a1 SymbolicInteraccionismSocialStructure SocFoc40-3WEB

SociologicalFocus

SOC

IOLO

GIC

AL FO

CU

SA

UG

UST 2

00

7vo

.40

,no

.3—

Pp.2

47

—3

42

Volume 40Number 3

August 2007

PA

RA

DIG

M P

UB

LISH

ER

S3360 M

itchell Lane, Suite E

Boulder, C

O 80301

PR

ESO

RT

STA

ND

AR

DU

.S. PO

STA

GE

PA

IDA

YE

R, M

AP

ER

MIT

NO

. 9

Guest Editor’s Introduction: The Intellectual Legacyof Herbert Blumer: Theory, Method, andContemporary Challenges

Scott Grills

2007 North Central Sociological AssociationPresidential Address: Teachingand Learning and the Culture of the Regional Association in American Sociology

Jay R. Howard

Causality, Agency, and Reality: Plato and AristotleMeet George Herbert Mead and Herbert Blumer

Antony J. Puddephatt and Robert Prus

Symbolic Interactionism and the Concept of Social Structure

Alex Dennis and Peter J. Martin

Extending and Broadening Blumer’sConceptualization of Fashion: Adding Notions of Generic Social Processand Semiotic Transformation

Phillip Vannini

Mind, Self, and Human-Animal Joint Action

Clinton R. Sanders

NCSA_Cover.qxd 7/20/07 10:57 AM Page 1

Edited and published at the Department of Sociology,University of Cincinnati, in association with

Paradigm Publishers, Boulder, Colorado

Volume 40 August 2007Number 3 ISSN 0038-0237Editor: Steve Carlton-Ford Associate Editor: Paula Dubeck

Guest Editor’s Introduction: The IntellectualLegacy of Herbert Blumer: Theory,Method, and Contemporary Challenges 247Scott Grills

2007 North Central Sociological AssociationPresidential Address: Teachingand Learning and the Culture of the Regional Association in American Sociology 250Jay R. Howard

Causality, Agency, and Reality: Plato andAristotle Meet George Herbert Mead and Herbert Blumer 265Antony J. Puddephatt and Robert Prus

Symbolic Interactionism and the Concept of Social Structure 287Alex Dennis and Peter J. Martin

Extending and Broadening Blumer’sConceptualization of Fashion:Adding Notions of Generic SocialProcess and Semiotic Transformation 306Phillip Vannini

SOCIOLOGICAL FOCUSQuarterly Journal of the

North Central Sociological Association

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page i

Mind, Self, and Human-Animal Joint Action 320Clinton R. Sanders

Copyright © 2007 North Central Sociological Association

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page ii

Call for Papers

A Special Issue of Sociological FocusRacial and Ethnic Inequality in Health and HealthcareJennifer Malat, Guest Editor

Sociological Focus solicits papers for a special issue titled “Racial and Ethnic Inequalityin Health and Healthcare.” Papers that address processes and mechanisms affectinghealth status and/or healthcare quality, understudied populations, or understudiedissues are of particular interest. The journal welcomes papers from a variety ofmethodological approaches.

For further information about the special issue, please contact Professor Malat atthe Department of Sociology, 1009 Crosley Tower, Box 210378, University ofCincinnati, Cincinnati, OH 45221-0378 or email her at [email protected].

Send complete submissions to Sociological Focus, Department of Sociology,Box 210378, University of Cincinnati, Cincinnati, OH 45221-0378, no later thanDecember 7, 2007. Submissions should contain two printed copies of your paper(in ASA format), an electronic copy of the paper, and a $15 submission fee.

Sociological Focus will forward manuscripts to the special issue editor. All manu-scripts will be peer reviewed. Submission requirements may be found in a current(August 2006 or later) issue of the journal or at the journal’s Web site:

http://www.ncsanet.org/sociological_focus/PUBINDEX.html

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page iii

Call for Papers

A Special Issue of Sociological FocusThe Globalization of Crime: Focus on East AsiaLiqun Cao and Shanyang Zhao, Guest Editors

Sociological Focus solicits papers for a special issue titled “The Globalization of Crime:Focus on East Asia.” This special issue focuses on the globalization of crime and itscontrol. Topics may include, but are not limited to, testing of criminological theorieswith comparative data, crime control in East Asian cultures, human and drug traf-ficking, and various forms of emerging crimes such as Internet crime in East Asia.

For further information about the special issue, please contact Professor Liqun Caoat the Department of Sociology, Anthropology, and Criminology, Eastern MichiganUniversity, Ypsilanti, MI 48197 (e-mail: [email protected]) or Professor ShanyangZhao at the Department of Sociology, Temple University, Philadelphia, PA 19122 (e-mail: [email protected]).

Send all manuscripts to Sociological Focus, Department of Sociology, Box210378, University of Cincinnati, Cincinnati, OH 45221-0378, by January 11,2008. Submit two printed copies of your manuscript (in ASA format), accom-panied by an electronic version of the paper on computer diskette, and a $15submission fee (payable to Sociological Focus).

Sociological Focus will forward manuscripts to the special issue editors. All manu-scripts will be peer reviewed. Specific submission requirements may be found in a cur-rent (August 2006 or later) issue of the journal or at the journal Web site:

http://www.ncsanet.org/sociological_focus/Pubindex.html

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page iv

Editors’ Note

Welcome to the August 2007 issue. We are very pleased to present this year’s NCSApresidential address and four articles all focused on the current relevance of HerbertBlumer’s legacy. The presidential address marks our continuing effort to bring theannual conference to members who either did not have the chance to hear the addressat the conference or who wish to revisit Jay Howard’s analysis of the changing role ofregional sociological associations. He presents strategies for enhancing the relevanceof regional meetings for high school teachers, undergraduates, graduate students, andprofessors at all levels. His essay on “Teaching-and-Learning and the Culture of theRegional Association in American Sociology” points out several ways to make sociol-ogy a more vital area for our discipline’s diverse constituents.

The four papers in this special issue, The Intellectual Legacy of Herbert Blumer:Theory, Method, and Contemporary Challenges are the direct outgrowth of an NCSAconference held in conjunction with the miniconference titled Symbolic Interactionand Ethnographic Research 2005: The Intellectual Legacy of Herbert Blumer. BobPrus’s and Billy Shaffir’s considerable efforts as conference organizers made the mini-conference a success; Kathy Rowell’s work as NCSA program chair resulted in a seam-less integration of the two conferences, enhancing both. These papers are the directoutgrowth of that joint conference.

Scott Grills took on the tremendous work required to shepherd submissions fromthe conference through the review process, correspond with the anonymous peerreviewers, and mold the most successful papers into a coherent special issue. We knowyou will enjoy his thoughtful introduction to these excellent papers.

Steve Carlton-FordPaula Dubeck

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page v

NCSA40.03_C00.qxd 7/20/07 11:01 AM Page vi

Guest Editor’s IntroductionThe Intellectual Legacy of Herbert Blumer: Theory, Method, and Contemporary Challenges

This special issue of Sociological Focus is dedicated to the work of Herbert Blumer andhis ongoing influence on our discipline and our practice. I trust that readers will findthe articles included in this volume of interest to members of the sociological com-munity more generally, rather than Blumer scholars more specifically. Representedherein is a diverse range of voices, perspectives, and conceptual inclinations. All arefirmly grounded in the range of questions that Blumer’s work encourages us to ask.Readers will find that collectively, these articles comment directly on concepts andthemes that have become central to the Blumerian contribution to the discipline: thecritique of variable analysis, the limits on importing natural science models into ourdiscipline, the analytical importance of an attention to generic social process, the con-cept of joint action, the sociological relevance of Meadian notions of self and mind,and the framing of social structure in action terms.

If one were to identify the single unifying theme of this volume, it would be theimportance of social process for contemporary sociology. I do not modify the termsociology here quite intentionally—that is, I do not limit the importance of theseworks to symbolic interaction or to interpretive thought. I join with Maines (2001),Prus (1996), and Stryker (1989), who, in different ways and in different contexts,argue for the realization within the discipline that interactionist ideas have permeatedsociological thought. Even those outside the tradition, and in fact those who mighthold themselves and their work to be critical of the tradition, have, nevertheless,incorporated Blumerian notions rather centrally into their work without necessarilyattending to intellectual debts that may have been lost within the development ofthe discipline over time. In this context it is fitting that the underlying theme of thevolume is social process. For it is perhaps Blumer’s insistence on the study of the activ-ities of people (and the interrelated emphasis on process, duration, situation, subcul-ture, and meaning) that has most changed our discipline. This influence is reflectedin the articles that follow.

In their article “Causality, Agency, and Reality: Plato and Aristotle Meet GeorgeHerbert Mead and Herbert Blumer,” Antony J. Puddephatt and Robert Prus framethe concept of causality in process terms. By drawing out some of the parallels andsympathies shared between classical Greek scholarship and the pragmatist-infusedtheories of Mead and Blumer, the authors encourage readers to consider models ofcausality that are not dependent upon antecedent variables but rather attend to theways in which people enter causal processes.

NCSA40.03_INTpg247-249.qxd 7/20/07 11:03 AM Page 247

Likewise, Alex Dennis and Peter J. Martin (“Symbolic Interactionism and theConcept of Social Structure”) encourage readers to recognize that Blumer’s work isnot so much astructural as it is part of a larger interactionist tradition that offers aprocessually located alternative to “structural” approaches. The authors extendBlumer’s critique of concept formation and apply it to the concept of structure. By sodoing they offer a challenge to sociological determinisms.

In “Extending and Broadening Blumer’s Conceptualization of Fashion: AddingNotions of Generic Social Process and Semiotic Transformation,” Phillip Vanniniattends to an aspect of Blumer’s research that is rarely the focus of the more casualreader—his work on fashion. This article addresses the two key Blumerian themes ofsocial process and the problem of meaning. By attending to the process of semiotictransformation foreshadowed in Blumer’s work, Vannini reframes and names thiswork explicitly in generic social process terms and extends the relevance of the origi-nal work.

In his article “Mind, Self, and Human-Animal Joint Action,” Clinton R. Sandersencourages us to seriously and thoughtfully consider the notion that nonhuman ani-mals engage in joint action. By so doing, Sanders encourages a critical reexamina-tion of a rather central interactionist notion. The result is a lively argument whereinpeople’s relationships with companion animals are understood in the context of aninteraction sequence where both parties “attempt to assume the perspective of theother.”

These articles share an interest in understanding human group life in action terms.A commitment to the fundamental premises of a Blumerian symbolic interactionholds consequences not only for how we approach the empirical world but also forthe concepts we use to make sense of the social worlds we share. In this volume arethought-provoking articles that have relevance and saliency far beyond the specificsubstantive interest that marks them.

I close with a word of thanks. I wish to thank Steve Carlton-Ford and the edito-rial board of this journal for supporting this project. To the authors, I very muchappreciate your commitment to this special issue and your willingness to share yourfine work with us. I must also thank the many anonymous reviewers that were a partof the behind-the-scenes work that is an essential part of any enterprise such as this.Likewise, my thanks to Michael Wehrman, editorial assistant with Sociological Focus,for exercising his organizational skill throughout.

Scott GrillsProfessor of Sociology and Vice President (Academic and Research)Brandon University

Scott Grills received his PhD from McMaster University. He is coauthor of TheDeviant Mystique (Praeger 2003) and editor of Doing Ethnographic Research (Sage1998). He has previously served on the editorial board of the Canadian Review ofSociology and Anthropology and the interdisciplinary journal Dianoia. He can bereached at Brandon University, Manitoba, Canada, R7A 6A9 ([email protected]).

248 SOCIOLOGICAL FOCUS


REFERENCESMaines, David. 2001. The Faultline of Consciousness: A View of Interactionism. New York:

Aldine de Gruyter.Prus, Robert. 1996. Symbolic Interaction and Ethnographic Research. Albany: State University

of New York Press.Stryker, Sheldon. 1989. “The Two Psychologies: Additional Thoughts.” Social Forces 68: 45–54.

GUEST EDITOR’S INTRODUCTION 249


In this essay, I examine the role of teaching and learning in the culture ofthe regional association in American sociology. I analyze the programs of(1) the 2007 joint meeting of the North Central Sociological Association(NCSA) and the Midwest Sociological Society (MSS); (2) the 2007 annualmeeting preliminary programs of the Eastern Sociological Society (ESS),the Pacific Sociological Association (PSA), and the Southern SociologicalSociety (SSS) along with the 2006 annual meeting programs of the MSSand NCSA, as well as the American Sociological Association (ASA); and(3) the 1991 NCSA and 1992 ASA annual meeting programs. I identify pro-gram trends with regard to teaching, professional development, under-graduate students, graduate students, and research on higher education.I conclude by identifying regional association annual meeting best prac-tices regarding each of these areas.

When trying to settle on a topic for my NCSA Presidential Address, I was a bit over-whelmed. I don’t remember my proseminar courses in graduate school addressing thisoccasion! Given the lack of guidance from my graduate training, I decided to take alook at what others have done with their presidential addresses. There seemed to beseveral approaches.

A first approach to presidential addresses is to take advantage of your captive audi-ence and subject it to a talk on your area of research specialization. There are somepresidential addresses of this type that have succeeded magnificently—for example,Kent Schwirian’s 2005 NCSA Presidential Address in which he analyzed the 2002SARS outbreak in the People’s Republic of China. He captivated us with stories of theoutbreak and how political decisions contributed to the infection of more than 8,000people and the deaths of 774 in just a few months. I thought about subjecting you tomy areas of research specialization. But a dirty little secret about presidential addressesthat take the “I’m going to make them listen to a talk on my area of research infocused detail” approach is that despite the occasional smashing success such asSchwirian’s SARS talk, most presidential addresses of this type are deadly dull.

2007 North Central Sociological AssociationPresidential Address: Teaching and Learning

and the Culture of the Regional Association inAmerican Sociology

Jay R. Howard*Indiana University–Purdue University Columbus

*Correspondence regarding this article should be sent to: Jay R. Howard, IUPUC, 4601Central Ave., Columbus, IN 47203, [email protected].

NCSA40.03_C01pg250-264.qxd 7/20/07 11:00 AM Page 250

A second approach to presidential addresses is to attempt to assess the state of soci-ology as a discipline. Now, I am not arrogant enough to try to assess the state of thediscipline as a whole. So I also scratched the “assess the state of sociology” approachoff the list of possibilities. Well, sort of. I’m not arrogant enough to address the stateof sociology as a discipline, but as NCSA president, I’ve had to spend a lot of timethinking about the role of the regional association in American sociology.

In reviewing previous NCSA presidential addresses, I took note of Bruce Keith’s2004 talk examining the contextual and historical relationship between the nationaland regional associations in American sociology. He and I have had many discussionsregarding the role of regional associations in general and the NCSA in particular. Oneoutcome of those discussions was the creation of the NCSA Future FacultyProgram—but more on that later.

So taking my lead from Keith (2004) and building on his findings, I want to takea look at one of the roles of regional associations—promoting effective teaching andgreater learning in sociology. I also want to examine the role of the regional associa-tion in encouraging the professional development of sociologists—a task that clearlyoverlaps with promoting more effective teaching and greater learning.

Why should we care about these issues? Aren’t professional conferences all aboutresearch? Why bother with a focus on teaching and professional development in theregional associations? There are a host of reasons why we need to pay more attentionto these issues. First, accreditation organizations, state legislatures, and universityadministrators are all paying increasing attention to the assessment of teaching andlearning. This year the governor of Texas proposed tying increases in funding forhigher education to exit exams for college graduates—sort of a No Child Left Behindprogram for higher education (Fisher and Hebel 2007). The students’ “Academic Billof Rights,” which purportedly seeks to ensure that faculty members do not use theirclassroom position for the “purpose of political, ideological, religious, or antireligiousindoctrination” has gained a hearing in several state legislatures (Horowitz 2006;Lipka 2006). Faculty members are increasingly being asked to provide assessment evi-dence regarding student learning outcomes. I think there is little doubt that others arebeginning to take a careful look at what is happening in college classrooms. Perhapswe should apply our sociological imagination to our own daily setting and provideevidence of how sociology contributes to students learning to think critically and tosee how social structures influence their lives.

Another of my favorite presidential addresses was Maxine Atkinson’s (2001)address to the Southern Sociological Society, which focused on the scholarship ofteaching and learning. Atkinson was a trailblazer in daring to use a regional associa-tion presidential address to talk about teaching! She stressed that in a culture of highereducation increasingly concerned with educational reform, we need to make sure wehave our house in order. A focus on effective teaching and the assessment of studentlearning can help us through the process of such initiatives as post-tenure review—one of those things that state legislatures frequently demand.

On the one hand, we should be concerned with teaching and learning because out-side forces are going to make us document our contribution to the university whether

2007 NORTH CENTRAL SOCIOLOGICAL ASSOCIATION PRESIDENTIAL ADDRESS 251


we like it or not; on the other hand, we should be concerned with teaching and learn-ing because it is the right thing to do. As the father of a college student, I face thosetuition bills. I want to be certain that my child is learning something as a result of mysignificant financial investment in her education. It has been instructive to be on theother side of the equation for a while. Admittedly, many of our students are comingto us so that they can get a high-paying job when they graduate. But part of the bar-gain is they should get an education along the way. We have an obligation to offerthem one—even when they aren’t always sure they really want to be bothered with it.

THE ROLE OF REGIONAL ASSOCIATIONS IN AMERICAN SOCIOLOGY

Keith (2004) noted that regional sociological associations largely emerged in the 1930sin an effort to provide sociologists access to professional conferences because travel dif-ficulties and expense prevented many sociologists from attending the national meet-ings of the American Sociological Association. In essence, the meetings of the regionalassociations attempted to be mini-ASA meetings providing sociologists the opportu-nity to present their work and learn about the latest research findings. Indeed, this con-tinues to be a major and important goal of regional associations today.

However, times have changed. Despite all the air-travel hassles that followedSeptember 11, 2001, travel to the national meetings of the ASA is certainly cheaperand easier than it was in the 1930s. Many more faculty members can and do partic-ipate in the national association today than did in the 1930s. In particular, facultymembers from research-oriented institutions have tended to focus their participationon the more prestigious national meetings. Keith (2004) suggests the result is that theregional associations, with some clear exceptions, largely serve the needs of facultyfrom institutions that do not grant PhDs in sociology. Keith (2004) examined fivemajor regional associations in American sociology: the Eastern Sociological Society(ESS), the Midwest Sociological Society (MSS), the North Central SociologicalAssociation (NCSA), the Pacific Sociological Association (PSA), and the SouthernSociological Society (SSS). He found these regional associations to have much incommon and to be diverging from the American Sociological Association (ASA).

One of Keith’s key findings was that regional associations underutilize the popula-tions they represent. The NCSA, squeezed in the area not claimed by the MSS andESS, draws primarily from only three states: Indiana, Ohio, and Michigan. Thisleaves the NCSA by far the smallest of the regional associations. Each of the otherfour regional associations is comparable in terms of membership, ranging from justunder 1,000 members to roughly 1,300 members in 2003. By estimating the popu-lation base of prospective faculty and student members, Keith (2004) revealed thatregional associations fall far short of their potential membership, with the NCSAreaching only about 11% of its potential population of faculty.

Another key finding of Keith’s (2004) analysis is that constituencies of the nationaland regional associations are diverging. Regional associations are increasingly serving



faculty from four-year colleges, rather than from institutions granting MAs or doctor-ates. At the same time, it has proven difficult to attract members from two-year colleges.

So what’s a regional association to do? One strategy is to “give up” on doctoralinstitutions and focus exclusively on meeting the needs of faculty from two- and four-year colleges, high school sociology teachers, and applied sociologists. However, asKeith (2004: 96) points out, 89% of graduates of those doctoral programs will endup with jobs at institutions that do not grant PhDs. Thus, regional associations havea role in the professional socialization of graduate students to prepare them for thetypes of jobs they are likely to find when they complete their doctoral work.Additionally, graduate students constitute a significant proportion of the membershipof regional associations—we would be foolish to ignore them.

So where do we go from here? I will offer some direction based on analysis of wherewe have been and where we are currently as regional associations. In particular, I aminterested in the role of professional associations in promoting effective teaching andlearning in sociology and in providing professional development opportunities forfaculty members and graduate students. I believe these are important foci if regionalassociations are truly going to appeal to and meet the needs of faculty at two-year andfour-year institutions and graduate students in doctoral programs. Clearly, everyonefrom state and national legislative bodies to accreditation agencies is interested inassessment of student learning. Likewise, given the seemingly always tight job marketfor sociology PhDs, anything that graduate students can do to better prepare them-selves for jobs at the types of institutions where they are likely to work is a good thing.In addition, an increasing number of institutions is looking for ways of documentingfaculty commitment to improvement and professional growth. Given these nationaltrends, regional associations have an opportunity to stake out new ground and meetthe needs of their constituents in ways we have not previously done.

RESEARCH METHODSFirst, I examine the program of the 2007 combined annual meeting of the NorthCentral Sociological Association and the Midwest Sociological Society. If we step backand look at the big picture, what are we doing? Second, I compare the preliminary pro-grams of the 2007 annual meetings of the ESS, PSA, and the SSS to the 2006 programsof the NCSA and MSS—their most recent programs as separate associations. Next, Icompare these with the 2006 program of the ASA. Finally, I make a comparison of theNCSA and ASA programs of 15 years ago with 2006 programs. By analyzing what weare doing with regard to teaching and learning as well as professional development inregional associations and the ASA, I hope to identify what we’re doing well, where wecould improve, and what people in K-12 circles call “best practices.”

I began by analyzing the conference programs to identify (1) teaching sessions,(2) professional development sessions, (3) sessions directed specifically toward under-graduate students, (4) sessions directed specifically toward graduate students, and(5) sessions focused on research pertaining to higher education. I defined “teaching



sessions” as any program session that focused on the teaching of a particular course(e.g., “Teaching Social Problems”), focused on a particular pedagogical strategy(e.g., service learning), or any meeting that focused on teaching included on the pro-gram (e.g., a meeting of the teaching committee). If a single session combined teach-ing and research, I counted it as a teaching session if 50% or more of the paper topicswere teaching-focused. For roundtable sessions with multiple tables with distinct top-ics, I counted each table with a topic that focused on teaching as a separate session.

“Professional development sessions” were defined as guidance or advice for facultymembers in areas other than teaching (e.g., “Surviving the First Few Years on the Job”or “How to Negotiate with Your IRB”). “Undergraduate sessions” were defined as ses-sions designed for the benefit of undergraduate students (e.g., “Applying to GraduateSchools”) or sessions designed to give undergraduates an opportunity to present theirwork (e.g., undergraduate paper or poster sessions). Most graduate students presenttheir research as a part of regular sessions at regional association meetings. However,some regional associations, such as the Pacific Sociological Association, have speciallydesignated sessions for graduate student research, if the students prefer to present ina session that includes only other graduate students. These sessions, along with ses-sions directed toward graduate students (e.g., academic job search sessions), werelabeled “graduate student sessions.” Finally, I identified sessions as “research on highereducation sessions” if they addressed topics relevant to higher education but were notclear about teaching or professional development. For example, this category includedsessions on such topics as “Socioeconomic Status and Achievement in College” and“21st Century Campus Race Relations.” Because these sessions on research on highereducation apply the sociological eye to our own institutional context, I believe theymake a contribution that is qualitatively distinct from sociological research on othersocial institutions and processes.

2007 MEETING OF THE NCSA/MSSAs Table 1 shows, there are 40 teaching sessions on this year’s NCSA/MSS program,accounting for roughly 12% of the total program. There are sessions on teaching par-ticular topics (e.g., “Strategies for Teaching Politically Charged Topics”), on teachingparticular skills (e.g., “Teaching Quantitative Literacy”), on teaching particularcourses (e.g., “Rethinking the Introductory Sociology Course”), and on the use ofparticular pedagogical strategies (e.g., “The Use of Music in Sociology Classrooms”).

This year’s program also shows a commitment to professional development for soci-ologists with sessions that cover such topics as publishing (e.g., “A Conversation withJournal Editors”), funding sources (“NSF Funding Opportunities”), research skills(e.g., “Social Analysis Tools from the Census Bureau”), and mentoring (e.g., “FeministMentoring”). There are 14 professional development sessions on the program, whichaccount for about 4% of the program sessions.

Our commitment to students, as judged by the 2007 NCSA/MSS program, isweaker. There are only three sessions specifically directed toward undergraduate



students—an undergraduate paper competition session, an undergraduate poster ses-sion, and a session called “Undergraduate Perspectives.” We also have only two sessionsfocused specifically on graduate students—a graduate student paper competition ses-sion and a session titled “Feminists in Graduate School: Negotiating a Chilly Climate.”The combined sessions specifically directed toward undergraduate and graduate stu-dents account for only 1.5% of the program.

Finally, we have research sessions on the topic of higher education. TheNCSA/MSS joint program features ten such sessions, about 3% of the program.These sessions include such topics as “Financial and Structural Issues in HigherEducation,” “Assessing the Collegiate Student Culture,” and “Self-Perceptions ofStudent Achievement.”

In sum, roughly one of every five sessions, 69 of the 338 sessions on the 2007NCSA/MSS program, cover one of these five topics: teaching, professional develop-ment, undergraduate students, graduate students, or research on higher education.This, of course, leaves about 80% of the program for sessions that are primarily tra-ditional sociological research sessions. The question is: Is this the ideal balance? Is this


Table 1. 2007 North Central Sociological Association/Midwest Sociological Society JointAnnual Meeting Program Sessions by Session Type

MSS/NCSASessions 2007 Sample Session Titles

Total 3381. Teaching 40 Strategies for Teaching Politically Charged Topics

(12%) Teaching Quantitative LiteracyRethinking the Introductory Sociology CourseThe Use of Music in Sociology Classrooms

2. Professional 14 A Conversation with Journal EditorsDevelopment (4%) NSF Funding Opportunities

Social Analysis Tools from the Census BureauFeminist Mentoring

3. Undergraduate 3 Undergraduate Research Poster(1%) Undergraduate Student Paper Competition

Undergraduate Perspectives4. Graduate 2 Graduate Student Paper Competition

(.5%) Feminists in Graduate School: Negotiating a Chilly Climate

5. Research on 10 21st Century Campus Race RelationsHigher Education (3%) Financial and Structural Issues in Higher Education

Assessing the Collegiate Student CultureSelf-perception of Student Achievement

1, 2, 3, 4, & 5 69Combined (20%)


balance in the long-term best interests of the NCSA, MSS, and other regional associ-ations? A way to begin answering that question is to compare the NCSA and the MSSwith other major regional associations, the ESS, the PSA, and the SSS. Are ourapproaches similar to or different from theirs?

COMPARING THE REGIONAL ASSOCIATION PROGRAMSWhile our 2007 joint meeting of the MSS and the NCSA was a tremendous success,with over 1,500 sociologists in attendance, it is unique. There were more sessions andmore participants in 2007 than in years past for both organizations. It is possible aswell that the nature of the program changed because of our collaborative efforts.Therefore, in order to compare the programs of the MSS and the NCSA with otherregional associations, I took the 2006 final programs of the MSS and the NCSA andcompared them with the 2007 preliminary programs of the ESS, the PSA, and theSSS. Table 2 presents the results of this comparison.

Teaching Sessions in the Regional Associations

The five regional associations committed an average of 12% of the program to teach-ing sessions. The NCSA had the most teaching sessions both in terms of the numberof sessions (30) and the percentage of the program focused on teaching (30%). The


Table 2. 2007 Eastern Sociological Society, 2007 Pacific Sociological Association, and2007 Southern Sociological Society Preliminary Program Sessions and 2006 North CentralSociological Association, 2006 Midwest Sociological Society, and 2006 AmericanSociological Association Final Program Sessions by Type

RegionalAssociation

NCSA MSS ESS PSA SSS Mean ASASessions 2006 2006 2007 2007 2007 Percentage 2006

Total 99 197 272 228 165 6081. Teaching 30 22 12 20 14 48

(30%) (11%) (4%) (9%) (8%) 12% (8%)2. Professional 5 16 9 6 5 34

Development (5%) (8%) (3%) (3%) (3%) 4% (6%)3. Undergraduate 4 10 4 14 14 5

(4%) (5%) (1%) (6%) (8%) 5% (1%)4. Graduate 4 4 0 11 2 5

(4%) (2%) (0%) (5%) (1%) 2% (1%)5. Research on 0 6 1 21 4 5

Higher Education (0%) (3%) (.004%) (9%) (2%) 3% (1%)1, 2, 3, 4, & 5 43 58 26 72 39 97Combined (43%) (29%) (10%) (32%) (24%) 28% (16%)


ESS (12 sessions, 4%) had the fewest teaching sessions and the lowest percentage oftotal sessions dedicated to teaching. The MSS (22 sessions, 11%), the PSA (20 sessions,9%), and the SSS (14 sessions, 8%) were clustered between these extremes. As with the2007 MSS/NCSA program, the teaching sessions fell into four categories: (1) teachingtopics covered in multiple courses, (2) teaching particular skills, (3) teaching specificcourses, and (4) pedagogical strategies. The topics covered in the NCSA, MSS, PSA,and SSS were quite similar to each other. However, the ESS stood out because five ofthe twelve teaching sessions at the ESS had titles that were generic. These included,“Pedagogy” (twice), “Classroom Practice and Teaching,” “Pedagogy in HigherEducation,” and “Teaching Substantive Topics.” The other associations tended to havenot only more teaching sessions than the ESS but sessions that were more focused.

There were some unique approaches. The NCSA included a day-long high schoolteachers’ workshop in the program. NCSA teaching sessions were arranged so that atleast one teaching session was scheduled in every regular time slot on the program,allowing members whose primary interest is teaching to attend those sessions. TheNCSA also presented the teaching sessions as a program, allowing graduate studentswho attended five teaching sessions the opportunity to earn an NCSA Future FacultyProgram Certificate. This program was created in recognition that contributing to theprofessional socialization of graduate students is a valid and necessary function ofregional associations. In both the call for papers and in the preliminary program,graduate students were encouraged to attend the annual meeting both to present theirresearch and to earn a Future Faculty Certificate by simply attending a certain num-ber of designated sessions and having the presiders initial a form. After the confer-ence, graduate students who attended the appropriate sessions received a certificatecertifying their completion of the NCSA Future Faculty Program.

Professional Development Sessions in the Regional Associations

Sessions emphasizing faculty professional development other than teaching, sessionsfocused on undergraduate or graduate students, and research sessions on higher educa-tion were less visible in the regional association programs than were teaching sessions.

The MSS had the most professional development sessions with sixteen, account-ing for 8% of the total program. Professional development sessions accounted foronly 3–5% of the programs of the other regional associations. Publishing research,balancing work and family, obtaining tenure, and concerns of women and minoritiesin higher education were the most common topics. None of the regional associationsattempted to offer professional development as a systematic program feature of theannual meeting.

Sessions for Undergraduate Students

Sessions for undergraduate students averaged 5% of the regional association programsessions. Both the PSA and the SSS had 14 sessions for undergraduate students—8%of the ESS sessions and 6% of the PSA sessions. Both had many undergraduate paper,poster, and roundtable sessions. The MSS program included ten sessions for under-graduate students, including multiple paper and poster sessions. The ESS program



primarily consolidated undergraduates into three poster sessions. The NCSAincluded undergraduate papers in three roundtable sessions. The SSS took what to anoutsider appears to be the most systematic approach. Of the 14 undergraduate ses-sions on the SSS program, 13 were in two time slots early on Saturday morning—thefinal day of the conference. This, in effect, places a mini-undergraduate research con-ference on the program of the SSS.

Sessions for Graduate Students

Because most graduate students participate in regular sessions at the regional associa-tions’ annual meetings, it is perhaps not too surprising that there were fewer sessionsdirected toward graduate students than undergraduate students. The regional associ-ation programs averaged only 2% of the program sessions being directed toward grad-uate students. The PSA had eleven sessions for graduate students. This was due in partto a practice of having research sessions set aside for those graduate students who donot wish to participate in sessions with faculty members. Eight of the eleven PSAgraduate student sessions were of this type. More typically, graduate student sessionsfeatured graduate paper competition winners, offered advice on surviving graduateschool, and presented information on the academic job search process.

Higher Education Research Sessions

Finally, research sessions on higher education typically accounted for only a small por-tion (3%) of the program sessions of the regional associations. The PSA was theexception with 21 sessions—9% of the program. The theme of the 2007 PSA annualmeeting was “Sociology and the Academy: Its Current and Prospective Position.”However, even without the 6 special “presidential sessions” on higher education, therewere an additional 15 sessions dedicated to sociological examinations of higher edu-cation. In contrast, the NCSA had no sessions on the topic in 2006 and the ESS hadonly one higher education session in 2007.

In total about 28% of the five major regional association programs are dedicatedto teaching, professional development, undergraduate students, graduate students,and research on higher education. How does this compare to what goes on at meet-ings of the American Sociological Association (ASA)?

COMPARING THE REGIONAL ASSOCIATIONS WITH THE ASAAt the 2006 ASA annual meeting in Montreal, there were a total of 608 sessions onthe program (see Table 2). Of these, 48 sessions, or 8%, were focused on teaching.This is two-thirds of the 12% average for the five major regional associations. But theNCSA is a statistical outlier with 30% of the 2006 program dedicated to teaching. Ifyou compare the ASA to the other four regional associations, the ASA has a compa-rable percentage of teaching sessions to the PSA (9%) and SSS (8%). It is not farbehind the MSS (11%) and doubles the ESS (4%).

In 2006 the ASA program featured 34 professional development sessions—6% ofthe total program. That is more than any of the regional associations except the MSS,



which dedicated 8% of the program to professional development. The regional associ-ations, on average, dedicated a higher percentage of the program (5%) to sessions forundergraduates than did the ASA (1%). The ASA program had a similar percentage ofsessions (1%) dedicated to graduate students to the regional associations (2%).Research on higher education did not receive much emphasis at the ASA meetingeither, with only 1% of the program dedicated to the topic.

In total, the 2006 ASA annual meeting dedicated 16% of the program to teaching,professional development, undergraduates, graduate students, and research on highereducation, in comparison to 28% of the program at the regional associations. Onlythe ESS dedicated a smaller portion of its program (10%) to these five areas. Perhapsit is logical for sociology’s national association to have a stronger research emphasis inits program (84%) than the regional associations (ranging from 57% to 90%).

THE NCSA AND THE ASA: YESTERDAY AND TODAYIn order to consider change over time in the percentage of the NCSA and ASA pro-grams dedicated to teaching, professional development, undergraduate students,graduate students, and research on higher education, I compared the 2006 NCSAprogram with the 1991 NCSA program of 15 years ago. I did not have easy access tothe 1991 ASA program, so I compared the 2006 ASA program with the 1992 ASAprogram of 14 years ago. I discovered that although the number of sessions in theNCSA annual meeting has stayed about the same (near 100 sessions), the ASA hasseen substantial growth from 377 program sessions in 1992 to 608 in 2006—a gainof 231 sessions! (See Table 3.)


Table 3. 1991 and 2006 North Central Sociological Association and 1992 and 2006American Sociological Association Final Program Sessions by Type

NCSA NCSA ASA ASA Sessions 1991 2006 Change 1992 2006 Change

Total 104 99 –5 377 608 �2311. Teaching 15 30 �15 17 48 �31

(14%) (30%) (�16%) (5%) (8%) (�3%)2. Professional 11 5 –6 28 34 �6

Development (11%) (5%) (–6%) (7%) (6%) (–1%)3. Undergraduate 5 4 –1 1 5 �4

(5%) (4%) (–1%) (.2%) (.8%) (�.6%)4. Graduate 0 4 �4 0 5 �5

(0%) (4%) (1%)5. Research on 4 0 –4 4 5 �1

Higher Education (4%) (0%) (1%) (1%) (na)1, 2, 3, 4, & 5 35 43 �8 50 97 �47Combined (34%) (43%) (�9%) (13%) (16%) (�3%)


During this period the NCSA doubled the percentage of its program dedicated toteaching (15% to 30%) while the ASA increased from 5% to 8%. That is especiallycommendable given the overall increase in the number of program sessions at theASA meeting. The real number of teaching-related sessions at the ASA went from 17to 48, nearly a threefold increase!

However, commitment to professional development has decreased. The NCSA hasseen the percentage of its program dedicated to professional development cut in half.As far as I am aware, this was not an intentional decision. The percentage of the ASAprogram declined somewhat from 7% to 6%—though the real number of sessionsincreased from 28 to 34.

The NCSA has maintained roughly the same number and percentage of sessionsdedicated to undergraduates—from five to four (5% to 4%). The ASA has increasedthe number of sessions focused on undergraduates, but the percentage of the programis still less than 1%. In 1991, the NCSA had no sessions targeting graduate studentsand the ASA had none in 1992. In 2006, 4% of the NCSA program and 1% of theASA program focused on graduate students.

The NCSA program featured research on higher education in about 4% of the 1991program, but no sessions in 2006. The ASA held steady with about 1% of the programin each year dedicated to research on higher education.

LESSONS LEARNEDSo now that we’ve taken this descriptive tour of the programs of the regional associa-tions in American sociology as well as that of the ASA, what can we learn from oneanother? Are there best practices? Are there opportunities and threats to take into con-sideration? I think there are several lessons to learn.

The first pertains to the significant growth of the ASA program. This is further evi-dence that more sociologists are choosing to participate in the national ASA meeting.That is good news for the ASA and the discipline as a whole. I support sociologistsparticipating in the ASA annual meeting. But given shrinking travel support in highereducation, the increased participation in the ASA presents a challenge for regionalassociations. We have to work harder to attract and keep participants. What canregional associations offer to sociologists other than a shorter distance to travel to themeetings? What do we have to offer to sociologists at various stages of their careers?

Let me emphasize that I am not suggesting that we quit presenting research atregional association meetings. Research has been and will always be an important partof our meetings. The problem occurs when research is treated as if it is the only thingthat matters at the meetings. So let’s continue to do good research and share it withour colleagues for critique and feedback at regional associations. But we need to doother things as well.

What are the various regional associations doing well? How can we learn from oneanother? With regard to teaching, regional associations can learn from the NCSA.Organizing the program so that there is at least one teaching session in every time slot



makes a lot of sense. Inevitably, at any professional conference there will be a limitednumber of sessions that pertain to the research agenda of any individual sociologist. Sowhat should we do with the rest of the day at the annual meetings? Focusing on improv-ing teaching and learning is one good use of that time. Every regional association shouldwork to have at least one teaching-related session in every time slot on the program.

I think the NCSA has also blazed some trails regarding teaching in a couple ofother ways. First, the High School Teachers Workshop helps to build bridges betweensociologists in higher education and those who teach sociology in high schools. If wewant high school sociology courses to be truly sociological, we had better take the ini-tiative to help those teachers develop effective courses.

Second, the NCSA, in response to the challenge former president Keith (2004) setforth, has sought to become more graduate student–friendly by organizing the teachingsessions on the program in such a way that graduate students can earn an NCSA FutureFaculty Program Certificate. By taking steps to ensure that a teaching session is availablein every time slot on the program, we can reward graduate students not only for pre-senting their research but for engaging in sessions that will help them be better teach-ers when we hire them as colleagues! The additional work to package what is alreadyfeatured in the program in a fashion that will benefit graduate students is minimal.And when we are hiring new colleagues, we need to give applicants credit for seekingto develop themselves as teachers as well as researchers.

The regional associations have some good things in place with regard to teaching.But with the possible exception of the MSS, there is a lot of room for improvementwith regard to professional development. If at the annual meeting of regional associ-ations, we don’t want sociologists to be “drive-by” participants who come in, presenttheir papers, and leave, we need to offer benefits that will encourage them to stay andengage in the meeting and the organization. What might do that? Every regionalassociation program could include professional development sessions such as“Surviving the First Years on the Job,” “Balancing Work and Family in HigherEducation,” “Keys to Obtaining Promotion and Tenure,” “Strategies for EffectiveDepartment Chairs,” “Post-Tenure Review,” “Benefits and Challenges for Sociologistsin Administration,” “Strategies for Successful Departmental Reviews,” “Publishingin Scholarly Journals,” and “Writing Effective Book Proposals.” Perhaps these ses-sions could be organized in such a way as to allow faculty members to earn aProfessional Development Program Certificate from the regional association.

Regional associations also need to take time to collaborate with community collegefaculty members when it comes to professional development. Community collegesare becoming the starting point for many students who will end up with a bachelor’sdegree from another institution. We need to seek to meet the needs of and to learnfrom community college faculty. How can we organize regional association meetingsto give community college faculty a reason to attend? First, I think an organizedemphasis on teaching helps. The first priority of virtually all community college fac-ulty members is classroom teaching. How can we help them, as well as ourselves,become better teachers? Community college faculty may also be more likely to haveparticipation in professional development activities as part of their job expectations.



Can we organize our professional development sessions in a way that makes docu-mentation of professional development easier for faculty who are expected to engagein it, regardless of the type of institution at which they teach? Second, we shouldavoid being paternalistic toward community college faculty members. We need to bewilling to learn from them. For example, community college faculty members aremore likely to have invested time and energy in assessing and documenting the con-tributions of sociology to general education. We need to ensure that these colleaguesfeel welcome and respected within our regional associations so that we can benefitfrom each other. It will take a concerted, intentional effort to invite and involvethese faculty members; since our associations have been focused so heavily onresearch in the past, many community college faculty members may not perceivethe annual meeting program as having anything for them.

Regarding undergraduate students, I think the regional associations could all borrowa good idea from the SSS. In 2007 the SSS, in effect, created a mini-undergraduateresearch conference on the last morning of the annual meeting. Undergraduate stu-dents could attend the rest of the meeting, taking in research sessions and keynotespeakers, and finish by presenting their own work. There are a couple of challengeshere. First, there will be a need to make undergraduate student membership afford-able. Many colleges and universities are emphasizing undergraduate participation inresearch and even providing some travel funding. If regional associations make par-ticipation both affordable and nonthreatening by having a mini-undergraduate con-ference within our annual meetings, we may be beneficiaries of this new emphasis inhigher education. A second problem is finding faculty members willing to stickaround for the last morning of the meeting to preside over and attend undergraduatepaper sessions. Perhaps a good group to begin with is the regional association’s teach-ing committee. Each teaching committee member could accept responsibility to chairone undergraduate research session as a part of his or her obligations as committeemembers. Alternatively, regional associations could create undergraduate committeesto coordinate the miniconference within the meeting. Regardless, faculty memberswill have to step up and commit their time and energy to make this proposal work.

Graduate students’ involvement in regional associations is unique. They are caughtin a liminal stage between student and faculty member. However, in most waysregional associations treat graduate students more like faculty colleagues than likegraduate students. But involving graduate students in the same way as faculty is notenough if we wish to entice them into joining regional associations and becominglong-term members. The aforementioned NCSA Future Faculty Program is one strat-egy for increasing graduate student involvement. Offering special sessions thatinclude research only by graduate students, as does the PSA, may also be an effectiveway to allow graduate students the opportunity to present at a conference—especiallyfor first-time presenters. But graduate students could also benefit by having specialsessions directed to them each year. Some possible sessions for graduate studentscould be: “What I Wish Someone Had Told Me during My First Year in GraduateSchool,” “Writing Effective Dissertation and Thesis Proposals,” “What to Look forWhen Selecting an Advisor,” “Negotiating the Academic Job Market,” and even a session



on “How to Get More Involved in Your Regional Association: Opportunities forGraduate Students.” These could be part of the Future Faculty Program or an addi-tional benefit of the program for graduate students.

What about research on higher education? Perhaps this is my bias showing, but Ithink sociologists have a lot to offer by bringing the sociological imagination to ourunderstanding of higher education. Most regional association members spend themajority of their waking hours in higher education institutions. We can take the toolsof our discipline and help everyone understand why some students don’t finish theirdegrees, why others don’t survive even the first year of college, and how race, class,and gender affect outcomes in higher education. It is ironic that sociologists are soeffective at bringing sociological insights to others’ daily lives but are not always inter-ested in turning a sociological eye upon their own world. The PSA is to be com-mended for making “Sociology and the Academy” the theme of its 2007 program.Perhaps other regional associations should follow its lead and occasionally make theinstitution where we live, higher education, the focus of our annual meeting.

THE ROAD AHEADIf I may be so bold as to think that you believe some of these proposals are worth pur-suing, where do we go from here? The next step is the hard part. It is up to the mem-bers of the NCSA and other regional associations to make changes happen. Regionalassociations are voluntary organizations. Things get done only because members wantthem done and are willing to work to see them accomplished.

Which of these initiatives grabs your attention? Which would you like to see hap-pen? Step up and volunteer. Talk to members of the NCSA Council or the leadershipof other regional associations. Talk to your NCSA presidents (past, present, and elect).Push to ensure that the NCSA evolves the way you want it to. If the regional associ-ation in American sociology is to survive and thrive, it must change and evolve. Itmust give sociologists from all types of institutions and at every point in their careersadditional and better reasons to participate. Consider this your call to action.

Jay R. Howard is Professor of Sociology and Head of Liberal Arts at IndianaUniversity–Purdue University Columbus. He served as the 2006–2007 President ofthe NCSA. His research interests range from the Scholarship of Teaching and Learningto religion and popular culture. He is a fellow of the P.A. Mack Center at IndianaUniversity for Inquiry on Teaching and Learning. He also served as the 2006–2007President of the Bartholomew Consolidated School Corporation Board of Trustees inColumbus, IN.

REFERENCESAtkinson, Maxine. 2001. “The Scholarship of Teaching and Learning: Reconceptualizing

Scholarship and Transforming the Academy.” Social Forces 79: 1217–1230.



Fisher, Karin and Sara Hebel. 2007. “Texas Governor Seeks Financing Overhaul.” Chronicleof Higher Education (February 16, 2007) available at http://chronicle.com/weekly/v53/i24/24a03203.htm.

Horowitz, David. 2006. “After the Academic Bill of Rights.” Chronicle of Higher Education(November 10, 2006) available at http://chronicle.com/weekly/v53/i12/12b02001.htm.

Keith, Bruce. 2004. “Disciplinary Culture and Organizational Dissonance: The RegionalAssociation in American Sociology.” Sociological Focus 37: 83–105.

Lipka, Sara. 2006. “‘Academic Bill of Rights’ Criticized.” Chronicle of Higher Education (June9, 2006) available at http://chronicle.com/weekly/v52/i40/40a01202.htm.

Schwirian, Kent. 2005. “Globalization, Plague, and the Local Community: HealthcareCapacity, Politics, and the Microbe War.” Sociological Focus 38: 151–170.



Causal theorizing in sociology is often presumed best handled in thedomain of multivariate, quantitative variable research. In contrast, weargue that such theorizing may be more authentically accomplished withinthe theoretical and methodological impetus of symbolic interactionism.Analyzing notions of causality in the work of Plato and Aristotle, we con-sider their relevance for contemporary accounts of human group life withreference to the work of George Herbert Mead and Herbert Blumer. Ratherthan envisioning causality as a concatenation of external forces acting onpeople to produce outcomes, we focus on how people actively enter intothe causal process through individual and joint action. Using the work ofPlato and Aristotle, we bolster contemporary accounts of agency by locat-ing the moving process of deliberation within the interplay of purposiveactivity, emotional states, bodily appetites, and changing character dispo-sitions. Further, we show that agency as a conceptual locus point of socialchange is found not only in the spontaneity of linguistically enabled indi-viduals but also in the emergent processes of deliberative interchange andcollective action in the world.

Interactionist research is often criticized as antiscientific, astructural, and subjectivistin its depiction of social life (e.g., Grenier 1992; Huber 1973; Zeitlen 1973).Demerath (2002) has gone so far as to assert that causal theorizing is the purview ofthose who use a multivariate research strategy and that ethnographic researchers areleft to generate merely descriptive theory. However, as Prus (1996, 1999) observes,these critiques are often based on partial and otherwise inadequate representations ofthe positions developed by George Herbert Mead (1934), Herbert Blumer (1969),Anselm Strauss (1993), and others working in Chicago-style interactionism (see Fine1993; Reynolds and Herman-Kinney 2003). Nonetheless, these criticisms also reflect

Causality, Agency, and Reality:Plato and Aristotle Meet George Herbert Mead

and Herbert BlumerAntony J. PuddephattMcMaster University*

Robert PrusUniversity of Waterloo

*We would like to thank Michael Adorjan, Robert Andersen, Tina Fetner, Steve Carlton-Ford, Scott Grills, Benjamin Kelly, Neil McLaughlin, Dorothy Pawluch, Aaron Segaert, andthe anonymous reviewers from Sociological Focus for their critical commentary and sug-gestions at various phases of this project. Please address correspondence to Antony J.Puddephatt, Department of Sociology, McMaster University, Kenneth Taylor Hall, Room 722,1280 Main Street West, Hamilton, ON L8S 4M4, Canada.


the unwillingness of interactionist scholars to incorporate the influence of external,antecedent causes into their formulations of human group activity. Indeed, manyinteractionists also explicitly avoid considerations of causality in their work. Taylor andBogdan’s (1984: 2) statement that “the search for social causes is neither what this bookis about nor where our research interests lie” seems indicative of the distance fromexplanatory theory within the interactionist paradigm. Likewise, it is almost impossi-ble to find scholars who broach the topic of causality in the major qualitative methodsmanuals.1 Neither Denzin and Lincoln (2000) nor Reynolds and Herman-Kinney(2003) engage the issue of causality in any serious way within their collections.

In contrast, the issue of causality has received a great deal of attention in the quan-titative methods literature (Berk 1988; McKim and Turner 1997). Lazarsfeld (1959)outlined three essential criteria for establishing causal relations in variable analysis insocial research. He lists these as: (1) antecedence; the cause must precede the effect intime, (2) correlation; the variables used must be empirically correlated together, and(3) there must be a control for spurious variables. Often added to these requirementsis (4) the causal relationship identified should represent a reasonable expectation fromsome set of theoretical propositions. Since then, other quantitative scholars haveinvoked more exacting criteria for establishing cause in variable analysis. Cox andWermuth (2001: 70) argue that (5) causal relations also ought to be “found repeat-edly in independent studies, especially if these are of somewhat different form.” Inthis way, implied causal associations would be less peculiar to one particular surveyand could be generalized across contexts. Cox and Wermuth further argue for (6) anincreased attention to path analysis, where “baseline” variables are seen to affect“intermediate” variables, which then bring about effects seen in the “final response”variable, to bolster proof of antecedence.

Still, mindful of the philosophy of David Hume, many quantitative scholars (7)reject the imputation of causation to variable relationships entirely (Arjas 2001), andinstead use the semantics of “correlations” and “tentative generalizations.”Nevertheless, the models constructed are still implicitly, if not explicitly, causal—leading to conceptions of human behavior that are built upon the supposed deter-ministic properties of external structural conditions.2 Goldthorpe (2001) hasobserved that many quantitative researchers now treat causal relationships (8) inmore probabilistic terms, such that the existence of causal variables do not guaran-tee a particular effect, but rather raise the probability of its occurrence. Further,Goldthorpe states that (9) the weakness of statistical methods is their inability toobserve processes as they happen in real time, and asserts (10) the theoretical neces-sity of a focus on temporal processes, testing these with longitudinal designs.Further, Goldthorpe (2001: 15) stresses (11) the need to branch out into deeper,microsocial methodologies, due to the “need for causal explanations of social phe-nomena to be grounded ultimately in accounts of the action and interaction of indi-viduals.” Arjas (2001: 59) agrees that “ordinary statistical analyses . . . lead to genericconclusions . . . about cause and effect that have little contact with the true behav-ior and reasoning of the individuals considered.” Given that some contemporaryquantitative researchers locate a great deal of potential in qualitative, face-to-face



approaches for their unique methodological abilities to study causal processes as theyunfold in real time, it is particularly ironic that most symbolic interactionists havebeen reluctant to address the topic head-on.

One reason for the neglect of causality in interactionist research may be thatChicago-styled interactionists have evolved from a tradition that has historically beenset against structuralist approaches, and so still view themselves as the “loyal opposi-tion” to positivist social science (Mullins 1973). The idea of studying “causal rela-tions” may have been too closely associated with the very structuralist frameworks theinteractionists were resisting.3 Blumer (1969) argued forcefully that such positivistmethods are inauthentic and fail to respect the nature of their human subject matter.Blumer’s rejection of traditional variable-centered structural analysis would serve to“rally the base” and have the biggest impact on the interactionist tradition for yearsto come. Indeed, many of the Chicago-style symbolic interactionists (e.g., Lofland1976, 1995; Prus 1996; Strauss 1993) maintain that the dominant sociologicalemphases on variable analysis is misguided. This does not mean these scholars haveignored causal reasoning or structural explanations entirely. However, mostBlumerian interactionists refuse to engage in causal analysis by reducing complex real-ity to survey variables and presumed or reified structural conditions that mysteriouslyact on people in deterministic ways (Prus 1999). Instead, the interactionists havetended to handle the issue of causality more implicitly, emphasizing the language of“developing theoretical linkages,” or attending to the evolution of “social processes.”Thus, although interactionists have actually laid out theories of causal relationshipsthat consider the ways that people engage structural conditions in concrete empiricalstudies (e.g., Ferraro 1983; Prus and Irini 1980; Wiseman 1970), they have beenreluctant to use the language of causality explicitly, probably because of its embed-dedness in positivist frameworks.

Despite this, some interactionists have been more explicit in the consideration ofcausality. Lindesmith (1981) argues that the pragmatist approach to scientific inquiry(specifically Dewey and Mead) sees causes and effects as interactional properties oflarger interdependent processes. To label something a cause, or an effect, dependsupon its place within this larger process. Lindesmith contends that although sociolo-gists often follow physics in their methodological constructs, biology is the bettermodel, since it is based much more on whole systems of interdependencies ratherthan the isolation and control of variables (see also Gotschalk 1941). For Lindesmith,this biological analogy makes much more sense than does physics for the study ofhuman societies, as the intention should be to study effects that stem from an irre-ducible interactional process as it goes on. More recently, Maines (2003: 10) haslamented the neglect of causation in symbolic interactionism, stating that “the inter-actionist heritage is grounded in concerns about causation,” using examples such asThomas and Znaniecki’s (1918) Polish Peasant and Blumer’s (1990) study of indus-trialization. Maines argues that interactionists have been emphasizing all along whatquantitative scholars have only recently begun to realize (e.g., Berk 1988; Britt 1997;Goldthorpe 2001)—that causality in relation to group life must be treated indeter-minately, with a central place for the role of human agency.

CAUSALITY, AGENCY, AND REALITY 267


In this spirit, we aim to develop a conceptualization of causality in more proces-sual, interactionist terms with a concerted emphasis on the role of social collectivesactively entering into the causal process. It must be recognized that humans are notpassive mediums through which social forces work, but that they have the ability toenter into the causal process themselves as agents of change in their own right(Blumer, 1969; Prus, 1996, 1999; Strauss, 1993). Although people are not alwayssuccessful or “wise” in their choices and activities, people have the ability to intervenein, and generate anew, causal sequences that may not be fully predictable by recourseto background conditions alone. To aid us in such a formulation, we draw not only onthe foundational work of Mead (1934, 1938) and Blumer (1969) but also the early for-mulations of cause offered by Plato (420–348 BCE) and Aristotle (384–322 BCE).Although this may seem a conceptual disjuncture, Prus (2003, 2004) has indicatedsome of the useful affinities between classical Greek scholarship and interactionistethnography. Scholars in the Greek era (circa 700–300 BCE) not only addressed theconcept of causation in consequential ways, but they also considered these issues inmaterialist, humanist, and legalist terms. Further, these scholars have an understand-ing of agency that is richer and more nuanced than contemporary accounts, as it con-nects motivation and decisionmaking practices to changing character dispositions,bodily desires, and emotion. In the final analysis, we hope to provide a conception ofcausation that indicates not only the ways in which people enter into the causalprocess but also how the agency of individuals and groups is contingent on the activ-ities and interchanges of the social collective.

PLATO ATTENDS TO CAUSALITYPlato is best known for his writings on theology, idealism, morality, and justice. Eventhough he is notably prescriptive in defining the moral order of the community in histwo most comprehensive works, Republic and Laws, Plato remains acutely attentive to theprocesses and problematics of establishing and maintaining an assortment of insti-tutions to regulate people’s lifestyles and participation within the community.Consequently, instead of merely prescribing and proscribing, as many moralists do,Plato frequently takes aspects of human group life apart, piece by piece, in more sec-ular, pluralist terms. Plato offers a sustained dialectic analysis of a number of areas ofhuman endeavor, such as politics, religion, poetics, friendship, rhetoric, education,scholarship, and science, often in ways that resemble American pragmatist emphases.Plato not only grounds much of his analysis in prototypic human situations, ten-dencies, practices, and deliberations but also examines these matters in extended,comparative detail. Thus, it is as a dialectician that Plato’s work is most significant tothe consideration of causality. Viewing dialectic analysis as the basis for developinga theory of knowledge, Plato takes a distinctively pluralist, generic approach to thematter. Despite the cynicism or totalizing skepticism often associated with Plato’sdialectics, Plato shows that although knowing may be problematic, it is essential foran understanding of the human condition. Although focused on causality more



specifically, the material we introduce here reflects Plato’s “philosophy of knowing”more broadly and, as will become evident, has remarkable contemporary relevance.

With Plato’s more distinctively pluralist, secular, and dialectic emphases in mind,we first consider how Plato views causation as it operates in the material world.Second, we ask how he approaches the notion of cause more specifically with regardto human behavior. As human cause is conceived to be intimately connected to delib-eration and choice, this necessarily leads to a consideration of Plato’s theory of moti-vation and human action. Rather than providing an exhaustive consideration of Plato’scontributions to causality, we highlight those features that may be of particular rele-vance to social scientists in their respective research agendas. Hence, it is instructiveto consider Plato’s explanations of physical changes in Statesman, where he contrasts“contributory” causes with the more “final” or “true” causes in the physical world:

Well then, let’s look at two sorts of expertise that there are in relation to all thethings that people do . . . one which is a contributory cause of production, onewhich is itself a cause. . . . Those which do not make the thing itself, but whichprovide tools for those that do—tools which, if they were not present, what hasbeen assigned to each expertise would never be accomplished: these are what Imean by the contributory causes, while those that bring the thing itself intocompletion are [true] causes. (281d-e)

Plato clearly recognizes that one could identify an endless number of causes for the gen-eration, existence, or depletion of anything, as the succession of causal chains continuebackward in time ad infinitum. As such, they are of little interest in the construction ofwhat is meant, for Plato, as a true explanation. Instead, Plato believes that the develop-ment of a true explanation should be sought by recourse to the more final and conse-quential phases of development. This argument relates to Lindesmith’s (1981) insistenceon the methodological decision to “close the system” at some point, and try to identifythose features deemed most relevant to a particular investigation. This becomes espe-cially relevant in the attempt to explain human behavior. As actions are most often thedirect result of, and follow through, the immediacy of the deliberative process, scholarswho are interested in the “true” causes of behavior ought to be more mindful of thismore final phase of deliberation that leads to and accompanies action.

Beyond recognizing that deliberation is often the most final cause sequentially,Plato considers the explanatory value of agency and the thought processes associatedwith choice in action that is distinctly human in character. The mere sequence oftemporal succession is not as relevant for prioritizing a true cause for Plato in the caseof human behavior. Rather, it is the more central moving force of deliberative actionby the actor who enters knowingly into the causal process. The following excerpt, inwhich Socrates is speaking, indicates this emphasis:

[I]n trying to tell the causes of everything I do, to say that the reason that I amsitting here is because my body consists of bones and sinews . . . the relaxationof and contraction of the sinews enable me to bend my limbs, and that is thecause of my sitting here with my limbs bent. . . . If someone said that withoutbones, and sinews, and such things, I should not be able to do what I decided,



he would be right, but surely to say that they are the cause of what I do, and notthat I have chosen the best course, even though I act with my mind, is to speakvery lazily and carelessly. Imagine not being able to distinguish the real causefrom that without which the cause would not be able to act as a cause. It is whatthe majority appear to do, like people groping in the dark; they call it a cause,thus giving it a name that doesn’t belong to it. (Plato, Phaedo 98c–99c)

At some point, one needs to strip away causal elements judged to be less central intrying to find the most consequential explanation for action. Plato stresses the neces-sity of attending to the most central determining causes of action, those of humanthought, deliberation, and choice. It is worth noting that Plato puts deliberation firstas the most central explanatory feature, over and above the finality of physicalmotion. This is justified in that muscular action can be seen as a sort of preexistingpotential, and as such, a contributory cause; musculoskeletal action is like a tool thatthe act of deliberative thought puts into action. Further, since the deliberative processfollows through the action temporally, deliberation is not entirely antecedent but alsocoexists with and follows after physical action.

If deliberation is seen as the most consequential determinant of human action, itis important to understand Plato’s model of deliberation. Plato’s account of delibera-tion reveals a multiplex model of human motivation, deliberation, and action (seealso Cooper 2000). Rather than taking a purely “rational choice” approach, Plato inRepublic (Book IV) divides the human soul (also psyche [Greek], de anima [Latin])into three separate spheres, all of which are seen to be active during the process ofhuman conduct. These are (1) reason (the faculty of rational thought); (2) appetite(inclinations toward food, drink, sex, bodily pleasure); and (3) spirit (energy and asso-ciated emotions such as anger and excitement). Still, Plato defines virtuous activitiesas those actions that are guided in such a way that the appetite and spirit are ade-quately moderated by the faculty of reasoning. Plato locates reasoning within theinseparable nexus of appetite and emotion. As such, Plato holds that bodily appetite,emotional energy, and reasoning constitute the synergy for the human psyche, andthus, provide a complex integrated model for human motivation, thought, andaction. For Plato, human deliberation involves an overlapping and often ongoingconflict between the three areas of the human psyche. It is the human capacity for rea-son Plato prizes most highly, as it is this capacity that overrides, tames, or moderatesappetite and spirit such that an individual can act justly in the world.

Mindful of these themes, let us consider what Plato offers contemporary consider-ations of causation in social analysis. First, Plato provides a justification for the pri-macy of isolating “final” or “true” explanations, especially in the case of humanbehavior. This is in opposition to questing after less relevant and potentially infiniteconceptions of “contributory causes,” which are shown only to allow for the “true” or“final” causes of a particular act to occur. We contend that the central failing of mostquantitative analyses is that, at best, they produce highly detailed, rigorous, and sophis-ticated models of contributory causes of human behavior. By drawing attention to cor-relations, time antecedence, and potential spurious variables, quantitative analysis



provides knowledge about some of the working parameters of human action.However, virtually all quantitative data, including those focusing directly on attitudesand preferences, miss the final and more determinative cause of human behavior—the process of minded, deliberative, and adjustive activity. To explain people’s activi-ties and the actual outcomes they experience with recourse to the most consequentialcauses, the final analysis requires the unique theoretical and methodological toolkit ofsymbolic interaction and ethnographic research.

We contend that to develop better, more thoroughgoing models of people’s behav-ior in certain situational settings, it is necessary to concentrate squarely on the con-scious processes by which these “contributory elements” may be taken into account,utilized, avoided, or ignored in the construction of action. As such, “external” elementscan be more connected to actor decisions (and given causal significance) insofar asthey become internal to deliberative action and enter into the minded thoughtprocesses of individuals and collectives as they construct lines of individual or jointaction. Plato allows us to reevaluate what this phase of deliberation entails, and howwe might reconceptualize our understanding of the human actor. Rather than pro-viding a calculating or rational, means-ends model of motivation, Plato envisionsdeliberation as a dynamic, humanly engaged process. Reasoning does not occur in avacuum as disinterested exercises in rationality (Descartes) or dialectic skepticism(Nietzsche) but unfolds within people’s concerns with purposive, meaningful activityamid the sway of human emotion and the temptation of bodily appetites. In short,deliberation is inseparable from the unity of reason, emotional energy, and the habits,preferences, and resistances that people develop. As such, these ought to be muchmore primary in the analysis of human action, which, as Plato teaches us, is extremelyconsequential in explanations of situational behavior.

ARISTOTLE ENGAGES CAUSATIONAlthough Aristotle’s considerations of the human condition are greatly informed byPlato’s scholarship, Aristotle’s approach to causality and the human condition differsin important respects. Whereas Plato emphasizes a dual-world metaphysics (sensateversus divine realities), Aristotle insists on the biological, material, and experientialconstitution of human thought. Accordingly, Aristotle views the development of peo-ple’s psyches as evolving in an integrated, processual fashion within the context ofcommunity life—wherein human knowing and acting presupposes the acquisition oflanguage. In relation to this, Aristotle’s notions of causation are informed by distinc-tions between inanimate objects, organisms, and humans as animals with the capac-ity for speech and recollectable memory. Like Plato, Aristotle envisions the conceptof cause as necessary for explaining change in the world. Further, Aristotle recognizesthat to understand how things work and how change occurs, the most relevant causesof change ought to be isolated and examined: “[W]e must proceed to consider causes,their character and number. Knowledge is the object of our inquiry, and men do notthink they know a thing until they have grasped the ‘why’ of it (which is to grasp the



primary cause)” (Physics 194b15–23). Viewing the mind as an emergent entity thatdevelops through experience with the sensate world and through instructed associa-tion with others (see Spangler 1998), Aristotle realizes that causes may be viewed asterms people invoke or assign to things in their quest to know (develop meanings)things. It is important, thus, to acknowledge Aristotle’s awareness of the constructedquality of human knowledge claims.

A simplified statement on Aristotle’s full conception of causality, his “doctrine ofthe four causes,” serves as a useful starting point for our analysis.4 The doctrine encap-sulates the following features: (1) the matter or substance of which something is con-stituted; (2) the shape or form that something assumes; (3) the mover of the processor the source of the effect (the “efficient” cause); and (4) the end or purpose of theproduct or outcome (the “final” cause). Although Aristotle clearly intends to encom-pass all physical instances in his statement on causation, he does not ignore the roleof human agency. Nevertheless, commentators often present these notions in highlytruncated forms and have tended to focus on purely physical or material notions ofcausality. We intend to consider what Aristotle’s conceptualization of the four causesdoes for an understanding of causality as a process and how human actors may enterinto this process as active agents of change.

Aristotle’s consideration of the first cause (materials or components) emphasizes theimportance of permanence and continuity in change: “that out of which a thing comesto be and which persists, is called a cause, e.g. the bronze of the statue, the silver of thebowl” (Aristotle Physics 194b24–26). This way, the cause can be seen not only as exist-ing antecedently but also as coexisting with the effect. This enables one to conceptual-ize causal relations as processes by which change does not happen in absolute terms,but rather occurs through a temporal phase, whereby continuities (e.g., the silver of thebowl) are as important as discontinuities (the changing of shape) in the larger movingprocess. This solves the problem of antecedence and concurrence, as materials andcomponents are seen to meet both requirements. Further, the emphasis on continuityand change, as will become apparent, links closely with contemporary interactionistformulations (Blumer 1969; Mead 1934; Prus 1996; Strauss 1993).

The second cause listed by Aristotle, that of the form or shape assumed, points to theconstructed nature of causal propositions: “the form or the archetype, i.e., the definitionof the essence, and its genera, are called causes (e.g., of the octave the relation of 2:1),and the parts of the definition” (Physics 194b27–29). Indeed, in order to recognizeeffects and attribute causes to these effects, people must first come to a definition of theessence to be explained. Aristotle was aware of the necessity of people imputing cate-gories, meanings, and forms to things such that they can be conceptualized and indi-cated. Whereas Plato discusses these forms as eternal, divinely given entities, Aristotleviews forms as constructed from the sensate experiences that take place in the process ofengaging instances of things and attributing meaning to events in the world.

The third “efficient” cause considered by Aristotle involves the primary cause, orwhat is considered the agent of change. This directly allows for the role of the humanactor to enter into the causal process. Aristotle defines this third cause as the “primarysource of the change or rest: e.g., the man who deliberated is a cause . . . what makes



of what is made and what changes of what is changed” (Physics 194b30–33). Aristotlegives explicit credence to the fact that the human being, through the act of delibera-tion, can be the efficient cause of change in the human and material world. He allowsthe process of deliberation, or the activation of agency and choice, to take a centralrole in humanly engaged causal relations. This becomes much more evident in hisconsiderations of responsibility and the attribution of blame in questions of moralityand justice (see Nichomachean Ethics [NE], Rhetoric). It is here that the “agent ascause” becomes more central as a way to justly attribute blame or innocence tooffenders.

The final cause considered by Aristotle is “the sense of end or that for the sake ofwhich a thing is done” (Physics 194b34–35), which also allows for explanations to beconstructed around human deliberation. An end may be purely materialistic anddevoid of human intention. For example, a flower in bloom is the end that explainsthe change from a bud. However, an end may also be understood teleologically, interms of human purpose. Thus, for Aristotle, the human definitions of purpose arenecessary in explaining the causes of voluntary (as opposed to involuntary) behavioron the part of individual or joint activity.

If Aristotle sees human action as constitutive of causal relations in social life, thenwhat does his model of human action look like? Aristotle provides a highly complextheory of behavior and shows that human action can be brought about by a numberof different causes, not all of which are voluntary. Aristotle recognizes, for example,that people are not always fully cognizant of, or able to fully control, their actions. Inhis consideration of ethical conduct, Aristotle (NE, Book III) contends that peoplenot only act voluntarily (with awareness and choice), but also nonvoluntarily (with-out an awareness of situational parameters) and involuntarily (without choice). InRhetoric (1369a5–10), Aristotle defines those forms of action that are not due to theagent in question. There is the explicit recognition that people often do things bychance (accidents, misunderstandings, unpredictability), compulsion (coercion by anexternal force), and nature (the activation of the human senses). Thus, it is importantto recognize that these elements of conduct may enter into the causal processes ofhuman group life in determinate ways and bring about unintended consequences.However, given our concerns with the role of actors intervening into causal relationsas agents, we will turn to Aristotle’s view of voluntary activity and then more centrallyto his account of deliberation and choice.

Acknowledging habits, passions, and extenuating circumstances, it is apparent thatnot all voluntary actions are a result of conscious deliberation or choice. Habitual activ-ity, for example, is voluntary, but does not require the exercise of deliberation to proceed.Still, Aristotle recognized the vital importance of habitual activity in generating one’scharacter, and hence, one’s predisposition to act in prescribed ways in the future:

[B]y doing the acts that we do in our transactions with other men we becomejust or unjust, and by doing the acts that we do in the presence of danger, andbeing habituated to feel fear or confidence, we become brave or cowardly . . .states arise out of like activities. This is why the activities we exhibit must be of



a certain kind. . . . It makes no small difference, then, whether we form habitsof one kind or another from our very youth; it makes a very great difference, orrather all the difference. (NE Book II: 1103b10–26)

Aristotle’s admonition that “people become what they do” is consistent with the inter-actionist position that people’s relationships, knowledge of the world, and sense of selfemerge from the intersubjective and corporeal web of mutually focused activity(Strauss 1993). In line with an interactionist motif, the Aristotelian self evolvesthrough active learning and adjustment through social and practical experience (seealso Spangler 1998). Aristotle also considers voluntary choices that are nonrational,and center more on appetite (food, drink, sex) or emotion (anger, fear). In sum, peo-ple often act in ways that reflect involuntary or nonrational habits, appetites, desires,emotions, accidents, compulsions, or instances of ignorance. Indeed, Aristotle’s modelof agency opens the way for social analysts to move beyond simplistic models of theself-aware actor who continuously makes use of a self-maximizing instrumental logic.

Aristotle (NE, Book III, iii) states that people do not deliberate about everything—but rather over those things over which they have some control that seem attainable,and that are defined as uncertain. Aristotle (NE, Book VI, ii) identifies three aspects ofthe human psyche that are integral to action as well as people’s definitions of the truth.These are sensation, thought, and desire. Aristotle contends that sensations cannot inthemselves generate rational action. He also observes that desire may provide direction,but that people’s desires are also inadequate to explain human behavior. The “efficient”cause of human action is to be found within the deliberative process. Still, for Aristotle,thought in itself moves nothing. Thought is consequential in causal terms only whenit is manifested in action. Aristotle states that people engage in deliberative activity byunifying desire and thought to direct or regulate their actions.

Equally consequential is Aristotle’s point that when people consider particular issuesimportant, they are more likely to seek counsel or otherwise involve others in theirdeliberations. Therefore, Aristotle’s theory of agency is highly relevant to sociologicalaccounts in that deliberation is not pursued solely on an individual basis, but occurs asa community phenomenon. Aristotle’s presentation of influence work and tactical per-suasion and resistance illustrates the ways in which deliberation, as an intersubjectiveprocess, may influence the views and decisions of others. Outlining a set of operationaltactics for embarking on influence work, Aristotle is highly attentive to the processualand problematic features of human definitional and choice-making activities.Aristotle’s Rhetoric also indicates with remarkable clarity and depth how people’s emo-tions are socially malleable and can be manipulated and reshaped by others. As such,Aristotle’s consideration of the emotions not only attends to the ways in which peo-ple’s emotional sensations affect their decisions but also shows how speakers can tacti-cally influence (intensify or neutralize) others’ emotional states. Since people’semotional states are inextricably linked to their reasoning processes, emotion is seen asboth an obstacle and a resource for effective influence work. More importantly, forAristotle, emotion is not understood merely as an individually given phenomenon, butis actively mediated and reshaped through the intersubjective process. The direct



integration of emotion into the tactics of persuasion and influence work seems to be amuch richer version than the model of interpersonal persuasion offered by ErvingGoffman (1959). Indeed, Aristotle’s more integrated notion of emotions as genuineconstituents of ongoing activity offers a much more vibrant picture of human group life.

In short, Aristotle provides us a model of causality that not only allows for con-tinuity and discontinuity in the unfolding of events but also for the direct involve-ment of human agents (as passionate, emotive, and biologically appetitive beings)in causal social relations. Finally, Aristotle has a remarkably sophisticated under-standing of the fact that by engaging others in influential terms, people may pro-foundly shape and reshape the experiences and definitions of reality and actions ofothers as individuals and collectivities. Thus, Aristotle effectively lays the founda-tion for a sociological theory that would root causality within the practices andactivities of those who constitute the human community, as subgroups and selvesevolve through the collectively based and situationally emergent processes ofthought and action.

GEORGE HERBERT MEAD: PROCESS AND EMERGENCELike Aristotle, Mead locates agency as a “prime mover” of change not only in thehuman actor’s individual psyche but also in the linguistic process of deliberation inthe social world. Of course, for Mead (1934), the act of deliberation is a fundamen-tally social process, in that the self-communication associated with reflective thoughtis a microcosm of the same process seen in the larger community. Conceptualizingagency within collective interchanges rather than individual selves providesresearchers many more potential departure points for conceptualizing causal processesas they take shape emergently in social life. As such, we will consider Mead’s varioussources of innovation and spontaneity that could be seen as “efficient causes” of socialchange in the interchanges of the community. We hope this will highlight some of thelines of convergence between Mead’s pragmatist theories of causation and the lineageof the aforementioned Greek thinkers. As well, this may elucidate some of the areasin which Mead’s contributions help formulate a sharper conceptualization of causal-ity in interactionist research, by identifying various sources of agency within the socialprocess generally.

Mead (1932, 1938) locates causality within process. Although many philosophers,most notably David Hume, have problematized the contradictory idea of antecedenceand concurrence, Mead adopts a position similar to Aristotle and contends that spe-cific changes can come about only through, and in relation to, a larger set of inter-connected processes: the “basis of this determination of the future by the past is foundin the fact that something is taking place which has a temporal spread—that realitycannot be reduced to instants. . . . Furthermore, the study of passage involves the dis-covery of events. . . . The relation of any event to the conditions under which it occursis what we term causation” (1932: 62). Also consistent with Aristotle’s insistence onsubstance (e.g., “the silver of the bowl”) for the necessity of continuity in explanations



of change, Mead argues that change cannot be disconnected from the processes lead-ing up to it. Instances of change can only occur within continuous processes or devel-opmental flows: “It is the enduring character of the experience that contains in it thecontinuity of nature, just the connection that Hume denied. There is something thatcontinues. If there is something that continues, that which is there at the present timeis responsible for what is going to be there in the future” (1938: 647). Drawing inspi-ration from Darwinian evolutionary theory, Mead argues that forms are always to beexplained as emerging from their connections (and continuities with) the variousenvironmental conditions that gave rise to them. Particular events are always consti-tuted by, and to be understood under, the larger processes that led to their creation.

Although Mead emphasized the continuity of process and the need to understandinstances as part of the larger processes with which they connect, he also placed agreat deal of emphasis on emergence (Mead 1932, 1936, 1938). Following thevitalism of Henri Bergson, Mead argued that nature advances in ways that cannotbe fully predicted from background conditions. Although science attempts toexplain the natural world in abstracted and predictable ways, emergence, unique-ness, and the unpredictable are all parts of reality that cannot be adequately sub-sumed in a scientific cause and effect logic, except in a post-hoc way to fit anomalousoccurrences within accepted systems of thought. Thus, the state of nature carrieswith it this unique tension for Mead. On the one hand, instances are to be under-stood best within the larger processes and conditions out of which they emerge; yetthe emergent event is often unpredictable and not fully reducible to backgroundconditions and developmental processes. It is the advent of emergent anomalies thatallows for innovation and creativity in science, technological design, and social life(see Puddephatt 2005). Emergent events that occur in the complex and interde-pendent “sociality” of the world (Mead, 1932) allow for the very locus points ofchange. As Lindesmith (1981: 88) observes, Mead saw discordant evidence and theunexpected to be the very growing points of science; it is in the nature of emergencethat the “growing points” of social life find their source as well. There is always anelement of the unpredictable to emergent, ongoing events, at all levels of society(Chang 2004). As Maines (2001: 50) observes “New forms are not found in oldforms; they are found in the conditioning of emergent events, which are the adjust-ments to novelty and which give rise to change.”

The fact that individuals and groups have the capacity for self-consciousness andself-directed behavior makes emergence a much more central concern for the studyof human life than in the natural sciences. As reflective entities, individuals andgroups are capable of projecting into the future through imaginatively role-taking inthe physical and social world (Mead, 1938), and as such, can interject into, andchange, would-be chains of events before they unfold in real time. The source of self-determination is found in Mead’s (1934) depiction of the “biologic individual” andhis emphasis on the dialectical relationship between the “I” and the “Me” in the actor’sconsciousness. The “I” refers to the actor operating on a set of prereflective organicimpulses in the present. Since these impulses are nonsymbolic and prelinguistic, theycannot be socialized and are biologically rooted. As time unfolds, the impulsive “I”



moves into the past and is later objectified with language and monitored and assessed asthe “Me,” then responded to by the impulses of the present “I” (Mead, 1934: 372).Thus, the “I” enters into the internal dialogue as an impulsive biological response tothe learned, socialized, and language-based “Me” that represents the self as an object(Kolb 1944; Lewis 1979). As such, the “I” can be seen as an internal emergent withinthe actor that will respond to conditions and provide spontaneous inclinations withineven the most routine and habitual contexts of action.5

Although Mead’s model stresses the spontaneous, intersubjective, reflective, andactivity-based nature of the human psyche, some weaknesses are apparent in compar-ing it with those cast by Plato and Aristotle. In comparison to these thinkers, Mead’sconception of the self seems to be almost entirely devoid of pleasure or emotionalexperience. For Mead, minds develop out of a natural tendency to adapt to problem-atic situations. We are left with a problem-solving social organism that is relativelyfree from affective concerns or bodily desires. This is where Aristotle’s notion of themind as operating within an interfusion of passions, appetites, nonrational desires,and habitual tendencies provides a much richer model of human activity. Both Plato’sand Aristotle’s conceptualizations of the psyche as a confluence of reason with pas-sions, emotions, and desires might provide a way to constitute Mead’s “I” with moredetail, giving it more specifiable substance. As suggested by Prus (2005), this opensup a fascinating arena of research in which scholars may begin to investigate the inter-play of emotional and bodily states as they serve to influence reasoning-relatedprocesses that go well beyond simplified images of the rational actor.

As well as locating the potential for innovative change within the emergent pro-gression of nature and the internal dialogue of creative individuals, Mead also locatesa source for agency in the nature of physical and social action. With his colleagueJohn Dewey, George Herbert Mead shares the rejection of a means-ends model ofinstrumental rationality that forms the basis for most economic as well as many soci-ological theories of human conduct. For Dewey and Mead, ends are not determinedat the outset, such that action is the simple execution of predefined and singular ends.Rather, ends are constantly emerging in the development of the act itself (see Mead, 1938).As Joas (1996) has shown, Dewey’s and Mead’s pragmatism can be seen as a debunk-ing of microeconomic models of the rational actor. Because of the emergent nature ofaction, not all information is available before the act proceeds. Further, people havethe capacity to monitor, assess, and adjust to their situations even as they are in theprocess of developing particular lines of action.

Within this embodied dialectic the “creativity of action” (Joas 1996) finds its root;it is in the spontaneous and unpredictable emergence of the new in social action.Mead (1938) argues that the phenomenon of uniqueness and unpredictability isinseparable from ongoing experience. However, the efficient cause of change forMead requires that we locate individual actors within emergent communities ofaction. Even when people act with direct intention, they cannot be fully aware ofwhat effect other people or objects may have, nor can people know all of the thingsthat will affect their own participation in the causal process. As objects unto them-selves, then, people may monitor, assess, and adjust their activities (and interactions)



even as they engage in particular lines of activity. By combining Mead’s recognitionof the developmental flow of human consciousness in the specific development of theact with Aristotle’s insights into the tactical and emotional foundations of rhetoricalactivity, one can more directly locate the efficient causes of change within ongoingrealms of collective interchange.

HERBERT BLUMER: PROVIDING THE METHODOLOGICAL ORIENTATION

[T]he failure of social scientists and psychologists to pay attention to the formation ofsocial action by the acting unit is astonishing in view of the fact that such social actionis what actually goes on in empirical social life. . . . The methodological position of sym-bolic interactionism is that social interaction must be studied in terms of how it isformed; its formation is a very different matter from the antecedent conditions that aretaken as the “causes” of the social action and is not covered by any specification of suchcauses. (Blumer 1969: 57)

Revisiting Blumer’s (1969, 1971, 1990) theoretical and methodological emphases, wecan see that he shares with Plato a dislike for diffuse, contributory causes as adequateexplanations of human conduct. Further, Blumer shares with Aristotle the concern oflocating efficient cause within the realm of deliberative practice. Like Mead, Blumeris acutely attentive to the concept of emergence and the human capacity for adjust-ing thought and activity even as one develops particular lines of action. As such,Blumer stresses the importance of analyzing the social act as the constitutive phe-nomenon of social life, and provides a theoretical and methodological agenda withwhich to do so within sociological research. Blumer’s emphasis on ethnographicinquiry allows for the situated isolation and study of efficient causes within the jointactivities of small groups but also within the dynamics of larger organizational set-tings. This provides interactionists a unique ability to uncover and examine the social-interactive locus points of the causal linkages found in social life.

Mindful of the socially constituted and emergent nature of human group life,Blumer directly opposes psychological and structuralist models of human interaction,and contends that the social act itself is formative in character. Like Mead, Blumerenvisions the ongoing interactive, adjustive process of the social act in formulating itsoutcome; the formative nature of the social act requires that it cannot be reduced toantecedent conditions alone. Thus, Blumer (1969: 71) emphasizes that “the essenceof society resides in an ongoing process of action—not in a posited structure of rela-tions. Without action, any structure of relations between people is meaningless.”Thus, the central unit of analysis for sociologists is the “social act.” Although envi-sioning all meaningful individual activities as social acts by virtue of their embedded,linguistically meaningful contexts, Blumer stresses that stable structural systems aredeveloped, maintained, and achieve viability only through deliberative action onthe part of the people who have aligned themselves with these particular realms ofactivity. In other words, organizational entities create real “effects” on human life,



or generate a more definite realization of action on the part of people, only whenthese essences are incorporated into the ongoing social process.

Blumer’s methodology does not deter us from studying the effects of institutionsand organizations. Rather, as Maines (1977, 1988), Morrione (1988), Strauss (1993),Prus (1999), and Maines and Morrione (2001) contend, Blumer’s methodology isuniquely suited to the study of organizations and institutions. Blumer does not denycorrelations between increased industrialization and certain features that arise insocial relations. However, he does argue that, far from being unilateral in direction,notions of causality in these cases are apt to be highly interactive, such that changingsocial relations may well bring about new forms of industrialization and vice versa.Beyond this, Blumer believes that it is mythical or fictional to claim that “industrial-ization” as a homogenous entity influences “social relations” as a diffuse effect. Nochange can occur, Blumer argues, without concrete contact points at which thesetransformations can take place. These contact points do not have an independentexistence but instead are mediated through the social process. Only by focusing onthese points of meaningful activity and interchange, Blumer (1990) contends, is itpossible to better understand the specific effects of increased industrialization onsocial relations and vice versa. By locating these contact points of experience, the qual-ity, form, direction, and magnitude of various organizational directives can be exam-ined at the particular points at which they enter into the deliberative processes ofindividuals or groups.

Whereas Blumer recognizes that structural dynamics do occur, they are alwaysachieved through interacting social collectives. Although acting in settings that havebeen created in various respects by their predecessors, it is people, through theiractivities, who generate, maintain, extend, readjust, disregard, dismantle, resist, anddestroy the “objectified” features of community life. Distancing himself from pre-suming the deterministic effects of such antecedent structural or organizationalconditions on the realization of human action, Blumer writes: “Taking each otherinto account in this mutual way not only relates the action of each to that of theother but intertwines the actions of both into what I would call, for lack of a bet-ter word, a transaction—a fitting of the developing action of each into that of theother to form a joint or overbridging action. . . . The transaction (which I think isthe real form of human interaction) is constructed or built up in the process ofoccurrence” (1969: 109–110).

Blumer’s emphasis on the dynamics of human interchange locates an emergentpotentiality of agency, as an efficient cause of change, within the social act. Thus,although people may anticipate certain matters and outcomes before entering partic-ular transactions (and may be correct in many respects), participants are unable tofully envision the actualized outcomes of formative social interchanges. As inSimmel’s (1978) analysis of exchange, Blumer’s transitive social act carries with it acreative and sui generis quality that is irreducible to the component parts contributedby individuals. As such, transitive social acts cannot be accurately anticipated throughthe previous intentions of individual actors alone (psychological reductionism), northe structural preconditions in place (structural reductionism).



Placing primary emphasis on the ongoing social process itself as the moving force ofsocial relations, Blumer provides a methodological approach equipped to uncover thecausal mechanisms of change as these take place within the assessments, actions, adjust-ments, deliberations, and interchanges of actors in particular settings. As a socialprocess, change (and causation) takes place within the minded deliberations of individ-uals, in the deliberative interactions of small assemblies, in the activities that enableorganizations to function, and the representative and media-enabled communicationsthat allow people to relate to others at the most extended national and global levels.

By focusing on the formative and developmental features of the social process asthe central unit of analysis, Blumer has isolated the essential features of causal rela-tions in the human community. Concentrating on the concrete contact points ofchange in social life, we may begin to examine the ways that the more enduring struc-tural elements, meanings, and ideologies people have created over time enter into themore immediate instances of social process and are engaged by people in the mainte-nance or adjustment of the social order. By implementing the methodologicalemphases of exploration and inspection, social researchers can begin to isolate themore consequential elements of causal relations by recourse to their location inhuman action, thought, or deliberation as it takes place in social life. Only throughinvestigations of the points at which human activity intersects with particularinstances of social structure, institutions, and technology can we examine the wayspeople take part in and actively shape the causal processes that permeate social life andmaintain or adjust the social order.

CONCLUSION: CAUSALITY IN PERSPECTIVERevisiting the concept of causality, we have examined and compared theapproaches to the study of human knowing and acting developed by Plato,Aristotle, Mead, and Blumer, asking how their conceptions of causality might bet-ter inform social scientific research. Plato’s work shows us that the contributorycauses of behavior, such as antecedent structural conditions (often measured bysurvey research) are much less relevant to the explanation of human behavior thanthe more immediate or “final” causes leading to action. Further, Plato emphasizesthat human deliberation and implementation are central for comprehending theproduction of human action. Aristotle enriches this position by stressing the cen-trality of human activity and, more specifically, considering when and how people(deliberatively as well as inadvertantly and nonrationally) enter into the causalprocess as the “efficient” causes or prime movers of change. Although not denyingthe place of antecedent conditions or structural factors, at best these represent thecontributory causes of human action. Because they miss or disregard the ways var-ious structural features of organizational entities are actually defined, assessed, andadjusted to by people who actually encounter and deal with these systems, struc-turalist approaches neglect the efficient causes at work in the maintenance andchange of the social order.



As Blumer (1969, 1990) asserts, the agents of change are to be found not in thestructural antecedents of action, but in the ongoing and formative process of socialaction. Moreover, for Mead and Blumer, as with Plato and Aristotle, the concept ofagency as a prime mover of change in social life is found not as a quality held by indi-vidual actors, but rather in the emergent interchanges, deliberations, and events thatarise in coordinated social activity. Actors often draw on or attend to structural andinstitutional parameters in constructing their lines of action, but these so-called fac-tors should not be mistaken for primary determinants of behavioral outcomes.

Since we stress that humans are able to enter into the causal process as agents ofchange, we also need to establish a workable model of human agency. Whereas Meadis attentive to people as biological creatures, his more general depiction of the humanas a problem-solving agent gives little explicit attention to the role of emotionality.Blumer also distances himself from explanations that reflect emotional dispositions,contending that such approaches reduce the social interactive process to pregivenpsychological states. Both Plato and Aristotle, in contrast, envision the human psychenot only as an active, linguistically informed essence but also as a biological phe-nomenon that is deeply emotional, appetitive, and often influenced by nonrationaltendencies. Further, both Plato and Aristotle view actors as struggling for self-controlamid their dispositions to act in a multiplicity of ways. In these respects, and mind-ful of his pronounced emphasis on activity, an Aristotelian model of the humanactor has much to offer to a viable model of agency. Indeed, the behavioral predis-positions, emotional energies, and appetitive desires claimed by actors in the con-struction of deliberative activity add considerable conceptual vitality to the Meadianconcept of the agent.

As a central feature of activity, Mead’s concept of emergence provides a solid locuspoint for change at all levels of social life (Chang 2004). As such, emergence ought tobe isolated and examined conceptually and methodologically, with awareness of theways that people as agents enter into this process. However, agency, as a prime moverof change, is found not only in people’s capacities for spontaneity with the reflectiveprocess but also emerges through activity and in the course of community interchange.Blumer (1969) recognizes that the social act always contains in it the potential forsocial change in his consideration of the “transitive” qualities of social interaction; whatemerges in discourse between actors is formative and thus impossible to fully predictfrom previous social conditions. Building on these thinkers, agency is conceptualizedas a locus point of change that cannot be comprehended on an individual level butfinds its way into the actions and interactions of the social collective.

Only by examining and analyzing the specific “contact points” of structural changein people’s life-worlds through thoroughgoing attentiveness to human group life “inthe making” can we begin to see how people engage or take structural elements intoaccount and how these elements constitute part of the ongoing social process(Maines and Morrione 2001). Even here, these developmental, emergent structuralelements ought to be considered contributory causes to the more immediate moving(causal) forces of the deliberative process surrounding social action. Thus, we con-tend it is deliberative activity, developed within the context of emergent processes of



human group life, that is most central to notions of causality and related explanatorymodels in the social sciences. Rather than subscribing to the more simplistic, oftenossified notion of structure that one finds in the social sciences, we must rememberthat structures achieve a consequential existence only insofar as they represent loci ofactivity.

We do not deny the great many “arrangements” people have developed over themillennia or the potency of these matters for enabling successive generations of peo-ple to frame things in predictable ways. However, we need to be much more exact indefining what we mean by terms such as “structure,” “culture,” “power,” and the like.Instead of waving these concepts around like flags that are thought to establishdominion over all human matters, we need to examine how people accomplish activ-ity in all of its known and enacted facets. We need to take activity apart, piece bypiece, both in more immediate situated terms and as it is embedded in historical anddevelopmental flows, to see exactly how people enter into the causal process andmediate emergent events in all of the manifestations of human group life. HerbertBlumer’s legacy reminds us to start with the examination of actual instances of socialaction, to investigate in more systematic and sustained terms how people encounter,adjust to, and reshape the ongoing flows of community life.

Antony J. Puddephatt is a postdoctoral fellow in the Department of Science andTechnology Studies at Cornell University. His areas of interest include sociologicaltheory, the sociology of culture, and the sociology of science and technology.

Robert Prus is a professor of sociology at the University of Waterloo. A symbolicinteractionist and pragmatist ethnographer, he intends to connect social theory withthe study of human action in a direct, experientially engaged (community life-world)sense. He has written several books on the ways people make sense of and engage thelife-worlds in which they find themselves. These include Road Hustler with C.R.D.Sharper; Hookers, Rounders, and Desk Clerks with Styllianoss Irini; Making Sales;Pursuing Customers; Symbolic Interaction and Ethnographic Research; SubculturalMosaics and Intersubjective Realities; Beyond the Power Mystique; and The DeviantMystique with Scott Grills. At present, Robert Prus is tracing the developmental flowsof pragmatist thought from the classical Greek era (c.700–300 BCE) to the presenttime. This involves a number of areas of western social thought—including rhetoric,poetics, religious studies, ethnohistory, education, politics, and philosophy.

NOTES1. It might be noted that Glaser and Strauss (1967) never ignored the quest for causality, asthey argued that by using the “constant comparative method,” variations in the group ordependent variable can be accounted for with recourse to conditional factors. Later work byStrauss and Corbin (1998) considers the causal linkages that occur in the “conditional matrix”of interactive responses from the group to changing structural conditions as they unfold as ele-ments of the ongoing group process. Dey (1999) builds on Strauss and Corbin’s formulations,



and adds that causal conditions might be usefully differentiated as those that are necessary, suf-ficient, or contingent to behavioral effects. Archer (2003) argues that when social actorsencounter elements of social structure directly, they enter into the “internal dialogue” vis-à-visthe pragmatist notion of the “me.” This plays out in the psyches of actors and collectives inthe deliberative process and serves to orient further social action that establishes or maintainsstructural conditions. 2. We do not deny the value of quantitative research for describing broad societal conditionsor for providing indications of patterns and variations among categories and groupings.Clearly, survey materials can be useful in a wider array of policy contexts where one tries todescribe and anticipate generalized patterns and trends (see Ulmer and Wilson, 2003). Still,we hope to show that the notion of causality is more authentically located within the theoret-ical and methodological impetus of symbolic interactionism and ethnographic research. 3. Contrary to this, Maines (2003: 11) argues that work by W. I. Thomas (1927), RobertPark (1925), and later, Irwin Deutscher (1973) as well as the Durkheimian aspects of ErvingGoffman (1967), demonstrate that the roots of the interactionist paradigm did not ignorethe impact of situational structural conditions, nor did Blumer (1990) ignore the impor-tance of structure. Although Maines is correct that the early interactionists did not all ignorestructural relations or causality in their analyses, there is no question that Blumer (1969)forged the interactionist paradigm in large measure in opposition to the more dominantstructuralist approaches that would attempt to reduce the study of human life to variableanalysis and quantitative measures of the presumed effects of social structure to attitudesand values.4. As stated in Physics, especially book II, 194b–196a; and Metaphysics, Book I: 980a–983b;Book V: 1013a–1014a.5. It is this more spontaneous element of minded behavior that would provide the Meadianreply to Dennis Wrong’s (1961) critique of the “oversocialized conception of man” in sociol-ogy. As fundamentally social as Mead envisions the human self, he recognizes that the humanalso is a biologic agent. As a result, the human self is never fully socialized, and constantlyengages the more impulsive, emotional, active aspects of the psyche or human life-energy.

REFERENCESArcher, Margaret S. 2003. Structure, Agency, and the Internal Conversation. Cambridge, UK:

Cambridge University Press.Aristotle. 1984. The Complete Works of Aristotle: The Revised Oxford Translation, edited by

Jonathan Barnes. Princeton, NJ: Princeton University Press.———. 1926. The Nicomachean Ethics. Translated by Harris Rackham. Cambridge, MA:

Harvard University Press.Arjas, Elja 2001. “Causal Analysis and Statistics: A Social Sciences Perspective.” European

Sociological Review 17: 59–64.Berk, Richard. 1988. “Causal Inference for Sociological Data.” Pp. 155–172 in Handbook of

Sociology, edited by N. Smelser. Newbury Park, CA: Sage.Blumer, Herbert. 1969. Symbolic Interactionism: Perspective and Method. Berkeley: University

of California Press.———. 1971. “Social Problems as a Collective Behavior.” Social Problems 1: 298–306.———. 1990. Industrialization as an Agent of Social Change. Hawthorne, NY: Aldine de

Gruyter.



Britt, David. 1997. A Conceptual Introduction to Modeling: Qualitative and QuantitativePerspectives. Mahwah, NJ: Lawrence Erlbaum.

Chang, Johannes Han-Yin. 2004. “Mead’s Theory of Emergence as a Framework for MultilevelSociological Inquiry,” Symbolic Interaction 27: 405–427.

Cooper, John M. 2000. “Plato’s Theory of Human Motivation.” Pp. 668–688 in Plato, editedby Gail Fine. Oxford, UK: Oxford University Press.

Cox, D. R. and Nanny Wermuth. 2001. “Some Statistical Aspects of Causality.” EuropeanSociological Review 17: 65–74.

Demerath, Loren. 2002. “Causal Theory vs. Descriptive Theory.” Perspectives 25: 5–6.Denzin, Norman K. and Yvonna S. Lincoln 2000. Handbook of Qualitative Research.

Thousand Oaks, CA: Sage.Deutscher, Irwin. 1973. What We Say/What We Do. Glenview, IL: Scott Foresman.Dey, Ian. 1999. Grounding Grounded Theory: Guidelines for Qualitative Inquiry. San Diego,

CA: Academic Press.Ferraro, K. 1983. “Negotiating Trouble in a Battered Women’s Shelter.” Urban Life 12:

287–306.Fine, Gary Alan. 1993. “The Sad Demise, Mysterious Disappearance, and Glorious Triumph

of Symbolic Interactionism.” Annual Review of Sociology 19: 61–87.Glaser, Barney and Anselm Strauss. 1967. The Discovery of Grounded Theory. Chicago, IL:

Aldine.Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Doubleday.———. 1967. Interaction Ritual: Essays in Face-to-Face Behavior. Chicago, IL: Aldine.Goldthorpe, John H. 2001. “Causation, Statistics, and Sociology.” European Sociological

Review 17: 1–20.Gotschalk, D. W. 1941. “Causality and Emergence.” Philosophical Review 51: 397–405.Grenier, Marc. 1992. “Review Essay: Social Order and the Public Philosophy—An Analysis

and Interpretation of the Work of Herbert Blumer.” Canadian Journal of Sociology 17:429–440.

Huber, Joan. 1973. “Symbolic Interaction as Pragmatic Perspective: The Bias of EmergentTheory.” American Sociological Review 38: 274–284.

Joas, Hans. 1996. The Creativity of Action. Chicago, IL: University of Chicago Press.Kolb, William. 1944. “A Critical Evaluation of Mead’s ‘I’ and ‘Me’ Concepts.” Social Forces

22: 291–296.Lazarsfeld, Paul F. 1959. “Problems in Methodology.” Pp. 39–78 in Sociology Today, edited by

Robert Merton. New York: BasicBooks.Lewis, J. David. 1979. “A Social Behaviorist Interpretation of the Meadian I.” American

Journal of Sociology 85: 261–287.Lindesmith, Alfred R. 1981. “Symbolic Interaction and Causality.” Symbolic Interaction 4:

87–96.Lofland, John. 1976. Doing Social Life. New York: Wiley.———. 1995. “Analytic Ethnography: Features, Failings, and Futures.” Journal of

Contemporary Ethnography 24: 30–67.Maines, David. 1977. “Social Organization and Social Structure in Symbolic Interactionist

Thought.” Annual Review of Sociology 3: 235–259.———. 1988. “Myth, Text, and Interactionist Complicity in the Neglect of Blumer’s

Macrosociology.” Symbolic Interaction 11: 43–57.———. 2001. The Faultline of Consciousness: A View of Interactionism in Sociology. New York:

Aldine de Gruyter.



———. 2003. “Interactionism’s Place.” Symbolic Interaction 26: 5–18.Maines, David and Thomas Morrione. 2001. “Herbert Blumer’s Theory of Industrialization

and Social Change.” Pp. 55–67 in The Faultline of Consciousness: A View of Interactionismin Sociology, by D. Maines. New York: Aldine de Gruyter.

McKim, Vaughn and Stephen Turner. 1997. Causality in Crisis? Statistical Methods and theSearch for Causal Knowledge in the Social Sciences. Notre Dame, IN: University of NotreDame Press.

Mead, George Herbert. 1932. Philosophy of the Present. Chicago, IL: University of Chicago Press.———. 1934. Mind, Self, and Society from the Standpoint of a Social Behaviorist. Chicago, IL:

University of Chicago Press.———. 1938. Philosophy of the Act. Chicago, IL: University of Chicago Press.Morrione, Thomas J. 1988. “Herbert G. Blumer (1900–1987): A Legacy of Concepts,

Criticisms, and Contributions.” Symbolic Interaction 11: 1–12.Mullins, Nicholas. 1973. Theories and Theory Groups in Contemporary American Sociology.

New York: Harper and Row.Park, Robert Ezra. 1925. The City. Chicago, IL: University of Chicago Press.Plato. 1997. Complete Works, edited by John M. Cooper. Indianapolis, IN: Hackett.Prus, Robert. 1996. Symbolic Interaction and Ethnographic Research: Intersubjectivity and the

Study of Human Lived Experience. Albany: State University of New York Press.———. 1999. Beyond the Power Mystique: Power as Intersubjective Accomplishment. Albany:

State University of New York Press.———. 2003. “Ancient Precursors.” Pp. 19–38 in Handbook of Symbolic Interactionism,

edited by Larry Reynolds and Nancy J. Herman-Kinney. Walnut Creek, CA: Altamira.———. 2004. “Symbolic Interaction and Classical Greek Scholarship: Conceptual

Foundations, Historical Continuities, and Transcontextual Relevancies.” AmericanSociologist 35: 5–33.

———. 2005. “Aristotle’s Nicomachean Ethics: Laying the Foundations for a PragmatistConsideration of Human Knowing and Acting.” Paper presented at the joint SymbolicInteraction and Ethnographic Research and North Central Sociological Association con-ference, April 7–10, Pittsburgh, Pennsylvania.

Prus, Robert and Styllianos Irini. 1980. Hookers, Rounders, and Desk Clerks: The SocialOrganization of the Hotel Community. Toronto, ON: Gage.

Puddephatt, Antony J. 2005. “Mead Has Never Been Modern: Using Meadian Theory toExtend the Constructionist Study of Technology.” Social Epistemology 19: 357–380.

Reynolds, Larry T. and Nancy J. Herman-Kinney, eds. 2003. Handbook of SymbolicInteractionism. Walnut Creek, CA: Altamira.

Simmel, Georg. 1978. The Philosophy of Money, translated by Tom Bottomore and DavidFrisby. London: Routledge and Kegan Paul.

Spangler, Sister Mary Michael. 1998. Aristotle on Teaching. Lanham, MD: University Press ofAmerica.

Strauss, Anselm. 1993. Continual Permutations of Action. New York: Aldine de Gruyter.Strauss, Anselm and Juliet Corbin. 1998. Basics of Qualitative Research: Techniques and

Procedures for Developing Grounded Theory. Thousand Oaks, CA: Sage.Taylor, Steven J. and Robert Bogdan. 1984. Introduction to Qualitative Research Methods: The

Search for Meanings. New York: John Wiley and Sons.Thomas, W. I. 1927. “Situational Analysis: The Behavioral Pattern and the Situation.”

Pp. 154–167 in W. I. Thomas on Social Organization and Social Personality, edited by M.Janowitz. Chicago, IL: University of Chicago Press.



Thomas, W. I. and Florian Znaniecki. 1918. The Polish Peasant in Europe and America.Boston, MA: Badger.

Ulmer, Jeffrey and Mindy S. Wilson. 2003. “The Potential Contributions of QuantitativeResearch to Symbolic Interactionism.” Symbolic Interactionism 26: 531–522.

Wiseman, Jacqueline. 1970. Stations of the Lost: The Treatment of Skid Row Alcoholics.Englewood Cliffs, NJ: Prentice-Hall.

Wrong, Dennis. 1961. “The Oversocialized Conception of Man in Modern Sociology.”American Sociological Review 26: 183–193.

Zeitlin, Irving. 1973. Rethinking Sociology: A Critique of Contemporary Theory. New York:Appleton Century Crofts.



Although Blumer asserts that to deny the existence of “structure” in humansociety is “ridiculous,” just such a denial has commonly been attributed tohim. The more conventional mainstream understanding of structure insociology, however, is theoretically incoherent, as demonstrated by classicand modern studies of, for example, stratification. Blumer’s sociology isshown, with particular reference to its bases in the pragmatist tradition, toprovide an alternative understanding of structure that is both theoreticallycoherent and capable of empirical investigation. Furthermore, it is capableof dissolving the dilemma of structure and agency in contemporary socio-logical theory.

In his remarks on the implications for sociology of the thought of George HerbertMead, Herbert Blumer emphasizes the point that, although Mead describes the socialorder as the outcome of collaborative “joint action,” such a position does not entail adenial of “the existence of structure in human society. Such a position would beridiculous” (Blumer 1969: 75). Blumer’s critics would disagree, arguing that symbolicinteractionism “prevents the understanding of social structures and their constrainingcharacteristics or of patterns of human organization such as class hierarchies or powerconstellations” (Coser 1976: 157). Even apparently sympathetic commentators areprepared to accept that the perspective suffers from an “astructural bias,” and displaysan “unconcern with social structure” (Meltzer et al. 1975: 113). More recently,Musolf (1992) has also accepted this criticism, arguing that only new directions ininteractionist thought will allow the perspective to address the macrosociological con-cerns of power, inequality, and social structure. Ridiculous or not, then, the notionthat Blumer denied the existence of structure has become widespread, from the pub-lication of Symbolic Interactionism (Blumer 1969) to the present day (see, for exam-ple, the discussions in Denzin 1992: 56; Gouldner 1970: 379; Maines 1977: 236;Morrione 2003: xiv; Sauder 2005: 286).

Our purpose here is to question the validity of this criticism, not only because itmisrepresents Blumer’s sociological position—although many critics do present a

Symbolic Interactionism and the Concept of Social Structure

Alex Dennis*University of Salford

Peter J. MartinUniversity of Manchester

*Communications should be sent to Alex Dennis, School of English, Sociology, Politics, andContemporary History, University of Salford, Salford, M5 4WT, United Kingdom. We aregrateful to participants at the 2005 Society for the Study of Symbolic Interaction annualmeetings for their comments on an earlier draft of this paper and to the anonymous review-ers for their criticisms.


curiously myopic version of this but also because it fails to acknowledge symbolicinteractionism’s status as a distinct and coherent alternative to more orthodox formsof sociological thought. At the risk of tautology, we will argue that the very concep-tion of “social structure” is a product of, and embedded in, orthodox “structural”sociology—the same orthodoxy to which Blumer’s work stands as a principled objec-tion. This conception cannot be transported from one perspective to another, becauseits meaning differs between the different theoretical traditions. Moreover, althoughsymbolic interactionism has indeed developed an approach for the analysis of pat-terned social organization, including hierarchical differentiation and asymmetries ofpower (see, for example, Hall 2003), interactionists have good reasons to regard “socialstructure” as a problematic concept.

Just as Mead concluded that “metaphysical problems were unnecessary ‘riddles’created by dualistic philosophies” (Baldwin 1986: 24), Blumer’s theoretical work aimsto overcome various problematic dualisms, among them the fruitless oppositionbetween realism and idealism in sociology. The concept of social structure is an exem-plary case of the former in three ways. First, it requires a reification of social processes,so that (as in Durkheim 1982) they become “things,” external to real people. Second,it facilitates a portrayal of social life as static rather than processual, understoodthrough categories with largely stable relationships instead of through social actorsactually doing things. Third, its analytic utility depends on its being seen as thecause of human behavior: it is necessarily deterministic in application. In short,then, the structural conception of sociology takes society to be an external systemthat is the “overall determinant of social action” (Blumer 1969: 74).

Often neglected in the secondary treatments of Blumer, however, is the fact thatBlumer was also concerned to show that an exclusive focus on the individual actorcould not form the basis of a theoretically satisfying sociological alternative: his rejec-tions of psychological reductionism, subjectivism, and solipsism were equally forth-right (Morrione 2004: xi). Indeed, one of Blumer’s neatest arguments is that structuralsociology is irredeemably subjective, as it inevitably requires analysts to impose theirown definitions of reality onto the social world they are investigating (Blumer 1969:74). Blumer stresses that such a position ignores or denies the “obdurate character” ofthe empirical world, which can “talk back in the sense of challenging and resisting, ornot bending to, our images or conceptions of it” (Blumer 1969: 22).

Blumer’s opposition to dualisms, then, should not be mistaken for a desire to showthat one side of a dualistic relationship (in this case, a commitment to Durkheimianstructural sociology) could be dissolved into the other (here, an individualistic or sub-jectivist position). Rather, his aim is to develop a sociological perspective that tran-scends the sterility of the realist—idealist or objective—subjective dualisms. In his ownwords, “Mead was a pragmatist in philosophical stance, and so am I. . . . [F]or prag-matism, reality does not exist in consciousness, nor is the reality eternally real, inde-pendent of human experience with it” (Blumer quoted in Morrione 2004: xii). What,then, is this reality that Blumer regards as the “obdurate character” of the social world?

Blumer’s answer cannot be clearer: “the empirical world of our discipline is the nat-ural social world of everyday experience” (Blumer 1969: 148). As Morrione puts it,



“Blumer located action and the creation of meaning in an ever-emergent situated presentthat is cognitively, behaviorally, and intersubjectively inseparable from the past andfuture” (Morrione 2004: xii, emphases in original). In these formulations, and in hisview of the social order as a situated accomplishment, Blumer echoes the Schutzianphenomenological tradition evident in such contemporaneous works as Berger andLuckmann’s (1966) The Social Construction of Reality and Garfinkel’s (1967) Studiesin Ethnomethodology. Unlike these approaches, however, Blumer’s focus on “thesocial world of everyday experience” has clear foundations in pragmatist philosophy.It is our intention to show that these foundations provided Blumer with the tools todevelop a coherent and distinct mode of sociological thought and inquiry that isincompatible with such notions as social structure and variable analysis as they areconventionally understood (Abbott 1997), and that represents a viable alternativeapproach to the dominant sociological orthodoxies. We will make this demonstra-tion with particular reference to the concept of social structure—a topic, it will bereadily recalled, that symbolic interactionism has often been regarded as “incapable” ofaddressing.

BLUMER AND THE PRAGMATIST ORIENTATION OF THE CHICAGO SCHOOL

Sociological classics tend to be reread through the lens of contemporary disciplinarydevelopments, and, because of its current theoretical centrality, structure/agencydualism tends to provide the means by which Blumer’s thought is understood.Commentators overwhelmingly locate symbolic interactionism at the agency end ofthis dualism (Dennis and Martin 2005). Our aim in this section is to question suchan attribution and to offer a preliminary clarification of what we understand symbolicinteractionism to be.

Blumer’s thought emerged from a radically different approach to doing sociology(usually glossed as the “Chicago School”) from that which currently commands main-stream recognition. Fundamentally empiricist in method, this approach stressed thepragmatist arguments of James, Dewey, and (particularly) Mead as epistemologicaltouchstones and methodological foundations. Although notions of structure andagency could be extracted from the Chicago School’s works, such an extraction wouldrepresent a redescription of those works in contemporary terms rather than a reasonablereconstruction of their authors’ own positions. In his exemplary excavation of Mead’sviews on what society might be, then, Athens (2005) is careful to demonstrate that themeaning of institution in Mead’s thought, common maxims (Athens 2005: 307), issimilar but irreducible to the same concept as used by structuralist sociologists.

What is distinctive about the Chicago approach is, as Abbott (1997: 1152) putsit, the presumption “that one cannot understand social life without understandingthe arrangements of particular social actors in particular social times and places.”This presumption is most corrosive to those forms of sociology based on the analysisof variables: the idea of context-independent Durkheimian social facts, capable of

SYMBOLIC INTERACTIONISM AND THE CONCEPT OF SOCIAL STRUCTURE 289


allowing different settings to be compared, contrasted, and organized on continua,is utterly alien to the Chicago tradition. Although facts can be extracted from set-tings, they remain the products of those settings, and the social property of partici-pants to their production—including those facts, like anomie, alienation, astructuralbias, or the masculine gaze, generated in the settings of professional sociologists.Settings are not, of course, hermetically sealed off from one another, but their rela-tionships require discovery and are not capable of being specified a priori. But, asAbbott (1997) points out, sociology remains committed to just the kinds of variableanalysis Blumer (1956) sought to undermine, his critique producing what has beencalled “one of the most telling 35-year silences in the recent history of academic life”(Watson 1995: 317).

Blumer’s critique of variable analysis did not come out of nowhere: it is a contri-bution to, and development of, the Chicago tradition’s emphasis on the specificity ofsituations. Although Blumer was often critical of this tradition’s orientations andorganization (see, in particular, Blumer 1939), the Chicago School provided both anintellectual home and a critical context within which he could pursue his method-ological and conceptual arguments. Although much has been written about the divi-sions between Blumer and other Chicago sociologists (see, for instance, Becker 1999;Abbott and Gaziano 1995), to the extent that there is a “Chicago School,” it is onethat shares many of the same characteristics and concerns of symbolic interactionism.These characteristics might very generally be characterized as a militantly empiricalsociological practice informed by pragmatist philosophical underpinnings. For thepurposes of our argument, we will focus on three themes that illustrate these charac-teristics: (1) that institutions are the products of interactions; (2) that sociologicalstudies are descriptions of the formal features of such institutions; and (3) that sta-bility and change are context-dependent. Wirth’s (1928) The Ghetto will be used toillustrate their application.

Institutions Are the Products of Interactions

Blumer drew on the Chicago School’s orientation to a fine-grained analysis of every-day activities. Although this did not preclude statistical or other larger-scale forms ofanalysis, it provided a background where some of the claims subsequently enunciatedon their behalf were treated with suspicion: institutions, organizations, classes, reli-gions, and other structural aspects of the social world were treated as the product ofpeople’s activities rather than the unquestioned bases for sociological investigations.This instantiates a key concept of pragmatist philosophy: rather than unquestioninglyaccepting or peremptorily denying the existence of a phenomenon (in this case context-independent features of social situations), that phenomenon becomes something to bediscovered, located empirically through processes of description and comparison.

Thus, in Wirth’s (1928) work, a particular synagogue was shown to be the prod-uct of the elaboration of secular, traditional (usually first-generation immigrant), and“Americanized” forms of Judaism. Its relative stability is not taken for granted but isused to say something about the nature of religious institutions at times of socialchange, and how political negotiations between different ethnic communities resolve



themselves through negotiation, compromise, and schism. For example, conservativecongregations were forced to postpone the Friday evening service from sundown untilafter supper to accommodate urban American working practices. Although such con-gregations professed to be orthodox, they were regarded as relatively secular by thestrictly orthodox on the one hand and relatively conservative by the reformed con-gregations on the other. What Wirth brings to our attention is the range of practicesthat count for members of a community as a congregation: having the status of a syn-agogue is not something that rests on either adherence to Judaic custom or the abil-ity to maintain a large and active community of believers. Rather, it is somethingthat is constantly worked out between groups of interested parties, more or less totheir satisfaction, with new circumstances and interested parties always having to betaken into consideration. Most important of all, it is absolutely not something thatcan be determined a priori by a sociologist or other disinterested observer.

Sociology Is the Description of the Formal Features of Institutions

Blumer shared with earlier Chicago scholars an orientation to formalism as a meansof organizing and presenting sociological findings (see Park and Burgess 1921 on therole of Simmel in Chicago sociology). Rather than seeking generalizations that canbe applied to larger and larger groups, an emphasis was placed on the ways in whichinstitutional features of one setting might reflect those of another. Perhaps the clear-est statement of this position is to be found in Hughes (1951: 320): “The compar-ative student of man’s work learns about doctors by studying plumbers; and aboutprostitutes by studying psychiatrists.” Instead of reducing the study of different jobsto the sociology of work, Hughes evinces the Chicago desire to find out what dif-ferent jobs have in common and how they differ—and, by asking questions of thiskind, finding out what the formal features of those different jobs might be in the firstplace.

This reflects the pragmatists’ rejection of neo-Platonism on the one hand and whatMills (1959) called “abstracted empiricism” on the other. There is no reason to believethat social phenomena necessarily have essential features in common, that these fea-tures could be specified on an a priori basis, or that they could be unproblematicallyapplied in different cases as a theoretical framework for understanding somethingnovel. Social facts cannot be assumed. This, however, does not imply a rejection of thenotion that different settings might have features in common, or, indeed, that such fea-tures could be used as sensitizing concepts, or spurs to the imagination, in other inves-tigations (Blumer 1954: 7). This focus on empirical investigation furnished the ChicagoSchool with the analytic resources to discover formal features of the social world.

In Wirth’s (1928) account of ghetto life, one such formal feature is community sol-idarity, mediated by the synagogue. “[D]eep bonds of sympathy” emerge through“colorful ritual,” thus providing “a place of dignity” for an individual who would bea “mere Jew” in the outside world. The centrality of the synagogue depends cruciallyon its capacity to provide families and households with the status they deserve, basedon their “learning, piety, the purity of family life, and services rendered to the com-munity,” rather than on the basis of their wealth. In turn, the community as a whole



“acquired a reputation . . . through its outstanding personalities, particularly throughits philanthropists and scholars” (Wirth 1928: 37). The synagogue is not, then, justan instance of a religious institution, or even an instance of a synagogue like anyother. What Wirth emphasizes is the ways in which the institution in this case is usedby members of the ghetto community to satisfy certain basic human needs that, giventheir circumstances, cannot be fulfilled elsewhere. Their marginal status means thattheir capacity to be afforded respect by others, and gain a reputation for their actionsrather than their ethnic and/or religious identity, cannot be taken for granted in theways members of other communities might legitimately expect. Formally, then, thesynagogue does not represent that which makes Jews Jewish, but rather allows themto maintain their status as human beings as well as Jews and so—over time—reducethe extent to which they are treated as different. Ironically, the very institution whichis at the heart of their religious beliefs becomes the thing that facilitates greater inte-gration with Gentile culture. The formal features of the synagogue, viewed from apragmatic Chicago School position, prove to almost reverse its institutional features,viewed from the perspective of structuralist sociology.

Stability and Change Are Context-Dependent

Notions of structural integrity and system maintenance were treated with analyticsuspicion by Blumer and his earlier Chicago counterparts: “stable and recurrent formsof joint action do not carry on automatically in their fixed form but have to be sus-tained by the meanings that people attach to the type of situation in which the jointaction reoccurs” (Blumer 1969: 59). Inertia alone will not maintain an orderly socialsituation, and one cannot treat such a situation as self-maintaining independently ofthe interpretative work of its members. Thus, just what something is—an institution,a class, an ethnic group, a religious movement—depends on the meanings people giveit and take from it. For example, an apparently context-independent feature ofhuman life is the nature of intergroup relations. “All societies of any great size havein-groups and out-groups; in fact, one of the best ways of describing a society is toconsider it a network of smaller and larger in-groups and out-groups. And an in-group is only one because there are out-groups” (Hughes 1962: 8). The formal fea-tures of a social institution, then, are the outcome of interactional work thatdetermines what the boundaries of that institution are, who is “inside” and who is“outside,” and how individuals can move between the two categories.

The pragmatist roots of this approach should by now be evident: rather than treat-ing stable social phenomena as the unquestioned starting points for analysis, or bydenying their stability altogether, the ways in which those phenomena are constitutedas stable becomes a topic of investigation. Neither stability nor change is assumed tobe a natural characteristic of human life or social phenomena, but both are to beregarded as the outcomes of meaningful action. In his careful critique of the conceptof industrialization, Blumer (1990) takes this concern to its logical conclusion bydemonstrating the radical incoherence of the idea that industrialization or its effects canbe treated as consistent or analytically coherent between settings—and, indeed, that theconcept of setting is itself one that requires further elaboration to be analytically useful.



The idea of a stable set of institutions being driven from one form to another by ahomogeneous set of social forces is revealed as a myth.

For Wirth, these concerns map literally onto the geography of the ghetto area.Individuals’ social and religious commitments shape where they choose to live andwork, and their domestic and vocational situations shape their outlooks. “Social typesrange themselves in constellations, each stellar figure with its little circle of satellitesseeking its place in the life of the group and changing its position and character as theculture of the area is transformed” (Wirth 1928: 250). But just as the ghetto inhabi-tants and their beliefs are transformed in the areas of “second settlement” (those areasinhabited by “Americanized” Jews), elsewhere they remain static and traditional: thedaily routine of the “ghetto Jew” “is confined largely to the narrow area of his imme-diate vicinity. Even when he drives his wagon through the other sections of the city,he does so with his eyes closed to the life that goes on” (Wirth 1928: 251). In short,the institution of Judaism is challenged both by internal differences and by externalpressures, and, self-consciously, members of the community locate and maintain them-selves in particular, but gradually changing, positions as participants and creators ofthis institution.

Blumer’s intellectual development is not, therefore, merely reducible to a commit-ment to ethnographic method or a suspicion of theory—or, indeed, to a continuousreinterpretation and analysis of Mead’s “social behaviorism.” Many of the key tropes inhis argument are just those of earlier Chicago sociologists, and, we will argue, his workis best understood with this in mind. Central to these arguments are a pragmatic rejec-tion of dualisms, in particular a rejection of the notion that one must either accept ordeny the existence of social structure as an a priori theoretical matter, and a commit-ment to finding out through investigation how orderly social institutions are producedand maintained. One of Blumer’s great contributions to sociological thought was todemonstrate the dangers of not maintaining such a tentative position with respect totheory: “Theoretical positions are held tenaciously, the concepts and beliefs in one’sfield are gratuitously accepted as inherently true, and the canons of scientific procedureare sacrosanct. It is not surprising, consequently, that the images that stem from thesesources control the inquiry and shape the picture of the sphere of life under study. Inplace of being tested and modified by firsthand acquaintance with the sphere of lifethey become a substitute for such acquaintance” (Blumer 1969: 37).

STUDIES OF SOCIAL STRATIFICATIONIn accordance with the pragmatist position, then, the Chicago School’s approach tothe concept of structure takes it as a possibly useful redescription of the ongoing flow ofhuman activities, rather than an ontological feature of the social world itself. Structureis (sometimes) a useful placeholder, a way of illustrating how the pattern of socialorganization is perpetuated and reproduced, but not that pattern itself—because sucha pattern is inevitably a way of facilitating sociological description, not somethingintrinsic to the setting in its own right. People orientate to different things, move out



of and into geographical areas and social circles, reinforce and change their institu-tional affiliations, and so on. In the course of their doing this the structure of a socialworld changes or remains the same—but that structure is not something that inheresin the social world but is an artifact of the questions we ask of it. It is a part of humanactivities in just these ways and no others—and to treat it as a real thing, somethingthat can be related to other real things, is to conflate a sociological description withthe thing being described. In this section we will suggest that this distinction is clearlyexemplified by the ways in which conventional sociologists have understood anddescribed social inequality and stratification. In doing so our aim is to substantiateBlumer’s point that what is claimed to be an “‘objective’ approach holds the dangerof the observer substituting his view of the field of action for the view held by theactor” (Blumer 1969: 74).

Sociological descriptions of societies in terms of a hierarchy of strata, or as a struc-ture of classes, derive from an understandable wish to represent societies as wholes,and to present the big picture of class inequalities as institutionalized features of suchsocieties. The problems of such approaches are well documented, from disputesabout what classes consist of (Martin 1987) to arguments over the utility of differ-ent methodological and measuring techniques (Crompton 1993). From a pragma-tist position, however, one would have to argue that such approaches cannot achievethe degrees of neutrality and objectivity to which they aspire. Such approachesinevitably fail to be literal descriptions of the social structure as they are, and have tobe selective representations that depend on the particular interests—practical andtheoretical—of researchers, and those researchers’ a priori theoretical and method-ological commitments (Scheff 1995). Putting it bluntly, the ways in which the classstructure has been represented in most sociological studies is an artifact of thosestudies’ theoretical and methodological preconceptions: “all class schemes are socialconstructs, or rather, the constructs of sociologists. Therefore different class schemes,when applied to the same occupational structure, can produce quite different ‘classmaps’” (Crompton 1993: 50).

The concept of social class is itself a source of theoretical problems. Indeed,although most European researchers have accepted the validity of the term, the dom-inant American approach has been to conceptualize social inequalities in terms of sta-tus (until challenged by neo-Marxian authors such as Bowles and Gintis, 1976, andWright 1978). The very existence of such a divergent approach suggests that the def-inition of social classes, and the assignment of people to them, is not a neutral orobjective process, but one in which theoretical perceptions of relevance are activelyinvolved. Even among those theorists who accept the validity of the concept of class,major disputes exist concerning where boundary lines between class strata can bedrawn. In the classic Marxian formulation, a tiny bourgeoisie comes to confront ahuge proletariat, and some authors (for example Braverman 1974) continued to arguefor this view. Others, however, attempted to incorporate an expanding middle classinto their models: for Poulantzas (1975), for example, the real working class has beenreduced to about a fifth of the overall working population, whereas for Wright (1985:45) those in the middle class occupy a contradictory class position.



Other sociologists have developed models of class that owe more to Weber thanto Marx, regarding the formation of social classes as an empirically open question,and seeking to identify the structural conditions under which real classes actuallycome into existence—or fail to do so (see, for example, Goldthorpe 1980). Inretaining the Marxian priority accorded to economic relations in the class structure,however, and in basing their analyses on the occupations of individual people, suchanalysts have been accused of neglecting the other factors Weber (1948) thoughtwere important in social stratification—in particular ethnicity, nationality, and reli-gion. Additionally, Garnsey (1978) argued that the conventional approach to classanalysis is “male-centered” both because it presupposes that women’s class positiondepends on that of husbands or fathers, and because it assumes that women’sdomestic labor is marginal to the system of production. By excluding women, theyoung, and the old, and by its inability to incorporate families, then, research inthis tradition yields a picture of social life that is hard to relate to any real, or pos-sible, societies. Similar points could be made about the neglect of, among otherthings, ethnic or national identification and their effects, or political organizationsand culture (Crompton 1993: 77). Inevitably, there have been more recent effortsto incorporate women and ethnic groups into studies of class and social mobility,but none has been successful—and it is hard to see how any could succeed: the ideaof producing a multidimensional model of the social structure cannot be realized.Our point here is that class is essentially a commonsense, rather than a scientific,concept (Schutz 1962). As such it “cannot simply be taken for granted as the basisof sociological analysis which aspires to be scientific, for the ‘objective’ researchends up simply reflecting the subjective preferences of the researcher” (Martin1987: 95).

Attempts to investigate the class structure have encountered equally intractablemethodological problems. Many of these concern the validity and reliability of datagenerated by sample surveys in which individual people are interviewed or sent ques-tionnaires about their occupational experiences. By restricting what counts as class towhat happens to people at work, and by treating those people as isolated experiencersof their structural position rather than members of social networks and other groups,such studies tend to reinforce the gap between individual experience and the socialstructure. Somehow what happens to individuals has to be reconciled with their classpositions—but how this is actually achieved in practice depends on the skillful use oftheory to bridge the gap. Because more than one theoretical reconciliation is possible,in each case the researcher has to select among alternative accounts to provide whatcan be construed as an objective account. Such attempts illustrate the pragmatist viewof knowledge as contextual—in this case, dependent on the a priori assumptions ofthe analyst.

Such studies are also problematic insofar as they seek to determine the life (or atleast employment) experiences of all the members of a particular population. Oneproblem with such an approach is that, if the theoretical justification for the conceptof class is correct, a tiny proportion of that population will have a disproportionateeffect on the social structure as a whole. More attention, then, should be paid to



members of powerful and wealthy elites—but these are precisely the groups that aresignificantly underrepresented in such studies: indeed, in the Oxford Mobility Studyof the 1970s no data whatsoever could be collected on these sections of the popula-tion (Martin 1984: 33).

There are also concerns about how occupational positions are, in fact, related tosocial classes: theorists have been increasingly concerned by people in the routinenonmanual class category (those doing clerical and lower administrative work). Arethey members of the new middle class, or are they becoming proletarianized? In fact,the great majority of people in such occupations are female, and so excluded frommost studies of class. Moreover, it has been shown that, for men, the significance ofoccupying such a position is variable: for those whose origins are in manual work thismay represent promotion to an office job, but to younger men it may be only a step-ping stone on the way to a more senior managerial post (Stewart et al. 1980). Thesample survey thus cannot reveal that the significance of the same occupational posi-tion is very different for the different people occupying it, and thus it cannot be con-sidered to be the same thing as far as class location is concerned.

Studies of stratification, then, demonstrate that in order to bridge the gap betweentheoretical claims and research practice, sociological approaches must deal with largenumbers of people, constant uncertainties, and continual change. One response tothis problem is to allow the availability of methodological techniques to determinedata collection and analysis—as in Blau and Duncan’s (1967) attempt to get the bigpicture of a social system’s stratification. The result, unfortunately, is a description ofthe American “occupational structure” that could not possibly describe any imagina-ble human society, being restricted, as we have suggested, to a sample survey of menbetween ages twenty and sixty-four. Others have foreclosed the debate by theoreticalfiat (Cicourel 1964). Giddens, for instance, correctly observed that a rigorousWeberian approach, taking all the different factors that could influence patterns ofstratification into account, would lead to the identification of a “cumbersome plural-ity of classes” (Giddens 1981: 104). He therefore asserted that there are “only a lim-ited number of classes” (Giddens 1981: 106) in any society. Just why this is so isunclear. There is surely a “cumbersome plurality” of chemical components in a com-plex protein, but a biologist or pharmacologist could not use this fact to decide apriori that they can be divided into discrete categories to simplify analysis. On thecontrary, he or she would necessarily have to deal with contingencies, uncertainties,and the impossibility of conventional measurement until the details of the phenom-enon under investigation could be adequately described. In some respects things areeasier in social life, as Weber (1978: 15) pointed out: “We can accomplish somethingwhich is never obtainable in the natural sciences, namely the subjective understand-ing of the actions . . . of individuals.” Certainly a wider scope of phenomena is readilyobservable.

There are, of course, patterns of inequality—of wealth, power, prestige, and soon—that are the outcomes of the activities of individuals and groups as they pursuetheir perceived interests. It is likely that our understanding of the processes of socialstratification will be best furthered by the direct investigation of these activities in the



real world rather than through the heavy-handed imposition of concepts such as the“class structure.” We will return to this point subsequently.

THE IDEA OF INSTITUTIONAL ANALYSISIf the concerns raised above are legitimate—and, as we mentioned, we are unawareof any coherent responses to them—then a central concern for nonstructural soci-ologies must be to radically separate description from the things being described: toremain faithful to the phenomena under investigation. Such concerns are elaboratedby, and find a radical expression in, the “mature” Chicago School’s emphasis oninstitutional analysis. The concept of institution, of course, is usually seen (cer-tainly in sociology textbooks) as one denoting collectivities basic to the perspectiveof mainstream structural sociology, as when, for example, the family, the church,the nation-state, armies, or business corporations are described as institutions. Ourargument, however, is that a close examination of the concept and its usage revealsnot only some (normally ignored) problems for the structural perspective but alsothe distinctiveness of the alternative vision of sociological work that animates inter-actionist studies.

Alexander and Giesen’s (1987) characterization of Mead stands as a fairly typicalassessment of his ideas and influence. Although they defend Mead against the(common) misreading that his ideas lead to a “microsociology devoid of macrorefer-ence,” and emphasize the ways in which his analysis is “open to more collectivist con-cerns,” they conclude that “Mead lacked an institutional theory,” and that his followersneglected the possibility of this “collective link” so that “experience, not individuallymediated structure . . . became the hallmark of interactionist microanalysis” (Alexanderand Giesen 1987: 9–10). This reading presupposes that social life can only be con-ceptualized in terms of interactional (micro) or institutional (macro) levels, and thatthe basic theoretical problem for sociology must be to provide a link between bothsides of the great divide. In fact, almost a third of Mead’s most famous work is con-cerned with society, and an entire section devoted to “The Community and theInstitution” (Mead 1934: 260–273). What emerges from Mead’s account is not anunderstanding of social life as operating on different levels, but—particularly as a resultof the ability to “take the role of the other” and the ways in which the “generalizedother” enters into the thinking of individuals (Mead 1934: 256)—a perspective thatsees organized patterns of social life as constituted and reproduced through the regularinteractions of real people in particular times and places. In this, Mead emphasizes theimportance of the uniquely human capacity to communicate through the medium oflanguage and the infinite possibilities for symbolic representation it facilitates.

Given this general perspective, and Mead’s (1934: 262) insistence on the pro-foundly social sources of the individual’s self, it is somewhat perplexing that socio-logical work in the interactionist tradition has been characterized by some of itscritics as “subjective and voluntaristic” (Lichtman 1970: 77), and as “an atheoreticalsociological theory [sic] that refuses in principle to transcend the peculiar characteristics



of social processes in the here and now” (Coser 1976: 156). Indeed, there is a certainirony in the fact that Mead has also been criticized for presenting an oversocializedconception of human action, exaggerating the degree to which people are led toconform to the perceived expectation of the (internalized) “generalized other”(Katovich and Reese 1993). It is hard to reconcile this Mead with the one who hasbeen accused of ignoring “social structures and their constraining characteristics”(Coser 1976: 157). In fact, Mead’s analysis is an attempt to transcend a sterile andfalse subjective-objective dualism, and to propose a genuinely dialectical model ofhuman conduct. Thus, for instance, although institutionalized patterns of activity areessential components of any form of social organization, institutions themselves neednot “crush or blot out individuality, or discourage any distinctive or original expres-sions of thought or behavior. . . . On the contrary, they need to define the social, orsocially responsible, patterns of individual conduct only in a very broad sense” (Mead1934: 262). Mead’s focus, then, is neither on coercion nor unconstrained volun-tarism, but on how people define situations in ways they consider appropriate, andinteract with others on that basis. Athens’ (2005) recent exposition of the relation-ships between selves, institutions, and maxims in Mead’s thought is an excellent clar-ification of these issues.

With this in mind, it is clear why, in one specific sense, Alexander and Giesen(1987) are right to suggest that Mead lacks a theory of institutions. For Mead, andfor the interactionists who followed him, institutions are not entities existing on a sep-arate level from that of the individual, nor are they ontologically distinct emergentproperties of social interaction. Mead does not need a theory of institutions, as theyhave no privileged place in his theoretical framework. Instead, institutions—as par-ticular instances of all orderly patterns of social organization—are simply complexesof human interaction, chains of self-other relations in which processes of symboliccommunication are of fundamental importance. From this perspective, we candevelop an empirical approach to what Blumer (1969: 148) called “the natural socialworld of everyday experience.”

We wish to emphasize that a focus on how the intersubjective world is constitutedand sustained leads to an analytical approach in which the familiar concepts of main-stream sociology—such as structures, systems, classes, organizations, institutions, andso on—appear not as objective entities but rather as modes of representing the complex-ity of human activity. From Blumer’s perspective, there is no higher level of social lifethat requires theorizing by sociologists. It makes little sense to argue that Mead lacks atheory of institutions or that interactionists ignore social structures, since their under-standing of the social world is based on premises fundamentally different to those ofconventional structural sociology. Only if the presuppositions of orthodox sociologyare accepted a priori can interactionists be accused of being antitheoretical. Moreover,it may be further argued that modern sociology has gradually (and reluctantly) had tocome to terms with the empirical inadequacies of the theoretical postulation of socialsystems or structures, so that “the collective concepts developed by early theorists havebeen reformulated in ways that recognize the processes through which people activelyconstitute their social worlds” (Martin 2004: 34; see also Jenkins 2002).



To anticipate a misunderstanding that has become tediously familiar, it should beemphasized that this perspective does not entail a denial of the reality of armies,bureaucracies, nation-states, and so on. Such complexes of human activity constitutethe very facticity of the social world, or, as Blumer (1969: 22) had it, its “obduratecharacter.” An interactionist approach need not deny the authority of the president,the reality of income differentials, the power of global corporations, and so on, butmust seek to explain such phenomena as the results of, and as sustained or challengedby, the activities of real people in particular situations, often over considerably longperiods of time. Moreover, a principled refusal to accept the ontological reality of struc-tures or systems, and a corresponding focus on the ways in which they are socially con-stituted, does not (as some critics would have it) mean that the organized patterns ofsocial life are to be seen as subjective, malleable, or open to arbitrary redefinition. Asindividuals, we are confronted with the social reality of established, organized, norma-tive patterns of activity and a stable environment of symbolic representations—the factthat the norms and the symbols are ultimately sustained through patterns of interactiondoes not make them any less resistant. We learn the language of our culture, for exam-ple, and although we may modify it in some ways, or use it more or less effectively, weare not at liberty to abandon it or invent a new one. As Maines (1977: 238) has putit, “Blumer’s message should not be lost . . . because these very constraining processesare composed of and expressed through interacting individuals.”

Putting the matter another way, we could say that although collective conceptsmay constitute powerful symbols in everyday discourse—indeed, they are indispensa-ble in the process of communication—this does not mean that they should thereforebe treated as real for the purposes of sociological analysis. To do so would be to ille-gitimately reify them, to commit the “fallacy of misplaced concreteness” (Whitehead1925: 72). It is inconceivable that we could sustain even routine conversations with-out making reference to such notions as, for example, political parties, churches, orheavy metal fans. Yet as any empirical investigation shows, and as our experience ofsocial participation confirms, no such grouping is a unified entity composed of par-ticipants sharing the same ideas and motivations—as our comments on Wirth, above,indicate. As Blumer (2004: 95, emphases in original) points out:

Group life exists in what people do. It is not a preestablished organization con-ceived in terms of the completed acts that it is believed or hoped the people willcarry out. However much people may conform in their acts to a preestablishedscheme, the scheme is not their action. Nor is group life a kind of product of theacts of the individual participants, such as the articulated arrangement intowhich one finds the completed acts of the individuals to have fallen. Group lifeconsists of the actual acting of people—not a conceived organization that ante-dates that acting, nor of an articulated product of that acting. . . . Group life . . .has an organized character, but that character exists in it, not before it or after it.

Blumer has also made explicit the direct link between this interactionist approach tothe empirical analysis of the social world and Mead’s pragmatist philosophy, pointingout that Mead was “primarily a philosopher. He differed from the bulk of philosophers



in believing that the cardinal problems of philosophy arose in the realm of humangroup life and not in a separate realm of an individual thinker and his universe”(Blumer 1981: 902). This intimate connection between pragmatism and the interac-tionist approach to sociological study is at the heart of our argument. The manner inwhich this connection manifests itself in sociological practice therefore requires expli-cation. To achieve this we will consider three areas of sociological concern: the cre-ation and maintenance of the social world, the negotiated nature of that world, andthe relationship between people’s values and that world’s constitution.

THE PRAGMATIST BASES OF INTERACTIONIST STUDIESThe Creation and Maintenance of the Social World

First, as we have mentioned, the analytical focus of interactionist studies is on the cre-ation and maintenance of a shared social world by real people in particular situations.If successful, such an analysis will make explicit the bases on which such situations aredefined by participants, including symbolic representations and the taken-for-grantedaspects that influence their interpretations of what is going on. What it does not aimto produce are decontextualized, objective descriptions that will reveal the intrinsicmeanings of cultural objects.

An example will make this point in a less abstract way. In what Becker (1982) callsthe “art world” of symphonic music, interactions take place against a background dis-course in which such terms as orchestra, composer, score, audience, critic, conductor,and so on are generally unproblematic. Such discourses and their conventionallyestablished concepts may reveal much to the sociologist about the assumptions, inter-ests, and purposes of the various people in different social worlds. It does not follow,though, that these concepts correspond to ontologically real entities that can be ulti-mately or unambiguously defined. The concept of “conductor,” for example, is not anobjective description of a particular individual, but rather something that tells usabout the role such a person is expected to play within the symphonic art world. Thereare many other ways in which the person could be defined, all of which are (poten-tially) relevant to other situations, other “worlds.” The person who in certain timesand places is the “conductor” may also, in other contexts, be “a father,” “the accused,”“a high coronary risk,” “a board member,” and so on. That is to say, there is no sin-gle, privileged, or transcendent way of describing such a person—any person infact—independently of the discursive contexts in which particular modes of repre-sentation are established. In seeking to understand social activities in such contexts,interactionist analyses reflect the pragmatists’ principled objections to essentialism.

The Social World Is a Negotiated World

As we argued above, the notion that structural features of society have stable prop-erties independent of the activities of members of that society is deeply problem-atic. In order to generate a model of those features—of that society—it is necessaryeither to systematically reduce the empirical complexity of the phenomena under



investigation (as in Blau and Duncan 1967) or to make a theoretical decision abouthow such phenomena must operate (as in Giddens 1981). Both strategies must beadopted as a priori matters as they do not represent analytical claims based on empir-ical data, but rather specify what is to count as an instance and what each instancemight be an instance of.

The interactionist alternative to such positions may be exemplified by the conceptof a “negotiated order” (Strauss et al. 1964; Strauss 1978). People in complex organ-izations operate not on the basis of their role alone, because many such people withdifferent and possibly contradictory roles are required to cooperate to get the work ofthe organization done. Instead, they negotiate with one another to get work done,partly on the basis of their role, but also on the basis of their personalities and per-sonal preferences, their relationships with other staff members, their extraorganiza-tional networks and affiliations, to whom they owe favors and who owes them favors,and so on. Through these negotiations work gets done. It is not the case, however,that a stable, objective structure preexists those negotiations or is produced by them.Such a structure simply does not exist in the ways Blau and Duncan, or Giddens,might imagine. Rules govern participants’ activities, but those rules are themselvesinterpreted and negotiated as part of the ongoing work of the organization. (Thisdoes not mean that they are capable of being interpreted however one likes—as withany social phenomena they do have an “obdurate character.”) Finding out what rulesmight apply and how activities can be shown to be conducted with proper respect totheir authority are themselves part of ongoing organizational work. The structuredand stable features of any organization are necessarily the products of the webs ofinteraction undertaken by that organization’s members. In accordance with pragma-tist ideas, then, what seem to be “things”—stable structures—are empirically investi-gated and found to be “activities”—human interactions—that maintain and developthe stable operations of the setting under investigation.

Values Shape How the Social World Is Constituted

In conventional structuralist sociology there is a tendency to treat activities as deter-mined by objective and external social facts. Where norms do appear they do so notas individuals’ personal, professional, and collective values but as the products ofsocial systems. Thus, classically, Durkheim’s (2006) analysis of suicide relocates indi-viduals’ feelings of alienation and hopelessness away from their psychological makeupand toward the normative order of the society they inhabit: these things become func-tions of religious affiliation, domestic circumstances, economic conditions, and so on.And, as Alexander (1988: 98–99) argues, “[f ]or Herbert Blumer, George Homans,the early Erving Goffman, the later Garfinkel, Ralph Turner, and Aaron Cicourel, theonly way to emphasize the importance of individual, contingent action was to neu-tralize the influence of values and prior social structure as such.”

Interactionist sociologists, however, do not reduce the nature of social norms andhuman values to mere psychological phenomena—indeed, the idea that they musteither be located at the level of a “prior social structure” or inside the heads of socialinteractants is entirely alien to their perspective. Rather, values are found to be



things that emerge through, and are modified by, the interactional work of partici-pants to a setting. Thus, in his careful analysis of the relationships between nativeCanadians and Canadian Somali immigrants, Maines (2001: 158–161) shows howeach group’s attitude to the other is shaped by their culturally specific values.Although most white Canadians orientate to other ethnic groups on the basis oftheir race, Somali migrants make such in-group/out-group distinctions on the basisof nationality. By being reduced to their racial characteristics, Somalis experiencemajority Canadian understandings as denigrating and respond with hostility, thusaccelerating and worsening divisions between the two groups. The point here is notthat these distinctions are structural, but that the structural features of differenceemerge from the two groups engaging in social activities with one another: culturalvalues are elaborated, strengthened, and crystallized through intergroup interac-tions, rather than predetermining the ways in which those interactions could occur.Just what values will be relevant in any situation is a function of the situation andits participants—without people doing things the “normative order” can be no morethan an analytical fiction.

This focus on what Blumer called the “natural social world of everyday experience”(Blumer 1969: 148) is, we believe, at the heart of interactionist sociology, and allowsthe analyst to avoid the determinist account of human behavior inherent—andinescapable—in structural approaches. This was certainly Blumer’s conclusion: “Thepoint of view of symbolic interactionism is that large-scale social organization has tobe seen, studied, and explained in terms of the process of interpretation engaged inby the acting participants” (Blumer 1969: 58).

CONCLUSIONWe have argued that, far from being deficient in its ability to deal with “macro” soci-ological phenomena, symbolic interactionism is a coherent alternative to structuralapproaches, and as a consequence the concept of structure (in any of the various waysin which it has been employed in mainstream sociology) cannot simply be importedinto, or imposed upon, the interactionist perspective. That is, from the latter point ofview the concept of structure is itself problematic; indeed, as Blumer argued, the sup-posedly objective concepts of “macro” sociology are themselves ultimately subjectivein that they are formulated on the basis of analysts’ assumptions and presuppositions.We have suggested that the Chicago tradition of research—strongly influenced bypragmatist currents of thought—has been, from the start, concerned with the waysin which social order emerges out of the dynamics of human interaction. In contrast,with reference to efforts to describe the class structure sociologically, we have arguedthat the (widely varying) results not only betray the theoretical presuppositions of theresearchers but also are representations of no conceivable societies. So, following thedirections indicated by Mead, the symbolic interactionist tradition has focused oninstitutional analysis in which—instead of presupposing that institutions are collec-tive entities—the regular, orderly patterns of social life (including inequalities of



wealth, power, and prestige) are viewed as the outcomes of the collaborative interac-tions of real people in real situations.

Our conclusion, then, is consistent with Blumer’s wish to differentiate the basicpremises of symbolic interactionism from those of the sociology and psychology thathave become orthodox—premises in which “the meanings of things for the humanbeings who are acting are either bypassed or swallowed up in the factors used toaccount for their behavior” (Blumer 1969: 3). Thus, “sociological views of humansociety are, in general, markedly at variance with the premises . . . underlying sym-bolic interactionism” (Blumer 1969: 83). By respecting Blumer’s message and hislegacy, we believe, it is possible both to demonstrate the problems inherent in struc-tural thinking and to rescue sociology from its continuing grip.

Alex Dennis is Lecturer in the Sociology of Deviance at the University of Salford. Hisprevious publications include Making Decisions about People (Aldershot: Ashgate, 2001).

Peter J. Martin is Senior Lecturer in Sociology at the University of Manchester. His pre-vious publications include Sounds and Society (Manchester: Manchester University Press,1995) and Music and the Sociological Gaze (Manchester: Manchester University Press,2006). The authors are currently working on an edited collection on the problem ofstructure and agency in contemporary sociology.

REFERENCESAbbott, A. 1997. “Of Space and Time: The Contemporary Relevance of the Chicago School.”

Social Forces 75: 1149–1182.Abbott, A. and E. Gaziano. 1995. “Transition and Tradition: Departmental Faculty in the Era

of the Second Chicago School.” Pp. 221–272 in A Second Chicago School? The Developmentof a Postwar American Sociology, edited by G. A. Fine. Chicago, IL: University of ChicagoPress.

Alexander, J. C. 1988. “Parsons’ ‘Structure’ in American Sociology.” Sociological Theory 6:96–102.

Alexander, J. C. and B. Giesen. 1987. “From Reduction to Linkage: The Long View of theMicro-Macro Link.” Pp. 1–42 in The Micro-Macro Link: Problems and Prospects, editedby J. C. Alexander, N. J. Smelser, and R. Munch. Berkeley: University of California Press.

Athens, L. 2005. “Mead’s Lost Conception of Society.” Symbolic Interaction 28: 305–325.Baldwin, J. D. 1986. George Herbert Mead: A Unifying Theory for Sociology. Newbury Park,

CA: Sage.Becker, H. S. 1982. Art Worlds. Berkeley: University of California Press.———. 1999. “The Chicago School, So-Called.” Qualitative Sociology 22: 3–12.Berger, P. and T. Luckmann. 1966. The Social Construction of Reality: A Treatise in the Sociology

of Knowledge. New York: Doubleday.Blau, P. and O. Duncan. 1967. The American Occupational Structure. New York: Wiley.Blumer, H. 1939. Critiques of Research in the Social Sciences: An Appraisal of Thomas and

Znaniecki’s “The Polish Peasant in Europe and America.” New York: Social ScienceResearch Council.



———. 1954. “What Is Wrong with Social Theory?” American Sociological Review 19: 3–10.———. 1956. “Sociological Analysis and the ‘Variable.’” American Sociological Review 21:

683–690.———. 1969. Symbolic Interactionism: Perspective and Method. Englewood Cliffs, NJ:

Prentice-Hall.———. 1981. “Review of David L. Miller’s George Herbert Mead: Self, Language, and the

World.” American Journal of Sociology 86: 902–904.———. 1990. Industrialization as an Agent of Social Change. Hawthorne, NY: Aldine de

Gruyter.———. 2004. George Herbert Mead and Human Conduct. Walnut Creek, CA: Altamira.Bowles, S. and H. Gintis. 1976. Schooling in Capitalist America: Educational Reform and the

Contradictions of Economic Life. New York: BasicBooks.Braverman, H. 1974. Labor and Monopoly Capital: The Degradation of Work in the Twentieth

Century. New York: Monthly Review Press.Cicourel, A. V. 1964. Method and Measurement in Sociology. New York: Free Press.Coser, L. A. 1976. “Sociological Theory from the Chicago Dominance to 1965.” Annual

Review of Sociology 2: 145–160.Crompton, R. 1993. Class and Stratification: An Introduction to Current Debates. Cambridge,

UK: Polity.Dennis, A. and P. J. Martin. 2005. “Symbolic Interactionism and the Concept of Power.”

British Journal of Sociology 56: 191–213.Denzin, N. K. 1992. Symbolic Interactionism and Cultural Studies: The Politics of Interpretation.

Oxford, UK: Blackwell.Durkheim, É. 1982 [1895]. The Rules of Sociological Method and Selected Texts on Sociology and

Its Method. London: Macmillan.———. 2006 [1897]. On Suicide. Harmondsworth, UK: Penguin.Garfinkel, H. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.Garnsey, E. 1978. “Women’s Work and Theories of Class Stratification.” Sociology 12:

223–243.Giddens, A. 1981. The Class Structure of the Advanced Societies. London: Hutchinson.Goldthorpe, J. H. 1980. Social Mobility and Class Structure in Modern Britain. Oxford, UK:

Clarendon.Gouldner, A. W. 1970. The Coming Crisis of Western Sociology. New York: BasicBooks.Hall, P. M. 2003. “Interactionism, Social Organization, and Social Processes: Looking Back

and Moving Ahead.” Symbolic Interaction 26: 33–55.Hughes, E. C. 1951. “Mistakes at Work.” Canadian Journal of Economics and Political Science

17: 320–327.———. 1962. “Good People and Dirty Work.” Social Problems 10: 3–11.Jenkins, R. 2002. Foundations of Sociology: Towards a Better Understanding of the Human

World. London: Palgrave Macmillan.Katovich, M. A. and W. A. Reese, II. 1993. “Postmodern Thought in Symbolic Interaction:

Reconstructing Social Inquiry in Light of Late-Modern Concerns.” Sociological Quarterly34: 391–411.

Lichtman, R. 1970. “Symbolic Interactionism and Social Reality: Some Marxist Queries.”Berkeley Journal of Sociology 15: 75–94.

Maines, D. R. 1977. “Social Organization and Social Structure in Symbolic InteractionistThought.” Annual Review of Sociology 3: 235–259.



Maines, D. R. 2001. The Faultline of Consciousness: A View of Interactionism in Sociology. NewYork: Aldine de Gruyter.

Martin, P. J. 1984. “Social Stratification.” Pp. 20–42 in Applied Sociological Perspectives, editedby R. J. Anderson and W. W. Sharrock. London: Allen and Unwin.

———. 1987. “The Concept of Class.” Pp. 67–96 in Classic Disputes in Sociology, edited byR. J. Anderson, J. A. Hughes, and W. W. Sharrock. London: Allen and Unwin.

———. 2004. “Culture, Subculture, and Social Organisation.” Pp. 21–35 in After Subculture,edited by A. Bennett and K. Kahn-Harris. London: Palgrave Macmillan.

Mead, G. H. 1934. Mind, Self, and Society from the Standpoint of a Social Behaviorist. Chicago,IL: University of Chicago Press.

Meltzer, B. N., J. W. Petras, and L. T. Reynolds. 1975. Symbolic Interactionism: Genesis,Varieties, and Criticism. London: Routledge and Kegan Paul.

Mills, C. W. 1959. The Sociological Imagination. New York: Oxford University Press.Morrione, T. J. 2003. “Preface.” Pp. ix–xvii in H. Blumer, George Herbert Mead and Human

Conduct, edited and introduced by T. J. Morrione. Walnut Creek, CA: Altamira.Musolf, G. R. 1992. “Structure, Institutions, Power, and Ideology: New Directions within

Symbolic Interactionism.” Sociological Quarterly 33: 171–189.Park, R. E. and E. Burgess. 1921. Introduction to the Science of Sociology. Chicago, IL: University

of Chicago Press.Poulantzas, N. 1975 [1973]. Classes in Contemporary Capitalism. London: New Left Books.Sauder, M. 2005. “Symbols and Contexts: An Interactionist Approach to the Study of Social

Status.” Sociological Quarterly 46: 279–298.Scheff, T. J. 1995. “Academic Gangs.” Crime, Law, and Social Change 23: 157–162.Schutz, A. 1962 [1953]. “Commonsense and Scientific Interpretation of Human Action.”

Pp. 3–47 in Collected Papers: The Problem of Social Reality, Volume I. The Hague:Martinus Nijhoff.

Stewart, A., K. Prandy, and R. M. Blackburn. 1980. Social Stratification and Occupations.London: Macmillan.

Strauss, A. L. 1978. Negotiations: Varieties, Contexts, Processes, and Social Order. San Francisco:Jossey-Bass.

Strauss, A. L., L. Schatzman, R. Bucher, D. Ehrlich, and M. Sabshin. 1964. PsychiatricIdeologies and Institutions. New York: Free Press.

Watson, D. R. 1995. “Some Potentialities and Pitfalls in the Analysis of Process and PersonalChange in Counseling and Therapeutic Intervention.” Pp. 301–335 in Professional andEveryday Discourse as Behavior Change, edited by Jürg Siegfried. Norwood, NJ: Ablex.

Weber, M. 1948 [1924]. “Class, Status, Party.” Pp. 180–195 in From Max Weber: Essays inSociology, edited by H. Gerth and C. W. Mills. London: Routledge and Kegan Paul.

———. 1978 [1914]. Economy and Society. Berkeley: University of California Press.Whitehead, A. N. 1925. An Enquiry Concerning the Principles of Natural Knowledge.

Cambridge, UK: Cambridge University Press.Wirth, L. 1928. The Ghetto. Chicago, IL: University of Chicago Press.Wright, E. O. 1978. Class, Crisis, and the State. London: New Left Books.———. 1985. Classes. London: Verso.



Herbert Blumer’s critics and followers have, for the most part, neglectedthe significance of his theory of fashion. In this paper I revisit Blumer’sthesis on fashion by identifying the fashion process as an instance of thegeneric social process of semiotic transformation. The concept of semi-otic transformation refers to the change in meanings of symbolic objectsover time. By identifying six elements of the generic social process ofsemiotic transformation (systemic permeability, systemic freedom, logo-nomic fluidity, logonomic openness, indeterminacy, and logonomic hege-mony) and by viewing semiotic transformation as a joint act grounded inemergence, I argue that the meaning of change over time is a processstructuring both the definition of the situation and symbolic objects. I con-clude the paper with a set of reflections on the significance of Blumer’swork on fashion for the development of a research agenda of semiotictransformation and for the legacy and reception of Blumerian theory.

Despite the great wealth of both receptive and dismissive criticism that the writingsof the late Herbert Blumer have raised over the better part of the twentieth century,few observers have found much of theoretical or methodological value in his writingson fashion (Blumer 1968, 1969a; for a notable exception see Davis 1991, 1994).1 Bytaking Blumer as a point of departure, and by deriving inspiration from one of hisinterpreters (Prus 1987, 1994, 1996, 1997), I argue that Blumer’s critics and follow-ers have missed the relevance of his thesis on fashion for the theoretical developmentof a generic social process of semiotic transformation. The concept of semiotic trans-formation refers to the succession of changes that link the meanings of signs and dis-courses over time. Semiotic transformation, in other words, is the process wherebymeanings are produced, exchanged, interpreted, and used in different and changingways throughout time. A more refined understanding of semiotic transformation holdsthe key to a better conceptualization of such central sociological phenomena as socialchange, public opinion change, politico-economic development, historical mutationin values, conduct, and beliefs, resocialization and identity change, and more. Becauseof interpretive sociologists’ lack of attention to the complexities of historical processes

Extending and Broadening Blumer’sConceptualization of Fashion:

Adding Notions of Generic Social Process andSemiotic Transformation

Phillip Vannini*Royal Roads University

*Please address all correspondence to Phillip Vannini, School of Communication and Culture,2005 Sooke Road, Victoria BC V9B 5Y2, Canada. E-mail: [email protected].


(Maines 1988) and because of their general mistrust and poor understanding of semi-otic theories (Denzin 1992), interpretive sociologists in general and symbolic inter-actionists in particular have failed to explore and inspect what semiotic transformationsconsist of and how they unfold.

Drawing from Blumer’s (1968, 1969a) paper on fashion, Prus’s (1987, 1994,1996, 1997) conceptualization of generic social processes, and from contemporarysocial semiotics (e.g., Hodge and Kress 1989; Van Leeuwen 2005), in the followingpages I extend and broaden the significance of Blumer’s writings on fashion not in anattempt to discuss fashion in itself, but with the goal of laying the foundations for asociological understanding of generic social processes of semiotic transformation. It isimportant to point out that I am less concerned with exegesis than with proposing acreative addition to Blumerian theory and to its interpretation by Prus, arguably inways not intended and certainly not envisioned by either of the two. The value of mycreation, I hope, lies in opening the way for broad application across sociology, semi-otics, and cultural studies. I begin by outlining the uniqueness of some of Blumer’sarguments contained in his work on fashion. Subsequently, I propose a way of extend-ing Blumer’s ideas by adding the concepts of generic social process and semiotic trans-formation. I conclude with a final set of reflections and a brief survey of possibleapplications.

BLUMER’S UNDERSTANDING OF FASHIONBlumer’s (1969a) Sociological Quarterly paper on fashion begins with his exhortationto sociologists to transcend the usual denotative association of the word “fashion”with the social world of clothing and elegance, or “costume and adornment” (Blumer1969a: 275). Fashion is everywhere for Blumer, as it works as a mechanism operating“in many diverse areas of human group life” including “the realm of the pure andapplied arts,” “the area of entertainment and amusement,” and the fields of medicine,industry, mortuary practice, literature, modern philosophy, political doctrine, and sci-ence (Blumer 1969a: 275–276). Fashion occurs, therefore, across different situationsand contexts.

In his traditional polemical style, Blumer proceeds to identify and criticize themain currents of social thought on the phenomenon of fashion. Blumer suggestswe should treat fashion as a form of collective and agentic selection arising out ofthe interaction of interested parties, rather than as an effect of the social will to classdifferentiation (e.g., Simmel 1957), the existential need for novelty, or of psychologi-cal motives, feelings, and impulses such as “escape from boredom or ennui,” “spirit ofadventure,” “sexual interests,” ego-enhancement, or “desires for personal prestige ornotoriety” (Blumer 1969a: 284–285). Davis’s (1991) evaluation of the usefulnessof Blumer’s thesis on fashion as collective disposition and selection is particularlyeffective in drawing attention to the importance of this aspect of Blumer’s theory. Myattention here goes elsewhere, namely to Blumer’s (1969a: 285) identification of the“generic character of fashion.”

EXTENDING AND BROADENING BLUMER’S CONCEPTUALIZATION OF FASHION 307


Blumer is adamant about the generic social qualities of the fashion process. Fashion isnot “an aberrant and irrational social happening, akin to a craze or mania” (1969a: 276).Instead, from the perspective of the individual, fashion constitutes the working of asystem of rules that “sets sanctions of what it is to be done” (1969a: 276) if the socialagent in a calculative manner intends “to make sure that he is ‘in style’” (1969a: 277).Fashion, therefore, works as a regulatory system of negotiable and emergent rules thatstructure the array of choices social agents have in determining the meaning and valueof symbolic objects. Blumer (1969a: 277) suggests that “the person who unwittinglyfollows a fashion does so because of a limitation of choice,” namely a limitation in thefreedom to act toward an object on the basis of the meanings the object has. Whatmatters, therefore, is not that an individual wills to be “in style,” but that an individ-ual wills to produce, exchange, or interpret the symbolic meanings of an object in linewith the system of rules for interpretation and use in effect within an interpretivecommunity.

At the level of collective choice, Blumer (1969a: 277) identifies a similar force inplay: “while people may become excited over a fashion they respond primarily to itscharacter of propriety and social distinction; these are tempering guides.” Collectiveselection and disposition processes then originate the rules by which people make andcommunicate meaning. The rules of fashion, for Blumer, are primarily lexical andsocial: they are lexical in that they stipulate what is meant by what symbolic object,and they are social in that they stipulate—at least to some extent—who has theauthority to define their meanings and the meanings of the circumstances underwhich they are produced, exchanged, used, and interpreted.

Finally, let us now pay closer attention to the temporal-structural elements of thisprocess and begin by evaluating once again how for Blumer (1969a: 282) the “fash-ion mechanism [is] a continuing process of collective selection.” By calling our atten-tion to the temporal movement of fashion as a collective and reflexive social processresponding to changes in preferences and sensitivity, Blumer effectively identifiesfashion as a “joint act.” Thus, whereas other students of fashion might view eitherinnate existential, psychological, or structural forces, or even arbitrary relations, as theorigin of transformation (Blumer 1969a: 276–280), Blumer views the process ofmeaning change in fashion as deeply interactional and social.

A joint act, Blumer’s term for the Meadian concept of the social act, is a collectivecourse of action undertaken by interacting individuals or groups (Blumer 1966: 540;1969b: 16). Joint or collective acts constitute the nature of human action in that theyrely on the individual’s capacity to make indications to him- or herself, based on inter-action with objects and others:

Joint or collective action constitutes the domain of sociological concern, as exem-plified in the behavior of groups, institutions, organizations, and social classes.Such instances of societal behavior, whatever they may be, consist of individualsfitting their lines of action to one another . . . [constituting] societal organizationof conduct of different acts of diverse participants. A joint action, while made upof diverse component acts that enter into its formation, is different from any one



of them and from their mere aggregation. The joint action has a distinctive char-acter in its own right, a character that lies in the articulation or linkage as apartfrom what may be articulated or linked. (1969b: 16–17)

Blumer’s view of fashion as a joint act is based on emergence, a concept that Blumerlearned from Mead’s (1934) philosophy. This view is shared by Davis (1991: 9) whofinds that for Blumer’s theory of fashion “nothing is either foreordained or altogetherimpossible: emergence is everything.”

Despite the attention Blumer paid to the structural and temporal/historical aspectsof social processes in the realm of fashion, industrialization, industrial relations, andother arenas of individual and collective conduct, criticisms of his alleged astructuraland ahistorical bias abound (see Maines 1988 for a review). These “myths” aboutBlumer are especially vexing in view of the fact that he made explicit and frequentremarks throughout his work over the years in regard to the centrality of joint actionfor the sociological enterprise. As Maines (1988) pointed out, Blumer’s statements onthe implications of joint action show precisely how the temporality of the process ofjoint action is at the basis of his theory of structure. Fashion, with its nature of being“caught in movement—movement from an outmoded past toward a dim, uncertain,but exploitable immediate future” (Blumer 1969a: 289), can work then—as I willargue—as a fantastic example of semiotic transformation and of the historical impli-cations of joint action.

EXTENDING BLUMER’S IDEAS: GENERIC SOCIAL PROCESSES OF FASHION

In order to understand semiotic transformations, one need not adopt a teleologicalview of history. Meaning change may have no ultimate directionality; nevertheless,meanings change throughout time. Following Mead (1934), we can understand semi-otic transformations as emergent and irregular, and, yet, if semiotic transformationspresented no regularity whatsoever, little of any theoretical usefulness could be saidabout them. The key analytical goal must be that of discovering existing patterns ofsemiotic transformations, or, more precisely, the links between semiotic transforma-tions across time and the particularities of the contexts in which they occur. Derivinginspiration from Prus’s interpretation of Blumerian theory, I argue that in order to doso we can envision semiotic transformations as generic social processes.

Generic social processes, according to Prus (1996: 142), refer to “the transitua-tional elements of interaction; to the abstracted transcontextual formulations of socialbehavior. Denoting parallel sequences of activity across diverse contexts, generic socialprocesses highlight the emergent, interpretive features of association. They focus ourattention on the activities involved in the ‘doing’ or ‘accomplishing of human grouplife.’” Although one often finds great variability in the symbolic meanings of objectsof action over time, it is possible to identify a chain of elements whose presence orabsence enables the occurrence of semiotic transformation while, at the same time,informing the very nature of variation of symbolic meaning. Take, for example, the



semiotic transformation of fashion in clothing over the years. One year may bethe year of pink, and the following year, black may be the new pink, only to be fol-lowed by the rediscovery of red, and so forth. Then, over the years, a color like pinkmay be associated with such divergent meanings and values as meekness, overt sexualseductiveness, childish innocence, sassiness, or a conservative gender ideology.Although prediction of future trends may be impossible, it is certain that the processof semiotic transformation in fashion occurs as a result of interacting elements whoseinterlinkage may be common to such diverse contexts as the social worlds of amuse-ment and entertainment, literature, arts, and even “political doctrine” (Blumer 1969a:275). Given Blumer’s broad understanding of the fashion process, it seems safe toargue that fashion can be studied as a generic social process.

As Prus (1997: 88) explained, the study of generic social processes must “denotepotential realms of inquiry and synthesis in the abstract” and therefore “it is not untilpeople act and interact” that we can speak of the actual “diversity and interconnect-edness (intersectedness and interwovenness) of the activities that constitute the(mosaic) essence of human group life.” Generic social processes, such as semiotictransformation, are then indeed generic, but “it is incumbent on social scientists to beattentive to the fundamentally intersubjective essence of every realm of humanendeavor they consider, to concentrate their emphasis on people doing things withinthe context of ‘minded’ (Mead 1934) enterprise” (Prus 1997: 88, italics in the origi-nal). With this in mind, let us examine how Blumer’s thesis on fashion can be broad-ened as a type of generic social process in order to further our understanding ofsemiotic transformation.

Blumer identifies six features or conditions of the fashion process. Social agents arenot necessarily aware of the existence of these features (Blumer 1969a: 286), much asthey are not necessarily immediately conscious of the existence of the logonomic rules(that is, rules that structure who has the power to produce, change, and consumemeaning, under what circumstances, and in what manner) that structure semiotictransformation (Hodge and Kress 1988).

Blumer describes these six features but stops short of naming them or expandingupon their significance; therefore, the six concepts presented below are merely mylexical choices made on the basis of their implicit relatedness to socio-semiotic theory.Throughout the following I use—one might even say that I strategically hijack—quotations by Blumer to support my argument, and thus ultimately the responsibilityfor the following argumentation is entirely mine and not Blumer’s.

The first feature of the fashion process is “logonomic fluidity.” The word “logo-nomic” refers to a discursive ordering mechanism (in Greek “logos” equals thoughtand the language or discourse used to present that thought; “nomos” equals controlor ordering mechanism). A logonomic system may be solid or fluid in relation to thepractices and “rules prescribing the conditions for production and reception of mean-ings” (Hodge and Kress 1988: 4). “Fluidity” refers to the existence of a constant“movement of change, with people ready to revise or discard old practices, beliefs, andattachments” (Blumer 1969a: 286). The absence of logonomic fluidity would meanthat there is a closure toward change, as may be the case in some radically orthodox



systems of religious belief, and that people are, therefore, not “poised to adopt newsocial forms” (Blumer 1969a: 286).

The second feature observed by Blumer is what I call “logonomic openness.” Thesocial world potentially undergoing semiotic transformation “must be open to the recur-rent presentation of models or proposals of new social forms” (Blumer 1969a: 286).There are similarities between Blumer’s arguments on fluidity and openness, but dif-ferences as well. Logonomic fluidity, as I conceptualize it, is marked by the existenceof a culture or collective disposition that “places a premium on being ‘up to date’ andwhich implies a readiness to denigrate given older forms of life as being outmoded”(Blumer 1969a: 286). Fluidity therefore refers to a “mindset,” whereas logonomicopenness refers to the structural openness of a semiotic system. Sociologists andsocio-semioticians interested in understanding how openness works as a conditionof the generic social process of semiotic transformation ought to be sensitive toinquiring about resistance to change. Wherever there are vested interests in uphold-ing traditions, or perhaps wherever there are material limitations to change, open-ness is hampered and semiotic transformation may not take place. The condition ofopenness, therefore, sensitizes us to pay attention to “who can claim to initiate (pro-duce, communicate) or know (receive, understand) meanings about what topicsunder what circumstances and with what modalities (how, when, why)” (Hodge andKress 1988: 4).

The third element is “systemic freedom”—or more simply availability of choicewithin a structure of signs. Whereas logonomic openness refers to the condition ofpermeability of the system of rules by which meanings and situations are definedand produced, systemic freedom refers to the condition of choice among availablesymbolic objects at the level of the user/interpreter. As Blumer (1969a: 286) states:“There must be a relatively free opportunity for choice” between symbolic objectsand semiotic models carrying competing meanings. Systemic freedom bespeaks ofboth basis of need and want that may be in place within a social world. When socialagents have the (relative) freedom to select among symbolic objects in order toachieve their goals, they relate to such objects on the basis of the changing semioticinstrumental potential they contain—or, in simpler words, on the basis of whatsymbols can possibly do for them. Symbolic objects, or signs, therefore becomeresources. Take, for example, the relative systemic freedom of a scholar in choosingcompeting theoretical models to interpret empirical data, or at least in choosinghow to formulate possible hypotheses in preparing a grant proposal. Although it isdoubtful that our scholar will select among competing theoretical models exclusivelyon the basis of their fashionableness, such fashionableness will definitely play a role,as any twenty-first century criminologist who has considered measuring people’sskulls to determine their propensity to commit crime will attest. In this case, it isevident how symbolic models and objects work just like technical resources amongwhich social agents will select in order to achieve their goals. Different and chang-ing arrays of availability for different people and differing goals will result in systemicchange, or semiotic transformation. It is imperative for the student of generic socialprocesses of semiotic transformation to understand agents’ changing motives for



their selection as well as the connection between differing motives and changingconfigurations of choice.

Symbolic objects work as resources under circumstances that are highly variable.Furthermore, as previously stated, different people with different goals will selectamong competing available resources on the basis of the potential they interpret suchresources to hold. For this reason, an important condition of the generic social processof semiotic transformation is the “indeterminacy” of the value of resources.Indeterminacy refers to the lack of an ultimate test of the “pretended merit or valueof the competing models” or resources (Blumer 1969a: 286). Of course, if there werefinal and absolute determinacy, there would be no door open to change, or at least noimmediately available justification for it. On the point of determinacy and indeter-minacy, Blumer seems to contradict himself. At one point Blumer (1969a: 286) statesthat fashion “is not guided by utilitarian or rational considerations.” This seems tocontradict his earlier statements that the “fashion-conscious person is usually quitecareful and discerning in his effort to identify the fashion in order to make sure thathe is ‘in style’” (1969a: 277) and that individuals’ fashion adoptions are done “delib-erately and not irrationally” (1969a: 277). Perhaps by refusing to identify fashion asa rational or utilitarian process, Blumer intends that what I call semiotic transforma-tions have no teleological directions, for they are but emergent from the continuouslychanging interplay of individual and group action. Notwithstanding this lack of clar-ity on the matter, the element of indeterminacy ought to sensitize the student of thegeneric social process of semiotic transformation to the changing methods by whichsymbolic resources are assessed over time.

The fifth element of the process of semiotic transformation is “logonomic hege-mony.” Hegemony is intended here as a configuration or alliance of groups and/orindividuals and the concerted action they exercise in order to preserve control overthe economic, political, cultural, and ideological domains of a logonomic system. ForBlumer (1969a: 287), no system of fashion can exist without the shared agreement of“prestige figures who espouse one or another of the competing models.” Therefore, acondition of hegemony, however partially and temporarily achieved as an unstableequilibrium, must be reached in order for logonomic systems to carry “weight as anassurance or endorsement of the superiority or propriety of a given model” (Blumer1969a: 287). Hegemony, as intended here, is not a form of domination of the subor-dinate, but more simply a form of institutionalization of meaning in the absence ofwhich there would be less “likelihood of adoption of the model” (Blumer 1969a: 287)or no “uniformity in what would otherwise be a markedly fragmented arrangement”(1969a: 289). At the collective level, semiotic transformations must be accepted oragreed upon by at least more than one individual, and following Blumer’s observa-tions of the haute couture social world in Europe, one can safely conclude that cer-tain individuals and groups have more power than others to promote change anddefine the situation. Nevertheless, as Blumer found, no hegemonic alliances are strongenough to determine semiotic transformation in any stable manner, as any high pro-file fashion designer who has failed to introduce an innovation despite great publicitywould promptly agree.



The sixth and final element identified by Blumer is “systemic permeability.”Permeability is the property of being penetrable, porous, and open to influxes. Forsemiotic transformation to take place, social worlds and their symbolic systemsmust be permeable to the influx of “new interests and dispositions in response to(a) the impact of outside events, (b) the introduction of new participants into thearea, and (c) changes in their inner social interaction” (Blumer 1969a: 287).Without systemic permeability there are no opportunities for collective tastes ordispositions to shift. Take, for example, the social world of a music subculture. Thatsocial world changes in relation to greater cultural and sociopolitical events in theworld, in relation to the influx of new followers and the exit of older members, andin relation to changing practices and conventions, or ways of scripting and per-forming the culture of the group.

Together, logonomic hegemony, fluidity and openness, systemic freedom and per-meability, and resource indeterminacy form the conditions under which semiotictransformation—following my reading of Blumer’s fashion thesis—will occur.Blumer (1969a: 288) explains that scrutiny of these components of the generic socialprocess of semiotic transformation will shed light on how social worlds turn “awayfrom old forms that are thought to be out of date; [and on] the introduction of newmodels which compete for adoption; a selection between them that is made not onthe basis of demonstrated merit or utility but in response to an interplay of prestigeendorsement and incipient taste; and a course of development in which a given typeof model becomes solidified, socially elevated, and imperative in its demands foracceptance for a period of time.” I now proceed to discuss in greater depth the tem-poral properties of semiotic transformation.

FASHION AND SEMIOTIC TRANSFORMATIONSemiotic transformations are of great interest to social semioticians. Social semioticsis not synonymous with general semiotics. Whereas most semioticians derive inspira-tion from the theory of Ferdinand de Saussure (1974), social semioticians are gener-ally highly critical of Saussurean thought and tend to favor an updated version ofPeircean theory (Peirce 1940). The differences between the two are significant and toonumerous to be considered here (for a review see Hodge and Kress 1988; Vannini2006); let it suffice to say that social semiotics rejects the structural determination ofmeaning and the relative fixity of meaning over time theorized by Saussure and laterby Levi-Strauss and Chomsky, drawing instead from Peirce’s conceptualization ofsemiosis (i.e., the process of constitution of meaning) as a human communicativeevent unfolding over time (Hodge and Kress 1988; Rochberg-Halton 1982). Muchlike interactionists, social semioticians also give great consideration to context, andthus to the material, ideological, historical, and sociopolitical relations that contributeto the definition of the situation in which semiosis takes place (see Hodge and Kress1988). Socio-semioticians then effectively straddle the continuum between semioticand sociological theory and constructionist theory in particular (Vannini 2004).



A central interest in social semiotics is the uncovering of ideological forces in rela-tion to semiosis. Because certain groups in our society have a vested interest in defin-ing situations and the meanings of symbolic objects in ways that favor their owneconomic or political agendas, and because they have relatively more power to do sothan others do, semiosis is always shaped by ideology and sociopolitical power.Therefore, although men and women agentically make their own semiotic history—to borrow from Marx—they do not necessarily make it under circumstances theyhave chosen, but instead under circumstances directly shaped by their past and by thesocial forces active in shaping that past. Fashion in clothing is directly illustrative ofthis. As Davis (1994) has shown, the semiotic transformation of the meanings of ele-gance is directly shaped by a historical dialectics of identity ambivalence. The samecould be said of fashion in scientific theory, since the succession of paradigms is, bynecessity, a response and reaction to previously dominating paradigms and to thesocial and political conditions that led to the emergence of consensus over them(Kuhn 1962). In both cases, and especially more so in the latter, old conventions andhabits are notoriously difficult to transform.

Following Peircean terminology, we could loosely define semiotic transformationsas successions of different habits of thought. The succession of habits of thought maythen be understood as the emergent variation of logonomic rules for the productionand interpretation of meaning over time. Emergence constitutes the basis of tempo-ral structure of semiotic transformation. In fact emergence—a process—structures thelogonomic rules by which meaning is produced, exchanged, interpreted, and used bysocial agents involved in joint acts. The emergence of fashion works because it “canonly build on a symbolically shared past of thoughts, images, and sentiments as theseflow into an undefined present and uncertain future” (Davis 1991: 9), or, in Blumer’s(1969b: 20) own words, the process of emergence structures collective action in that

any instance of joint action, whether newly formed or long established, has nec-essarily arisen out of a background of previous actions of the participants. A newkind of joint action never comes into existence apart from such a background.The participants involved in the formation of the new joint action always bringto that formation the world of objects, the sets of meanings, and the schemes ofinterpretation that they already possess. Thus, the new form of joint actionemerges out of and is connected with a context of previous joint action.

Emergence, therefore, allows at once for innovation and tradition to be part of thetemporal process of semiotic transformation. Furthermore, by focusing on the tem-poral perspective of emergence, the student of semiotic transformation can pay atten-tion to both processual and structural elements.

The importance of the principle of emergence in semiotic transformation cannotbe stressed enough. Students of semiotic transformation in the Saussurean traditionhave either discounted diachrony2 or treated it as a by-product of forces external tosemiosis (Rochberg-Halton 1982; Vannini 2004). A theorization of semiotic transfor-mation and social change grounded in pragmatism, instead, can allow us to viewdiachronic change as a complex dynamic outcome of the habituated and nonhabituated



sides of life (see Han-Yin Chang 2004; Snow 2001). Understood this way, semiotictransformation depends on the interplay of joint acts that give rise to emergent mean-ings and the nature of preexisting conditions that make this interaction meaningfuland possible.

The conceptualization of semiotic transformation as emergent brings us to viewBlumer’s thesis on fashion in a new light. When we understand fashion as a constantmovement of “reaching out for new models which will answer to as yet indistinctand articulate newer tastes” (Blumer 1969a: 282), we can see how for Blumer trans-formation in meaning arises “from the diversity of experience that occurs in socialinteraction in a complex moving world” (1969a: 282). This view of symbolic changeparallels Blumer’s theory of industrialization as an agent of social change (Blumer1990). In industrialization, much like in semiotic transformation, variation overtime does not occur as a result of the determining causal force of specific agents orsingle structural components, but instead arises out of the tension toward movementtypical of the interaction system itself. In sum, therefore, fashion, semiotic transfor-mation, industrialization, and any form of symbolic change occur in this worldbecause of a continuous tension toward becoming—a central, but somewhat under-theorized philosophical component of Blumerian theory and symbolic interaction-ism in general.

Such tension toward becoming can be traced back to Peircean philosophy (Peirce1940). For Peirce, semiosis is always a process in tension toward a temporal infinite.This is because interpretation always gives rise to a new sign, which gives rise to a newinterpretation, and so forth. For Rochberg-Halton (1982) this semiotic tensiontoward the inevitability of transformation constitutes the very basis of “structure” ininteractionist and pragmatist theory. Indeed, because meaning always tends to itsregeneration, it also always tends to the restructuring of the conditions (Peirce’s“habits of thought”) upon which it depends, or, in other words, the restructuring ofits logonomic rules. This ought to explain precisely why for Blumer (1969a: 283) “thehistory of fashion shows clearly that new fashions are related to, and grow out of, theirimmediate predecessors” with a “character of a cultural drift.” Therefore, even thoughtransformations may have “no line of historical continuity” (1969a: 283) they stilldrift, tend toward movement, or shift in trends “woven deeply into the texture ofmodern life” (1969a: 283).

Blumer’s (1969b) famous three premises of symbolic interactionism are also deeplygrounded in time. Symbolic interaction unfolds as an emergent process whose fluid-ity is made possible by the vast diversity of individual and group forms of conduct,and by the polysemy3 of the objects toward which symbolic action is directed overtime. Because pragmatist philosophy recognizes the variability of meaning over timeand the social origin of meaning, semiotic transformation stands as the most funda-mental social and semiotic process of all. Semiotic transformation, therefore, is at oncea social and semiotic process.

Returning to semiotic transformation as a generic social process, the following canbe said about the temporal structure of this process. First, semiotic transformationtends toward “unanimity and uniformity in what would otherwise be a markedly



fragmented arrangement” (Blumer 1969a: 289). The temporal generic process of semi-otic transformation, in other words, tends toward logonomic hegemony; however,because of the continuously changing and vastly diverse arrangement of joint acts,hegemony always remains unstable.

Second, “fashion serves to detach the grip of the past in a moving world” (Blumer1969a: 289). Therefore, because of its very nature of movement and change, the forceof semiotic transformation, much like a will to semiotic becoming, pushes against allrestrictions, thus ensuring the conditions of systemic permeability, systemic freedom,logonomic fluidity, and logonomic openness. Emergence, therefore, works both as a con-dition and an outcome of semiotic transformation.

Third, “by allowing the presentation of new models but by forcing them throughthe gauntlet of competition and collective selection the fashion mechanism offers acontinuous means of adjusting to what is on the horizon” (Blumer 1969a: 290).Selection, competition, and instrumental choice, as joint acts emerging out of the dif-fering value of resources, thus ensure the indeterminacy of semiotic value and trans-formation over time. The importance of this point cannot be overstated. In Blumer’s(1969a: 290) own words: “the value of a pliable and reforming body of common tasteto meet a shifting and developing world should be apparent.”

CONCLUSIONThroughout this article I have argued that Blumer’s thesis on fashion can be broadenedand extended to guide us through the study of semiotic transformation. The features ofthe fashion process identified by Blumer, I have suggested, are contiguous with the ele-ments of a generic social process of semiotic transformation. It is important to keep inmind, however, that fashion and semiotic transformation are not synonymous. Fashionrepresents a case of semiotic transformation, and only one of many possible cases andforms at that. Fashion is unique as a case of semiotic transformation, because it seem-ingly changes so drastically and so quickly and varies greatly among different groups ofusers and consumers. Semiotic transformation is liable to occur in less rapid ways inarenas less subject to the forces of fashion, such as language use (then again, if we wereto think of teenage slang we might wish to rethink this), or jurisprudential interpreta-tion. Nevertheless, even in those arenas it seems obvious that in order for the process ofsemiotic transformation to occur, such features as systemic permeability, systemic free-dom, logonomic fluidity, logonomic openness, indeterminacy, and logonomic hege-mony should be present. The process in such arenas less subject to fashion—and yet,still subject to it—is then likely to differ in terms of intensity of its features, but arguablystill likely to occur. Such is the core of my extension of Blumer’s argument and my appli-cation of Prus’s concept of generic social process: fashion is a generic social process anda case of the generic social process of semiotic transformation.

Obviously, it is not my intention to put words in Blumer’s mouth. Blumer’s thesis onfashion was meant, if we are to believe his words, to transcend widely the social worldof elegance, but it is my interpretation to suggest that fashion can serve as a sensitizing



tool for our understanding of semiotic transformation in general. In this sense, myextension of Blumer’s thesis on fashion is purely strategic and less concerned withpurity and exegesis and more with pragmatic applicability.

Viewing fashion as a lens for the generic social process of semiotic transformationopens the doors to the diachronic study of the changing meanings of anything that“can be conceptualized, envisioned, indicated, pointed to, referenced, talked about,or even thought about” (Prus 1997: 93). Symbolic objects whose semiotic changeover time could be studied include “all environmental and biological phenomena; alltools, appliances, and devices of all sorts; as well as all clothing and other apparel,and all realms of fashion, ideas, concepts, and beliefs—in short, all things of humanawareness” (Prus 1997: 93–94). Following Prus’s (1997: 93–96) lead (and arguablyin spite of his intentions for application), it is not difficult to envision how thegeneric social process of semiotic transformation—whose elements have been iden-tified above—can be applied to such diachronic semiotic processes as defining andredefining the use value, exchange value, sign-exchange value, and symbolic value ofobjects; the change in development, production, distribution, exchange, and acqui-sition of objects; the maintenance, repair, recycling, and disposing of objects, and soforth. Despite the limited attention sociologists and semioticians have paid to semi-otic transformation, the range of applicability of this concept shows that it is indeedof the utmost importance in the sociological, anthropological, and sociolinguisticenterprise.

The importance of Blumer’s thesis on fashion also transcends the phenomenon offashion per se in that it contains the genes of Blumerian thought on emergence, and,in particular, on the structuring power of the temporal process of emergence. As Davis(1991, 1994) also noted, Blumer’s thesis on fashion clearly shows that the apostle ofsymbolic interactionism has been but wrongly accused of fostering a vision of thisperspective marred by lack of attention to structure and history. Following Mead,Blumer found the fashion process interesting precisely because it clearly showed howa seemingly irrational, semidecontextualized, and subjective mechanism is, in reality,contingent on an emergent collective action directly shaped by, and in turn directlyshaping, a changing historical structure. Nevertheless, general ignorance of his vasttheoretical agenda as well as interactionist complicity in the reception of his theorieshave resulted in pushing forth an impartial, unaware, and often wrongful under-standing of the significance of Blumerian thought (Maines 1988). It is precisely forthis reason that all of his writings, including some of the most seemingly esoteric,such as his work on fashion, contain great potential for the development of interpre-tive sociological theory and research.

Phillip Vannini is assistant professor in the School of Communication and Cultureat Royal Roads University in Victoria, British Columbia. Together with DennisWaskul, he is editor of Body/Embodiment: Symbolic Interactionism and the Sociology ofthe Body (2006 Ashgate) and author of numerous articles in such journals as SymbolicInteraction, Qualitative Inquiry, Critical Discourse Studies, Cultural Studies, CriticalMethodologies, Journal of Popular Culture, and Studies in Symbolic Interaction.



NOTES1. More precisely, Blumer’s writings on fashion have often been acknowledged by researchersof the fashion (i.e., clothing) industry, but many have missed Blumer’s (1969a) analyticalpoints on “the wide range of operation of fashion.”2. Diachrony, the opposite of synchrony, is a term within semiotic theory that refers to his-torical unfolding.3. Polysemy is a concept that refers to the idea that signs have multiple meanings.

REFERENCESBlumer, Herbert. 1966. “Sociological Implications of the Thought of George Herbert Mead.”

American Journal of Sociology 71: 535–548.———. 1968. “Fashion.” Pp. 341–345 in International Encyclopedia of the Social Sciences, vol. 5,

edited by David Sills and Robert King Merton. New York: Macmillan.———. 1969a. “Fashion: From Class Differentiation to Collective Selection.” Sociological

Quarterly 10: 275–291.———. 1969b. Symbolic Interactionism: Perspective and Method. Berkeley: University of

California Press.———. 1990. Industrialization as an Agent of Social Change. Hawthorne, NY: Aldine de Gruyter.Davis, Fred. 1991. “Herbert Blumer and the Study of Fashion: A Reminiscence and a

Critique.” Symbolic Interaction 14: 1–21.———. 1994. Fashion, Culture, and Identity. Chicago, IL: University of Chicago Press.Denzin, Norman K. 1992. Symbolic Interactionism and Cultural Studies: The Politics of

Interpretation. Cambridge, MA: Blackwell.Han-Yin Chang, Johannes. 2004. “Mead’s Theory of Emergence as a Framework for

Multilevel Sociological Inquiry.” Symbolic Interaction 27: 405–427.Hodge, Robert and Gunther Kress. 1988. Social Semiotics. Ithaca, NY: Cornell University

Press.Kuhn, Thomas. 1962. The Structure of Scientific Revolutions. Chicago, IL: University of

Chicago Press.Maines, David. 1988. “Myth, Text, and Interactionist Complicity in the Neglect of Blumer’s

Macrosociology.” Symbolic Interaction 11: 43–57.Mead, George H. 1934. Mind, Self, and Society. Chicago, IL: University of Chicago Press.Patzer, Gordon. 1985. The Physical Attractiveness Phenomenon. New York: Plenum Press.Peirce, Charles S. 1940. Collected Papers. Cambridge, MA: Belknap Press of Harvard

University.Prus, Robert. 1987. “Generic Social Processes: Maximizing Conceptual Development in

Ethnographic Research.” Journal of Contemporary Ethnography 16: 250–291.———. 1994. “Generic Social Processes: Intersubjectivity and Transcontextuality in the

Social Sciences.” Pp. 393–412 in Doing Everyday Life: Ethnography as Human LivedExperience, edited by Mary Lorenz Dietz, Robert Prus, and William Shaffir. Toronto,ON: Copp Clark/Longman.

———. 1996. Symbolic Interaction and Ethnographic Research: Intersubjectivity and the Studyof Human Lived Experience. Albany: State University of New York Press.

———. 1997. Subcultural Mosaics and Intersubjective Realities: An Ethnographic ResearchAgenda for Pragmatizing the Social Sciences. Albany: State University of New York Press.



Rochberg-Halton, Eugene. 1982. “Situation, Structure, and the Context of Meaning.”Sociological Quarterly 23: 455–476.

Saussure, Ferdinand de. 1974. Course in General Linguistics. London: Fontana.Simmel, Georg. “Fashion.” American Journal of Sociology 62: 541–558.Snow, David A. 2001. “Extending and Broadening Blumer’s Conceptualization of Symbolic

Interactionism.” Symbolic Interaction 24: 367–377.Van Leeuwen, Theo. 2005. Introducing Social Semiotics. London: Routledge.Vannini, Phillip. 2004. “Toward an Interpretive Analytics of the Sign: Interactionism, Power,

and Semiosis.” Studies in Symbolic Interaction 27: 151–176.———. 2006. “Social Semiotics and Fieldwork: Method and Analytics.” Qualitative Inquiry

12: 1–28.



Because Herbert Blumer maintained that symbolic interactionism was use-ful in examining all realms of social behavior, and advocated what MartinHammersley refers to as “critical commonsensism,” this paper focuseson one of the most common contemporary social relationships—thatbetween people and companion animals. I first examine the basis forBlumer’s (like Mead before him and many interactionist scholars today)exclusion of nonhuman animals from consideration as “authentic” socialactors. Primarily employing the recent work of interactionists EugeneMyers, Leslie Irvine, Janet and Steven Alger, and Clinton Sanders, thispaper advocates the reasonableness of regarding nonhuman animals as“minded,” in that mind, as Gubrium emphasizes, is a social constructionthat arises out of interaction. Similarly, I maintain that animals possess anadmittedly rudimentary “self.” Here I focus special attention on Irvine’s dis-cussion of those “self experiences” that are independent of language andarise out of interaction. Finally, I discuss “joint action” as a key element ofpeople’s relationships with companion animals as both the animal andhuman attempt to assume the perspective of the other, devise relatedplans of action and definitions of object, and fit together their particular(ideally, shared) goals and collective actions. I stress the ways in whichanalytic attention to human-animal relationships may expand and enrichthe understanding of issues of central sociological interest.

As constructed by Herbert Blumer on a foundation largely established by GeorgeHerbert Mead, symbolic interactionism offers a unique1 and compelling perspectiveon social life. Central to interactionist thought are Blumer’s key propositions: “humanbeings act toward things on the basis of the meanings that things have for them”; “themeaning of such things is derived from, or arises out of, the social interaction that onehas with one’s fellows”; and “meanings are handled in, and modified through, an inter-pretive process used by the person in dealing with the things he encounters” (Blumer1969: 2, emphasis mine).2 In these elemental propositions Blumer makes the assump-tions that only “human beings” engage in the construction of meaning; that “humanbeing” and “person” are synonymous; and that social interaction based on the com-munication, cooperative shaping, and utilization of shared meaning is possible onlybetween and among humans. In so doing, Blumer relegates nonhuman animals to thecategory of “physical” rather than “social” objects (Blumer 1969: 10), and excludes

Mind, Self, and Human-Animal Joint ActionClinton R. Sanders*University of Connecticut

*An earlier version of this paper was presented at the joint Symbolic Interaction andEthnographic Research and North Central Sociological Association Conferences,Pittsburgh, PA, April 7–10, 2005. I appreciate the assistance of Steve Carlton-Ford, ScottGrills, and three anonymous reviewers. Please contact the author at the Department ofSociology, University of Connecticut; his email address is [email protected].


from consideration one of the most historically long-term, culturally relevant, andcontemporarily commonplace forms of association—that between humans and non-human animals.

Blumer’s lack of attention to animals is understandable given the central theoreti-cal place of language in his perspective and the foundational role Mead’s preexistingdiscussions played in his theoretical orientation. For Mead, animals, lacking the abil-ity to employ linguistic symbols, acted and interacted solely on the more elementaland limiting basis of gesture. As he put it:

Gestures may be either conscious (significant) or unconscious (nonsignificant).The conversation of gestures is not significant below the human level, becauseit is not conscious, that is, not self-conscious (though it is conscious in the senseof involving feelings or sensations). An animal as opposed to a human form, inindicating something to, or bringing out a meaning for, another form, is not atthe same time indicating or bringing out the same thing or meaning to or forhimself; for the animal has no mind, no thought, and hence there is no mean-ing here in the significant or self-conscious sense. A gesture is not significantwhen the response of another organism to it does not indicate to the first organ-ism what the second organism is responding to. (1964: 168)

Mead believed that animals were mindless since thought was dependent on theactor’s ability to say. Consequently, animals lacked a self since they were unable toengage in the processes of defining situations and taking the role of the other thatconstituted “authentic” social interactions that gave rise to the actor’s self definition.For Mead, human-animal exchanges were premised on anthropomorphic delusion.As he maintained:

We, of course, tend to endow our domestic animals with personality, but as weget insight into their conditions we see there is no place for this sort of impor-tation of the social process into the conduct of the individual. They do not havethe mechanism for it—language. So we say that they have no personality; theyare not responsible for the social situation in which they find themselves. . . . Weput personalities into the animals, but they do not belong to them. . . . And yetthe common attitude is that of giving them just such personalities as our own.We talk to them and in our talking to them we act as if they had the sort ofinner world that we have. (1962: 182–183)3

Blumer’s (and Mead’s) anthropocentric exclusion of animals from the realms ofminded action, meaning construction, and “authentic” interaction has recently beenchallenged both by interactionist scholars actively engaged in investigating human-animal relationships and by writers not directly associated with human-animal stud-ies. Hammersley (1989: 128–129), for example, criticizes Blumer and interactionismfor disregarding animal behavior and ignoring its potential for expanding an under-standing of human motivation. More recently, Prus (1997: 96) has advocated atten-tion to “life form objects” and human definitions of and consequent interactionswith them. Nonetheless, criticism of attending to human-animal exchanges contin-ues both outside of the symbolic interactionist community (e.g., Perrow 2000) and

MIND, SELF, AND HUMAN-ANIMAL JOINT ACTION 321


within it. Recently, Meltzer has dismissed attention to the interactional, mental, andbehavioral abilities of animals as an exercise in anthropomorphizing since animalsengage in behavior, he maintains, that is merely “automatic, immediate, and directreaction of a stimulus-response or conditioned response type” (2003: 258). Similarly,Dennis Waskul, in a listserve exchange focused on animal abilities, recently observed:

[P]ersonally, I simply cannot accept this notion of animals (other than humans)possessing a mind or self. I find some of the recently published works on thisinteresting, but I remain deeply skeptical. It is the human mind that conjuresthese fictions . . . and THAT is the key difference. Because animals do not, andcannot (I argue), wonder what humans (or any other organism) are “thinking,”they do not (and cannot) posses a mind or self. If animals could do these thingsthey wouldn’t act like they do. . . . I think the classic portrayal of these things asuniquely human is quite central and significant. Sure we can interact with ani-mals and animals can act back—but so does my computer, my . . . pickuptruck, and a whole host of utterly inanimate objects (emphases in the original).4

In the following discussion I take at face value Blumer’s (1969: 67) position that“symbolic interactionism is able to cover the full range of the generic forms of humanassociation” and question the linguicentric interactionist orthodoxy that routinelyexcludes animals and their interactions with human coactors from serious attention.Drawing primarily on the ethnographic work of Janet and Steven Alger, Leslie Irvine,Eugene Myers, and Clinton Sanders—analysts of human-animal relationships whoovertly employ an interactionist perspective—I extend Blumerian interactionism toconstruct elements of an interactionist “sociozoology”5 that casts animals as “actingunits,” expands the concept of the “person,” presents mind and self as not based solelyon linguistic capabilities or as uniquely human, and examines forms of joint actionthat involve intersubjective exchanges between humans and other animals. In sodoing, I take to heart what Hammersley (1989: 131–135) refers to as Blumer’s “crit-ical commonsensism,” which regards commonsense as providing a useful and reliableentry into the world of everyday experience. I maintain that it is through the analyticand methodological valuing of commonsense and the practical ways in which itshapes our understanding of situations and the interactions that take place withinthem that we can best construct a view of social life that is immediate, comprehensi-ble, and includes all of the associations that constitute it.

ANIMAL MINDIn constructing an understanding of mind and mindedness, Blumerian interaction-ism rejects both dualism (the analytic separation of the actor’s body and mind) andindividualistic psychologies in which social processes are viewed as produced by mind(Meltzer 2003: 253–254). Further, interactionism rejects a computational or infor-mation-processing view of mind that focuses on the manipulation of abstract mentalrepresentations to the exclusion of emotions, body awareness, and other elements ofminded experience (Dutton and Williams 2004: 212, 217). Finally, interactionism



rejects the separation of mind from self and society in that mindedness presupposes aself that arises out of the social exchanges that constitute society (Blumer 1969:78–89; Meltzer 2003: 254–255; Prus 1996: 52–54). Nonetheless, Blumer, like Meadbefore him and most interactionists who followed, largely regards mind as an inter-nal conversation premised on the minded actor’s ability to employ significant symbolsas the medium of interaction with the self. As Mead (1962: 47) put it: “Only in termsof gestures as significant symbols is the existence of mind or intelligence possible; foronly in terms of gestures which are significant symbols can thinking—which is sim-ply an internalized or implicit conversation of the individual with himself by meansof such gestures—take place.”

For over a decade, a growing number of interactionist sociologists have sought toestablish an orientation toward mind that deemphasizes this view of mindedness as alinguistic phenomenon and returns to an understanding of mind as the outcome ofsocial interaction and social experience. Examining the interactional worlds of peoplewith Alzheimer’s disease (Gubrium 1986), those with severe physical and mental dis-abilities (Bogdan and Taylor 1989; Goode 1994), infants (Kaye 1982; Stern 1985), andnonhuman animals, researchers have called into question the centrality of language useto mindedness and have emphasized the interactional process of “doing mind” (Duttonand Williams 2004; Sanders 1993). As Dutton and Williams (2004: 215–216) observe:

The attribution of meaning or intention to behavior hinges crucially on theextent to which such behavior is considered meaningful within the context of thesocial relationship. Social relationships actually provide rather clear conditionsand parameters for what constitutes “mindful” behavior in contrast to thosebehaviors that do not seem to merit an intentional explanation because theyseem inappropriate within the context of relationship. . . . To see (doing mind)as simply folk psychology or a useful social heuristic would be to ignore theimportance of the social relationship in structuring and scaffolding intersubjec-tive understanding (emphasis in the original).

Mind, therefore, as it arises from shared experience, is an element of the meaningstructure that those who interact with alingual others devise in understanding andconstructing their interactions—an eminently Blumerian position. Caretakers of ani-mals, people with severe disabilities, infants, and Alzheimer’s patients construct a “the-ory of mind” that allows them to understand the thinking, emotions, preferences,desires, and intentions of the other (see Alger and Alger 1997; Myers 1998: 99–102;Sanders 1993).6 Once established in the context of an ongoing relationship, these care-takers commonly use their knowledge of the other to give voice to the mental contentof their alingual coactors. Arluke and Sanders (1996: 61–81) offer a variety ofinstances of this intersubjective activity, a process they refer to as “speaking for.”7 Here,for example, is a quote from Sanders’s ethnographic observations in a veterinary clinic.

A young male shepherd with long hair is brought into the exam room by anolder couple. . . . The woman goes on at some length about the dog’s long hairand how they hadn’t anticipated this when the dog was a puppy. . . . The dog liesdown with his head on the woman’s feet and she says, “Oh, I’m so tired. I just



have to lie down here.” Later, as the dog is having his nails trimmed, she (again)speaks for the dog in observing, “Oh, I have such nice nails.” During much ofthe time she holds the dog’s head tenderly and stares into his eyes. (69)8

The skeptic would, most likely, acknowledge that humans attribute mental activi-ties and use these attributions to give voice to the “presumed” thought of a wide vari-ety of nonhuman “objects.” Disagreement would revolve around whether animalspossess a theory of mind to orient their interactions with human and nonhumancoactors. Cognitive ethologists working with primates have amassed considerable evi-dence that these “border animals” do indeed hold representations of the intentions,goals, and perspectives of their colleagues (e.g., Cheney and Seyfarth 1990; Povinelli,Nelson, and Boysen 1990). It is difficult to understand how primates, or other highlysocial beings such as the animals we accept as our companions, could survive verylong were they unable, at least at some rudimentary level, to “take the role of theother” and use this theory of mind to shape action and interaction.

Take this simple example. My dogs have learned to ring a bell hanging from thedoor to their fenced-in yard to indicate that they want to go outside.9 It is notuncommon for me not to be immediately responsive to their signaled demands if Iam working at the computer or otherwise immediately occupied. After ringing thebell with increased vigor, they will pad down the hall, look fixedly at me as I sit at mydesk, and return to the door to ring the bell once again. If this tactic proves ineffec-tive, they commonly come back to my study and poke me with a nose as if to say(speaking for), “Are you stupid or deaf?” One could, of course, understand thisexchange as premised on simple principles of stimulus and response.10 However, if Itake Blumer (1969: 153–170) at his word and agree that commonsense offers thepragmatic understanding of social experience central to symbolic interactionism, it isentirely sensible for me to see my dogs’ actions as demonstrating a clear definition ofthe situation, a minded ability to ascertain and affect the content of my own mind,the recognition that they can affect my action by acting in a particular way, and thecapability to devise and act upon a plan of action directed toward achieving anintended outcome.11 In short, from a commonsense perspective, “minding animals,”to use Bekoff ’s (2002) multilayered term, involves the cooperative construction ofnonhuman mind as an intersubjective accomplishment that is not dependent uponthe animal’s ability to employ human language.12 As the animal’s “wordless mind”(Terrace 1987) is produced and understood as an outcome of interaction, bothhuman and animal coactors are able to engage in the practical and elemental socialdance that Blumer (1969: 70–77) refers to as “joint action” (a topic upon which Ifocus in a later section of this discussion).

ANIMAL SELFIn order for minded activity to exist, the minded actor must possess a self-object withwhom to communicate. Symbolic interactionism locates the generation of the self,like the construction of mind, in the context of social interaction. Conscious awareness



entails the actor indicating “things” to his or her self, devising an understanding of theobjects so indicated, and using this understanding to shape lines of action (Blumer1969: 12–13). Even analysts skeptical of the ability of commonsense to provide validand/or practical representations of experience (e.g., Hilbert 1994) would have a dif-ficult time denying that animals are conscious in this sense and routinely involved inthis type of interpretive process (see Damasio 1999; Dawkins 1998).

Although the animal self is a focus of attention for most interactionists involved inexploring human-animal relationships (e.g., Alger and Alger 2003: 76–89; Sanders1999: 9–15, 137–140), Irvine (2004a; 2004b) and Myers (1998) offer the mostdetailed discussions of the issue. Both locate the animal self in the context of the phys-ical body (cf., Cooley 1964: 168–200) and emotional experience; see it as emergingfrom, and displayed through, goal-directed behavior (Irvine 2004b: 8; Myers 1998:49–50); and deemphasize the importance of linguistic ability (Irvine 2004b: 9; Myers1998: 6–8). Like mind, the self is presented as a product that emerges through rou-tine interactions within the context of relationships.

Using a conceptual foundation provided by William James (1961 [1892]) and fol-lowing the lead of Stern (1985) and Myers (1998: 74–78), Irvine (2003; 2004a:126–146; 2004b) details four “self-experiences” that are independent of language,arise out of interaction, and generate the “core self.” As she puts it:

In the model of self I am using, in order for animals to [participate in the cre-ation of self ], they must themselves be subjective Others. How can we sensetheir subjective presence? As with other people, we cannot observe subjectivitydirectly. We perceive it indirectly, during interaction. . . . [T]he case can bemade for the presence of these [self ] experiences among animals, who have thesame structures of the brain, nervous system, musculature, and memory [as dohumans]. Whereas human development takes us into a stage of language acqui-sition that adds to these basic experiences, the experiences themselves are pre-verbal. (2004b: 9, emphasis in the original)

Irvine’s first self-experience is “agency,” the actor’s recognition that he/she/it hascontrol over his/her/its actions. This awareness leads to an understanding that one isdifferentiated from the other and that, through purposive action, one can have impacton others’ actions. For example, my Newfoundlands, like most dogs, are fond of rou-tine. An academic schedule allows me on most days to take my dogs for their dailywalk between ten and noon. Around this time, the dogs become restive and begin tofollow me around as I move about the house, all the while watching closely for indi-cations that I am making walk preparations. As it grows later, the youngest dog willescalate the exchange by occasionally poking me with her nose and whining. The antic-ipation and attempts to hurry me along persuasively demonstrate that the dogs notonly have some sense of time but that they also recognize that their actions can shapemy behavior in ways they define as desirable. In short, they provide evidence that theyrecognize their own agency and are able to take the role of the (human) other.

Irvine’s second self-experience, “coherence,” focuses attention on the actor’s physi-cal self, “the entity to which . . . agency belongs” (2004a: 133). She emphasizes that



coherence is demonstrated by the animal’s ability to recognize, and orient actiontoward, specific animal and human others and by the activity of self-concealment.Hiding is an “indication that an animal is aware of itself as an object in the world sep-arate from the physical environment and from others” (Sanders 1999: 137). Withinthe human-animal relationship, the coherence of the unique animal self is solidifiedand symbolized through the cultural process of naming (Irvine 2004b: 12; Sanders2003: 411).

“Affectivity”—possessing feelings/emotions associated with experience—is the thirdelement of the core self discussed by Irvine. Here she distinguishes between “categor-ical affects” (self-experiences such as happiness, sadness, anger, fear, and so forth thatwe typically think of as emotion) and “vitality affects” (ways of feeling such as moodand vitality). Based on knowledge of the companion animal’s common emotionalresponses to particular situations and the ability to read the bodily gestures associatedwith these responses, the caretaker is able to understand and behave appropriatelytoward the animal and, over time, construct an understanding of what the animal is“like” (Irvine 2004a: 137–139). Through interactional experience with the animal-other’s affective self, caretakers come to see the animal has having a unique and iden-tifiable “personality.” As one of Sanders’s (1999: 30) interviewees put it: “The onlyway I can describe the relationship [with my dog] is to use human terms. He certainlyhas as much personality as any of the people I know. I just use many of the same indi-cators to know what he is thinking and feeling that I use to understand my wife orother people.”

Finally, Irvine emphasizes “self-history” as a central organizing element of the animal’score self. Past events and experiences, knowledge of others and places are, she maintains,preserved in nonverbal memory and provide the basis for self-in-relationship. Inanswering the presumed behaviorist criticism that responses to past experience, suchas an animal’s display of fear when entering the veterinarian’s office, are simply theresult of conditioning, Irvine calls upon Allen and Bekoff ’s (1997: 56–62) distinctionbetween behavior that is “stimulus bound” (i.e., motivated by external stimuli) andthat which is “stimulus free” (i.e., prompted by internal factors). Despite the seduc-tive simplicity of stimulus-bound explanations of animal behavior, she argues thatmany of the experiences we have with animals persuasively indicate that their behav-ior is frequently prompted by memory and evaluation of the immediate situation. Asan example, she observes that her cats prefer sleeping on a fleece blanket and acknowl-edges that the “comfort” they associate with the blanket is, to some degree, the con-sequence of conditioning. However, were she to lay out the blanket in the veterinaryexamining room (a setting in which her cats exhibit considerable fear), she doubtsthat they would respond to it in the same way they do when it rests on the floor inher home (Irvine 2004b: 14–15).

Shared routines also are a key element of self-history. Routine activities—especiallyfeeding and play—constitute temporal high points in the shared flow of events thatconstitute the regular interactions between caretakers and companion animals. Theseroutines, then, come, over time, to be anticipated, remembered, and ritualized withinthe relationship. For example, I always feed my own dogs between seven and eight in



the morning. As in the case of going for walks described above, when this timeapproaches the dogs begin to watch me carefully and to follow me from place toplace. If I am late, they will begin to make “impatient” noises and look longingly atthe metal food bowls kept on the kitchen counter. When I take the bowls into thebasement where their food is stored, they wait expectantly at the top of the stairs andthen mill about and bark with anticipation as I prepare their meals. Over time,Cynder, the older of my Newfoundlands, has devised a unique postprandial routine.After I put the other dog out to take care of elimination, Cynder takes her bowl overto a place by the couch and waits expectantly. I go to her, rub her ears, and say a fewkind words. After this special attention she lies down and chews on her bowl for a fewminutes until she falls asleep. I then return her bowl to the counter and let the otherdog back into the house. It is clear from her demeanor that this is an enjoyable andexpected routine. If I am distracted or busy and do not follow the established pattern,she will continue to sit with the bowl in her mouth and watch me until I realize myfailure to meet her expectations. This routine, together with a number of others, is acomponent of our shared history. We all remember how routines have played out inthe past and structure present interactions on the basis of these memories, we antici-pate the exchanges in the future, and Cynder’s personal shaping of the routine is partof my knowledge of her unique selfhood (cf. Shapiro 1990).

The animal self, therefore, is constituted by the animal-actor’s recognizing he orshe is distinct from others and can engage in volitional behavior; that moods and feel-ings are connected to experiences, places, and relationships; and that past events arerelevant to the emotional experience of and physical response to current situations. Inthe context of an ongoing relationship, caretakers construct an understanding of theanimal’s self and use this understanding to interact appropriately and practically. Likehuman-with-human interaction, exchanges between people and their animal com-panions is a mutual interplay of discrete selves. Seen in this way, the self is a processwith a past, present, and future and an intersubjective accomplishment rather thansimply an object that an actor “indicates.” As Clark (1984: 47) explains:

To have . . . a concept [of self ] one must recognize other beings in the world,distinguish what is self and what is not-self, and lay claim to pasts and futuresof the single being one is. To have a concept of self is, in part at least, to do thesethings: to show awareness of the world as being more than one’s immediate per-ception of it, to recognize other beings as being the same as some past acquain-tances, to admit responsibility for past action, and to intend some future. (Seealso Myers 1998: 58–61.)

ANIMALS AND JOINT ACTIONFor Blumer, society is constituted by interaction (Blumer 1969: 7–8). FollowingMead, Blumer distinguishes between symbolic and nonsymbolic interaction. In thelatter type cointeractants engage in a “conversation of gestures” in which there isshared meaning and mutual role-taking, but interactants respond directly to each



other’s behavior in the process of cooperatively constructing joint action. Accordingto Blumer (1969: 9): “the gesture has meaning for both the person who makes it andfor the person to whom it is directed. When the gesture has the same meaning forboth, the two parties understand each other. . . . [The gesture] signifies what the per-son to whom it is directed is to do; it signifies what the person who is making the ges-ture plans to do; and it signifies the joint action that is to arise by the articulation ofthe acts of both.”

Blumer, like Mead, would maintain that animal-with-animal, human-with-animal,and some simple forms of human-with-human interaction are premised on this com-municative foundation. But, for Blumer, most human interaction is symbolic in thatit involves interactants subjectively interpreting the meaning of each other’s behaviorsand acting (rather than responding) on the basis of this understanding. The conse-quence of this symbolic exchange is a more complex form of “interlinked” action thatis repetitive, stable, and grounded on emergent rules (Blumer 1969: 19). This“authentic” form of joint action is made possible as interactants share basically simi-lar definitions of the situation and possess an understanding of each other’s goals,plans of action, and definitions of objects within the situational context (Blumer1969: 62, 96–97). In other words, for Blumer joint action is a dance of mindedselves—an interpretive fitting together of lines of action (Blumer 1969: 70–77)—operating in the context of shared perspectives (cf. Prus 1997: 6–8).

Again, I maintain that Blumer’s exclusion of nonhuman animals from the realm of“authentic” (i.e., human-like) joint action is based simply on anthropocentric andlinguicentric articles of faith. Belief in the unique character of human joint action ison shaky ground given the growing body of research-based literature being producedby interactionist scholars concerned with human-animal relationships and interac-tions. This work clearly demonstrates that, within the context of the ongoing, histor-ically constructed, emotionally rich relationships between people and their animalcompanions, humans and animals do share perspectives and are routinely involved injoint action. The ability of animals to take the role of the other and act accordinglyin order to achieve certain goals is illustrated by a veterinarian’s description of how hisdog deceptively defines a particular definition of the situation for personal gain.

I believe that dogs think. My dogs play a game called “bone.” One of them willget the rawhide bone and take it over to the other one and try to get him to tryand get it. Or one will try to get the bone if the other one has it. One day I waswatching and the youngest one was trying to get the bone without much luck.So he goes over to the window and begins to bark like someone is coming upthe driveway. The other dog drops the bone and runs over to the window andthe puppy goes and gets the bone. There wasn’t anyone in the driveway, it wasjust a trick. (Sanders 1999: 24)

The ability of dogs and other companion animals to share perspectives and sit-uational definitions in both animal-with-animal (as in the above) and human-with-animal situations is demonstrated by what Myers (1998: 93–95) refers to as“interattentionality,” the sharing of attention and bodily experience. One indication



of animals’ abilities to assume the perspective of the other is seen in the common phe-nomenon of gaze-following. Researchers have frequently noted this empathetic activ-ity in observations of peoples’ interactions with dogs. For example, Sanders (1999:114) notes the importance of gaze in his autoethnographic observations of his expe-rience with his own dogs.

When Emma and Isis look at me they usually pay attention to my eyes. I havenoticed on walks how important looking is to them. A common way that onewill communicate to the other that she wants to play is by staring. During playthey have a variety of ways of signaling “time out.” In addition to stopping andavidly sniffing someplace, a player can effectively suspended the game by star-ing fixedly off into the middle distance. The other dog typically responds to thismove by looking to see if there is actually anything important to look at. Theydo the same with me. If on the walk I stop and look in a particular direction,they will stop, glance at me, and gaze off in the direction I am looking. Thisseems a fairly clear indication of their elemental ability to put themselves intomy perspective. In a literal sense they attempt to assume my “point of view.” IfI look at something they conclude that it is probably something important.13

In addition to the “social force” (Dutton and Williams 2004: 218) of mutual gazeand shared attention, sociologists involved in human-animal studies have focused onplay as the interactional situation in which joint action is readily apparent. Playinvolves interaction that is apparently purposeless; mutually regarded as separatedfrom everyday, practical, “serious” endeavors; and pursued in order to achieve intrin-sic pleasure instead of being directed toward instrumental ends (see Irvine 2001:153–154; 2004a: 166–171).14 Mechling (1989: 313), for example, emphasizesanimal-human play as a communicative interaction when he describes playing withhis dog Sunshine.

The game of fetch was truly interactive. I was not always in control of the game.Sometimes Sunshine would fetch the ball but stop on the way back to me someten feet away. He would begin a slow retriever stalk, then drop the ball in frontof him and assume the familiar canine “play bow”—forepaws extended flat onthe ground, the body sloping upward toward his erect hindquarters, tail wag-ging. This is the canine invitation to play. In this case, however, we were alreadyengaged in a game, so his message to me was that he, too, could exert somepower and control in the game.15

We can see that within the mutually defined situational “frame” (Goffman 1974)of play—in addition to feeding (Alger and Alger 2003: 57–58, 63–64; Irvine 2004a:153–155), the solicitation and acceptance of physical contact, and the other rou-tinized exchanges between people and their animal companions show the basic ele-ments of joint action described by Blumer. Here human and animal cointeractants“[align their] individual actions . . . by interpreting or taking into account each other’sactions” (Blumer 1969: 82). In so doing, people and animals develop and play outwhat Collins (1989) calls “natural interaction rituals” as they share attention, mutualawareness, emotions, and obligations.



Within the context of this joint action people and animals create, in turn, a “pri-vate culture” (Fine 1981: 267) consisting of established routines, knowledge of andfeeling for each other, and expectations about the course of interaction. As Smuts(2001: 302) observes, “with . . . prolonged exposure, members of two different speciescan co-create shared conventions that help to regulate interspecies encounters.” Whatis of central importance to this interactionally constructed culture and the collectiveaction it encompasses is that, although these structures and processes involve sym-bolic meanings shared by both humans and animals, they exist exclusive of the abil-ity of all interactants to employ linguistic symbols.

CONCLUSIONBased on the basic precepts offered by Blumer, symbolic interactionism offers both ananalytic window through which to view social life and methodological guidelines foropening that window. Through direct personal engagement with the social worldconstituted by interaction (Blumer 1969: 24–26), taking the role of “acting units”(be they human or nonhuman), and attending to personal experiences within inter-actional situations (Blumer 1969: 39, Prus 1996: 68–69), we build a descriptive(albeit tentative) understanding of how social exchanges are structured and played out(Blumer 1969: 18; Prus 1997: 9). Of central importance is the Blumerian principlethat everyday (“reliable”) commonsense (Hammersley 1989: 126), when combinedwith intimate familiarity with the actors and phenomena of interest, provides an ade-quate foundation for this understanding.16

Given Blumer’s faith in what Hammersley (1989: 131–135) refers to as “criticalcommonsensism” and his belief that symbolic interactionism provides an adequateground for understanding all forms of social relationships (Blumer 1969: 67), it seemseminently reasonable to conclude that building an interactionist view of human-animalassociations is both a viable endeavor and a means by which the interactionist perspec-tive may be developed, enlarged, and changed (Hammersley 1989: 219). Central to thisworthy endeavor are the methodological approach and analytic perspective related toBlumerian commonsensism. Interactionists working in human-animal studies consis-tently advocate and employ what cognitive ethologists call “critical” or “interpretive”anthropomorphism (Burghardt 1991). This orientation acknowledges that all system-atic understanding of social others flows from the analyst’s adopting the perspective ofthose with whom he or she interacts. Through imaginatively taking the role of animal-others and critically employing insights gained from one’s personal experience with andknowledge of the intersubjective groundings of everyday (human) social life, we maybuild a rich and nuanced understanding of how nonhuman minds and selves are con-structed and operate within the context of intimate exchanges with their human com-panions. For, as the ethologist Margaret Nice succinctly puts it: “A necessary conditionfor success . . . is a continuous sympathetic observation of an animal under as naturalconditions as possible. To some degree, one must transfer oneself into the animal’s situ-ation and inwardly partake in its behavior” (quoted in Lawrence 1989: 118).17



Interactionist scholars who focus on human relationships with other animals have,I submit, taken Blumer and his interpreters at their word. Possessing “informal andintimate knowledge” (Costall 1998) of their own and others’ everyday experiencewith animals—what Cox and Ashford (1998) refer to as “epistemic authority”—thesescholars are constructing an enlarged understanding of group life within a mixed-species social world. What has emerged, as demonstrated above, is a portrait of non-human animals as minded and self-aware participants in collective action with theirhuman associates.

But exploring human interactions with nonhuman animals holds promise formore than simply shedding substantive light on human-animal relationships.Analytic attention to people’s interactions with animals offers a route to expandingthe conventional sociological conception of mind as a language-dependent, inter-nalized conversation. Instead, mind may be regarded as an intersubjective accom-plishment, the outcome of interaction between intimates (Sanders 2003). Similarly,exploration of human-animal interaction has already reshaped, and will continue toexpand, the sociological understanding of the self. The process of constructing a selfand the emotionally rich experience of self-awareness can no longer be seen asdepending upon a social actor’s ability to use words. Instead, the self may be under-stood as arising from the actor’s recognition of his/her/its ability to control the phys-ical and social world, experience emotions, and realize that purposive action comesfrom a unique and ongoing entity. In addition, as self aware social actors, nonhumananimals come to act as what Irvine (2004a: 162–166) refers to as “resources for selfconstruction.” Interaction with animals provides people with information and expe-riences they can use to define the roles they enact and the situations in which theyfind themselves.

Seeing animals as minded selves—as participants in the interactional process of“doing mind” (Dutton and Williams 2004) and as social actors whose personhood isconstructed within this process (Sanders 1995)—also demands an ethical reevalua-tion of our relationships with them.18 Knowing animals and their central place in ourculture and everyday experiences leads, of necessity, to appreciating their abilities,including them in sociological discussion, and affording them the respect and care weoffer to the others with whom we share our lives.

Clinton R. Sanders is a professor in the sociology department at the University ofConnecticut. He has served as president (2002–2003) and vice president(1994–1995) of the Society for the Study of Symbolic Interaction. Sanders’ workfocuses on cultural production, deviant behavior, ethnographic research, and socio-zoology. His most recent books are Understanding Dogs: Living and Working withCanine Companions (Temple University Press 1999) and Regarding Animals (TempleUniversity Press 1996, coauthored with Arnold Aluke), both of which received theCharles Horton Cooley Award given by the Society for the Study of SymbolicInteraction. Sanders is coeditor of the Temple University Press series on “Animals,Culture, and Society” and an associate editor of Society and Animals. He was the recip-ient of the University of Connecticut Provost’s Award for Research Excellence in 2004



and received the Distinguished Scholarship Award from the Animals and Society sec-tion of the American Sociological Association in 2006.

NOTES1. Maines (2001) persuasively argues that symbolic interactionism is not a “unique” perspec-tive in the sense that its theoretical precepts are central to sociology.2. Prus (1997:11–17) expands these basic principles.3. For extended critical discussions of Mead’s perspective on animals and his single-mindedfocus on language see Alger and Alger 1997: 67–71, 2003: 203–206; Irvine 2003, 2004a:120–125, 2004b; Myers 1998: 120–125, 2003; Sanders 1999: 117–119.4. Retrieved January 4, 2005, from http://venus.soci.niu.edu/~achives/SSITALK/aug04/0276.html.5. Arluke (2003) prefers the older term “ethnozoology.” I see “sociozoology,” “anthrozoology,”and “ethnozoology” as equally descriptive and serviceable.6. The concept of “theory of mind” was originated by the primatologists Premack andWoodruff (1978).7. Goffman (1981: 86–87) calls this activity “say-foring.” He recognizes that say-foring is ageneral feature of relationships between those who are competent and those who are definedas less, or minimally, competent. He uses infants and companion animals as examples.

[W]e can with impunity address words in public to a pet, presumably on the grounds thatthe animal can appreciate the affective element of the talk, if nothing else. We extend thesame sort of regard to infants. . . . Moreover, special forms of talk are involved: for exam-ple, . . . mimicked babytalk projected as the talk the incompetent would employ were itable to speak [“say-foring”].

With reference to the relationship between parents and infants, Kaye (1982: 182) refers to thisminding interaction as the “he says phenomenon.” For an extended discussion of the ethicaland policy issues involved in speaking for incompetents, see Buchanan and Brock (1989).8. For examples of this process within the context of adult-infant interaction see Brazelton(1984) and in interactions between caretakers and people with severe disabilities see Pollnerand McDonald-Wikler (1985). Myers (1998) provides a variety of examples of children speak-ing for the animals they encountered in a nursery school. Also see Irvine’s (2001) recent dis-cussion of this “minding” phenomenon in the context of human-animal play.9. Interestingly, for new dogs in my household this knowledge is not the result of my train-ing. Young dogs have learned from observing more experienced animals what the bell “means”and the typical human activity ringing it produces.10. But, as Hearne (1987:58) observes:

To the extent that the behaviorist manages to deny any belief in the dog’s potential forbelieving, intending, meaning, etc., there will be no flow of intention, meaning, believing,hoping going on. . . .The behaviorist’s dog will not only seem stupid, she will be stupid.

11. For extended discussions of animal mind, see Crist 1999; Griffin 1992; Hauser 2000; andRogers 1997.12. The psychologist and dog trainer Stanley Coren observes that his own dogs have a “recep-tive vocabulary” of around sixty-five words or phrases (Coren 1994:114). On the basis of the



dog intelligence tests he has devised he finds that, relative to humans, dogs have an IQ some-where in the 60s.13. This is related to the gaze-related activity that Irvine (2004a:155–158) refers to as “check-ing back.” See also Dutton and Williams 2004:214–215.14. Erving Goffman (1997) emphasizes the central role of playing games and having fun inhuman social life. He describes games as socially sanctioned, nonserious, face-to-faceexchanges in which interactants share situational definitions. 15. For additional discussion and examples of human-animal play see Irvine 2004a: 168–171;Mitchell and Edmonson 1999; Myers 1998: 117–143. Shapiro (1990: 186) emphasizes theinteractional complexity and intersubjective nature of human-animal play in his observationsof playing with his own dog:

At first glance, this activity hardly seems worth mentioning—a simple game of keep-away and chase. However, on reflection, the invitation ritual, the play itself, the implicitrules and regulations governing the legitimate play area, permitted and prohibitedmoves, and the object of the game, as well as the conditions ending it, were all quiteintricate and yet easily maintained by both players. Even more subtle were the postures,feints, and deceits, the half-executed moves during what might look, to the uninitiated,like time-outs from the game.

16. Similarly, Silverman (1997) advocates replacing the anthropocentric “commonsense of sci-ence” with “ordinary commonsense” derived from everyday experience when attempting tounderstand the perspectives of nonhuman animals and our interactions with them. 17. For discussions of “critical anthropormorphism” as it is advocated and used within socio-logical studies of human-animal relationships, see Irvine (2004a:68–73) and Sanders (1999:140–141). See also Shapiro’s (1990) related discussion of “kinesthetic empathy.” 18. For discussions of the moral and ethical implications of an interactionist understanding ofnonhuman animals see Alger and Alger (2003: 199–211); Arluke and Sanders (1996: 52–57);and Irvine (2004a: 172–184).

REFERENCESAlger, Janet and Steven Alger. 1997. “Beyond Mead: Symbolic Interaction between Humans

and Felines.” Society and Animals 5: 65–81.———. 2003. Cat Culture: The Social World of a Cat Shelter. Philadelphia, PA: Temple

University Press.Allen, Colin and Marc Bekoff. 1997. Species of Mind. Cambridge: Massachusetts Institute of

Technology Press.Arluke, Arnold. 2003. “Ethnozoology and the Future of Sociology.” International Journal of

Sociology and Social Policy 23: 26–45.Arluke, Arnold and Clinton Sanders. 1996. Regarding Animals. Philadelphia, PA: Temple

University Press.Bekoff, Marc. 2002. Minding Animals: Awareness, Emotions, and Heart. Oxford, UK: Oxford

University Press.Blumer, Herbert. 1969. Symbolic Interactionism. Englewood Cliffs, NJ: Prentice-Hall.Bogdan, Robert and Steven Taylor. 1989. “Relationships with Severely Disabled People: The

Social Construction of Humanness.” Social Problems 36: 135–148.



Brazelton, T. Berry. 1984. “Four Stages in the Development of Mother-Infant Interaction.”Pp. 19–34 in The Growing Child in Family and Society, edited by Noboru Kobayashi andT. B. Brazelton. Tokyo: University of Tokyo Press.

Buchanan, Allen and Dan Brock. 1989. Deciding for Others: The Ethics of Surrogate DecisionMaking. Cambridge, UK: Cambridge University Press.

Burghardt, Gordon. 1991. “Cognitive Ethology and Critical Anthropomorphism: A Snakewith Two Heads and Hog-Nosed Snakes that Play Dead.” Pp. 53–90 in CognitiveEthology, edited by Caroline Ristau. Hillsdale, NJ: Lawrence Erlbaum.

Cheney, Dorothy and Robert Seyfarth. 1990. How Monkeys See the World. Chicago, IL:University of Chicago Press.

Clark, Stephen. 1984. The Nature of the Beast. New York: Oxford University Press.Collins, Randall. 1989. “Toward a Neo-Meadian Theory of Mind.” Symbolic Interaction 12: 1–32.Cooley, Charles Horton. 1964. Human Nature and the Social Order. New York: Shocken.Coren, Stanley. 1994. The Intelligence of Dogs. New York: Free Press.Costall, Alan. 1998. “Lloyd Morgan and the Rise and Fall of ‘Animal Psychology.’” Society and

Animals 6: 13–29.Cox, Graham and Tony Ashford. 1998. “Riddle Me This: The Craft and Concept of Animal

Mind.” Science, Technology, and Human Values 23: 425–438.Crist, Eileen. 1999. Images of Animals: Anthropomorphism and Animal Mind. Philadelphia, PA:

Temple University Press.Damasio, Antonio. 1999. The Feeling of What Happens: Body and Emotion in the Making of

Consciousness. New York: Harcourt Brace.Dawkins, Marian Stamp. 1998. Through Our Eyes Only? The Search for Animal Consciousness.

Oxford, UK: Oxford University Press.Dutton, Diane and Carl Williams. 2004. “A View from the Bridge: Subjectivity, Embodiment,

and Animal Minds.” Anthrozoös 17: 210–224.Fine, Gary A. 1981. “Friends, Impression Management, and Preadolescent Behavior.”

Pp. 257–272 in Social Psychology through Symbolic Interaction, 2nd ed., edited byG. Stone and H. Farberman, New York: Wiley.

Goffman, Erving. 1974. Frame Analysis. Cambridge, MA: Harvard University Press.———. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press.———. 1997. “Fun in Games.” Pp. 129–139 in The Goffman Reader, edited by Charles

Lemert and Ann Branaman. Malden, MA: Blackwell.Goode, David. 1994. A World without Words: The Social Construction of Children Born Deaf

and Blind. Philadelphia, PA: Temple University Press.Griffin, Donald. 1992. Animal Minds. Chicago, IL: University of Chicago Press.Gubrium, Jaber. 1986. “The Social Preservation of Mind: The Alzheimer’s Disease

Experience.” Symbolic Interaction 9: 37–51.Hammersley, Martyn. 1989. The Dilemma of Qualitative Method: Herbert Blumer and the

Chicago Tradition. New York: Routledge.Hauser, Marc. 2000. Wild Minds. New York: Holt.Hearne, Vickie. 1987. Adam’s Task. New York: Knopf.Hilbert, Richard. 1994. “People Are Animals: Comment on Sanders and Arluke.” Sociological

Quarterly 35: 533–536.Irvine, Leslie. 2001. “The Power of Play.” Anthrozoös 14: 151–160.———. 2003. “George’s Bulldog: What Mead’s Canine Companion Could Have Told Him

about the Self.” Sociological Origins 3: 46–49.



———. 2004a. If You Tame Me: Understanding Our Connection with Animals. Philadelphia,PA: Temple University Press.

———. 2004b. “A Model of Animal Selfhood: Expanding Interactionist Possibilities.”Symbolic Interaction 27: 3–22.

James, William. 1961 [1892]. Psychology: A Briefer Course. New York: Harper.Kaye, Kenneth. 1982. The Mental and Social Life of Babies. Chicago, IL: University of Chicago

Press.Lawrence, Elizabeth. 1989. “Wild Birds: Therapeutic Encounters and Human Meanings.”

Anthrozoös 3: 111–118.Maines, David. 2001. The Faultline of Consciousness. New York: Aldine de Gruyter.Mead, George Herbert. 1962. Mind, Self, and Society. Chicago, IL: University of Chicago Press.———. 1964. George Herbert Mead on Social Psychology, edited by Anselm Strauss. Chicago,

IL: University of Chicago Press.Mechling, Jay. 1989. “‘Banana Cannon’ and Other Folk Traditions between Human and

Nonhuman Animals.” Western Folklore 48: 312–323.Meltzer, Bernard. 2003. “Mind.” Pp. 253–266 in Handbook of Symbolic Interactionism, edited

by Larry Reynolds and Nancy Herman-Kinney. Walnut Creek, CA: Altamira.Mitchell, Robert and Elizabeth Edmonson. 1999. “Functions of Repetitive Talk to Dogs dur-

ing Play: Control, Conversation, or Planning?” Society and Animals 7: 55–81.Myers, Eugene. 1998. Children and Animals. Boulder, CO: Westview.———. 2003. “No Longer the Lonely Species: A Post-Mead Perspective on Animals and

Self.” International Journal of Sociology and Social Policy 23: 46–68.Perrow, Charles. 2000. “An Organizational Analysis of Organizational Theory.” Contemporary

Sociology 29: 469–476.Pollner, Melvin and Lynn McDonald-Wikler. 1985. “The Social Construction of Unreality: A

Case of a Family’s Attribution of Competence to a Severely Retarded Child.” FamilyProcess 24: 241–254.

Povinelli, Daniel, K. E. Nelson, and S. T. Boyson. 1990. “Inferences about Guessing andKnowing by Chimpanzees.” Journal of Comparative Psychology 104: 203–210.

Premack, David and G. Woodruff. 1978. “Does the Chimpanzee Have a Theory of Mind?”Behavioral and Brain Sciences 1: 515–526.

Prus, Robert. 1996. Symbolic Interaction and Ethnographic Research. Albany: State Universityof New York Press.

———. 1997. Subcultural Mosaics and Intersubjective Realities. Albany: State University ofNew York Press.

Rogers, Lesley. 1997. Minds of Their Own: Thinking and Awareness in Animals. Boulder, CO:Westview.

Sanders, Clinton. 1993. “Understanding Dogs: Caretakers’ Attributions of Mindedness inCanine-Human Relationships.” Journal of Contemporary Ethnography 22: 205–226.

———. 1995. “Killing with Kindness: Veterinary Euthanasia and the Social Construction ofPersonhood.” Sociological Forum 10: 195–214.

———. 1999. Understanding Dogs: Living and Working with Canine Companions.Philadelphia, PA: Temple University Press.

———. 2003. “Actions Speak Louder than Words: Close Relationships between Humans andNonhuman Animals.” Symbolic Interaction 26: 405–426.

Shapiro, Kenneth. 1990. “Understanding Dogs through Kinesthetic Empathy, SocialConstruction, and History.” Anthrozoös 3: 184–195.



Silverman, Paul. 1997. “A Pragmatic Approach to the Inference of Animal Mind.”Pp. 170–188 in Anthropomorphism, Anecdotes, and Animals, edited by Robert Mitchell,N. Thompson, and H. L. Miles. Albany: State University of New York Press.

Smuts, Barbara. 2001. “Encounters with Animal Minds.” Journal of Consciousness Studies 8:293–309.

Stern, Daniel. 1985. The Interpersonal World of the Infant. New York: BasicBooks.Terrace, Herbert. 1987. “Thoughts without Words.” Pp. 123–137 in Mindwaves: Thoughts on

Intelligence, Identity, and Consciousness, edited by Colin Blakemore and S. Greenfield.Malden, MA: Blackwell.



Notice to Contributors of Manuscripts Submitted to Sociological Focus

Submission of a manuscript to Sociological Focus clearly implies a commitment topublish in this journal. Previously published papers and papers under review byanother journal are unacceptable.

A submission must include the following:

• Two (2) hard copies of the manuscript and one (1) identical copy on disk in Word orWordPerfect, including title page, abstract, text, notes (if absolutely necessary), refer-ences, tables, figures, and illustrations with captions. Manuscripts should be preparedfollowing the American Sociological Association “Style Guide.” Copies must be clearreproductions.

• A cover letter giving the name, mailing address, and e-mail address of the corre-sponding author, title of the manuscript and any other important information. Youmay cite your own work but not in a way that identifies you as the author of thecurrent manuscript. Authors may suggest individuals as potential reviewers, as longas these individuals are not close colleagues or friends. Authors may also suggest indi-viduals to whom they would prefer the manuscript not be sent. You will receiveacknowledgment, via e-mail, of the receipt of your manuscript in the SociologicalFocus office.

• A $15 processing fee, by check or money order payable to Sociological Focus.

Manuscript Preparation

Format and Length: All copy (including cover page, abstract, text, notes, and refer-ences) must be typed, double-spaced on 8-1/2 inch by-11 inch white opaque paper.Left and right margins must be at least 1 inch; top and bottom margins must be atleast 1-1/2 inches. (Bold and italic characters should be noted if not apparent in thetext.) Papers should be a maximum of 35 pages, including text, notes, references,tables, and figures or illustrations. The title page, abstract, text, notes (if any), refer-ences, tables, figures, and illustrations appear, in that order, in separate sections asdescribed below.

1. Title Page: Include the full title, the name(s) of author(s), and the institutionalaffiliation of each author, and a running head. Use an asterisk (*) to add a foot-note to the title to identify the author, with full address, to whom communica-tions about the article should be sent. In the same footnote, provideacknowledgements, credits, grant numbers, and other pertinent informationabout the article.

2. Abstract: Place on a separate page, following the title page, headed by the articletitle and omitting author identification. The abstract should be in italics.

3. Text: Begin text on a new page headed by the title. Omit author identification.


a. Headings and subheadings in the text organize the content. Generally, three levelsof headings are sufficient for an article. Heading styles are shown below (referto the ASA style manual or an ASA publication for additional examples).

THIS IS A LEVEL ONE HEADINGLevel 1 headings are centered on the page, in all capital letters.

This Is a Level Two Heading

Level 2 headings are flush left, in italic capital/lowercase (all words capped exceptprepositions, articles, and coordinating conjunctions).

This is a level three heading. Level 3 headings are indented, run in to the paragraphfollowing them, and end with a period. Capitalize only the first letter and propernouns.

b. References in the text must be listed in the reference section, and all referenceslisted in the reference section must be cited in the text. Cite the last name(s)of the authors and year of publication. Include page references wheneverpossible. References in text must be listed consistently either in chronologicalor alphabetical order. Subsequent citations of the same source are cited in thesame way as the first, with the exception of a reference with three or moreauthors (see the example below). Some examples are listed below (refer torecent issues of Sociological Focus, the ASA Style Guide, or ASA publicationsfor other examples):

• If author’s name is in the text, follow it with the year of publication in parenthe-ses: Fishbein (1975).

• If author’s name is not in the text, enclose last name and year of publication inparentheses: (Benedict 1938).

• Page citation follows year of publication: Antonovski (1979:142) or (Antonovski1979:142).

• Give both names of joint authors (Jessor and Jessor 1977); for three authors, listall three last names in the first citation (Jessor, Chase, and Donovan 1980), and insubsequent citations use “et al.” (Jessor et al. 1980). For citations with four or moreauthors, always use the first author's last name followed by “et al.”

• For institutional authorship, supply minimum identification from the beginningof the complete citation: (U.S. Bureau of the Census 1990:216).

• Separate references in a series with semicolons: (Yamaguchi and Kandel 1984;Johnston 1985; Thompson 1989).

• For unpublished materials: Robin and Robin (forthcoming) or White (unpub-lished) or Johnston (personal communication).

c. Notes in the text should be numbered consecutively throughout the articlewith superscript Arabic numerals. If a note is referred to again later in thetext, use a parenthetical insertion “(see note 3).”


d. Equations in the text must be typed. Important equations should be identi-fied by consecutive Arabic numerals in parentheses to the right of the equa-tion. Expressions should be aligned and subscripts and superscript clearlymarked. Use notes in the margin to clarify symbols.

e. Notes are not required in an article; when possible incorporate the informa-tion into the text.

f. Statistical Terms: All statistical terms, such as t, n, R 2, SD are italicized.

4. Notes: If notes cannot be avoided, list the notes in a section entitled “Notes”directly following the text. Begin each note with the Arabic numeral with whichit is keyed in the text, e.g., “1 This is the first note.” Sociological Focus does notuse footnotes except for the one on the first page about correspondence.

5. Reference List: Authors are responsible for ensuring that publication informationfor each reference must be complete and correct. Listed below are guidelines andexamples for citations in your reference list (see recent issues of Sociological Focus, theASA Style Guide, or ASA publications for further examples):

• List the references alphabetically by the first author's last name. • The author's last name should be inverted. In the case of multiple authors,

only the first author's name should be inverted. (e.g., Dryfoos, Joy D., DavidJ. Hawkins, Denise M. Lishner, and Matthew O. Howard). Use first and lastnames of all authors, unless only initials are given on the cited publication.List all authors in the references: the use of et al. is not permitted in the ref-erence section.

• If your reference list includes more than one item by the same author, listthem in order of the year of publication. For listing more than one work bythe same author(s) within the same year, distinguish them (in alphabeticalorder by title of article) by adding the letters a, b, c, etc. to the year (or toforthcoming, unpublished, etc.).

• If the material listed has been accepted for publication but has not yet been pub-lished, use “forthcoming” in place of the date and give the journal or publisher.

Books Dryfoos, Joy D. 1991. Adolescents at Risk. New York: OxfordUniversity Press.

Periodicals Johnson, Robert H., and Howard B. Kaplan. 1990. “Stability ofPsychological Symptoms: Drug Use Consequences andIntervening Processes.” Journal of Health and Social Behavior31:277–298.

Robin, Stanley S., and Eric O. Johnson. Forthcoming. “Early OnsetDrug Use and the Gateway Phenomenon.” Journal of Health andSocial Behavior 33:157–166.

Yamaguchi, Kazuo and Denise Kandel. 1984a. “Patterns of DrugUse from Adolescence to Young Adulthood: II. Consequencesof Progression.” American Journal of Public Health 74:668–672.


———. 1984b. “Patterns of Drug Use from Adolescence to YoungAdulthood: III. Predictors of Progression.” American Journal ofPublic Health 74:673–681.

E-References Roberts, Les, Pascal Ngoy, Colleen Mone, Charles Lubula, LucMwezse, Mariana Zantop, and Michael Despines. 2003.“Mortality in the Democratic Republic of the Congo: Resultsfrom a Nationwide Survey.” International Rescue Committee.Retrieved November 23, 2004 (http://intranet.theirc.org/docs/drc_mortality_iii_full.pdf ).

Collections Greenspan, Stanley L. 1986. “Research Strategies to IdentifyDevelopmental Vulnerabilities for Drug Abuse.” Pp. 136–154in Etiology of Drug Abuse: Implications for Prevention, edited byCarly L. Jones and Robert J. Battjes. Washington, DC:National Institute on Drug Abuse Series. NIDA ResearchMonograph 56.

6. Tables: Number tables consecutively as referenced in the text, and place each ona separate sheet at the end of the paper. Avoid the use of lines between rows intables. Insert a note in the text indicating the approximate placement of eachtable: {Table 3 about here}. Each table must have a descriptive title and headingsfor all rows and columns (avoid abbreviations). Gather footnotes to tables at thebottom of the respective tables as “Note(s)” and designate each note as a, b, c, etc.,with corresponding designations within and at the bottom of the tables.Asterisks indicate statistical significance [*p � .05; **p � .01; etc.]. All tablesshould provide easily interpreted titles and labels and contain the minimumamount of formatting necessary. Authors are responsible for removing anysuperfluous lines, boxes, or other formatting prior to publication.

7. Figures and Illustrations: Should be numbered in the same way as tables; placeeach on a separate sheet at the end of the paper. Insert a note in the text indi-cating the approximate placement of each: {Figure 3 about here}. Figures andillustrations submitted with the final draft must be camera-ready, executed inblack ink on white paper or vellum; artistic standards must be observed in theproduction of figures and illustrations.



NCSA40.03_C05pg320-342.qxd 7/20/07 3:07 PM Page 342

Documents

a1 SymbolicInteraccionismSocialStructure SocFoc40-3WEB