. In representative sampling, it doesnot.)

Probability samplingThe techniques of probability sampling includerandom sampling, systematic random sam-pling, stratified random sampling and clustersampling:

• Random sampling is the most basic form ofprobability sampling. Under this scheme, everyindividual or unit in a population has an equalchance of being selected. For this purpose,researchers may use a table of random numbersor a computer-based system that taps into elec-tronic databases comprising census data or tele-phone numbers.

• With systematic random sampling, a criter-ion is fixed to select every nth person or unitfrom a population. To exemplify, a decisionmight be taken to select one in ten members ofa population totalling 1000. A random startingpoint is then chosen, and from there, every tenthmember is selected for a total sample of 100.

• With stratified random sampling, furtherrestrictions are placed upon the selection pro-cedure, although the fundamental element ofrandomness is retained. If, for example, the aimis to exactly match the sample’s demographicdistribution with that of the population as awhole in terms of gender, age and socio-economic levels, this aim can be built into thesampling frame. Hence, if 51 per cent of thepopulation is female, random selection offemales to the sample will cease once that targethas been reached, namely 510 females of atarget sample of 1000. Sampling may also be

stratified, or disproportionate, when studiesselect certain demographic groups in larger pro-portions than their population distributionsbecause the end-users of the research have aspecial interest in particular population sub-groups. This is typically the case with advertis-ers whose products are aimed at particulartarget markets.

• Cluster sampling involves a special case ofstratification. A population may be divided interms of its geographical distribution betweendifferent regions, districts and postal codes. The random sampling process is accordinglyconducted in a progressive and hierarchicalfashion. First, regions are randomly selected,next districts are randomly selected from withinregions, then postal code districts are selectedfrom within larger districts, and finally indi-viduals are randomly selected from withinpostal codes. The weakness of this procedurelies in the fact that the postal code districtswhich are thus selected may represent particu-lar kinds of neighbourhood, as defined by theage or class of their residents. Since other neigh-bourhoods that were not selected might repre-sent different types of people, the final samplecould be demographically distorted.

Time span

Surveys can be distinguished into cross-sectional or one-off studies that obtain, forinstance, opinions at one point in time, andlongitudinal or repeat studies which can beconducted with the same or different groups ofpeople over time. With attitudes, beliefs andperceptions that are prone to change, repeatsurveys are largely more informative than one-off surveys. It should be noted, however,that regardless of the survey schedule that isfollowed, all surveying involves the collectionof self-report data from respondents in whichthey provide verbal accounts of their opinionsor behaviour at specific times.

Cross-sectional surveysThis first type of survey attempts to establishan aspect of public opinion or behaviour at thetime when the study is conducted. A sample oftelevision viewers may be questioned about

Searching for media–effect associations: survey research 217

1111234567891012341115678920123456789301234567894012345678950111

synchronicstudies

their current viewing patterns or their opinionsabout the standards of programmes. Duringpolitical election campaigns, surveys are con-ducted to find out who respondents would votefor if the day of interview was polling day. Across-sectional survey may also be used toinvestigate correlations between the extent andtype of media use that is claimed by differentsegments of the public and their knowledge oropinions about issues.

Cross-sectional surveys have been used,among other things, to investigate the effects ofmass media. In the context of the media violencedebate, for example, respondents have beenasked to identify or recall details about theirtelevision viewing and their aggressive disposi-tions. They may be given lists of programmetitles, asked to report the programmes they likewatching best of all, or complete viewing diariesto provide the researcher with some indicationof how much they watch and what they watch(e.g., Greenberg 1975; Hartnagel et al. 1975;McCarthy et al. 1975; McLeod et al. 1972;Robinson and Bachman 1972). Within thisapproach, assumptions are made about thecontents of named television programmes, butrarely are these assumptions tested by analysingthe programmes themselves. Instead, it is takenas axiomatic that action-adventure or crime-drama programmes contain violence. Hence, ifa respondent nominates such a programmeamong his or her favourites, that is taken as evidence of exposure to televised violence. In addition, respondents’ personal aggressiontendencies have been assessed through self-report measures and sometimes also throughreports from other people (e.g., parents, teach-ers, friends, peers). However, such correlationalsurveys do not measure actual behaviour; theymerely examine degrees of statistical associationbetween verbally described behaviour.

Cross-sectional surveys have not beenrestricted to samples of the general public, evenif these are by far the most frequent type.Surveys have been conducted among mediaprofessionals� to obtain data, for instance, onjournalists’ working practices, job satisfactionamong people employed in media industries,

and their opinions concerning the impact ofnew communication technologies on the futureof their businesses (Bergen and Weaver 1988;Demers and Wackman 1988; Ross, 1998).

Longitudinal researchLongitudinal surveys are an efficient procedurefor examining long-term relationships betweenselected variables because they permit the col-lection of responses over time. One particularstrength of longitudinal methodology is that itenables researchers to examine the plausibilityof different types of causal hypotheses. First,researchers can begin to untangle the poten-tial bidirectional causal relationships betweenmedia and audience attitudes or behaviour. Inother words, exposure to media violence mayincrease the likelihood of aggressive behaviour,but an aggressive predisposition might alsocause individuals to favour watching pro-grammes containing violence (Huesmann et al.1984; Lefkowitz et al. 1972; Milavsky et al.1982). Second, research using longitudinalmethodology can determine whether exposureto media is associated with long-term changesin audiences’ attitudes and behaviour.

Three types of longitudinal research may bedifferentiated:

1 Trend studies. A given population may besampled and studied at different points in time,so that different respondents are questioned ineach survey, but each sample is drawn from thesame population. This type of study is oftenused during election campaigns (e.g., Gunter et al. 1986). Here, samples of respondents aresurveyed about their voting intentions before,during and at the end of a political campaign.Furthermore, trend studies can be conductedusing data from secondary sources. For in-stance, researchers have conducted historicalanalyses of the relationship over time betweenthe penetration of television sets in a popula-tion and its crime rates, using existing statisti-cal data (e.g., Centerwall 1989; Hennigan et al.1982).

2 Cohort studies. A cohort study focuses onthe same specific subset of a population eachtime data are collected, although the samplesmay be different. Normally the individuals


� surveys with media professionals – Chapter 5, p. 89

diachronicstudies

secondaryanalysis ofstatistical data

participating are linked in some way, perhapsby having the same birth date (a birth cohort),or because they have experienced the same sig-nificant life event. For example, a study mightsurvey all children aged 5 to 6 in a communitybefore the introduction of television. Then, twoyears later, after television transmission hasbegun, another survey may be conducted withthose same children at age 7 to 8.

Cohort analysis is a technique that is espe-cially suited to monitor changes in attitudes andbehaviour which are associated with matura-tion. However, since exactly the same peoplemay not be surveyed on each occasion (butdifferent samples from the original cohort), anychanges may be attributable to unidentifieddifferences between the actual respondentgroups, in addition to age and maturation differences.

To exemplify, Rentz et al. (1983) conducteda cohort analysis of consumers born in fourtime periods: 1931 to 1940, 1941 to 1950,1951 to 1960 and 1961 to 1970. Soft drinkconsumption was measured in all the samplesthat were taken from these cohorts at inter-vals, and a range of potential predictors of thisconsumption assessed. The results indicated alarge cohort effect, in the sense that the level ofsoft drink consumption which had been estab-lished early in life in each cohort tended toremain stable later in life, compared with theother cohorts. Rosengren and Windahl (1989)also used cohort analysis as part of their in-depth longitudinal study of television usage bySwedish youngsters. Among other things, theyfound a slight similar cohort effect, but con-cluded that age was the prime determinant ofhabitual television viewing.

3 Panel studies. Trend and cohort studiespermit the analysis of process and change overtime, which is rarely possible in a cross-sec-tional survey. Yet, a limitation of these two typesof longitudinal study is that, on each occasion,different people are surveyed. As a result, it isnot possible to track changes in attitudes orbehaviour over time for specific individuals. Incomparison, panel studies involve the collectionof data over time from the same sample ofrespondents – this sample is called a ‘panel.’ For

example, in a study to test for the effects of tele-vised violence on viewers’ aggression, repeatedsurveys were carried out with the same indi-viduals at intervals ranging from one to tenyears in order to assess whether an earlier dietof violent programmes was associated not onlywith aggressiveness at that time, but withaggressive tendencies in later life (Milavsky et al. 1982).

Panel studies, however, also have their diffi-culties and limitations. For one thing, they needto be based on original data collection from thespecific panel, whereas trend or cohort studiesmay be conducted through secondary analysisof previously collected data. For another, aspecial problem is the loss of panel membersover time. People interviewed in the first surveywave may be unavailable for, or unwilling toparticipate in, the second or third waves. Inother cases, people move house, die or becomeuntraceable. Consequently, it is quite commonthat such ‘attrition’ causes the panel to gradu-ally diminish in size as the study progresses.

Survey studies, whether cross-sectional orlongitudinal, can only demonstrate correla-tional links between variables. Their reliance onself-report or other-report measures is fraughtwith potential inaccuracies. These may arisefrom the respondents’ memory failure, ill judge-ment or inadequate knowledge, but as import-antly from a form of questioning that providesnon-valid verbalised representations of theaspects of everyday reality under study.Through experimental methodologies, quanti-tative researchers have examined relationshipsbetween media and audience variables moredirectly, as a later section explains in detail.

SURVEYING MEDIA OUTPUT

Survey principles may also be applied to theanalysis of media content. This quantitativeassessment of media output – content analysis�

– may be traced back to the 1940s whenwartime intelligence units monitored radiobroadcasts for their music and news content as

Surveying media output 219

1111234567891012341115678920123456789301234567894012345678950111

� content analysis – example analysis, Chapter 7, p. 101

indicators of the morale and movements of theenemy (Wimmer and Dominick 1994). Contentanalysis was soon taken up by social scientiststo monitor more general social and economictrends. As early as 1910, Max Weber hadsuggested launching a study to monitor presscoverage of political and social issues alongsidesurveys of public opinion, thus anticipatingagenda-setting research (see Beniger 1978).�In the second half of the twentieth century, the methodology was increasingly applied to awide range of media issues. Prominent applica-tions have examined patterns of news coveragein order to ascertain the agenda-setting role ofmedia (McCombs and Shaw 1972) and patternsin the representation of social groups andevents in order to assess cultivation effects ofmedia upon public perceptions of social reality(Gerbner 1972).

An early definition of content analysis con-ceived it as ‘a research technique for theobjective, systematic, and quantitative descrip-tion of the manifest content of communi-cation’ (Berelson 1952: 18). Krippendorf(1980) defined it as a research technique formaking replicable and valid references fromdata to their context. Kerlinger (1986) sug-gested that content analysis is a method ofstudying and analysing communication in a sys-tematic, objective and quantifiable manner forthe purpose of measuring variables. This lastdefinition particularly encapsulates the definingingredients of any traditional form of quanti-tative analysis of media output. Content analy-sis is, first, systematic in that it utilises aprincipled form of media output sampling andcontent coding. Second, it is objective in thatthe researcher’s own idiosyncracies and biasesshould not affect the analysis. Operationaldefinitions and rules for the classification ofvariables should be explicit, so that otherresearchers might repeat the procedure. Finally,content analysis is quantifiable in that its mainfocus is on counting occurrences of predefinedentities in a media text. On this last point,purely quantitative forms of content analysishave been challenged for displaying a lack ofsensitivity to hidden meanings that may be

conveyed by media texts (see Merten 1996).Thus, counting and quantifying may need to be supplemented by interpretive procedureswhich can clarify the weight and implicationsof singular media messages in terms of theirpotential impact upon the audience (Gunter1985a; Hodge and Tripp 1986; Potter andSmith 1999).

Five main purposes of content analysis havebeen identified (Wimmer and Dominick 1994):

1 Describing patterns or trends in media portrayals

2 Testing hypotheses about the policies oraims of media producers

3 Comparing media content with real-worldindicators

4 Assessing the representation of certaingroups in society

5 Drawing inferences about media effects.

In each case, studies must return to and departfrom the basics. A quantitative content analysisis designed to provide a descriptive account ofwhat a media text (film, TV programme, adver-tisement, newspaper report, etc.) contains, andto do so in a form that can be repeated byothers. In putting together a content analysis,then, the researcher must work through anumber of stages of measuring and sampling.

Measuring media content

Having decided upon the general topic of inves-tigation, it is necessary first to define the ‘things’to be measured. Here, the basic concept is the‘unit of analysis’ – the textual element that isto be counted. In addition to the unit of analy-sis as a whole, there may be features or attrib-utes of that unit about which data are alsocollected. The data collection process as suchproceeds by relying on a ‘coding frame’. Thisis a form on which the occurrences of the differ-ent categories relating to the unit of analysis,its features and attributes, can be numericallycatalogued.

To illustrate: a content analysis of genderrepresentation in television advertising wouldfocus on appearances by males and females.The unit of analysis in this context would be


� agenda-setting research – Chapter 9, p. 145

unit ofanalysis

coding frame

an appearance by a male or female character,either on screen or as a voice-over. Further ana-lytical categories may then be deployed todescribe in more detail the nature of theseappearances. For instance, the researcher mightbe interested in establishing not only how oftenmen and women appear at all in televisionadvertisements, but also whether they appearwith differing frequencies in different socialroles or in connection with different producttypes (Furnham and Schofield 1986; Furnhamand Skae 1997; Furnham et al. 1997).

Some studies have identified theoreticallyrelevant patterns simply by analysing mediacontent in terms of its major themes. In aninvestigation of the extent to which filmsreleased in Britain between 1945 and 1991 werecharacterised by crime themes, Allen et al.(1997) assigned films to one of ten genre cat-egories (western, crime, war, romance, fantasy,sex, farce, adventure, drama, other) on the basisof details in the film synopses. The analysis hadtwo parts. First, the presence of crime contentin each synopsis was assessed by examining itfor mentions of a crime, criminals, or the crim-inal justice system. Second, the film was classi-fied according to the ten genre categories. In thecase of crime films, this classification meant thatthe ‘primary focus of the narrative is on thecauses or consequences of illegal activities,central characters include criminals, victims andthose who work in the criminal justice system(e.g., private eyes, amateurs, police, courts,gangsters)’ (Allen et al. 1997: 92).

In the US National Television ViolenceStudy, units of analysis were defined at morethan one level (National Television ViolenceStudy 1997; see also Potter and Smith 1999).Whereas previous television violence contentanalyses had emphasised counting violent actsas such, this study included more global mea-sures of violence that would better represent thecontextual features of violence. Three levels of measurement were devised: the ‘PAT’; thescene; and the programme. A PAT representedan interaction between a perpetrator (P), an act(A) and a target (T). A sequence of PATs, eithercontinuous or separated by brief cutaways or scene changes, might together make up aviolent scene, and such sequences afforded an

opportunity to examine relationships betweendiscrete acts of violence and their meaningswithin the scene. Finally, the researchers arguedthat larger meanings seemed to be conveyed bythe pattern of violence as a whole within a pro-gramme, and that this meaningful pattern couldbe effectively interpreted only when analysedwithin the full context of the programme.

In addition to such incidents or actions,other units of analysis that are often codedinclude the agents in either fictional output ornews. In the case of news, it is particularly ananalysis of the sources of quotes, comments orother material that can yield insights (Lasorsaand Reese 1990). In fictional media content, ananalysis of the attributes of actors or characterscan yield evidence about the proportional rep-resentation of different social groups. Further-more, studies of the presence of different typesof sources in news or other factual output canprovide evidence for assessing the balance, neu-trality, thoroughness and impartiality of report-ing. This form of analysis may examine therange of sources used; which groups, organisa-tions or institutions they represent; and thecontext (interview, official meeting, press con-ference) in which they appear. Such analysismay also examine the kind of informationsought and obtained from different sources,and any indications of the status of the source(Ericson et al. 1991).

Sampling media content

Once the body of content to be considered hasbeen specified with reference to theoreticalpurpose, and the units of measurement havebeen selected, the researcher has to determinehow much of that content to analyse. In someinstances, the universe may be small enough to be analysed in its entirety. Generally, re-searchers must sample a subset of content fromthe total universe, since it is too large to beanalysed in full. In contrast to mass publics,which will have been surveyed in their entiretyin population censuses, thus enabling a con-struction of sampling frames based on knownpopulation parameters, such a point of com-parison is not available for surveys of mediaoutput.

Surveying media output 221

1111234567891012341115678920123456789301234567894012345678950111

‘population’ ofcontents tobe sampled

Partly for this reason, sampling in contentanalysis often takes place in more than onestep. A first step may be to specify whichcontent sources are to be sampled. For exam-ple, in a study of newspaper coverage of currentevents, the first step is to decide which par-ticular national or local newspapers are to beanalysed. Then a decision must be taken abouthow many editions of each newspaper toanalyse, and over what period of time. Afurther step may be to decide how much, orwhich parts of the newspaper to analyse. At thislevel, one has to consider how many ‘stories’ toanalyse and how these should be defined anddelimited. Finally, the analyst will considerwhether there are specific story ingredients thatneed to be measured.

Previous content studies have establishedsome rough guidelines for sampling. Stempel(1952) drew separate samples of six, twelve,eighteen, twenty-four and forty-eight issues of a newspaper and compared the averagecontent of each sample size in a single subjectcategory against the corresponding total for the entire year. He found that each of the five sample sizes was adequate, and thatincreasing the sample beyond twelve issues didnot significantly improve upon the accuracy offindings. In the longstanding content researchinto television violence, some studies haverestricted programme samples to a single week(e.g., Gerbner et al. 1977), while others haveopted for samples of up to four weeks. In somecases, chronological weeks are sampled, whileothers have compiled composite weeks byselecting one day from each of seven differentweeks (Gunter and Harrison 1998; Gunter et al. 1996).

The issue of content sampling was examinedvery closely by the National Television ViolenceStudy (1997). This study analysed a far largersample of programme output from a largernumber of television channels than any previ-ous American content study, and the totalproject sample, unusually, covered program-ming broadcast throughout the day. What dis-tinguished this research most of all, however,was its use of a random sampling frame toselect programmes over a period of twentyweeks each year. The programmes were chosen

with a modified version of random sampling,as described above. Two half-hour time slots(defined by hour of day and day of week) wererandomly selected for each channel during eachweek that sampling occurred. Once a time slothad been selected, the TV Guide was consulted,and the programme corresponding to that timeslot was entered into a scheduling grid severaldays before the actual broadcast in the targetweek, so that recording and coding could beprepared. This procedure was repeated until afull composite week of programmes for eachchannel had been compiled for analysis.

Limits to quantitative content analysis

Quantitative content analyses tend to be purelydescriptive accounts of the characteristics ofmedia output, and often make few inferencesin advance about the potential significance oftheir findings for what they may reveal aboutproduction ideologies or impact on audiences.In order to support conclusions about mediaprocesses and effects, theoretically informeddecisions have to be made about which aspectsof media content to analyse and classify. Themost informative content analyses, therefore,will be produced by analysts who choose theircontent categories with reference to an explicittheoretical framework. Purely descriptive, a-theoretical applications of this technique mayyield reliable indicators of the manifest contentof the media, but will contribute in only alimited way to a better understanding, either ofthe forces which lie behind that content, or ofits eventual impact upon audiences.

TESTING CAUSALITY DIRECTLY:EXPERIMENTAL RESEARCH

Like surveys, experiments have been used inmedia research for more than fifty years.Experimental research usually involves a quan-tification of the effects of media upon their audi-ences, although experimental methods have alsobeen used to investigate the way people usemedia. In the 1940s and 1950s, experimentalmethodologies were employed particularly toinvestigate the impact of media messages onopinions about the enemy in wartime (Hovland


chronologicaland compositesample weeks

et al. 1953), and to study the effects of produc-tion techniques on learning from informationalmedia (Belson 1967; Tannenbaum 1954;Tannenbaum and Kernick, 1954; Trenaman1967). In the 1960s, experiments came to beused, not least, to study the effects of media vio-lence (Bandura and Walters 1963; Berkowitz1964; Feshbach 1961). Useful sources of infor-mation about experimental practice and designsare Bailey (1994) and Neuman (1994).

An experiment generally begins with ahypothesis about a likely outcome following anevent, or set of events, that can be controlledor manipulated by the researcher. In mediaresearch, an experiment will typically create aset of conditions under which an individual, orgroup of individuals, are exposed to a mediastimulus, and are then invited to respond insome way. The conditions are constructed insuch a way that the media may be said to havecaused a particular kind of audience response.For instance, if one wishes to test a hypothesisthat a media depiction of violence may have aneffect upon the behaviour of an audience, aminimum of two situations will be created inwhich different audience groups watch either aviolent or a non-violent portrayal. Subse-quently, both groups will be placed in anotherkind of situation in which there is an opportu-nity to behave aggressively. The researchhypothesis might predict that those who havebeen exposed to media violence will behavemore aggressively than those who have beenshown non-violent material. Quantitative mea-surements would be taken to establish whetherthat prediction is borne out by the actual,observed behaviour of the two groups.

Experiments tend to investigate smallernumbers of respondents than surveys, becausethey operate on a different logic. Participantsin an experiment are allocated to either exper-imental or control groups. The former areexposed to the manipulated independent vari-able(s), while the latter are not. However, par-ticipant samples may be non-representative.One reason for this is that much experimentalresearch on media derives from psychologicalresearch traditions which have conceived ofmany of the psychological processes of interestas constants across individuals. Non-represen-

tative samples are compensated for, in part, by the ‘random assignment’ of participants toeither experimental treatments or control con-ditions – a key concept in experimentation.Thus, even if the participants do not representthe wider population, within the confines of theexperiment, the random assignment of partici-pants to the two sets of conditions controlsagainst biases in the findings. This procedurealso aids replication. The conditions and stepsof an experimental study are so minutelyspelled out that another investigator couldreadily repeat the study to find out if the orig-inal results stand up.

The main advantage of experimentalresearch is that it enables research to test forevidence of direct cause–effect relationshipsbetween variables. Hence, if one variable is pre-sented at a particular point in time, there willbe a measurable impact upon a second variableobserved at a later point. In addition, experi-mental research allows the investigator to exer-cise control over some or all of these variables.In a study of the impact of media on an audi-ence, the researcher can determine the contentto which individuals are exposed, the contextin which the exposure occurs, and the ways inwhich they are asked to respond subsequently.

One weakness of experiments is that theytend to be carried out in artificial conditions. Alaboratory environment is quite different fromthe everyday, social environment. In mediaresearch, the need to control for the inherentcomplexities of media content when examiningthe effect of one of its aspects may result bothin the respondents’ consumption of mediacontent under highly artificial conditions, andin a selection of media content extracts that failto reproduce their normal media experience. Aninvestigation of the impact of televised violenceon viewers in which violent extracts of a fewminutes’ duration are shown to viewers thusremoves the violence from its original pro-gramme context, which under ordinary viewingconditions might influence how viewersrespond to violence.

A second weakness is that participants may‘second-guess’ what the study is about, andwhat the experimenter expects to happen. Byresponding accordingly, they may give the

Testing causality directly: experimental research 223

1111234567891012341115678920123456789301234567894012345678950111

experimentaland control

groups

randomassignment of subjects togroups

relations of directcausality

experimenter what he or she wants. Such‘demand characteristics’ can bias the results ofan experiment, because participants no longerbehave the way they might otherwise have doneunder the specified conditions. A related sourceof bias is ‘experimenter bias’. In this case, theexperimenter may unwittingly give away cluesabout the hypothesis by giving strongerencouragement to participants to behave in oneway than in another. To counteract such prob-lems of bias during the interaction, a double-blind technique may be used,� in which neitherthe person running the experimental sessionnor the participants know whether a given par-ticipant is in an experimental group or a controlgroup.

Experimental designs

Experimental designs differ principally in termsof the number of stages and conditions theycomprise. Some experiments employ testsbefore and after a manipulation of an inde-pendent variable; others use tests following themanipulation only. In some experiments, onlyone group of participants is studied, while inothers, two or more groups may be studied.

Classic experimental designThis design is also known as a ‘pre-test–post-test with control group’ design. Such experi-ments use at least two groups, of which one isa control group. Participants are randomly allo-cated to these groups, and then tested prior to,and following, an experimental manipulation.

Pre-experimental designsA classic experimental design is not alwaysattainable if too few participants or resourcesgenerally are available. Researchers may thensettle for lesser designs with fewer controls. Aone group, post-test only design uses just onegroup and no pre-test. For example, a group ofviewers may be shown a television news pro-gramme after which their knowledge of thenews stories in question is tested. In this case,without a pre-test, it is impossible to determinehow much knowledge they already held about

these stories before seeing the programme, evenif the findings of such a design may reflect onthe processing of information from media or ontheir forms of presentation.

A slightly more advanced design is the one-group, pre-test–post-test design with only onegroup of participants, but including a pre-testas well as a post-test. Extending the examplegiven above, participants would here be testedfor their knowledge of relevant issues before thenews programme, and then for any changes tothat knowledge after viewing the programme.In the absence of a control group, who wouldbe ‘pre-tested’ and ‘post-tested’ without seeingthe news programme, it is again difficult toinfer that any change in the experimentalgroup’s knowledge resulted from exposure tothe programme. One alternative explanationmight be that the pre-test itself encouraged par-ticipants to rehearse their knowledge of thenews, so that by the second test they were ableto perform better. An additional explanation ofany knowledge change over time, if there is agap of hours or days between the two tests,could be participants’ exposure to other newsmedia that also contained information aboutthe stories in question.

A third version of a pre-experimental designis a static group comparison, in which a post-test only is employed, but applied to twogroups. In this case, one group might be showna news programme while another group wouldnot. Any differences between the two groups intheir post-test knowledge scores, however,might be due to pre-existing knowledge differ-ences between them, and should not simply be explained in terms of exposure to a newsprogramme.

Quasi-experimental designsWhile quasi-experimental designs do not reachthe control standards of the classical design,because they do not include a pre-test stage,unlike pre-experimental designs, they do atleast employ a control group. With the post-test only with control group design, partici-pants are randomly allocated to experimentaland control groups, but are tested only after theexperimental manipulation has been imple-mented. Although random assignment reduces


� double-blind techniques – Chapter 16, p. 290

pre-test–post-test withcontrol group

the chances that the groups will differ beforetreatment, a researcher cannot be sure withouta pre-test.

It should be added that, even with a pre-test,as in the classical design, measurement issuescan arise. Thus, a pre-test may influence theway participants react during and after theexperimental treatment because the initial teststage has given away clues as to the purpose ofthe experiment, or has given participants theopportunity to practise a relevant skill. The‘Solomon four-group design’ offers a solutionto this problem. Here, for some participants inboth the experimental and control groups, pre-tests and post-tests are run, while for others,only post-tests are used. The aim is to controlfor possible effects of the pre-test as well as ofthe experimental manipulation on the post-testscores. If the groups who were pre-tested differin their post-test performance from those whowere not pre-tested, the researcher can concludethat the pre-test itself had an effect on post-testresults, thus potentially biasing the findingsconcerning the effect of the central experimen-tal manipulation.

Factorial designsThe experimental designs considered so far areall set up to investigate the effect of a singleindependent variable per group, regardless ofwhether control groups and a pre-test wereused or not. In many studies, however, experi-menters are interested in examining the effectsof more than one independent variable upon adesignated dependent variable among the samegroup of individuals. It is possible that two ormore independent variables produce joint anddistinctive effects upon a dependent variable. Itis also possible that their effects are inter-dependent, so that one independent variable hasan influence upon a dependent variable only inthe presence of a second independent variable.

In factorial designs, accordingly, two or moreindependent variables or ‘factors’ are manipu-lated. Factors, in addition, may have two ormore aspects to be considered in the analysis.For example, in an experimental study of theeffects of various attributes of both televisionprogrammes and advertising on viewers’ recallof the advertising, there may be three types of

programming (holiday programme, car pro-gramme, cooking programme) and three typesof advertisement (holidays, cars, cookery prod-ucts). In a factorial design, these attributes,which amount to experimental treatments,would produce nine conditions, as each type ofadvertisement is embedded, in turn, in eachtype of programme. A full experimental studywould thus require nine groups of participants.

Repeated measures designsAs suggested by the previous example, one ofthe main disadvantages of a factorial design isa practical one, since it requires different setsof participants for each experimental condi-tion. Each time a new independent variable isintroduced, the number of cells or groups thatis generated within the design increases, requir-ing an increase also in numbers of participants.One way of resolving this problem is to obtainmeasures concerning more than one indepen-dent variable from the same group of partici-pants. Such a repeated measures design, ifapplied to the factorial advertising experimentdescribed above, could examine the samenumber of variables with just three groups. The groups would be defined by the type ofprogramme they are exposed to, but each pro-gramme would be embedded with all threetypes of advertisement. This solution, of course,raises further questions of whether the presen-tation of the advertisements together, or theirsequence, may affect the audience response.

Experimental contexts: the problem withlaboratory research

To sum up, experiments may be equipped to test causal hypotheses, but are not withoutserious limitations. An important shortcomingof experiments testing for media effects on audi-ence attitudes and behaviour stems from theconditions they create for examining such links(Cook et al. 1983; Stipp and Milavsky 1988).Laboratory conditions do enable researchers toexert control over the behaviour of their partic-ipants as well as over various environmentalfactors which might influence their behaviour in the real world. However, such research maylack external validity; its findings may not begeneralisable beyond the laboratory.

Testing causality directly: experimental research 225

1111234567891012341115678920123456789301234567894012345678950111

Solomon four-group design

issues ofexternalvalidity

In research on behavioural effects, for example, researchers have frequently createdartificial measures of behaviour, especially whenstudying the effects of media violence on audi-ence aggression (Berkowitz 1964; Berkowitzand Geen 1966; Donnerstein and Berkowitz1981). While there are sound ethical reasons forthis, the responses measured in experimentsoften fail to resemble what would be more com-monly seen as ‘aggressive behaviour.’ The labo-ratory creates a social environment all of itsown in which the usual sanctions against be-having in particular ways (e.g., aggressively) aresuspended (Comstock 1998). At the same time,experimenters artificially constrain the res-ponses participants might make. In aggressionexperiments, participants are normally givenonly one behavioural response option – one ofaggression. Yet, in a laboratory experiment thatwas designed to test the effects of exposure toviolent pornography on males’ behaviouralreactions towards a female target, the menunder study showed little inclination to useaggression against this female if a non-aggres-sive response alternative was made available tothem. Even when the female had earlier beeninsulting towards them, they still chose the non-aggressive alternative (Fisher and Grenier1994).

Studies that utilise the repeated measuresdesigns within a laboratory setting face specialproblems which arise from ‘priming’ or condi-tioning effects. An earlier exposure to similarstimulus materials may affect respondents’reactions to materials which are later presentedby providing points of comparison or a frameof reference for judging those later materials.Again, participants may receive clues from theearlier materials as to the experimental hypo-thesis, which may lead them to behave inaccordance with that hypothesis.

Research in naturalistic settingsTo overcome problems of ‘ecological’ validity,experiments may be carried out in more natu-ralistic settings. Here, the participants in exper-iments are observed in surroundings where theymay not be aware of the research going on.Two such categories of real-world experimentmay be distinguished: those in which the

researcher manipulates a set of conditions in anaturalistic environment, and those in whichthe researcher takes advantage of some natu-rally occurring event or change of circum-stances, the effects of which can be measured.The first type of study is commonly referred toas a field experiment, and the second type as anatural experiment (see MacBeth 1998):

1 In field experiments, researchers often studypre-existing groups, but they assign these groupsto different conditions of media exposure. Thegroups may be observed, first, during a baselineperiod to establish their similarity, and againafter the exposure period. An example would,again, be an experiment to study the effects oftelevision programmes on viewers’ aggressive-ness. In a cable television environment in whichthe flow of programmes into people’s homescould be controlled by the supplier, it would bepossible to create two different groups: one thatreceives programmes containing violence, andanother that receives violence-free entertain-ment. Over a number of days or weeks before,during and after this treatment, the viewerscould be monitored by someone in their familyfor any changes of mood or behaviour thatoccurred as a result of this manipulation of theirtelevision diet (see Gorney et al. 1977).

2 In natural experiments, researchers takeadvantage of a naturally occurring change inthe availability of media in order to assess theimpact of this change on the people in thatenvironment. For example, a community withaccess to television could be compared withanother community that has less or no accessto television. If the second community has tele-vision introduced to it for the first time, pre-TVand post-TV observations and tests may beconducted to assess the impact of television onthat community – a type of research that hasbeen ongoing since the early days of themedium (see Charlton 1997; Williams 1986).

The problems faced by field experiments aremainly ethical and practical. Some researchissues would be unethical and socially irrespon-sible to investigate in the field.� For example,


priming effects

� research ethics – Chapter 16, p. 289

studies of the effects of sexually violent porno-graphy upon men’s attitudes towards femalesexuality, their propensity to commit rape them-selves and their sympathy for rapists must beconducted under controlled conditions. The‘effects’ measures cannot take the form of realbehaviour (i.e. rape), but must rely on simula-tions in which changes of attitude or perceptionmay be monitored, and then immediately coun-tered through elaborate debriefing sessions withparticipants.

On the practical front, field experiments canbe difficult to run because they occur in envi-ronments over which the experimenter’s controlis restricted. As a result, it may not be easy tocreate all the conditions that satisfy the require-ments of a classic experimental design or of asound factorial or repeated measures design.Participants may not always agree to makethemselves available for observation in thesame way they would in a laboratory experi-ment. Moreover, researchers may need to gainpermission to make clandestine observations ofparticipants, or to secretly manipulate aspectsof their social environment in an attempt toinstigate some change in their behaviour. Thislast consideration is especially important inresearch involving children.

SURVEYS OR EXPERIMENTS: HOW DOTHEY COMPARE?

This chapter has focused on the two principalforms of quantitative research – surveys andexperiments – which have been predominantlyconcerned with audiences, either by demon-strating cause–effect relationships or by estab-lishing the possibility, through evidence ofassociation, that such relationships might existbetween media and audiences. The two ap-proaches have, however, also been used to exam-ine aspects of media production (see Wimmerand Dominick 1994). As general methodologies,surveys and experiments have their own inher-ent advantages and limitations, to be weighed indesigning concrete empirical studies.

Experiments, in sum, are designed to exam-ine cause–effect relationships between variablesin a direct sense. For this purpose, researchersmanipulate media and audiences under artifi-

cial conditions, or else take systematic meas-urements of phenomena that occur withinnatural environments where a specific event hastaken place to bring about a radical change.There is nearly always an element of artificial-ity about experiments, because researchersmust be able to relate measurable changes inone variable and measurable changes in an-other variable in such a way that a causalconnection can confidently be inferred. Thepossibility that the ‘criterion’ or dependent vari-able was changed by some factor other than the independently manipulated causal variablemust be reduced to a minimum.

The weaknesses of experiments are:

• their use of non-representative samples;• the degree of artificial control over the

environment being studied;• the contrived nature of many of the media

and audience measures that are deployed;• the difficulty of controlling totally for extra-

neous factors that could have affected thecriterion variables.

The results of experimental studies may there-fore lack any validity, in the predictive orexplanatory senses, in the real world beyondthe controlled environment of the study.

Surveys, by comparison, enable researchersto study media and audiences in their naturalenvironments. Relationships between media,communications and audiences are not manip-ulated in any artificial sense, but are observed asthey occurred, unencumbered by experimentalrestrictions. Surveys also tend to involve muchlarger samples of people and media output than do experiments. Furthermore, whereasexperiments tend to be dependent upon con-venience or volunteer samples because they are usually more demanding of participantsthan other research methods, surveys are able todraw far larger samples that are representativeof the general populations from which they aredrawn in terms of important social and psycho-logical characteristics. This means that theirresults can be more readily generalised to thewider populations from which samples weredrawn.

Surveys or experiments: how do they compare? 227

1111234567891012341115678920123456789301234567894012345678950111

The weaknesses of surveys are:

• their dependence on post hoc, self-reports ofphenomena, which may suffer from inaccu-racies of detail;

• their use of verbal measures of observableevents that may similarly fail fully to repre-sent what actually occurred;

• their reporting only of degrees of associa-tion, or correlation, between variables,which cannot on their own conclusivelydemonstrate causality.

To exemplify, survey respondents’ verbalreports of how much time they normally spendwatching television in hours per day may becharacterised by a significant degree of error.Another means of surveying television viewing,used widely by broadcasting systems around the world, is the TV meter. A metering systemusually has two components: one automaticallyregisters when the TV set is switched on andwhich channel it is tuned to, and the otherrequires viewers to report their viewing, either ina paper diary or, more often these days, via aremote control handset. While technologicallysophisticated, such measurement methods infact merely indicate the presence of one or more persons in the room in which a TV set hasbeen switched on (and only if the respondentsremember to check in), and not whether they areactually watching the screen. Methodologicaltests have shown that when compared againstdirect observations, self-report measures areinaccurate to a degree which may render themuseless for anything other than a very broadindication of whether someone is a relativelyheavy or light viewer (Anderson and Burns1991; Bechtel et al. 1972; Gunter et al. 1995).

In media effects studies, surveys are limitedin the degree of detail with which they are ableto measure a person’s media consumption. Evenif measures relying on respondents to recall, forinstance, programmes they watched on televi-sion in the past week, or newspapers or maga-zines they read over the past month, arereasonably accurate per se, it may be necessaryto know much more about the nature of thespecific content of these media in order to assessits effects on audiences’ knowledge, beliefs, atti-tudes or behaviours.

In one particular realm of research, whereboth surveys and experiments have been promi-nent, their appropriateness must be weighedwith considerations other than the purity ofmeasurement. Studies of children’s fright reac-tions to films and television programmes havevariously measured young viewers’ responses tohorror and other suspenseful content in labo-ratory settings; have questioned young respon-dents about their memories of scary movies;and have interviewed parents about theirobservations of their children’s reactions tofrightening films and television (Cantor 1994).Although survey interviews can yield interest-ing insights, they are, again, dependent uponchildren’s or parents’ recollections. Experi-mental methods can, with greater precision,explore whether specific kinds of portrayalcause children to become scared, and they canshed light on how such reactions change withmaturity. Experimental studies have shown, forinstance, that children under 8 years old arefrightened by scary monsters seen on screen,whereas older children become more anxiousabout unseen dangers lurking off screen (Sparks and Cantor 1986). Whether an on-screen character is attractive or ugly, andwhether it behaves cruelly or kindly, can haveindependent and interdependent effects uponhow children react (Hoffner and Cantor 1985).These studies have only been able to demon-strate the significance of these mixtures of char-acteristics by experimentally manipulating astory to create different versions of the samecharacter.

There are occasions, however, when re-searchers must temper their enthusiasm forusing experiments and limit themselves tosurvey interviews, not least with children. Eventhough the evidence will be less powerful in anexplanatory sense, the experimental manipula-tion of fear in children may be deemed uneth-ical. The creation of genuinely adverse reactionsamong children, in response to a speciallydesigned or selected horror scene, might add tothe wealth of scientific knowledge, but couldpossibly cause undue harm that may not beeasily undone. Accordingly, where there is apossibility that fright reactions may be detri-mental to a child’s development, such responses


TV metersystems

studies ofchildren

rightly tend to be studied after the fact throughsurvey interviews.

Quantitative methodologies, like otherresearch approaches, have strengths and weak-nesses that must be taken into account byresearchers, both before deciding upon their useand during their implementation. While thereare those who would question the validity ofany quantitative approach on epistemologicalgrounds, such techniques have a useful contri-bution to make, particularly for the under-standing of media output, its consumption and

effects, provided that the interpretation of datanever loses sight of their characteristic limita-tions. In this respect they are no different fromthe qualitative methodologies preferred byother epistemologies.

HANDLING QUANTITATIVE DATA

Quantitative research methodologies generatenumerical data. Surveys (whether of audiencesor content) and experiments are the basic‘methods’ of the data collection, but they enter

Handling quantitative data 229

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 13.1 SURVEYS AND EXPERIMENTS COMPARED: THE CASE OFCULTIVATION RESEARCH

A prominent example of the difference between surveys and experiments is research into thecultivation effects of television.Through survey evidence, this research has found correlational linksbetween the amount of viewing that is claimed by respondents and certain patterns in their socialbeliefs, perceptions and levels of anxiety (Gerbner et al. 1977, 1979, 1986). Such global measuresof television viewing may lack the necessary sensitivity to the significant variations in the messagecontent of television.

More detailed measures of viewing habits, using diaries, have indicated that certain socialperceptions may be sensitive to influences from particular types of programmes, but not fromothers (Gunter 1987a; Wober and Gunter 1988). An analysis of British viewers found no linkbetween their perception of personal victimisation in their local neighbourhood and any aspectof their television viewing, while a corresponding analysis of viewers in Los Angeles found thesame perception to be associated with their reported viewing of US-produced crime drama shows(Gunter 1987a).This suggests, among other things, that if the information of certain programmesis seen by viewers to have a direct relevance to their immediate social context, it may affect theirparticular perceptions of that context. However, surveys can explore this link only in a very generalfashion.

In comparison, the application of experimental methodology to cultivation research has madeit possible to explore such links in greater detail. Just as certain general features of a televisionseries, such as its cultural setting, may render its content especially pertinent to how some viewersform judgements about certain aspects of their own or other societies, such effects may also beinfluenced, at a more detailed level, by how, for instance, conflicts are resolved in the series.Experiments have shown that the same television drama can have different effects upon viewers’perceptions of crime and their associated anxiety reactions if the ending is manipulated, so that,in the version shown to one group, criminals are brought to justice, while in another version theyare not (Bryant et al. 1981). Further, reality programmes that depict crime on television can have a more powerful impact on viewers’ perceptions of crime than fictional depictions (Tamboriniet al. 1984).While surveys may also reveal such differential degrees of association between socialperceptions and the exposure to particular types of television content (Gunter and Wober 1982),they are less appropriate for establishing whether viewers were especially attentive to certainmessages within programmes. If, instead, an experimental methodology is chosen, programmes canbe edited to include or exclude specific ingredients, so that differential audience reactions may besystematically measured in post-viewing tests.

into theoretically informed ‘methodologies’ ofanalysis and interpretation.� Once numericaldata have been collected, they need to beanalysed through statistical techniques. Thesemathematical techniques are used to describe,organise, as well as explore relationships withinthe data. In epistemological terms, quantitativeresearch is typically grounded in a hypothetico-deductive approach, in which investigatorsmount hypotheses (or predictions) about theexpected associations or cause–effect relation-ships between variables. The aim of quantita-tive data collection and analysis, then, is toproduce findings which lead to the acceptanceor rejection of a specified hypothesis. Numeri-cal data analysis through statistical procedures,as reviewed in this section, represents a sys-tematic and objective way of determiningwhether significant patterns of relationshipsexist among those phenomena that have beenmeasured in data collection.

Describing data

Data collected via either survey questionnaires,content coding frames or experimental instru-ments are coded numerically, extracted from(what are still mostly) paper formats, andentered into a computerised database, uponwhich various forms of statistical tests maythen be performed. Both the accuracy withwhich this data transfer process takes place,and the application of statistical proceduresthat are appropriate for the particular type ofdata, are crucial to the entire quantitativeresearch project – errors in these early stagescan invalidate the final results.

Often, quantitative data analysis begins byadopting a simple descriptive approach in orderto establish some initial patterns in the findings.A survey of public opinion about the compe-tence of, for instance, national political leadersmight first present the percentages of respon-dents who agreed, and respectively disagreed,that particular political figures were performingwell. A further computation might produce thepercentages who agreed with such sentiments,as broken down by the gender, age, social class

and political affiliations of respondents. Suchresults may be presented for visual display in abar chart or summarised in a table.�

A different type of study might ask a surveysample of 1000 respondents to state how manyhours of television they watch each week. Here,descriptive statistics might be applied to showhow many respondents viewed nothing, lessthan one hour a week, between one and twohours, two to four hours, or more than fourhours. Next, a ‘frequency distribution’ could begenerated, which shows how the respondentswere distributed across these different volumesof viewing. Such data may be visually repre-sented in a line graph or bar chart.

Central tendency and varianceData may be further analysed in terms ofsummary statistics, which render large amountsof data more manageable. Summary statisticsmeasure two basic aspects of the distribution of‘scores’ or measurements in a dataset: centraltendency and dispersion, or variability. Acentral tendency measure indicates which outof a range of scores is the typical one. Thistypical score, in turn, may be defined in threedifferent ways:

1 The mode is the most frequently occurringscore in a range of scores. If, in a set of tenscores, five score ‘4’, three score ‘2’ and twoscore ‘1’, the mode is ‘4’.

2 The median score is the mid-point in a rangeof scores. In the following set of scores, themedian is ‘8’: 2 4 5 6 (8) 10 12 15 16. Thescore ‘8’ lies at the exact half-way point inthis distribution of scores. In other cases,where there is an even number of scores andtherefore no exact mid-point, the medianmust be calculated by averaging between thetwo centre scores: 3 3 5 8 (8.5) 9 10 13 14.Here, the median is 8.5, or the average of‘8’ and ‘9’.

3 The mean score is the average of the totalrange of scores. In the last example, theeight scores totalled 65, which, divided by8, gives a mean score of 8.125.


� methods and methodologies – Chapter 15, p. 258� forms of representing research findings – Chapter 16,p. 275

frequencydistribution

Another fundamental descriptive measure isthe degree of dispersion or variation in a set ofscores. While central tendency measures indi-cate the typical score of a distribution, disper-sion measures capture the extent to which thescores vary around that central point.

• Range, which is the simplest expression ofdispersion, is the difference between the highestand lowest scores in a particular distribution.

• Variance provides a mathematical index ofthe degree to which scores deviate from themean score, and tends to be expressed not interms of the original scores, but as squareddeviations from the mean. To compute the vari-ance, one subtracts the mean of a distributionfrom each score, and then squares the result.These squared scores are then summed anddivided by the number of original scores minusone. Variance is a powerful and widely appliedmeasure, like the standard deviation, and bothare illustrated below.

• The standard deviation is a third measure ofdispersion that utilises the original units ofmeasurement. The standard deviation is com-puted as the square root of the variance. If thevariance of a distribution of scores is 100, thestandard deviation (SD) for that distributionequals 10.

The normal distributionThe standard deviation and the mean can beused to further compute standard scores (zscores). Standard scores permit comparisons tobe made between two or more distributions or groups of scores, because all the scores arestandardised to the same metric, whereby themean is zero and the standard deviation is one.Within a given group, the z score expresses thevarious scores on a frequency distribution interms of a number of standard deviations fromthe mean. The point is that scores can thus be expressed in terms of their relative positionwithin a distribution, and not as absolutevalues. To exemplify, suppose two groups ofchildren, one of average age 10 years andanother of average age 8 years, are found todisplay average reading ability scores of 85 and68 respectively. While the older children clearly

have better reading scores than the youngerchildren, the most relevant comparisons are notbetween the two groups, but internally betweenmembers of these groups, and between eachgroup and a national average for that age. If,in fact, the average reading score for the 10-year-olds can be shown, relying on standardscores, to equal the average known readingscore for their age group nationally (givingthem a z score of zero), and the same is true ofthe 8-year-olds, then both groups may be con-sidered average for their respective age groups.

Standard scores are also used in conjunctionwith another fundamental statistical concept –the normal distribution curve (Figure 13.3). Ifa distribution of scores is normally distributed,its graphical curve should be symmetrical andachieve its maximum height at its mean, whichis also its median and its mode. One of the mostimportant features of the normal curve is thata fixed proportion of the area below the curve– representing a known proportion of the pop-ulation or other phenomena under investigation– lies between the mean and any unit of stan-dard deviation. The normal distribution is animportant analytical instrument because anumber of natural as well as social phenomenaare normally distributed, or nearly so. Not onlythe scores of mathematical tests, but pheno-mena such as the heights and weights of indi-viduals, and their IQ scores, all have normaldistributions. This means that if the average IQscore is 100 and the standard deviation is 15points, the proportions of people with scoresfalling between 85 (SD = �1.0) and 100, andbetween 100 and 115 (SD = +1.0) should bethe same. Likewise, the proportions of peoplewith IQ scores between 70 (SD = �2.0) and85, and between 130 (SD = +2.0) and 115should also be the same. However, the propor-tions of people with IQ scores within one stan-dard deviation of the mean will be much greaterthan the proportions with IQ scores betweenone and two standard deviations of the mean.

Testing hypotheses

Much quantitative research goes beyond thesimple description of data and their distribu-tions. In hypothesis testing, the researcher is


1111234567891012341115678920123456789301234567894012345678950111

normaldistributioncurve

interested in establishing whether two or morevariables are associated, or whether the scoresobtained in two or more groups are different –in both cases in an unambiguous or ‘significant’way. When deciding whether to accept or rejecta hypothesis, then, the researcher must examinethe statistical significance of the results. Thestarting point is to set up a null hypothesis, ora hypothesis which asserts that any statisticaldifferences or relationships that emerge withinthe dataset are due entirely to chance fluctua-tions or random error. The research hypothesisputs forward the alternative viewpoint: that thestatistical relationships or differences are notdue to chance, but represent real phenomenawhich can be explained theoretically in theresearch.

In determining whether a statistical test hasupheld the research hypothesis, a probabilitylevel must be set, so that the null hypothesiscan be tested against it. If the results indicate aprobability level lower than this level, the nullhypothesis may be rejected in favour of theresearch hypothesis. Conversely, if the proba-bility level indicated by the statistical test ishigher than the pre-set level, the null hypothe-sis must be accepted. The usual probabilitythreshold is .05. This means that there is a 5per cent chance that a relationship between twovariables, or a difference between two groups,occurred randomly; but there is a 95 per centprobability that the result demonstrates a realrelationship or difference.

In testing hypotheses, it is important toselect the appropriate statistical technique for

the data type in question. A broad distinctionis made between parametric and non-paramet-ric tests. Non-parametric tests are appropriateonly for nominal and ordinal data, while para-metric tests are appropriate for interval andratio data. Parametric results can be general-ised to the population as a whole, while non-parametric results cannot. Parametric testsassume normally distributed data, while non-parametric statistics do not depend on assump-tions about the precise distribution of thesampled population.

Using non-parametric statistics, researcherstest whether frequencies of the phenomenaobserved match the frequencies that might beexpected to occur by chance. A range of non-parametric tests are available for categorical ornominal (binomial test, Chi-square, Mcnemartest, Cochran Q test) and rank or ordinal data(Kolmogorov–Smirnov test, Sign test, Wilcoxonmatched-pairs, signed-ranks test, Friedmantwo-way analysis of variance, Kruskal-Wallisone-way analysis of variance) (see Bryman andCramer 1997; Siegel 1956). Probability tablesfor these tests may be found in Siegel (1956)and Hays (1973).

Parametric statistics assume that the dataare normally distributed, with means, variancesand standard deviations readily calculable. Aswith nonparametric tests, distinctions may bemade among different types of parametric sta-tistics, depending on whether the comparisonsare being made between groups or samples (t-test for groups in the sense of pairs, analysis ofvariance for three or more groups, or forsamples of individuals). In addition, parametricanalyses may explore whether correlations existamong scores on separate variables for one ormore groups (using the Pearson correlationcoefficient). While simple correlation measuresa relationship of association between two vari-ables, more complex regression analysis may be used to determine the degree to which onevariable changes, given a change in anothervariable. An even more sophisticated form of analysis, multiple regression, enables the re-searcher to find out the extent to which one(criterion or dependent) variable changes asmore than one other (predictor or independent)variable changes. In the latter case, each inde-


0 +1 +2–1–2

68%

95%

Figure 13.3 The normal distribution curve

the nullhypothesis

probabilitylevels

non-parametricstatistics

parametricstatistics

(multiple)regressionanalysis

pendent variable is examined in turn for itsimpact upon the dependent variable while, ineach step of the analysis, statistical controls are introduced for all the other independentvariables.

Pritchard and Hughes (1997) used multipleregression analysis in a content analysis of hownine characteristics of homicides might deter-mine whether and how they were covered innewspapers. Newspaper stories were classified,and numerically coded, in terms of the presenceor absence of specific attributes. There werefour dependent variables: the average length ofhomicide stories; the number of news itemsabout a homicide; the proportion of homicideitems on the front page; and whether a homi-cide story was accompanied by a photograph.

The results are summarised in Table 13.1. Itwas the average story length and the numberof items that were best predicted by the nine

independent variables. The scores displayed areso-called Beta weights, which indicate the indi-vidual predictive strength of each predictorvariable. The scores with asterisks are thesignificant predictor variables. The R2 scoreindicates the percentage of variance in each cri-terion variable (along the top of the table)which is accounted for by the combination ofall the predictor variables (down the side of thetable) that entered into that analysis.

In this analysis, the results show that homi-cides tend to get more newspaper coverage(number of items and average story length)when the event involves white participants, afemale suspect, a female victim, or a child orelderly victim. The further likelihood of such astory appearing on the front page increaseswhen the homicide involves whites, and whenthe victim is a child or elderly person. The like-lihood of a photograph being published with


1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 13.2 TESTING FOR STATISTICAL SIGNIFICANCE: CHI-SQUARE

A popular technique for the analysis of nominal or categorical data, such as results from muchcontent analysis, is Chi-square. In a study of the representation of violence on British television,Gunter and his colleagues examined the distribution of male and female aggressors (Gunter et al.1999). Focusing on aggressors acting on their own (as distinct from in groups, gangs or crowds),a total of 1282 acts of male-perpetrated violence and 385 acts of female-perpetrated violencewere found in drama programming. Together, these two categories comprised a total of 1667violent acts. Is this difference in the gender distribution of aggressiveness on television statisticallysignificant?

To find out, a Chi-square analysis was computed. If there is no inherent gender difference, itwould be hypothesised that 50 per cent of acts would be male-perpetrated and 50 per centfemale-perpetrated. Out of the current total, each gender would be responsible for 833.5 violentacts (the expected frequency). To compute a Chi-square, the expected frequency is subtractedfrom each observed frequency and squared.These squared results are then summed, and the totalis divided by the expected frequency (e.g., [1,282-833.5]2 + [38-833.5]2 / 833.5 = 482.7).

With reference to this value of the coefficient, a goodness-of-fit test can be carried out. It isnecessary, first, to determine the probability level (usually 5 per cent or .05) and, second, thenumber of degrees of freedom.The latter refers to the number of scores in a test that are free tovary in value.This is calculated as the number of groups of scores being compared, minus one (inthe current example, the degrees of freedom would therefore equal one). The researcher nextconsults a probability table that indicates whether the calculated coefficient is significant. Such atable displays probability levels along the top and numbers of degrees of freedom down the side.If the Chi-square coefficient exceeds the number displayed in the appropriate cell (here, one degreeof freedom by .05 level of probability), it is accepted as a significant result. If not, it is rejected asnon-significant. A Chi-square with one degree of freedom must exceed 3.84 to be significant atthe .05 level. The current result is, therefore, highly significant.

degrees offreedom

probabilitytables

the story increases when the victim is a child orelderly person, but decreases when there hasbeen a police information ban on the detailsthat could be released and included in the story.

CONCLUSION

This chapter has examined quantitative method-ologies as they are applied in media and communication research. These methodologieshave been used particularly to study media audi-ences and media content, and they do so byexamining either associations between variablesor cause–effect relationships. The guiding prin-ciple of quantitative research is the hypothetico-deductive approach – certain hypothetical ex-pectations are proposed, and then accepted orrejected through the collection and analysis of scientific data.

The essential characteristic of quantitativeresearch is that it reduces phenomena to numer-ical codes. Numerical measurement, however,can occur at more than one level, and an under-standing of the different levels of measurement

is crucial, both to the correct use of statisticalmethods of data analysis and to the properinterpretation of data. At the simpler levels,data are used merely to categorise and rankphenomena. At more complex levels of analy-sis, data may be used to measure relation-ships among phenomena and to establish causalconnections.

As a theoretical enterprise, much quantita-tive research aims to enhance knowledge bydemonstrating both causal links between phe-nomena and the universality of such relations.In practice, research must frequently make atrade-off between an ideal research design thatmight effectively demonstrate the nature andcausality of social phenomena, and a designthat is feasible, given the available resourcesand the necessary ethical and other considera-tions. Also for this reason, quantitative studies– like other traditions of research – should becarefully scrutinised for their methodologicallimitations and the quality of their data beforereaders and users attach weight and credibilityto their findings.


Table 13.1 Summary of a regression analysis for variables predicting four measures of homicidenewsworthiness (Pritchard and Hughes 1997)

Average Number of Proportion of Newspaperstory length news items items on published

published front page photograph

White participants .42** .38*** .25* .12Female suspect .33** �.25* �.18 �.19Female victim .26* .30** .15 .23Victim child/senior .33*** .33*** .31*** .31**Census tract income �.02 .13 .01 �.03Suspect knew victim �.11 �.02 �.12 �.10Risky behaviour �.06 �.02 .04 .01Police information ban �.15 �.16 �.15 �.25*Race/gender interaction �.25 �.25* �.22 �.01R2 .29*** .31*** .18* .17*

Notes:N = 100*p < .05, **p < .01, ***p < .001

INTRODUCTION

Into the 1980s, it was still common for human-istic and other qualitative researchers to referto their own contributions to media studies asbeing ‘non-scientific’ (Farrell 1987: 123), per-haps to secure a relatively underdefined nichefor something other than social ‘science.’ Whilethe distinction between ‘hard’ science and ‘soft’scholarship is still occasionally rehearsed in the new millennium (Rosengren 2000), the pasttwo decades have witnessed a shift from passivetolerance to active dialogue between self-defined qualitative and quantitative researchers.Simultaneously, the field has witnessed a wide-spread effort to specify the requirements ofqualitative research. This chapter reviews thecurrent state of qualitative methodologies, as employed in media and communication re-search, emphasizing the systematics and opera-tional stages of the qualitative research process.

First, the chapter identifies certain key con-cepts, originating from anthropology, sociology,

as well as the humanities, which inform theconduct of contemporary qualitative studies.Second, the planning of empirical projects isdescribed in terms of the several strategic, tacti-cal, and technical choices that must be madeduring their conception and implementation.Third, these overviews lead into a review, withexamples, of three prototypical methodologiesin qualitative media studies, as defined by theirmeans of data collection – in-depth interview-ing, participating observation, and document ordiscourse analysis. In each case, data analysispresents a particular challenge for future quali-tative research. The fourth section accordinglyoutlines procedures for the analysis and inter-pretation of qualitative data.

BASIC CONCEPTS IN QUALITATIVERESEARCH

Compared to a prototypical quantitativeresearch project, the ‘basic’ constituents of aqualitative project are best thought of as a

1111234567891012341115678920123456789301234567894012345678950111

The qualitative research processKlaus Bruhn Jensen14

• an overview of basic concepts within qualitative media research, including meaning as produced ineveryday settings and examined by researchers as interpretive subjects

• an argument for the importance of systematic design and sampling in qualitative research• a review of interviewing, observation, and other empirical approaches to media and communication studies,

with examples• a discussion of coding, discourse analysis, and other forms of qualitative data analysis as special challenges

for qualitative media research• a presentation of computer-supported qualitative research.

‘middle range’ of concepts (Merton 1968: 39).These concepts mediate between concreteresearch techniques and more grand theoreticalframeworks. Compared to much humanistic-textual research, or to classic anthropologicalfieldwork, qualitative studies have gone a long way toward specifying the procedures ofsuch a middle range of research. Qualitativeresearchers tend to conceive of their studies,most generally, as an iterative or repeated pro-cess, which allows for the flexible applicationof theoretical concepts and analytical proce-dures to a wide variety of empirical domains.

Qualitative research is a heterogeneous area,as documented through a growing number ofreference works (e.g., Bryman and Burgess1999; Denzin and Lincoln 2000; Miles andHuberman 1994). Nevertheless, it is possible toidentify at least three distinctive features thatare shared by most current work. First, it is the concept of meaning, its embedding in andorientation of social action, which serves as acommon denominator for different schools of thought. Human agents experience boththeir ordinary lives and extraordinary events asmeaningful. Cultural artefacts and other vehi-cles of meaning provide people with a sense ofidentity, a position from which to exerciseagency, and a means of orienting oneself insocial interaction. The textual contents of thetechnological media, but also their materiality,scheduling, and social uses, are studied by qual-itative research in order to explore empiricallyhow the media generate meaning. It is theconnection between meaning and action – asperformed inside media organizations (e.g.,Tuchman 1978) as well as by audiences (e.g.,Scannell 1988) – which, in part, distinguishesrecent qualitative work from earlier aestheticand historical studies of mediated culture.

A second assumption of qualitative researchis that meaningful actions should be studied, asfar as possible, in their naturalistic contexts. Inits strong form, this assumption calls for theclassical variety of anthropological fieldwork,in which a researcher’s lengthy immersion in awhole culture enables him or her to ultimatelygrasp in full ‘the native’s perspective’ on reality(Malinowski 1922: 25). In a more modestform, qualitative studies involve a weighing of

theoretical aims with practical constraints. Atissue are not just restrictions on time andmoney, but also epistemological questions (cana culture or context ever be known in full?) and ethical considerations (is it appropriate toimpose oneself on the people that make upcultures and contexts?).

Qualitative researchers also perform sam-pling – of cultural settings, communities, infor-mants, locales, periods, and activities. Thishappens as they transform theoretical conceptsinto empirical, researchable phenomena, forexample, television viewers’ definition of theircitizenship as articulated in interviews (e.g.,Monteiro and Jayasankar 1994), or their con-ception of the reality of melodrama as stated insolicited letters (e.g., Ang 1985). As elaboratedbelow, there are several possible strategiesbesides probability sampling and the totaldescription of all manifestations of a single caseor community. The naturalistic ambition,rather, stipulates that the phenomena of inter-est should not be parsed prematurely, or iso-lated from the contexts which facilitate theirinterpretation.

A third common feature of qualitativeresearch concerns the role of the researcher,who is defined emphatically as an interpretivesubject. In one sense, all research depends onthe human subject as a primary instrument.What distinguishes qualitative from quantita-tive projects in this regard is a global and con-tinuous form of interpretation. In quantitativestudies, interpretive agency tends to be exer-cised in a sequential and delegated form – seg-regating the phases of operationalization andanalysis from interpretation and discussion,and delegating certain moments of a study tocollaborators as well as machines, above all incomputerized data analysis. The qualitativeambition, in comparison, has been for a singleresearcher to interpret ‘meaning in action.’

In order to define the nature of such inter-pretation, much research has referred to a pairof concepts – etic and emic approaches toculture and society (Pike 1967). The ideaderives from linguistics, but has been widelyapplied in anthropology and, to a degree, inother social sciences. A ‘phonetic’ approach tolanguage assumes a continuum of sounds, as

The qualitative research process236

qualitativeresearch as

iterativeprocess

meaning

naturalisticcontexts

the native’sperspective

researcher as humaninterpretivesubject

emic and eticanalysis

measured on acoustic scales. A ‘phonemic’approach focuses on the distinctive set ofsounds that constitute one particular language.By analogy, other cultural forms than languagemay be understood as either a sound engineeror a native speaker would, and they may bestudied, accordingly, from either external orinternal perspectives.

The analytical question, as debated sincePike (1967), is how the two aspects of inter-pretation and understanding relate to eachother (see the contributions in Headland et al.1990). For one thing, the emic, internal per-spective is not simply found, but must be ascer-tained from some etic, external perspective. Foranother, this analytical perspective is itself emic, in that it represents both a historicalperiod and an academic subculture. Muchrecent work has questioned the distinction, andhas gone on, mostly from a postmodernist posi-tion, to redescribe research ‘findings’ as simplyone more narrative about the culture in ques-tion (see Clifford and Marcus 1986; James et al. 1997; Marcus and Fischer 1999; VanMaanen 1988; also Chapter 16, this volume).Ultimately, this position would do away withdistinctions between scientific and other formsof discourse, focusing on rhetorical and poeticaspects of research.�

As part of empirical procedures, however,the distinction is clearly important for specify-ing different levels and stages of analysisthroughout a research project. In qualitativemedia research, cross-cultural studies, forinstance, have described distinctive local expe-riences (emic) of globally disseminated genressuch as news and melodrama, as studied alsoby etic categories in content analysis and dis-course studies (e.g., Jensen 1998; Liebes andKatz 1990). For both qualitative and quantita-tive studies, moreover, there is no way aroundthe emic–etic consideration when devisingcodes and scales, as examined below, for theanalysis of mediated meaning.

DESIGNING QUALITATIVE STUDIES

Formating the field

To design an empirical study is to identify anddelimit a portion of reality – which is to be exam-ined with reference to a theoretically informedpurpose, or conceptualization, and according toa systematic procedure of data collection andanalysis. Whether in qualitative or quantitativetraditions, an operationalization of conceptsand issues in this broad sense is required (S. L.Schensul et al. 1999: 50). Only an empirical‘microcosm’ may be studied in any detail inorder to substantiate theoretical, ‘macrocosmic’inferences and conclusions (see further Alex-ander and Giesen 1987).� Planning, then, is anecessary part of that ‘construction’ of the fieldwhich most qualitative researchers recognize astheir epistemological premise. An early reminderof the importance of planning comes fromMalinowski (1922: 9), who concluded thatwhile preconceived ideas can get in the way ofresearch, the foreshadowing of problems in thefield is an important preamble to empiricalwork.

In order to identify key features of qualita-tive designs, it is useful to distinguish threeaspects of research design – strategy, tactics,and techniques (Gorden 1969). Notwithstand-ing any militaristic connotations, these aspectshelp to make explicit some of the premises andpriorities that inform any empirical study:

• Strategy, first, refers to a general plan forentering a particular social setting, and forestablishing means of communication andpoints of observation which may generate rel-evant evidence. Strategy builds a particular setof social relations between researchers andinformants in a designated time and space – atemporary structure for preparing and reflect-ing upon new knowledge.

• Tactics, next, refers to attempts by the re-searcher to anticipate and prestructure, to vary-ing degrees, the social interactions which willyield evidence. A case in point is the definition

Designing qualitative studies 237

1111234567891012341115678920123456789301234567894012345678950111 � rhetoric and poetics of research – Chapter 16, p. 287

conceptual-ization

operational-ization

� empirical microcosm and theoretical macrocosm –Chapter 15, p. 270

of people to be interviewed. They may be under-stood either as well-placed sources (informants,e.g., within a media organization) or represen-tatives of a particular position in the socialsystem in question (respondents, e.g., from anaudience segment) (Lindlof 1995: 170–172).

• Techniques, finally, are the researcher’sconcrete means of interacting with and docu-menting the field. Verbal language is a mainconstituent of all the methodologies reviewedbelow, but other sign systems and artefactsequally serve as sources of qualitative data.Whereas it is the choice of techniques that mostclearly distinguishes qualitative from quantit-ative designs, these instruments will be prefig-ured by strategy as well as tactics. These proce-dures, in turn, are anticipated by theoreticalpurpose and empirical substance of the givenstudy. In sum, the ‘why’ and the ‘what’ ofresearch come before the ‘how’ (Kvale 1987).

Sampling cases

The general formating, or operationalization,of the field is a first step in making a portionof reality accessible for empirical study. Anecessary next step is the sampling of elementsor constituents within that field. While mediaresearchers may most commonly think of ‘sam-ples’ as subgroups of ‘populations’ whichconsist of either people or texts, qualitativestudies frequently sample other units of analy-sis, for instance, settings, activities, and events(Lindlof 1995: 125). Furthermore, the samplingprocedure differs from probability sampling,�which is most familiar from opinion polls,surveys, and other quantitative studies.

Qualitative studies often sample in two ormore steps, first determining the relevant con-text of certain meaningful events, which, next,are singled out for detailed study. In an earlyqualitative classic of media research, Kurt and Gladys Lang (1953) examined the 1951MacArthur Day parade by relying on thirty-oneobservers on-site in Chicago as well as twoobservers monitoring the television coverage.From each of these contexts, representations

and accounts (including recorded observations,overheard remarks, and content elements) were compared. The findings indicated that on-site participants and television viewers wouldencounter two distinctively different perspec-tives on the same event. (As the authors havelater emphasized, that particular and import-ant finding was ‘entirely serendipitous’ (Lang and Lang 1991: 211). Interestingly, so was an-other classic finding, from a quantitative study,namely the two-step hypothesis concerningmedia impact (Lazarsfeld et al. 1944).)

Two-step and multi-step sampling is in keep-ing with the contextual orientation of qualit-ative research. This orientation implies, first,that statements and actions must always beinterpreted with reference to their context(s).Contexts are of a variable nature and scope –from a long discursive sequence, to familial andcommunity settings, and nation-states or entirecultures. Second, qualitative studies require, atleast ideally, that such contexts remain access-ible to the researcher in the process of analysisand interpretation. The qualitative researchprocess amounts to a continuous operational-ization and refinement of theoretical conceptswith reference to empirical evidence generatedthrough several analytical stages. As the Langs(1991) noted, some preliminary suggestions of how the media buildup to MacArthur Dayhad affected the public’s expectations anddemeanor in the streets could be tested againstother evidence, including ‘badges and behav-ioral cues’ (p. 212).

Regarding sampling procedures, much re-cent qualitative work has offered descriptionsof, and rationales for, distinctive approaches(e.g., Lindlof 1995: 126–131; Miles and Huber-man 1994: 28). These contributions followed a long period when qualitative samples weremostly defined en bloc, and in negative terms, as‘non-random.’ In this regard, qualitative sam-pling normally operates within contexts – whichwill have been preselected, like the ‘populations’of quantitative studies, according to theoreticalcriteria. Three procedures are of particularimportance:

1 Maximum variation sampling aims tocapture as wide a range of ‘qualities’ or pheno-


informantsand respond-

ents

units ofanalysis

� probability sampling – Chapter 13, p. 217

multi-stepsampling: of

and withincontexts

samplingprocedures

mena as possible. This range is normally sug-gested by certain other, established character-istics, such as the popularity ratings of radioprograms or the age of their listeners. In a studyof how American viewers evaluated televisionand possible ways of reforming it, I comparedtwo groups of viewers: below 35 and above 55years of age. What distinguished the groupswas that they had, and respectively had not,grown up with television as a cultural given(Jensen 1990). The study found, among otherthings, that the two groups relied on differenttypes of metaphor to describe how more, andmore diverse, programming could be madeavailable. The older group referred to a ‘library’of programs, whereas the younger group con-ceived of a continuous ‘flow’ of specializedchannels.

2 Theoretical sampling can also be performedwithin contexts, for example, in studies ofmedia organizations.� Here, the purpose oftenis, in a sense, to sample structures, i.e., differ-ent layers of the hierarchy which generates agiven product. The strategic junctures may behierarchical or spatial (in terms of the structureof the organization), or they may be temporal(in terms of the work process regarding thatmedia product). Theoretical sampling can relyon what is sometimes called ‘(proto-)typical’and ‘critical cases.’ In order to explore a theo-retical category, such as a movie or rock fan-dom (e.g., Lewis 1991), a project may focus oncore members of an established organization,or it may examine borderline instances – a com-peting fan formation or devotees of a-tonalmusic – as limiting or test cases.

3 Convenience sampling is sometimes used asa derogatory term for studying those individualsmost easily available to the researcher. However,a well-documented convenience sample cangenerate both valid and reliable insight into asocial setting or event. An example is the earlystudy by Cantril et al. (1940) which tried toexplain a public panic in response to OrsonWelles’s 1938 radio production of War of theWorlds. The study relied on a variety of quanti-

tative as well as qualitative approaches, and alsoemployed convenience samples of respondentswho were asked to recollect their experience ofbeing frightened. Moreover, given the notoriousdifficulty of gaining entry to certain socialarenas, convenience in the sense of physical andsocial accessibility is a legitimate consideration.(The alternative is to rule out research aboutsuch arenas – which may be motivated byethical concerns.) A variant is ‘snowball sam-pling,’ in which initial contact with an infor-mant generates further contacts. For each roll ofthe snowball, it is, in fact, possible to specifyhow a given choice relates to the overall purposeand design of a study, and, importantly, howearlier stages and insights enabled a moreinformed choice.

The three types of qualitative sampling allgenerate and document a set of meaningfulevents, instances, or cases for further study.One type of (primarily) qualitative design – casestudy – particularly explores delimited entities,such as communities and organizations, butalso singular individuals and events (see Gommet al. 2000; Yin 1994). In addition to theirinherent interest, a purpose of case studies isnormally to arrive at descriptions and typolo-gies which have implications for other, orlarger, social systems. What case studies sharewith other qualitative research is the detailedattention given, first to phenomena within theireveryday contexts, and second to their struc-tural or thematic interrelations with other phe-nomena and contexts.

Having formated the field of theoreticalinterest, and having sampled its empirical con-stituents, a qualitative project faces centralchoices concerning techniques or ‘methods’ –its concrete means of interacting with the field.These choices are normally posed in terms ofthe main instruments and procedures of datacollection. (Again, such choices are anticipatedby the domain and purpose of the study, andthe choices have consequences for the subse-quent analyses and for the potential social usesof the findings.) The following sections reviewthe three main instruments of qualitativeresearch: interviewing, observation, and docu-ment or discourse analysis. Chapter 15 returns

Designing qualitative studies 239

1111234567891012341115678920123456789301234567894012345678950111

� qualitative studies of media organizations – Chapter4, p. 66

prototypicaland critical

cases

snowballsampling

case study

to the combination of several methods and tothe relations between ‘methods’ and ‘method-ology.’�

INTERVIEWING

Interviewing is one of the most widely useddata collection methods, also in media andcommunication research. A commonsensicaljustification for this fact is that ‘the best wayto find out what the people think about some-thing is to ask them’ (Bower 1973: vi).Particularly in-depth interviewing, with itsaffinities to conversation, may be well suited totap social agents’ perspective on the media,since spoken language remains a primary andfamiliar mode of social interaction, and onethat people habitually relate to the technologi-cal media. The difficulty, of course, is thatpeople do not always say what they think, ormean what they say. In research as in othersocial practice, communication has its purposesand contexts, which must be teased out byresearchers, as by other communicators.

It is crucial, then, to realize that statementsfrom focus groups or biographical interviews(or from surveys) are not simple representa-tions, true of false, of what people think. Allinterview statements are actions, arising froman interaction between interviewer and inter-viewee. Interview statements are, in a strongsense of the word, ‘data,’ and they becomesources of information only through analysisand interpretation. For one thing, interviewstudies ask people to ‘discursify’ things whichcommonly fall into the category of practicalconsciousness (Giddens 1984). For another,interviewers themselves have no perfect aware-ness either about their own performance or ofthe responses which they must process in a splitsecond. The disambiguation of interview dis-courses (or the conclusion that an ambiguity isunresolvable) is the outcome of data analysis,and will remain an inference. This is in spite ofthe occasional suggestion in textbooks that ahighly competent interview may stand fullyinterpreted when it ends (Kvale 1996: 189).

These points reiterate a classic insight of

rhetoric and of the ‘hermeneutics of suspi-cion’:� There is no way around language as amedium of access to social and cultural phe-nomena. Language is a permanent condition ofresearch, not a removable obstacle. On the onehand, respondents’ self-conceptions, opinions,and worldviews must be inferred from theirlanguage (and other systems of communication)and their argumentative structures, culturalthemes, and narratives. On the other hand, lin-guistic categories offer means of quality controlregarding the ‘language work’ of researchers, asexplained below.

Figure 14.1 summarizes the key role of lan-guage in qualitative research, both as a tool ofdata collection and as an object of analysis.(The double notation concerning language asan object of analysis in observational studiesindicates that language is often not documentedor analyzed in systematic detail, as explained inthe section on observation.)

In media research, qualitative studies haveprimarily employed three types of interviewing.The types reflect the basic options of inter-viewing one or more persons, with or withouta pre-established relationship with each other:

1 Respondent interviews. In contrast to infor-mant interviews (less common in media studies),the interviewee is here conceived as a represen-tative of one or more social and cultural cate-gories. The assumption is that these categoriesare inscribed in, and can be recovered from, the respondent’s discourses with reference to the media. Respondent interviews have beenprevalent not least in reception studies of the decoding of media.

2 Naturalistic group interviews. In order toexplore, to the extent possible, what ‘normally’goes on in a social setting, qualitative studieshave examined naturally occurring groups inthe case of both media production and recep-tion. In production studies, interviews have com-monly been an integral part of more inclusiveobservational methodologies, although specific(especially individual) interview methodologies


� methods and methodology – Chapter 15, p. 258� interview analysis as hermeneutics of suspicion –Chapter 2, p. 22

have been employed (e.g., Newcomb and Alley1983). For audience studies, household inter-views may produce several conflicting perspec-tives (e.g., by children and parents) on what is a ‘normal media day’ in the home (e.g., Jensenet al. 1994). Interviews with fan groups (e.g.,Spigel and Jenkins 1990) and with children andyouth peer groups (e.g., Livingstone and Bovill2001) can also supplement the observation oftheir distributed activities.

3 Constituted group interviews. Groups thatare constituted specifically for the purpose ofresearch represent a compromise between therespondent and naturalistic strategies. Groupmembers remain bearers of particular demo-graphics, while entering into an approximatednatural group dynamic. An example is Liebesand Katz’s (1990), which relied on householdsto invite acquaintances to their home in orderto watch and discuss the Dallas television series. The classic inspiration for group inter-views was the work by Robert K. Merton andhis associates beginning in the early 1940s,which resulted in ‘the focused interview’ (Mer-ton and Kendall 1955; see also Merton 1987).The general idea became especially influential inthe shape of ‘focus groups’ in marketing and,later, in media research (for an overview seeMorgan and Krueger 1998). Morrison (1998) isa key source to the development, and laterabuses, of the method. In media studies, focusedinterviews with relatively homogeneous groupshave proven useful for gaining access to theirdistinctive experiences of media content, some-times in conjunction with other methods (e.g.,Schlesinger et al. 1992).

It should be added that a wide variety ofcommunicative and other techniques are avail-able for eliciting both mediated and lived expe-rience (e.g., Bryman and Burgess 1999;Marshall and Rossman 1999; Punch 1998).The tradition of ‘oral history’ (Dunaway andBaum 1996), relying on the lengthy oral testi-mony not least from ‘ordinary’ people, who cantell history ‘from below,’ has made its mark alsoon media studies. In addition, while most qual-itative as well as quantitative methodologiesask people to respond to society and culture asthey now exist, some approaches invite groupsto anticipate and evaluate forms of communi-cation that are not yet in existence. Such studiescould be said to stimulate the sociological imag-ination (Mills 1959). The study cited aboveabout older and younger viewers’ conception oftelevision (Jensen 1990) relied on such amethod, ‘workshops on the future’ (Jungk andMüllert 1981). In a critical reflection on focusgroups, Barbour and Kitzinger (1999) noteother action-oriented varieties of interviewing,such as Delphi and consensus groups.

Across the different qualitative interviewforms, three issues normally call for theresearcher’s attention: duration, structure, anddepth:

1 Duration. Interviews range from brief dia-logues that establish the meaning of a term ata production site, to repeated hour-long ses-sions with an individual or family, to even morecomprehensive life-historical interviews (e.g.,Bertaux 1981; Prue et al. 2000). Duration isdetermined largely by the overall purpose of astudy, in addition to practical circumstances.

Interviewing 241

1111234567891012341115678920123456789301234567894012345678950111

Methodology

Interviewing

Observation

Documents/artefacts

Tool of data collection

+

+

–

Object of analysis

+

+/–

+

Language

Figure 14.1 The role of language in qualitative methodologies

focusedinterviews and

focus groups

workshops onthe future

oral history

2 Structure. Probably the main challenge inplanning and administering interviews is how toprestructure the interaction. Depending againon the theoretical purposes, the exchange maycover a predefined set of themes from severalperspectives, but in no particular order, or itmay follow a sequenced and subdivided inter-view guide (see, e.g., S. L. Schensul et al. 1999:121–164). In all events, it is the responsibility ofthe researcher to justify choices in this regard,and to make explicit the procedures which ledto particular inferences and conclusions.

3 Depth. The researcher’s responsibility be-comes even more acute when assessing therelevant ‘depth’ of a qualitative interview.Although the aim will be to elicit the respon-dent’s terminology, or to probe a conceptualstructure, the research interview process hassimilarities with the therapeutic interview, andmay articulate tacit or repressed insights (Kvale1996: 74–80). Depth, which is generally con-sidered the strength of qualitative research, thusalso poses serious issues for the ethics andpolitics of research.�

OBSERVATION

‘Observation’ refers inclusively to a set ofresearch activities that involve the continuousand long-term presence, normally of oneresearcher, and generally in one delimitedlocale. An observer is able to emphaticallybecome the central instrument of research,relying on several sensory registers and ondiverse media information. One difficulty ofsuch immersion in the field – beyond the dangerof ‘going native’ and conflating emic and eticperspectives – is documentation. This is import-ant both for the ongoing collection and analy-sis of data, and for making transparent thesteps from an initial observation to a laterconclusion. Unless documentation is presentedin an explicit, reflexive, and intersubjectivemanner, a fieldworker may become akin to anartist. Holistic interpretations of some possibleworld may be inspired and inspiring, but willoften be non-negotiable and, hence, beyond

research. Within anthropology, debates aboutsuch issues have a long history. For example, itis only recently that the sharing of fieldnoteswith other researchers has become a commonpractice (S. L. Schensul et al. 1999: 226).

One of the most influential metaphors, alsoin media studies (Deacon et al. 1999; Jensenand Jankowski 1991), for what observationaccomplishes is Geertz’s (1973) ‘thick descrip-tion.’ (Geertz derived the concept from thephilosopher, Gilbert Ryle (1971: 465–496).)The point is that a very detailed analysis of asetting is necessary in order to establish theimplications of what people do or say, forexample, when they use irony. Rather thanspreading one’s resources thinly across a largefield, and predefining the phenomena of inter-est, efforts should be focused on a smaller fieldwhich is explored both for relevant phenomenaand for descriptive categories. This approach isin keeping with the qualitative ambition ofsearching out one’s analytical categories in thefield itself – even though research questions andpurposes inevitably orient a study. Some crit-iques have identified a rather vague notion ofwhat thick description requires, leading in factto ‘thin’ descriptions (Murdock 1997). Never-theless, the open, inclusive approach is helpfulespecially in the early stages of a study, and iscomparable, to a degree, to pilot studies inother research designs.

A common way of distinguishing varieties offieldwork is to refer to its relative componentsof participation and observation, respectively.Summarizing earlier accounts of participatingobservation, Hammersley and Atkinson (1995:104) indicate a scale from full observation to full participation, suggesting that there is normally a degree of each element in any fieldwork. Just as any interview question, in asense, is leading in presuming a particular scopeof relevant answers, so observers participate,and participants observe, as they try to interpret‘what is really going on.’ In studies of media usein households and other private settings, forexample, it is evidently difficult to maintain theobserving role of a fly-on-the-wall. However,the study of media use in public places can alsobe said to involve a measure of participation(Lemish 1982).


� research ethics and politics – Chapter 16

thickdescription

participatingobservation

One advantage of the reference to degreesof observation and participation, and to inter-faces with other data collection methods, is thatit avoids the controversial, and questionable,terminology of ‘ethnography,’ which was im-ported into media and communication researchparticularly during the 1980s.� It is doubtfulwhether ‘ethnographic’ studies are different inkind from other multimethod media studies.

Compared to interview studies, which havebeen able to rely on affordable audiotape-recorders since around 1950 (Fielding and Lee1998: 28), observational studies have contin-ued to face special problems of documentation,relying on heterogeneous types of data, butwith handwritten fieldnotes as the naturalcenter-piece. Thus, while observers rely onseveral sensory registers, the resulting observa-tions are only to a limited extent documentedin several discursive registers, such as visualmedia and non-prose genres. In a summaryvolume on the subfield of visual anthropology,Margaret Mead concluded that anthropologyhad remained ‘a discipline of words’ (Hockings1995: 3), a characterization that still rings truealso in (visual) media studies (but see Bauer andGaskell 2000; J. J. Schensul et al. 1999).

Regarding fieldnotes, there have been fewconsensual, full-fledged procedures. This is incomparison to detailed conventions concerninginterview transcript notations (for an overviewsee Potter 1996: 233–234) and content analy-sis. The presentation and analysis of verbatimnotes in publications are still exceptions, cer-tainly in observational media studies. This stateof affairs has been exacerbated by the assump-tion that fieldnotes serve only as extensions of more essential ‘head notes.’ The furtherassumption – that interpretations could only beconducted adequately by the researcher whowas present at the scene – has helped to per-petuate the view of the fieldworker as a creativeloner, an artist. (When it emerged, Malin-owski’s (1967) ‘secret’ diary, which, amongother things, revealed his contempt for thepeople being studied, undercut his descriptivemonograph and questioned both the validityand the legitimacy of his accounts.)

In the past few decades, more attention hasbeen given to systematic production of fieldrecords (e.g., Ellen 1984: 278–293; Lindlof1995: 200–215; Spradley 1979: 69–77). Ahelpful typology distinguishes three purposesand genres:

1 substantive notes, which capture represen-tations of the scene under study;

2 logistical notes, which add informationabout the circumstances under which thesedata were gathered;

3 reflexive notes, which initiate the process ofanalysis and theorizing on the basis ofobservations and other data (adapted fromBurgess 1982).

An additional rule of thumb is to focus on sub-stance (what) and logistics (how) in the field,and to reserve the main reflexive activity (why)for later stages of the research process. Theresearcher can only fulfill his/her role as themain instrument of research through a differ-entiated and staggered process of analysis,interpretation, and self-reflexivity.

The wider lesson for empirical studies is toapproach fieldnotes not as self-contained rep-resentations, but as working documents fromone stage or level of study. As elaborated in alater section, these documents, like interviewrecordings and transcripts as well as variousother artefacts, feed into an equally differenti-ated analytical process.

DOCUMENTS, ARTEFACTS, ANDUNOBTRUSIVE MEASURES

The third group of approaches to data collec-tion is more heterogeneous than either observa-tion or interviewing. What unites them is theirrelatively naturalistic or unobtrusive (Webb et al. 2000) nature – the data are ‘found’ ratherthan ‘made’ through the researcher’s interven-tion in the field. On the one hand, data such asfeature films and computer games, governmentreports and executive memos, are produced aspart of the normal business of media, and thusare not ‘biased’ by the researcher. On the otherhand, the data may therefore have a morelimited or indirect explanatory value for theresearch question.

Documents, artefacts, and unobtrusive measures 243

1111234567891012341115678920123456789301234567894012345678950111 � ethnography – Chapter 10, p. 164

visualanthropology

field notes asworkingdocuments

Overall, these approaches represent a poten-tially valuable, and somewhat neglected, sourceof evidence. In certain respects, they addressissues which have fallen outside, or between,the most common social-scientific and human-istic methodologies. The humanistic traditionhas remained largely text-centric, focusing onpre-existing texts; social-scientific studies haveemphasized the production of new data regard-ing the contexts of media. In this last respect,it should be noted that media research has repli-cated a tendency in other social research toprivilege new datasets, and to de-emphasize, forinstance, written documents (Atkinson andCoffey 1997: 45; Hammersley and Atkinson1995: 157–174). A case in point is biographies(but see Chamberlayne et al. 2000). This maybe attributed, in part, to the original focus ofanthropology on oral cultures, and of sociologyon modernity as lived, perhaps reinforced by acertain romanticism for ‘authentic’ data.

As suggested by the headline of this section,the approaches are characterized each by theircentral object of analysis, rather than any uni-fying method. First of all, the textual output ofmedia has naturally been a central object ofanalysis in qualitative media studies, being thevehicles of cultural forms and historical world-views.� Moreover, qualitative research hasbeen instrumental in extending the notion of‘text’ to processes of intertextuality and ofmeaningful social interaction. The texts includeadvertising, publicity, as well as audience ex-changes with and about media.�

A second type of text or document relates toproduction and reception. Like other modernorganizations, the media are highly prolific gen-erators of documents that prepare and feed intothe media texts proper. Like other cultural insti-tutions, they attract written responses such asaudience letters (e.g., Collins 1997), but alsodiaries, autobiographies, and fan fiction reflecton the social role of media (see Hammersley andAtkinson 1995: 159). Film studies, for instance,have relied on diverse data sources regardingearly movie audiences (e.g., Stokes and Maltby1999a).

Figure 14.2 suggests a further distinctionbetween evidence that is associated with, andprimarily originates from, sources in the private(intimate) or public domain. In each case, it isnecessary to keep in mind that, like observa-tional records, interview transcripts, as well asaudience measurements, the documents thus‘inherited’ are not ready-made representationsof social facts. They are the outcome of previ-ous social interactions, whose circumstancesmust be part of the object of analysis and inter-pretation.

Third, artefacts and various physicalarrangements pertaining to media are also rele-vant sources of evidence. Media are most com-monly thought of as means of representation, orvehicles of meaning, but media are also physicalobjects and constitutive elements of other socialaction. One example is the use of a newspaperor other reading material as a resource forwarding off unwanted contact, typically in apublic context such as mass transit, but also inthe home as a means of insisting on one’s per-sonal time or space (Radway 1984: 61).�Another example is the design of media hard-ware such as early television sets and theirembedding in the home setting, as studied bySpigel (1992) with reference to their depictionin magazines. Media hardware also constitutesobjects-to-think-with.� Similarly, cinema archi-tecture establishes an experiential setting, notonly for movie viewing, but for the collectiveactivity of ‘going to the movies’ (e.g., Gomery1992).

It should be added that computer media andnetworked communication pose new difficultiesas well as opportunities, with the researchstrategies still in the making.� For data collec-tion, digital technologies offer increasinglyminiaturized equipment, for example, digitalcameras, for use on-site or on-line. (The rele-vance of computers for qualitative data analysisis taken up below.) In general, computer-medi-ated communication has problematized the dis-


� media texts – Chapters 7–8


productionand reception

documents

� physical arrangements – Chapter 10, p. 169

� media artefacts as objects-to-think-with – Chapter 1,p. 6

� methodologies of computer media studies – Chapter11, p. 183

tinction between media and their embeddingsocial context. The dispersed, customizable,and interactive features of computer media allcontribute to their integration into social con-texts of action (and into interpersonal commu-nication), the study of which requires severalforms of evidence.

To sum up, this third group of approachesholds untapped opportunities for media andcommunication research. One of the relativelyfew sustained efforts has addressed children’sexperience of their media use, which often lendsitself better to, for instance, drawings thanverbal evidence (e.g., Rydin 1996). In the classicon unobtrusive measures (Webb et al. 2000),the references to media are many, includinginfra-red recordings of audiences in darkenedcinemas (p. 154), and fingerprints used to deter-mine which advertisements have been seen orread (p. 44) – the ethical concerns here becomeacute. Beyond the growing, but still limitedattention to images and artefacts as researchevidence, some work has gone on to study ‘the seen’ (Emmison and Smith 2000: ix) – theobservable two- and three-dimensional con-texts of social and cultural life, including publicspace and material culture, shopping malls andgraffiti. The growth of ‘virtual’ worlds in com-puter-mediated communication is likely todraw renewed attention also to the spatial,architectural aspects of media in society.

DATA ANALYSIS

Coding and analysis

The question of what constitutes ‘analysis’ hasbeen a central area of controversy betweenqualitative and quantitative traditions of media

studies. From one perspective, a textual analy-sis or case study might be said to perform, notan ‘analysis’ (e.g., an explicit segmentation andsubsequent categorization of component partsto anticipate later inferences and conclusions),but a new ‘synthesis,’ a general reinterpretationof the object of analysis. Qualitative data analy-sis has, indeed, suffered from an insufficientspecification and documentation of its pro-cedures and stages. As of the present, however,a whole range of approaches and techniques arebeing applied. Studies rarely employ standard-ized logical or mathematical models, but cer-tainly rely on systematic procedures. In fact,recent qualitative research has helped to reopenthe dialogue with quantitative research con-cerning the definition and status of ‘description’and ‘interpretation,’ ‘findings’ and ‘discussion,’‘analysis’ and ‘synthesis.’

Like any scientific enterprise, a qualitativestudy is committed to carrying out an analysiswhose elements, procedures, and stages areexplicit, documented, and can be argued about.A hallmark of any competent research report isthat it delivers a qualified basis for disagreement– for resolving, or specifying, differences regard-ing methodology, theory, or epistemology.

What distinguishes qualitative research, inparticular, is that key concepts and other mini-mal constituents are defined and redefined aspart of the research process itself. Moreover,the contexts in which such constituents are tobe interpreted equally remain open to redefini-tion throughout the study. Accordingly, syn-thesis is not a single concluding act, but acontinuous activity of assessing data and artic-ulating concepts. Far from justifying unspeci-fied references to ‘patterns’ that ‘emerge’through a ‘spiral of insight,’ the qualitative

Data analysis 245

1111234567891012341115678920123456789301234567894012345678950111

Private

Public

Production

Autobiographies by film stars, journalists,or other media practitioners

Organizational archives, from policypapers to rewrites and worksheets

Reception

Clippings and other memorabiliaconcerning a media personality or genre

Letters to the editor, fan magazines, orcomputer-mediated news groups

Figure 14.2 Examples of documents relating to media production and reception

public spaceand material

culture

children’sdrawings

analysis andsynthesis

research process requires systematic and dis-tinctive formats of presentation and argument.

Several practical issues can be laid out withreference to the notion of ‘coding.’ Some tradi-tions treat coding as a necessary and self-evidentingredient of, even a synonym for, ‘analysis.’Importantly, the term covers two differentunderstandings of how words, numbers, andmental categories can be matched to pheno-mena in reality. In slightly different inflections,the two conceptions have recently beendescribed as heuristic and factual codes (Silver-man 2000: 170), and as indexing and represen-tational devices (Fielding and Lee 1998: 176).

On the one hand, a code may be taken asan account or representation of a portion of thefield of study, capturing and fixating certainqualities of a person, event, text, or other unitof analysis for the purpose of later comparison.The aim is to arrive at exhaustive and mutuallyexclusive categories. Coding in this sense, more-over, aims to establish a standard by which suchqualities may be conferred across contexts and,in the end, quantified.

On the other hand, a code may be under-stood as a resource or instrument for identify-ing and retrieving a given portion of the field.In a next step, this unit of analysis may beexamined either for its immanent structure andspecific qualities, or with reference to someadditional portion of its context. Here, theambition of research in each case is to examinea wider setting, which is said to always circum-scribe and embed the central object of analysisand its meaning.

The two conceptions of coding are clearlythe legacy of quantitative and qualitativeresearch traditions, respectively. At the sametime, coding offers an interface between the twotraditions, since an index may be developed into a representation, and vice versa. Typically,a qualitative study will identify subsets andsequences of data which are related thematicallyor structurally, and which can be singled out for detailed and perhaps comparative analysis.In this connection, it bears repeating that themovement from fieldnotes, tapes, and tran-scripts to final research reports comprisesseveral steps – memoing, modeling, drafting,recontextualizing – each of which lends itself to

documentation, if not necessarily standardiza-tion. Referring to financial audits of firms andorganizations, Lincoln and Guba (1985: ch. 11)developed the term ‘U’ to refer also to ways ofkeeping the research process transparent to theresearcher in question, and accountable to col-leagues and readers (see also Atkinson andCoffey 1997).

Variants of data analysis

The history of qualitative data analysis spansrhetoric, hermeneutics, and semiotics, as wellas several strands of social science (for an over-view, see Fielding and Lee 1998: 21–55;Chapter 2, this volume). Much early socialscience was based not on codes, but on thestudy of cases. This was, in part, a response tothe search for general categories and theoreti-cal frameworks within emerging disciplines.One approach was outlined by Znaniecki(1934), who, under the heading of analyticinduction, wanted to replace enumerativeinduction and probabilistic statements with theintensive study of single cases in a stepwiseattempt to arrive at general, even universal cat-egories of social phenomena. Despite examplesof the approach also in media studies (Lang andLang 1953), it has not gained wide acceptance,partly because of its time-consuming proce-dures and limited practical applications, partlybecause of epistemological doubts concerningthe universality of its categories.

Occasionally, proposals have been made forformalizing qualitative analysis into matrices or truth tables. For example, there have beenattempts at accounting for cases (social events,processes, or units) with reference to the com-bination of historical and cultural conditionsunder which particular cases occur (e.g., Ragin1987, 1994). A second, more broadly definedapproach to studying cases was prevalent fromthe 1920s onward in the Chicago Schoolstudies particularly of urban life, but also ofmedia as in the Payne Fund studies.� Third,grounded theory, below, took part of its inspi-ration from analytic induction.

By the 1950s, a noticeable shift had


heuristic andfactual codes

audit trail

� Payne Fund case studies – Chapter 10, p. 158

occurred in social science that could be summedup as a ‘transition from case-based to code-based analysis’ (Fielding and Lee 1998: 27).�The growing professionalization of research aswell as the increasing availability of techno-logies such as tape-recorders, typewriters, and, subsequently, computers are among theexplanations for this shift. The question was,and remains, what exactly this entails – for example, where on a continuum from heuristicto factual coding does an analysis situate itself?In qualitative studies, three attempted responsesare of particular interest.

Thematic coding

A widespread approach in media research hasbeen a loosely inductive categorization of inter-view or observational extracts with reference tovarious concepts, headings, or themes. Theprocess comprises, to varying degrees, the com-paring, contrasting, and abstracting of the con-stitutive elements of meaning. It is the veryoccurrence of a particular theme or frame in acontext of communication which is of primaryinterest to qualitative research. Such catego-rizations of data, in turn, can support infer-ences concerning, for example, the selectivereception and reconstruction of news items bydifferent audience groups (e.g., Morley 1980).

Some qualitative research has relied on con-sensual or group coding of themes in order tosubstantiate their conclusions (e.g., Neuman et al. 1992: 32–33). In addition, some studieshave emphasized the development of modelsand nonverbal tools as integral elements of astepwise analytical procedure. A main contri-bution has come from Miles and Huberman(1994), who place special emphasis on ‘datadisplay’ in the form of figures and graphics asaids in interpretation. Perhaps surprisingly in afield which otherwise tends to work throughmodels of communication and representation,such approaches are rarely encountered in qual-itative media studies.

To a degree, this inclusive understanding ofcategorical analysis is compatible with coding

as performed in content and survey studies. Inan important early contribution to qualitativeresearch, Lazarsfeld and Barton (1951) ex-plored ways of transforming classifications ofempirical units into typologies, indices, andmodels. Even though their inclination, in keep-ing with the period, was toward formalization,the article highlighted the many possible vari-eties and stages of data analysis. The authorsalso acknowledged the necessary interpretivelabor of the researcher throughout the analyti-cal process.

Grounded theory

The second main variety of qualitative dataanalysis is ‘grounded theory’ (Glaser andStrauss 1967). It became influential in the socialsciences from the 1960s as a framework thatwould name and legitimate an alternative, par-ticularly to survey research. As suggested by thename, the approach is a methodology whichtends to assume that theory can be ‘found’ inthe field, if the research activity is sufficiently‘grounded’ in the categories of that field.

Relying on a ‘constant comparative method,’researchers in this tradition have developed adetailed set of procedures for collecting andanalyzing empirical data. A characteristic fea-ture is the several, and often repeated, stages ofsampling, analyzing, memoing, and interpretingmaterials; another feature is a stepwise processof coding data at different levels of abstraction.Most important, perhaps, is the assumption that this sequence may ultimately produce theo-retical ‘saturation’ – an equilibrium betweenempirical evidence and explanatory concepts.Compared to other qualitative traditions which will often rely on theoretical sampling,grounded theory has emphasized an iterative,inductive variety of both sampling and theorydevelopment.

As part of the more recent return to quali-tative methodologies, the approach has beendisseminated widely through textbooks (Strauss1987; Strauss and Corbin 1990) and includedin reference works (e.g., Denzin and Lincoln2000). It should be mentioned, however, thatthe terminology appears to be more widespreadthan the practice, and is ‘sometimes invoked

Data analysis 247

1111234567891012341115678920123456789301234567894012345678950111

� from case-based to code-based research – Chapter 15,p. 256

consensualcoding

data display

constantcomparativemethod

theoreticalsaturation

. . . to legitimize an inductive approach’(Fielding and Lee 1998: 178). This descriptionapplies to media studies, despite occasional ref-erences to a more elaborate use of the approach(Lindlof 1995: 223).

Part of the explanation for this state ofaffairs may be problems inherent in the foun-dations of grounded theory. In particular, other researchers have questioned the apparentpremise of grounded theory that a researchermay enter the field without theoretical pre-suppositions (for an overview see Alvesson and Sköldberg 2000: 12–36). Not only is thatposition epistemologically dubious, but it mayencourage and justify researchers in not reflect-ing on the scientific as well as social conditionsshaping their work. Instead, later textbookswithin the tradition have suggested increasinglycumbersome procedures, designed to ensure thegrounding of categories in the field itself. Infact, the origins of the procedures are often notclear, nor is it evident how they differ fromother available qualitative methods. A furtherconcern is that, at least in Strauss’s version(Strauss and Corbin 1990), the analytical pro-cedures cut off social events from their contextof other events, as each event is analyzed, rean-alyzed, and condensed within increasinglyabstract categories.

The uncertain status of the theoretical cat-egories of grounded theory and the question ofhow to relate them to empirical phenomena has been part of the backdrop to an unusuallyvehement confrontation between the two foun-ders of the tradition. Glaser (1992) attacked his co-author for an ‘immoral undermining’ (p. 121) of their original joint contribution and demanded the withdrawal of a volume co-authored by Strauss (Strauss and Corbin 1990).In the end, it remains unclear what distinguishesgrounded theory beyond a rather commonsensi-cal notion of sampling, comparing, and reflect-ing in a reiterated sequence, as conducted bymost (qualitative) researchers. The frequent ref-erencing of the tradition may testify to the gen-eral weakness of some earlier qualitative dataanalysis, and to the need to further develop suchapproaches as thematic coding and discourseanalysis.

Discourse analysis

A third approach to data analysis draws itsinspiration from discourse studies.� Specialattention is given to it here because of its poten-tial for further qualitative research. On the onehand, discourse analysis offers a systematic andoperational approach to different kinds of cod-ing, including consensual and thematic pro-cedures. On the other hand, it avoids thecondensation and decontextualization of mean-ing which is implicit in grounded theory as wellas most quantitative versions of coding.

A way of tapping the advantages of bothcoding and close analysis is to distinguishbetween two moments of data analysis:

1 Heuristic coding – a preliminary and itera-tive assignment of verbal and other codes to different portions and levels of a dataset.Examples range from metaphors, expressed ina single word or image, to complex narrativesabout particular issues.

2 Discourse analysis – a more definitive anddetailed categorization of a data extract and itsconstituents. As exemplified in Box 14.1,aspects of form as well as of thematic contentoffer relevant units of analysis, from a respon-dent’s use of pronouns in an interview, to thenarratives constructed jointly by a circle ofviewers as observed around a television set.

Heuristic coding allows researchers to producea working document, or summary, of elementsand structures occurring in a dataset. This firstanalytical step does not preclude, but rathersupports a later detailed analysis of discursivesegments in context. One purpose of this laterstage of analysis may be to test hypotheses thathave emerged, including the identification ofpossible counterexamples. An advantage of thistwo-pronged approach is that it promotes multiple forms and levels of analysis. A relatedadvantage is that both heuristic codes and lin-guistic categories are compatible with contentanalysis and survey techniques. This is exempli-fied by studies of so-called keywords-in-contextresearch (KWIC) (Fielding and Lee 1998: 53), a


� discourse studies – Chapter 7, p. 104

keywords incontext

Data analysis 249

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 14.1 DISCOURSE ANALYSIS OF QUALITATIVE INTERVIEW DATA

In a study of the reception of television news by American men at different educational levels, acombination of heuristic coding and discourse analysis was employed to explore the experientialqualities and social uses of the news genre (Jensen 1986).The data analysis was structured accord-ing to a set of themes which had been operationalized in the interview guide.The heuristic codingidentified references to these themes, both in response to the dedicated questions, and elsewherein each interview. Next, the relevant textual sequences were examined in further detail throughdiscourse analysis. The following illustration pays special attention to the level of continuous dis-course (argument), but also considers the levels of speech acts and interaction. Of special inter-est at the discourse level are:

• generalizations – summary statements, often signaled by adverbials, conjunctions, and verbalor ‘do’-emphasis, and by initial or final placement in the speaker’s turn;

• substantiations – the supporting reasons or examples given for a generalization;• implicit premises – the unquestioned point of departure for an argument, either a logical

presupposition or a natural assumption in context;• implications – what follows from a statement with varying degrees of certainty, depending

on the speech community, the immediate context, as well as the wider social and culturalsetting.

At issue in the following quotation is the theme of flow – whether the respondent carries overfrom another television program to the news. Asked by the interviewer (I) whether he does that,this respondent, a junior university professor, says:

‘No. (I: No?). I don’t think so, because I have a real thing, when I was living at home mysister is one who just always has some appliance on (I: H-hm), and I really, something deepin me, I really dislike that (I: H-hm), so that, no, if I get up and if the first story didn’t catchme, or maybe even if I was done with that program I’d turn it off (I: H-hm) and not keepit on just because it had been on the hour before’ (p. 177).

The generalization – a denial of this possibility – is expressed both in the initial emphatic ‘no’ andin the summarizing ‘so that, no.’ However, two quite different substantiations are offered.The firstreference is to a situation where the first news story does not catch him; the implied premise ofthe uncompleted sentence is that if that story did catch him, he would or might keep watching.The second substantiation, perhaps on second thought, asserts that if the previous program hadended, he would turn off the TV.

The disjuncture between the two substantiations served as one occasion to reconsider otherresponses by this and other respondents. One conclusion of the study, as suggested also by otherresearch, was that the respondents aimed to project an image – to offer an ‘implication’ – of them-selves as rational citizens, not least in relation to a ‘serious’ genre such as news. Presumably, they‘ought to’ select news specifically as part of their media diet. In the study, the theme of watchingnews as a way of participating in political democracy tied in with other themes, particularly thefeeling of being a legitimated member of an imagined political community (Anderson 1991). Thisgeneralized use of news was opposed, in different ways, to any concrete, instrumental value ofnews as a political resource.

At the level of discourse, this brief quotation also contains the rudiments of a narrative regard-ing family life, specifically the respondent’s sister, with implications for gender-specific media use.The narrative is constituted, in part, at the level of the speech acts:

cont.

technique which facilitates enumeration as wellas contextual interpretation.

It is a characteristic feature of discourse andother language studies since the 1970s that theyhave emphasized the understanding of commu-nication as social action. A case in point is thecommunication or ‘language work’ of research.In his influential functional linguistics, Halliday(1973, 1978) identified three aspects of thesocial uses of language:�

Speech actsUtterances serve as means of representing, andof interacting with, social reality. All utterancesmay be categorized as a type of speech act – aperformance in a particular context for a par-ticular purpose.� Orders will be negotiatedwithin a media production team; journalisticsources may be offered a promise of anonymityin return for a statement. In addition to such

types of speech acts, three elements in particu-lar lend themselves to discourse analysis as indi-cators of how people (including a researcher inthe empirical field) interact:

1 personal pronouns (I, you, one, they, etc.) –especially in reference to themselves, people usepronouns to signal different degrees of distancefrom a topic or opinion, and of solidarity withothers;

2 impersonal grammar – through the use ofpassive sentences and other linguistic forms,both media and people are often less thanprecise regarding ‘who does what,’ and some-times these structures imply a worldview wherethings just happen;�

3 metaphors – alone, in pairs, or in severalvarieties, metaphors can serve as an organizingprinciple behind a description or argument and,hence, as a key to a respondent’s conception ofparticular phenomena.


• Personal pronouns – a characteristic feature is the consistent and insistent use of ‘I,’apparently a self-assured and self-aware position, which the respondent applies equally toliving at home and to his current living conditions.

• Impersonal grammar – in reference to his sister, the respondent seems to suggest that shedoes not actively use the media, but merely ‘has on’ a semi-autonomous technology, what isderogatorily termed ‘some appliance.’ (It is worth adding, however, that, also in his owncase, the first news story is something that must ‘catch him,’ not vice versa.)

• Metaphor – is not a strong feature in the example. Still, metaphors serve to emphasize therespondent’s distanciating evaluations (‘deep in me,’ ‘have a real thing’), and the reference tomedia as ‘appliance’ (and to a program ‘catching’ a person) similarly helps to establish thedistance between the speaker and particular uses of the medium.

At the level of interaction, the respondent re-emphasizes his personal position:

• Turn-taking – the specific design gave respondents the opportunity to elaborate at length,which this respondent did; this was a short reply. Here, he responded to the verbatimquestion, ‘If you were watching the program right before the news program, would thatmake you more inclined to watch the news?’ It is worth noticing that the relativity of‘inclined to watch’ is canceled by the respondent, who instead frames a clear-cut choice assignaled initially by ‘no.’

• Semantic networks – a longer sequence, preferably a full interview or other interaction, isneeded to assess the interrelations between concepts. Still, the lexical choices, includingvarious emphatic formulations (‘a real thing,’ ‘just always,’ ‘really,’ ‘something deep in me’),are in line with the respondent’s profiling of himself as a rational citizen.

cont.

� functional linguistics – Chapter 7, p. 105

� speech acts – Chapter 2, p. 38 � statements without an agent – Chapter 7, p. 105

InteractionThe interaction between two or more peoplewhich makes up most empirical media researchis important for two different purposes. First,the gradual development of viewpoints in, forinstance, focus groups lends itself to closeanalysis, which may trace important conceptualdistinctions and continuities. Second, the re-searcher’s own role in the interaction calls forquality control – what is also considered underthe intersubjectivity or reliability of research.�In qualitative research, such analyses can assesswhether studies, in fact, explore the respon-dents’ perspective. For both purposes, twoaspects of discourse provide indicators:

1 turn-taking – a description of the structure,order, and length of interventions by respon-dents and researcher(s) (Sacks et al. 1974);

2 semantic networks – an examination ofcentral terms and concepts, as introduced andredeveloped by both researcher and respon-dent(s) (e.g., Corley and Kaufer 1993).

DiscoursesDefined as meaningful wholes with argument-ative, narrative, and other purposes, discoursesrepresent the largest unit of qualitative dataanalysis. In addition to the rhetorical and liter-ary traditions, a source of inspiration for dis-course analysis in this regard has been the studyof informal argument (Toulmin 1958). Toexemplify, the actant model,� which was devel-oped from folktales, has proven applicable tohow people narrate and argue about themselves– the ‘Story of Me’ (Jensen 1995: 137).

The analysis in Box 14.1 illustrates the prin-ciples of all three levels of discourse analysiswith reference to an argumentative interviewdiscourse. To clarify, the purpose is not to con-duct a comprehensive formal analysis of inter-views and other data as language – the analysisof language (and other signs) here is an auxil-iary operation to better understand researchevidence as it arises from communication. It is

important, then, to distinguish between dis-course as ‘structure’ and as ‘evidence’ (Gee et al.1992: 229–230).

Nevertheless, language and its componentsare main vehicles of evidence in qualitativeresearch, and require the attention of qualita-tive researchers. Language is comparable, inquantitative research, to mathematical symbolsand procedures. Numbers have a different rel-evance for quantitative media researchers thanthey do for mathematicians, but still requiretheir attention.

In response to a final concern – that dis-course analysis requires specific linguistictraining in order for media researchers to trans-fer it to a qualitative project – this may be saidto raise a rather more general issue for muchinterdisciplinary research. Interdisciplinary ap-proaches are premised on the assumption that the transfer and integration of methods or, alternatively, collaboration and the use ofexpert consultants, is feasible. Quantitativeprojects routinely consult with statisticians; it islikely that qualitative projects will increas-ingly involve discourse-analytical consultants,or include their competences in interdisciplinarygroups. Discourse analysis presents itself as a candidate for a ‘statistics’ or systematics ofqualitative research.

To sum up, the three variants of qualitativedata analysis are displayed and compared inFigure 14.3:

1 Thematic coding represents an attempt toidentify, compare, and contrast meaning ele-ments, as they emerge from and recur in severaldifferent contexts. What distinguishes thematiccoding from much quantitative content analy-sis is the emphasis on defining each of these ele-ments in relation to their context.

2 Grounded theory seeks to condense or com-press its central categories of meaning throughthe research process, partly through iterativesampling, partly through repeated analyses.

3 Discourse analysis proposes a combinationof coding, primarily for heuristic purposes,with in-depth linguistic analysis of selectedmeaning elements.

Data analysis 251

1111234567891012341115678920123456789301234567894012345678950111

� reliability and other quality control – Chapter 15, p. 266

� the actant model – Chapter 8, p. 127

discourse asstructure andas evidence

a ‘statistics’ ofqualitativeresearch

Computer interfaces

Over the past decade, the ambition of develop-ing more systematic approaches to qualitativeresearch has been evident also in an increasedreliance on computers. Dedicated software fordata analysis and for the administration of theresearch process has become available (Weitz-man 2000; Weitzman and Miles 1995), and itspotentials (and problems) in research designand theory development are being explored(Fielding and Lee 1998). The main practicalvalue so far has been more efficient means ofheuristic coding, in addition to facilities forretrieving evidence in different configurations,including tabular and graphic formats. Increas-ingly, qualitative software also supports theanalysis of sound, still, and moving images.Such resources are likely to encourage a greateruse of nonverbal data forms.

One prospect is that ‘hypertext’ systems, invarious forms, will support the research processas a whole – from theoretical exploration, lit-erature search, and networking, through datacollection, annotation, and analysis, to publi-cation, debate, and revision. Not least, the com-puter allows for extensive data sharing andcollaborative qualitative projects, which havebeen comparatively rare in the past. Thesharing of evidence may involve not only col-leagues but respondents, and, in time, segmentsof the public. For one thing, a computerizedresearch process facilitates ‘member checks’(Lindlof 1995: 168) – conferring with respon-dents about the interpretation of data (althoughconclusions remain the researcher’s right andresponsibility). For another, computer-mediatedcommunication, at least in principle, providesa link between qualitative research and public


DATA SETS– as analyzedby differentprocedures

Thematiccoding

Groundedtheory

Discourseanalysis

TIME

Figure 14.3 Three variants of qualitative data analysis

data sharing

memberchecks

debate, which has been more of an arena forquantitative research traditions in the past.

Concretely, the bulk of many qualitativereports, including close analyses of transcripts,has sometimes served as an argument againsttheir publication in the major journals which,for better or worse, define a field. More differ-entiated publication formats can assign fulldigital documentation to (what is now knownas) an Internet site or to another medium. Atthe same time, computer-mediated publicationmay provide one more interface between qual-itative and quantitative research traditions. It

seems likely that the meta-medium (Kay andGoldberg 1999 [1977]), which both main-streams increasingly rely on, could support thedevelopment also of more common analyticalterms and concepts.

Fifty years ago, Lazarsfeld and Barton(1951: 155) provided a reassuring argument,and an agenda of complementarity, which ispursued in the next chapter: ‘There is a directline of logical continuity from qualitative clas-sification to the most rigorous form of mea-surement.’

Data analysis 253

1111234567891012341115678920123456789301234567894012345678950111

INTRODUCTION

The legacy for media and communicationresearch from a century of sometimes intenseconflict within the general theory of science hasbeen a set of conceptual dichotomies that usedto divide the field into two main camps withsome additional internal fronts and alliances.Even while studies of scientific practice haveshown that the reality of laboratories andlibraries may rarely conform to textbookmodels (e.g., Hacking 1983; Latour 1987),these conceptions have had real and largelycounterproductive consequences for how dif-ferent research traditions have understood both themselves and their ‘others.’ The percep-tion of fundamental difference has generatedvarious kinds of response – from ‘imperialism,’seeking to subordinate or delegitimate otherapproaches, to ‘apartheid,’ protecting one’sown worldview through insulation from those of others (see Jensen 1995: 141–145).This chapter presents an overview of thecurrent, ongoing convergence in the practice of research, and provides examples of the

complementarity of different types of empiricalmedia studies.

Departing from the received conceptualdichotomies regarding qualitative and quant-itative research, the chapter first takes themodest position that the two mainstreams havedifferent strengths, and might proceed in par-allel. To begin a more detailed comparison, thechapter distinguishes several forms and levelsof ‘analysis,’ which also involve differentnotions of ‘method’ and ‘methodology.’ Next,the different types of analysis are related to thetwo classic forms of inference – induction anddeduction – to which is added a third – abduc-tion. The presence of each in qualitative as wellas quantitative media studies is exemplified,and their respective requirements lead into areconsideration of criteria of validity, reliabil-ity, and other key concepts in empirical work.These various comparisons suggest the promiseof a realist model of science, which may accom-modate diverse research traditions. Finally, thechapter presents three concrete ways of com-bining qualitative and quantitative methodolo-gies with a view to further research.

The complementarity of qualitativeand quantitative methodologies inmedia and communication researchKlaus Bruhn Jensen

15

• an overview of the two main paradigms which inform media and communication research• a presentation of six levels of analysis that are shared across different research traditions• a comparison and exemplification of three forms of inference from the philosophy of science• a reassessment of reliability, validity, generalization, and probability in empirical media studies• a discussion of realism as a framework for convergence.

neither‘imperialism’

nor ‘apartheid’

FROM CONFLICT TO COMPLEMENTARITY

Two paradigms

Figure 15.1 displays a list of interrelated con-ceptual dichotomies, as still received in muchmedia research (see further Jensen 1995: ch. 9).The two columns qualify as ‘paradigms’ (Kuhn1970), defined as particular configurations of ontological, epistemological, and method-ological assumptions about reality and how tostudy it (Guba and Lincoln 1994: 108). In thepresent context, it is their methodologicalassumptions, and their background in thetheory of science, that are the focus of interest.(For an overview on paradigms, see Conant andHaugeland 2000.)

The differences appear most concretely atthe level of methodology. It is commonly recog-nized that quantitative research instruments are

particularly suited to establish the recurrence ofevents or objects (e.g., the expression of a par-ticular opinion, or a specific content feature).Qualitative approaches, in their turn, explorethe singular occurrence of meaningful pheno-mena, but with reference to their full context(e.g., a film narrative or everyday media uses ina household). Defined thus as an indivisiblewhole, human experience calls for an exegesis –a reiterated and continuous interpretation ofmeaningful elements whose context may beredefined as the interpretation proceeds. In con-trast, the whole of human experience may alsobe defined as the sum of its parts, divisible atleast in operational terms and, hence, manipu-lable in an experiment, whose findings areexpressed in measurements. It is in this latterregard that the quantitative conception ofmeaning is sometimes referred to as a product –

From conflict to complementarity 255

1111234567891011112341115678920123456789301234567894012345678950111

paradigms

Methodology

Recurrence

Experiment

Measurement

Product

Theory of science

Gesetzeswissenschaften

Naturwissenschaften

Nomothetic

Erklären

External

Information

Epistemology

Nature

Causes

Objects

Occurrence

Experience

Exegesis

Process

Ereigniswissenchaften

Geisteswissenschaften

Idiographic

Verstehen

Internal

Meaning

History

Intentions

Subjects

Figure 15.1 Two paradigms of social and cultural research

methodology

a delimited vehicle that may be conferred with-out ambiguity to other meaning products inother contexts. Qualitative methodologies, incomparison, tend to approach meaning as aprocess which is tied to its shifting contexts.

The historical background to these method-ological variants was attempts within the phil-osophy or theory of science to stipulate pro-cedures for the study of society and culture.Such procedures took on renewed importancefollowing the development of social sciencesfrom the late 1800s. One famous statement of purpose was Weber’s (1964: 88) definition ofsociology as ‘a science which attempts the inter-pretive understanding of social action in orderthereby to arrive at causal explanation of itscourse and effects,’ referring specifically to bothinterpretations and causes. Many contemporarydebates were anticipated in the Methodenstreit(struggle over methods) of the period, centeredin German thought, which confronted histori-cist and naturalist notions of the ‘facts’ of sociallife (for an overview see Hammersley 1995).

On the one hand, one may search for oneor more laws (Gesetz), as previously establishedin sciences of nature (Naturwissenschaften).One thus takes a generalizing or nomotheticattitude and tries to explain (erklären) society.On the other hand, one may study one or moresingular events (Ereignis), as associated withphilosophy, aesthetics, and other disciplines ofthe ‘human spirit’ (Geist). Here, one takes anindividualizing or idiographic attitude with theaim of understanding (verstehen) society. Withthe advent of information and communicationtheory (which has restated classic issues of phil-osophy), several related distinctions emerged.Media and communication studies have tendedto take either an external perspective on inform-ation as a technical, neutral carrier, or an internal perspective on meaning as an alwaysinterpreted and interested construct.

These ingredients of the theory of sciencehave mediated between the specific answerssuggested by empirical research, and those epis-temological questions which appear ‘eternal,’ inthe sense that they have refused to go away formillennia of documented research. As such,they have returned, sometimes with a ven-geance, in new scientific fields.

The media field has faced general questionsof whether technologies and other aspects ofmatter or nature may determine particularforms of communication within society andhistory. A recurring issue has been whetherexplanatory models from the natural sciencesmight be transferred, more or less directly, to thestudy of society and culture. Media practition-ers, legislators, as well as publics all bring inten-tions into mediated communication. Hence, itbecomes necessary to ask specifically how –through which social structures and culturalprocesses – their motivated actions are coordi-nated, and become causes.

In the end, media and communicationresearch is faced with the fundamentaldichotomy of subject and object. This relation-ship is both a substantial feature of communi-cation and a condition of research. In com-municative interaction, the key ‘object’ of inter-est is normally another subject or social group,as suggested by Mead’s (1934) concept of thesignificant other. In empirical studies, the pur-pose is to describe, interpret, and explain suchinteraction, as seen inevitably from the re-searcher’s perspective, but through reference tothe categories of understanding that make upthe participating subjects’ perspective.

Code-based and case-based analysis

A first step in bridging the apparent abyssbetween the two paradigms is to ask whethersome division of labor can be sustained. At leastin principle, most contemporary research wouldrecognize that its choice of methods mustdepend on what aspect of mediated communi-cation is being examined, and on the purpose ofstudy – the ‘how’ of research depends on its‘what’ and ‘why’ (Kvale 1987). Few researcherswould want to argue (or admit) that theirconceptualizations and designs are driven bycertain methods that are preferred as such.Certainly, different approaches would seem tobe required to account for editorial decision-making processes or for the employment pat-terns of women and men in news media; for thestructure of metaphor in newspaper headlines as opposed to the coverage of a particular eventin different newspapers; and for audience

Methodologies in media and communication research256

the philosophyor theory of

science

Methodenstreit

epistemology

the significantother

decodings of, versus their exposure to, particu-lar items of information. More generally, par-ticular stages, for example, of media receptionand impact, are natural candidates for study byparticular methodologies.

A justification for matching particular re-search questions with particular methodologiesis found at the level of concrete analyticalprocedures. The two paradigms may be dis-tinguished, in broad terms, by their reliance oneither case-based or code-based operations(Fielding and Lee 1998: 27). On the one hand,code-based analysis assumes that, for instance,a survey response may be assigned unequivo-cally to one category for analytical purposes. Aclassic example is the opinion polls leading upto a general election. The polls owe much oftheir predictive value to being clearly defined,not only because the options are mutuallyexclusive (one designated party, abstention, orundecided), but also because the act of voting(and of responding to a poll) is a familiar prac-tice and a cultural convention.

On the other hand, case-based analysisseeks, at least initially, to avoid reduction of thecomplexity of data, for instance, from a depthinterview. Instead, the process of interactionwith respondents may be seen to carry over intothe process of analysis. The categories of under-standing and meaning may be identified, rede-fined, and clarified throughout the researchprocess as a whole. To exemplify, a case-basedanalysis might explore conceptions of what isa ‘political’ issue in the media and how, accord-ing to respondents, it relates to institutionalizedpolitical activities.

In procedural terms, then, a code-basedanalysis relies on predefined categories whichboth decontextualize and disambiguate the unitsof meaning. A case-based analysis allows itscategories to be informed and modified by thecontext in an iterative fashion. It also remainsopen to ambiguities in the delimitation andinterrelation of the units of meaning throughoutthe analytical process.

One particular issue which has continued todivide the paradigms is whether code-basedanalysis could, and should, gradually replacecase-based analysis. This might be the idealwithin single studies, but also, more controver-

sially, for entire disciplines as they mature. Thequalitative, case-oriented emphasis of earlysocial science has been explained, in part, by itspreliminary search for ‘global, overall perspec-tives’ (Jankowski and Wester 1991: 46), overand above its roots in philosophy and otherhumanities.

In the post-1945 period, code-based analy-sis became established as the scientific norm. Bythe same token, qualitative research was largelyassigned the role of performing ‘pilot studies’that would pre-test codes and, to a degree,develop theory. One question which has re-mained controversial, then, is whether case-based, qualitative analysis may be said to havean independent explanatory value. To clarifythe concrete disagreements, and to assess theprospects for convergence, it is necessary, next,to differentiate the several stages and levels ofanalysis which qualitative and quantitativestudies share.

SIX LEVELS OF ANALYSIS

Figure 15.2 distinguishes six levels of research,as associated with different stages of planning,conducting, documenting, and interpreting anempirical project. Here, the levels are describedin terms of their ‘discourses’ or symbolic instru-ments – the varied social uses of language,mathematical symbols, graphical representa-tions, and other instruments which enter intothe practice of research:

1 The empirical object of analysis will typicallybe everyday and institutional discourses arisingfrom some media-related social interaction. The particular materials range widely, fromorganizational memoranda and media texts totest responses in an experimental setting. Whilesome materials are ‘found’ (e.g., radio programsfrom a sound archive), and others ‘made’ (e.g.,observation of a production site), each consti-tutes a possible source of evidence regarding aparticular purpose of inquiry.

2 Also in the case of evidence that is ‘found,’the second level of data collection methods –from content sampling frames to interviewguides – serves to demarcate and document a

Six levels of analysis 257

1111234567891012341115678920123456789301234567894012345678950111

pilot studies

particular portion of reality. The evidence gath-ered will next be examined from a particulartheoretical perspective, again according to thepurpose of inquiry.

3 Data analysis methods entail diverse opera-tions of categorizing, segmenting, and inter-preting the evidence or dataset. This level ofanalysis typically involves comparisons. In abasic sense, all analysis is comparison of units;in a wider sense, comparative analysis, for in-stance, of contents and audience responses, is a main ingredient of many media studies. Inaddition, empirical studies commonly include a ‘meta-analytical’ component in the form ofstatistical tests or an ‘audit trail’ of the steps of a qualitative analysis.�

4 It is at the level of methodology, defined as a theoretically informed plan of action in rela-tion to an empirical field, that the distinctionbetween qualitative and quantitative researchbecomes most apparent. Even though, for example, a quantitative and a qualitative inter-

view study share certain techniques, in each casethe full set of techniques amounts to a distinc-tive mapping of theoretical concepts onto empir-ical phenomena. As elaborated below, it is thespecific linking of a necessarily selective empiri-cal ‘microcosm’ with a theoretical ‘macrocosm’which, above all, distinguishes quantitative andqualitative research as traditions and paradigms.

5 A theoretical framework of interrelated con-cepts lends relevance and meaning to the con-figuration of empirical data, or findings. Theorymay be thought of as a ‘frame,’ as referred toin several traditions (Eysenck and Keane 1995;Goffman 1974), which focuses attention andallows for certain interpretations, while dis-couraging others. Crucially, quantitative andqualitative approaches may be subsumed underthe same theoretical framework. It is method-ology, not theory, that distinguishes qualitativeand quantitative research.

6 Theoretical frameworks are ‘substantive,’ in the sense that they account for a particulardomain of reality, here communication and cul-ture. Theoretical choices, further, are supportedby meta-theoretical or epistemological argu-ments and assumptions. The epistemologicallevel of analysis provides preliminary definitionsof the ‘object’ of study, as well as justificationsconcerning the nature of the ‘analysis.’

Debates across the qualitative–quantitativedivide have generally focused on these indi-vidual levels. In particular, the levels of datacollection and analysis is where the two main-streams have parted company. In fact, the rela-tionship between the paradigms is betterunderstood if one notes in detail how eachparadigm joins – and separates – these levels ofanalysis.

On the one hand, quantitative research tendsto assume that a separation of the moments ofconceptualization, design, data collection, dataanalysis, and interpretation is both possible anddesirable. This practice derives from a distinc-tion between ‘the context of discovery’ and ‘thecontext of justification,’ as originally associatedwith logical positivism (for an overview seePassmore 1972). The underlying argument isthat scientific research is characterized, above


Empirical object of analysis

Data collection methods

Data analysis methods

Methodology

Theoretical framework

Epistemology

Figure 15.2 Six levels of empirical research

� meta-analysis as quality control, including statisticaltests and audit trails – Chapter 14, p. 246

methodology,not theory,distinguishesqualitative andquantitativeresearch

contexts ofdiscovery andjustification

all, by its procedures for testing and justifyingbeliefs, and that the process of discovering andhypothesizing beliefs is not part of the scientificmethod proper. The resulting emphasis on for-malization and operationalization is reflected,for example, in the structure of the standardresearch article with its sections on ‘previousresearch,’ ‘hypotheses,’ ‘method,’ ‘analysis,’ and‘discussion.’�

On the other hand, qualitative researchholds that at least certain phenomena call fora research process that moves liberally betweenall the analytical levels in order to articulateand elaborate analytical categories. This gen-eral procedure of ‘in-depth’ analysis is evidentboth in data collection and analysis. Certainsteps of analysis are repeated, for example,through ‘snowball sampling’ or in developingthematic categories for the textual analysis ofan interview. Here, the minimal act of analysisis a constitutive part of a sequence. As a whole,the sequence reconfigures empirical data, anddevelops theory to account for them, ratherthan matching predefined categories with dataelements. It is these characteristics which makefor the sort of flexible data collection andanalysis that is associated with qualitativeresearch, and it is this same absence of hard-and-fast techniques that has motivated recentdevelopments toward more systematic andtransparent forms of qualitative analysis.�

Apart from the very early stages and instru-ments of research, an additional focus of thedebates between qualitative and quantitativepositions has been the very final stages of theresearch process – its epistemology and overalllogic. Reference is traditionally made to twoforms of logical inference – induction anddeduction. Recent theory of science, however,has recovered a third type, which helps to spell out some differences as well as simi-larities between qualitative and quantitativemethodologies.

THREE FORMS OF INFERENCE

Induction, deduction, and abduction

Standard accounts of the theory of science stillassume that research infers either from ageneral principle or law to individual instances(deduction), or from the examination of severalinstances to a law (induction). The require-ments of each type have been at the core of thedevelopment of natural sciences, and of moderntheory of science, since the eighteenth century.Their relevance for the humanities, the socialsciences, as well as interdisciplinary fields hasbeen debated particularly since the late nine-teenth century (for an overview see Pitt 1988).In comparison, the third form of inference –abduction – has rarely been considered explic-itly as a model of scientific reasoning. First ofall, then, each of the types should be described.

While Aristotle had identified abduction asa type of inference (Blaikie 1993; Hanson1958), it was reintroduced into modern phil-osophy by Charles Sanders Peirce in an 1878article and related to the other two main types.The basic idea is that there are three compo-nents to an inference – a rule which, whenapplied to a single case, produces a conclusionor result. These components yield three poss-ible combinations:

DEDUCTIONRule. All the beans from this bag arewhite.Case. These beans are from this bag.Result. These beans are white.

INDUCTIONCase. These beans are from this bag.Result. These beans are white.Rule. All the beans from this bag arewhite.

[ABDUCTIONResult. These beans are white.Rule. All the beans from this bag arewhite.]Case. These beans are from this bag.

(Peirce 1986: 325–326)

Formally, only the deduction is a valid infer-ence. Here, given the meaning of the constituent

Three forms of inference 259

111123456789101234111567892011123456789301111234567894012345678950111

� conventions of publication – Chapter 16, p. 274

� systematic qualitative data analysis – Chapter 14, p. 245

terms, the rule can be applied without anyuncertainty to the case, so that the result followsas a matter of course. In the induction, theimplication is that, if one examines a sufficientnumber of beans (cases), one may be willing toconclude that they are all white. This way ofreasoning appears commonsensical, and entersinto both everyday life and sciences. The pointof the abduction, finally, is that it introduces arule which may explain why one encountersparticular (surprising) facts in a particularcontext.

The bean example is, of course, trivial. Inother cases, the newly devised rule may repre-sent an exceptionally bright idea, as in SherlockHolmes’s solution of crime mysteries (Sebeokand Umiker-Sebeok 1983). Scientific discoverycan also be seen to rely on abduction.

The three forms of inference are rarelyfound in a pure form in any given empiricalstudy. More often, different combinations willserve the purpose of inquiry, and it may beargued that an aspect of each type is requiredto produce new insight. An example is the pro-totypical form of social-scientific study. Such astudy departs from a relatively specific hypoth-esis that has been derived from a more generalpremise (deduction), and which is then testedagainst a large number of concrete instances(induction). The outcome of the data analysismay be a pattern of findings that is only partlyin accordance with the hypothesis, and whichgives rise to the formulation of a new rule(abduction) to be investigated in further re-search. (The original premise of the study mightin itself have been the outcome of a bright ideaor bold conjecture – abduction.)

One advantage of such a combinatorialunderstanding of scientific inference is that itleaves open the question of whether researchprojects actually conform to the models ofeither logic or methods textbooks. A closerexamination of how researchers, in fact, inter-pret and infer from their data can serve to makeexplicit some hidden premises. It can also helpto identify more options for the design of futureresearch.

An inductive heritage

Induction has a long heritage, both in thehistory of science and in human evolution. Inthe latter respect, one may assume that thehuman capacity for generalizing and abstract-ing from single events has been a key factor innatural evolution, being an instrument of adap-tation and, ultimately, survival (Megarry 1995).(Indeed, the many cultural forms of expressionof humans may be described, functionally, interms of induction that serves to accumulate,externalize, and communicate experience for a range of practical purposes.) In addition the common-sense or lay theories (Furnham1988) which guide people through a day ofpractical activities have important ingredientsof induction.�

Historically, induction has been a centralproblem for philosophers and empirical re-searchers alike. While David Hume had noted,in the mid-1700s, that an induction from‘some’ to ‘all’ can never, strictly speaking, belogically valid, an inductive ideal of scienceremained attractive throughout the nineteenthcentury, as associated, for example, with JohnStuart Mill’s A System of Logic (1843). Thiswork sought to detail how inferences could beaccumulated in various scientific fields, depend-ing on their specific subject matter (Mill 1973–1974). In the twentieth century, an inductiveideal of science rose to new prominence, andthen fell definitively, in the shape of logical pos-itivism. Nevertheless, references to ‘positivism’(especially critical ones) still abound also inmedia research, suggesting the need to clarifythe basic issues and positions.

Taking its general cue from Mill’s contem-porary, Auguste Comte, and his call for a ‘pos-itive philosophy’ that would be non-speculativeand applicable to real human concerns, logicalpositivism developed into an influential schoolof thought during the interwar years. Much of the specific, technical inspiration came fromthe linguistic and formal turn of philosophy at the time. The linguistic turn started from the assumption that the structure of sen-tences corresponds to the structure of facts in


induction inculture andevolution

� lay theories – Chapter 16, p. 275

logicalpositivism

reality.� Moreover, logical positivism upheldan absolute distinction not only between factsand values, but between empirical observationsand theoretical conceptions of reality. Accord-ingly, any meaningful statement about theworld is either elementary in itself (reducible tosense impressions in a given space and time), ormay be decomposed into such elementarypropositions which are either true or false. It isfrom this base of atomistic facts that morecomplex regularities or laws may be inferred byverifying that a larger set of propositions cor-responds to the way the world is.

Within such a reductionist understanding ofknowledge, most of the topics of both social-scientific and humanistic research fall outsidethe realm of science. One response to positivism– the ‘apartheid’ option – has been to contrastthis (assertedly) natural-scientific understand-ing of research with an interpretive counterpart(Figure 15.1.) The resulting dichotomies haveaffected the social sciences throughout theirhistory, and have reappeared in media studies,as noted above.

A second, ‘imperialistic’ response has beenan (often implicit) attempt to transfer some ofthe ideals of logical positivism to research onculture and society. Hammersley (1989: 17)identified three features of such an unacknow-ledged positivism, as it continues to operate inmuch social science:

1 the view of science as the identification ofuniversal laws, or at least context-indepen-dent regularities;

2 the grounding of science in elementary sen-sations;

3 the requirement that all scientific researchproceed according to the same methodolog-ical principles.

This last requirement in particular still tends to divide research into different ‘camps,’ suchas the two paradigms of media studies, and toproduce allegations of (quasi-)positivism. Infact, both positivism and inductivism, if under-stood as epistemological programs, may betreated as positions of the past. This is not

to deny that the two positions, to a degree, still inform particular features of research practice.

The following sections describe the twomain epistemological positions which can beseen to have replaced positivism and induc-tivism in current media and communicationresearch. First, however, two varieties of con-temporary research should be noted for theirmore or less explicit commitment to induction.First, grounded theory� became influentialespecially among qualitative researchers for itsinsistence on generating explanatory categoriesfrom the field itself and its informants – induc-ing theory from empirical data. Whereas thisposition may blind researchers to the theoreti-cal frameworks which they inevitably bringwith them into the field, it has served animportant purpose by attracting renewed atten-tion to the practical, lived categories of under-standing with which people engage media andother social interaction.

The second example of an inductive orienta-tion is descriptive studies, often with an appliedor administrative purpose.� Among the mainexamples are marketing studies of media audi-ences and evaluation research which supportsgovernment policy decisions. Although the aimhere is normally not to state or test partic-ular theories, nevertheless the findings are com-monly seen as offering a general account, forexample, of media infrastructures and theirimpact on society. At least, companies and stateagencies are willing to base investments andlegislation on the descriptions. In this connec-tion, Wober (1981) has called for a distinctionbetween audience ‘research’ with some explan-atory or interpretive ambition, and audience‘measurement,’ which provides baseline figures.An implication for readers of descriptiveresearch reports is that their findings are alsotheory-loaded (Hanson 1958), so that theclaims of such research must be weighed equally– or more – carefully against its more or lessimplicit theory. This is all the more import-ant, since the majority of all media studies are


1111234567891012341115678920123456789301234567894012345678950111 � the linguistic turn – Chapter 2, p. 26

� grounded theory – Chapter 14, p. 247

� descriptive and administrative research – Chapter 16,p. 281

research vs.measurement

probably of the administrative and ambiguouslyinductive variety.

A deductive mainstream

The quantitative mainstream of internationalmedia and communication research is normallydescribed as ‘hypothetico-deductive.’ It pro-poses to test hypotheses which have beendeduced from some general ‘law.’ In a first step,it is deduction which ensures that a hypothesisis neither logically inconsistent nor tautological– which would make it irrelevant for empiricalresearch. If, next, a relevant hypothesis may beseen to contradict or perhaps specify anaccepted law, it calls for further study. Again,it is deduction (from ‘all’ to ‘some’) whichserves to predict what the researcher will findunder certain specified empirical circumstances.If, finally, the findings correspond to the pre-diction, the hypothesis is confirmed, and can beadmitted into the body of accepted and cumu-lated theories of the field.

Importantly, however, confirmation doesnot equal ‘verification’ in the stronger senseassociated, not least, with logical positivism.The hypothetico-deductive position, as elabo-rated especially by Popper (Popper 1972[1934]; see also Hempel and Oppenheim 1988[1948]), instead assumes that scientists mustseek to falsify their hypotheses. Only if falsifi-cation fails is one justified in still holding thehypothesis, and only preliminarily. Furtherstudies, by oneself or by others, may end upfalsifying it after all (which, in effect, admitsinductivism through the back door in a multi-step research process).

What might appear to be a ‘philosophical’nuance nevertheless has important conse-quences, both for what is accepted as know-ledge in a field, and for the procedures of science– also regarding society and culture. Becausesocieties and cultures are ‘open’ systems com-pared to the more closed and, hence, predictablesystems studied by the natural sciences, mostsocial and cultural research of the hypothetico-deductive variety cannot unequivocally falsifyor verify a hypothesis. Instead, studies arebacked by measures of statistical probability –which accounts for their ‘quantitative’ nature.

In formal terms, then, a prototypical quan-titative study will pose a ‘null hypothesis,’which suggests that any relations between thevariables, as anticipated by the theory anddocumented by the data, are due to chance fluc-tuations or random error.� If this is the case,as defined by a conventionally agreed level of probability, a falsification of the originalhypothesis has occurred. Thus, the quantitativeresearch process operates on a principle ofdeduction, even if the ‘laws’ in questions areascertained in a stochastic rather than a deter-minist sense (Hempel and Oppenheim 1988[1948]: 13–18).�

The agenda-setting hypothesisA classic contribution to media studies – agenda-setting research – exemplifies the general deduc-tive logic. McCombs and Shaw (1972) departedfrom previous findings regarding politicalcommunication, suggesting that, ‘although theevidence that mass media deeply change atti-tudes in a campaign is far from conclusive, theevidence is much stronger that voters learn fromthe immense quantity of information availableduring each campaign’ (p. 176). While thusrecognizing previously established regularities in the audience response, the authors went on,first, to deduce a conceptual distinction between‘attitudes’ and ‘agendas,’ and, second, to opera-tionalize this distinction in an empirical com-parison of news content and voter statements.Their hypothesis stated that ‘the mass media set the agenda for each political campaign,influencing the salience of attitudes toward thepolitical issues’ (p. 177).

The study, accordingly, matched ‘what . . .voters said were key issues of the campaignwith the actual content of the mass media usedby them during the campaign’ (p. 177). Theauthors further deduced two particular condi-tions under which the hypothesis might beexamined empirically: only voters who wereundecided on who to vote for, and hence mightbe more open to campaign information, enteredinto the sample interviewed, and the respon-


hypothetico-deductiveresearch

falsificationandverification

� null hypothesis – Chapter 13, p. 232

� stochastic and other models of meaning, Chapter 2,p. 30

dents were sampled randomly from lists of reg-istered voters in a particular community inorder to limit other sources of variation, forexample, regional differences in media cover-age. (Yet, following a pre-test, major nationalsources such as television network news, TheNew York Times, Time, and Newsweek wereincluded in the content sample.)

The resulting datasets consisted in respon-dents’ answers regarding ‘major problems asthey saw them’ (p. 178), and in news as wellas editorial comments during a specified periodoverlapping with the interview period. Each ofthese datasets was coded into predefined cate-gories concerning both political issues andother aspects of a campaign. As is commonlythe case in content studies, the intercoder reli-ability� of the procedure was tested, and layabove .90, which bears witness to well-definedand operational categories. In sum, the appli-cation of these various analytical categoriesamounted to a mapping of conceptually de-duced distinctions onto instances of politicalinformation in a particular empirical field, asoffered by media and taken by voters.

Two findings, in particular, illustrate thelogic. First, the design aimed to establish causality, and found that the media had ‘exerteda considerable impact’ (p. 180) on the respon-dents’ perceptions of the political issues. Thecoding of content had distinguished between‘major’ and ‘minor’ coverage of topics (referringto their prominence in terms of position andtime or space allowed). In both instances, theanalysis found very strong correlations betweenmedia emphasis and voters’ judgments (+.967and +.979).

Second, in order to determine whethervoters might be attending to, and reproducing,the agenda that their preferred candidates pre-sented in the media, a further analysis wasmade of those respondents who had a prefer-ence for one candidate (without being finallycommitted). Here, a comparison was madebetween these respondents’ references to theissues associated with a particular party or can-didate, and their references to issues in all thenews. Both for major and minor issues, the

findings suggested that ‘the voters attend rea-sonably well to all the news, regardless ofwhich candidate or party issue is stressed’ (p. 182). The fact that ‘the judgments of votersseem to reflect the composite of the mass mediacoverage’ (p. 181) again lends support to theagenda-setting hypothesis.

In their discussion of the findings overall,McCombs and Shaw (1972) were careful toqualify their conclusions regarding the originalhypothesis. Most importantly, they acknow-ledged that the correlations reported in theirstudy do not prove the hypothesis. However,‘the evidence is in line with the conditions that must exist if agenda-setting by the massmedia does occur’ (p. 184). Put differently, thecarefully deduced design failed to falsify thehypothesis. In addition, the agenda-settinghypothesis presents itself as a more justifiedalternative than, for example, a theory concern-ing selective perception, which would have beensupported if voters had been found to attendparticularly to their preferred candidates. Theauthors also described the study as ‘a first test . . . at a broad societal level’ (p. 184), and called,among other things, for further research on asocial-psychological level regarding individualattitudes and uses of media.

An additional lesson from the discussion inMcCombs and Shaw (1972) is that the weigh-ing of competing hypotheses and interpreta-tions ultimately takes place at a theoreticallevel, not at the level of measurements or otheranalytical procedures (Figure 15.2). The dis-cussion of ‘probability,’ below, reconsiders thetextbook maxim (often violated) that correla-tion does not equal causation. In this example,the correlations between media coverage andvoter judgments are indicative of causality, butthe nature of the causation must be accountedfor in terms of a conceptual framework ortheory. The theory itself may be falsified infuture empirical studies, in addition to beingchallenged by alternative theories.

An abductive substream

Abduction is the third form of inference, andhas been afforded much less attention thaneither deduction or induction in the theory of


1111234567891012341115678920123456789301234567894012345678950111 � intercoder reliability – Chapter 13, p. 212

science as well as in empirical research. Thissection describes abduction as the (unacknowl-edged) logic of much qualitative research, andgoes on to consider the place of abduction inrelation to other methodologies as well.

Since inductivism has come to appear unten-able as an epistemological program, the inter-pretive paradigm has been hard pressed, eitherto defer to the hypothetico-deductive model, orto specify an alternative. In direct opposition towhat he termed Hempel and Oppenheim’s (1988[1948]) ‘covering-law model,’ Dray (1957) spec-ified how historical events cannot be examinedas a variant of natural events (which may all be‘covered’ by one law), but require a differenttype of ‘rational explanation.’ In another influ-ential contribution, Danto (1965) suggested thatnarratives provide a model for understanding,and for empirically studying, historical eventsand human actions (see also Bruner 1986).Further, Ginzburg (1989) identified an ‘eviden-tial paradigm,’ within which, for example,Sigmund Freud and Sherlock Holmes were bothable to uncover an underlying structure, respec-tively in dreams and crimes.

While no consensus comparable to thehypothetico-deductive model has yet emerged,one advantage of abduction is that it lends itselfto comparison and combination with the trad-itional types of inference. Since Peirce’s earlyformulation, its relevance has occasionally beenconsidered in both philosophy and other disci-plines, including mainstream sociology (Merton1968: 158). It was reasserted by, among others,Hanson (1958) as part of the post-1945 ques-tioning of both inductive and hypothetico-deductive prototypes of research. Recently,abduction has been reintroduced into socialand cultural research as an alternative to theinductivist self-understanding of groundedtheory (Alvesson and Sköldberg 2000), as astrategy of interpretive social science (Blaikie1993), and as a characteristic of qualitativemedia research (Jensen 1995). In fact, qualita-tive empirical studies that tend to refer to them-selves as interpretive and broadly inductiveoften rely on an abductive procedure.

To illustrate, Radway’s (1984) study ofromance reading was an exemplary contribu-tion to early empirical reception studies (for

later work see Radway 1997). One quality wasits combination of a history of the genre withan account of its system of production and dis-tribution, as well as in-depth research into itsfemale readers’ decodings and social uses of thetexts. The field study of readers comprisedinterviewing and observation in an ‘interpretivecommunity’� which had formed around a par-ticular bookstore. This case-based analyticalstrategy entailed a stepwise and highly contex-tualized articulation of meaningful categories,which this study shares with much other qual-itative work.� The researcher went throughseveral stages of establishing the meaning ofromance reading for these women, relying onboth face-to-face interviews and questionnaires.By probing motivations such as ‘education’ and‘escape,’ and by taking cues from some espe-cially articulate informants and presenting theirresponses to others in later interviews, theanalysis came to differentiate a blanket termsuch as ‘escape’ into additional categories ofrelaxation and time for self-indulgence.

Radway (1984) summed up the attrac-tion of romance reading in concluding that ‘itcreates a time or space within which a womancan be entirely on her own, preoccupied with her personal needs, desires, and pleasure’(p. 61). A central implication is a shift ofemphasis, away from romances as texts thatoffer a more or less escapist universe for thereader’s identification and gratification, andtoward an understanding of the very activity ofreading as a social practice which enables thesewomen to position themselves in, and apartfrom, the rest of everyday life. Media discoursesare not only representations of the world, butalso resources in the world.

An inductivist interpretation might takethese categories to emerge from the field itself. Itis more appropriate, however, to describe themas the outcome of abduction by the researcher,as she repeatedly engaged the informants’worldviews. In trying to grasp their perspective‘from within,’ in ‘emic’ terms,� Radway can be


covering-lawmodel

narrative asresearchmodel

evidentialparadigm

� interpretive communities – Chapter 10, p. 167

� case-based and code-based analysis – p. 256

� emic and etic analytical categories – Chapter 14, p. 236

said to have performed a sequence of abductiveinferences. Her purpose was to account for thesewomen’s experiences of their reading, and oftheir lives, by introducing concepts or rules thatwould made the informants’ statements mean-ingful and relevant in the context where theywere made.

As noted by Peirce, the essence of abductivereasoning may be expressed in the form of asyllogism. Here, first, is Radway’s recapitula-tion of one of her main findings, namely thatromance reading is a specific kind of social act:

In summary, when the act of romancereading is viewed as it is by the readersthemselves, from within a belief system thataccepts as given the institutions of hetero-sexuality and monogamous marriage, it canbe conceived as an activity of mild protestand longing for reform necessitated by thoseinstitutions’ failure to satisfy the emotionalneeds of women. Reading therefore func-tions for them as an act of recognition andcontestation whereby that failure is firstadmitted and then partially reversed. Hence,the Smithton readers’ claim that romancereading is a ‘declaration of independence’and a way to say to others, ‘This is my time,my space. Now leave me alone.’

(Radway 1984: 213)

Next, the central and somewhat surprisingnotion, that romance reading is a ‘declarationof independence,’ can be explicated in the formof an abduction:

Romance reading is a declaration of independence.

All uses of texts by readers to claim theirown time are declarations of independence.

Conclusion: Romance reading is a use oftexts by readers for claiming their own time.

Whereas the first premise registers a puzzlingfact from the empirical universe of romancereaders (puzzling to the extent that the romancegenre tends to represent women in dependentroles), the second premise introduces the con-ception or rule that texts are resources in every-

day life. At the same time, the second premisemay be seen to sum up the research processwhich served to gradually articulate variousconceptions of the romance genre and of theact of reading. In this regard, the singular state-ment arguably builds on several previousabductions from an iterative process.

One quality of formalization is that it facil-itates a systematic comparison of the severaldifferent kinds of reasoning that enter into thepractice of empirical research. In the case ofqualitative research projects, the category ofabductive inference can help to explicate theirvarious contextual and empathetic forms ofinterpretation. For the further development ofqualitative research this is an important chal-lenge, both in the field and in the several laterstages of analysis.

Abduction is at the core of that interchangebetween researcher and informants whichserves to establish – infer – relevant categoriesand concepts for the specific purpose of inquiry.By reconsidering and reformulating their per-spectives on romance reading, Radway’s (1984)informants produced new insights, not only forresearch, but presumably for themselves, asthey verbalized their conceptions, perhaps forthe first time. Abduction (like induction) is a common aspect of everyday reasoning as well. By formalizing or systematizing the com-ponents of their analysis and interpretation,qualitative studies gain a resource, first, forstructuring key moments of the analyticalprocess and, second, for deciding who (infor-mants, researchers, both) originates whichabductions. Third, the category of abductioncan help to document the steps of the researchprocess, and to make them transparent toothers in the name of intersubjectivity andinterdisciplinarity.

Several kinds of abduction occur in scien-tific as well as everyday reasoning. Eco (1984)has outlined a preliminary typology:

• Overcoded abduction is a basic form ofcomprehension that works semi-automatically.‘When someone utters /man/, I must first assumethat this utterance is the token of a type ofEnglish word’ (Eco 1984: 41). No complexinference is needed to first establish the fact that


1111234567891012341115678920123456789301234567894012345678950111

people may speak different languages, and nextthat English is the appropriate choice in context.

• In performing an undercoded abduction,however, one must choose between several pos-sible interpretations of a word or statement. InEco’s words, ‘when one utters /this is a man/,we have to decide whether one says that this isa rational animal, a mortal creature, or a goodexample of virility, and so on’ (p. 42).

• Creative abduction, finally, occurs when thevery rule of interpretation has to be inventedfor the specific purpose, for example, in the caseof poetic language, as it enters into both poetryand advertising. In science, Darwin’s interpre-tation of humans as the latest animal to enterthe evolutionary chain was an (unusually) cre-ative abduction.

By thus treating abduction as a specificallyinnovative element – a creative potential thatmay be administered in research in several dif-ferent ways – it becomes possible to relate itback to deduction and induction, and to con-sider their combination in various forms ofresearch. In this perspective, the hypotheses ofquantitative projects arguably are not merelydeduced. They are, in part, the outcome ofundercoded abductions which articulate newconceptual relations or configurations with ref-erence to earlier research. Qualitative projects,in their turn, perform a sequence of (iterative)undercoded abductions, which ideally accumu-late at the end as a robust idea or theory. In a sense, the abductive labor of hypothetico-deductive research is focused in the earlyphases, while in qualitative studies the sequenceof abductions is distributed along and acrossthe entire research process.

Overcoded abduction, moreover, arguablyenters into the administration of (already famil-iar) analytical categories and procedures. This isthe case for both qualitative and quantitativemethodologies. Creative abductions, finally, arethe kind of unusual event and scarce resourcewhich both mainstreams can only hope toproduce once in a while. They emerge in theoperationalization of innovative hypotheses,and through the exploration of new meanings inthe field.

In conclusion, the three types of inferencepoint to several different types of relationshipbetween empirical analysis and theory develop-ment in media and communication research.While many, perhaps most, actual research pro-jects may be said to include aspects of abduc-tion, deduction, as well as induction, two mainvarieties have been identified – hypothetico-deductive reasoning in combination with quan-titative types of data analysis, and iterativeabductive reasoning employing qualitative dis-course and other data analysis. The reviews inthis chapter indicate that a unification of thesetraditions in the first instance – in their concretemeans of empirical analysis – has not occurred,and is unlikely to occur in the future. However,in addition to sometimes joining their empiri-cal forces within a shared theoretical frame-work, qualitative and quantitative traditionshold a further potential for unification as com-ponents of interdisciplinary fields of research,such as media studies. Exactly how this poten-tial may be tapped is one of the most import-ant and difficult questions for future research,and is outlined in the next section.

UNIFICATION IN THE FINAL INSTANCE

Validity and reliability reconsidered

A final aspect of most research projects – sup-porting their primary analytical operations andinferences – is to perform various types of‘quality control.’ The purpose is to assess analy-ses and conclusions according to the standardsinvoked by the study itself and, in a next step,to make both the standards and the findingsaccessible for collegial, public scrutiny.

Through categories of validity and relia-bility,� quantitative research has provided themost elaborate set of measures and proceduresfor evaluating empirical findings and proce-dures (see also Black 1999). At the same time,the specific techniques have been perceived asless relevant to the concerns of qualitativeresearch (Kirk and Miller 1986). In some cases,this perception arguably has led to a neglect of


abductiveaspects of

hypotheses

qualitativeresearch as

iterativeabduction

� quantitative concepts of reliability and validity –Chapter 13, p. 212

the issues involved, as noted by, among others,Höijer (1990). In other cases, qualitative re-searchers have proposed new terminologieswhich would recognize the processual and con-textual nature of qualitative research, forinstance, trustworthiness, credibility, depend-ability, transferability, and confirmability (Lin-coln and Guba 1985). So far, however, suchalternative terminologies have not taken hold,certainly in media research. Instead, the con-ceptual and operational definitions of ‘validity’and ‘reliability’ may be extended.

In brief, reliability addresses the consistencyof descriptions and interpretations over time,typically in the form of repeated measurements.In the example from agenda-setting researchgiven above (McCombs and Shaw 1972), theintersubjective agreement of coders was ex-pressed in a measure of intercoder reliability.Validity, in turn, addresses the extent to whicha research ‘instrument’ measures what it wasintended or is claimed to measure – all researchaims for truth in some such sense. An addi-tional distinction is made between internalvalidity (evaluating the consistency of the con-cepts and procedures that are applied to theempirical context) and external validity (assess-ing whether the findings from this context maybe generalized to other situations or popula-tions). In the agenda-setting example, both theconception of ‘agenda,’ and the relationshipbetween the community studied and the largerelectorate were considered.

Validity as well as reliability have tradition-ally been expressed in measures that serve assummary assessments of the research. It is thissummary feature, in particular, which has ledqualitative researchers to call for more contin-uous and contextual assessments of both theresearch process and its findings. Some have,for example, advocated ‘validation’ as a con-tinuous activity throughout the researchprocess (e.g., Kvale 1996).

Figure 15.3 presents a schematic overviewof validity and reliability, as they relate to eachother and to other components of the researchprocess. In both qualitative and quantitativestudies, reliability concerns the intersubjectivecomponent of research in a general sense. Beingan inherently interactional, social phenomenon,

intersubjectivity is established not only by com-paring minimal measurements in the earlystages of a study, but by addressing emergingfindings, forms of documentation, and issues ofinterpretation. For instance, while the use oftwo independent coders for categorizing thesame dataset is standard procedure in severalquantitative traditions, intercoder reliabilitycan also be ensured by consensual coding andother collective approaches to analysis andinterpretation (e.g., Neuman et al. 1992). Simi-larly, informants may be involved in somestages of analysis, through several waves ofinterviewing, and by ‘member checks,’ as inRadway’s (1984) study of romance reading.

Most generally, reliability remains an issueafter the conclusion of the research processproper – in reanalyses of data and discussionsof findings, both among researchers and inother social contexts. The category of reliabil-ity, then, leads into questions concerning thesocial organization and uses of research and itsknowledge interests. ‘Reliable’ findings amountto knowledge which individuals and institu-tions are prepared to act on.�

The category of validity, equally, opens onto a wider scientific as well as social terrain.Compared to various specific, quantitative mea-sures of validity, a qualitative study typicallyattends to the internal validity of its catego-ries by exploring and comparing, for instance,

Unification in the final instance 267

111123456789101234111567892011123456789301234567894012345678950111

validation

Reliability

SignsSubject

Signs

Object

Signs

Subject

Validity

Figure 15.3 Dimensions of validity and reliability

� reliability as social relevance – Chapter 16, p. 274

reliability as inter-subjectivity

informants’ conceptual repertoires, while itsexternal validity must normally be left unde-cided, until other cases or larger samples havebeen examined. In addition, ‘triangulation,’which combines several analytical perspectiveson the same empirical context, has beeen givenspecial attention by qualitative researchers.�These different conceptions of validity arise, inpart, from two distinctive concepts of ‘general-ization’ within quantitative and qualitativeresearch, respectively (e.g., Yin 1994: 30).Empirical or statistical generalization refers tothe capacity of quantitative research to applypredefined (hypothetically deduced) categoriesto a representative set of empirical instances,thus supporting external validity. Theoreticalor analytical generalization refers to the articu-lation (abduction) of new concepts or interrela-tions which conceive empirical instances in amore consistent or insightful manner, henceprioritizing internal validity.

The general topic of quality control in em-pirical research thus represents a fertile area fortheory development and interchange betweenparadigms. For instance, aspects of validity andreliability have been brought together in notionsof pragmatic or communicative validity, empha-sizing the process of validation (see Flick 1998:221–240; Kvale 1996: 244–251). Before intro-ducing a framework which may satisfy severalcriteria of generalization, reliability, and valid-ity, it remains to consider the other constituentsof Figure 15.3.

The figure acknowledges that differentresearch traditions rely on different ‘signs’ – re-search instruments, analytical procedures, meansof documentation – which enable researchers as‘subjects’ to engage their ‘objects’ of analysis. In social and cultural research, these signs or dis-courses take on a special significance as commu-nications, because here the ‘objects’ of analysisare more commonly subjects who provide inter-pretations of themselves and their circumstances– of which research performs a second-orderinterpretation. Communication theory faces and rearticulates many issues from the theory ofscience: How is knowledge of other minds, andof reality as such, possible?

An important task for the theory of scienceis to devise meta-discourses which allow forcomparison of, and reflection on, the differentkinds of knowledge that the signs of differenttheoretical and epistemological traditions areable to communicate. Signs make realityresearchable, communicable, and disputable. InMead’s (1934) terms, researchers and theirinformants, at least for a moment, becomesignificant others for each other through com-municative interaction. Researchers equallyconduct a specific kind of social interactionamong themselves, and the research institutionas such ultimately engages the rest of society ina double hermeneutic concerning the meaningof social life (Giddens 1984).

Realism reasserted

Realism has become an influential position inrecent theory of science. Pavitt (1999), for one,has suggested that this is now the dominantposition; indeed, that it already informs thepractice of much current media and communi-cation research. The positions of ‘logical em-piricism’ (from logical positivism throughPopper) and constructivist ‘perspectivism’ (fromKuhn to poststructuralism and beyond) havesometimes been perceived as absolute opposites,or as the two legs of an unresolvable dilemma.Realism presents itself as a likely candidate fora framework within which convergence mayproceed.

While several varieties of a contemporaryrealist position may be singled out under head-ings of scientific, transcendental, and criticalrealism (see further Archer et al. 1998), its mainimplications for empirical media studies can belaid out with particular reference to the earlywork of Bhaskar (1979) (who later turned tomore metaphysical as well as political con-cerns). At its most general level, Bhaskar’s (crit-ical) realism has three premises:

1 Ontological realism. Rejecting skepticistand nominalist positions which variously haveheld that no knowledge of the empirical worldis possible, or that reality is nothing but the sumof our descriptions of it, realism reverses theburden of proof, in a sense. A realist would


validity:empirical vs.

theoreticalgeneralization

� triangulation – p. 272

pragmaticvalidity

argue that we must assume the existence ofreality as a limit condition or regulatory idealin order to account for the sorts of natural andcultural phenomena one encounters in scienceas well as in everyday life.

2 Epistemological relativism. From a moder-ate constructivist position, realism assumes thatknowledge of both nature and other mindsdepends on a reiterated sequence of percep-tions, cognitions, and inferences, all of whichmay be questioned, rejected, and revised.

3 Judgmental rationality. Like other humanpractice, science must depend on the exercise ofrationality which, at some point, must end in(fallible) judgments about what to do next – asan individual researcher, a scientific field, or asociety. Up to that point, the business of scienceis to continuously compare and contrast alter-native accounts of the same reality with refer-ence to a wide range of means and criteria forrepresenting and interpreting reality.

In moving from such general premisestoward specific research procedures, Bhaskarand other realists have questioned a certain pre-sumptuous ‘anthropocentrism’ in the theory ofscience: ‘Copernicus argued that the universedoes not revolve around man. And yet in phil-osophy we still represent things as if it did’(cited in Archer et al. 1998: 45). Concretely,logical positivism, for one, proposed to reducereality, as legitimately studied by science, towhat is immediately accessible to the humansenses. By contrast, the mechanisms of realitymust be said to operate prior to their discoveryby humans. Reality is ‘intransitive’ (in gram-matical terms), but it is made ‘transitive’ as anobject of discovery in a variety of forms.

The argument concerning transitivity leadsinto a related argument concerning transfactu-ality, namely that facts of several kinds exist.These facts are not reducible to each other, and they call for different forms of inquiry.Both of these arguments may be taken tosupport a further point regarding the stratifica-tion of reality. On the one hand, aesthetic ex-perience or microsociological order are realphenomena, as established by interpretive tradi-tions of research. On the other hand, they do

not represent a separate mode of reality. A rela-tionship of emergence obtains between theseand other aspects of reality, such as the bio-logical capacities of humans and the structuralconditions of social life.

In methodological terms, this line of argu-ment implies a distinction between threedomains of reality, as addressed at differentlevels of research (Figure 15.4).

1 The empirical domain is the source of con-crete evidence – experience of the world. Byexperiencing and documenting, for example,how journalists collect information, and howreaders respond to it as news, researchersprocure a necessary, though not sufficient con-dition of empirical research.

2 The actual status of this documentation is amatter of inference. It is by characterizing and conceptualizing the empirical materials asevidence of events (e.g., reporter-source inter-actions or decodings) that one can infer theiractuality and their association with mediatedcommunication.

3 The domain of the real is more inclusivethan both the empirical and actual. Researchseeks to establish the mechanisms which mayaccount for events (e.g., a system of politicalcommunication which operates according toeconomic prerogatives, professional routines,as well as ideals of citizenship). In this manner,research is able not only to perceive, document,and categorize, but to explain a portion ofreality.

In sum, experiences, events, and mechanisms areall real. Experiences may seem to ‘push’ them-selves upon researchers as evidence of events.However, the task of research is to mount acountervailing ‘pull’ – to infer the underlyingmechanisms by a great deal of methodologicaland theoretical labor.

The realist framework is of special interest in a convergence perspective. It indicates thatwhile different empirical procedures (e.g., ex-periments or depth interviews) document and,in a sense, privilege particular kinds of events(the recall or decoding of media content), they may nevertheless bear witness to related

Unification in the final instance 269

1111234567891012341115678920123456789301234567894012345678950111

transitivity,transfactuality,

andstratification

emergence

mechanisms. Instead of engaging in conflict overa singular definition of the empirical domain, arealist strategy thus proposes to take advantageof several methodologies in order to documentvarious aspects of mediated communication. Itis the overriding task of research to inquire intothe actuality of the phenomena as documented,and to explore whether they may be subsumedunder similar explanatory mechanisms.

Probabilities

An important example of a concept which lendsitself to such meta-theoretical development anddialogue across the qualitative–quantitativedivide is ‘probability.’ As shown by Hacking’s(1975) historical analysis, the concept gradu-ally acquired an ambiguous meaning in modernphilosophy and empirical sciences. Indeed,much thinking about probability has conflatedtwo distinct meanings, with problematic conse-quences for research practice:

1 Stochastic probability has to do with stablerelative frequencies, as established by statisticalprocedures. Here, the purpose is to rule out,beyond reasonable doubt, that the particularconfiguration of (empirical) findings could haveoccurred by chance (the null hypothesis).�

2 Epistemological probability concerns ‘thedegree of belief warranted by evidence’ (p. 1).Here, the concept of probability applies to the(actual) events and (real) mechanisms to whichthe empirical evidence is said to bear witness.

The implication of the distinction is sometimessummed up in the dictum that ‘correlation doesnot equal causation.’ In other words, statisticalmeasurements and other formalisms do not inthemselves warrant conclusions about causalityand other types of association. Theoretical and substantive arguments are always requiredto assign explanatory value to the empiricalmeasures.

Relating Hacking’s (1975) analysis to themedia field, Ritchie (1999) has suggested thatmuch current empirical media research fails onthis crucial point. The slippage occurs when‘the statistical probabilities associated with thenull hypothesis are . . . used to support infer-ences about the epistemological probabilities ofa preferred interpretation’ (p. 7). Put differ-ently, the fact that the null hypothesis, whichassumes random findings, is sufficiently im-probable (statistically), is mistaken for evidencethat a specific alternative hypothesis, namelythe one deduced at the outset of a study, is more probable (epistemologically). The logic of hypothesis testing may thus contribute to aconflation and confounding of two distinctlevels of scientific analysis and argument. Likevalidity and reliability, the category of prob-ability lends itself to refinement in a conver-gence perspective.

All empirical research necessarily examinesan empirical ‘microcosm’ with reference to atheoretical ‘macrocosm’ (for a history of theconcepts in social research see Alexander andGiesen 1987). Qualitative and quantitativemethodologies may be defined, in part, withreference to their conception of and approachto the empirical microcosm. Figure 15.5 indi-cates (top right-hand corner) how the popula-tions and samples of quantitative projects make


Experiences

Events

Mechanisms

The real

x

x

x

The actual

x

x

The empirical

x

Figure 15.4 Three domains of reality, incorporating three types of phenomena

� null hypothesis and stochastic probability – Chapter13, p. 232

up two levels of an empirical universe, in whicheach level represents a set of relatively self-con-tained units. In qualitative projects, contexts,cases, and informants are made into objects ofanalysis according to different principles ofsampling.� It is the reciprocal conditioning andconstitution of such units within networkswhich motivate both their selection and thequalitative research questions.

The various modes of data collection andanalysis that are associated with qualitative andquantitative research traditions represent differ-ent means of gaining experience of particularaspects of the media field – and not others. Thechapters in this volume may be taken to con-clude that both quantitative and qualitativemethodologies are best suited for studying par-ticular kinds of social and cultural events – andnot others. The matching of research questionsand methodologies will depend not least on the degree to which the relevant theoretical

categories and issues are already well estab-lished and familiar – both to respondents andresearchers – in public discourse. What unitesthe two main forms of empirical research onmedia production, texts, reception, and con-texts is that they address a middle range ofsocial and cultural phenomena which call forboth detailed documentation (empirical experi-ence) and grand theorizing (theoretical mecha-nisms). Ultimately, it is the mechanisms ofmodern culture and mediated communication,as defined and related in a theoretical macro-cosmos, that both main traditions will beworking, and cooperating, to uncover in furtherresearch.

CONVERGENCE IN PRACTICE: THREEAPPROACHES

Whereas convergence is in evidence in several of the media-related literatures summarized inthis volume, much work is required to integrateand consolidate elements from the theory of

Convergence in practice: three approaches 271

111123456789101234111567892012345678930123456789401112345678950111

Events

Experiences

Empiricalmicrocosm

MechanismsTheoreticalmacrocosm

Theoretical frameworks ofconceptual configurations

Methodological designs andsubstantive selections

Middle range ofsocial and cultural

phenomena

Cases in contextSample

Population

Figure 15.5 Empirical microcosms, theoretical macrocosms

� sampling in qualitative research – Chapter 14, p. 238

science and communication theory, as well asfrom diverse methodologies. Methodologiesrepresent a strategic area of dialogue and coop-eration, because they join theoretical concernswith the practical requirements of empiricalwork. In fact, over the past decade, a growingnumber of publications have outlined ways ofcombining or ‘mixing’ qualitative and quantita-tive methodologies (e.g.,Bernard 1995; Brannen1992; Creswell 1994; Miles and Huberman1994; Tashakkori and Teddlie 1998). Several ofthese approaches have transfer value for mediaand communication research.

In conclusion, it is possible to single outthree principal forms of combining qualitativeand quantitative methodologies (Hammersley1996: 167–168):

1 Facilitation. The common practice tradi-tionally has been to treat qualitative and quan-titative components as relatively separate stagesof a sequence. The norm in survey research and other quantitative studies of, for example,media consumption is to conduct a qualitative‘pilot’ study. The aim is to arrive at categoriesand formulations that are both understandableand meaningful to respondents. Less frequently,a quantitative phase, for instance, a mapping of a social and lifestyle segment, can facilitatesubsequent in-depth analyses of the interpretivecategories that make up the lifestyle.�

2 Triangulation. As developed especially byDenzin (1970, 1989), triangulation is a general

strategy for gaining several perspectives on thesame phenomenon. In attempting to verify andvalidate findings, the strategy addresses aspectsof both reliability and validity. Triangulationmay be performed through several datasets,several investigators, and several methodolo-gies, in the latter case combining, for instance,experimental and observational approaches to human–computer interaction.� (Denzin’s(1989: 239) additional suggestion that theoriesmight also be triangulated seems more prob-lematic, since theoretical interpretation norm-ally involves closure around a single perspec-tive, however preliminary. Methodological tri-angulation also leads up to a weighing andinterpreting of the several sets of evidence fromone position.)

3 Complementarity. The most challenging,and so far the least common, option is com-plementarity. Different methodologies may bebest suited to examine different aspects of aresearch question, and not necessarily in thesame concrete empirical domain. In the finalinstance, methodologies and findings can bejoined with reference to a common theoreticalframework. The existence of scientific ‘camps’has often worked against exploring even thepotential of complementarity. Neverthless, bothclassic studies such as Cantril et al. (1940) andrecent developments in media studies (for anoverview see Potter 1996) suggest that com-plementarity can and should be a significantfeature of research practice in the future.


� lifestyle research – Chapter 9, p. 153 � human-computer interaction – Chapter 11, p. 184

THEORIES INTO PRACTICE

Making public

This final chapter returns to a number of theissues which motivate media students and re-searchers in the first place. Indeed, why studythe media? (Silverstone 1999). Individual re-searchers are prompted, in part, by the sameconcerns which bring major economic and political agents to focus on the area. The mod-ern media are sources of power as well as of meaning – mediated meanings can have powerful social consequences. Accordingly, thischapter examines the three-way relationshipbetween the media, research, and the rest of thesocial system. Media studies, like their object ofanalysis, originate from a particular social andhistorical setting. Part of the relevance of mediastudies is that they may contribute to the socialconditions under which communication willtake place in the future.

Like the media themselves, then, universitydepartments and other research organizationsmay be understood theoretically as institu-tions-to-think-with, enabling (second-order)

reflexivity about the role of media in society.�In offering their perspectives, researchers parti-cipate in a double hermeneutic (Giddens 1984):they reinterpret the ‘lay theories’ (Furnham1988) of ‘ordinary’ social agents, and feed those reinterpretations back into society. Forexample, citizens’ lay theories of the category of ‘public opinion’ vary widely (Herbst 1993), may differ from those of political andmedia theorists, but are, nevertheless, informedover time by scientific theory, in part, throughmedia. In addition, programming decisions by television executives have been centrallyshaped by new research on the ‘active’ audience(Eastman 1998).

Alvesson and Sköldberg (2000: 248) havegone on to suggest several additional types ofhermeneutics. For instance, a ‘triple’ hermeneu-tics would be performed by critical theory withthe specific aim of exposing and ending rela-tions of social dominance. The general point ofhermeneutics in this regard is that all social

1111234567891012341115678920123456789301234567894012345678950111

The social origins and uses of media and communication researchKlaus Bruhn Jensen

16

• a presentation of the several types of ‘theory’ which connect research with social practice• a review of the normative theories of media• a description of media and communication research as a social institution• a comparison of the main applications of media research in policy and politics• a discussion of both ethics and logistics as aspects of the relations between researchers, the academic

community, respondents, and wider social communities.

� media research as second-order institution-to-think-with – Chapter 1, p. 6

double andtriple hermen-eutics

practice is informed by ‘theories,’ defined asgeneralized conceptions of what the world islike, and how things can be done. As concludedby one early founder of the media field, KurtLewin, ‘Nothing is as practical as a goodtheory’ (Greenwood and Levin 1998: 19). It isparticularly through an interchange with othertypes of theory – as held by media profession-als, legislators, and the general public – thatcommunication theory and research may makea difference in practice.

Having thought matters through withintheir institutions, media researchers publishtheir findings. Publication in academic settings,however, is only one way of making public the knowledge which is generated by researchprojects. To make public is to make findingsavailable for social practice. The interested con-stituencies may be large or small, specialized orgeneral, representing dominant or oppositionalinterests, and may operate in either the short or the long historical term. One central con-stituency is media businesses that are willing topay not only to have studies conducted, but tohave the results kept away from any significantpublic for competitive reasons.

An extended category of ‘publication’ clearsthe ground for an examination of how differ-ent forms of research relate to other social insti-tutions, as allies as well as adversaries. In orderto assess the present and future relevance ofmedia research, it is important to consider bothits scientific methodologies and its social prac-tices. While publication in journals, conferenceproceedings, books, as well as more popularformats remains a necessary infrastructure anda procedure legitimating the scientific institu-tion, researchers and students are well advisedto consider other channels and contexts ofimpact – including the unanticipated conse-quences of their research activity.

Following a definition of several forms of‘theory,’ this chapter elaborates on the so-callednormative theories, which fueled the early riseand subsequent transformations of the modernmedia, and which still motivate much contem-porary research. An overview of current mediaand communication research as an institution,next, notes its main academic and commercialvarieties and the intellectual cultures driving

them. A distinction between ‘politics’ and‘policy’ further helps to specify different kindsof interchange between research and media –from an instrumental alliance to an adversarialrelation challenging the media as they nowexist. While both quantitative and qualitativemethodologies enter into studies from policy aswell as political perspectives, the methodologi-cal boundaries, to a degree, dissolve in thesocial uses of media research. In conclusion, thechapter reviews a number of practical, ethical,and professional issues in planning student pro-jects as well as research programs.

Four types of theory

Applying an extended understanding of theoryas plans of action to the media field, McQuail(2000: 7–9) has made a useful distinctionbetween four types of ‘theory’:

1 Scientific theory. The most common under-standing of the term is that theory consists in‘general statements about the nature, workingand effects of mass communication, based onsystematic and objective observation of mediaand other relevant factors’ (p. 7). In this sense,both qualitative and quantitative studiesprovide explanatory concepts and models thatapply to a range of empirical instances.

2 Normative theory. While scientific theorieshave also normally been underpinned, in oneway or another, by ideals of how the mediaought to operate in society, the normative the-ories represent a separate area of inquiry intothe ways and means of organizing and financ-ing mediated communication. Dating from the seventeenth century, normative theory hasremained central through to current debatesabout public service broadcasting and theInternet.

3 Operational theory. Media professionalscommand a repertoire of procedures thatamount to operational theory, from practicalrules-of-thumb to ingrained ethical and ideo-logical positions regarding the purpose andstatus of their work. Reflecting its interested andpractical nature, professional knowledge over-laps with both normative and everyday theory.

Social origins and uses of research274

publication:knowledgeinto social

practice

4 Everyday theory. Finally, the public’s inter-action with media as consumers, citizens, and,occasionally, sources of information is guidedby a number of everyday or common-sensenotions of what media are, how they operate,and in whose interests.

Other chapters in this volume have emphasizedthe scientific theories which constitute the aca-demic media field. The professional and every-day theories that enter into media productionand reception have also, in part, been addressed.This chapter shifts the emphasis toward norma-tive theory, examining, as well, its interrelationsparticularly with scientific theory.

Being a strategic cultural resource, a major

economic sector, and a political institution intheir own right, the technological media havegenerated a large proportion of commerciallymotivated as well as socially concerned re-search. From its initiation, the field was part andparcel of the emergence by the 1930s of whatBeniger (1986) termed ‘the control society’ –characterized by a greatly intensified surveil-lance of society, both by individuals through themedia, and by private as well as public agenciesthrough, for example, market research andopinion polling. Several of the early ‘milestones’of media research� were produced in responseto perceived social problems (e.g., violence,

Theories into practice 275

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 16.1 THE SIGNS OF SCIENCE

Media and communication researchers rely on varied means of representation and expression inorder to arrive at an understanding of the empirical field themselves, to share findings and issueswith colleagues, and to present their studies to an interested public. While other sections of thischapter consider both different ways of ‘making public’ and the relevant audiences, the purposeof this brief element is to reflect on the signs of science – the concrete means of communicat-ing research.This can be an important way of keeping research self-reflexive, scientifically as wellas socially. As noted, for instance, in the discussion of rhetoric in Chapter 2, signs and numbersare never innocent, but carry implicit premises (see further Barrow 1992).

Models and other means of display have been integral to the development of media studies(for additional references on scientific illustration, see Lee and Mandelbaum 1999; McQuail andWindahl 1993; Shore 1998).Whereas tables and figures may be associated specifically with quan-titative traditions of research, Chapter 14 noted how visual display, coding, and modeling are alsoan important part of qualitative analysis.

Throughout this volume, a number of verbal, mathematical, and graphic forms have been usedto communicate different points. In review, some of the main types may be described as follows:

• literature reviews and theoretical arguments in all chapters in verbal discourse;• conceptual models of a portion of the field, as represented in either graphic displays (e.g.,

Figure 5.2) or in multiple-field matrices (e.g., Figure 7.2);• analytical examples (e.g., the analysis boxes, such as Box 14.1), employing prose, graphic

displays, and images to represent the object of analysis and aspects of the analytical process(e.g., Chapter 8 on The Big Sleep);

• tables summarizing findings in terms of a numerical distribution (e.g.,Table 13.1);• scattergram, indicating correlations between data elements concerning, for instance, opinions

and media preferences (Figure 9.4);• time line, locating shifting technologies and institutions of communication in relation to each

other (Figure 2. 1).

(Other common formats of presentation include bar charts, histograms, line graphs, and pie charts.(See further Deacon et al. 1999: 93–98.) )

the controlsociety

� milestones of media research – Chapter 10, p. 158

propaganda) which were associated with themedia as well as with the new urban and inter-national realities embedding them – a differentculture of time and space (Kern 1983).

The mediatization of (Western) societies isperhaps best understood as one element of theprocesses of bureaucratization and rationaliza-tion that were then taking place as aspects ofmodernization in politics and economy.� A keyrole of media became that of facilitating theoverall stability and integration of increasinglycomplex social systems. Cui bono? The nor-mative theories provided a framework of ideasand ideals for addressing issues of social power,personal identity, and political and culturalrights in relation to mediated communication.

NORMATIVE THEORIES

The entire set of normative theories illustrateshow ideas with a long history are sometimesmobilized for particular historical purposes.Certain of their constituents date from Renaissance and Enlightenment ideals and,indeed, from the Socratic dialogue as a meansof true insight, and of doing what is morallygood. However, the normative theories werearticulated specifically to the media field in the context of the Cold War. Also in this emerg-ing academic field, the period pitted differ-ent models of society against each other. Theclassic publication identified four theories, with special reference to the printed press(Siebert et al. 1956):

1 Authoritarian theory. A traditional model ofpublicity took for granted a social and religiouscosmology which may be described as apyramid or chain of being.� Here, everythinghad its righful place, and ‘information’ flowedtop–down from the monarch, being the repre-sentative of divine authority on earth. Far frombeing a means of oppression, the pyramid couldbe understood as a framework that enabledindividuals to flourish on the road to theirdestiny. Only especially reliable persons were

allowed to disseminate information on anysocial scale, being subject, as well, to censor-ship, and their audiences were just that – recip-ients of messages from political and religiousauthorities who knew better. While rarelyadvocated as such, being the unspoken doxa ofthe medieval, feudal order, the authoritariantheory provided the contrast against whichmost later theories defined themselves.

2 Libertarian theory. It was liberal theorywhich, in the areas of both politics and com-munication, challenged authoritarian models.Liberalism informed the larger shift from trad-itional to modern social structures, as epito-mized in the public sphere.� Not only werehumans defined as ends in themselves, withcertain inalienable political, economic, and cul-tural rights; they were also conceived as ratio-nal animals with the ability collectively todefine and administer such rights. One unifyingmetaphor became the ‘marketplace of ideas,’suggesting that the market for goods andservices would also empower individuals topromote their political interests and culturalideals through the press (or to establish onethemselves). The resulting competition of ideas,presumably, would benefit society as such.

3 Totalitarian theory. The occasion for formu-lating the normative theories, as noted above,was the Cold War, specifically the implementa-tion of a totalitarian or communist theory of thepress in a number of countries following WorldWar II. The distinction between totalitarian and authoritarian theory (and their relation tofascism during the 1930s) can be, and has been, debated. Still, it was characteristic that thecentral control of communist regimes overmedia was officially conceived as a means offundamentally restructuring society, rather thanpreserving any social pyramid. Centralizedcontrol, moreover, equaled state or governmentcontrol over all means of production, whetherit was meaning or material goods being pro-duced. Following the breakup of communistsystems in Europe from 1989, in the People’sRepublic of China the development of newjournalistic practices still takes place within a


� modernization – Chapter 11, p. 143

� the great chain of being – Chapter 2, p. 21 � the public sphere – Chapter 1, p. 7

relatively fixed political-economic system (e.g.,Pan 2000).

4 Social responsibility theory. A shift inemphasis, from liberal ideals toward an under-standing of the press and other media astrustees or representatives of the public, hastaken place in the Western world particularlysince 1945. For one thing, the growing con-centration and conglomeration of the mediasector increasingly undercut any simple notionof a ‘free press.’ For another, some new mediaforms, especially radio and television, at leastfor a period were limited in number for tech-nological reasons. Furthermore, all technologi-cal media require economic resources andprofessional skills on a scale which promoteslarge organizations and concentration gener-ally. Whereas references to social responsibilityaccordingly have been witnessed in severalmedia types, European public service broad-casting represents a particularly elaborate andinstitutionalized expression of social responsi-bility theory.�

Apart from the inherently controversial statusof normative theories, later commentators havenoted that the four types fail to capture severaldevelopments in media over the past fourdecades. In particular, media systems in thedeveloping world and the growth of mediaforms with increased public participation –from local radio to the Internet – have led tothe formulation of two further positions (seealso McQuail 1983):

5 Development theory. In the context of decol-onization, the 1960s witnessed intensifyingdebates about media in relation to the ‘ThirdWorld’ (while the other two ‘worlds’ were con-fronting each other in the Cold War). The issuesincluded an imbalance in the flow of news in theworld and the possible international as well asnational, local means of redressing it. Attemptsat developing a comprehensive theory in thisregard had to weigh conflicting interests – agenerally desirable ‘free’ flow of information in the world versus the right of individual cultures

to gain a hearing in world media, and to deter-mine the shape of their own media systems.Debate was further complicated by the fact thatreferences to ideals such as ‘free flow’ and ‘self-determination’ could serve as fronts, either foreconomic expansionism or for governmentspromoting themselves abroad and repressingtheir citizens at home. As suggested by the work of the MacBride (1980) Commission, theissues have proven difficult to formulate in anycomprehensive normative theory, but continueto generate international debate as well asresearch. For example, Husband (1996) hasintroduced the notion of a ‘right to be under-stood’ in the multi-ethnic public sphere.

6 Democratic-participant theory. Particularlyin the Western world, the 1960s witnessed asecond type of upheaval around media, associ-ated with the political mobilization and culturalcritique by anti-authoritarian movements.� Onthe one hand, the social responsibility of themainstream media, their political and culturaldiversity in practice, was challenged. On theother hand, information and communicationtechnologies appeared to offer the means of anovel form of political as well as culturaldemocracy. Moving beyond the liberal andresponsibility theories, democratic-participanttheory proposed steps toward ensuring publicinvolvement, by structural means and notmerely by individual initiative (Enzensberger1972 [1970]). It is this participatory ambitionwhich, in part, has fueled ‘grass-roots’ media(e.g., Downing 2000; Glessing 1970), and itcontinues to inform ideals regarding Internetcommunication (e.g., Rheingold 1994).

In continuing debate and research, several additional varieties of normative theory havebeen outlined (see Nerone 1995; Nordenstreng1997), some of which are outgrowths or closeallies of scientific media theories, for example,on intercultural or postcolonial issues.�Broadly speaking, however, most normative(and many scientific) theories today emphasize

Normative theories 277

1111234567891012341115678920123456789301234567894012345678950111 � public service broadcasting – p. 283

the right tobe under-stood

� anti-authoritarian movements – Chapter 3, p. 56

� intercultural and North–South communication –Chapter 11, p. 177

either critical-reformist or pluralist-functional-ist criteria – a conflictual or consensual modelof society – in evaluating media performance. Arelated opposition between ‘state’ and ‘market’is commonly referred to in policy discussions ofhow to ensure the ‘freedom’ of media.

Part of the difficulty of debating the realconflicts and high stakes in the area beyondsimple oppositions has been the ambiguity ofthe concept of freedom in sociopolitical and,later, communication theory. Different accountstend to assume either a negative definition(freedom from state interference in communi-cation) or a positive definition (freedom todemand certain media provisions as a civicright). Habermas (1989 [1962]) traced thisambiguity to shifting notions of how state orgovernment agencies should interact withvarious sectors of economic and other sociallife. The modern period was inaugurated by the negative definition of freedom, as the newmiddle classes asserted their political and eco-nomic rights vis-à-vis the state. However, a pos-itive redefinition of rights, involving economicregulation and social services along Keynes-ian principles, followed from world crises in the late nineteenth century and especially the1930s. It was this reorganization of the society–state nexus which presumably preserved thelarger system of industrial capitalism and repre-sentative government.

One should constantly keep in mind that cur-rent debates about normative media theory takeplace in the context of highly regulated politicaleconomies – at least in the Western world, anddespite measures of deregulation particularlysince the 1980s. Any reference to a negativedefinition of media freedom (‘less state interfer-ence, more freedom of expression’) is likely toserve rhetorical, not analytical purposes. Thesubstantial point of contention is the particularforms in which the regulation of technologicallymediated communication will take place – fromnational laws affecting film production to theinternational assignment of Internet domainnames. Also in the future, the questions thatmedia researchers will be asked to study andcomment on involve conflicts and compromisesregarding who will benefit most from the defacto positive definition of social freedom.

Against this historical background, the ‘four-plus-two’ normative theories remain relevantpoints of reference by articulating political,economic, and cultural ideals which still enterinto contemporary public and policy debates.(On issues of justice, with largely unexaminedimplications for media, see Rawls 1999.)

The normative ‘theories’ have been sup-ported primarily by abstract reasoning andprincipled argument. Nevertheless, they amountto strongly held beliefs on which whole societieshave been prepared to act, to use pragmatist ter-minology (Joas 1993), and, indeed, to plan their entire system of communication. A centralrole of media research, approximately since the1956 statement of the normative theories, hasbeen to differentiate and strengthen the socialbases of reasoning, argument, and action inrelation to the media. Media and communica-tion research has developed at the crossroads ofseveral social sectors and intellectual currents.

MEDIA RESEARCH AS A SOCIALINSTITUTION

Intellectual cultures

A modern locus classicus regarding the relation-ship between theory and practice (Lobkowicz1967) – between knowing that something is thecase and knowing how to act accordingly – wasthe statement by Karl Marx in his Theses onFeuerbach (1845), that ‘The philosophers haveonly interpreted the world, in various ways; thepoint, however, is to change it.’ Whether or notindividual scholars have drawn either revolu-tionary or reformist consequences from such aview, it is undeniable that, at the institutionallevel, research participates in actively shapingand maintaining modern societies in countlessways. This has been evident not least withinscience and technology during the twentiethcentury (for an overview see Biagioli 1999). In a structural sense, then, basic research also eventually comes to be applied. Compared,however, to a widespread nineteenth-centurynotion of sciences as means of both materialand cultural progress, much twentieth-centuryresearch found itself struggling to come toterms with its sense of a mission. The complic-


from negativeto positive

definitions offreedom

basic andappliedresearch

ity of research in world wars, colonialism, andquestionable forms of social and cultural engin-eering meant that its legitimacy was not a given.

Media and communication research emerg-ed within the university after 1945 at the junc-ture between several intellectual cultures (seethe historical overviews in Resource box 16.1).The history of the modern university (Rudy1984) is, in one sense, a history of reality beingpartitioned into researchable and manageabledomains, to be studied through increasinglyspecialized theory and methods, and to bemanaged in practical life by specialists gradu-ating in each of these specialties. Following thefounding of the research university in the early1800s,� the late nineteenth century had wit-nessed the establishment of the social sciencesas a separate ‘faculty’ alongside the humani-ties.� A recognizable specialty of (mass) mediastudies emerged from the 1930s, as witnessedby the ‘milestones’ of Figure 10.1. (Resourcebox 11.1 provides references regarding nationalvariations of this development.)

It was not until the 1950s, however, that anactual institutionalization occurred, notably inAmerican social science (for an overview seeSchramm 1997), which began to accumulate aseparate body of theory and methodology, andto consolidate its social contexts of influence.During these same decades, humanities depart-ments, including film, literature, linguistics, andhistory, similarly prepared important contribu-tions to media research by expanding thesubject matters of the established disciplines farbeyond high arts and official history and, in theprocess, by reinvigorating and revising theiranalytical procedures and theoretical frame-works. Following the political as well as epis-temological upheaval and stock-taking betweenthe 1960 and 1970s, a process of convergencebetween social-scientific and humanistic tradi-tions gained momentum during the 1980s, asreviewed in this volume.

One of the continuing debates is whethermedia studies constitute an established oremerging discipline, or an interdisciplinary field(for an overview see Levy and Gurevitch 1994).

There is no doubt that media research has a per-manent presence as a social institution. Theindicators are university departments, journals,conferences, consultancies, and continuousinterventions in public debate. Nevertheless, itshould be emphasized that the area remains het-erogeneous. First, bibliometric research suggeststhat ‘communication study’ still consists of tworelatively separate subspecialties of research,focusing on interpersonal and mass communi-cation respectively (Rogers 1999). Second, alsowithin mass communication research, analysesof journal publications suggest the existence ofthree relatively self-contained literatures, repre-senting social science, interpretive studies, andcritical analysis (Fink and Gantz 1996). Withthe growth of computer-mediated communica-tion, which, in certain respects, rejoins the twotypes, however, a move toward integrationseems likely in coming decades. Indeed, it doesnot appear too far-fetched to imagine that fac-ulties of communication and culture whichcomprise social-scientific, humanistic, and tech-nological components will be the structuralresponse of universities, once again, to a chang-ing social context (Jensen 2000b).

In his often cited account of divisions withinacademia, Snow (1964) identified two intellec-tual cultures, as represented by ‘the literaryintellectuals’ and ‘the physical scientists’ (p. 4).Perhaps surprisingly in view of its enabling tech-nologies and engineering aspects, the media fieldso far has had a negligible natural-scientificelement. Instead, it is the humanities and socialsciences that have provided the theoreticalbuilding blocks, compatible or not. Still, ‘thescientific method,’ as associated above all withthe natural sciences, has figured prominentlywithin methodology and theory of science, andperhaps even more so in debates over the poli-tics of doing research. These latter debates havefrequently centered on the question of whetherparticular social interests are served by particu-lar kinds of scientific knowledge.

Knowledge interests

The concept of knowledge interests implies that an ulterior purpose is ingrained in the veryprinciples and procedures of scientific inquiry.

Media research as a social institution 279

1111234567891012341115678920123456789301234567894012345678950111

� the modern research university – Chapter 2, p. 15

� social sciences as separate faculty – Chapter 3

field ordiscipline?

interpersonalvs. masscommuni-cation

Importantly, these interests are different in kindfrom, albeit related to, researchers’ personalconvictions and questions of how research isfunded or situated institutionally. The categoryof knowledge interests provides a frameworkfor examining the relationship between socialends and scientific means in a more nuancedfashion than is often the case, for example, inaccounts simply contrasting commercial andacademic research. Knowledge interests beginto address the relative autonomy of different

kinds of media research, and to clarify issues ofknowledge and power which have come to thefore in recent decades.

The concept was formulated as such byHabermas (1971 [1968]), who distinguishedthree types of knowledge interest. Each is asso-ciated with the characteristic subject mattersand social functions of three faculties of study:

1 Control through prediction. In the case ofnatural sciences, a central purpose of inquiry is


RESOURCE BOX 16.1 HISTORIES OF MEDIA AND COMMUNICATION RESEARCH AS A FIELD

Over the past decade, more historical accounts of the development of media studies have begunto appear. Presumably, this follows, in part, from the fact that media and communication researchis now an established institution in society and a relatively mature field of inquiry. The secondgeneration of researchers, now occupying university chairs, may also perceive a need to revisitand reassess their roots, partly in response to convergence.

At one level, the development of the field is the outcome of interventions from the social sci-ences and the humanities, as traced in Chapters 2 and 3, in response to the central role of mediain modem societies. At a more specific level, different national cultures – in universities, in poli-tics, and in the media themselves – have produced a range of forms in which research and teach-ing are organized.The references in Box 11.1 cover some of these aspects.

Given its longer history as well as its centrality and resources generally, the North Americanresearch community has produced some of the more elaborate historical accounts (e.g., Delia1987). At the same time, different accounts of the U.S. experience bear witness to quite differentconceptions of both historiography and politics. As such, they are instructive, not only regardingcontroversial issues within U.S. research, but also for the writing of histories of the field else-where. The following references are indicative, both of relatively more administrative or criticalperspectives, and of their different sources in the history of ideas:

• Hardt 1992 – a monograph with review and discussion, emphasizing critical and interpretiveaspects of U.S. communication studies, and linking these to pragmatism and the widerintellectual history of the U.S.A.

• Dennis and Wartella 1996 – an edited collection with contributions from several of thecentral figures of U.S. research, including accounts of its roots in Europe and in ChicagoSchool sociology. In a review, Hardt (1999) found that this ‘remembered history’ by keyindividuals served more as a professional position statement than as an analyticalhistoriography, implying that it may be a Whig history written from the still largelyfunctionalist perspective of the victors.

• Schramm 1997 – a retrospective account by the researcher who is normally considered tobe the central figure in institutionalizing communication studies in the U.S.A., supplementedwith perspectives supporting this conclusion, by Steven M. Chaffee and Everett M. Rogers.Hardt (1999), in his turn, argued that ‘Wilbur Schramm had failed to forge a discipline,’ andthat instead ‘mass communication research was legitimated intellectually by the centrality ofcommunication in social theory and cultural studies’ (p. 239).

to be able to plan future activities in the materialworld, in detail and with confidence. Predictionsand hypotheses make for human interventioninto nature under controlled circumstances. By developing and accumulating criteria foranticipating physical, chemical, and biologicalphenomena and processes, the modern scienceshave mastered the natural environment to anunprecedented degree. This has facilitated thehuman management of resources, time, andspace as well as extensive social planning,notably in agriculture and industrial production.(Media example: Quantitative surveys predict-ing the preferences of audiences.)

2 Contemplative understanding. In the human-ities, scholarship has revolved around culturalforms of expression which are subjected tocontemplation – interpretation through intro-spection. Aesthetic works, for one, could beunderstood as ends in themselves that should be analyzed for their inherent meaning andvalue. Historical events, for another, might bearwitness to universal, even eternal aspects of thehuman condition, even if the religious overtonesof contemplation have gradually been down-played. By disseminating their (re)interpreta-tions of culture and history to a wider public,humanistic scholars came to serve, not least, asthe professional keepers of cultural tradition.(Media example: Qualitative textual studies ex-ploring media representations of social reality.)

3 Emancipation through critique. If the naturalsciences procured the material and collectivebases of modern society, whereas the humanitiesaddressed the individual’s life experiences, thesocial sciences were called upon to examinematerial as well as experiential, collective as wellas individual conditions of social life. While thisambiguous status is in evidence in the two para-digms of media research,� Habermas suggestedthat social-scientific inquiry does have a distinc-tive knowledge interest, at least potentially,namely emancipation. By performing a critiqueof the prevailing forms of social organization,and by clarifying alternatives, the social sciencescan promote the emancipation of humans from

living conditions that are not of their own mak-ing. (Media example: Participatory models ofcommunication.)

These three forms of knowledge interestmust be understood as ideal types that are sub-ject to variation and combination in scientificpractice. However, Habermas (1971 [1968])further argued that the different methodologicaland theoretical requirements do not transferwell from one domain of research to another. In particular, he concluded that the emancipa-tory potential of social sciences tends to be lostif one imports, and gives priority to, the ‘techni-cal’ knowledge interest of the natural sciences.The argument is familiar from some mediaresearch which has depicted audience surveysand television meters as (quantitative) means ofcultural control (e.g., Ang 1991).

It is important to emphasize that the elementof critique does not follow from the political-ideological attitude of the individual researcher.While a commitment to the emancipation of specific socioeconomic groups will be thetypical personal motivation, the distinctive fea-tures of critical research are found in its prac-tices, its epistemologies, and in the institutionsensuring its relevance to the rest of society. A competent critical study, thus, adheres sys-tematically to particular methodological andtheoretical approaches that are likely to havean emancipatory potential. Critical research isalso concerned with researchable, rather thanmerely debatable or normative issues.

Sectors of research

The different intellectual currents which may besummed up, for convenience, as knowledgeinterests are found to varying degrees in thesocial institutions and sectors that perform orrely on media research. From the early begin-nings of the field, an awareness of the differentpurposes of research has been reflected in itsterminology. The classic distinction was intro-duced by Lazarsfeld (1941):

• Administrative research refers to the kind ofgoal-oriented and instrumental studies whichresolve specific issues, typically for the purpose


1111234567891012341115678920123456789301234567894012345678950111

� two paradigms of media research – Chapter 15, p. 255

of planning some media production or activity.Studies in this vein ‘solve little problems, gen-erally of a business character’ (p. 8).• Critical research addresses the wider soci-etal, cultural, and historical issues of mediatedcommunication, often in a reception perspec-tive, from which ‘the public interest’ may beassessed. Here, studies take up ‘the general roleof our media of communication in the presentsocial system’ (p. 9).

When Lazarsfeld (1941) described the criticalvariety of research, he did so, in part, under theinfluence of the ‘first generation’ of FrankfurtSchool scholars who had fled Nazi Germany for the U.S.A. While highly suspicious of the‘culture industry’ (Adorno and Horkheimer1977 [1944]) they encountered there, theirresponse went beyond a normative rejection.One of the points that they introduced to mediastudies was an analytical, Kantian notion of critique that seeks to explicate the conditions of belief, which are themselves one of the con-ditions of the social status quo (Hammersley1995: 30). By reflecting on the media as theynow exist, and by uncovering alternatives, crit-ical studies outline what might be. In thisregard, Lazarsfeld recognized the creative,theoretical potential of critical research. Haber-mas, who is normally seen as the main repre-sentative of a ‘second generation’ of the Frank-furt School, in his turn specified critique as oneof several knowledge interests.

When making the distinction between criti-cal and administrative research, Lazarsfeld(1941) found that the two types, largely syn-onymous in his description with basic andapplied research, could and should cross-fertil-ize. His own accomplishments, centered in theBureau of Applied Social Research at ColumbiaUniversity, seemed to suggest as much. In addi-tion to early ‘milestones’ in media research, heand his collaborators pioneered several generalapproaches, from the panel method to focusedinterviewing. Many critical researchers, how-ever, including European expatriates who, likeTheodor Adorno, found a temporary home inBureau projects, were highly unsympathetic tothe implications of administrative research(Delia 1987: 52). Commercial and other instru-

mental interests might narrow the theoreticalscope of projects, curtail their later uses, and,in the long term, undermine the intellectualfreedom of researchers to choose their researchquestions and methods. Readers of the last sen-tence of Lazarsfeld’s article may have felt con-firmed that critical research was being assignedthe role of generating bright ideas to beexploited (financially and ideologically) in theadministrative mainstream of research: ‘there is here a type of approach which, if it wereincluded in the general stream of communica-tions research, could contribute much in termsof challenging problems and new conceptsuseful in the interpretation of known, and inthe search for new, data’ (Lazarsfeld 1941: 16).

On closer examination, the two varietiesexhibit similarities as well as differences, andhave combined in various research traditionsand organizations. Both rely on qualitative aswell as quantitative methodologies. (The mar-ginalizing of qualitative studies as preliminarypilots, perhaps surprisingly, seems more preva-lent in academic contexts.) In both cases, more-over, projects may be reactive or proactive,evaluating what already is, or shaping what isnot yet. Critical projects can be the most instru-mental of all, since they design research ques-tions and methods, for example, to exposeinequalities in the availability of communica-tion resources, or to develop such resources.Recently, researchers within cultural studieshave advocated more focused social uses of thistradition in policy contexts (Bennett 1992), anda greater reliance on quantitative evidence aswell (Lewis 1997).

In all cases, research projects and their find-ings should be assessed with some reference totheir social infrastructure – their funding, organ-ization, time frames, and anticipated uses –over and above their theoretical models andmethodological approaches. This infrastructureconditions the reflexivity which researchersmay exercise on behalf of themselves, theircolleagues or clients, particular sociopoliticalconstituencies, or the public at large. Despitenational and cultural variations, it is possibleto identify certain main types of media researchinstitutions, as displayed in Figure 16.1.

A central divide separates private enterprise


the FrankfurtSchool

reactive andproactiveresearch

and public service, also in the world of research.This is suggested by the first two types – uni-versity departments and commercial companies.Although reliable measures of the relative sizeof each of these main sectors are difficult to cal-culate, it is safe to say that commercial projectsoutdistance academic ones in terms of bothfinancial resources and the number of singlestudies. Any simple divide between public andprivate research, however, is complicated by thefact that university departments, in many coun-tries, increasingly depend on commercial spon-sorship to fund their research. Furthermore,commercial research is frequently subcon-tracted outside the media organization in ques-tion, sometimes to academic institutions. Theresearch entities of public service media occupyan additional middle ground.

The third type – independent researchinstitutes – has been a staple feature of mediaresearch since Lazarsfeld’s Bureau, avoidingsome of the negative connotations of both

‘state’ and ‘market’ and attracting clients fromboth sides of the divide. The fourth type – doc-umentation centers – has more commonly beenassociated with historical, arts, and otherhumanistic archives than with empirical re-search on contemporary culture and society(although some film institutes have filled thisrole). At present, such entities are gaining impor-tance, both as a strategic resource in media pro-duction and planning, and as support for theaffiliated research activities.�

(It is worth adding that public service broad-casting and public domain research representcomparable conceptions of the social organiza-tion and dissemination of knowledge. In bothcases, knowledge is understood as a ‘publicgood’ (Samuelson 1954), in relative autonomyfrom market forces. In the case of broadcasting,it is this understanding that has been challenged


1111234567891012341115678920123456789301234567894012345678950111 � museums and archives for media research – p. 285

Funding

Organization ofresearch activity

Time frames

Anticipated usesof results

Examples

Commercialcompany

Income fromclients

Managementhierarchy

Days to years

Strategicplanning andproductdevelopment

Marketingsections;Advertisingagencies;Consultancies

Universitydepartment

Public funding

Autonomousresearcherswithincollegialgovernment

Years to decades

Descriptionand critique ofpast andpresent mediaforms

Media studiesdepartments;Schools ofcommunication

Independentresearch institute

Commercial incomeand/or publicfunding

Board of trusteesand managementhierarchy

Days to decades

Descriptive as wellas proactiveanalyses

Research bureausand ad hoc centers;Thinktanks

Documentationcenter

Commercial incomeand/or publicfunding

Board of trusteesand managementhierarchy

Years to centuries

Description anddocumentation ofmedia contentsand uses

Archives withproprietary and/orpublic (museum)access

Figure 16.1 Types of media research organization

under the heading of deregulation (see, e.g.,Blumler and Gurevitch 1995; Garnham andLocksley 1991).)

In the end, it is the different time frameswhich, most of all, distinguish the social roles ofeach type of research organization. Whereascommercial projects typically are scheduled for short-term instrumental purposes, academicstudies may suggest a course of action in the(very) long term. If research is defined summar-ily as the representation of reality for a purpose,the practical question becomes when, where, andhow this purpose is enacted. Short- and long-term purposes may also be expressed in terms ofeither policy or political uses of research.

POLITICS VERSUS POLICY

Policy contexts

Policies are codified plans of action. The impor-tance of policy in both public administrationand commercial companies is one structuralconsequence of increased complexity, internallyin modern organizations as well as in the larger social context. Collective and coordi-nated action requires deliberation and plan-ning, and, because of their scale and cost, theresulting policies further call for evaluation andadjustment. Both the nature of the deliberationsand the criteria of evaluation follow largelyfrom predefined organizational goals. Accord-ingly, policy research is focused within existinginstitutions, and on agendas set by those insti-tutions. The area has been growing since 1945,one key figure being the communicationscholar, Harold D. Lasswell (e.g., Lerner andLasswell 1951). The expanding sector of eval-uation research may be understood as one sub-variety of policy studies in this broad sense(e.g., Patton 1990).

From its inception, media research has con-tributed to planning and evaluating the media’sperformance. Because these uses are relativelyfamiliar, they can be described more brieflythan the following types. At least three policycontexts can be identified:

1 Business administration. Within privateenterprise, practically all media employ in-

house as well as commissioned research tosupport their business. Studies address not onlyaudiences, but also the internal development ofcontent and the strategic placement of theorganization in relation to competitors, regula-tors, and the general public (e.g., Grunig 1992).

2 Public planning. Compared to the specificpolicies of media businesses, public policy de-lineates the general framework in which mediaoperate. A typical arena of influence for mediaresearchers has been commission work leadinginto decisions within the political system, assometimes supported by specially fundedstudies. To exemplify, most European countriesduring the 1980s and 1990s witnessed a greatdeal of commission work and research regard-ing satellite and cable technologies and theirimplications for public service broadcasting.

3 Non-governmental organizations. Beyondand between the business and state contexts,citizens’ groups, thinktanks, and other organi-zations regularly develop or advocate particu-lar media policies. They do so with a view notonly to legislative frameworks, but also to therole of media, for example, in the educationalsystem. In addition to commissioning research,these organizations serve as audiences for polit-ically motivated and socially concerned mediastudies.

It is not by coincidence that the three policycontexts correspond to elements in the domi-nant model of contemporary society, as laid outin Figure 1.3 – the spheres of private businesses,state agencies, and civil society. It is by engag-ing these institutions that research can addressthe structural conditions of meaning produc-tion. It is in these main contexts, for better orworse, that the future of mediated communi-cation is being shaped most concretely.

Political processes

A second set of approaches to applying mediaresearch socially bracket present institutionalagendas and look to the future. Compared to thedelimited contexts of policy, these approachesshift the emphasis toward less well-defined, but potentially more far-reaching processes of


evaluationresearch

organizationalcommuni-cation

statecommissions

change. (As in the case of policy contexts, bothqualitative and quantitative methodologies areof relevance for such political processes.)

By insisting on the autonomy of the researchinstitution, and by resisting a hegemony ofother institutional logics, much academic workmay be said to adopt a long-term strategy ofinterchange with other social institutions. Insome cases, the strategy entails the counteringof specific policies with alternative politicalviewpoints. For example, media studies mayexpose the hidden or unacknowledged interestsof either commercial or government policies. Inthis regard, academic media and communica-tion researchers carry on aspects of the classic,critical role of the intelligentsia (Mannheim1976 [1922]: 136–146).

Corner (1991) has identified two intellectualprojects in recent media research, originallywithin reception analysis, but with equal rele-vance for other areas of study. On the one hand, the field as such has been committed to Enlightenment ideals concerning the demo-cratic accessibility of public knowledge throughvarious factual genres. From propaganda re-search to decoding studies of news, an import-ant research question has been how wellaudiences are able to process mediated infor-mation, and to employ it in the political process.On the other hand, the last couple of decades inparticular have witnessed both textual andaudience research rehabilitating the value andrelevance of popular culture, especially fictiongenres.

Compared to the policy contexts ofresearch, its wider political arenas of influenceare centered around the public sphere as aforum of social reflexivity and intervention. Inaddition, the interventions of media researchaddress related institutions within, for instance,education and politics:

• Public debate. Most generally and uncon-troversially, media research contributes to (andoccasionally initiates) debates in the publicsphere, its political as well as cultural compo-nents (Figure 1.3). The contributions rangefrom popular publications at the conclusion ofa project to syndicated commentaries. In theprocess, researchers may promote the general

self-reflexivity of media, as they address issuessuch as political ‘spin doctors’ or lifestyleadvertising. In a case such as the public jour-nalism movement in the U.S.A., a more ambi-tious aim has been for research to support areinvigoration of both the press and politicalparticipation (for an overview and referencessee Haas 1999).

• Media education. Beyond their own gradu-ate and undergraduate students, media andcommunication researchers have contributed tothe democratization (or relativization) of thecultural standards and ‘texts’ of curricula atmost educational levels. In addition, the fieldhas been successful, in a number of countries,in arguing the need for a component of ‘medialiteracy’ in general education (Masterman1985; Messaris 1994; Potter 1998). This is inspite of the fact that the exact purpose andplacement of media education (as a separatesubject or within other subjects) remaindebated. With the introduction of computer-supported learning, a redefinition of (media) lit-eracy is again likely to occur.

• Museums and archives. As suggested inFigure 16.1, documentation centers constitutean increasingly strategic resource for mediaproduction as well as research. Also from polit-ical and public perspectives, the preservationand documentation of contemporary mediapose important issues (Jensen 1993b). Thepoint is not only to enable future scholars to(re)write media history, or to assess contempo-rary research models and findings. Only if thebreadth and depth of media, including theireveryday uses and audience experiences, remainavailable and documented – alongside the highcultural forms that still reign supreme inmuseums and archives (and among employedarchivists) – will coming generations have thepossibility of assessing and learning from theirpast, our present. The Payne Fund (Jowett etal. 1996) and Mass Observation (Richards andSheridan 1987) studies of the 1930s providedindications of the kinds of evidence needed. Thechallenge has been taken up in at least somerecent work (e.g., Day-Lewis 1989; Gauntlettand Hill 1999).

Politics versus policy 285

1111234567891012341115678920123456789301234567894012345678950111

critical role of the

intelligentsia

popularcultureproject

publicknowledge

project

publicjournalism

media literacy

Standpoint interventions

A third, heterogeneous group of research strat-egies – beyond policy and politics – are unitedby an explicit commitment to social and culturalchange, perhaps on a revolutionary scale. Thisis in comparison to the more reformist orienta-tion of the political processes reviewed in thelast section. (Because they tend to explore polit-ical and epistemological alternatives, studieshere rely largely on qualitative approaches.)

The orientation of these strategies might be defined as being versus institutions – theynot only bracket social agendas, but activelyoppose major social institutions and, occasion-ally, the very institutionalization of society.Furthermore, their analyses sometimes call fora radical break with the methodological andepistemological premises of other research. Afrequent implication is that alternative socialarrangements must be outlined through alter-native epistemologies. As theorized most influ-entially by Foucault (1972), this implicationsometimes leads on to a stronger position ofquestioning the legitimacy of any and all formsof knowledge.

Within media research, it is especially thecultural studies tradition� which has advancedthese issues and generated debate with otherresearch traditions (e.g., Ferguson and Golding1997). Since the 1970s, much work has takenprevious social and cultural research to task forarticulating and promoting interests associatedwith the Western world, the economic middleclass, the political mainstream, and with a per-vasive patriarchal mode of social interaction.One counter-strategy has been to treat know-ledge in the plural, also terminologically, ex-ploring alternative ‘knowledges’ in the interestof the disempowered.

Some of these debates have crystallized inthe notion of ‘political correctness.’ The impli-cation is that research questions, findings, andexplanatory frameworks are perhaps not beingassessed by their scientific or professionalmerits, but by their immediate relevance andlegitimacy, for instance, in making up for pastand current silences and injustices in the wider

social system (for an overview see Levy 1992).Against such more traditionalist concerns, crit-ical researchers have reiterated that neither‘professionalism’ nor ‘science’ are innocent cat-egories. Asante (1992: 141), for one, has coun-tered that the charge of ‘political correctness’represents ‘a peculiar mixture of notions aboutfreedom of speech, affirmative action, sexism,and the decline of white male privilege in a plu-ralistic society.’

At least three types of interventions involvea concerted commitment to social changethrough research.

Feminist methodologyThe term ‘standpoint’ has been related tomethodology and theory of science by feministresearch, as elaborated especially by Harding(1986). The point of departure is that all know-ledge is produced from a socially situated stand-point, and that the life experiences of women –arguably silenced in much of the history of ideas– provide a necessary corrective to other science.Thus, feminism might enable a ‘strong objectiv-ity’ by allowing both women and men to tran-scend classic canons of objectivism and to takea (more) reflexive position in the researchprocess (see also Alcoff and Potter 1993;Harding 1987).

Compared to the biological essentialism ofIrigaray (1997 [1977]) and others, standpointfeminism represents an attempt to historicizethe nexus between knowledge and power. Inpractice, however, the pendulum swings easilyto another extreme of ‘sociological essential-ism.’ Some writings seem to suggest that femi-nism and other cognate traditions, standing onthe shoulders of the disempowered, necesssar-ily offer more insightful theories as well asbetter empirical bases of change. In mediaresearch, this tendency is found in work assert-ing that, compared to certain recognized clas-sics, studies of women’s culture, of ethnicminorities, and of marginalized youth subcul-tures not only have a contribution to make, butare unrecognized origins of key ideas regardingthe place of media in everyday life (Drotner1996: 41).

The orientation of feminist methodologyand epistemology is toward the very long term.


� cultural studies – Chapter 2, p. 39

politicalcorrectness

standpointepistemology

By reshaping the research institution, feministstudies may, in turn, help to reshape those othersocial institutions that are affected by research.In doing so, feminist research links up with thewider feminist movement and its struggle forequal rights.�

Textual deconstructionismThis second position for change shares withmuch feminism the ambition of challenging aunified concept of knowledge. Deconstruc-tionism, however, is almost entirely focused ontexts as the locus of both dominance andchange. Departing from poststructuralist theo-ries of discourse and postmodernist conceptionsof culture,� many textual media studies of, forinstance, film and television seek to exposeeither misrepresentations or reified representa-tions of reality.

Neither empirical audiences, however, northe concrete social contexts in which texts takeeffect, have much of a place in these studies.Instead, deconstructionist works may be seento perform an interpretive reworking of mediatexts, and to offer their reinterpretations to no constituency or context in particular. Atleast some key publications leave readers withthe impression of an inverse relationship be-tween quite revolutionary ambitions and lessthan concrete political strategies (e.g., Kristeva1984). (For a critique of the associated termi-nology concerning the ‘social construction’ ofreality, see Hacking 1999.)

Deconstructionism dissolves boundaries,not only between text and reality, but betweenthe text being studied and the study itself. Likesome art criticism, media criticism in this veinmight be understood as another artwork in itsown right. The work of Richard Rorty has been influential by promoting a definition ofscience as story-telling (e.g., Rorty 1979, 1998).Research may be taken as one of many contri-butions to ‘the conversation of humankind.’Some recent qualitative work has also proposedto expand the range of genres in which sciencetells stories (e.g., Denzin and Lincoln 2000).

Given its disregard for other social practices,textual deconstructionism is most likely tomake an impact within the educational system,which may trickle down or disseminate theideas. A second area of impact is the mediathemselves, where notions of postmodernismhave been widespread, and where programdevelopers have taken specific inspirations fromdeconstructionism (e.g., Caldwell 1995).

Action researchA third type of intervention may be describedbriefly as action research, although the tradi-tion recognizes several participatory andapplied variants (for overviews see Greenwoodand Levin 1998; Reason and Bradbury 2001).Like the two positions above, action studieschallenge established institutions, normally in cooperation with, or initiated by, interestgroups within the institution. Unlike the othertypes, action research presents an operationalstrategy for social change, for example, withinindustrial democracy or community develop-ment. Ideally, studies become ‘co-generative’(Greenwood and Levin 1998: 109) by having‘everyday theorists’� as partners in the formu-lation of research questions as well as empiri-cal strategies.

In media and communication studies, actionresearch has apparently been less widespreadthan in some other fields. However, ‘develop-ment communication’ programs included someinvolvement of the communities and culturesthat were end-users of the technologies beingdiffused.� More recently, some communitymedia work (e.g., Jankowski 1991) and studiesof computer-mediated communication (e.g.,Preece 1994) have relied on participatory strategies.

Like other political and policy strategies,these various ‘standpoint’ interventions leave anumber of political and epistemological ques-tions to be addressed in future research prac-tice and debate. To conclude, the final sectionpresents some of the additional practical as well as ethical considerations which arise in

Politics versus policy 287

1111234567891012341115678920123456789301234567894012345678950111

� feminism – Chapter 2, p. 34

� poststructuralism and postmodernism – Chapter 2, p. 33

� everyday theory – p. 275

� development communication – Chapter 11, p. 179


RESOURCE BOX 16.2 KEY STUDIES AND REFERENCE WORKS FOR INDIVIDUAL MEDIA

Books• Vincent 2000 – a comprehensive overview of the development of literacy in modern

Europe, with reference to books and other media of communication, and to their socialuses;

• Radway 1997 – a recent study of books as a medium that continues to serve as a centralingredient of popular culture.

Newspapers• Habermas 1989 [1962] – still an essential resource regarding the historical development

and current functions of the press;• Schudson 1978 – a classic social history of the U.S. press, in this respect complementing

the European focus of Habermas;• Curran and Seaton 1997 – a historical analysis of the British press, with many references.

Film• Andrew 1976 – a solid introduction to the main classic film theories;• Mast et al. 1990 – an anthology of classical texts in film studies, covering the range of

traditions;• Bordwell and Carroll 1996 – an important intervention contrasting cognitive and other film

theory.

Radio• Crisell 1994 – one of the first comprehensive introductions to an under-researched

medium;• Hendy 2000 – another broad overview of the medium, including its changing technologies

and its place in, for instance, African and Latin American cultures;• Scannell and Cardiff 1991 – an exemplary social history of radio, with implications for

current studies of broadcasting.

Television• Williams 1974 – the classic study which defined television (and radio) in terms of their

characteristics of ‘flow’;• Ellis 1982 – a redevelopment of the understanding of television flow, and a comparison of

television with film, also in terms of the audience experience;• Newcomb 1997 – a three-volume encyclopedia and study resource on most aspects of

television.

Computer media• Mayer 1999 – a collection of key texts contributing to the definition and study of the

computer as a medium;• Castells 1996 – a comprehensive analysis of the relations between information and

communication technologies and the social and cultural practices which they enable.

planning both student projects and researchprograms.

THE SOCIAL TRIAD OF RESEARCHPRACTICE

Research practice may be understood as a par-ticular kind of rule-governed social interaction.It involves, minimally:

• the researcher• his/her respondents (or other sources of

evidence)• the community of colleagues who, sooner or

later, will assess the quality of findings andof professional conduct.

This triad operates across all the differentsectors and intellectual cultures of mediastudies. It cumulates a body of theory and evi-dence, and it enacts the institution of researchon a daily basis.

Most of this volume has, in the nature ofthe matter, emphasized the researcher’s per-spective as a professional and social subject.The encounter with ‘ordinary’ people, and withcolleagues, presents additional and concretechallenges and choices. Such encounters acutelyinvolve the perspectives and rights of others.

A common denominator for researchers’relations with respondents and colleagues isresearch ethics (for an overview see Resnik1998). The category of ethics blends intopolitics: politics formulates rules of collectiveconduct in society, whereas ethics addressesstandards of individual conduct. The ethics thatapply to a given domain of social activity,further, overlap with the more general stan-dards of morality that are prevalent in a his-torical period and cultural setting.

Some aspects of research ethics are codifiedin legislation (as is politics and morality), andare subject to enforcement by legal as well asexecutive authorities. Whereas the ethical andlegal frameworks of media and communicationresearch vary between countries, the mainissues may be laid out, as they relate to researchsubjects and the research community (see alsoDeacon et al. 1999: 365–387; Priest 1996:207–230).

Research subjects

To begin with a seemingly trivial premise,research subjects are just that – subjects. Unlikerocks or books, subjects may be harmed a greatdeal, both socially and emotionally, by empiri-cal media studies. And, while self-determinationis the prerogative of any (research) subject, it isthe responsibility of the researcher to anticipateand prevent harm. Research competence thusrequires an awareness of ethical pitfalls, asestablished in previous research, as well asempathy and respect. (Perhaps surprisingly,organizations such as the International Associa-tion for Media and Communication Researchand the International Communication Asso-ciation have not developed codes of researchethics.)

In preliminary terms, the requirements maybe stated as practical rules of thumb: ‘Do asyou would be done by’ and ‘Leave things asyou find them’ (Deacon et al. 1999: 385). The first principle is a variation on the Kantiancategorical imperative, which accepts a stan-dard of conduct only if it amounts to areversible or general rule. The second principlespecifies the right of research subjects to self-determination. Researchers should not inter-vene proactively unless this has been part of anexplicit agreement as in, for instance, actionresearch.

Depending on the research question and theconcrete field of analysis, empirical researchersface several dilemmas. First, ‘harm’ may be ofdifferent types, and may only manifest itself inthe longer term. Although empirical mediastudies rarely become life-threatening, com-pared to, for instance, medical research, the dis-closure of, for instance, documents concerningproduct development from within a mediaorganization may produce significant losses,financially or in terms of legitimacy. Similarly,the publication of politically charged debateswithin focus groups without sufficient anonym-ization may result in participants losing socialstatus or ‘face’ in their community.

Second, these two examples suggest howresearch subjects have different potential levelsof vulnerability, and should normally be treatedaccordingly. In their reception study of women

The social triad of research practice 289

1111234567891012341115678920123456789301234567894012345678950111

researchethics, politics,

morality

the categoricalimperative

viewing mediated violence, Schlesinger et al.(1992) exercised special care in screening anddebriefing their respondents, some of whomhad been physically abused in their own lives. By comparison, in production and otherorganizational studies, informants are normallyapproached as representatives of an organiza-tion or profession, who are aware of this roleand, to a degree, of the nature of research.Nevertheless, ethical dilemmas also ariseregarding what information should be providedor withheld in the attempt to complete a study.A case in point is the Glasgow University MediaGroup’s undercover work within British tele-vision (1976, 1980).

A standard procedure in professional codesof ethics is ‘informed consent’ (Resnik 1998:133). Its purpose is to enable subjects to agreeor decline to participate in a study. Their deci-sion should be based on information about thecomponents of the study, its potential conse-quences for themselves, and its likely socialuses. Informed consent is one key element ofthe Nuremberg Code (1949), a set of protocolsfor research on human subjects which wasestablished during the Nuremberg Trials onNazi war crimes that also included scientificexperimentation. While debated as to its suffi-ciency and practical implementation, informedconsent represents an important and applicableprinciple concerning what (not) to do withresearch subjects. Compared to the procedureof ‘double-blind’ experiments in, for instance,medical science (neither patient nor therapistknows who gets the active drug being tested,and who gets the placebo), informed consentmay be said to aim for a procedure of ‘doubleinsight’ involving, ideally, both sides.

In next reporting their studies, researchersface issues of confidentiality in general, andanonymity in particular. These issues may bestated in terms of information and communi-cation theory. Most important, it is the ethicalobligation of empirical researchers to preserverespondents’ anonymity by withholding infor-mation. In quantitative studies, a potentialabuse is the recycling and recombination ofseveral datasets, so that individuals may be tar-geted in subsequent marketing campaigns. Inqualitative research, harm is more likely to

result because readers of a report are able toidentify a particular individual through a rich, contextual description. In both of thesecases, the problem is not so much that the infor-mation is publicized. (Such problems arisewhen information is proprietary, typically forcommercial reasons, and hence confidential.)The problem is that the information can belinked to its original source. In fact, neitherqualitative nor quantitative methodologiesdepend for their explanatory value on such a link being made. From the researcher’s perspective, the source is of interest not as aunique entity with biographical (and biological)characteristics, but as a social and cultural prototype, or as a representative specimen of a social segment.

From the sources’ perspective, they have aright not to be associated with the informationthey offered in the context of research. Theirright may be understood as a ‘reverse copy-right’: the social contract of most media studiesimplies that research subjects speak as types,not tokens – as anybody, not somebody. Theprinciples of reverse copyright, and of doubleinsight, can be understood as responses to theambivalent status of media and other social andcultural research between nomothetic and idio-graphic research – between laws and cases.�

Analysis box 16.2 outlines a set of guide-lines for empirical student projects, which alsohave relevance for the planning of largerresearch programs.

The research community

Another set of issues is located along the secondleg of the research triad, which connectsresearchers with their professional peers. As an abstract ideal, science calls for the completedissemination of all potentially relevant infor-mation among the international community ofscholars. As an interested social practice,research requires that individual scholars weighthis ideal against, for instance, anonymityrequirements, but also against more materialconsiderations, such as intellectual property


informedconsent

double-blindand double-

insightprocedures

confidentialityand anonymity

reversecopyright

� idiographic and nomothetic research – Chapter 15, p. 255

rights and the protection of their own careers ortheir clients. The research community is itself asocial system of checks and balances, privilegesand sanctions. The chief issues in this area canbe reviewed with reference to the differentstages of the research process.

Especially in academic research, an earlyand decisive juncture is the approval of anempirical project by a national or institutionalreview board (IRB). In some countries, suchapproval is required before any study can beundertaken with human subjects, and officiallyon behalf of a university. While debated alongthe same lines as the informed consent proce-dure, an initial review process provides someassurance that gross ethical misconduct will

not occur. Next, the basic criteria of ethicalresearch include intellectual honesty in the pre-sentation of the sources of ideas, a completeaccounting of successes and failures in datacollection and analysis, and a systematic docu-mentation of evidence and the bases of theo-retical inference (see Resnik 1998: 53–95). Anoverriding concern is the link between the con-texts of discovery and justification,� explicat-ing and substantiating the relationship betweenresearch as process and as product.

At their conclusion, most research projectsface a second review, typically in the form of a

The social triad of research practice 291

1111234567891012341115678920123456789301234567894012345678950111

ANALYSIS BOX 16.2 TEN RULES FOR EMPIRICAL STUDENT PROJECTS

1 First make sure to ascertain the rules and procedures in your social and cultural contextand academic institution regarding review of research involving human subjects.

2 Always treat the people under study as people.They are neither things nor texts. Astandard procedure for ensuring their rights and preventing harm is ‘informed consent.’

3 Exercise caution and concretion. Be prepared to give up a question (or an entire study) if, incontext, it violates the ethical, cultural, or personal limits of the people involved. Beprepared to explain concretely the relevance of any question to informants and others.

4 Practice reflexivity.The analysis of (cautiously collected and concrete) data begins in theempirical field. In qualitative as well as quantitative projects, supplementary evidence andnotes will support both the respectful use of respondents’ contributions and theexplanatory value of later interpretations.

5 Safeguard the anonymity of people and the confidentiality of information throughout theresearch process.

6 Be honest about the sources of ideas informing a study and the contributions of peers indeveloping and conducting it.

7 A research report includes accounts both of process and outcome, and of successes andfailures in each respect.

8 Two key requirements of a research report are a systematic documentation of evidence andan explication of the bases of theoretical inference.

9 Explore several different publication formats, including a means of feedback to the peoplecontributing to a study.

10 Consider what’s next – further research, the social relevance of findings, and the possibleunanticipated consequences of the research.

institutionalreview boards

� contexts of discovery and justification – Chapter 15,p. 258

peer review – an anonymous (double-blind)evaluation by experienced researchers. This hasbeen standard procedure in many fields sincethe mid-twentieth century, and determineswhether a study will be published in majorjournals. Once again, peer review has been theobject of criticism, for example, for favoringentrenched traditions. Still, the procedure pre-sents itself as one of the least worst alternativesin the inevitably controversial enterprise ofevaluating the standards of research projects. In a next step, access to the empirical datasetsof other researchers is a way of keeping theresearch community critically reflective and indialogue. An example of debates arising fromsuch a secondary data analysis was Hirsch’s(1980, 1981) questioning of Gerbner and col-leagues’ cultivation hypothesis.�

Despite such specific conflicts, and despitestand-offs between qualitative and quantitativetraditions, it may be concluded that mediastudies have entered the new millennium as afield in communication with itself. Certainly,what is sometimes denoted ‘the culture wars’over the aesthetic ideals of traditional Westernhigh art has continued to occupy and occa-sionally divide media studies. Moreover, ‘thescience wars,’ which flared up following AlanSokal’s hoax in a 1996 paper (reprinted inSokal and Bricmont 1998) – the publication ofa deliberately nonsensical article in a majorjournal in order to expose the lack of scientificrigor in postmodern cultural studies and relatedtraditions – have been in evidence in some ofthe political and epistemological debates out-lined in this chapter. In both types of ‘wars,’however, some of the most vehement interven-tions have originated within the traditions inquestion. Morris (1990), for one, put culturalstudies on trial for its banality, its solemntextual paraphrasing of the fact that culturesand societies are complex and contradictoryphenomena. Ritchie (1999), for another, ques-tioned the validity of much survey and otherquantitative research because of its ambiguousconcepts of ‘probability.’

Intellectual conflict with social implicationsis part of the ongoing business of media and

communication research. A recent internationalexample centered on the work of Noelle-Neumann (1984) on ‘the spiral of silence.’ Thepoint is that if people perceive their views to bein the minority, as represented, for instance, inthe media, they are less likely to state theseviews. As a result, they participate in a circleor spiral that is potentially vicious for democ-racy. One of Noelle-Neumann’s specific argu-ments has been that a predominance of left-wing views among German journalists andmedia has helped to silence right-wing politicalviews in that country.

Noelle-Neumann herself has been quiteexplicit regarding her right-wing political posi-tion and her work as a strategy adviser for theGerman Christian Democratic Party. An acri-monious debate, however, began when Simpson(1996) drew attention to her apparent sym-pathies with the Nazi party during World WarII. Most centrally, Simpson linked her conclu-sions in the present with a set of theoreticalassumptions and methodologies which wereoriginally developed for her research during thewar. In a similarly fierce response, Kepplinger(1997) suggested, in essence, that the critiquewas ad hominem. His counter-argument wasthat the quality of methodologies as well as find-ings can be judged independently of their his-torical origins and contemporary applications.

This particular debate may be especiallycontroversial and painful against the back-ground of world war atrocities. However, theunderlying questions are of a general nature.They concern the relations between knowledge,interest, and power. They have no simpleanswers, and will require continuous engage-ment in media studies.

THE END OF COMMUNICATION

The political and epistemological issues of‘doing research’ may be summed up in a fantasyinvolving two key figures in American sociol-ogy: C. Wright Mills and Paul F. Lazarsfeld(cited in Gitlin 1978: 223). The first sentenceof Mills’s The Sociological Imagination (1959)reads, ‘Nowadays men often feel that theirprivate lives are a series of traps.’ The fantasyimagines Lazarsfeld as replying: ‘How many


peer review

� cultivation research – Chapter 9, p. 15

culture wars

science wars

men, which men, how long have they felt thisway, which aspects of their private lives botherthem, do their public lives bother them, whendo they feel free rather than trapped, whatkinds of traps do they experience, etc., etc., etc.’

What these two different scientific andpolitical temperaments arguably share, never-theless, is a realization that research respondsto contemporary concerns and realities. Mills’ssweeping statements and Lazarsfeld’s moremundane operationalizations both assumedthat, by describing and interpreting social andcultural conditions, research can make a differ-ence in social practice.

The orientation toward social action issomething that research shares with communi-cation. Both media research and mediated com-munication have ends, whether explicitly orimplicitly. This realization is an importantcorrective to one of the most influential con-temporary theorists of communication, JürgenHabermas. Following his early historical work,Habermas’s (1984 [1981]; 1987 [1981]) latertheory of communication came to insist on anabstracted ‘ideal speech situation’ as the sourceof genuine human interaction and insight.

What Habermas came to neglect is the fact thatall communication inevitably comes to an end(see further Jensen 1995: 185–191).

Political democracy is a case in point. Assuggested by Schudson (1997), it is necessaryto distinguish between ‘sociable’ and ‘problem-solving’ communication in modern societies.‘As far as democracy is concerned, the responseto the familiar rhetorical question, “Can’t wejust talk about it?”, should be: No’ (Jensen2001: 94). It is the conclusion of mediated com-munication and its regulated transformationinto concerted social action that is the hallmarkof democracy, not an interminable process ofcommunication. The end of communicationserves the ends of democracy.

A similar argument may be made for researchabout mediated communication. Researchinevitably comes to an end. The end of theresearch process is the beginning of other socialpractices. By keeping itself aware of its origins,its uses, and its potential unanticipated conse-quences, media and communication researchcan finally claim the status of a scientificallymature and socially relevant field of study.

The end of communication 293

1111234567891012341115678920123456789301234567894012345678950111

Aarseth, E. J. (1997) Cybertext: Perspectives onErgodic Literature, Baltimore, MD: The JohnsHopkins University Press.

Abercrombie, N. and Longhurst, B. (1998) Audi-ences: A Sociological Theory of Performance andImagination, London: Sage.

Abrams, M. H. et al. (1962) The Norton Anthologyof English Literature (Vol. 2), New York: Norton.

Adams, W. C. (ed.) (1982) Television Coverage ofInternational Affairs, Norwood, NJ: Ablex.

Adorno, T. and Horkheimer, M. (1977) ‘The CultureIndustry: Enlightenment as Mass Deception,’ in J. Curran, M. Gurevitch, and J. Woollacott (eds),Mass Communication and Society, London:Edward Arnold [orig. publ. 1944].

Ahern, T. J. (1984) ‘Determinants of ForeignCoverage in U.S. newspapers,’ in R. Stevenson andD. L. Shaw (eds) Foreign News and the NewWorld Information Order, Ames, IA: Iowa StateUniversity Press.

Alasuutari, P. (ed.) (1999) Rethinking the MediaAudience: The New Agenda, London: Sage.

Alcoff, L. and Potter, E. (eds) (1993) FeministEpistemologies, New York: Routledge.

Alexander, A., Owers, J., and Carveth, R. (eds)(1998) Media Economics: Theory and Practice,Mahwah, NJ: Lawrence Erlbaum Associates.

Alexander, J. and Giesen, B. (1987) ‘From Reductionto Linkage: The Long View of the Micro–MacroLink,’ in J. Alexander, B. Giesen, R. Münch, andN. Smelser (eds) The Micro–Macro Link,Berkeley: University of California Press.

Alexander, J., Boudon, R., and Cherkaoui, M. (eds)(1997) The Classical Tradition in Sociology: TheAmerican Tradition, London: Sage.

Allen, J., Livingstone, S., and Reiner, R. (1997) ‘TheChanging Generic Locations of Crime in Film: AContent Analysis of Film Synopses, 1945–1991,’Journal of Communication, 47 (4): 89–101.

Altenloh, E. (1913) Zur Soziologie des Kino [TheSociology of the Cinema], Leipzig: SpamerscheBuchdruckerei.

Altheide, D. L. (1976) Creating Reality, BeverlyHills, CA: Sage.

Altheide, D. L. and Snow, R. P. (1979) Media Logic,Beverly Hills, CA: Sage.

Althusser, L. (1977) For Marx, London: Verso [orig.publ. 1965].

Altman, R. (1986) ‘Television/Sound,’ in TaniaModleski (ed.) Studies in Entertainment, Bloom-ington, IN: Indiana University Press.

Altman, R. (ed.) (1980) Cinema/Sound, Yale FrenchStudies, 60.

Altman, R. (ed.) (1992) Sound Theory/Sound Prac-tice, New York: Routledge.

Altman, R. (1999) Film/Genre, London: British FilmInstitute.

Alvarado, M. and Buscombe, E. (1978) Hazell: TheMaking of a TV Series, London: British FilmInstitute/Latimer New Dimensions.

Alvesson, M. and Sköldberg, K. (2000) ReflexiveMethodology: New Vistas for Qualitative Re-search, London: Sage.

Anderson, B. (1991) Imagined Communities:Reflections on the Origin and Spread of Nation-alism (2nd edn), London: Verso [orig. publ. 1983].

Anderson, D. R. and Burns, J. (1991) ‘PayingAttention to Television,’ in J. Bryant and D.Zillmann (eds) Responding to the Screen:Reception and Reaction Processes, Hillsdale, NJ:Lawrence Erlbaum Associates.

Anderson, J. (1996) Communication Theory:Epistemological Foundations, New York: TheGuilford Press.

Andrén, G., Ericsson, L. O., Ohlsson, R., andTännsjö, T. (1978) Rhetoric and Ideology inAdvertising, Stockholm: Liber Förlag.

References

Andrew, J. D. (1976) The Major Film Theories: AnIntroduction, New York: Oxford University Press.

Ang, I. (1985) Watching Dallas, London: Methuen.Ang, I. (1991) Desperately Seeking the Audience,

London: Routledge.Anon (1935–1958) The History of the Times (Vols

1–5), London.Appadurai, A. (1996) Modernity at Large: Cultural

Dimensions of Globalization, Minneapolis: Uni-versity of Minnesota Press.

Archer, M., Bhaskar, R., Collier, A., Lawson, T., andNorrie, A. (eds) (1998) Critical Realism: EssentialReadings, London: Routledge.

Aristotle (1994) Categories and De Interpretatione(trans. J. L. Ackrill), Oxford: Clarendon Press.

Arnheim, R. (1974) Art and Visual Perception,Berkeley: University of California Press.

Aron, D. (1998) ‘Towards a Link Between Produc-tion and Reception in Television Documentary,’Paper presented at the 48th Annual Conference ofthe International Communication Association,Jerusalem.

Asante, M. K. (1992) ‘The Escape into Hyperbole:Communication and Political Correctness,’ Jour-nal of Communication, 42(2): 141–147.

Asante, M. K. and Gudykunst, W. B. (eds) (1989)Handbook of International and InterculturalCommunication, Newbury Park, CA: Sage.

Ashcroft, B., Griffiths, G., and Tiffin, H. (eds) (1995)The Post-Colonial Studies Reader, London: Rout-ledge.

Atkinson, P. and Coffey, A. (1997) ‘AnalysingDocumentary Realities,’ in D. Silverman (ed.)Qualitative Research: Theory, Method, andPractice, London: Sage.

Austin, J. L. (1962) How to Do Things with Words,Oxford: Oxford University Press.

Babbie, E. (1990) Survey Research Methods (2ndedn), Belmont, CA: Wadsworth Publishing Com-pany.

Bailey, K. D. (1994) Methods of Social Research (4thedn), New York: The Free Press.

Bakhtin, M. M. (1981) The Dialogic Imagination,Austin: University of Texas Press.

Bal, M. (1985). Narratology: Introduction to theTheory of Narrative, Toronto: University ofToronto Press [orig. publ. 1977].

Ball-Rokeach, S. (1985) ‘The Origins of IndividualMedia–System Dependency,’ Communication Re-search, 12(4): 485–510.

Bambach, C. R. (1995) Heidegger, Dilthey and theCrisis of Historicism, Ithaca, NY, and London:Cornell University Press.

Bandura, A. and Walters, R. (1963) Social Learningand Personality Development, New York: Holt,Rinehart & Winston.

Barbour, R. S. and Kitzinger, J. (eds) (1999) Develop-ing Focus Group Research: Politics, Theory, andPractice, London: Sage.

Barker, M. and Brooks, K. (1998) Knowing Audi-ences: Judge Dredd, Its Friends, Fans, and Foes,Luton, Beds: University of Luton Press.

Barnhurst, K. G. and Wartella, E. (1998) ‘YoungCitizens, American TV Newscasts, and the Collec-tive Memory,’ Critical Studies in Mass Communi-cation, 15(3): 279–305.

Barnouw, E. (1966–1970) A History of Broadcastingin the United States (Vols 1–3), New York: OxfordUniversity Press.

Barnouw, E. (ed.) (1989) International Encyclopediaof Communications, New York: Oxford Univer-sity Press.

Barrow, J. D. (1992) Pi in the Sky: Counting,Thinking, and Being, London: Penguin.

Barthes, R. (1964a) Elements of Semiology, NewYork: Hill & Wang, 1967 [English trans.]

Barthes, R. (1964b) ‘Rhetoric of the Image,’ in R.Barthes, Image, Music, Text, London: FontanaPress, 1984 [English trans.].

Barthes, R. (1966) ‘Introduction to the StructuralAnalysis of Narratives,’ in S. Sontag (ed.) ABarthes Reader, London: Jonathan Cape, 1982.[English trans.]

Barthes, R. (1970) S/Z, Paris: Seuil.Barthes, R. (1973) Mythologies, London: Paladin

[orig. publ. 1957].Barthes, R. (1980) Camera Lucida: Reflections on

Photography, New York: Hill & Wang [repr. in1981].

Batscha, R. M. (1975) Foreign Affairs News and theBroadcast Journalist, New York: Praeger.

Baudrillard, J. (1988) Selected Writings, Cambridge:Polity Press.

Bauer, M. W. and Gaskell, G. (eds) (2000) Quali-tative Researching with Text, Image, and Sound:A Practical Handbook, London: Sage.

Baym, N. K. (2000) Tune In, Log On: Soaps, Fan-dom, and Online Community, Thousand Oaks,CA: Sage.

Bazin, A. (1967–1971) What Is Cinema?, Berkeley:University of California Press.

Bechtel, R. B., Achelpohl, C., and Akers, R. (1972)‘Correlates between Observed Behaviours andQuestionnaire Responses on Television Viewing,’in E. A. Rubinstein, G. A. Comstock, and J. P.Murray (eds) Television and Social Behavior, Vol.4.Television in Day-to-Day Life: Patterns of Use,Washington, DC: U.S. Government Printing Office.

Beck, U. (1999) World Risk Society, Cambridge:Polity Press.

Beck. U. (1994) ‘The Reinvention of Politics:Towards a Theory of Reflexive Modernization,’ in

References 295

1111234567891012341115678920123456789301234567894012345678950111

Beck, U. et al. (1994) Reflexive Modernization:Politics, Tradition and Aesthetics in the ModernSocial Order, Cambridge: Polity Press.

Beck, U., Giddens A., and Lash, S. (1994) ReflexiveModernization. Politics, Tradition and Aestheticsin the Modern Social Order, Cambridge: PolityPress.

Becker, H. (1982) Art Worlds, Berkeley: Universityof California Press.

Behlmer, R. (ed.) (1981) Memo From David O.Selznick, New York: Viking Press [reprint. orig.publ. 1972].

Bell, A. and Garrett, P. (eds) (1998) Approaches toMedia Discourse, Oxford: Blackwell.

Bell, M. M. and Gardiner, M. (eds) (1998) Bakhtinand the Human Sciences: No Last Words,London: Sage.

Bellour, R. (1973) ‘The Obvious and the Code,’ in P.Rosen (ed.) Narrative, Apparatus, Ideology: AFilm Theory Reader, New York: ColumbiaUniversity Press, 1986.

Bellour, R. (1979) L’analyse du film [Film Analysis],Paris: Albatros.

Belson, W. A. (1967) The Impact of Television,London: Crosby Lockwood & Son.

Benedikt, M. (ed.) (1991) Cyberspace: First Steps,Cambridge, MA: MIT Press.

Benhurst, K. and Mutz, D. (1997) ‘AmericanJournalism and the Decline of Event-CenteredReporting,’ Journal of Communication, 47(4):27–53.

Beniger, J. R. (1986) The Control Revolution,Cambridge, MA: Harvard University Press.

Beniger, J. R. (1978) ‘Media Content as SocialIndicators: The Greenfield Index on Agenda-setting,’ Communication Research, 5: 437–453.

Benjamin, W. (1977) ‘The Work of Art in the Age of Mechanical Reproduction,’ in J. Curran, M.Gurevitch, and J. Woollacott (eds) Mass Com-munication and Society, London: Edward Arnold[orig. publ. 1936].

Bennett, T. (1992) ‘Putting Policy into CulturalStudies,’ in L. Grossberg, C. Nelson, and P.Treichler (eds) Cultural Studies, New York:Routledge.

Bennett, T. (1995) The Birth of the Museum,London: Routledge.

Bennett, T. and Woollacott, J. (1987) Bond andBeyond, London: Methuen.

Berelson, B. (1949) ‘What “Missing the Newspaper”Means,’ in P. F. Lazarsfeld and F. M. Stanton (eds)Communications Research 1948–9, New York:Duell, Sloan, and Pearce.

Berelson, B. (1952) Content Analysis in Com-munication Research, Glencoe, IL: The Free Press.

Bergen, L. A. and Weaver, D. (1988) ‘Job Satis-faction of Daily Newspaper Journalists andOrganization Size,’ Newspaper Research Journal,9(2): 1–14.

Berger, C. R. and Chaffee, S. H. (eds) (1987) Hand-book of Communication Science, Beverly Hills,CA: Sage.

Berger, J. (1972) Ways of Seeing, Harmondsworth:Penguin.

Berger, P. L. and Luckmann, T. (1966) The SocialConstruction of Reality, London: Allen Lane.

Berkowitz, L. (1964) ‘The Effects of ObservingViolence,’ Scientific American 21: 35–41.

Berkowitz, L. and Geen, R. G. (1966) ‘Film Violenceand the Cue Properties of Available Targets,’Journal of Personality and Social Psychology, 3:525–530.

Berman, M. (1982) All That Is Solid Melts into Air:The Experience of Modernity, London: Verso.

Bernard, H. R. (1995) Research Methods in Anthro-pology: Qualitative and Quantitative Approa-ches, Walnut Creek, CA: Alta Mira Press.

Bertaux, D. (ed.) (1981) Biography and Society,Beverly Hills, CA: Sage.

Bhabha, H. K. (1994) The Location of Culture,London: Routledge.

Bhaskar, R. (1979) The Possibility of Naturalism,Brighton: Harvester Press.

Biagioli, M. (ed.) (1999) The Science Studies Reader,New York: Routledge.

Billig, M. (1995) Banal Nationalism, London: Sage.Biltereyst, D. (1991) ‘Resisting American Hege-

mony: A Comparative Analysis of the Receptionof Domestic and US Fiction,’ European Journal ofCommunication, 6(4): 469–497.

Biltereyst, D. (2001) ‘Global News Research andComplex Citizenship: Towards an Agenda forResearch on Foreign/International News and Audi-ences,’ in S. Hjarvard (ed.) News in a GlobalizedSociety, Gothenburg: Nordicom.

Black, T. R. (1999) Doing Quantitative Research inthe Social Sciences, London: Sage.

Blaikie, N. (1993) Approaches to Social Enquiry,Cambridge: Polity Press.

Blankenberg, N. (1999) ‘In Search of a RealFreedom: Ubuntu and the Media,’ Critical Arts13(2): 42–65. Durban: University of Natal (www.und.ac.za/und/ccms).

Blumer, H. and Hauser, P. H. (1933) Movies,Delinquency, and Crime, New York: The Mac-millan Company.

Blumler, J. G. (ed.) (1992) Television and the PublicInterest: Vulnerable Values in Westem EuropeanBroadcasting, London, and Newbury Park, CA:Sage.

References296

Blumler, J. G. (1979) ‘The Role of Theory in Usesand Gratifications Studies,’ CommunicationResearch, 6(1): 9–36.

Blumler, J. G. and Gurevitch, M. (1995) The Crisisof Public Communication, London: Routledge.

Blumler, J. G., McLeod, J. and Rosengren, K. E. (eds)(1992) Comparatively Speaking: Communicationand Culture across Space and Time, NewburyPark, CA: Sage.

Boddy, W. (1990) Fifties Television: The Industryand Its Critics, Urbana, IL: University of IllinoisPress.

Bogdanovich, P. (1967) John Ford, Berkeley:University of California Press.

Bolter, J. D. (1991) Writing Space: The Computer,Hypertext, and the History of Writing, Hillsdale,NJ: Lawrence Erlbaum.

Bolter, J. D. and Grusin, R. (1999) Remediation:Understanding New Media, Cambridge, MA:MIT Press.

Boorstin, D. (1961) The Image: A Guide to Pseudo-Events in America, New York: Atheneum.

Booth, W. (1961) The Rhetoric of Fiction, Chicago,IL: University of Chicago Press.

Borde, R. and Chaumeton, E. (1955) Panorame dufilm noir americain [Outline of the American‘Black’ Film], Paris: Les Éditions de Minuit.

Borden, D. L. and Harvey, K. (eds) (1998) TheElectronic Grapevine: Rumor, Reputation, andReporting in the New Online Environment,Mahwah, NJ: Lawrence Erlbaum.

Bordewijk, J. and Kaam, B. (1986) ‘Towards a NewClassification of Tele-information Services,’Intermedia, 14(1): 16–21.

Bordwell, D. (1985a) Narration in the Fiction Film,London: Methuen.

Bordwell, D. (1985b) ‘The Classical HollywoodStyle, 1917–60,’ in D. Bordwell et al. TheClassical Hollywood Cinema: Film Style andMode of Production to 1960, New York: Colum-bia University Press.

Bordwell, D. (1989) ‘Historical Poetics of Cinema,’in R. Barton Palmer (ed.) The Cinematic Text:Methods and Approaches, New York: AMS Press.

Bordwell, D. and Carroll, N. (eds) (1996) Post-Theory: Reconstructing Film Studies, Madison,WI: University of Wisconsin Press.

Bordwell, D. and Thompson, K. (1997) Film Art: AnIntroduction, New York: McGraw-Hill.

Bordwell, D., Staiger, J., and Thompson, K. (1985)The Classical Hollywood Cinema: Film Style andMode of Production to 1960, New York:Columbia University Press.

Borzekowski, D. L. G. and Robinson, T. N. (1999)‘Viewing the Viewers: Ten Video Cases of

Children’s Television Viewing Behaviors,’ Journalof Broadcasting and Electronic Media, 43(4):506–528.

Boudon, R., Cherkaoui, M., and Alexander, J. (eds)(1997) The Classical Tradition in Sociology: TheEuropean Tradition, London: Sage.

Bourdieu, P. (1977) Outline of a Theory of Practice,Cambridge: Cambridge University Press.

Bourdieu, P. (1984) Distinction, Cambridge, MA:Harvard University Press [orig. publ. 1979].

Bourdieu, P. (1988) Homo Academicus, Cambridge:Polity Press.

Bower, G. H. and Cirilo, R. K. (1985) ‘CognitivePsychology and Text Processing,’ in T. A. van Dijk(ed.) Handbook of Discourse Analysis (Vol. 1),New York: Academic Press.

Bower, R. T. (1973) Television and the Public, NewYork: Holt, Rinehart & Winston.

Boyd-Barrett, O. (1980) The International NewsAgencies, London: Constable.

Boyd-Barrett, O. and Rantanen, T. (eds) (1999) TheGlobalization of News, London: Sage.

Boyd-Barrett, O. and Thussu, D. (1992) Contra-flowin Global News. International and Regional NewsExchange Mechanisms, London: John Libbey.

Branigan, E. (1992) Narrative Comprehension andFilm, London: Routledge.

Brannen, J. (ed.) (1992) Mixing Methods: Qualita-tive and Quantitative Research, Aldershot:Avebury.

Braudel, F. (1972, 1974) The Mediterranean and theMediterranean World in the Age of Philip II (Vols1–2), New York: Harper & Row.

Braudel, F. (1980) On History, London: Weidenfeld& Nicolson.

Breed, W. (1955) ‘Social Control in the Newsroom,’Social Forces, 33: 326–335.

Bremond, C. (1979) [1966] ‘The Logic of NarrativePossibilities,’ New Literary History, 11: 387–411.

Briggs, A. (1961–1995) The History of Broadcastingin the United Kingdom (Vols 1–5), Oxford:Oxford University Press.

Briggs, A. (1980) ‘Problems and Possibilities in theWriting of Broadcasting History,’ Media, Cultureand Society 2(1): 5–13.

Brodkey, L. (1987) Academic Writing as Social Prac-tice, Philadelphia, PA: Temple University Press.

Brooker, P. and Brooker, W. (eds) (1997) Post-modern After-Images: A Reader in Film, Tele-vision, and Video, London: Edward Arnold.

Brown, M. E. (1994) Soap Opera and Women’sTalk: The Pleasure of Resistance, Thousand Oaks,CA: Sage.

Bruner, J. (1986) Actual Minds, Possible Worlds,Cambridge, MA: Harvard University Press.

References 297

1111234567891012341115678920123456789301234567894012345678950111

Bruner, J. (1994) ‘The Narrative Construction of“Reality,”’ in M. Ammaniti and D. N. Stern (eds)Psychoanalysis and Development, New York:New York University Press.

Brunsdon, C. (1989) ‘Text and Audience,’ in E.Seiter, H. Borchers, G. Kreutzner, and E.-M.Warth (eds) Remote Control: Television, Audi-ences, and Cultural Power, London: Routledge.

Bryant, C. and Jary, D. (eds) (1991) Giddens’ Theoryof Structuration: A Critical Appreciation,London: Routledge.

Bryant, J., Carveth, R. A., and Brown, D. (1981)‘Television Viewing and Anxiety: An Experi-mental Examination,’ Journal of Communication,31: 106–119.

Bryce, J. (1966) ‘The Nature of Public Opinion,’ inB. Berelson and M. Janovitz (eds) Reader in PublicOpinion and Communication, New York: TheFree Press [orig. publ. 1900].

Bryman, A. and Burgess, R. G. (eds) (1999)Qualitative Research, London: Sage.

Bryman, A. and Cramer, D. (1997) QuantitativeData Analysis with SPSS for Windows, London:Routledge.

Bryson, N. (1991) ‘Semiology and Visual Inter-pretation,’ in N. Bryson et al. (eds) Visual Theory:Painting and Interpretation, Cambridge: PolityPress.

Buckingham, D. (ed.) (1993) Reading Audiences:Young People and the Media, Manchester:Manchester University Press.

Buckland, W. (2000) The Cognitive Semiotics ofFilm, Cambridge: Cambridge University Press.

Burgess, R. G. (1982) ‘Keeping Field Notes,’ in R. G.Burgess (ed.) Field Research: A Sourcebook andField Manual, London: Unwin Hyman.

Burke, K. (1950) A Rhetoric of Motives, New York:Prentice-Hall.

Burke, K. (1957) The Philosophy of Literary Form,New York: Vintage.

Burke, P. (ed.) (1991) New Perspectives on His-torical Writing, Cambridge: Polity Press.

Burke, P. (1992) History and Social Theory,Cambridge: Polity Press.

Burns, R. (ed.) (1995) German Cultural Studies: AnIntroduction, Oxford: Oxford University Press.

Butsch, R. (2000) The Making of American Audi-ences: From Stage to Television, 1750–1990,Cambridge: Cambridge University Press.

Butterworth, H. (1965) The Whig Interpretation ofHistory, New York and London: W. W. Norton &Co.

Caldwell, J. T. (1995) Televisuality: Style, Crisis, andAuthority in American Television, New Bruns-wick, NJ: Rutgers University Press.

Calhoun, C. (ed.) (1992) Habermas and the PublicSphere, Cambridge, MA: MIT Press.

Cantor, J. (1994) ‘Fright Reactions to Mass Media,’in J. Bryant and D. Zillmann (eds) Media Effects:Advances in Theory and Research, Hillsdale, NJ:Lawrence Erlbaum Associates.

Cantor, M. (1971) The Hollywood TelevisionProducer: His Work and His Audience, NewYork: The Free Press. Revised version 1988, NewBrunswick, NJ: Transaction Books.

Cantril, H., Gaudet, H., and Herzog, H. (1940) TheInvasion From Mars, Princeton, NJ: PrincetonUniversity Press.

Capella, J. N. (1996) ‘Symposium: Biology andCommunication,’ Journal of Communication,46(3): 4–84.

Carey, J. W. (1989) Communication as Culture,Boston, MA: Unwin Hyman.

Carey, J. W. and Kreiling, A. W. (1974) ‘PopularCulture and Uses and Gratifications: NotesToward an Accommodation,’ in J. G. Blumler andE. Katz (eds) The Uses of Mass Communications,Beverly Hills, CA: Sage.

Carringer, R. (1996) The Making of Citizen Kane,Berkeley: University of California Press.

Castells, M. (1996) The Rise of the Network Society,Oxford: Blackwell.

Cater, D. (1959) The Fourth Branch of Government,Boston, MA: Houghton-Mifflin.

Cayrol, R. and Mercier, A. (1998) ‘Political Com-munication Scholarship in France,’ PoliticalCommunication, 15: 383–421.

Centerwall, B. S. (1989) ‘Exposure to Television asa Cause of Violence,’ Public Communication andBehaviour, 2: 1–58.

Chaffee, S. H. and Hochheimer, J. (1985) ‘TheBeginnings of Political Communication Researchin the United States: Origins of the “LimitedEffects” Model,’ in E. M. Rogers and F. Balle (eds)The Media Revolution in America and WesternEurope, Norwood, NJ: Ablex.

Chamberlayne, P., Bornat, J. and Wengraf, T. (eds)(2000) The Turn to Biographical Methods inSocial Science, London: Routledge.

Charlton, T. (1997) ‘The Inception of BroadcastTelevision: A Naturalistic Study of Television’sEffects,’ in T. Charlton and K. David (eds) ElusiveLinks: Television, Video Games and Children’sBehaviour, Tewkesbury: Park Published Papers.

Chatman, S. (1989) Story and Discourse: NarrativeStructure in Fiction and Film, Ithaca, NY: CornellUniversity Press (5th paperback edn) [orig. publ.1978].

Chatman, S. (1990) Coming to Terms: The Rhetoricof Narrative in Fiction and Film, Ithaca, NY:Cornell University Press.

References298

Chion, M. (1985) Le son au cinéma [Sound inCinema], Paris: Éditions de l’Étoile.

Chion, M. (1988) La toile trouée [The PerforatedScreen], Paris: Seuil.

Chion, M. (1994) Audio-vision. Sound on Screen,New York: Columbia University Press [orig. publ.1990].

Chion, M. (1995) La musique au cinéma [Music inCinema], Paris: Fayard.

Chion, M. (1998) Le son [Sound], Paris: Nathan.Chion, M. (1999) The Voice in Cinema, New York:

Columbia University Press [orig. publ. 1982].Chomsky, D. (1999) ‘The Mechanisms of Manage-

ment Control at the New York Times,’ Media,Culture and Society, 21(5): 579–599.

Chomsky, N. (1965) Aspects of the Theory ofSyntax, Cambridge, MA: MIT Press.

Christians, C. and Traber, M. (eds) (1997) Com-munication Ethics and Universal Values, Thou-sand Oaks, CA: Sage.

Clark, A. (1983) Raymond Chandler in Hollywood,New York: Proteus.

Clarke, D. S. (1990) Sources of Semiotic, Carbon-dale: Southern Illinois University Press.

Clifford, J. and Marcus, C. E. (eds) (1986) WritingCulture: The Poetics and Politics of Ethnography,Berkeley: University of California Press.

Coakley, J. (ed.) (2000) Handbook of Sports Studies,London: Sage.

Cohen,A.,Levy,M.,Roeh, I.,andGurevitch,M.(1996)GlobalNewsrooms,LocalAudiences.A Study of theEurovision News Exchange, London: John Libbey.

Cohen, B. (1963) The Press and Foreign Policy,Princeton, NJ: Princeton University Press.

Collins, C. (1997) ‘Viewer Letters as AudienceResearch: The Case of Murphy Brown,’ Journal ofBroadcasting and Electronic Media, 41(1): 109–131.

Comstock, G. A. (1978) Television and HumanBehavior, New York: Columbia University Press.

Comstock, G. A. (1998) ‘Television Research: PastProblems and Present Issues,’ in J. K. Asamen andG. L. Berry (eds) Research Paradigms, Television,and Social Behavior, Thousand Oaks, CA: Sage.

Comstock, G. A., Murray, J. P., and Rubinstein, E.A. (eds) (1971) Television and Social Behavior,Washington, DC: Government Printing Office.

Conant, J. and Haugeland, J. (eds) (2000) ThomasS. Kuhn: The Road Since Structure, New York:Basic Books.

Converse, J. M. (1987) Survey Research in theUnited States: Roots and Emergence, 1890–1960,Berkeley: University of California Press.

Conway, J. C. and Rubin, A. M. (1991) ‘Psycho-logical Predictors of Television Viewing Motiva-tion,’ Communication Research, 18(4): 443–463.

Cook, N. (1998) Analysing Musical Multimedia,Oxford: Oxford University Press.

Cook, P. (1999) The Cinema Book, London: BritishFilm Institute.

Cook, T. D., Kendziersky, D. A., and Thomas, S. A.(1983) ‘The Implicit Assumptions of TelevisionResearch: An Analysis of the 1982 NIMH Reporton Television and Behaviour,’ Public OpinionQuarterly, 47: 161–201.

Cooley, C. H. (1894) ‘The Theory of Transport-ation,’ Publications of the American EconomicAssociation, 9(3): 221–370.

Cooper, B. (1999) ‘The Relevancy and GenderIdentity in Spectators’ Interpretations of Thelmaand Louise,’ Critical Studies in Mass Com-munication, 16(1): 20–41.

Corley, K. M. and Kaufer, D. S. (1993) ‘SemanticConnectivity: An Approach for Analyzing Sym-bols in Semantic Networks,’ CommunicationTheory, 3(3): 183–213.

Corner, J. (1991) ‘Meaning, Genre, and Context:The Problematics of “Public Knowledge” in theNew Audience Studies,’ in J. Curran and M.Gurevitch (eds) Mass Media and Society, London:Edward Arnold.

Cottle, S. (2000) ‘New(s) Times: Towards a “SecondWave” of News Ethnography,’ Communications:The European Journal of Communication Re-search, 25(1): 19–41.

Coulthard, M. (1985) An Introduction of DiscourseAnalysis (2nd edn), London: Longman.

Cowan, G. (1979) See No Evil: The Backstage BattleOver Sex and Violence on Television, New York:Simon & Schuster.

Coward, R. and Ellis, J. (1977) Language andMaterialism, London: Routledge & Kegan Paul.

Creswell, J. W. (1994) Research Design: Qualitativeand Quantitative Approaches, Thousand Oaks,CA: Sage.

Crisell, A. (1994) Understanding Radio (2nd edn),London: Routledge.

Croteau, D. and Hoynes, W. (2001) The Business ofMedia: Corporate Media and the Public Interest,Thousand Oaks, CA: Pine Forge Press.

Crouse, T. (1973) The Boys on the Bus, New York:Random House.

Crow, B. K. (1986) ‘Conversational Pragmatics inTelevision Talk: The Discourse of “Good Sex,”’Media, Culture and Society, 8: 457–444.

Cunningham, S. (1988) ‘Kennedy-Miller: “HouseStyle” in Australian Television,’ in S. Dermodyand E. Jacka (eds) The Imaginary Industry: Aus-tralian Film in the late ’80s, North Ryde,Australia: Australian Film, Television and RadioSchool.

References 299

1111234567891012341115678920123456789301234567894012345678950111

Curran, J. (ed.) (2000) Media Organizations inSociety, London: Edward Arnold.

Curran, J. and Seaton, J. (1997) Power WithoutResponsibility: The Press and Broadcasting inBritain (5th edn), London: Routledge.

Curtin, M. (1996) ‘On Edge: Culture Industries inthe Neo-network Era,’ in R. Ohmann, G. Averill,M. Curtin, D. Shumway, and E. Traube (eds)Making and Selling Culture, Hanover, NH:Wesleyan University Press.

Curtin, M. (1999) ‘Feminine Desire in the Age ofSatellite Television,’ Journal of Communication,49(2): 55–70.

D’Acci, J. (1994) Defining Women: The Case ofCagney and Lacey, Chapel Hill: University ofNorth Carolina Press.

Dahl, H. F. (1976) ‘The Art of Writing BroadcastingHistory,’ Gazette. International Journal for MassCommunication Studies, 3: 130–137.

Dahl. H. F. (1994) ‘The Pursuit of Media History,’Media Culture & Society, 16(4): 551–564.

Danto, A. C. (1965) Analytical Philosophy ofHistory, Cambridge: Cambridge University Press.

Dayan, D. and Katz, E. (1992) Media Events,Cambridge, MA: Harvard University Press.

Day-Lewis, S. (1989) One Day in the Life ofTelevision, London: Grafton.

Deacon, D., Fenton, N., and Bryman, A. (1999)‘From Inception to Reception: The Natural His-tory of a News Item,’ Media, Culture & Society,21: 5–31.

Deacon, D., Pickering, M., Golding, P., andMurdock, G. (1999) Researching Communica-tions: A Practical Guide to Methods in Media andCultural Analysis, London: Edward Arnold.

Dearing, J. and Rogers, E. M. (1996) Agenda-Setting, Thousand Oaks, CA: Sage.

DeFleur, M. L. (1964) ‘Occupational Roles as Por-trayed on Television,’ Public Opinion Quarterly,28: 57–74.

DeFleur, M. L. (1975) Theories of Mass Com-munication, New York: David McKay.

DeFleur, M. L. and Larsen, O. N. (1987) The Flowof Information, New Brunswick, NJ: TransactionPress [orig. publ. 1958].

Delia, J. (1987) ‘Communication Research: AHistory,’ in C. R. Berger and S. H. Chaffee (eds)Handbook of Communication Science, NewburyPark, CA: Sage.

Demers, D. P. and Wackman, D. B. (1988) ‘Effect ofChain Ownership on Newspaper ManagementGoals,’ Newspaper Research Journal, 9(2): 59–68.

Dennis, E. E. and Wartella, E. (eds) (1996) AmericanCommunication Research: The Remembered His-tory, Mahwah, NJ: Lawrence Erlbaum.

Denton, R. E. (ed.) (1993) The Media and thePersian Gulf War, Westport, CT: Praeger.

Denzin, N. (1970) The Research Act: A TheoreticalIntroduction to Sociological Methods, EnglewoodCliffs, NJ: Prentice-Hall.

Denzin, N. (1989) The Research Act: A TheoreticalIntroduction to Sociological Methods (3rd edn),Englewood Cliffs, NJ: Prentice-Hall.

Denzin, N. and Lincoln, Y. S. (eds) (2000) Hand-book of Qualitative Research (2nd edn),Thousand Oaks, CA: Sage.

Derrida, J. (1976) Of Grammatology, Baltimore,MD: Johns Hopkins University Press [orig. publ.1967].

Dewey, J. (1926) The Public and its Problems,London: George Allen & Unwin.

Dichter, E. (1947) ‘Psychology in Market Research,’Harvard Business Review, 25(4): 432–443.

Dickey, S. (1993) Cinema and the Urban Poor inSouth India, Cambridge: Cambridge UniversityPress.

Dissanayake, W. (1988a) ‘The Need for AsianApproaches to Communication,’ in W. Diss-anayake (ed.) Communication Theory: The AsianPerspective, Singapore: Asian Mass Communi-cation Research and Information Center.

Dissanayake, W. (ed.) (1988b) CommunicationTheory: The Asian Perspective, Singapore: AsianMass Communication Research and InformationCenter.

Doherty, T. (1999) Pre-code Hollywood: Sex,Immorality, and Insurrection in American Cin-ema, New York: Columbia University Press.

Donnerstein, E. and Berkowitz, L. (1981) ‘VictimReactions in Aggressive Erotic Films as a Factor inViolence against Women,’ Journal of Personalityand Social Psychology, 36: 180–188.

Donohue, G., Olien, C., and Tichenor, P. (1987)‘Media Access and Knowledge Gaps,’ CriticalStudies in Mass Communication, 4(1): 87–92.

Donohue, G., Tichenor, P., and Olien, C. (1975)‘Mass Media and the Knowledge Gap,’ Com-munication Research, 2: 3–23.

Dorfman, A. and Mattelart, A. (1975) How to ReadDonald Duck: Imperialist Ideology in the DisneyComic, New York: International General [orig.publ. 1971].

Douglas, M. (1987) How Institutions Think,London: Routledge & Kegan Paul.

Douglas, M. (1996) Thought Styles: Critical Essayson Good Taste, London: Sage.

Douglas, M. (1997) ‘The Depoliticization of Risk,’in R. Ellis and M. Thompson (eds) CultureMatters: Essays in Honour of Aaron Wildavsky,Boulder, CO: Westview Press.

References300

Downey, G. L. (1998) The Machine in Me: AnAnthropologist Sits Among Computer Engineers,New York: Routledge.

Downing, J. D. H. (2000) Radical Media: RebelliousCommunication and Social Movements, Thou-sand Oaks, CA: Sage.

Doyle, B. (1989) English and Englishness, London:Routledge.

Dray, W. (1957) Laws and Explanation in History,London: Oxford University Press.

Drew, P. and Heritage, J. (1992) ‘Analyzing Talk atWork: An Introduction,’ in P. Drew and J.Heritage (eds) Talk at Work: Interaction inInstitutional Settings, Cambridge: CambridgeUniversity Press.

Drotner, K. (1988) English Children and theirMagazines 1751–1945, Yale, CT: Yale UniversityPress.

Drotner, K. (1994) ‘Ethnographic Enigmas: “TheEveryday” in Recent Media Studies,’ CulturalStudies, 8(2): 341–357.

Drotner, K. (1996) ‘Less Is More: Media Ethnog-raphy and Its Limits,’ in P. I. Crawford and S. B.Hafsteinsson (eds) The Construction of theViewer, Højbjerg, Denmark: Intervention Press.

Drotner, K., Jensen, K. B., Poulsen, I. and Schrøder,K. C. (1996) Medier og kultur [Media andculture], Copenhagen, Denmark: Borgen.

Duke, L. (2000) ‘Black in a Blonde World: Race andGirls’ Interpretations of the Feminine Ideal in TeenMagazines,’ Journalism and Mass Communica-tion Quarterly, 77(2): 367–392.

Dunaway, D. K. and Baum, W. K. (eds) (1996) OralHistory: An Interdisciplinary Anthology (2ndedn), Walnut Creek, CA: Alta Mira Press.

Duncan, H. D. (1968) Symbols in Society, NewYork: Oxford University Press.

During, S. (ed.) (1999) The Cultural Studies Reader(2nd edn), London: Routledge.

Eagleton, T. (1983) Literary Theory: An Intro-duction, Minneapolis: Minnesota UniversityPress.

Eastman, S. T. (1998) ‘Programming Theory underStress: The Active Industry and the ActiveAudience,’ in M. Roloff (ed.) CommunicationYearbook (Vol. 21), Thousand Oaks, CA: Sage.

Eco, U. (1968) La struttura assente. Introduzionealla ricerca semiologica [The Absent Structure.Introduction to Semiological Research], Milan:Bompiani.

Eco, U. (1976) A Theory of Semiotics, Bloomington:Indiana University Press [orig. publ. 1968, revised1975].

Eco, U. (1984) Semiotics and the Philosophy ofLanguage, London: Macmillan.

Eco, U. (1987) ‘Narrative Structures in Fleming,’ inU. Eco (ed.) The Role of the Reader, London:Hutchinson [orig. publ. 1965].

Edelman, M. (1971) Politics as Symbolic Action,Chicago, IL: Markham.

Edwards, D. and Potter, J. (1992) Discursive Psych-ology, London: Sage.

Eisenstein, E. L. (1979) The Printing Press as anAgent of Change: Communication and CulturalTransformation in Early Modern Europe (Vols1–2), Cambridge: Cambridge University Press.

Eisler, H. and Adorno, T. W. (1947) Composing forthe Films, London: Dobson.

Ellen, R. F. (ed.) (1984) Ethnographic Research: AGuide to General Conduct, London: AcademicPress.

Elliott, P. (1972) The Making of a Television Series,London: Constable.

Elliott, P. (1974) ‘Uses and Gratifications Research:A Critique and a Sociological Altemative,’ in J. G.Blumler and E. Katz (eds) The Uses of MassCommunications, Beverly Hills, CA: Sage.

Ellis, J. (1975) ‘Made at Ealing,’ Screen, 16(1):78–127.

Ellis, J. (1982) Visible Fictions, London: Routledge& Kegan Paul.

Emery, M. and Emery, E. (1988) The Press inAmerica, Englewood Cliffs, NJ: Prentice-Hall.

Emmison, M. and Smith, P. (2000) Researching theVisual: Images, Objects, Contexts, and Inter-actions in Social and Cultural Inquiry, London:Sage.

Enzensberger, H. M. (1972) ‘Constituents of aTheory of the Media,’ in D. McQuail (ed.)Sociology of Mass Communications, Harmonds-worth: Penguin [orig. publ. 1970].

Epstein, E. J. (1973) News from Nowhere, NewYork: Random House.

Ericson, R. V., Baranak, P. M., and Chan, J. B. L.(1987) Visualizing Deviance: A Study of NewsOrganization, Toronto: University of TorontoPress.

Ericson, R. V., Baranek, P. M., and Chan, J. B. L.(1991) Representing Order. Crime, Law andJustice in the News Media, Milton Keynes: OpenUniversity Press.

Erlich, V. (1955) Russian Formalism: History,Doctrine, The Hague, Mouton.

Ettema, J. S. and Whitney, D. C. (eds) (1994)Audiencemaking: How the Media Create theAudience, Thousand Oaks, CA: Sage.

Evans, J. and Hall, S. (eds) (1999) Visual Culture:The Reader, London: Sage.

Eysenck, M. W. and Keane, M. T. (1995) CognitivePsychology: A Student’s Handbook (3rd edn),Hove, Sussex: Psychology Press.

References 301

1111234567891012341115678920123456789301234567894012345678950111

Fairclough, N. (1992) Discourse and Social Change,Cambridge: Polity Press.

Fairclough, N. (1993) ‘Critical Discourse Analysisand the Marketization of Public Discourse: theUniversities,’ Discourse & Society, 4(2): 133–168.

Fairclough, N. (1995) Media Discourse, London:Edward Arnold.

Fairclough, N. (1998) ‘Political Discourse in theMedia: An Analytical Framework,’ in A. Bell andP. Garrett (eds) Approaches to Media Discourse,Oxford: Blackwell.

Fallon, D. (1980) The German University: A HeroicIdeal in Conflict with the Modern World, Boulder:Colorado Associated University Press.

Farrell, T. B. (1987) ‘Beyond Science: HumanisticContributions to Communication Theory,’ in C. R. Berger and S. H. Chaffee (eds) Handbook ofCommunication Science, Newbury Park, CA:Sage.

Fenby, J. (1986) The International News Services,New York: Schocken Books.

Ferguson, M. and Golding, P. (eds) (1997) CulturalStudies in Question, London: Sage.

Feshbach, S. (1961) ‘The Stimulating vs CatharticEffects of a Vicarious Aggressive Experience,’Journal of Abnormal and Social Psychology, 63:381–385.

Feuer, J., Kerr, P. and Vahamagi, T. (eds) (1984)MTM: ‘Quality Television,’ London: British FilmInstitute.

Fielding, N. G. and Lee, R. M. (1998) ComputerAnalysis and Qualitative Research, London: Sage.

Filmer P., Philipson, M., Silverman, D. and Walsh,D. (1972) New Directions in Sociological Theory,London: Collier-Macmillan.

Findahl, O. (1985) ‘Some Characteristics of NewsMemory and Comprehension,’ Journal of Broad-casting and Electronic Media, 29(4): 379–396.

Fink, A. (1995a) The Survey Handbook, ThousandOaks, CA: Sage.

Fink, A. (1995b) How To Ask Survey Questions,Thousand Oaks, CA: Sage.

Fink, E. J. and Gantz, W. (1996) ‘A Content Analysisof Three Mass Communication Research Tradi-tions: Social Science, Interpretive Studies, andCritical Analysis,’ Journalism and Mass Com-munication Quarterly, 73(1): 114–134.

Finnemann, N. O. (1999) Thought, Sign, andMachine: The Computer Reconsidered. Available:http://www.hum.au.dk/ckulturf/DOCS/PUB/nof/TSM [accessed December 15, 1999].

Fish, S. (1979) Is There a Text in This Class? TheAuthority of Interpretive Communities, Cam-bridge, MA: Harvard University Press.

Fisher, W. A. and Grenier, G. (1994) ‘Violent Pornog-raphy, Antiwomen Thoughts, and AntiwomenActs: In Search of Reliable Effects,’ Journal of SexResearch 31(1): 23–38.

Fishman, M. (1980) Manufacturing the News,Austin: University of Texas Press.

Fiske, J. (1987) Television Culture, London:Methuen.

Fiske, J. (1990) Introduction to CommunicationStudies, London: Routledge.

Fiske, J. and Hartley, J. (1978) Reading Television,London: Methuen.

Flick, U. (1998) An Introduction to QualitativeResearch, London: Sage.

Flitterman-Lewis, S. (1983) ‘The Real Soap Operas,’in E. A. Kaplan (ed.) Regarding Television: Criti-cal Approaches – An Anthology, Frederick, MD:University Publications of America.

Flourney, D. M. and Stewart, R. K. (1997) CNN.Making News in the Global Market, Luton, Beds:University of Luton Press.

Forbes, J. and Kelly, M. (eds) (1995) French CulturalStudies: An Introduction, Oxford: OxfordUniversity Press.

Foster, H. (1985) Postmodern Culture, London:Pluto Press.

Foucault, M. (1972) The Archaeology of Know-ledge, London: Tavistock.

Fowler, R. (1985) ‘Power,’ in T. van Dijk (ed.)Handbook of Discourse Analysis (Vol. 4), Lon-don: Academic Press.

Fowler, R., Hodge, B., Kress, G., and Trew, T. (eds)(1979) Language and Control, London: Rout-ledge & Kegan Paul.

Fredin, E. S., Monnett, T. H., and Kosicki, G. M.(1994) ‘Knowledge Gaps, Social Locators, andMedia Schemata: Gaps, Reverse Gaps, and Gapsof Disaffection,’ Journalism and Mass Communi-cation Quarterly, 71(1): 176–190.

Furnham, A. (1988) Lay Theories: Everyday Under-standing of Problems in the Social Sciences,Oxford: Pergamon Press.

Furnham, A. and Schofield, I. (1986) ‘Sex-roleStereotyping in British Radio Advertisements,’British Journal of Social Psychology, 25: 165–171.

Furnham, A. and Skae, E. (1997) ‘Changes in theStereotypical Portrayal of Men and Women inBritish Television Advertisements,’ European Psy-chologist, 2: 44–51.

Furnham, A., Abramsky, S., and Gunter, B. (1997)‘A Cross-cultural Content Analysis of Children’sTelevision Advertisements,’ Sex Roles, 37(l/2):91–99.

References302

Gadamer, H.-G. (1975) Truth and Method, NewYork: The Seabury Press [(orig. publ. 1960].

Galtung J. (1971) ‘A structural theory of imperial-ism,’ Journal of Peace Research, 8: 81–147.

Galtung, J. (1973) ‘Galtung-Ruge Revisited’ (with atranscription of the subsequent debate at the con-ference ‘Nordisk konferens om konflikt – ochmassmediaforskning’ [Nordic conference on con-flict and mass media research]), Department ofHistory, Lund University, Sweden, 6–7 April,unpublished.

Galtung, J. and Ruge, M. H. (1962) Presentasjonenaf utenriksnyheter. En undersøkelse av nyhets-meldingene fra Cuba og Kongo juli 1960 [ThePresentation of Foreign News. An Analysis ofNews Reports from Cuba and Congo, July 1960],Report no. 14–1, Avdeling for Konflikt- ogFredsforskning, Oslo, Norway.

Galtung, J. and Ruge, M. H. (1965) ‘The structureof foreign news,’ Journal of Peace Research, 2:64–91.

Gamson, W. A. (1992) Talking Politics, New York:Cambridge University Press.

Gans, H. J. (1957) ‘The Creator–Audience Relation-ship in the Mass Media: An Analysis of MovieMaking,’ in B. Rosenberg and D. White (eds)Mass Culture: The Popular Arts in America, NewYork: Free Press.

Gans, H. J. (1979) Deciding What’s News, NewYork: Vintage.

Gardner, G. (1987) The Censorship Papers: MovieCensorship Letters From the Hays Office: 1934–1968, New York: Dodd Mead.

Gardner, H. (1985) The Mind’s New Science: AHistory of the Cognitive Revolution, New York:Basic Books.

Garfinkel, H. (1967) Studies in Ethnomethodology,Englewood Cliffs, NJ: Prentice-Hall.

Garnham, N. and Locksley, G. (1991) ‘TheEconomics of Broadcasting,’ in J. G. Blumler andT. J. Nossiter (eds) Broadcasting Finance in Tran-sition: A Comparative Handbook, New York:Oxford University Press.

Garrison, B. (2000) ‘Journalists’ Perceptions of On-line Information-Gathering Problems,’ Journalismand Mass Communication Quarterly, 77(3):500–514.

Gauntlett, D. and Hill, A. (1999) TV Living: Tele-vision, Culture, and Everyday Life, London:Routledge.

Gay, P. du, Hall, S., Janes, L., Mackay, H., andNegus, K. (1997) Doing Cultural Studies: TheStory of the Sony Walkman, London: Sage.

Gaziano, C. (1997) ‘Forecast 2000: WideningKnowledge Gaps,’ Journalism and Mass Com-munication Quarterly, 74(2): 237–264.

Gee, J. P., Michaels, S. and O’Connor, M. C. (1992)‘Discourse Analysis,’ in M. D. LeCompte, W. L.Millroy, and J. Preissle (eds) The Handbook ofQualitative Research in Education, San Diego,CA: Academic Press.

Geertz, C. (1973) ‘Thick Description,’ in The Inter-pretation of Cultures, New York: Basic Books.

Gelder, K. and Thornton, S. (eds) (1997) TheSubcultures Reader, London: Routledge.

Genette, G. (1980) Narrative Discourse, Oxford:Basil Blackwell [orig. publ. 1972].

Genette, G. (1997) Palimpsests: Literature in theSecond Degree, Lincoln: University of NebraskaPress [orig. publ. 1982].

Genova, B. K. L. and Greenberg, B. S. (1979)‘Interests in News and the Knowledge Gap,’Public Opinion Quarterly, 43(1): 79–91.

Gerbner, G. (1972) ‘Violence in Television Drama:Trends and Symbolic Functions,’ in G. A. Com-stock and E. A. Rubinstein (eds) Television andSocial Behaviour, Vol. 1: Media Content andControl, Washington, DC: U.S. GovernmentPrinting Office.

Gerbner, G. and Gross, L. (1976) ‘Living withTelevision: The Violence Profile,’ Journal of Com-munication, 26(3): 173–199.

Gerbner, G., Gross, L., Morgan, M., and Signorielli,N. (1986) ‘Living with Television: The Dynamicsof the Cultivation Process,’ in J. Bryant and D.Zillmann (eds) Perspectives on Media Effects,Hillsdale, NJ: Lawrence Erlbaum.

Gerbner, G., Gross, L., Morgan, M., Signorielli, N.,and Jackson-Beeck, M. (1979) ‘The Demon-stration of Power: Violence Profile No. 10,’Journal of Communication, 29: 177–196.

Gerbner, G., Gross, L., Eleey, M. F., Jackson-Beeck,M., Jeffries-Fox, S., and Signorielli, N. (1977)‘Television Violence Profile No. 8: The Highlights,’Journal of Communication, 27: 171–180.

Gerth, H. and Mills, C. W. (1954) Character andSocial Structure: The Psychology of SocialInstitutions, London: Routledge & Kegan Paul.

Giddens, A. (1984) The Constitution of Society,Berkeley: University of California Press.

Giddens, A. (1990) The Consequences of Modernity,Cambridge: Polity Press.

Giddens, A. (1991) Modernity and Self-Identity,Cambridge: Polity Press.

Giddens, A. (1994) ‘Living in a Post-traditionalSociety,’ in U. Beck et al. (eds) Reflexive Modern-ization: Politics, Tradition, and Aesthetics in theModern Social Order, Cambridge: Polity Press.

Gilbert, G. N. and Mulkay, M. (1984) OpeningPandora’s Box: A Sociological Analysis of Sci-entists’ Discourse, Cambridge: Cambridge Uni-versity Press.

References 303

1111234567891012341115678920123456789301234567894012345678950111

Gillespie, M. (1995) Television, Ethnicity, andCultural Change, London: Routledge.

Ginneken, J. van (1998) Understanding GlobalNews: A Critical Introduction, London: Sage.

Ginzburg, C. (1989) ‘Clues: Roots for an EvidentialParadigm,’ in C. Ginzburg (ed.) Clues, Myths, andthe Historical Method, Baltimore, MD: JohnsHopkins University Press.

Gitlin, T. (1978) ‘Media Sociology: The DominantParadigm,’ Theory and Society, 6: 205–253.

Gitlin, T. (1979) ‘Prime Time Ideology,’ SocialProblems, 26: 251–266.

Gitlin, T. (1980) The Whole World Is Watching,Berkeley: University of California Press.

Gitlin, T. (1983) Inside Prime Time, New York:Pantheon.

Gitlin, T. (1997) ‘The Anti-political Populism ofCultural Studies,’ in M. Ferguson and P. Golding(eds) Cultural Studies in Question, London: Sage.

Glaser, B. G. (1992) Basics of Grounded TheoryAnalysis: Emergence vs. Forcing, Mill Valley, CA:Sociology Press.

Glaser, B. G. and Strauss, A. L. (1967) The Dis-covery of Grounded Theory: Strategies forQualitative Research, Chicago: Aldine.

Glasgow University Media Group (1976) Bad News,London: Routledge & Kegan Paul.

Glasgow University Media Group (1980) More BadNews, London: Routledge.

Glenn, T. C., Sallot, L. M. and Curtin, P. A. (1997)‘Public Relations and the Production of News,’Communication Yearbook, 20: 111–115.

Glessing, R. J. (1970) The Underground Press inAmerica, Bloomington: Indiana University Press.

Goffman, E. (1974) Frame Analysis, Cambridge,MA: Harvard University Press.

Goffman, E. (1979) Gender Advertisements,London: Macmillan [orig. publ. 1976].

Goldhaber, G. M. and Barnett, G. A. (eds) (1988)Handbook of Organizational Communication,Norwood, NJ: Ablex.

Golding, P. and Elliott, P. (1979) Making the News,London: Longman.

Golding, P. and Harris, P. (eds) (1997) BeyondCultural Imperialism, London: Sage.

Golding, P. and Middleton, S. (1982) Images ofWelfare, Oxford: Martin Robertson.

Golding, P. and Murdock, G. (1991) ‘Culture,Communication, and Political Economy,’ in J.Curran and M. Gurevitch (eds) Mass Media andSociety, London: Edward Arnold.

Gombrich, E. H. (1960) Art and Illusion, Princeton,NJ: Princeton University Press.

Gomery, D. (1992) Shared Pleasures: A History ofMovie Presentation in the United States, Madison:University of Wisconsin Press.

Gomm, R., Hammersley, M. and Foster, P. (eds)(2000) Case Study Method: Key Issues, Key Texts,London: Sage.

Goode, W. J. and Hatt, P. K. (1952) Methods inSocial Research, New York: McGraw-Hill.

Goody, J. (ed.) (1968) Literacy in TraditionalSocieties, Cambridge: Cambridge UniversityPress.

Goody, J. (1987) The Interface between the Writtenand the Oral, Cambridge: Cambridge UniversityPress.

Goody, J. (2000) The Power of the WrittenTradition, Washington, DC: Smithsonian Institu-tion Press.

Goody, J. and Watt, I. (1963) ‘The Consequences ofLiteracy,’ Comparative Studies in Society and His-tory, 5: 304–345.

Gorbman, C. (1987) Unheard Melodies. NarrativeFilm Music, Bloomington and Indianapolis:Indiana University Press.

Gorden, R. L. (1969) Interviewing: Strategy, Tech-niques, and Tactics, Homewood: The DorseyPress.

Gorney, R., Loye, D., and Steele, G. (1977) ‘Impactof Dramatised Television Entertainment on AdultMales,’ American Journal of Psychiatry, 134(2):170–174.

Gottlieb, S. (ed.) (1995) Hitchcock on Hitchcock:Selected Writings and Interviews, Berkeley: Uni-versity of California Press.

Grabe, M. E., Lang, A., Zhou, S. and Bolls, P. D.(2000) ‘Cognitive Access to Negatively Arous-ing News: An Experimental Investigation of theKnowledge Gap,’ Communication Research,27(1): 3–26.

Graber, D. (1984) Processing the News: How PeopleTame the Information Tide, New York: Longman.

Graham, H. and Labanyi, J. (eds) (1995) SpanishCultural Studies: An Introduction – The Strugglefor Modernity, Oxford: Oxford University Press.

Gramsci, A. (1971) Selections from the PrisonNotebooks, New York: International Publishers.

Grant, B. K. (1986) Film Genre Reader, Austin:University of Texas Press.

Grant, B. K. (1995) Film Genre Reader II, Austin:University of Texas Press.

Gray, A. (1992) Video Playtime: The Gendering ofa Leisure Technology, London: Routledge.

Greatbatch, D. (1998) ‘Conversation Analysis:Neutralism in British News Interviews,’ in A. Belland P. Garrett (eds) Approaches to MediaDiscourse, Oxford: Blackwell.

Greenacre, M. J. (1993) Correspondence Analysis inPractice, London: Academic Press.

References304

Greenberg, B. S. (1964) ‘Person-to-Person Com-munication in the Diffusion of a News Event,’Journalism Quarterly, 41: 489–494.

Greenberg, B. S. (1975) ‘British Children and Tele-vised Violence,’ Public Opinion Quarterly, 38:531–547.

Greenwood, D. J. and Levin, M. (1998) Intro-duction to Action Research: Social Research forSocial Change, Thousand Oaks, CA: Sage.

Greimas, A. J. (1966) Structural Semantics. AnAttempt at a Method, Lincoln, NE: University ofNebraska Press [English trans. 1983].

Gripsrud, J. (1995) The Dynasty Years: HollywoodTelevision and Critical Media Studies, London:Routledge.

Grodal, T. K. (1997) Moving Pictures: A NewTheory of Film Genres, Feelings, and Cognition,New York: Oxford University Press.

Gross, A. G. (1990) The Rhetoric of Science,Cambridge, MA: Harvard University Press.

Grossberg, L., Nelson, C., and Treichler, P. (eds)(1992) Cultural Studies, London: Routledge.

Grunig, J. E. (ed.) (1992) Excellence in PublicRelations and Communication Management,Hillsdale, NJ: Lawrence Erlbaum.

Guba, E. G. and Lincoln, Y. S. (1994) ‘CompetingParadigms in Qualitative Research,’ in N. Denzinand Y. S. Lincoln (eds), Handbook of QualitativeResearch, Thousand Oaks, CA: Sage.

Gudykunst, W. B. and Kim, Y. Y. (1992) Com-municating with Strangers: An Approach toIntercultural Communication (2nd edn), NewYork: McGraw-Hill.

Gudykunst, W. B. and Kim, Y. Y. (1997) Com-municating with Strangers: An Approach toIntercultural Communication (3rd edn), NewYork: McGraw-Hill.

Gulich, E. and Quasthoff, U. M. (1985) ‘NarrativeAnalysis,’ in T. A. van Dijk (ed.) Handbook ofDiscourse Analysis (Vol. 2), New York: AcademicPress.

Gumpert, G. and Cathcart, R. (eds) (1986)Inter/Media: Interpersonal Communication in aMedia World, New York: Oxford UniversityPress.

Gunter, B. (1979) ‘Recall of Television News Items:Effects of Presentation Mode, Picture Content andSerial Position,’ Journal of Educational Television,5: 57–61.

Gunter, B. (1985a) Dimensions of TelevisionViolence, Aldershot, Hants: Gower.

Gunter, B. (1985b) ‘News Sources and News Aware-ness: A British Survey,’ Journal of Broadcasting,29: 397–406.

Gunter, B. (1987a) Television and the Fear of Crime,London: John Libbey.

Gunter, B. (1987b) Poor Reception: Misunder-standing and Forgetting Broadcast News,Hillsdale, NJ: Lawrence Erlbaum.

Gunter, B. (2000) Media Research Methods:Measuring Audiences, Reactions, and Impact,London: Sage.

Gunter, B. and Harrison, J. (1998) Violence on Tele-vision: An Analysis of Amount, Nature, Locationand Origin of Violence in British Programmes,London: Routledge.

Gunter, B. and Wober, J. (1982) ‘Television Viewingand Public Trust,’ British Journal of SocialPsychology, 22: 174–176.

Gunter, B., Furnham, A., and Lineton, Z. (1995)‘Watching People Watching Television: WhatGoes on in Front of the TV Set?,’ Journal ofEducational Television, 21(3): 165–191.

Gunter, B., Svennevig, M., and Wober, M. (1986)Television Coverage of the 1983 General Election,Aldershot, Hants: Gower.

Gunter, B., Harrison, J., Arundel, J., and Osborn, R.(1999) ‘Female Victimisation on Television: Ex-tent, Nature and Context of On-screen Por-trayals,’ Communication, 24(4): 387–405.

Gunter, B., Harrison, J., Arundel, J., Osborn, R., andCrawford, M. (1996) Violence on Television inBritain: A Content Analysis, Report to the BBC,BSC, BSkyB, Channel 4, ITC, and ITV, Sheffield:University of Sheffield, Department of JournalismStudies.

Gurevitch, M. (1992) ‘The Globalization ofElectronic Journalism,’ in J. Curran and M.Gurevitch (eds) Mass Media and Society, London:Edward Arnold.

Haas, T. (1999) ‘What’s “Public” About PublicJournalism? Public Journalism and the Lack of aCoherent Public Philosophy,’ CommunicationTheory, 9(3): 346–364.

Habermas, J. (1971) Knowledge and Human Inter-ests, Boston, MA: Beacon Press [orig. publ. 1968].

Habermas, J. (1984) The Theory of CommunicativeAction (Vol. 1), Boston, MA: Beacon Press [orig.publ. 1981].

Habermas, J. (1987) The Theory of CommunicativeAction (Vol. 2), Cambridge: Polity Press [orig.publ. 1981].

Habermas, J. (1989) The Structural Transformationof the Public Sphere, Cambridge, MA: MIT Press[orig. publ. 1962].

Hacker, K. L., Coste, T. G., Kamm, D. F., and Bybee,C. R. (1991) ‘Oppositional Readings of NetworkTelevision News: Viewer Deconstruction,’ Dis-course and Society, 2(2): 183–202.

Hacking, I. (1975) The Emergence of Probability,London: Cambridge University Press.

References 305

1111234567891012341115678920123456789301234567894012345678950111

Hacking, I. (1983) Representing and Intervening:Introductory Topics in the Philosophy of NaturalScience, Cambridge: Cambridge University Press.

Hacking, I. (1999) The Social Construction ofWhat?, Cambridge, MA: Harvard UniversityPress.

Hall, S. (1973a) ‘The Determination of NewsPhotographs,’ in S. Cohen and J. Young (eds) TheManufacture of News, London: Constable.

Hall, S. (1973b) Encoding and Decoding in theTelevision Discourse (Stencilled Occasional Paper7), Birmingham: Centre for ContemporaryCultural Studies.

Hall, S. (1983) ‘The Problem of Ideology – Marxismwithout Guarantees,’ in B. Matthews (ed.) Marx:A Hundred Years On, London: Lawrence &Wishart.

Hall, S. (1986) ‘Cultural Studies: Two Paradigms,’ inR. Collins, J. Curran, N. Garnham, P. Scannell, P.Schlesinger, and C. Sparks (eds) Media, Culture,and Society: A Critical Reader, London: Sage[orig. publ. 1980].

Hall, S. and Jefferson, T. (eds) (1975) ResistanceThrough Rituals, London: Hutchinson.

Hall, S., Connell, I., and Curti, L. (1976) ‘The“Unity” of Current Affairs Television,’ WorkingPapers in Cultural Studies, No. 9: 51–94, Centrefor Contemporary Cultural Studies, University ofBirmingham.

Hall, S., Hobson, D., Lowe, A., and Willis, P. (eds)(1980) Culture, Media, Language, London:Hutchinson.

Halliday, M. A. K. (1973) Explorations in theFunctions of Language, London: Edward Arnold.

Halliday, M. A. K. (1978) Language as SocialSemiotic: The Social Interpretation of Languageand Meaning, London: Edward Arnold.

Hallin, D. C. (1986) The ‘Uncensored War’: TheMedia and Vietnam, New York: Oxford Univer-sity Press.

Hallin, D. C. (1992) ‘Sound Bite News: TV Coverageof Elections 1968–1988,’ Journal of Communi-cation, 42(2): 5–24.

Halliwell, L. (1999) Halliwell’s Film & Video Guide2000, New York: Harper-Perennial.

Halloran, J. (1970) The Effects of Television,London: Panther.

Halloran, J., Elliott, P., and Murdock, G. (1970)Communications and Demonstrations: A CaseStudy, Harmondsworth: Penguin.

Hammersley, M. (1989) The Dilemma of Qualita-tive Method: Herbert Blumer and the ChicagoTradition, London: Routledge.

Hammersley, M. (1995) The Politics of SocialResearch, London: Sage.

Hammersley, M. (1996) ‘The Relationship BetweenQualitative and Quantitative Research: ParadigmLoyalty Versus Methodological Eclecticism,’ in J.T. E. Richardson (ed.) Handbook of QualitativeResearch Methods for Psychology and the SocialSciences, Leicester: British Psychological Society.

Hammersley, M. and Atkinson, P. (1995) Ethno-graphy: Principles in Practice (2nd edn), Londonand New York: Routledge.

Hannerz, U. (1996) Transnational Connections:Culture, People, Places, London: Routledge.

Hannington, W. (1977) Unemployed Struggles,1919–1936, London: Lawrence & Wishart.

Hansen, A., Cottle, S., Negrine, R., and Newbold,C. (1998) Mass Communication ResearchMethods, London: Macmillan.

Hansen, M. (1990) ‘Early Cinema – Whose PublicSphere?,’ in T. Elsaesser (ed.) Early Cinema:Space, Frame, Narrative, London: British FilmInstitute.

Hanson, N. R. (1958) Patterns of Discovery: AnInquiry into the Conceptual Foundations ofScience, Cambridge: Cambridge University Press.

Haraway, D. J. (1997) [email protected]_Oncomouse:Feminism and Technoscience, New York:Routledge.

Harding, S. (1986) The Science Question inFeminism, Ithaca, NY: Cornell University Press.

Harding, S. (ed.) (1987) Feminism and Method-ology: Social Science Issues, Bloomington: IndianaUniversity Press.

Hardt, H. (1979) Social Theories of the Press: EarlyGerman and American Perspectives, London:Sage.

Hardt, H. (1992) Critical Communication Studies:Communication, History, and Theory, London:Routledge.

Hardt, H. (1999) ‘Review of Dennis and Wartella,’Communications: The European Journal of Com-munication Research, 24(2): 239–240.

Harrison, C. and Wood, P. (eds) (1992) Art inTheory – 1900–1990: An Anthology of ChangingIdeas, Oxford: Blackwell.

Hartley, J. and Pearson, R. (eds) (2000) AmericanCultural Studies: A Reader, Oxford: OxfordUniversity Press.

Hartnagel, T. F., Teevan, J. J. Jr., and McIntyre, J. J.(1975) ‘Television Violence and Violent Behav-iour,’ Social Forces, 54: 341–351.

Harvey, D. (1989) The Condition of Postmodernity,Oxford: Blackwell.

Hauben, M. and Hauben, R. (1997) Netizens: Onthe History and Impact of Usenet and the Inter-net, Los Alamitos, CA: IEEE Computer SocietyPress.

References306

Hauser, A. (1951) The Social History of Art (Vols1–4), New York: Vintage.

Have, P. Ten (1999) Doing Conversation Analysis: A Practical Guide, London: Sage.

Havelock, E. A. (1963) Preface to Plato, Oxford:Blackwell.

Hays, W. L. (1973) Statistics for the Social Sciences,New York: Holt, Rinehart & Winston.

Headland, T., Pike, K., and Harris, M. (eds) (1990)Emics and Etics: The Insider/Outsider Debate,Newbury Park, CA: Sage.

Heath, S. (1975) ‘Films and System: Terms ofAnalysis, Part I,’ Screen, 16(1): 7–77; ‘Part II,’Screen 16(2): 91–113.

Heath, S. B. (1980) ‘The Functions and Uses ofLiteracy,’ Journal of Communication, 30(1): 123–133.

Heider, F. (1958) The Psychology of InterpersonalRelations, New York: Wiley.

Heim, M. (1987) Electric Language, New Haven,CT: Yale University Press.

Held, D. and Thompson, J. B. (eds) (1989) SocialTheory of Modern Societies: Anthony Giddensand His Critics, Cambridge: Cambridge Uni-versity Press.

Held, D., McGrew, A., Goldblatt, D. and Perraton,J. (2000) Global Transformations: Politics, Eco-nomics, and Culture, Cambridge: Polity Press.

Hempel, C. G. and Oppenheim, P. (1988) ‘Studies inthe Logic of Explanation,’ in J. Pitt (ed.) Theoriesof Explanation, New York: Oxford UniversityPress [orig. publ. 1948].

Hendy, D. (2000) Radio in the Global Age,Cambridge: Polity Press.

Hennigan, K. M., Del Rosario, M. L., Heath, L.,Cook, T. D., Wharton, J. D., and Calder, B. J.(1982) ‘Impact of the Introduction of Televisionon Crime in the United States: Empirical Findingsand Theoretical Implications,’ Journal of Person-ality and Social Psychology, 42: 461–477.

Herbst, S. (1993) ‘The Meaning of Public Opinion:Citizens’ Constructions of Social Reality,’ Media,Culture & Society, 15(3): 437–454.

Heritage, J. (1985) ‘Analyzing News Interviews:Aspects of the Production of Talk for an Over-hearing Audience,’ in T. van Dijk (ed.) Handbookof Discourse Analysis (Vol. 3), London: AcademicPress.

Heritage, J. and Greatbatch, D. L. (1991) ‘On theInstitutional Character of Institutional Talk: TheCase of News Interviews,’ in D. Boden and D. H.Zimmerman (eds) Talk and Social Structure:Studies in Ethnomethodology and ConversationAnalysis, Oxford: Polity Press.

Herman, E. S. and Chomsky, N. (1988) Manu-facturing Consent: The Political Economy of theMass Media, New York: Pantheon Books.

Hermes, J. (1995) Reading Women’s Magazines: AnAnalysis of Everyday Media Use, Oxford: PolityPress.

Herzog, H. (1941) ‘On Borrowed Experience: AnAnalysis of Listening to Daytime Sketches,’ Stu-dies in Philosophy and Social Science, 9: 65–95.

Herzog, H. (1944) ‘What Do We Really Know aboutDaytime Serial Listeners?,’ in P. F. Lazarsfeld (ed.)Radio Research 1942–3, New York: Duell, Sloan,& Pearce.

Hilmes, M. (1997) Radio Voices: American Broad-casting, 1922–1952, Minneapolis: The Universityof Minnesota Press.

Himmelweit, H. T., Vince, P., and Oppenheim, A. N.(1958) Television and the Child, London: OxfordUniversity Press.

Hine, C. (2000) Virtual Ethnography, London: Sage.Hirsch, P. (1980) ‘The “Scary World” of the Non-

Viewer and Other Anomalies – A Reanalysis ofGerbner et al.’s Findings in Cultivation Analysis,’Communication Research, 7(4): 403–456.

Hirsch, P. (1981) ‘On Not Learning from One’sMistakes, Part II,’ Communication Research,8(1): 3–38.

Hjarvard, S. (1993) ‘Pan-European Television News:Towards a European Political Public Sphere?,’ inP. Drummond, R. Paterson, and J. Willis (eds)National Identity and Europe, London: BritishFilm Institute.

Hjarvard, S. (1995) Internationale TV-nyheder. Enhistorisk analyse af det europæiske system forudveksling af internationale TV-nyheder [Inter-national TV News], Copenhagen, Denmark:Akademisk Forlag.

Hjarvard, S. (1999) ‘TV News Exchange,’ in O.Boyd-Barrett and T. Rantanen (eds) The Global-ization of News, London: Sage.

Hjarvard, S. (ed.) (2001) News in a GlobalizedSociety, Gothenburg, Sweden: Nordicom.

Hjelmslev, L. (1963) Prolegomena to a Theory ofLanguage, Madison: University of WisconsinPress [orig. publ. 1943].

Hobson, D. (1980) ‘Housewives and the MassMedia,’ in S. Hall, D. Hobson, A. Lowe, and P.Willis (eds) Culture, Media, Language, London:Hutchinson.

Hobson, D. (1982) Crossroads: The Drama of aSoap Opera, London: Methuen.

Hockings, P. (ed.) (1995) Principles of Visual Anthro-pology (2nd edn), Berlin: Mouton de Gruyter.

Hodge, B. and Kress, G. (1988) Social Semiotics,Cambridge: Polity Press.

References 307

1111234567891012341115678920123456789301234567894012345678950111

Hodge, B. and Tripp, D. (1986) Children and Tele-vision, Cambridge: Polity Press.

Hoffner, C. and Cantor, J. (1985) ‘DevelopmentalDifferences in Responses to a Television Char-acter’s Appearance and Behaviour,’ Develop-mental Psychology 21: 1065–1074.

Hoggart, R. (1957) The Uses of Literacy, Har-mondsworth: Penguin.

Höhne, H. (1984) Report überNachricten-Agent-uren [Report on News Agencies] (2nd edn),Baden-Baden, Germany: Nomos Verlagsgesell-schaft.

Höijer, B. (1990) ‘Studying Viewers’ Reception of Television Programmes: Theoretical andMethodological Considerations,’ European Jour-nal of Communication, 5(1): 29–56.

Holsti, O. R. (1969) Content Analysis for the SocialSciences and Humanities, Reading, MA: Addison-Wesley.

Holub, R. C. (1984) Reception Theory: A CriticalIntroduction, London: Methuen.

Hoskins, C., McFadyen, S., and Finn, A. (1997)Global Television and Film: An Introduction tothe Economics of the Business, Oxford: Cla-rendon, and New York: Oxford University Press.

Hovland, C. I., Janis, I. L., and Kelley, H. H. (1953)Communication and Persuasion, New Haven, CT:Yale University Press.

Hovland, C. I., Lumsdaine, A. A., and Sheffield, F.D. (1949) Experiments on Mass Communication,Princeton, NJ: Princeton University Press.

Høyer, S., Lauk, E., and Vihalemm, P. (eds) (1993)Towards a Civic Society: The Baltic Media’s LongRoad to Freedom, Tartu: Nota Baltica.

Huesmann, L. R., Eron, L. D., Lefkowitz, M. M.,and Walder, L. O. (1984) ‘Stability of Aggressionover Time and Generations,’ DevelopmentalPsychology, 20(6): 1120–1134.

Hughes, H. M. (1937) ‘Human Interest Stories andDemocracy,’ The Public Opinion Quarterly,April: 73–87.

Hughes, H. M. (1940) News and the Human InterestStory, Chicago, IL: University of Chicago Press.

Hughes, R. (1991) The Shock of the New: Art andthe Century of Change (updated and enlargededn), London: Thames & Hudson.

Hur, K. K. (1984) ‘A Critical Analysis of Inter-national News Flow Research,’ Critical Studies inMass Communication, 1: 365–378.

Husband, C. (1996) ‘The Right to be Understood:Conceiving the Multi-Ethnic Public Sphere,’Innovation, 9(2): 205–215.

Huyssen, A. (1986) After the Great Divide: Modern-ism, Mass Culture, and Postmodernism, London:Macmillan.

Hyuhn, S. B. (2000) ‘Product Differentiation inNational Television Newscasts,’ Journal of Broad-casting and Electronic Media, 44(1): 62–77.

Inglehart, R. (1990) Culture Shift in AdvancedIndustrial Society, Princeton, NJ: Princeton Uni-versity Press.

Inglehart, R., Nevitte, N., and Basanez, M. (1996)The North American Trajectory: Cultural, Eco-nomic, and Political Ties Among the UnitedStates, Canada, and Mexico, New York: Aldine deGruyter.

Innis, H. A. (1956) The Fur Trade in Canada: An Introduction to Canadian Economic History(2nd edn), Toronto: Toronto University Press[orig. publ. 1930].

Innis, H. A. (1972) Empire and Communications,Toronto: University of Toronto Press [orig. publ.1950].

Innis, H. A. (1964) The Bias of Communication,Toronto: University of Toronto Press [orig. publ.1951].

International Motion Picture Almanac, New York:Quigley Publishing.

International Television and Video Almanac, NewYork: Quigley Publishing.

Irigaray, L. (1997) ‘This Sex Which Is Not One,’ inL. Nicholson (ed.) The Second Wave: A Reader inFeminist Theory, London: Routledge [orig. publ.1977].

Irwin, W. (1969) ‘The American Newspaper: AStudy of Journalism in its Relation to the Public,’Collier’s January–June, reprinted with commentsby Clifford F. Weigle and David G. Clark, Ames,Iowa: The Iowa State University Press [orig. publ.1911].

Iser, W. (1974) The Implied Reader. Patterns ofCommunication in Prose Fiction from Bunyan toBeckett, Baltimore, MD: Johns Hopkins Uni-versity Press [orig. publ. 1972].

Iser, W. (1978) The Act of Reading. A Theory ofAesthetic Response, London: Routledge & KeganPaul [orig. publ. 1976].

Iser, W. (1989) Prospecting: From Reader Responseto Literary Anthropology, Baltimore, MD: JohnsHopkins University Press.

Ito, Y. and Tanaka, N. (1992) ‘Education, ResearchInstitutes, and Academic Associations in Journ-alism and Mass Communications in Japan,’ KEIOCommunication Review, 14: 15–35.

Iyengar, S. and Kinder, D. R. (1987) News thatMatters, Chicago, IL: University of Chicago Press.

Jakobson, R. (1960) ‘Closing Statement: Linguisticsand Poetics,’ in T. A. Sebeok (ed.) Style inLanguage, Cambridge, MA: MIT Press.

References308

James, A., Hackey, J. and Dawson, A. (eds) (1997)After Writing Culture: Epistemology and Praxis inContemporary Anthropology,London:Routledge.

Jameson, F. (1991) Postmodernism, Or, The Cul-tural Logic of Late Capitalism, London: Verso.

Jandt, F. E. (2001) Intercultural Communication: AnIntroduction (3rd edn), Newbury Park, CA: Sage.

Jankowski, N. W. (1991) ‘Qualitative Research and Community Media,’ in K. B. Jensen and N. W. Jankowski (eds) A Handbook of Qualita-tive Methodologies for Mass CommunicationResearch, London: Routledge.

Jankowski, N. W. and van Selm, M. (in press)Researching New Media, London: Sage.

Jankowski, N. W. and Wester, F. (1991) ‘TheQualitative Tradition in Social Science Inquiry:Contributions to Mass Communication Research,’in K. B. Jensen and N. W. Jankowski (eds) AHandbook of Qualitative Methodologies for MassCommunication Research, London: Routledge.

Jankowski, N. W., Prehn, O. and Stappers, J. (eds)(1992) The People’s Voice, London: John Libbey.

Janowitz, M. (ed) (1966) W. I. Thomas on SocialOrganisation and Social Personality: SelectedPapers, Chicago, IL: The University of ChicagoPress.

Janowitz, M. (1967) The Community Press in anUrban Setting (2nd revised edn), Chicago, IL:University of Chicago Press [orig. publ. 1952].

Janson, H. W. (1991) History of Art (4th edn),Englewood Cliffs, NJ: Prentice-Hall.

Jauss, H. R. (1982) Toward an Aesthetic ofReception, Brighton, Sussex: Harvester Press.

Jenkins, H. (1992) Textual Poachers: Television Fansand Participatory Culture, New York: Routledge.

Jensen, E., Braxton, G., and Calvo, D. (2000) ‘NBC,NAACP In Pact to Boost Minorities in TV,’ LosAngeles Times, 6 January, A-1.

Jensen, J. F. (1999) ‘Interactivity: Tracking a NewConcept in Media and Communication Studies,’in P. A. Mayer (ed.) Computer Media and Com-munication: A Reader, Oxford: Oxford UniversityPress.

Jensen, K. B. (1986) Making Sense of the News:Towards a Theory and an Empirical Model ofReception for the Study of Mass Communication,Aarhus, Denmark: University of Aarhus Press.

Jensen, K. B. (1988) ‘News as Social Resource,’ Euro-pean Journal of Communication, 3(3): 275–301.

Jensen, K. B. (1990) ‘Television Futures: A SocialAction Methodology for Studying InterpretiveCommunities,’ Critical Studies in Mass Com-munication, 7(2): 129–146.

Jensen, K. B. (1993a) ‘One Person, One Computer:The Social Construction of the Personal Com-puter,’ in P. B. Andersen, B. Holmqvist, and J. F.

Jensen (eds) The Computer as Medium, Cam-bridge: Cambridge University Press.

Jensen, K. B. (1993b) ‘The Past in the Future:Problems and Potentials of Historical ReceptionStudies,’ Journal of Communication, 43(4): 20–28.

Jensen, K. B. (1994) ‘Reception as Flow: The “NewTelevision Viewer” Revisited,’ Cultural Studies,8(2): 293–305.

Jensen, K. B. (1995) The Social Semiotics of MassCommunication, London: Sage.

Jensen, K. B. (ed.) (1998) News of the World. WorldCultures Look at Television News, London andNew York: Routledge.

Jensen, K. B. (1999) ‘Local Empiricism, GlobalTheory: Problems and Potentials of ComparativeResearch on News Reception,’ Communications:The European Journal of CommunicationResearch, 23(4): 427–445.

Jensen, K. B. (2000a) ‘Interactivities: Constituents of a Model of Computer Media and Com-munication,’ in I. Bondebjerg (ed.) MovingImages, Culture, and the Mind, Luton, Beds:University of Luton Press.

Jensen, K. B. (2000b) ‘On the Edge: A Meta-Analysisof the State of Media and CommunicationResearch,’ Nordicom Review, 21(2): 23–30.

Jensen, K. B. (2001) ‘Why Virtuality Can Be Goodfor Democracy,’ in S. Hjarvard (ed.) News in aGlobalized Society, Gothenburg, Sweden: Nordi-com.

Jensen, K. B. and Jankowski, N. W. (eds) (1991) AHandbook of Qualitative Methodologies for MassCommunication Research, London: Routledge.

Jensen, K. B., Schrøder, K. C., Stampe, T., Sønder-gaard, H., and Topsøe-Jensen, J. (1994) ‘SuperFlow, Channel Flows, and Audience Flows: AStudy of Viewers’ Reception of Television asFlow,’ Nordicom Review, 2: 1–13.

Joas, H. (1993) Pragmatism and Social Theory,Chicago, IL: University of Chicago Press.

Johansen, J. D. (1993) Dialogic Semiosis, Bloom-ington: Indiana University Press.

Johnson, L. (1988) The Unseen Voice: A CulturalHistory of Early Australian Radio, London:Routledge.

Johnson, M. (1987) The Body in the Mind, Chicago,IL: University of Chicago Press.

Jones, S. G. (ed.) (1998) Cybersociety 2.0, ThousandOaks, CA: Sage.

Jones, S. G. (ed.) (1999) Doing Internet Research,Thousand Oaks, CA: Sage.

Jordan, A. (1992) ‘Social Class, Temporal Orienta-tion, and Mass Media Use within the FamilySystem,’ Critical Studies in Mass Communication,9(4): 374–386.

References 309

1111234567891012341115678920123456789301234567894012345678950111

Jowett, G. S., Jarvie, I. C., and Fuller, K. H. (1996)Children and the Movies: Media Influence and thePayne Fund Controversy, Cambridge: CambridgeUniversity Press.

Jungk, R. and Müllert, N. (1981) Zukunft-werkstätten [Workshops on the Future], Ham-burg, Germany: Hoffmann & Campe.

Kalinak, K. (1992) Settling the Score: Music and theClassical Hollywood Film, Madison: University ofWisconsin Press.

Kaminsky, S. M. and Mahan, J. H. (1985) AmericanTelevision Genres, Chicago, IL: Nelson-Hall.

Kamwangamulu, N. M. (1999) ‘Ubuntu in SouthAfrica: a Sociolinguistic Perspective to a Pan-African Concept,’ Critical Arts 3 (2): 24–41,Durban: University of Natal (www.und.ac.za/und/ccms).

Kaniss, P. (1991) Making Local News, Chicago, IL:University of Chicago Press.

Kaplan, E. A. (1988) Rocking Around the Clock:Music Television, Postmodernism, and ConsumerCulture, New York: Routledge.

Katz, E. and Lazarsfeld, P. F. (1955) PersonalInfluence, Glencoe, IL: Free Press.

Katz, E. and Wedell, G. (1977) Broadcasting in theThird World: Promise and Performance, Cam-bridge, MA: Harvard University Press.

Katz, E., Blumler, J., and Gurevitch, M. (1974)‘Utilization of Mass Communication by theIndividual,’ in J. G. Blumler and E. Katz (eds) TheUses of Mass Communications, Beverly Hills, CA:Sage.

Katz, E., Gurevitch, M., and Haas, H. (1973) ‘Onthe Use of Mass Media for Important Things,’American Sociological Review, 38(2): 164–181.

Kay, A. and Goldberg, A. (1999) ‘Personal DynamicMedia,’ in P. A. Mayer (ed.) Computer Media andCommunication: A Reader, Oxford: Oxford Uni-versity Press [orig. publ. 1977].

Kaye, B. K. and Sapolsky, B. S. (1997) ‘ElectronicMonitoring of In-Home Television RCD Usage,’Journal of Broadcasting and Electronic Media, 41:214–228.

Kelly, C. and Shepherd, D. (eds) (1998) RussianCultural Studies: An Introduction, Oxford:Oxford University Press.

Kennedy, G. A. (1980) Classical Rhetoric and ItsChristian and Secular Tradition from Ancient toModern Times, Chapel Hill: University of NorthCarolina Press.

Kepplinger, H.-M. (1997) ‘Political Correctness andAcademic Principles: A Reply to Simpson,’ Jour-nal of Communication, 47(4): 102–117.

Kerlinger, F. N. (1986) Foundations of BehaviouralResearch (3rd edn), New York: Holt, Rinehart &Winston.

Kern, S. (1983) The Culture of Time and Space,1880–1918, Cambridge, MA: Harvard UniversityPress.

Kincaid, D. L. (ed.) (1987) Communication Theory:Eastern and Western Perspectives, San Diego, CA:Academic Press.

Kinder, M. (1 984) ‘Music Video and the Spectator:Television, Ideology, and Dream,’ Film Quarterly,38: 2–15.

Kingsford, P. (1982) The Hunger Marchers in Britain,1920–1940, London: Lawrence & Wishart.

Kirk, J. and Miller, M. (1986) Reliability andValidity in Qualitative Research, Beverly Hills,CA: Sage.

Kittler, F. A. (1999) Gramophone, Film, Typewriter,Stanford, CA: Stanford University Press.

Kjørup, S. (1996) Menneskevidenskabeme: Prob-lemer og traditioner i humanioras videnskabsteori[The Human Sciences: Problems and Traditions inthe Theory of Science of the Humanities],Roskilde, Denmark: Roskilde Universitetsforlag.

Klapper, J. (1960) The Effects of Mass Com-munication, Glencoe, IL: Free Press.

Klingender, F. D. and Legg, S. (1937) Money Behindthe Screen, London: Lawrence & Wishart.

Knapp, M. L. and Miller, G. R. (eds) (1994) Hand-book of Interpersonal Communication (2nd edn),Thousand Oaks, CA: Sage.

Koss, S. (1981, 1984) The Rise and Fall of thePolitical Press in Britain (Vols 1 and 2), London:Hamish Hamilton.

Kozloff, S. (1988) Invisible Storytellers: Voice-overNarration in American Fiction Film, Berkeley, CA:University of California Press.

Kozloff, S. (2000) Overhearing Film Dialogue,Berkeley, CA: University of California Press.

Kracauer, S. (1953) ‘The Challenge of QualitativeContent Analysis,’ Public Opinion Quarterly,16(2): 631–642.

Kracauer, S. (1960) Theory of Film: The Redemptionof Physical Reality, London: Oxford UniversityPress.

Kracauer, S. (1995) The Mass Ornament: WeimarEssays, Cambridge, MA: Harvard UniversityPress [orig. publ. 1963].

Kraus, S. and Davis, D. (1976) The Effects of MassCommunication on Political Behavior, UniversityPark, PA: Pennsylvania State University Press.

Kress, G. and van Leeuwen, T. (1996) ReadingImages: The Grammar of Visual Design, London:Routledge.

Kress, G., Leite-García, R. and van Leeuwen, T.(1997) ‘Discourse semiotics,’ in T. van Dijk (ed.)Discourse as Structure and Process, London: Sage.

References310

Krippendorf, K. (1980) Content Analysis: AnIntroduction to its Methodology, Beverly Hills,CA: Sage.

Kristeller, P. O. (1961) Renaissance Thought: TheClassic, Scholastic, and Humanist Strains, NewYork: Harper.

Kristeva, J. (1984) Revolution in Poetic Language,New York: Columbia University Press [orig. publ.1974].

Kroeber, A. L. and Kluckhohn, C. (1952) Culture: ACritical Review of Concepts and Definitions,Cambridge, MA: Peabody Museum of AmericanArchaeology and Ethnology.

Kubey, R. and Csikszentmihalyi, M. (1990) Tele-vision and the Quality of Life: How ViewingShapes Everyday Experience, Hillsdale, NJ:Lawrence Erlbaum.

Kuhn, T. S. (1970) The Structure of ScientificRevolutions (revised edn), Chicago, IL: Universityof Chicago Press [orig. publ. 1962].

Kuiper, K. (1996) Smooth Talkers: The LinguisticPerformance of Auctioneers and Sportscasters,Mahwah, NJ: Erlbaum.

Kvale, S. (1987) ‘Validity in the Qualitative ResearchInterview,’ Methods, 1: 37–72.

Kvale, S. (1996) InterViews: An Introduction toQualitative Research Interviewing, ThousandOaks, CA: Sage.

Labov, W. and Waletzky, J. (1967) ‘NarrativeAnalysis,’ in J. Helm (ed.) Essays on the Verbaland Visual Arts, Seattle: University of WashingtonPress.

Lacan, J. (1977) The Four Fundamental Concepts ofPsychoanalysis, Harmondsworth: Penguin.

Lacey, K. (1994) ‘From Plauderei to Propaganda: OnWomen’s Radio in Germany, 1924–35,’ MediaCulture & Society, 16(4): 589–608.

Lakoff, G. and Johnson, M. (1999) Philosophy inthe Flesh: The Embodied Mind and Its Challengeto Western Thought, New York: Basic Books.

Landow, G. P. (1997) Hypertext 2.0: The Converg-ence of Contemporary Critical Theory andTechnology (2nd edn), Baltimore, MD: JohnsHopkins University Press.

Lang, G. E. and Lang, K. (1981) ‘Watergate: AnExploration of the Agenda-Building Process,’ inG. C. Wilhoit and H. D. Bock (eds) MassCommunication Review Yearbook (Vol. 2),Beverly Hills, CA: Sage.

Lang, K. and Lang, G. E. (1953) ‘The UniquePerspective of Television and Its Effects: A PilotStudy,’ American Sociological Review, 18: 3–12.

Lang, K. and Lang, G. E. (1991) ‘Studying Events inTheir Natural Settings,’ in K. B. Jensen and N. W.Jankowski (eds) A Handbook of Qualitative

Methodologies for Mass Communication Re-search, London: Routledge.

Lange, D. L., Baker, R. K., and Ball, S. J. (1969)Mass Media and Violence, Washington, DC: U.S.Government Printing Office.

Langer, J. (1998) Tabloid Television: Popular Jour-nalism and the ‘Other’ News, London: Routledge

Larsen, P. (1974) ‘Analyse af TV-avisen’ [Analysis ofTelevision News], in M. B. Andersen and J.Poulsen (eds) Mediesociologi. Introduktion tilmassekommunikationsforskning [Media Socio-logy: Introduction to Mass CommunicationResearch], Copenhagen, Denmark: Rhodos.

Larsen, P. (1989) ‘Beyond the Narrative. RockVideos and Modern Visual Fictions: Readings,Experiences,’ in M. Skovmand (ed.) Media Fictions, Aarhus, Denmark: Aarhus UniversityPress.

Larsen, P. (2000) ‘The Sound of Images: ClassicalHollywood and Music,’ in I. Bondebjerg (ed.)Moving Image, Culture and The Mind, Luton,Beds: University of Luton Press.

Lashley, K. S. and Watson, J. B. (1922) A Psycho-logical Study of Motion Pictures in Relation toVenereal Disease Campaigns, Washington, DC:Interdepartmental Social Hygiene Board.

Lasorsa, D. and Reese, S. (1990) ‘News Source Usein the Crash of 1987: A Study of Four NationalMedia,’ Journalism Quarterly, 167: 60–71.

Lasswell, H. D. (1938) Propaganda Technique in theWorld War, New York: Alfred A. Knopf [orig.publ. 1927].

Lasswell, H. D. (1966) ‘The Structure and Functionof Communication in Society,’ in B. Berelson andM. Janowitz (eds) Reader in Public Opinion andCommunication, Glencoe, IL: The Free Press[orig. publ. 1948].

Latour, B. (1987) Science in Action, Milton Keynes:Open University Press.

Lazarsfeld, P. F. (1934) ‘The Psychological Aspect ofMarket Research,’ Harvard Business Review,13(1): 54–71.

Lazarsfeld, P. F. (1941) ‘Remarks on Administrativeand Critical Communications Research,’ Studiesin Philosophy and Social Science, 9: 2–16.

Lazarsfeld, P. F. and Barton, A. H. (1951)‘Qualitative Measurement in the Social Sciences:Classification, Typologies, and Indices,’ in D.Lerner and H. D. Lasswell (eds) The PolicySciences: Recent Developments in Scope andMethod, Stanford, CA: Stanford University Press.

Lazarsfeld, P. F. and Merton, R. K. (1960) ‘MassCommunication, Popular Taste, and OrganisedSocial Action,’ in W. Schramm (ed.) MassCommunications, Urbana: University of IllinoisPress [orig. publ. 1948].

References 311

1111234567891012341115678920123456789301234567894012345678950111

Lazarsfeld, P. F., Berelson, B., and Gaudet, H. (1944)The People’s Choice, New York: Duell, Sloan &Pearce.

Lee, J. B. and Mandelbaum, M. (1999) Seeing isBelieving: 700 Years of Scientific and MedicalIllustration, New York: New York Public Library.

Lefkowitz, M., Eron, L. D., Walder, L. O., andHuesmann, L. R. (1972) ‘Television Violence andChild Aggression,’ in E. Rubinstein, G. Comstockand J. Murray (eds) Television and AdolescentAggressiveness, Washington, DC: U.S. Govern-ment Printing Office.

Leiss, W., Kline, S., and Jhally, S. (1986) SocialCommunication in Advertising: Persons, Productsand Images of Well-Being, London: Methuen.

Lembo, R. (2000) Thinking Through Television,Cambridge: Cambridge University Press.

Lemert, C. (1993) Social Theory, San Francisco, CA:Westview Press.

Lemish, D. (1982) ‘The Rules of Viewing Televisionin Public Places,’ Journal of Broadcasting andElectronic Media, 26: 757–791.

Lerner, D. (1958) The Passing of Traditional Society,New York: Free Press.

Lerner, D. and Lasswell, H. D. (eds) (1951) ThePolicy Sciences: Recent Developments in Scopeand Method, Stanford, CA: Stanford UniversityPress.

Lévi-Strauss, C. (1958) Structural Anthropology,New York: Penguin.

Lévi-Strauss, C. (1991) Totemism, London: MerlinPress [orig. publ. 1962].

Levine, L. W. (1988) Highbrow/Lowbrow: TheEmergence of Cultural Hierarchy in America,Cambridge, MA: Harvard University Press.

Levy, M. (ed.) (1992) ‘Symposium: CommunicationScholarship and Political Correctness,’ Journal ofCommunication, 42(2): 56–149.

Levy, M. and Gurevitch, M. (eds) (1994) DefiningMedia Studies: Reflections on the Future of theField, Oxford: Oxford University Press.

Levy, M. and Windahl, S. (1985) ‘The Concept ofAudience Activity,’ in K. E. Rosengren, P. Palm-green, and L. Wenner (eds) Media GratificationsResearch: Current Perspectives, Beverly Hills, CA:Sage.

Lewis, J. (1983) ‘The Encoding/Decoding Model:Criticisms and Redevelopments for Research onDecoding,’ Media, Culture & Society, 5(2): 179–197.

Lewis, J. (1985) ‘Decoding Television News,’ in P.Drummond and R. Paterson (eds) Television inTransition, London: British Film Institute.

Lewis, J. (1991) The Ideological Octopus: AnExploration of Television and Its Audience, NewYork: Routledge.

Lewis, J. (1997) ‘What Counts in Cultural Studies,’Media, Culture & Society, 19(1): 83–97.

Lewis, L. (ed.) (1991) The Adoring Audience,London: Routledge.

Lichtenstein, A. and Rosenfeld, L. B. (1983) ‘Usesand Misuses of Gratifications,’ CommunicationResearch, 10(1): 97–109.

Licklider, J. C. R. and Taylor, R. W. (1999) ‘TheComputer as a Communication Device,’ in P. A.Mayer (ed.) Computer Media and Communi-cation: A Reader, Oxford: Oxford UniversityPress [orig. publ. 1968].

Liebes, T. and Katz, E. (1990) The Export ofMeaning, New York: Oxford University Press.

Lievrouw, L. and Livingstone, S. (eds) (2002)Handbook of New Media: Social Shaping andSocial Consequences, London: Sage.

Lincoln, Y. S. and Guba, E. G. (1985) NaturalisticInquiry, London: Sage.

Lindlof, T. R. (1987) ‘Ideology and Pragmatics ofMedia Access in Prison,’ in T. R. Lindlof (ed.)Natural Audiences: Qualitative Research ofMedia Uses and Effects, Norwood, NJ: Ablex.

Lindlof, T. R. (1995) Qualitative CommunicationResearch Methods, Thousand Oaks, CA: Sage.

Lippmann, W. and Merz, C. (1920) ‘A Test of theNews: An Examination of the News Reports inthe New York Times on Aspects of the RussianRevolution of Special Importance to AmericansMarch 1917–March 1920,’ The New Republic[Special Supplement], Vol XXIII, Part II, No 296,4 August.

Livingstone, S. (1990) Making Sense of Television:The Psychology of Audience Appreciation,Oxford: Pergamon Press.

Livingstone, S. (1998) Making Sense of Television:The Psychology of Audience Interpretation (2ndedn), London: Routledge.

Livingstone, S. and Bovill, M. (eds) (2001) Childrenand their Changing Media Environment: A Euro-pean Comparative Study, Hillsdale, NJ: LawrenceErlbaum.

Livingstone, S. and Lunt, P. (1994) Talk on Tele-vision, London: Routledge.

Lobkowicz, N. (1967) Theory and Practice: Historyof a Concept from Aristotle to Marx, NotreDame: University of Notre Dame Press.

Lothe, J. (2000) Narrative in Fiction and Film: AnIntroduction, Oxford: Oxford University Press.

Lovejoy, A. O. (1936) The Great Chain of Being,Cambridge, MA: Harvard University Press.

Lowenthal, L. (1961) ‘The Triumph of Mass Idols,’in L. Lowenthal, Literature, Popular Culture, andSociety, Englewood Cliffs, NJ: Prentice-Hall [orig.publ. 1941].

References312

Lowery, S. A. and DeFleur, M. L. (1983) Milestonesin Mass Communication Research: Media Effects,(2nd edn), New York: Longman.

Lowery, S. A. and DeFleur, M. L. (1995) Milestonesin Mass Communication Research: Media Effects(3rd edn), New York: Longman.

Lukács, G. (1989) The Historical Novel, London:Merlin Press [orig. publ. 1916].

Lull, J. (1980) ‘The Social Uses of Television,’Human Communication Research, 6: 197–209.

Lull, J. (1988a) ‘Critical Response: The Audience asNuisance,’ Critical Studies in Mass Communi-cation, 5: 239–243.

Lull, J. (ed.) (1988b) World Families WatchTelevision, Newbury Park, CA: Sage.

Lull, J. (1991) China Turned On: Television, Re-form, and Resistance, London: Routledge.

Lutz, C. A. and Collins, J. L. (1993) ReadingNational Geographic, Chicago, IL: University ofChicago Press.

Lynd, R. S. and Lynd, H. M. (1929) Middletown: AStudy in American Culture, London: Constable.

Lyotard, J.-F. (1984) The Postmodern Condition,Minneapolis: University of Minnesota Press [orig.publ. 1979].

MacBeth, T. M. (1998) ‘Quasi-experimentalResearch on Television and Behavior: Natural andField Experiments,’ in J. K. Asamen and G. L.Berry (eds) Research Paradigms, Television andSocial Behavior, Thousand Oaks, CA: Sage.

MacBride, S. (ed.) (1980) Many Voices, One World,Paris: UNESCO.

MacCabe, C. (ed.) (1986) High Theory/LowCulture: Analysing Popular Television and Film,Manchester: Manchester University Press.

MacDonald, M. R. (ed.) (1998) Traditional Story-telling Today: An International Sourcebook,Chicago, IL: Fitzroy Dearborn.

McAdams, D. (1999) ‘Opportunity of a Lifetime,’Broadcasting and Cable, 18 October: 23–28.

McAvoy, K. (2000) ‘TV Economics,’ Broadcastingand Cable, 27 November: 54–66.

McCarthy, E. D., Langner, T. S., Gerstein, J. C.,Eisenberg, V. G., and Orzeck, L. (1975) ‘Violenceand Behaviour Disorders,’ Journal of Communi-cation, 25(4): 71–85.

McChesney, R. (1999) Rich Media, Poor Demo-cracy: Communication Politics in Dubious Times,Urbana: University of Illinois Press.

McCloud, S. (1994) Understanding Comics, NewYork: HarperCollins.

McCombs, M. E. and Shaw, D. L. (1972) ‘TheAgenda-Setting Function of the Press,’ PublicOpinion Quarterly, 36: 176–187.

McCombs, M. E., Lopez-Escobar, E. and Llamas, J.P. (2000) ‘Setting the Agenda of Attributes in the1996 Spanish General Election,’ Journal ofCommunication, 50(2): 77–92.

McGuire, W. J. (1973) ‘Persuasion, Resistance, andAttitude Change,’ in I. D. S. Pool (ed.) Handbookof Communication, Chicago, IL: Rand McNally.

McLeod, J. M., Atkin, C. K., and Chaffee, S. H.(1972) ‘Adolescents, Parents and Television Use:Adolescent Self-report Measures from Marylandand Wisconsin Samples,’ in G. A. Comstock andE. A. Rubinstein (eds) Television and SocialBehaviour, Vol. 3, Television and AdolescentAggressiveness, Washington, DC: U.S. Govern-ment Printing Office.

McLuhan, M. (1962) The Gutenberg Galaxy,Toronto: Toronto University Press.

McLuhan, M. (1964) Understanding Media, NewYork: McGraw-Hill.

McNelly, J. T. (1959) ‘Intermediary Communicatorsin the International Flow of News,’ JournalismQuarterly, 36: 23–26.

McQuail, D. (1976) Toward a Sociology of MassCommunications, London: Collier-Macmillan.

McQuail, D. (1983) Mass Communication Theory:An Introduction, London: Sage.

McQuail, D. (2000) McQuail’s Mass Communi-cation Theory (4th edn), London: Sage.

McQuail, D. and Windahl, S. (1993) Communi-cation Models for the Study of Mass Com-munication (2nd edn), London: Longman.

McQuail, D., Blumler, J. G., and Brown, J. R. (1972)‘The Television Audience: A Revised Perspective,’in D. McQuail (ed.) Sociology of Mass Com-munications, Harmondsworth: Penguin.

McRobbie, A. (1991) Feminism and Youth Culture:From ‘Jackie’ to ‘Just Seventeen’, Basingstoke,Hants: Macmillan.

Malinowski, B. (1922) Argonauts of the WesternPacific, London: Routledge.

Malinowski, B. (1967) A Diary in the Strict Sense ofthe Term, London: Routledge & Kegan Paul.

Manheim, J. B. (1998) ‘The News Shapers: StrategicCommunication as Third Force in Newsmaking,’in D. Graber, D. McQuail, and P. Norris (eds) The Politics of News: the News of Politics,Washington, DC: Congressional Quarterly Press.

Mann, C. and Stewart, F. (2000) Internet Com-munication and Qualitative Research: A Hand-book for Researching Online, London: Sage.

Mannheim, K. (1976) Ideology and Utopia,London: Routledge & Kegan Paul [orig. publ.1922].

Manzella, J. C. (1997) Mediating Newspeople:Cultures of Writing and the Mechanisms of

References 313

1111234567891012341115678920123456789301234567894012345678950111

Change at Three Daily Newspapers, unpublishedPh.D. dissertation, University of Connecticut.

Marcus, G. E. and Fischer, M. M. J. (1999)Anthropology as Cultural Critique (2nd edn),Chicago: University of Chicago Press.

Marks, E. and de Courtivron, I. (eds) (1981) NewFrench Feminisms, Brighton: Harvester Press.

Marshall, C. and Rossman, G. B. (1999) DesigningQualitative Research (3rd edn), Thousand Oaks,CA: Sage.

Martin, J. K. and Nakayama, T. K. (1999) ‘ThinkingDialectically About Culture and Communication,’Communication Theory, 9(1): 1–25.

Martin-Barbero, J. (1993) Communication, Culture,and Hegemony, London: Sage.

Martinet, A. (1964) Elements of General Linguis-tics, London: Faber and Faber [orig. publ. 1960].

Mast, G., Cohen, M., and Braudy, L. (eds) (1999)Film Theory and Criticism: Introductory Readings(5th edn), New York: Oxford University Press.

Masterman, L. (1985) Teaching the Media, London:Comedia.

Mathieson, M. (1975) The Preachers of Culture: A Study of English and its Teachers, London:George Allen & Unwin.

Mayer, J. P. (1948) British Cinemas and TheirAudiences: Sociological Studies, London: DenisDobson.

Mayer, P. A. (ed.) (1999) Computer Media and Com-munication: A Reader, Oxford: Oxford UniversityPress.

Mead, G. H. (1934) Mind, Self, and Society,Chicago, IL: University of Chicago Press.

Meadel, C. (1994) Histoire de la radio dans lesannées trente [History of Radio During the1930s], Paris: Anthropos/INA.

Megarry, T. (1995) Society in Prehistory: TheOrigins of Human Culture, London: Macmillan.

Meinhof, U. H. and Richardson, K. (eds) (1999)Worlds in Common?, London: Routledge.

Melnik, S. R. (1981) Eurovision News and theInternational Flow of Information, Bochum:Studienverlag Dr. N. Brockmeyer.

Merleau-Ponty, M. (1962) Phenomenology ofPerception, London: Routledge & Kegan Paul[orig. publ. 1945].

Merten, K. (1996) ‘Reactivity in Content Analysis,’Communications, 21(1): 65–76.

Merton, R. K. (1949) ‘Patterns of Influence: A Studyof Interpersonal Influence and CommunicationsBehaviour in a Local Community,’ in P. F.Lazarsfeld and F. Stanton (eds) CommunicationsResearch, 1948–49, New York: Harper Brothers.

Merton, R. K. (1968) Social Theory and SocialStructure (enlarged edn), New York: Free Press.

Merton, R. K. (1987) ‘The Focussed Interview andFocus Groups: Continuities and Discontinuities,’Public Opinion Quarterly, 51(4): 550–566.

Merton, R. K. and Kendall, P. L. (1946) ‘TheFocused Interview,’ American Journal of Socio-logy, 51: 541–557

Merton, R. K. and Kendall, P. L. (1955) ‘TheFocused Interview,’ in P. F. Lazarsfeld and B.Rosenberg (eds) The Language of Social Research,Glencoe, IL: Free Press [orig. publ. 1946].

Merz, C. (1926) ‘When American Movies GoAbroad,’ Harpers, January: 159–165.

Messaris, P. (1994) Visual ‘Literacy’: Image, Mind,and Reality, Boulder, CO: Westview Press.

Messaris, P. (1997) Visual Persuasion, ThousandOaks, CA: Sage.

Metz, C. (1974) [1968] Film Language: A Semioticsof the Cinema, New York: Oxford UniversityPress.

Metz, C. (1982) The Imaginary Signifier: Psycho-analysis and the Cinema, Bloomington: IndianaUniversity Press.

Meyrowitz, J. (1985) No Sense of Place: The Impactof Electronic Media on Social Behavior, NewYork: Oxford University Press.

Meyrowitz, J. (1993) ‘Images of Media: HiddenFerment – and Harmony – in the Field,’ Journal ofCommunication, 43(3): 55–66.

Meyrowitz, J. (1994) ‘Medium Theory,’ in D.Crowley and D. Mitchell (eds) CommunicationTheory Today, Cambridge: Polity Press.

Mick, D. G. and Buhl, C. (1992) ‘A Meaning-BasedModel of Advertising Experience,’ Journal ofConsumer Research, 19: 317–338.

Middleton, R. (1990) Studying Popular Music,Milton Keynes: Open University Press.

Milavsky, J. R., Kessler, R., Stipp, H., and Rubens,W. S. (1982) Television and Aggression: Results ofa Panel Study, New York: Academic Press.

Miles, M. B. and Huberman, A. M. (1994) Qualita-tive Data Analysis: An Expanded Sourcebook(2nd edn), Thousand Oaks, CA: Sage.

Mill, J. S. (1973–1974) Collected Works of JohnStuart Mill (Vols 7 and 8), Toronto: University ofToronto Press.

Miller, D. and Slater, D. (2000) The Internet: AnEthnographic Approach, Oxford: Berg.

Mills, C. W. (1959) The Power Elite, Oxford:Oxford University Press.

Mills, C. W. (1970) The Sociological Imagination,Harmondsworth: Penguin [orig. publ. 1959].

Millum, T. (1975) Images of Woman: Advertising inWomen’s Magazines, London: Chatto & Windus.

Moi, T. (1985) Sexual/Textual Politics: FeministLiterary Theory, London: Methuen.

References314

Molotch, H. and Lester, M. (1974) ‘News as Pur-posive Behavior,’ American Sociological Review,39: 101–112.

Monteiro, A. and Jayasankar, J. P. (1994) ‘TheSpectator-Indian: An Exploratory Study on theReception of News,’ Cultural Studies, 8(1): 162–182.

Montgomery, K. (1989) Target: Prime Time: Advo-cacy Groups and the Struggle Over EntertainmentTelevision, New York: Oxford University Press.

Montgomery, M. (1986) ‘DJ Talk,’ Media, Cultureand Society, 8: 421–440.

Moores, S. (1988) ‘“The Box on the Dresser”:Memories of Radio and Everyday Life,’ MediaCulture & Society, 10(1): 23–40.

Moores, S. (1993) Interpreting Audiences: TheEthnography of Media Consumption, London:Sage.

Moran, A. (ed.) (1996) Film Policy: International,National, and Regional Perspectives, London andNew York: Routledge.

Morgan, D. L. and Krueger, R. A. (1998) The FocusGroup Kit, Thousand Oaks, CA: Sage.

Morley, D. (1980) The ‘Nationwide’ Audience,London: British Film Institute.

Morley, D. (1981) ‘“The Nationwide Audience”: ACritical Postscript,’ Screen Education, 39: 3–14.

Morley, D. (1986) Family Television, London:Comedia.

Morley, D. (2000) Home Territories: Media, Mobil-ity, and Identity, London: Routledge.

Morley, D. and Chen, K.-H. (eds) (1996) Stuart Hall:Critical Dialogues in Cultural Studies, London:Routledge.

Morley, D. and Robins, K. (1995) Spaces of Iden-tity: Global Media, Electronic Landscapes, andCultural Boundaries, London and New York:Routledge.

Morris, M. (1990) ‘Banality in Cultural Studies,’ inP. Mellencamp (ed.) Logics of Television,Bloomington: Indiana University Press.

Morrison, D. (1992) Television and the Gulf War,London: John Libbey.

Morrison, D. (1998) The Search for a Method: FocusGroups and the Development of Mass Com-munication Research, Luton, Beds: University ofLuton Press.

Morrison, D. and Tumber, H. (1988) Journalists atWar, London: Sage.

Morse, M. (1998) Virtualities: Television, MediaArt, and Cyberculture, Bloomington: IndianaUniversity Press.

Mortensen, F. (1977) ‘The Bourgeois Public Sphere:A Danish Mass Communications ResearchProject,’ in M. Berg, P. Hemanus, and J. Ekecrantz

(eds) Current Theories in Scandinavian MassCommunication, Grenaa, Denmark: GMT.

Mosco, V. (1996) The Political Economy of Com-munication, London: Sage.

Moser, C. A. and Kalton, G. (1971) Survey Methodsin Social Investigation, Aldershot, Hants: Gower.

Mowlana, H. (1993) ‘The New Global Order andCultural Ecology,’ Media, Culture & Society,15(1): 9–27.

Mowlana, H. (1997) Global Information and WorldCommunication (2nd edn), London: Sage.

Mowlana, H. and Wilson, L. J. (1990) The Passingof Modernity, New York: Longman.

Mulvey, L. (1992) ‘Visual Pleasure and NarrativeCinema,’ in G. Mast, M. Cohen, and L. Braudy(eds) Film Theory and Criticism: IntroductoryReadings (4th edn), New York: Oxford UniversityPress [orig. publ. 1975].

Münsterberg, H. (1970) The Film: A PsychologicalStudy, New York: Dover Publications [orig. publ.1916].

Murdock, G. (1982) ‘Large Corporations and theControl of Communications Industries,’ in M. Gurevitch, T. Bennett, J. Curran, and J.Woollacott (eds) Culture, Society, and the Media,London: Methuen.

Murdock, G. (1997) ‘Thin Descriptions: Questionsof Method in Cultural Analysis,’ in J. McGuigan(ed.) Cultural Methodologies, London: Sage.

Murdock, G. (2000) ‘Talk Shows: DemocraticDebates and Tabloid Tales,’ in J. Wieten, G.Murdock, and P. Dahlgren (eds) Television AcrossEurope: A Comparative Introduction, London:Sage.

Murdock, G. and Golding, P. (1977) ‘Capitalism,Communication, and Class Relations,’ in J.Curran, M. Gurevitch, and J. Woollacott (eds)Mass Communication and Society, London:Edward Arnold.

Murdock, G. and Pickering, M. (eds) (forthcoming)Machineries of Experience: Media and The Mak-ing of Modernity, Oxford: Blackwell.

Murray, J. H. (1997) Hamlet on the Holodeck, NewYork: Free Press.

Myers, G. (1994) Words in Ads, London: EdwardArnold.

Nagel, T. (1986) The View from Nowhere, Oxford:Oxford University Press.

Naremore, J. (ed.) (2000) Film Adaptation, London:Athlone Press.

National Television Violence Study (Vol.1) (1997)Thousand Oaks, CA: Sage.

Neale, S. (1980) Genre, London: British FilmInstitute.

References 315

1111234567891012341115678920123456789301234567894012345678950111

Neale, S. (2000) Genre and Hollywood, New York:Routledge.

Negroponte, N. (1995) Being Digital, London:Hodder & Stoughton.

Negt, O. and Kluge, A. (1993) Public Sphere andExperience, Minneapolis: University of Minne-sota Press [orig. publ. 1972].

Nelson, J. S., Megill, A., and McCloskey, D. M. (eds)(1987) The Rhetoric of the Human Sciences,Madison: University of Wisconsin Press.

Nerone, J. C. (ed.) (1995) Last Rights: RevisitingFour Theories of the Press, Urbana: University ofIllinois Press.

Neuman, W. L. (1994) Social Research Methods:Qualitative and Quantitative Approaches (2ndedn), Boston, MA: Allyn & Bacon.

Neuman, W. R. (1976) ‘Patterns of Recall amongTelevision News Viewers,’ Public OpinionQuarterly, 40(1): 115–123.

Neuman, W. R. (1989) ‘Parallel Content Analysis:Old Paradigms and New Proposals,’ in G. A.Comstock (ed.) Public Communication andBehavior (Vol. 2), Orlando, FL: Academic Press.

Neuman, W. R., Just, M., and Crigler, A. N. (1992)Common Knowledge: News and the Constructionof Political Meaning, Chicago, IL: University ofChicago Press.

Newcomb, H. (1974) Television. The Most PopularArt, Garden City, NY: Anchor Press.

Newcomb, H. (1978) ‘Assessing the Violence ProfileStudies of Gerbner and Gross: A HumanisticCritique and Suggestion,’ CommunicationResearch, 5(3): 264–282.

Newcomb, H. (ed.) (1997) Encyclopedia ofTelevision, London: Fitzroy Dearborn.

Newcomb, H. and Alley, R. (1983) The Producer’sMedium: Conversations with Creators of Ameri-can TV, New York: Oxford University Press.

Newcomb, H. and Hirsch, P. (1984) ‘Television as aCultural Forum: Implications for Research,’ in W. D. Rowland and B. Watkins (eds), InterpretingTelevision, Beverly Hills, CA: Sage.

Ngugi, W. (1986) Decolonizing the Mind, London:James Currie and Heinemann.

Nicholson, L. (ed.) (1997) The Second Wave: AReader in Feminist Theory, London: Routledge.

Nnaemeka, T. and Richstad, J. (1980) ‘StructuredRelations and Foreign Newsflow in the PacificRegion,’ Gazette, 26(4): 235–257.

Noelle-Neumann, E. (1984) The Spiral of Silence,Chicago, IL: University of Chicago Press.

Nofsinger, R. E. (1991) Everyday Conversation,Newbury Park, CA: Sage.

Nohrstedt, S. A. and Ottosen, R. (eds) (2000)Journalism and the New World Order: Gulf War,

National News Discourses, and Globalization,(Vol. 1), Gothenburg, Sweden: Nordicom.

Nora, S. and Minc, A. (1980) The Computerizationof Society, Cambridge, MA: MIT Press.

Nordenstreng, K. (1972) ‘Policy for News Trans-mission,’ in D. McQuail (ed.) Sociology of MassCommunications, Harmondsworth: Penguin.

Nordenstreng, K. (1984) ‘Defining the New Inter-national Information Order,’ in G. Gerbner andM. Siefert (eds) World Communication: AHandbook, New York: Longman.

Nordenstreng, K. (1997) ‘Beyond Four Theories ofthe Press,’ in J. Servaes and R. Lie (eds) Media andPolitics in Transition, Leuven, Belgium: Acco.

Nordenstreng, K. and Schiller, H. I. (eds) (1979)National Sovereignty and International Com-munication, Norwood, NJ: Ablex.

Nordenstreng, K. and Varis, T. (1974) TelevisionTraffic – A One-way Street? A Survey andAnalysis of the International Flow of TelevisionProgramme Material, Reports and Papers onMass Communication, no. 70, Paris: UNESCO.

O’Donnell, H. (1999) Good Times, Bad Times: SoapOperas and Society in Western Europe, London:Leicester University Press.

O’Sullivan, T. (1991) ‘Television, Memories and theCulture of Viewing,’ in J. Corner (ed.) PopularTelevision in Britain, London: British Film Insti-tute.

Ong, W. (1982) Orality and Literacy: The Tech-nologizing of the Word, New York: CornellUniversity Press.

Oppenheim, A. N. (1992) Questionnaire Design,Interviewing and Attitude Measurement, London:Pinter.

Outhwaite, W. (1975) Understanding Social Life:The Method Called Verstehen, London: GeorgeAllen & Unwin.

Palmgreen, P. and Rayburn, J. D. (1985) ‘AnExpectancy–Value Approach to Media Gratifi-cations,’ in K. E. Rosengren, L. Wenner, and P.Palmgreen (eds) Media Gratifications Research:Current Perspectives, Beverly Hills, CA: Sage.

Palmgreen, P., Wenner, L. and Rayburn, J. D. (1980)‘Relations Between Gratifications Sought andObtained: A Study of Television News,’ Com-munication Research, 7(2): 161–192.

Pan, Z. (2000) ‘Spatial Configuration in InstitutionalChange: A Case of China’s Journalism Reforms,’Journalism, 1(3): 253–281.

Parameswaran, R. (1999) ‘Western Romance Fictionas English-Language Media in Postcolonial India,’Journal of Communication, 49(3): 84–105.

Park, H.-W. (1998) ‘A Gramscian Approach toInterpreting International Communication,’ Jour-nal of Communication, 48(4): 79–99.

References316

Park, R. E. (1922) The Immigrant Press and ItsControl, New York: Harper.

Parkin, F. (1971) Class Inequality and PoliticalOrder, London: MacGibbon & Kee.

Parsons, T. (1951) The Social System, Glencoe, IL:Free Press.

Passmore, J. (1972) ‘Logical Positivism,’ in P.Edwards (ed.) The Encyclopedia of Philosophy(Vol. 5), New York: Macmillan.

Paterson, C. (2001) ‘Media Imperialism Revisited:The Global Public Sphere and the News AgencyAgenda,’ in S. Hjarvard (ed.) News in aGlobalized Society, Gothenburg, Sweden: Nordi-com.

Patterson, T. (1998) ‘Political Roles of the Journ-alist,’ in D. Graber, D. McQuail, and P. Norris(eds) The Politics of News: the News of Politics,Washington, DC: Congressional Quarterly Press.

Patton, M. Q. (1990) Qualitative Evaluation andResearch Methods (2nd edn), Newbury Park, CA:Sage.

Pavitt, C. (1999) ‘The Third Way: Scientific Realismand Communication Theory,’ CommunicationTheory, 9(2): 162–188.

Pearl, D., Bouthilet, L., and Lazar, J. (eds) (1982)Television and Behavior: Ten Years of ScientificProgress and Implications for the Eighties, Wash-ington, DC: U.S. Government Printing Office.

Pearson, R. and Uricchio, W. (eds) (1990) The ManyLives of the Batman, New York: Routledge.

Peirce, C. S. (1931–1958) Collected Papers, Cam-bridge, MA: Harvard University Press.

Peirce, C. S. (1985) ‘Logic as Semiotic: The Theoryof Signs,’ in R. Innis (ed.) Semiotics: An Intro-ductory Anthology, London: Hutchinson.

Peirce, C. S. (1986) Writings of Charles S. Peirce(Vol. 3), Bloomington: Indiana University Press.

Peirce, C. S. (1992) The Essential Peirce (Vol. 1),Bloomington: Indiana University Press.

Peirce, C. S. (1998) The Essential Peirce (Vol. 2),Bloomington: Indiana University Press.

Pelfrey, R. (1985) Art and Mass Media, New York:Harper & Row.

Perelman, C. (1979) The New Rhetoric and theHumanities, Dordrecht, The Netherlands: Reidel.

Peters, J. D. (1999) Speaking into the Air: A Historyof the Idea of Communication, Chicago, IL:University of Chicago Press.

Philo, G. (1990) Seeing and Believing: The Influenceof Television, London: Routledge.

Pike, K. L. (1967) Language in Relation to a UnifiedTheory of the Structure of Human Behavior (2ndedn), The Hague, The Netherlands: Mouton.

Pitt, J. (ed.) (1988) Theories of Explanation, NewYork: Oxford University Press.

Pollock, J. C. and Guidette, C. L. (1980) ‘MassMedia, Crisis, and Political Change: A Cross-National Approach,’ in D. Nimmo (ed.) Com-munication Yearbook (Vol. 4), New Brunswick,NJ: Transaction Books.

Popper, K. R. (1972) The Logic of Scientific Dis-covery, London: Hutchinson [orig. publ. 1934].

Porat, M. (1977) The Information Economy:Definition and Measurement, Washington, DC:U.S. Government Printing Office.

Porter, D. (ed.) (1997) Internet Culture, London:Routledge.

Porter, R. (1991) ‘The History of the Body,’ in P.Burke (ed.) New Perspectives on HistoricalWriting, Cambridge: Polity Press.

Posner, R., Robering, K. and Sebeok, T. A. (eds)(1997–1998) Semiotics: A Handbook of theSign–Theoretic Foundations of Nature and Cul-ture, Berlin: Walter de Gruyter.

Postman, N. (1985) Amusing Ourselves to Death,New York: Viking Press.

Potter, J. (1996) Representing Reality. Discourse,Rhetoric and Social Construction, London: Sage.

Potter, J. and Wetherell, M. (1987) Discourse andSocial Psychology: Beyond Attitudes and Behavi-our, London: Sage.

Potter, J. and Wetherell, M. (1996) ‘DiscourseAnalysis,’ in J. A. Smith, R. Harré, and L. VanLangenhove (eds) Rethinking Methods in Psy-chology, London: Sage.

Potter, W. J. (1996) An Analysis of Thinking andResearch about Qualitative Methods, Mahwah,NJ: Lawrence Erlbaum.

Potter, W. J. (1998) Media Literacy, Thousand Oaks,CA: Sage.

Potter, W. J. and Smith, S. (1999) ‘Consistency ofContextual Cues about Violence across NarrativeLevels,’ Journal of Communication, 49(4): 121–133.

Pound, R. and Frankfurter, F. (eds) (1922) CriminalJustice in Cleveland, Cleveland, OH: The Cleve-land Foundation.

Preece, J. (1994) Human–Computer Interaction,Wokingham: Addison-Wesley.

Press, A. (1991) Women Watching Television:Gender, Class, and Generation in the AmericanTelevision Experience, Philadelphia: University ofPennsylvania Press.

Press, A. (2000) ‘Recent Developments in FeministCommunication Theory: Difference, PublicSphere, Body, and Technology,’ in J. Curran andM. Gurevitch (eds) Mass Media and Society (3rdedn), London: Edward Arnold.

Priest, S. H. (1996) Doing Media Research: AnIntroduction, Thousand Oaks, CA: Sage.

References 317

1111234567891012341115678920123456789301234567894012345678950111

Pritchard, D. and Hughes, K. D. (1997) ‘Patterns ofDeviance in Crime News,’ Journal of Communi-cation, 47(3): 49–67.

Project (1996) The Media History Project, Available:www.mediahistory.com (accessed 2001, March 9).

Propp, V. (1958) Morphology of the Folktale,Bloomington: Research Center, Indiana University[orig. publ. 1928].

Prue, C., Bornat, J., and Wengraf, T. (eds) (2000)The Turn to Biographical Methods in SocialScience, London: Routledge.

Punch, K. F. (1998) Introduction to Social Research:Quantitative and Qualitative Approaches, Lon-don: Sage.

Radway, J. (1984) Reading the Romance: Women,Patriarchy, and Popular Literature, Chapel Hill:University of North Carolina Press.

Radway, J. (1988) ‘Reception Study: Ethnographyand the Problem of Dispersed Audiences andNomadic Subjects,’ Cultural Studies, 2(3): 359–376.

Radway, J. (1997) A Feeling for Books: The Book-of-the-Month Club, Literary Taste, and Middle-Class Desire, Chapel Hill: University of NorthCarolina Press.

Ragin, C. R. (1987) The Comparative Method:Moving Beyond Qualitative and QuantitativeStrategies, Berkeley: University of CaliforniaPress.

Ragin, C. R. (1994) Constructing Social Research:The Unity and Diversity of Method, ThousandOaks, CA: Pine Forge Press.

Rawls, J. (1999) Collected Papers, Cambridge, MA:Harvard University Press.

Ray, L. and Sayer, A. (eds) (1999) Culture andEconomy after the Cultural Turn, London: Sage.

Reason, P. and Bradbury, H. (eds) (2001) Handbookof Action Research: Participative Inquiry andPractice, London: Sage.

Reeves, B. and Nass, C. (1996) The Media Equation:How People Treat Computers, Television, andNew Media Like Real People and Places, NewYork: Cambridge University Press.

Reinharz, S. (1992) Feminist Methods in SocialResearch, New York: Oxford University Press.

Remnick, D. (1998) ‘Bad News: How BroadcastJournalism Is Downsizing the World,’ The NewYorker, 74: 4–5.

Renn, O. (1991) ‘Risk Communication and theSocial Amplification of Risk,’ in R. Kasperson and P. Stallen (eds) Communicating Risks to thePublic: International Perspectives, Dordrecht:Kluwer Academic.

Rentz, J., Reynolds, F., and Stout, R. (1983)‘Analysing Changing Consumption Patterns with

Cohort Analysis,’ Journal of Marketing Research,20: 12–20.

Resnik, D. B. (1998) The Ethics of Science: AnIntroduction, London: Routledge.

Rheingold, H. (1994) The Virtual Community,London: Minerva.

Richards, J. and Sheridan, D. (eds) (1987) MassObservation at the Movies, London: Routledge &Kegan Paul.

Richardson, K. and Meinhof, U. H. (1999) Worldsin Common? Television Discourse in a ChangingEurope, London: Routledge.

Ricoeur, P. (1981) Hermeneutics and the HumanSciences: Essays on Language, Action and Inter-pretation, Cambridge: Cambridge UniversityPress.

Ricouer, P. (1983) Time and Narrative (Vol. 1),Chicago and London: University of Chicago Press.

Riley, J. W. and Riley, M. W. (1965) ‘Mass Com-munication and the Social System,’ in R. K.Merton, L. Broom, and L. S. Cottrell (eds)Sociology Today, New York: Harper & Row.

Rimmon-Kenan, S. (1983) Narrative Fiction: Con-temporary Poetics, London: Methuen.

Ritchie, D. (1999) ‘Probably, Probably Not:Rhetoric and Interpretation in CommunicationResearch’, Paper presented at the InternationalCommunication Association Conference, SanFrancisco, CA.

Roach, C. (1997) ‘The Western World and theNWICO: United They Stand?,’ in P. Golding andP. Harris (eds) Beyond Cultural Imperialism:Globalization, Communication, and the NewInternational Order, London: Sage.

Robertson, R. (1992) Globalization: Social Theoryand Global Culture, London: Sage.

Robertson, R. (1995) ‘Glocalization: Time–Spaceand Homogeneity–Heterogeneity,’ in M. Feather-stone, S. Lash, and R. Robertson (eds) GlobalModernities, London: Sage.

Robinson, A. (1995) The Story of Writing, London,Thames & Hudson.

Robinson, J. P. and Bachman, J. (1972) ‘TelevisionViewing Habits and Aggression,’ in G. A.Comstock and E. A. Rubinstein (eds) Televisionand Social Behaviour (Vol. 3), Television andAdolescent Aggressiveness, Washington, DC: U.S.Government Printing Office.

Robinson, J. P. and Converse, P. (1972) ‘The Impactof Television on Mass Media Usage,’ in A. Szalai(ed.) The Use of Time, The Hague, The Nether-lands: Mouton.

Robinson, J. P. and Levy, M. (1986) The MainSource, Beverly Hills, CA: Sage.

References318

Roe, K. and De Meyer, G. (2000) ‘Music Television:MTV-Europe,’ in J. Wieten, G. Murdock, and P.Dahlgren (eds) Television Across Europe: AComparative Introduction, London: Sage.

Rogers, E. M. (1962) The Diffusion of Innovations,Glencoe, IL: Free Press.

Rogers, E. M. (1986) Communication Technology,New York: Free Press.

Rogers, E. M. (1999) ‘Anatomy of Two Sub-disciplines of Communication Study,’ HumanCommunication Research, 25(4): 618–631.

Rogers, E. M. (2000) ‘Reflections on News EventDiffusion Research,’ Journalism and MassCommunication Quarterly, 77(3): 561–576.

Rogers, E. M. and Storey, J. D. (1987) ‘Com-munication Campaigns,’ in C. R. Berger and S. H. Chaffee (eds) Handbook of CommunicationScience, Newbury Park, CA: Sage.

Rorty, R. (ed.) (1967) The Linguistic Turn, Chicago,IL: University of Chicago Press.

Rorty, R. (1979) Philosophy and the Mirror ofNature, Princeton, NJ: Princeton University Press.

Rorty, R. (1991) Objectivity, Relativism, and Truth,Cambridge: Cambridge University Press.

Rorty, R. (1998) Truth and Progress, Cambridge:Cambridge University Press.

Rose, B. G. (1985) TV Genres: A Handbook andReference Guide, Westport, CT: Greenwood Press.

Rosengren, K. E. (1970) ‘International News: Intraand Extra Media Data,’ Acta Sociologica, 13(1):96–109.

Rosengren, K. E. (1974) ‘International News:Methods, Data and Theory,’ Journal of PeaceResearch, 11: 145–156.

Rosengren, K. E. (2000) Communication: AnIntroduction, London: Sage.

Rosengren, K. E. and Windahl, S. (1989) MediaMatter: TV Use in Childhood and Adolescence,Norwood, NJ: Ablex.

Rosengren, K. E., Arvidson, P., and Sturesson, D.(1978) ‘The Barsebäck Panic,’ in C. Winick (ed.)Deviance and Mass Media, Beverly Hills, CA:Sage.

Rosengren, K. E., Wenner, L., and Palmgreen, P.(eds) (1985) Media Gratifications Research: Cur-rent Perspectives, Beverly Hills, CA: Sage.

Ross, A. (1910) ‘The Suppression of ImportantNews,’ Atlantic Monthly, CV: 303–311.

Ross, S. S. (1998) ‘Journalists’ Use of On-lineTechnology and Sources,’ in D. L. Borden and K.Harvey (eds) The Electronic Grapevine: Rumor,Reputation, and Reporting in the New On-LineEnvironment, Mahwah, NJ: Lawrence Erlbaum.

Rowbottom, S. (1974) Hidden From History,Harmondsworth: Pelican Books.

Ruby, J. E. (1965) The Daily News ExchangeScheme of the European Broadcasting Union,unpublished Master’s thesis, Indiana University.

Rudy, W. (1984) The Universities of Europe 1100–1914: A History, London: Associated UniversityPresses.

Rush, M. (1999) New Media in Late Twentieth-Century Art, London: Thames & Hudson.

Rydin, I. (1996) Making Sense of TV Narratives:Children’s Readings of a Fairy Tale, Linköping,Sweden: Linköping University.

Ryle, G. (1971) Collected Papers (Vol. 2), London:Hutchinson.

Sabin, R. (1993) Adult Comics: An Introduction,London: Routledge.

Sacks, H., Schegloff, E. A., and Jefferson, G. (1974)‘A Simplest Systematics for the Organisation ofTurn-Taking in Conversation,’ Language, 50:696–735.

Sahlins, M. (1985) Islands of History, Chicago, IL:University of Chicago Press.

Saïd, E. (1978) Orientalism, New York: RandomHouse.

Sallach, D. (1974) ‘Class Domination and Ideo-logical Hegemony,’ in G. Tuchman (ed.) The TVEstablishment, Englewood Cliffs, NJ: Prentice-Hall.

Samuelson, P. A. (1954) ‘The Pure Theory of PublicExpenditure,’ The Review of Economics andStatistics, 36(4): 387–389.

Sarris, A. (1968) The American Cinema: Directorsand Directions, 1929–1968, New York: Dutton.

Saussure, F. de (1959) Course in General Linguistics,London: Peter Owen [orig. publ. 1916].

Sayer, A. (2000) Realism and Social Science,London: Sage.

Scannell, P. (1986) ‘“The Stuff of Radio”: Develop-ments in Radio Features and Documentariesbefore the War,’ in J. Corner (ed.) Documentaryand the Mass Media, London: Edward Arnold.

Scannell, P. (1988) ‘Radio Times. The TemporalArrangements of Broadcasting in the ModernWorld,’ in P. Drummond and R. Paterson (eds)Television and its Audiences, London: British FilmInstitute.

Scannell, P. (1990) ‘Public Service Broadcasting: The History of a Concept,’ in A. Goodwin and G. Whannel (eds) Understanding Television,London: Routledge.

Scannell, P. (ed.) (1991) Broadcast Talk, London:Sage.

Scannell, P. (1996a) Radio, Television, and ModernLife, Oxford: Blackwell.

References 319

1111234567891012341115678920123456789301234567894012345678950111

Scannell, P. (1996b) ‘Public Service Broadcasting:From National Culture to Multiculturalism 1923–1995,’ in M. Raboy (ed.) Public Broadcasting for the Twenty-First Century, London: JohnLibbey.

Scannell, P. (2000) ‘For-anyone-as-someone Struc-tures,’ Media Culture & Society, 22(1): 5–24.

Scannell, P. and Cardiff, D. (1991) A Social Historyof British Broadcasting: 1922–1939: Serving theNation (Vol. 1), Oxford: Blackwell.

Schatz, T. (1988) The Genius of the System, NewYork: Pantheon.

Schensul, J. J. and LeCompte, M. D. (eds) (1999)The Ethnographer’s Toolkit, Walnut Creek, CA:Alta Mira Press.

Schensul, J. J., LeCompte, M. D., Nasasti, B. K., andBorgatti, S. P. (1999) Enhanced EthnographicMethods: Audiovisual Techniques, FocusedGroup Interviews, and Elicitation Techniques(Vol. 3), Walnut Creek, CA: Alta Mira Press.

Schensul, S. L., Schensul, J. J., and LeCompte, M. D.(1999) Essential Ethnographic Methods: Obser-vations, Interviews, and Questionnaires (Vol. 2),Walnut Creek, CA: Alta Mira Press.

Scheufele, D. (1999) ‘Agenda-Setting, Priming, andFraming Revisited: Another Look at CognitiveEffects of Political Communication,’ Paper pre-sented at the International Communication Asso-ciation Conference, San Francisco, CA.

Schiller, D. (1981) Objectivity and the News,Philadelphia: University of Pennsylvania Press.

Schiller, H. I. (1969) Mass Communications andAmerican Empire, New York: Kelley.

Schiller, H. I. (1976) Communication and CulturalDomination, White Plains, NY: International Artsand Science Press.

Schlesinger, P. (1978) Putting ‘Reality’ Together:BBC News, London: Constable.

Schlesinger, P., Dobash, R. E., Dobash, R. P., andWeaver, C. (1992) Women Viewing Violence,London: British Film Institute.

Schmitt, R. (1972) ‘Phenomenology,’ in P. Edwards(ed.) The Encyclopedia of Philosophy (Vol. 6),New York: Macmillan [orig. publ. 1967].

Schramm, W. (1964) Mass Media and NationalDevelopment, Stanford, CA: Stanford UniversityPress.

Schramm, W. (1997) The Beginnings of Communi-cation Study in America: A Personal Memoir,Thousand Oaks, CA: Sage.

Schramm, W. and Lerner, D. (eds) (1976) Com-munication and Change: The Last Ten Years – andthe Next, Honolulu: University Press of Hawaii.

Schramm, W., Lyle, J., and Parker, E. (1961)Television in the Lives of Our Children, Palo Alto,CA: Stanford University Press.

Schrøder, K. C. (1997) ‘Cynicism and Ambiguity:British Corporate Responsibility Advertisementsand their Readers in the 1990s,’ in M. Nava, A.Blake, I. MacRury, and B. Richards (eds) Buy ThisBook: Studies in Advertising and Consumption,London: Routledge.

Schrøder, K. C. (1999) ‘The Best of Both Worlds?Media Audience Research Between Rival Para-digms,’ in P. Alasuutari (ed.) Rethinking the MediaAudience: The New Agenda, London: Sage.

Schudson, M. (1978) Discovering the News, NewYork: Basic Books.

Schudson, M. (1984) Advertising, the UneasyPersuasion, New York: Basic Books.

Schudson, M. (1991) ‘News Production Revisited,’in J. Curran and M. Gurevitch (eds) Mass Mediaand Society (2nd edn), London: Edward Arnold.

Schudson, M. (1997) ‘Why Conversation Is Not theSoul of Democracy,’ Critical Studies in MassCommunication, 14(4): 297–309.

Schulz, W. (1976) Die Konstruktion von Realität inNachrichtenmedien [The Construction of Realityin News Media], Freiburg and Munich: KarlAlber.

Schulz, W. (1997) ‘Political Communication Schol-arship in Germany,’ Political Communication, 14:113–145.

Schütz, A. (1956) The Phenomenology of the Social World, London: Heinemann [orig. publ.1932].

Schwartz, S. H. and Bilsky, W. (1990) ‘Toward aTheory of the Universal Content and Structure ofValues: Extensions and Cross-cultural Replica-tions,’ Journal of Personality and SocialPsychology, 58(5): 878–891.

Scott, J. (1991) ‘Women’s History,’ in P. Burke (ed.)New Perspectives on Historical Writing, Cam-bridge: Polity Press.

Scribner, S. and Cole, M. (1981) The Psychology ofLiteracy, Cambridge, MA: Harvard UniversityPress.

Searle, J. R. (1969) Speech Acts, London: CambridgeUniversity Press.

Sebeok, T. A. (ed.) (1986) Encyclopedic Dictionaryof Semiotics (2nd edn), Berlin: Mouton de Gruyter.

Sebeok, T. A. and Umiker-Sebeok, J. (1983) ‘“YouKnow My Method”: A Juxtaposition of CharlesS. Peirce and Sherlock Holmes,’ in U. Eco and T. A. Sebeok (eds) The Sign of Three, Bloom-ington: Indiana University Press.

References320

Seiter, E., Borchers, H., Kreutzner, G., and Warth, E.-M. (eds) (1989) Remote Control: Television,Audiences, and Cultural Power, London: Rout-ledge.

Sellitz, C., Jahoda, M., Deutsch, M., and Cook, S.(1976) Research Methods in Social Relations,New York: Holt, Rinehart & Winston.

Sennett, R. (1974) The Fall of Public Man, Cam-bridge: Cambridge University Press.

Sepstrup, P. (1989) ‘Research into International TVFlows,’ European Journal of Communication,4(4): 393–408.

Seymour-Ure, C. (1989) ‘Prime Ministers’ Reactionto Television: Britain, Australia, and Canada,’Media, Culture & Society, 11(3): 307–325.

Shanahan, J. and Morgan, M. (1999) Television andIts Viewers: Cultivation Theory and Research,Cambridge: Cambridge University Press.

Shannon, C. and Weaver, W. (1949) The Mathemati-cal Theory of Communication, Urbana, IL:University of Illinois Press.

Sharpe, J. (1991) ‘History from Below,’ in P. Burke(ed.) New Perspectives on Historical Writing,Cambridge: Polity Press.

Sharrock, W. and Anderson, B. (1986) The Ethno-methodologists, London: Tavistock.

Shimp, T. A. and Gresham, L. G. (1983) ‘An Infor-mation Processing Perspective on Recent Adver-tising Literature,’ Current Issues and Research inAdvertising, 5: 39–75.

Shoemaker, P.J. and Reese, S. D. (1991) Mediatingthe Message, New York: Longman.

Shoemaker, P. J. (1991) Gatekeeping, NewburyPark, CA: Sage.

Shore, B. (1998) ‘Models Theory as a Framework forMedia Studies,’ in B. Höijer and A. Werner (eds)Cultural Cognition: New Perspectives in AudienceTheory, Gothenburg, Sweden: Nordicom.

Shrum, L. J. (1996) ‘Psychological Processes Under-lying Cultivation Effects: Further Tests of Con-struct Accessibility,’ Human CommunicationResearch, 22(4): 482–509.

Siebert, F., Peterson, T., and Schramm, W. (1956)Four Theories of the Press, Urbana, IL: Universityof Illinois Press.

Siegel, P. M. and Hodge, R. W. (1968) ‘A CausalApproach to the Study of Measurement Error,’ inH. M. Blalock, Jr. and A. B. Blalock (eds)Methodology in Social Research, New York:McGraw-Hill.

Siegel, S. (1956) Nonparametric Statistics for theBehavioural Sciences, Tokyo: McGraw-Hill.

Sigman, S. J. and Fry, D. L. (1985) ‘DifferentialIdeology and Language Use: Readers’ Recon-structions and Descriptions of News Events,’

Critical Studies in Mass Communication 2:307–322.

Silj, A. (1988) East of Dallas: The European Chal-lenge to American Television, London: BritishFilm Institute.

Silverman, D. (2000) Doing Qualitative Research: APractical Handbook, London: Sage.

Silverstone, R. (1994) Television and Everyday Life,London: Routledge.

Silverstone, R. (1999) Why Study the Media?,London: Sage.

Simon, H. W. (ed.) (1989) Rhetoric in the HumanSciences, London: Sage.

Simpson, C. (1996) ‘Elisabeth Noelle-Neumann’s“Spiral of Silence” and the Historical Context of Communication Theory,’ Journal of Com-munication, 46(3): 149–173.

Sinclair, J., Jacka, E., and Cunningham, S. (1996)New Patterns in Global Television: PeripheralVision, New York: Oxford University Press.

Sinclair, U. (1920) The Brass Check: A Study ofAmerican Journalism, Long Beach, CA: Publishedby the author.

Slaatta, T. (1999) Europeanisation and the Norwe-gian News Media: Political Discourse and News Production in the Transnational Field,Department of Media and Communication,Report No. 36, Oslo, Norway: University of Oslo.

Slemon, S. (1995) ‘The Scramble for Post-Colonial-ism,’ in B. Ashcroft, G. Griffiths, and H. Tiffin(eds) The Post-Colonial Studies Reader, London:Routledge.

Slevin, J. (2000) The Internet and Society, Cam-bridge: Polity Press.

Smith, D. (1974) ‘Theorizing as Ideology,’ in R.Turner (ed.) Ethnomethodology, Baltimore, MD:Penguin.

Smulyan, S. (1994) Selling Radio: The Com-mercialization of American Broadcasting, 1920–1934, Washington, DC: Smithsonian InstitutePress.

Smythe, D. W. (1977) ‘Communications: Blindspotof Western Marxism,’ Canadian Journal ofPolitical and Social Theory, 1(3): 1–27.

Snow, C. P. (1964) The Two Cultures and a SecondLook, Cambridge: Cambridge University Press.

Sokal, A. and Bricmont, J. (1998) FashionableNonsense: Postmodern Intellectuals’ Abuse ofScience, New York: Picador.

Solso, R. L. (1994) Cognition and the Visual Arts,Cambridge, MA: MIT Press.

Sonesson, G. (1989) Pictorial Concepts: Inquiriesinto the Semiotic Heritage and its Relevance to theInterpretation of the Visual World, Lund, Sweden:Lund University Press.

References 321

1111234567891012341115678920123456789301234567894012345678950111

Sparks, C. (1997) ‘Post-Communist Media inTransition,’ in J. Corner, P. Schlesinger, and R.Silverstone (eds) International Media Research: ACritical Survey, London: Routledge.

Sparks, C. and Tulloch, J. (2000) Tabloid Tales:Global Debates over Media Standards, Boulder,CO: Rowman & Littlefield.

Sparks, G. G. and Cantor, J. (1986) ‘DevelopmentalDifferences in Fright Responses to a TelevisionProgramme Depicting a Character Transforma-tion,’ Journal of Broadcasting and ElectronicMedia, 30: 309–323.

Spigel, L. (1992) Make Room for TV: Television andthe Family Ideal in Postwar America, Chicago, IL:University of Chicago Press.

Spigel, L. and Jenkins, H. (1990) ‘Same Bat Channel,Different Bat Times: Mass Culture and PopularMemory,’ in R. Pearson and W. Uricchio (eds) TheMany Lives of the Batman: Critical Approaches toa Superhero and His Media, London: British FilmInstitute.

Spivak, G. C. (1988) ‘Can the Subaltern Speak?,’ inC. Nelson and L. Grossberg (eds) Marxism and the Interpretation of Culture, London: Mac-millan.

Spradley, J. P. (1979) The Ethnographic Interview,Fort Worth, TX: Harcourt Brace Jovanovich.

Sreberny, A. (2000) ‘The Global and the Local inInternational Communications,’ in J. Curran andM. Gurevitch (eds) Mass Media and Society (3rdedn), London: Edward Arnold.

Sreberny, A. and Stevenson, R. L. (eds) (in prepara-tion) ‘Rethinking International News: Theory andResearch at the end of the Millennium,’ Creskill,NJ: Hampton Press.

Sreberny, A., Nordenstreng, K., Stevenson, R., andUgboajah, F. (eds) (1985) Foreign News in theMedia: International Reporting in 29 Countries,Paris: UNESCO.

Stacey, J. (1994) Star Gazing: Hollywood Cinemaand Female Spectatorship, London: Routledge.

Stam, R., Burgoyne, R., and Flitterman-Lewis, S.(1992) New Vocabularies in Film Semiotics:Structuralism, Post-structuralism and Beyond,London: Routledge.

Standage, T. (1998) The Victorian Internet, London:Weidenfeld & Nicolson.

Stempel, G. H. (1952) ‘Sample Size for ClassifyingSubject Matter in Dailies,’ Journalism Quarterly,29: 333–334.

Stevenson, R. L. and Shaw, D. L. (eds) (1984)Foreign News and the New World InformationOrder, Ames, IA: Iowa State University.

Stipp, H. and Milavsky, R. (1988) ‘US TelevisionProgramming Effects on Aggressive Behaviour of

Children and Adolescents,’ Current Psychology:Research and Review, 7: 76–92.

Stokes, M. and Maltby, R. (eds) (1999a) AmericanMovie Audiences: From the Turn of the Centuryto the Early Sound Era, London: British FilmInstitute.

Stokes, M. and Maltby, R. (eds) (1999b) IdentifyingHollywood’s Audiences: Cultural Identity and theMovies, London: British Film Institute.

Stone, A. R. (1991) ‘Will the Real Body Please StandUp? Boundary Stories about Virtual Cultures,’ inM. Benedikt (ed.) Cyberspace: First Steps, Lon-don: MIT Press.

Stone, L. (1979) The Family, Sex and Marriage,1500–1800, Harmondsworth: Penguin.

Stone, L. (1987) The Past and the Present Revisited,London and New York: Routledge & Kegan Paul.

Strauss, A. L. (1987) Qualitative Analysis for SocialScientists, Cambridge: Cambridge UniversityPress.

Strauss, A. L. and Corbin, J. (1990) Basics ofQualitative Research, Newbury Park, CA: Sage.

Street, A. T. (1909) ‘The Truth about Newspapers,’Chicago Tribune, 25 July.

Striphas, T. (ed.) (1998) The Institutionalization ofCultural Studies, London: Routledge.

Sumpter, R. S. (2000) ‘Daily Newspaper Editors’Audience Construction Routines: A Case Study,’Critical Studies in Mass Communication, 17(3):334–346.

Surgeon-General’s Scientific Advisory Committee onTelevision and Social Behavior (1972) Volumes I-VI, Washington, DC: U.S. Government PrintingOffice.

Swales, J. M. and Rogers, P. (1995) ‘Discourse andthe Projection of Corporate Culture: The MissionStatement,’ Discourse and Society, 6(2): 223–242.

Tamborini, R., Zillmann, D., and Bryant, J. (1984)‘Fear and Victimisation: Exposure to Televisionand Perceptions of Crime and Fear,’ in R. N.Bostrum (ed.) Communication Yearbook (Vol. 8),Beverly Hills, CA: Sage.

Tannenbaum, P. H. (1954) ‘Effect of Serial Positionon Recall of Radio News Stories,’ JournalismQuarterly, 31: 319–323.

Tannenbaum, P. H. and Kemick, J. (1954) ‘Effects ofNewscast Item Leads upon Listener Inter-pretation,’ Journalism Quarterly, 31: 33–37.

Tashakkori, A. and Teddlie, C. (1998) MixedMethodology: Combining Qualitative and Quan-titative Approaches, Thousand Oaks, CA: Sage.

Thomas, W. I. and Znaniecki, F. (1927) The PolishPeasant in Europe and America, New York:Knopf.

References322

Thompson, E. P. (1968) The Making of the EnglishWorking Class, London: Gollancz [orig. publ.1963].

Thompson, J. B. (1995) The Media and Modernity,Cambridge: Polity Press.

Thompson, K. (1998) Moral Panics, London:Routledge.

Thompson, M. (1999) Forging: War. The Media inSerbia, Bosnia and Hercegovina (2nd edn), Luton,Beds: University of Luton Press.

Thorson, E. (1990) ‘Consumer Processing of Adver-tising,’ Current Issues and Research in Advertis-ing, 12: 197–230.

Thorson, E. (1994) ‘Using Eyes on Screen as aMeasure of Attention to Television,’ in A. Lang(ed.) Measuring Psychological Response to Media,Hillsdale, NJ: Lawrence Erlbaum.

Thrasher, F. M. (1927) The Gang, Chicago, IL:University of Chicago Press.

Tichenor, P., Olien, C., and Donohue, G. (1970)‘Mass Media Flow and Differential Growth inKnowledge,’ Public Opinion Quarterly, 34(2):159–170.

Todorov, T. (1971) ‘The Two Principles ofNarrative,’ Diacritics, 1: 37–44.

Todorov, T. (1990) Genres in Discourse, Cambridge:Cambridge University Press [orig. publ. 1978].

Tomlinson, J. (1999) Globalization and Culture,Cambridge: Polity Press.

Tönnies, F. (1974) Community and Association,London: Routledge & Kegan Paul [orig. publ.1887].

Toulmin, S. (1958) The Uses of Argument,Cambridge: Cambridge University Press.

Trenaman, J. S. M. (1967) Communication andComprehension, London: Longman.

Trenaman, J. S. M. and McQuail, D. (1961) Tele-vision and the Political Image, London: Methuen.

Trew, T. (1979) ‘Theory and Ideology at Work,’ inR. Fowler, Hodge, B., Kress, G. and Trew, T. (eds)Language and Control, London: Routledge &Kegan Paul.

Tuchman, G. (1978) Making News: A Study in theConstruction of Reality, New York: Free Press.

Tuchman, G., Daniels, A. K., and Benet, J. (eds)(1978) Hearth and Home: Images Of Women inMass Media, New York: Oxford University Press.

Tulloch, J. (1989) ‘Approaching the Audience: TheElderly,’ in E. Seiter, H. Borchers, G. Kreutzner,and E.-M. Warth (eds) Remote Control: Tele-vision, Audiences, and Cultural Power, London:Routledge.

Tunstall, J. (1977) The Media are American: Anglo-American Media in the World, London:Constable.

Tunstall, J. and Machin, D. (1999) The Anglo–American Media Connection, Oxford: OxfordUniversity Press.

Turkistani, A. S. (1988) News Exchange via Arabsatand News Values of Arab Television News People,Indiana University, unpublished Ph.D. thesis.

Turkle, S. (1995) Life on the Screen: Identity in theAge of the Internet, New York: Simon & Schuster.

Umiker-Sebeok, J. (ed.) (1987) Marketing Signs,Berlin: Mouton de Gruyter.

Underwood, D. (1993) When MBAs Rule the News-room, New York: Columbia University Press.

UNESCO (1953) The Flow of News, Paris:UNESCO.

UNESCO (1985) Foreign News in the Media:International Reporting in 29 Countries, Reportsand Papers on Mass Communication No. 93,Paris: UNESCO.

UNESCO (1989) World Communication Report,Paris: UNESCO.

Urry, J. (2000) Sociology Beyond Societies: Mobili-ties for the Twenty-First Century, London:Routledge.

van Dijk, J. (1999) The Network Society, London:Sage.

van Dijk, T. (ed.) (1997) Discourse Studies: AMultidisciplinary Introduction (Vols 1–2), Lon-don: Sage.

van Dijk, T. A. and Kintsch, W. (1983) Strategies ofDiscourse Comprehension, New York: AcademicPress.

van Maanen, J. (1988) Tales of the Field, Chicago:University of Chicago Press.

van Zoonen, L. (1992) Feminist Media Studies,London: Sage.

van Zoonen, L., and Wieten, J. (1994) ‘“It Wasn’tExactly a Miracle”: The Arrival of Television inDutch Family Life,’ Media Culture & Society,16(4): 641–660.

Varis, T. and Jokelin, R. (1976) Television News inEurope: A Survey of the News-Film Flow inEurope, Tampere, Finland: Institute of Journalismand Mass Communication.

Venturi, R., Brown, D. S., and Izenour, S. (1972)Learning from Las Vegas, Cambridge, MA: MITPress.

Vestergaard, T. and Schrøder, K. (1985) The Lan-guage of Advertising, Oxford: Basil Blackwell.

Vincent, D. (2000) The Rise of Mass Literacy:Reading and Writing in Modern Europe, Cam-bridge: Polity Press.

Viswanath, K. and Finnegan, J. R. (1996) ‘TheKnowledge Gap Hypothesis 25 Years Later,’Communication Yearbook, 19: 187–227.

References 323

1111234567891012341115678920123456789301234567894012345678950111

Volkmer, I. (1999) CNN – News in the GlobalSphere: A Study of CNN and Its Impact on GlobalCommunication, Luton, Beds: University of LutonPress.

Volosinov, V. N. (1973) Marxism and the Phil-osophy of Language, New York: Seminar Press[orig. publ. 1929].

Volosinov, V. N. (1985) ‘Verbal Interaction,’ in H. A.Innis (ed.) Semiotics: An Introductory Anthology,Bloomington: Indiana University Press.

Walker, J. A. (1994) Art in the Age of Mass Media,London: Pluto Press.

Wallerstein, I. (1974) The Modern World-System(Vol. 1), New York: Academic Press.

Wallerstein, I. (1980) The Modern World-System(Vol. 2), London: Academic Press.

Wallerstein, I. (1989) The Modern World-System(Vol. 3), London: Academic Press.

Watson, J. B. (1924) Psychology from the Stand-point of a Behaviorist, Philadelphia, PA: J. B.Lippincott.

Watson, J. B. (1930) Behaviorism, Chicago, IL:University of Chicago Press.

Watson, J. and Hill, A. (1999) Dictionary of Mediaand Communication Studies (5th edn), London:Edward Arnold.

Weaver, D. and Wilhoit, G. C. (1996) The AmericanJournalist in the 1990s, Mahwah, NJ: LawrenceErlbaum.

Webb, E. J., Campbell, D. T., Schwartz, R. D., andSechrest, L. (2000) Unobtrusive Measures (revisededn), Thousand Oaks, CA: Sage.

Weber, M. (1958a) ‘Politics as a Vocation,’ in H. H.Gerth and C. W. Mills (eds) From Max Weber,New York: Oxford University Press [orig. publ.1918].

Weber, M. (1958b) The Protestant Ethic and theSpirit of Capitalism, New York: Charles Scribner’sSons [orig. publ. 1930].

Weber, M. (1964) The Theory of Social and Eco-nomic Organization, New York: The Free Press.

Webster, D. (1988) Looka Yonder! The ImaginaryAmerica of Populist Culture, London: Routledge.

Webster, J. G. and Phalen, P. F. (1997) The MassAudience: Rediscovering the Dominant Model,Mahwah, NJ: Lawrence Erlbaum.

Weitzman, E. (2000) ‘Software and QualitativeResearch,’ in N. Denzin and Y. S. Lincoln (eds)Handbook of Qualitative Research (2nd edn),Thousand Oaks, CA: Sage.

Weitzman, E. and Miles, M. B. (1995) ComputerPrograms for Qualitative Data Analysis,Thousand Oaks, CA: Sage.

Wetherell, M. and Potter, J. (1992) Mapping theLanguage of Racism: Discourse and the Legiti-

mation of Exploitation, London: HarvesterWheatsheaf.

Wetherell, M., Taylor, S., and Yates, S. (eds) (2001)Discourse Theory and Practice: A Reader, Lon-don: Sage.

Whannel, G. (1992) Fields in Vision: Television Sportand Cultural Transformation, London: Routledge.

White, D. M. (1950) ‘The “Gate Keeper”: A CaseStudy in the Selection of News,’ JournalismQuarterly, 27: 383–390.

Wieten, J., Murdock, G., and Dahlgren, P. (eds)(2000) Television Across Europe: A ComparativeIntroduction, London: Sage.

Wilcox, D. F. (1900) ‘The American Newspaper: A Study in Social Psychology’, Annals of theAmerican Academy of Political and Social Science,16: 56–92.

Wilke, J. (1984) Nachrichtenauswahl und Medien-realität in vier Jahrhunderten [News Selection andMedia Reality in Four Centuries], Berlin: Walterde Gruyter.

Williams, R. (1962) Communications, Harmonds-worth: Penguin.

Williams, R. (1974) Television: Technology andCultural Form, London: Fontana.

Williams, R. (1975) Culture and Society 1780–1950,Harmondsworth: Penguin [orig. publ. 1958].

Williams, R. (1977) Marxism and Literature,London: Oxford University Press.

Williams, R. (1983) Keywords, London: Fontana.Williams, T. M. (ed.) (1986) The Impact of

Television: A Natural Experiment in ThreeCommunities, New York: Academic Press.

Williamson, J. (1978) Decoding Advertisements.Ideology and Meaning in Advertising, London:Marion Boyars.

Wimmer, R. D. and Dominick, J. R. (1994) MassMedia Research: An Introduction (4th edn),Belmont, CA: Wadsworth.

Wimsatt, W. (ed.) (1954) The Verbal Icon, London:Methuen.

Wimsatt, W. and Brooks, C. (1957) Literary Criti-cism: A Short History, New York: Alfred A.Knopf.

Winch, P. (1963) The Idea of a Social Science,London: Routledge.

Windahl, S., Signitzer, B., and Olson, J. (1992) UsingCommunication Theory, Newbury Park, CA:Sage.

Winston, B. (1998) Media, Technology, and Society– A History: From the Telegraph to the Internet,London: Routledge.

Winthrop-Young, W. and Wutz, M. (1999) ‘Trans-lators’ Introduction,’ in F. A. Kittler Gramophone,Film, Typewriter, Stanford, CA: Stanford Uni-versity Press.

References324

Witmer, D. F., Colman, R. W., and Katzman, S. L.(1999) ‘From Paper-and-Pencil to Screen-and-Keyboard: Toward a Methodology for SurveyResearch on the Internet’ in S. G. Jones (ed.)Doing Internet Research: Critical Issues andMethods for Examining the Net, Thousand Oaks,CA: Sage.

Wittgenstein, L. (1953) Philosophical Investigations,London: Macmillan.

Wittgenstein, L. (1972) [1921] Tractatus Logico-Philosophicus, London: Routledge & Kegan Paul.

Wober, J. M. (1981) ‘Psychology in the Future ofBroadcasting Research,’ Bulletin of the BritishPsychological Society, 34: 409–412.

Wober, M. and Gunter, B. (1988) Television andSocial Control, Aldershot, Hants: Avebury.

Wolfe, T. (ed.) (1973) The New Journalism, NewYork: Harper & Row.

Wood, R. (1965) Hitchcock’s Films, London:Zwemmer.

Wood, R. (1968) Howard Hawks, London: Secker& Warburg.

Wright, C. R. (1959) Mass Communication: A Socio-logical Perspective, New York: Random House.

Wu, W., Weaver, D., Owen, D., and Johnstone, J. W.L. (1996) ‘Professional Rules of Russian and USJournalists: A Comparative Study,’ Journalismand Mass Communication Quarterly, 73(3): 534–548.

Yin, R. K. (1994) Case Study Research: Design andMethods (2nd edn), Thousand Oaks, CA: Sage.

Zillmann, D. and Bryant, J. (1994) ‘Entertainmentas Media Effect,’ in D. Zillmann and J. Bryant(eds) Media Effects: Advances in Theory andResearch, Hillsdale, NJ: Lawrence Erlbaum.

Zillmann, D. and Vorderer, P. (eds) (2000) MediaEntertainment: The Psychology of its Appeal,Mahwah, NJ: Lawrence Erlbaum.

Znaniecki, F. (1934) The Method of Sociology, NewYork: Farrar and Rinehart.

Zuckerman, M. (1994) Behavioural Expressionsand Biosocial Bases of Sensation Seeking, Cam-bridge: Cambridge University Press.

References 325

1111234567891012341115678920123456789301234567894012345678950111

abduction, 259–60, 263–6abstracted empiricism, 56abstracts, 3accountability, 147actants, 127–8, 251action research, 176, 287activists, 41–4adaptations, 126address, 85, 106, 131, 134advertising, 41, 112age, 163agency, human, 1–3agenda-setting, 87, 103, 145–7, 163, 220,

262–3agents, 65–6, 79, 80analysts, 41–4anchorage, 28, 111, 135Annales School, 193anonymity, 290anthropology, 48, 179Any Day Now, 67–72apartheid (in research), 254applied research, 278–9archives, 63, 201, 203–4, 283, 285Aristotle, 16, 19art history, 27, 120artefacts, 243–5article formats (newspapers), 83articulation, 189association relations, 214attributes, 210audience/s, 60, 138–70, 182; active – , 51–4,

160–4; media-audience-society nexus, 169–70;overhearing – , 110; as product, 72, 143; andtext, 167–9

audiovisual media, 145audit trail, 246auteur theory, 65author, implied, 129

authoritarian theory, 275autonomy, 62, 89

Bakhtin, M. M., 186Barthes, Roland, 25, 50, 111, 115, 135basic research, 278–9BBC, 201behaviourism, 51Bellour, Raymond, 123Berelson, B., 103The Big Sleep, 122, 124–5, 133, 134, 244biographies, 202biological esssentialism, 36Birmingham Centre for Contemporary Cultural

Studies, 39, 176books, 288Bordwell, David, 119, 126, 128, 129BP advertisement, 112Branigan, Edward, 130–1Braudel, Fernand, 193Briggs, Asa, 200–1British press, 199broadcasting histories, 200–4budgets, 67bureaucratization, 174

camera, 132camera narration, 132campaigns, 147–8canonical story, 128Cantril, Hadley, 52capitalization, 174Carey, James, 7–8, 9case studies, 67–72, 239case-based research, 67–72, 164, 239, 247, 256–7,

264categorical imperative, 289categories of analysis, 103–4causality (experimental research), 222–7

Index

census, 214central tendency, 230character narration, 127–8, 131Chatman, S., 129chi-square, 233Chicago School of Sociology, 43, 176children, 228, 245chronological table, media and communication,

18, 199class, 163closed texts, 167CNN International, 96code-based research, 247, 256–7, 264codes, 8, 120, 131coding, 100–4, 162–3, 220–2, 229–34, 245,

247–8, 251coding frames, 220cognitivism, 33, 36–8, 116, 122cohort studies, 218–19Cold War, 95, 178comics, 115commercial media, 63, 89commodities, culture as, 174communication, 171–3; definitions, 172;

development, 179; the end of, 292–3; history,194–8; levels of, 8, 9; models, 7–10; modes,185; societies as systems of, 185

communication research, 1–11, 13, 273–93competition, 82complementarity of methodologies, 254–72computer media studies, 244, 252, 288computer-mediated cultures, 4, 182–90concepts, 210conceptualization, 237confidentiality, 290connotation, 25, 103constant comparative method, 247constructionism, 42constructs, 210consumerism, 47consumption, 143–4contextual action, 160–2, 163–4, 184contextualization, 171–2control groups, 223control society, 275convergence, 1–11, 13, 183, 271–2conversation analysis, 108–10copyright, 290corporate control, 46corporate records, 66covering-law model, 264creativity, 66Cressey, Paul, 51–2crew, 70critical discourse analysis, 106–8

critical linguistics, 104–6critical political economy, 56cultivation research, 100, 119, 151, 229, 292cultivation theory, 150–1cultural capital, 154cultural contexts, 17, 40–57, 171–90cultural forum, 155cultural imperialism, 93, 177–8cultural studies, 39, 162, 286culturalism, 39culture, 4–6, 172–3culture wars, 292Curran, James, 199cyberspace, 189cyborgs, 189

dailiness, 199data analysis, 218, 229–34, 245–53, 258data collection methods, 183, 257, 258data display, 247data sharing, 252decodings, 162–3deconstructionism, 33–4, 287dedicated monogeneric news channels, 96deduction, 259–60, 262–3defamiliarisation, 29democracy, 44–6democratic participant theory, 275denotation, 25, 103descriptive discovery, 116, 261determinism, 16, 196–7development communication, 140, 179, 277, 287diachronic philology, 30–1diachronic studies, 218diaspora, 180diegesis, 130diffusion of innovations, 140–2digital culture, 17digital divide, 149directors, 70discipline see field-discipline debatediscourse analysis, 31; BP advertisement, 112–16;

critical, 106–8; and data analysis, 248–51;qualitative, 104–10, 119–20

discourse, 98–116, 126, 166–9, 182; see alsotexts

discovery, context of, 258–9, 291discursive consciousness, 2discursive psychology, 108–10disembedding, 2, 182distinction (position in society), 27distribution, 231, 232distribution technologies, 73documents, 243–5dominant paradigm, 138

Index 327

1111234567891012341115678920123456789301234567894012345678950111

double articulation, 188double hermeneutic, 9, 22double-blind procedures, 224, 290double-insight procedures, 290drama, 132

economic capital, 154education for media, 285effects research, 138–55, 214–19electronic culture, 17emergence, 269emic analysis, 236, 264empirical research, 257–8encyclopedias, 3epic, 132epistemology, 36, 255, 256, 258, 269epoché, 23essentialism, 36ethics, 289ethnicity, 75, 163ethnography, 72, 164–6, 179, 242ethnomethodology, 56, 81, 108etic analysis, 236, 264evaluation research, 94, 102, 284everyday theory, 275, 287evidential paradigm, 264evolution, 36executive producers, 75; assistants to, 71experience sampling method, 144experimental contexts, 225–7experimental designs, 224–5experimental groups, 223experiments, 154, 222–9exposure, 143eye-tracking, 143

fabula, 126facilitation, 272factorial designs, 225factual narratives, 131Fairclough, Norman, 106–8falsification, 262family resemblances, 133feminism, 34–6, 35, 67, 286–7fictional narratives, 62–77, 96, 117–37, 131field notes, 243field research, 67, 243field-discipline debate, 9, 13, 279film production, 76film studies, 31–3, 37, 199films, 64, 119, 133, 288fine arts, 27Fiske, J., 186–7flows: bilateral, 175; in and between countries,

173

flow, as level of analysis, 168, 169, 187;multilateral, 175; national, 175; TV news –studies, 92–3, 178

focus groups, 241formalist tradition, 32framing, 37, 86, 149–50, 163Frankfurt School, 282freedom, 278frequency distribution, 230

Galtung, Johan, 93–5Garfinkel, Harold, 56gatekeeping research, 79, 91the gaze, 33Gemeinschaft, 176gender, 34–5, 132, 163genres, 132–4, 142–3, 168, 201Gerbner, George, 119German historicism, 192Gesellschaft, 176Giddens, A., 1–3, 4, 9Gitlin, T., 86–7globalization, 95–7, 181, 196glocalization, 182grammar, 250gratification, 142–3, 144great chain of being, 21, 276The Green Magazine (BP advertisement), 112,

244Greimas, A. J., 127grounded theory, 247–8, 251, 261Gulf War, 95

Habermas, J., 6, 280, 293Hall, Stuart, 39handbooks, 3hermeneutic circle, 21–2hermeneutics, 21–2, 104, 203; double/triple

hermeneutic, 9, 273–4; of suspicion, 22, 106,162, 240

historical research, 203–4historical-biographical approaches (literary

criticism), 29historiography, 191history, 191–4Hollywood, 64, 119homology, 177horizon, 22, 23, 167horizontal integration, 73house style, 65humanities, 13, 15–39Husserl, Edmund, 22–3hypermedia, 187hypertexts, 168, 169, 187, 188–9hypothetical interpretation, 116

Index328

hypothetico-deductive research, 211–12, 231–4,262, 266

icon, 111ideas, history of, 19–26ideology critique, 34, 178images, 110–16, 121, 121–3imagined communities, 174, 189imperialism, (in research), 254impersonal grammar, 250implicature, 105implied author, 129implied reader, 129index, 111induction, 259–62industrial contexts, 64inference, 259–66informants, 238information society, 188informed consent, 290Innis, Harold, 46, 47, 195innovations, diffusion of, 140–2, 175, 193institutional review boards, 291institutionalization, 152–5, 163institutions-to-think-with, 6integration, 73, 181intellectual cultures, 278–9intelligentsia, 285interaction, 251interactivity, 184intercultural communication, 179, 277interest groups, 64intermediality, 188–90interpersonal communication, 8, 140, 279interpretive communities, 167, 264interpretive repertoires, 109, 167intersubjectivity, 16, 267intertextuality, 186–8; and campaign research,

148; contribution of structuralism, 30; andcritical discourse analysis, 107; and integratedproduction, 77; and media environments,168–9; and qualitative research, 244; self-referentiality, 99

interviewing, 215, 240–2, 249–50inverted pyramid, 84invisible narration, 132iterative abduction, 266iterative process, 236

joint operating agreements, 83journalism, 84, 89, 200, 285journals, 3, 205justification, context of, 258–9, 291

keywords in context, 248

knowledge, 15, 212, 274knowledge gaps, 148–9knowledge interests, 87, 100, 279–81Krippendorff, K., 103

laboratory research, 225–7Lang, Kurt 238Lang, Gladys, 238Language: emic/etic, 236, 264; generative model,

29–30, 188; of the news, 88, 98–116; role inqualitative methodologies, 241

langue, 24, 120Lazarsfeld, Paul, F. 43–4, 56, 214, 281–2leitmotifs, 136lexical processes, 105libertarian theory, 275lifestyles, 152, 153Lifetime Television, 73–4lifeworld, 22linguistic turn, 20, 26, 261linguistics, 30–1, 104–6, 250literacy, 16–18, 193, 195–6 literary criticism, 29–30literature, 27living conditions, 153localities, 175–7log files, 183longitudinal research, 218longue durée, 193Lynd, Helen, 48–9Lynd, Robert, 48–9

macrocosms, theoretical, 237, 271McLuhan, Marshall, 16, 18, 196–7McQuail, D., 139mass communication, 8, 279material culture, 245mean score, 230meaning, 30, 119–20, 188, 236, 262measurement, 213–14, 220–1, 261media, 1–4; as cultural forum, 146; and

democracy, 44–6; history, 17, 63, 194, 198–205;institutionalization by, 152–5; and publicculture, 49–51; public perception, 142–3; asrepresentation or resource, 163; researchorganizations, 283; secondary use, 143;socialization by, 151–2; three degrees, 3–4

media effects, 138–55, 214–19, 222–7 media events, 155 media literacy, 285 media production, 62–97, 182, 245; fiction,

62–77; film v. television, 76; and globalization,182; individual artefacts, 65; international news,91–7; macro- and micro-theories, 78, 92; news(newspapers, TV) 78–90

Index 329

1111234567891012341115678920123456789301234567894012345678950111

media research: history, 280; institutions, 88, 198,278–84; milestone studies, 156–60; socialorigins, 40–8, 156, 273–93; unification, 207;uses, 273–93

median score, 230mediation, 86medium theory, 16–18, 155, 197melodrama, 163member checks, 252Merton, Robert, K. 53metamedia, 184metaphor, 24, 250Methodenstreit, 256methodologies, 10–11, 146, 183, 230, 240, 244,

254–72 metonymy, 24, 99, 111metropolitanism, 83Metz, Christian, 32, 47, 121–2microcosms, empirical, 237, 271middle range theories, 18, 55, 79, 96milestone studies, 156–60, 275Mills, C. Wright, 55–6mise-en-scène, 134modality, 105mode, 230models of communication, 7–10modern period, 2–3, 40–57, 173–7, 181montage, 32, 123morality, 289Morley, D., 162–3moving images, 121–3multi-step sampling, 238multiple personae, 183multiple regression, 232–3, 234museums, 27, 283, 285music, 27, 136, 177

narratee, 129narrative analysis, 29, 123–32, 264narrative history, 192national research traditions, 39, 190National Television Violence Study, 222nations, 173–5native’s perspective, 236natural evolution, 36naturalistic contexts, 226–7, 236network society, 188networks, 64–5, 74, 84New Criticism, 29, 120New World Information and Communication

Order, 93, 178news, language of, 88, 98–116news media, 78–97, 98–116, 288; see also pressnewspapers, 83–4, 199–200, 288normative theories, 89, 155, 276–8

note-taking, 71, 243null hypothesis, 232, 262, 270

objectivity, 87, 192, 200objects of analysis, 257, 258observation, 242–3ontologies, 179, 268open texts, 167operational theory, 274operationalization, 237oral culture, 17, 194–8, 241organizational communication, 8, 202, 284orientalism, 180the other, 180

panel studies, 219paradigms, 99, 120; dominant –, 138;

evidential –, 264; (in structuralist analysis) 24, 99, 120; (of science) 39, 138, 255–6, 264, 281

parallel content analysis, 146parole, 24, 120Parsons, Talcott, 55Payne Fund studies, 152, 158, 246peer review, 292Peirce, Charles Sanders, 23–4, 114, 115personal reference, 106phenomenology, 22–3, 35, 81philology, 30–1pilot season, 71poetry, 132point-of-hearing, 135point-of-view shot, 131policy, 284political correctness, 286political economy, 56, 63, 72, 81, 143, 178politics, 242, 284–9polysemy, 104, 167positivism, 22, 27, 209, 260–1postcolonial theory, 179–80, 205postmodernism, 33–4, 287postmodernity, 33poststructuralism, 32, 33, 180, 287Potter, J., 108–9practical consciousness, 2pragmatics, 31, 104–10, 134press cuttings, 202press histories, 199–200priming, 146print, 17, 145prior knowledge, 145proactive research, 282probability, 232, 233, 238, 270product development, 87product differentiation, 84

Index330

production companies, 64–5production documents, 244programme recordings, 202pronouns, 99, 106, 250propaganda studies, 46, 103, 147, 159, 174psychology, 36–8, 51, 108–10psychosemiotics, 33public perceptions, 142–3public relations, 41public service media, 63, 89, 204, 277, 283public settings, 161public sphere, 6, 174, 276, 285publication formats, 20, 259, 274

qualitative discourse analysis, 104–10, 119–20,249–51

qualitative research, 202, 235–53, 259, 266quantitative content analysis, 100–4, 118–19,

219–22; categories, 103–4; Hollywood cinema,119; measurement, 220–1; mediarepresentations studies, 79, 93, 118–19, 219;Review of the Year 1998, 100–3

quantitative research, 209–34, 266questionnaires, 215, 216

radio, 200–1, 288Radway, J., 164–5, 264–5random assignment, 223range, 231ratings, 73re-embedding, 2, 182reactive research, 282reader, implied, 129reading (the act of), 164realism, 268–70recall studies, 37, 144–5, 168, 211reception studies, 37, 129, 156–70reception aesthetics, 167; documents relating to,

244–5; framing, 150; main areas of analysis,163; two varieties of, 166

reference works, 3, 288reflexivity, 2, 172relay, 28, 111, 135reliability, 212–13, 251, 263, 266–8repeated measures design, 225reporters, 41–4, 79–82research community, 290–2research design, 214–27, 237–40research ethics, 69, 183–4, 227, 242, 289–92research subjects, 289respondents, 238revenues, 73reverse copyright, 290Review of the Year 1998 (BBC1), 101–2rhetoric, 19–20, 237, 287

ritual model, 8, 20, 162Ruge, Mari Holmboe, 93–5Russian formalism, 29, 126

sampling, 101, 215–17, 221–2, 238–40, 271; – cases, 238–40; media content – , 221–2; non-probability – , 216; probability – , 217

Saussure, Ferdinand de, 23, 24, 25, 120schedules, 202Schudson, Michael, 199–200science wars, 292scribal culture, 17secularization, 21, 38, 174self-referentiality, 99self-report information, 214semantic networks, 251semantics, 134, 251semiology, 25semiosis, 24semiotics, 23–4, 26, 32, 119–20, 167, 186sensationalism, 49sequences of events (in narrative), 125–6sets (in media production), 70shot-to-shot analyses, 123significant other, 256signified (signifié), 25signifier (signifiant), 25, 120–3snowball sampling, 239social responsibility theory, 275social sciences, 13, 40–57, 279socialization, 88, 151–2, 163sociodemographic categories, 167sociology, 55, 79–80Solomon four-group design, 225sound, 135–7sources, 89space, 182, 195speech, 4, 18, 194speech-act theory, 38, 105, 250sports, 182standard deviation, 231standpoint epistemology, 286state commissions, 284statistics, 66, 101, 230–4stimulus/response, 51, 145structural functionalism, 55, 78structuralism, 29, 30, 39, 193structuration theory, 1student projects, 291studio system (films), 65subcultures, 175–7suburbs, 83super-themes, 150survey research, 214–22, 227–9symbol, 111

Index 331

1111234567891012341115678920123456789301234567894012345678950111

symbolic diffusion, 141synchronic linguistics, 30–1syntactic transformations, 105syntagms, 24, 32, 99, 120, 121–3syntax, 134syuzhet, 126

tape-recording, 71, 243taste, 154technological determinism, 16, 196–7telephone interviews, 215television: house style, 65; news, 84–5; reference

works, 288; sound, 136–7; structural andrelational uses of, 161

textbooks, 3texts: and audience, 167–9; as general category,

28, 117; as heterogeneous constructions, 134–7; primary, secondary and tertiary, 186

textual blanks, 167thematic coding, 247, 251theory, 133, 179, 258, 273–8thick description, 51, 242Third World, 178time, 182, 192–8, 217time-in culture, 5time-out culture, 5time-space distanciation, 2, 182totalitarian theory, 275trade press, 67transcripts, 202transmission model, 7–8, 79

trend studies, 153, 218triangulation, 268, 272turn-taking, 105, 108, 184, 251TV meter systems, 228

ubiquitous computing, 183ubuntu (“humanness”), 205the unconscious, 2units of analysis, 220, 238unobtrusive measures, 243–5

validation, 267–8validity, 94, 183, 212–13, 225, 266–8variables, 164, 210, 211, 213, 234variance, 230, 231verification, 262vertical integration, 73virtual reality, 189visual anthropology, 243visual arts, 27voice-over narration, 131

Watson, John, 51Weber, Max, 49Whig interpretation of history, 199wholesale news exchange, 92Wittgenstein, Ludwig, 27women, images of, 35working documents, 243workshops on the future, 241writers, 71, 84writing, 21, 194–5

Index332