Creating a Democratic Public Sphere Through Political Discussion: A Case Study of Abortion Conversation on the Internet

http://ssc.sagepub.com/Social Science Computer Review

http://ssc.sagepub.com/content/14/4/373The online version of this article can be found at:

DOI: 10.1177/089443939601400401

1996 14: 373Social Science Computer ReviewSteven M. Schneider

Abortion Conversation on the InternetCreating a Democratic Public Sphere Through Political Discussion: A Case Study of

Published by:

http://www.sagepublications.com

can be found at:Social Science Computer ReviewAdditional services and information for

http://ssc.sagepub.com/cgi/alertsEmail Alerts:

http://ssc.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://ssc.sagepub.com/content/14/4/373.refs.htmlCitations:

What is This?

- Dec 1, 1996Version of Record >>

at DALHOUSIE UNIV on November 9, 2014ssc.sagepub.comDownloaded from at DALHOUSIE UNIV on November 9, 2014ssc.sagepub.comDownloaded from

http://ssc.sagepub.com/

http://ssc.sagepub.com/content/14/4/373

http://www.sagepublications.com

http://ssc.sagepub.com/cgi/alerts

http://ssc.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://ssc.sagepub.com/content/14/4/373.refs.html

http://ssc.sagepub.com/content/14/4/373.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml



373

Creating a Democratic Public SphereThrough Political DiscussionA Case Study of Abortion Conversation on the Internet

STEVEN M. SCHNEIDER

State University of New York

Computer-mediated discussion is an increasingly popular method of engaging in political talk withother citizens. This article presents a case study of a Usenet newsgroup focused on abortion, anddiscusses the creation of a public sphere by the conversants. The notion of the public sphere isdiscussed, and measures allowing an assessment of its democratic character are proposed. A formulafor estimating entropy is developed and applied to data obtained from the case study of an ongoingdiscussion. A high level of inequality in participation among conversants is noted, with very few ofthe discussants responsible for an extraordinarily high proportion of the content. This inequality,though tempered by analysis of the newsgroup on a day-to-day basis, calls into question the democraticcharacter of the public sphere represented by this conversation.

Keywords: democracy, public sphere, computer-mediated communication, politicaldiscussion, political conversation, abortion, Internet, Usenet, newsgroups,entropy

he emergence of computer-mediated political discussion represents an enhancementof opportunities for individual citizens to talk with each other about politics. Thisenhancement has important consequences for the structure of what can be termed the publicsphere. This article explores these consequences through the structural analysis of an ongoingconversation about a political issue taking place in a Usenet newsgroup.

This article presents a part of an analysis from a case study of a year-long politicalconversation focused on abortion and conducted on computer networks. The project drawson three separate topics in the scholarly literature, each of which is briefly introduced in thefollowing section: public sphere, computer-mediated communication, and abortion dis-course. Part 2 of the article describes a technique for assessing the structure of the publicsphere created through computer-mediated communication, and presents results from theanalysis. The focus of the present work is on the structure of the arena; forthcoming analyseswill examine the content of the conversation and the behavior of the conversants. The final

section outlines conclusions, suggests implications of the findings to date, and raises issuesfor further analysis.

AUTHOR’S NOTE: Please address correspondence to SUNY Institute of Technology, Utica, NY 13504; e-mail:s teve@ sunyi t . edu. An earlier version of this article was presented at CSS/IASSIST 1996 Annual Conference.Additional research from this project will be available at http: / /nextwave. sunyit. edu/steve/ publicsphere.

Social Science Computer Review, Vol. 14 No. 4, Winter@ 1996 Sage Pubhcations, Inc.

at DALHOUSIE UNIV on November 9, 2014ssc.sagepub.comDownloaded from


374

THEORETICAL PERSPECTIVES: THE PUBLIC SPHERE, COMPUTER-MEDIATED DISCUSSION, AND ABORTION DISCOURSE

Conversation is the foundation on which all political behavior is built (Ackerman, 1989;Barber, 1984; Dewey, 1959; Huspek & Kendall, 1991). Through political conversation, theabstract world of issues and events is transformed into meaningful reality with personalconsequences. &dquo;Democracy begins in conversation,&dquo; wrote John Dewey (1959). BruceAckerman (1989, p. 6) agreed: &dquo;Dialogue is the first obligation of citizenship.&dquo;

Recent analyses of the state of democracy in the United States lament the lack ofopportunities for ordinary Americans to talk with each other in public settings about publicissues (Barber, 1984; Bellah, Madsen, Sullivan, Swidler, & Tipton, 1991; Oldenburg, 1989).These settings include those places occupying &dquo;associational space&dquo; encompassing the &dquo;coresettings of informal public life&dquo; (Oldenburg, 1989). Oldenburg (1989) and Bellah et al.(1991) have identified the scarcity of associational space as one of the primary ills facingour polity. These opportunities have declined in recent decades as the public sphere hasbecome increasingly focused on satisfying commercial and other private interests(Habermas, 1989). Furthermore, the few opportunities that remain for public conversationhave become less meaningful as citizens adopt a consumerist approach to politics and viewthe public interest as a majoritarian or pluralist construct (Dahl, 1989).

Creating new settings for public interaction and public life may revive our democraticspirit andrenew our public consciousness. I believe the changing structure of communicationin the late 20th century offers this possibility. By shifting the focus from communicationbased on transactions with large, homogenous, undifferentiated, and anonymous audiences(Lasswell, 1948; Wright,1975) to communication based on exchange among small, hetero-geneous, distinct, and identifiable participants (Abramson, Arterton, & Orren, 1988;Neuman, 1991; Rogers, 1986; Stevens, 1981), it may be possible to reclaim the resourcesnecessary to support opportunities for effective political discussion.Among the recent developments in communication technology that offer this shift in

focus is the rapidly emerging medium of computer-mediated discussion, in which individu-als use computers and computer networks to engage in conversations that take place overtime. I believe this type of communication, when focused on politics, can be an importantcontributor to developing the skills necessary for political discussion, and in fact can serveas the platform for a meaningful forum for democratic political discussion.

The question facing those creating and using emerging forms of associational space ishow to ensure that these spaces satisfy desired democratic values. One theoretical domainto which we can turn for answers is concerned with the notion of the public sphere. Scholarlydebate on and discussion about the public spherehas been reinvigorated since thepublicationof Jurgen Habermas’s Structural Transformation of the Public Sphere (Habermas, 1989).This concept provides a fertile domain in which to frame a discussion about associationalspace.

The public sphere is defined by Habermas as &dquo;a realm of our social life in which somethingapproaching public opinion can be formed&dquo; (1974, p. 49). Keane (1984, p. 2) suggested that&dquo;a public sphere is brought into being whenever two or more individuals, who previouslyacted singularly, assemble to interrogate both their own interactions and the wider relationsof social and political power within which they are always and already embedded.&dquo;

Functionally, a public sphere can be thought of as those arenas in which members of a society&dquo;consider what they are doing, settle how they will live together, and determine, within theestimated limits of the means available to them, how they might collectively act within theforeseeable future&dquo; (Keane, 1984, p. 2).



375

Democratic theorists who hold to participatory maximsl strive for a public sphere as closeto Habermas’s (1989) &dquo;idealized&dquo; formulation as possible. In this view, all members of thepolity, as independent and private citizens, have an opportunity to participate in rational-critical debate of topics relevant to their public lives. These debates and deliberations resultin consensus2 among the members of a public on those actions for which it is necessary forthe state to engage, and produce a sense of democracy among the participants in the publicdiscussion. Public authorities respond by enacting this consensus into law. In contrast,democratic theorists who view the primary function of participation as protection3 argue fora public sphere that seeks only to regulate conflict in a fair and equitable manner. It isunnecessary and inefficient for the public sphere to enable citizens to discover what theyhave in common. The purpose of the public sphere is to provide legitimacy to both state andprivate market actions seeking to allocate resources within society.

One of the primary issues to resolve when considering the public sphere is how tounderstand the factors that influence the nature and style of discourse within them. Someargue that structural characteristics ofpublic spheres are responsible for the nature of politicalconversation supported within them. As part of this argument, characteristics of a publicsphere that will support democratic conversation are identified. One view is that the structuralcharacteristics of what Stanley (1988) called the &dquo;forum,&dquo; and what I prefer to call the&dquo;conversational arena,&dquo; determine the nature of discourse within it. Closely connected to thisliterature is a utopian strand suggesting that an arena properly constructed would yielddemocratic discourse.

American policymakers, especially during periods of technological emergence similar tothe present time, have for the past 200 years expressed the view that democratic communi-cation is made possible by structural characteristics. Discussions of postal policy in the late18th and early 19th centuries justify the creation of a national infrastructure and below-costrates to facilitate communication among citizens and between citizens and their government(U.S. Congress, 1990). The initial foray of the government into telegraphy in the late 1830swas in part justified on the necessity of citizen interaction to democracy (Schneider, 1988).The U.S. government explicitly allowed a telephone monopoly to emerge to provideuniversal service to citizens, at least in part because of a belief in the democratic potentialof the telephone network (U.S. Congress, 193 9). The presence of local voices in the broadcastspectrum was valued more highly than the number of voices; trade-offs were made inspectrum allocation that provided fewer local channels rather than a greater number ofnational channels (McChesney, 1993; Noll, Peck, & McGowan, 1973) In policy decisionsconcerning cable television, the commitment to democratic communication was exhibitedin requirements for local access and locally originated programming (Pool, 1983). Keyaspects of the debate on &dquo;media access&dquo; argue that economic and political structure areresponsible for the democratic character of the public sphere (Barron, 1967; Caristi, 1988).More recently, the utopian ideal of democratic communication has been present in thedebates considering the national information infrastructure and computer networks(Fisher, Margolis, & Resnick, 1994; Surman, 1994a, 1994b). Twenty-five years of discussionconcerning the structure of computer networks, starting with Sackman (1970, 1972), havebeen infused with the ideal that, if properly constructed, a democratic public sphere willemerge; this is the common thread linking Kapor’s ( 1994) &dquo;Jeffersonian&dquo; information policyand Dyson, Gilder, Keyworth, and Toffler’s (1994) &dquo;Magna Carta for the Knowledge Age.&dquo;

It is my contention that political discussion conducted on computer networks cancontribute to a democratic enhancement of the public sphere. Others suggest that the mostimportant impact on politics of computer networks is the ability to link the elite policymakerswith what I would call the &dquo;quasi-elite&dquo; members of the public. More optimistic, perhaps,



376

are those who focus on the increased ability of grassroots groups to network amongthemselves and create strategic links among leaders and activists enabling social movementorganizations to function more effectively and efficiently (Agre, 1994; Downing, 1989;Holland, 1990). I find both of these possibilities less interesting than those raised bycomputer-mediated discussion. Even if these outcomes occurred, politics as we know itwould hardly be transformed-perhaps the balance of power would be slightly shifted for ashort time, but over the long haul, little would change.

Computer-mediated discussion may be different. With computer-mediated discussion,there is a true transfer of power from one group to another. Individuals who have no voice

or, more precisely, no audience for their voice may suddenly find themselves in a publicforum, offering their opinions, andhaving other people listen and respond to them. This formof social interaction bridges the gap between the two dominant modes of communicationwith which we are familiar. On the one hand, we have the dyadic, interpersonal, face-to-faceor one-to-one communication-two people talking to each other. One the other hand, wehave mass or one-to-many communication-one person speaking, many people listening.In between is &dquo;group participation&dquo; (Neuman, 1991) or &dquo;mini-communication&dquo; (Gumpert,1983) or &dquo;many to many&dquo; communication (Stevens, 1981), in which each participant has anopportunity to be both a speaker and a listener. Computer network technology expands thepossibilities for group participation media.

Computer-mediated discussion is a specific form of computer-mediated communication(CMC). CMC can be defined as &dquo;the use of a computer to create, address, route, distribute,or receive messages sent from one individual to another, from a group to an individual, froman individual to a group, or from one group to another group&dquo; (Murphy, 1994). CMC can beeither synchronous or asynchronous (Walther, 1992). These specifications of CMC blend avariety of communication modes by emphasizing the computer aspect of the communication;if we were to follow a similar strategy in defining, for example, print communication, as theuse of paper and ink to exchange of messages between individuals or groups, we would beunable to distinguish between letter writing and newspaper publishing.

It is essential that we distinguish computer-mediated discussion from other types of CMC.Following Rice (1984), I suggest the following definition:

Computer-mediated discussion is a computer-facilitated mechanism for recording and using atextual transcript of a series of messages written by more than one individual over varyinglengths of time, with participation by individuals who may be geographically dispersed, andwho may interact with the transcript either simultaneously or at times of their own choosing.

This definition excludes from consideration as computer-mediated discussion the following:a single author writing a series of messages read by many individuals, and two authors whoeach write a single message (the discussion aspect begins with the third message).**

Computer-mediated discussions are both technologically simple (Krol, 1992) and linguis-tically complex (Bolter, 1989, 1991; Ess, 1992; Ferrara, 1991; Shank, 1993) 5 The mostfamiliar analogy to the noncomputer world might be a bulletin board in a common area. Anindividual happening by the bulletin board might decide one day topost a message, addressedto no one in particular, about a specific subject. Another individual happening by might readthe posting, and decide to post a reply to the original message, or to post a message of hisown. A third individual happening by might read all of the postings, and reply as desired.The first individual might happen by a few days later, and respond to the responders, andadd a new posting on a different subject. And so it goes: individuals can enter or leave thediscussion at any time, responding to both recent and not-so-recent messages, starting newthreads in the discussion. In this example, the size of the discussion would be limited by the



377

physical dimensions of the bulletin board and the participants would be limited to those whophysically happened by the board. The computerized version is very similar, with thephysical location of the board replaced by one or many physical locations in a network ofcomputers who are connected to each other, and the physical location of the participantsreplaced by their presence on a computer that can access the board. The size of the discussionis limited by the much less constraining limitations of the network, and the participants arelimited to the much less constraining limitations of network users.

Different types of computer-mediated discussion emerge when contrasted with othermodels of interpersonal conversation. Conversational models include the monologue, thedialogue, and the discussion. Amonologue involves one message sender and multiplepassivereceivers; a dialogue involves two participants taking turns and exchanging roles as senderand receiver. Adiscussion involves a singleperson who starts as the sender and retains controlof the conversation with multiple receivers, some of whom rotate (with the initiator) the roleof the sender. These models of conversation do not describe a fourth alternative, which can

potentially be realized in the computer-mediated mode:

We have the starter, or the initial sender, who starts the &dquo;thread.&dquo; ... Once a thread has beenstarted though, it is no longer under sender control. This is because the mechanics of Netresponse do not require turn taking.... [I] t is as if everyone who is interested in talking can alljump in at once, but still their individual voices can be clearly heard.... [I]t is as if someonehad started writing a piece, but before he/she gets too far, people are there magically in print toadd to, correct, challenge, or extend the piece. Therefore, what we have is a written quasi-discussion that has the potential to use the strengths of each form.... I think it is best to callthis form of communication &dquo;multiloguing.&dquo;(Shank, 1993)

Thus computer-mediated discussion might take on characteristics of a conversationaldiscussion, in which a single person retains control of the conversation. These kinds ofdiscussions can currently be found on a variety of computer networks, including somemoderated newsgroups on the Usenet, as well as a number of commercial services, such asAmerica OnLine, CompuServe, Prodigy, and Delphi. The form of discussion referred toabove as a multilogue is found within the unmoderated newsgroups on Usenet, as well asmany independent bulletin boards.

Although the use of computers for discussion is a relatively recent phenomenon, scholarshave begun systematic investigations of its characteristics. Some have examined the rela-tionship between access to information and the distribution of power in society, suggestingthat computer-mediated discussion could ease differential access, distribute power morebroadly, and thus lead to greater democratization (Ess, 1992). Others, mostly concerned witheducational applications, have noted the potential to foster creativity and cooperation amongparticipants (Kiesler, 1984; McCormick, 1992). The absence of social cues to status andgender have suggested the potential to create more democratic communication (Graddol &Swann, 1989). Users of Usenet newsgroups have been found to be more active in theircontacting activities than average citizens (Fisher et al., 1994).

One of the most extensive collections of ongoing computer-mediated discussions isUsenet. Usenet links tens of thousands of computers and potentially millions of users througha series of discussions called newsgroups. The range of topics addressed by Usenet news-groups can be assessed by a brief look at the conventions for naming groups. Newsgroupsalways have a name that includes at least two words separated by a period. The first wordindicates the classification, and the second and subsequent word further identifies the group.There are seven types of Usenet groups: comp groups discuss computer science, software,hardware, systems, and the computing profession; sci groups include discussions of a



378

technical nature related to a specific science or discipline; news groups are concerned withUsenet news administrative issues; soc groups are addressed to socializing and social issues;rec groups are dedicated to recreational topics; talk groups are intended, as their nameimplies, for people to talk about various topics; and misc groups include those who do notfit into other categories (Spafford, 1994). In addition, an alternative naming hierarchy existsto include the &dquo;unofficial&dquo; newsgroups, including the alt (alternative) groups. Among thiscollection are about 50 newsgroups whose title suggests a dedication to discussion of

political issues, covering a wide range of topics. One of these newsgroups-talk.abortion-includes &dquo;all sorts of discussions and arguments on abortion,&dquo; and provides the case studyfor this project.

Abortion is, at this time, a relatively permanent feature of our ongoing politicalconversation. Although some issues receiverelatively fleeting attention in political conversation,others, like abortion, develop a sustained and seemingly permanent place on the politicalagenda. For 30 or so years, abortion has been a permanent feature of American politicaldiscourse. Though it has had peaks and valleys in terms of the relative amount of attentiondevoted to it, the issue remains at the forefront ofpublic discourse. Millions of conversations,thousands of books and magazine articles, hundreds of films and television specials, and ahalf dozen Supreme Court cases have engraved the issue deeply into the political agenda,and there is little prospect of ending discussion and debate anytime in the near future.

Abortion has not always been a public or political issue in the United States. Prior to 1865,many states, and the federal government, were silent on the legality of abortion; it was largelya private issue, with little public activity and virtually no public discourse. During the secondhalf of the 19th century, every state outlawed abortion, allowing exceptions only when thelife of the mother was endangered. The factors leading to the criminalization of abortioninclude a rise in professionalism of the medical field, and advances in the biological sciencesthat shed new light on fetal development. Discourse during the criminalization period isdominated by professional voices, including doctors and clergy; noteworthy is the almosttotal absence of women’s voices from this discourse. (Craig & O’Brien, 1993; Davis, 1985;Gordon, 1989; Luker, 1975, 1984; Olasky, 1988; Tatalovich & Daynes, 1981; Tribe,1990).

Once the pattern of criminalization was established, abortion again receded from publicview, and public discourse was largely excluded from mainstream media (Olasky, 1988).Not until the 1960s, as part of the burgeoning women’s movement, did abortion again occupya significant part of the public sphere. The &dquo;structure of political opportunities&dquo; (Eisinger,1973; McAdam, 1988; McCarthy & Zald, 1977) supported the emergence of a socialmovement in support of abortion law reform. The reformist movement was supported by (a)a &dquo;critical discourse moment&dquo; (Gamson & Modigliani, 1989), the Finkstein case, and theassociated thalidomide scare; (b) the success of other social movements, including the civilrights and antiwar initiatives (Evans & Boyte, 1986; Huntington, 1981; Staggenborg, 1991);and (c) the reemergence of gender as an organizing force (Evans, 1978; Freeman, 1975). Thehigh point of the reformist movement was the Roe v Wade decision by the Supreme Court,which invalidated all state laws prohibiting abortion during the early stages of pregnancy.During this period (1960s to early 1970s), two significant changes occurred with regard topublic discourse concerning abortion. First, women’s voices became, for the first time, asignificant part of the discourse. Second, the discourse began to focus on rights, contrastingthe rights of women to the rights of embryos (Blake, 1971; Condit, 1990; Ginsburg, 1989;Luker, 1975; Olasky, 1988; Tribe, 1990).

Following Roe v. Wade, a highly polarized era emerged. Abortion became, for the firsttime, a major issue in which large portions of the public participated through their interac-tions with public discourse, across the entire domain of the public sphere (Baehr, 1992;



379

Condit, 1990; Craig & O’Brien, 1993; Ginsburg, 1989; Gordon, 1989). For the first time,abortion discourse was accessed by members of the general public, through widespreadcoverage in the news media and by the entertainment media. However, discourse aboutabortion was much more polarized than in previous eras, and consensual voices (Barber,1984; Bellah et al., 1991; Oldenburg, 1989; Stanley, 1983, 1988) lost their access to thepublic sphere. More recently, the polarization of the post-Roe era has intensified. Someprolife advocates have essentially given up trying to change general attitudes throughreasoned discourse, and are instead using or advocating (or at least not criticizing) the useof terror as a weapon against those providing abortion services (Donovan, 1985; Forrest &

Henshaw, 1987; Grant, 1991; Terry, 1988; Wills, 1989). This has resulted in a substantialshift in abortion discourse, away from the prolife, prochoice dimension and toward thedimension of tactics and strategies of the prolife movement. Discourse about abortion is noless widespread than during the years following Roe v. Wade, although its focus and emphasisare rather different.

The abortion issue is particularly well-suited to this analysis, with its focus on theconstruction of a democratic public sphere. The abortion issue is an important ongoing publicissue, which continues to have unresolved aspects featured in public discourse (Ferree &

Gamson, 1993). Furthermore, there is well-established literature on all aspects of the abortionissue, including the political and social history of abortion, the philosophical and legalchallenges posed by the issue, and the nature of public discourse concerning abortion. Thestandard works on the discourse of abortion include Condit (1990), Olasky (1988), andGinsburg (1989). In addition, the nature of the abortion issue requires a well-constructedpublic sphere for a democratic resolution to emerge. It is not neatly characterized by familiarideological distinctions; there are direct links between personal beliefs, personal behavior,and public actions; and the policy-making process is widely distributed across a range ofpublic and private institutions.

METHODS: ANALYZING COMPUTER-MEDIATED DISCUSSIONAS A CONVERSATIONAL ARENA

Developing a technique for estimating the degree to which a public sphere satisfies therequirements of participatory democracy requires that the conditions or criteria of whatcritical theorists call a &dquo;counterfactual ideal&dquo; (Dryzek, 1990) be specified. Habermas’s idealspeech situation, and its operationalization in the idealized public sphere, serves here as themodel against which real-world experience can be criticized. The public sphere created bya computer-mediated discussion includes three components, and is fully examined in threedimensions. The components of the public sphere include the conversational arena, theconversants, and the conversation. The conversational arena refers to the forum or space inwhich the conversation takes place; the conversants refer to those individuals who engagein discourse in the public sphere; and the conversation is the actual discourse that takes placeamong the conversants in the conversational arena. The three dimensions I identify thatembody the spirit of the idealized public sphere presented by Habermas are equality,diversity, and reciprocity. Equality in the conversational arena refers to rates of participationamong participants. Diversity concerns the patterns of discussion and participation. Reci-procity concerns opportunities to gain knowledge of the perspectives of others and thedegree to which these opportunities are realized. This analysis is focused on equality in theconversational arena; forthcoming work examines the additional dimensions in other com-ponents of the public sphere.



380

Before launching into a discussion of results, some readers might find a brief explanationof how Usenet newsgroups actually work helpful.6 This is best done by analogy to familiarmedia. One of the many features we use to differentiate some media messages from othersis addressability. Addressability, as a feature of a specific message, concerns the specificityof the intended recipient. Some traditional media require specific addresses to be delivered:telephone calls are routed to specific telephone lines; personal letters are routed to specificmailboxes. Other traditional media do not require specific addresses: broadcast televisionand radio programs are accessible by tuning into a specific frequency. Finally, some mediaare produced without a specific addressee in mind (books, newspapers, junk mail), butrequire specific addresses for delivery (i.e., many copies of identical messages are deliveredto identifiable recipients).

Computer-mediated communication has the same types of messages. Fully addressedmessages are called electronic mail. Individuals receive mail from other individuals that is

specifically addressed to them, or by virtue of being on a mailing list, to which they havesubscribed. Subscribing to a mailing list is analogous to subscribing to a magazine. Theanalogous computer-mediated communication to broadcast television is the newsgroup.Unlike electronic mail, which requires a user to open her mailbox to retrieve, newsgroupsrequire users to &dquo;turn on the set&dquo; and &dquo;tune into a specific channel.&dquo; The set is typically acomputer system (as opposed to a personal computer) shared by many users. Thoughindividuals might use a personal computer to connect to the system, they are unlikely todirectly access newsgroups on these devices. Individuals often log on to a computer systemand use a newsreader to access the newsgroups. Newsreader software identifies the news-

groups available to the reader, and the user selects a newsgroup to read (equivalent toselecting a channel on a television) and is informed about the number of articles availableon the user’s host system.An article or posting7 is a block of text (generally stored as a file on the user’s host system)

written by another individual, and intended to be included as part of an existing newsgroup.The two major portions of a posting are the header and the mess age body. Theheader includesinformation in a group of standard fields, such as the author of the message, the subject ofthe message, a unique identification symbol, the date the message was written, and so on.From the perspective of the user, the two most important fields are probably the subject andthe author fields. Most newsreaders allow for threading of messages using the data in thesubject field. Threading links postings sharing the same subject and allows readers to viewthem sequentially. Many newsreaders, in fact, will initially display a list of threads withcurrent postings, and the number of associated messages. Participants may choose to readany of the postings currently available on the newsgroup (most news sites discard messagesafter a fixed number of days, usually 3 to 10). They may choose to add a posting of theirown to the group, either selecting a new subject for their posting, or adding to the ongoingconversation under an existing subject heading. Many participants who choose to add to theongoing conversation use a feature of most newsreader software that allows them to includethe text of previous messages into their new message; text from previous messages isidentified by a special symbol (often a > or a I) preceding each line, and the author of theprevious text is usually indicated.

During the 12-month study period (April 1,1994, to March 31,1995), all of the discussionthat took place within talk.abortion was archived, with each posting saved as a record in adatabase. A total of 46,592 articles were posted to the newsgroup by 3,276 different authors,and the articles were distributed across 7,831 threads or subjects.



381

RESULTS : EQUALITY IN THE CONVERSATIONAL ARENA

The dimension of equality is reflected by rates of participation among participants.Although this aspect of equality is only implicitly included in Habermas’s (1989) conceptionof the public sphere, it is familiar in democratic theory. One indicator of equality is equalprotection of interests of members. Equality of interests could be measured by the distribu-tion of participation among participants. In plebiscites, equality is achieved when eachparticipates equally, the &dquo;one person, one vote&dquo; maxim, universally recognized as a desirableend (Wolin, 1960). Applying plebiscitary values to a public sphere would suggest thatequality is obtained when each participates or contributes to the discussion in equal measure.Participation in a plebiscitary sense could be measured by both frequency of expression andtotal quantity of expression.A measure of the degree of equality obtained in the public sphere is the concentration

of participation. Concentration is a term borrowed from economics that summarizesindustry structure by measuring the number of firms competing in a market and theirrespective market shares (Scherer, 1980). In a conversational arena, concentration is ameasure of the number of participants and each participant’s respective share of thetotal participation. The most common graphic technique for summarizing concentrationis a Lorenz curve. This technique ranks participants from the most participatory to leastparticipatory along a horizontal axis, and the percentage of total participation on thevertical axis. Figure 1 indicates the Lorenz curve summarizing degree of concentrationof posts by contributors to the talk.abortion newsgroup. As is evident from this figure,posting behavior is extraordinarily concentrated: very few authors account for a very highproportion of posts. Examining the concentration curve for the 50 most frequent con-tributors (Figure 2) indicates that the top contributor is responsible for 11 % of the posts,and that the top 10 contributors wrote nearly 40% of the posts. The top authors posted 60%of the articles.

There are a variety of single parameters to summarize Lorenz curves and allow compari-son across concentration levels (Scherer, 1980). When accurate data summarizing participa-tion are available for all participants, the entropy coefficient is an appropriate measure(Horowitz, 1971).

Concentration of participation can be summarized using the entropy coefficient, definedas:

where Si is the percentage of total participation of the ith participant. The entropy coefficientsummarizing concentration of participation by posts for the entire sample is .67, indicatinga high level of concentration. In the formulation of entropy above, the value of E whenparticipation among all participants is equal will be log2N. Thus, to compare the degree ofentropy among populations in which N varies, E can be normalized as E



382

øoptt!x14

a.s!5::sb

c.80oQ.w’3copbic!iJr.



383

8ft

ix....19«

assa:I.abc

9s

A.I’aAll,

iA.*3C0p

Iic.N!iit



384

so that E is the proportion of the maximum entropy possible given the number of participants.Figure 3 and Figure 4 show the decline in entropy as the number of authors consideredincreases from 10 to 3,276. Much of the drag on entropy can be accounted for by the highproportion of low-contributing authors. Ignoring the articles and authors posted bylow-contributing author groups results in normalized entropy levels among the mostfrequent contributors of nearly .9, and levels .75 or above even when accounting for the morethan 90% of the articles posted by the most prolific 20% of the authors. This analysishighlights the relative equality at the higher levels of participation and indicates thesubstantial difference between the higher and lower level participants.

These findings suggest a categorization of authors by frequency of contribution for furtheranalysis. Eight categories of authors are suggested, developed from the number of articleswritten by an author or from the rank ordering of authors by number of articles written. Twocategories are based on the number of articles written: One-time authors are those writing asingle article (1,740 or 53% of all authors); two-time authors include those writing twoarticles (475, 15 %). The remaining authors are categorized by ranking all authors from mostto least prolific and establishing cutoffs at fixed points. Fanatic authors include the mostprolific one half of 1% of authors (15 authors, each of whom wrote at least 483 articles).Subsequent groups of authors were formed from those in the top 2.5%, 5%, 10%, 20%, and30% of authors, ranked by number of articles written. The complete categorization ispresented in Table 1.

The 11 fanatic authors are responsible for an overwhelming 43% of the posts to thenewsgroup during the study period (see Table 1). Combined with the frequent authors, morethan two thirds of the posts are accounted for. The one-, two-, and three- to five-time authorscontributed less than 10% of the total postings. Similar findings are shown when examiningwords posted by author type. However, the average length of posts, measured in originalwords, does not vary significantly across author types (p > .05), indicating some minimallevel of equality.

More frequent authors tended, on average, to be active in the newsgroup on a greaternumber of different days (p <.01), and for much longer periods of time (p <.01). The spreadof days, calculated as the number of days elapsing between the first recorded post and thelast recorded post, declines precipitously as the number of frequency of posting declines.The fanatic authors posted messages on an average of 170 different days, across an averagespread of 335 days. In comparison, the occasional authors (those who posted between 33and 77 messages) posted articles an average of 23 different days. This finding is somewhattempered when the intensity of participation is considered. Intensity is the percent of daysin which an author posted an article, using the spread of days as the base. In other words, ifthe spread of days was 5, and an author posted articles on 5 different days, intensity was100%. Although Table 1 indicates that the most frequent authors had the highest levels ofintensity, the differences among the middle-range contribution groups are not as substantialas in other measures. This finding suggests a small degree of equality in participation.

Rates of participation in the talk.abortion newsgroup during the entire study period areanything but equal. Concentration of participation among the most frequent contributors wasextreme. More than half of the contributors wrote only one message, whereas the mostfrequent 15 authors wrote more than 40% of the messages. The only limited support forequality found when examining data over the course of the entire study period is theconsistency across authors of the average length of their contributions.



385

85s0T

E!I().51!oa8tgcW0)

0AM!::sQ¡¡:



386

eaNCf).s8E

iIu

.51!o:54ItgB0

:5

i.!0)u.



387

CL

:s20)

Idz

ø0.¡J

f- k

uiJm.l~4- 1-1 0

.¡J

.5co:;:ICISa..ü1:CISa.



388

Another approach to the data is to examine activity in the newsgroup across time. Mostnews servers &dquo;expire&dquo; (i.e., delete) messages from newsgroups in 3 to 10 days, dependingon the site. In this part of the analysis, data are presented assuming a 7-day expiration.Assuming, then, that articles expire after 7 days, the entropy level of contributing authorscan be calculated on a daily basis. Figure 5 shows the trend in entropy, normalized to accountfor the varying number of contributing authors, during the course of the study period.Recalling from Figure 4 that the entropy level for the group as a whole was shown as .67,this analysis indicates that, on any given day, a much higher level of entropy was observed,as entropy varies from .7 to .85 during the study period. Although entropy varies during thestudy period-there were several substantial periods outside the boundaries of one standarddeviation of mean entropy-it is clear that participation is considerably more equal on aday-to-day basis than over the entire year.

DISCUSSION

Based on the analysis presented in this article, the public sphere created by participantsin the talk.abortion newsgroup only partially satisfies the requirement of equality. Over thecourse of the period observed, substantial inequalities in the rate of participation wereobserved. Though this inequality is somewhat tempered by examining the newsgroup on aday-to-day basis, this finding suggests that the conversational arena associated with thisparticular newsgroup is far removed from the idealized public sphere envisioned byHabermas.

Usenet newsgroups share all of the characteristics of the idealized public sphere, and arefree from the weaknesses of Habermas’s degenerate public sphere. They are responsive toneither commercial nor state pressures; they exist in a space that is essentially owned andcontrolled by the participants; there are relatively low barriers to entry and no obviousdistortion effects on the communication. It is curious, therefore, that the conversational arenadoes not appear to encompass the kind of democratic communication that might be expected.This finding is probably not surprising to those who have spent much time on Usenetnewsgroups; complaints about the low signal-to-noise ratio are common, especially withregard to the politically oriented groups. Nevertheless, current efforts by policymakers andmarketers to shape the next generation of communication services, at least partially basedon the &dquo;Field of Dreams&dquo; mentality that merely providing a technology will ensure its desireduse, might be partially guided by the findings here.

Research on newsgroups and other forms of computer-mediated discussion has not yetcombined analyses of structure, content, and participants. Future efforts associated with thisproject examine both the actual conversation created by the newsgroup, using traditionalmethods of content analysis, and the conversants themselves, using a series of electronicmail interviews with a small sample of participants. In addition, the dimensions of diversityand reciprocity will be measured. When complete, this project will provide observers witha robust assessment of the democratic nature of the public sphere created by a Usenetnewsgroup over the course of a full year.

NOTES

1. Classical theorists include J. S. Mill and Rousseau. Contemporary theorists include Barber (1984), Pateman(1970), and Mansbridge (1983).



389

’Õ=’4I!0Ca.

gew.5&dquo;

!Ii)!0)i:



390

2. But see McCarthy (1993) and Manin (1993) for discussion on the need for consensus.3. Represented classically by J. Mill and in contemporary theory by Schumpeter (1942) and Dahl (1961).4. Rice (1984, p. 131) suggested the following definition: "computer-facilitated mechanism for recording and

using a textual transcript of a group discussion over varying lengths of time, by group members who may begeographically dispersed and who may interact with the transcript either simultaneously or at times of their ownchoosing." This definition highlights the dimensions of time, geography, and asynchronicity, and does not imposethe restraints suggested by a traditional definition of discussion (Shank, 1993). It is instructive, however, toconcentrate on the notion of "group" implied in this definition. In the first place, the discussion is said to be groupdiscussion, and the participants in the discussion are said to be group members. Computer-mediated discussion,especially in the variation distributed by Usenet and examined in this article, does not support the logic of "group"in the traditional sense of the word, and certainly rejects the notion of group "membership." A person becomes aparticipant in the discussion by, in Rice’s words, interacting with the transcript. However, there is not necessarilya record of having done so, and neither entry to nor exit from the discussion is necessarily a public event of whichothers would be aware. Individuals can interact with the transcript by reading the words of others without leaving"evidence" of their actions. Furthermore, the notion of membership suggests some level of commitment to thegroup as a whole. I would suggest that the "group," constituted as a structure that exists in the minds of its members,does not necessarily exist in a computer-mediated discussion. In addition, to the extent that such identification doesexist, it is not necessary that there be any commitment on the part of the participants.

5. See also Herring (1993); Kiesler, Seigel, and McGuire (1984); McCormick and McCormick (1992); Rafaeli(1986); and Smith and Balka (1991). Early descriptions of computer-mediated discussion (sometimes calledcomputer conferencing) appear in Hiltz (1984); Hiltz and Turoff (1978); Johansen, Vallee, and Spangler (1979);Kerr and Hiltz (1982); and Rice (1980, 1984).

6. For a more detailed description, see Krol (1992).7. I prefer the term "posting" to "article." Article, to me, conjures up images of written and published text that

has gone through an editorial process. Aposting, on the other hand, seems to me more similar to a letter, which thesender has complete control over, in terms of composition and entry into the delivery system.

REFERENCES

Abramson, J., Arterton, C., & Orren, G. (1988). The electronic commonwealth. Cambridge: Harvard UniversityPress.

Ackerman, B. (1989). Why dialogue? Joumal of Philosophy, 86(1), 5-22.Agre, P. (1994). The Network Observer, 1(1). Hostname: weber. ucsd. edu.Baehr, N. (1992). Abortion without apology: A radical history for the 1990s. Boston: South End Press.Barber, B. (1984). Strong democracy: Participatory politics fora new age. Berkeley: University of California Press.Barron, J. (1967). Access to the press—A new first amendment right. Harvard Law Review, 80, 1641-1657.Bellah, R N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, S. M. (1991). The good society. New York:

Knopf.Blake, J. (1971). Abortion and public opinion: The 1960-1970 decade. Science, 171, 540-549.Caristi, J. (1988). The concept of a nght of access to the media: A workable alternative. Suffolk University Law

Review, 86, 945-958.Condit, M. (1990). Decoding abortion rhetoric. Champagne, IL: University of Illinois Press.Craig, B. H., & O’Brien, D. M. (1993). Abortion and American politics. Chatham, NJ: Chatham House.Dahl, R (1961). Who governs? New Haven: Yale University Press.Dahl, R A. (1989). Democracy and its critics. New Haven: Yale University Press.Davis, N. (1985). From crime to choice: The transformation of abortion to America. Westport, CT: Greenwood.Dewey, J. (1959). Dialogue on John Dewey. New York: Horizon.Donovan, P. (1985). The holy war. Family Planning Perspectives, 17(1), 5-9.Downing, J. D. (1989). Computers for political change: PeaceNet and public data access. Journal of Communica-tion, 42, 153-162.

Dryzek, J. S. (1990). Discursive democracy: Politics, policy, and political science. New York: CambridgeUniversity Press.

Dyson, E., Gilder, G., Keyworth, G., & Toffler, A. (1994, December). Cyberspace and the American dream: Amagna carta for the knowledge age. Host address: http : //www.pff. org.



391

Eisinger, P. K. (1973). The conditions of protest behavior in American cities. American Political Science Review,67, 11-28.

Evans, S. (1978). Personal politics: The roots of women’s liberation in the civil rights movement and the new left.New York: Knopf.

Evans, S. M., & Boyte, H. C. (1986). Free spaces: The sources of democratic change in America. New York:Harper & Row.

Ferree, M., & Gamson, W. A. (1993). Social movements and public discourse: The construction of abortiondiscourse in Germany and the United States: Proposal to the National Science Foundation. Unpublishedmanuscript.

Fisher, B., Margolis, M., & Resnick, D. (1994). A new way of talking politics: Democracy on the Internet. Paperpresented at the Annual Meeting of the Amencan Political Science Association. Available FTP: Hostname:weber. uc sd . edu Directory: pub/rre/democracy.

Forrest, J. D., & Henshaw, S. K. (1987). The harassment of U.S. abortion providers. Family Planning Perspectives,19(1), 9-13.

Freeman, J. (1975). The politics of women’s liberation. New York: McKay.Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist

approach. American Journal of Soczology, 95, 1-37.Ginsburg, F. (1989). Contested lives: The abortion debate in an American community. Berkeley: University of

California Press.

Gordon, L. (1989). The struggle for reproductive freedom: Three stages of feminism. In Z. R. Eisenstein (Ed.),Capitalist patriarchy and the case for socialist feminism (pp. 137-161). New York: Monthly Review Press.

Grant, G. (1991). Third time around: A history of the pro-life movement from the first century to the present. NewYork: Wolgemuth & Hyatt.

Habermas, J. (1974). The public sphere: An encyclopedia article. New German Critique, 3, 44-55.Habermas, J. (1989). The structural transformation of the public sphere. Cambridge, MA: MIT Press.Herring, S. C. (1993). Gender and democracy in computer-mediated communication. Electronic Joumal of

Communication, 2.Hiltz, S. R. (1984). Online communities: A case study of the office of the future. Norwood, NJ: Ablex.Hiltz, S. R., & Turoff, M. (1978). The network nation: Human communication via computer. Reading, MA:

Addison-Wesley.Holland, G. M. (1990) Computer networks for social change: The social re-shaping of telematics. Unpublished

master’s thesis, University of Wollongong, Department of Science and Technology Studies.Horowitz, I. (1971). Numbers-equivalents in U.S. manufactunng industries. Southern Economic Journal, 37,

396-408.

Huntington, S. P. (1981). American politics and the promise of disharmony. Cambridge, MA: Harvard Belknap.Huspek, M., & Kendall, K. E. (1991). On withholding political voice: An analysis of the political vocabulary of a

"nonpolitical" speech community. Quarterly Journal of Speech, 77, 1-19.Johansen, R., Vallee, J., & Spangler, K. (1979). Electronic meetings: Technical alternatives and social choices.

Reading, MA: Addison-Wesley.Kapor, M. (1994). Where is the digital highway really heading? The case for a Jeffersonian information policy.

Wired Magazine, [On-line], 1(3).Keane, J. (1984). Public life and late capitalism: Toward a socialist theory of democracy. New York: Cambridge

University Press.Kerr, E., & Hiltz, S. R. (1982). Computer-mediated communication systems: Status and evaluataon. New York:

Academic Press.

Kiesler, S., Seigel, J., & McGuire, T. W. (1984). Social psychological aspects of computer-mediated communica-tion. American Psychologist, 39, 1123-1134.

Krol, E. (1992). The whole Internet user’s guide and catalog. Sebastopol, CA: O’Reilly & Associates.

Lasswell, H. (1948). The structure and function of communication in Society. In L. Bryson (Ed.), The communi-cation of ideas (pp. 37-52). New York: Harper.

Luker, K. (1975). Taking chances: Abortion and the decision not to contracept. Berkeley: University of CaliforniaPress.

Luker, K. (1984). Abortion and the politics of motherhood. Berkeley: University of California Press.McAdam, D. (1988). Micromobilization contexts and recruitment to activism. International Social Movements

Research, 1.



392

McCarthy, J. D., & Zald, M. N. (1977). Resource mobilization and social movements: Apartial theory. AmericanJournal of Sociology, 82, 1212-1241.

McChesney, R. W. (1993). Telecommunication, mass media, and democracy: The battle for the control of U.S.broadcasting, 1928-1935. New York: Oxford University Press.

McCormick, N. B., & McCormick, J. W. (1992). Computer friends and foes: Content of undergraduate’s electronicmail. Computers in Human Behavior, 8, 397-405.

Murphy, D. J. (1994). Computer-mediated communication in the aerospace industry. Unpublished doctoraldissertation, Rensselear Polytechnic Institute.

Neuman, W. R (1991). The future of the mass audience. New York: Cambridge University Press.Noll, R. G., Peck, M. J., & McGowan, J. J. (1973). Economic aspects of television regulation. Washington, DC:

Brookings Institution.Olasky, M. (1988). The press and abortion, 1838-1988. Malvern, PA: Lawrence Erlbaum.

Oldenburg, R. (1989). The great good place: Cafes, coffee shops, community centers, beauty parlors, general stores,bars, hangouts, and how they get you through the day. New York: Paragon.

Pool, I. (1983). Technologies offreedom: Onfree speech in an electronic age. Cambridge, MA: Harvard Belknap.Rafaeli, S. (1986). The electronic bulletin board: A computer-driven mass medium. Computers and the Social

Sciences, 2, 123-136.Rice, R. E. (1980). Computer conferencing. In B. Dervin & M Voigt (Eds.), Progress in communication sciences

(Vol. 2, pp. 215-240). Norwood, NJ: Ablex.Rice, R. E. (1984). Mediated group communication. In R.E. Rice Associates (Eds.), The new media: Communica-

tion, research & technology (pp. 129-154). Beverly Hills, CA: Sage.Rogers, E. M. (1986). Communication technology: The new media in society. New York: Free Press.Sackman, H., & Boehm, B. (Ed.). (1972). Planning community information utilities. Montvale, NJ: AFIPS Press.Sackman, H., & Nie, N. (Ed.). (1970). The information utility and social choice. Montvale, NJ: AFIPS Press.Scherer, F. M. (1980). Industrial market structure and economic perfornance, 2nd Edition. Chicago: Rand McNally.Schneider, S. M. (1988) The telegraph as a technology offreedom: Congressional politics and policy, 1837-1845.

Unpublished master’s thesis, University of Pennsylvania, Annenberg School for Communication.Schumpeter, J. (1942). Capitalism, socialism and democracy. New York: Harper.Shank, G. (1993). Abductive multiloguing: The semiotic dynamics of navigating the net. Arachnet Electronic

Journal on Virtual Culture [On-line], 1.Smith, J., & Balka, E. (1991). Chatting on a feminist computer network. In C. Kramerae (Eds.), Technology and

women’s voices (pp. 82-97). New York: Routledge & Kegan Paul.Staggenborg, S. (1991). The pro-choice movement: Organization and activism in the abortion conflict. New York:

Oxford University Press.Stanley, M. (1983). The mystery of the commons: On the indispensability of civic rhetoric. Social Research, 50,

851-883.

Stanley, M. (1988). The rhetoric of the commons: Forum discourse in politics and society. Teacher’s College Record,89, 1-45.

Stevens, C. H. (1981). Many-to-many communication. Cambridge, MA: Sloan School of Management, Massachu-setts Institute of Technology.

Surman, M (1994a). From VTR to cyberspace: Jefferson, Gramsci and the electronic commons. Host address:http://www.web.apc.org/&sim;msurman/ecommons.html.

Surman, M. (1994b). Wired worrls: Utopia, revolution and the history of electronic highways. Host address:http://www.web.apc.org/&sim;msurman/ecommons.html.

Tatalovich, R., & Daynes, B. W. (1981). The politics of abortion. New York: Praeger.Terry, R. A. (1988). Operation rescue. Springdale, PA: Whitaker House.Tribe, L. H. (1990). Abortion: The clash of absolutes. New York: Norton.U.S. Congress, House of Representatives. (1939). H.R. Doc. No. 340, 76th Congress, 1st Session (Report on the

Investigation of the Telephone Industry in the United States).U.S. Congress, Office of Technology Assessment. (1990). Critical connections: Communication for the future.

Washington, DC: U.S. Government Printing Office.Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A relational perspective. Communi-

cation Research, 19, 52-90.Wills, G. (1989, May 1). Operation rescue: A case study in galvanizing the antiabortion movement. Time Magazine,

26-28.



393

Wolin, S. (1960). Politics and vision. New York: Little Brown.Wright, C. R. (1975). Mass communication: A sociological perspective (2nd ed.). New York: Harper & Row.

Steven M. Schneider is a political science instructor at the State University of New York Institute ofTechnology at UticaIRome, and a Ph.D. candidate in political science at the Massachusetts Institute ofTechnology.



Documents

Creating a Democratic Public Sphere Through Political Discussion: A Case Study of Abortion Conversation on the Internet