Part 4 - Identity and the Internet
Connections and Research Directions
In doing research on CMC, the specific questions I will ask should focus upon and integrate three theoretical elements: 1) the symbolic interactionist concepts of using significant symbols (language) in a dynamic relationship between selves, a process of meaningful exchange, consciousness formation, identity formation, and (reciprocally) societal formation; and 2) the identity concepts of salience, commitment and relevance and the attendant social support concept of the importance of role-position and relationships to psychological well-being. Furthermore, each of these elements can be investigated empirically using a social network perspective, where the properties of the social network such as density, multiplexity and intensity of relationships form the basis for understanding the symbolic interaction that is important to identity formation and psychological well-being. Each of these concepts is tied together by the empirical circumstance of contemporary society (namely the introduction of a new medium of communication - CMC). The way in which questions and hypotheses are formulated regarding this "lifeworld" will be determined by the content that is found therein. This is an exploratory study and exploration is research that is tractable and that involves a sharpening of focus as the research proceeds. "A variety of techniques may be needed to achieve [focus]," say Lauer and Handel (1977:314). "In his own work on movies, Blumer used five different methods: autobiographical data, personal interviews, a collection of conversations about movies, direct questionnaires, and direct observation of children." This study may also need many methods, including content analysis, survey methods, and non-participant observation to fully understand the ramifications of the internet upon society.
An examination of any studies of this nature performed in the past will inform the development of hypotheses, as well as contribute to efforts at replication. Unfortunately, there is not much precedent for such studies. Prior to the mid-1980s, Fischer describes "the sorry state of the sociology of technology, examin[ing] its decline, theoretical confusion, and empirical vacuum. No cumulation of empirical research on the social consequences of technology [exists]" (Fischer, in Castells, 1985: 284, 290). More recently, a researcher at Stanford University indeed found only one behavioral study involving human-computer interaction prior to 1985.
Fortunately, largely as a response to recent increases in home computer networking,30 much more work has been done since the mid-1980s. Rogers (1987) makes many good suggestions for conducting network research on communications technologies. One strategy he proposes rests on the fact that network analysis focuses upon interaction or "interpersonal relationships of information-exchange" as the units of analysis, rather than on individuals. Thus, "the investigation of interactive communication systems best combines network analysis (a) with content analysis of the messages that are exchange [sic], and (b) with time , so that the behaviour can be studied as a process" (1987: 303; original emphasis). Studying the process of message content is precisely the kind of research that is supported by the tenets of symbolic interactionist theory.
A foundational element of symbolic interaction is the exchange of significant symbols, language. "The seemingly objective and taken-for-granted character of the social definitions of reality can be seen most clearly in the case of language itself, but it is important to keep in mind that the latter forms the base and instrumentality of a much larger world-erecting process" (Berger and Kellner, in Farberman 1973: 72). Examining in detail the character of language use in cyberspace should be foremost on the list of data inquiry procedures for this study. This leads to the question, how is the communication being conducted over the internet demonstrably different than that which has accommodated face-to-face communication or telephony?31 One of the most apparent and most important differences is the technique that is used. The primary act involved in CMC is that of writing. The character and content of communication can be altered by technique. How? Jones (1995) quotes James Clifford on the technique of writing: "When we write we do so from a necessarily local setting." In other words, writing takes place in isolation from immediate response and negotiation. The act of writing is "intensely local, for, although we may be certain of an audience, we are unable to verify its existence just as we are unable to verify its interpretation of our writing" (Jones, 1995:12; original emphasis).
The language of computer communications itself leads to altered or conjugated meanings for individuals. One of the most useful techniques for making computers more "user friendly" is to employ familiar language; for example, adding only the prefix appendage "e" (for electronic), as in e-mail or e-conferencing, or to employ casual, non-threatening language which identifies mental spaces as "chat rooms." The use of analogy to the present language of communications undermines the fear of an unfamiliar communicative realm. Despite the fact that it is a medium predominately of writing, CMC is often thought of as an oral medium, a conversational space like the telephone, rather than the more formal space of print. Nicholas Johnson thinks of looking at his e-mail conferences as going to the "pub" and entering an oral conversation; as he says "I dont yet have the new language I need to think and behave with computer communications. And, until such language exists, thinking about computer communications as oral seems to work better than thinking of them as print." By analyzing the language shared in e-mail and chat rooms, the ways in which such semantics (as described above) will effect message content may be illuminated and may help to explain why these particular techniques of communication have become so wildly popular over the course of a few short years. It may also help to explain how CMC mediates the social construction of identity and creates a unique cultural community that shares given meanings throughout cyberspace. Analyzing the language of CMC is also important to the network aspect of CMC. By examining the content of specific relationships that are formed via CMC, I can assess the intensity and intimacy of such relations qualitatively, and I will be able to examine in detail the multiplexity, or role context, of given relations based upon this content.
The second aspect of Rogers suggestions toward a network study of electronic communications is the aspect of time. An analysis of this dimension of cyberspace - whether it be the synchronous time of chat, or the asynchronous time of e-mail - will help to reveal the maintenance of identity salience and also reveal the hierarchy of identities commitment. Recall that identity salience is operationalized as the likelihood that a given identity will be invoked in diverse situations. In CMC, diverse situations arise often due to the "open" nature of the medium - anyone can enter into a "conversation." Reviewing transcripts of chat or e-mail sessions at many different points in time will hopefully allow for the recognition of changes and/or stability in participants psychosocial status. Thoits (1982: 31): "Social identities provide actors with existential meaning and behavioral guidance and these qualities are essential to psychological well-being." And by using a "reverse process" and examining the behavior and the existential meaning that is revealed through content analysis over time, the concomitant social identities should become obvious.
Such a study that reviewed transcripts in this way was done by Danowski and Edison-Swift (1985). They studied the 2,592 messages on the e-mail bulletin board of a state university-connected extension service that were exchanged over the course of one year. Part of their analysis was an investigation of the frequency of occurrence of different words in the network messages that were exchanged before, during and after a hiring/firing crisis (Rogers 1987) (note that this technique has become the basis of qualitatively oriented data organizing programs such as HyperResearch and N.U.D.I.S.T. and could be easily reproduced within this study). There are a number of benefits to using computer-monitored data, such as e-mail conferencing, in this way as opposed to using standard sampling methods (Rogers 1987:305). First of all, the research resources needed for the collection of a large amount of data are significantly fewer than mailing questionnaires or doing personal interviews. Second, response bias is cut down because of the relatively unobtrusive character of the data collection. Third, as has been mentioned, the reports received from ongoing sessions between the same participants facilitates a collection of continuous or frequently discrete longitudinal data.
An example of another study which collected data in this manner, albeit not unobtrusively, is Corrells Ethnography of an Electronic Bar (1995). Based on her participant observation, Correll was able to note many ways in which the electronic cafe she studied (the "Lesbian Cafe") allowed for the construction of lesbian identities and the negotiated, ritualistic meaning of the "space" of the bar which they created and serves to reinforce their structural position, despite its cultural devaluation.32 Furthermore, Correll identifies different typologies, or identity options, for people "entering" the cafe: the Regulars, the Newbies, the Lurkers, and the Bashers, all described using the same language developed within the network itself. These typologies, and others, exist (under different names; e.g. Bashers=Flamers) throughout the realm of Internet Chat and BBSs and are frequently mentioned within . These typologies should act as another guide to content analysis. Such role-identity types in CMC may be unveiled by the presence of strong, emotion-laden communication.
Another issue centering around identity salience is Gergens hypothesis regarding the "saturated self." According to Gergen, "one detects amid the hurly-burly of contemporary life a new constellation of feelings or sensibilities, a new pattern of self-consciousness" (1991: 73). As mentioned, this syndrome he terms "multiphrenia."33 Multiphrenia is exacerbated by the increasing "number and variety of relationships in which we are engaged, potential frequency of contact, and expressed intensity of a relationship" (1991:61). The onset of the computer revolution is seen to be one of the most dramatic ways in which these increases have occurred, and it seems logical that some semblance of this "new pattern of self-consciousness" will be evident upon an investigation of that medium.
How are people adapting to increased role-possibilities and intense relationships and what possible effects is this having upon psychological well-being? This can be studied using the concepts of identity salience, as well as network concepts. A connection exists between identity salience and this process of modernization (contemporarily known as "globalization"). As more and more relationships are in contention and more role-identities come into existence (due to greater capability to communicate and a great variety of possible relations to one another), the self becomes "saturated." This could be interpreted empirically as greater interconnectedness among social groups, or, in network terms, in high density and multiplexity among those employing CMC. As I have maintained, a network approach can aid in the analysis of computer-monitored data to discover the social relationships that benefit or detract from psychological well-being. "Since the early 1980s networks have played an increasingly important role in such studies. By using the network paradigm, social support researchers have been able to extend their studies far beyond just the ego in an ego-centered, personal network."
I have suggested that the dimension of multiplexity can be observed through a content analysis of CMC; does this same opportunity exist for the examination of network density? In fact, computer-monitored data is an excellent source for cutting down on the methodological problems of investigating network density. "One past problem in measuring network links in sociometric surveys," according to (Rogers 1987:305), "is that the number of names of network partners that can be provided by a typical respondent is only three or four. Unfortunately, such data only reflect respondents strong ties, leaving weak ties,34 which are often crucial in the communication of certain information, unreported and hence unanalyzed."35
The interconnectedness within and between members of given chat or e-mail groups can be thoroughly studied using logs of the "conversations" that take place there. The communications between members of these groups can be studied to find out the extent to which the participants within a given subset or boundary of chat rooms or e-mail lists have some proportion of relationship to one another. One benefit of using data such as chat logs is that the data can be scrupulously analyzed, modeled and applied to a network analysis. This will entail defining the positions that exist within such a medium, for positional relatedness is fundamental to network analysis. However, it should be understood that positions do not have an a priori existence in society. Rather, they are constructed through the dialectic that occurs between people as they attempt to fulfill the requirements of socialization. This construction involves a mediation between persons using the technology of communication. As the technology of communication has changed, the social construction of societal "positioning" also may change; thus it will be important to have a clear definition of the social positions that exist in CMC. Correll (1995), among others, has begun such a typification; this study will build upon and continue that kind of work.
The question of where to draw the lines of a network boundary is a difficult one, regardless of the network involved, due to the fact that networks (like communities) seem to blend seamlessly into one another. Knoke and Kuklinski ask: "Where does one set the limits when collecting data on social networks that in reality may have no obvious limits? Sometimes the answer is not obvious" (in Rogers 1987: 297). In some instances the boundaries of a CMC network may be easier to define due to the limited nature of the space that has been intentionally created by the users. However, to what extent should one study the relations outside of those avenues of relationship to get a sense of a users whole social network and the place that cyberspace takes within it? This might be attempted using surveys or personal interviews with a number of users that seem to have particular relevance to the study (being remarkable in terms of demonstrating recognizable role-identities or psychological state as found in an initial content analysis). A model of survey sampling in network research that might be adapted to this study can be found in Fischer (1982). Another possible survey technique is Kuhn and McPartlands open-ended "Who Am I?" measure, helpful in revealing role-identities, or self-conceptions of ones position in the social structure of the network (Thoits, 1991: 104).
Much network analysis, including Fischers and Wellmans, has been done in urban sociology. To the extent that computer networks embody the latest technology of modernization, they are equivalent to urban environments, only a much more acute case. Given this fact, it will be important to find out the extent to which the urban phenomenon coincides with the CMC phenomenon. What findings in urban network analysis might provide clues as to what to look for when studying CMC?
The use of content analysis and surveys in this study will help to understand not only the state of personal identity hierarchies as they appear in CMC, but also will help to examine some fundamental questions regarding the state of community on-line (the crux of most urban network studies). Based on his work on urbanism, Barry Wellman, in 1973, presciently forecasts that "there will be increasing use of non-face-to-face media of communication. The possibilities [for this communication are far-reaching], including intense emotional relationships, such as love relationships and the mourning of a loved ones death."36 Twenty-five years later these possibilities have come to pass.37 One of the tasks that this research project will undertake is to attempt to find out why.
While not conventional, the integration of symbolic interaction, identity theory, social support and a qualitative method of network analysis will likely provide a unique window on the recent phenomenon of computer-mediated interaction, specifically chat-rooms and e-mail conferencing. One potential problem that I have not yet mentioned is the ever-present ethical problem inherent to social research of confidentiality. While this problem is ameliorated somewhat by the fact that personal information which could identify a user may not be present (especially in chat rooms, where one can share as much or as little as one cares to regarding their objective personal identity), e-mail conferencing will present a greater difficulty, for acceptance will have to be given by all present and future members of the "bulletin board" before mass downloads of transcripts can be made. However, given the "openness" of the internet community previously cited, this may not be as difficult as it seems. Still, this may restrict the population studied to a convenience sample and selection bias may be a concern in this study. But I feel this to be a minor concern, given the dearth of research yet done on CMC. The greater project here, the guiding influence of this study, is to explain how cyberspace allows us to know our selves better and to recognize the burgeoning relational qualities of self and society through the use of computer-mediated communications. In order to attain the goal of this project, several research questions need to be addressed using the quantitative and qualitative methods I described above.
First of all, according to Gergen, CMC leads to the "saturated self;" one research task is to look at the density and multiplexity in users of CMC which, if generally higher than the density and multiplexity of non-users (as described in Fischers 1982 or Wellmans 1983,1989 research), might indicate that people really are inundated with a greater number of relationships Another way to examine this is through either private e-mail or chat interviews, or through collective interviews in chat rooms.
A second research question might center around Fischers understanding of the community question. According to Fischer (1977), "limitation on the number of potential social relations available to individuals leads to more communal social relations." Conversely, then, when the number of potential social relations is extended to a global extent through the use of the internet, does an entire breakdown of communal social relations ensue? If so, what character does this breakdown take? If not, how do communal relations retain a significance to individuals? What issues do people organize around? Again, these questions can be organized around both a social network analysis and a review of qualitative data. Castells (1997) work on the power of identity also provides clues as to how to answer such questions.
Regarding Thoits (1982, 1991, 1992) work on identity-relevance, the question arises, how does identity become "relevant" on the internet? What perceivable effects might these identity-relevant experiences have upon psychological well-being? A quantitative perspective could rely upon the survey data collected from an on-line questionnaire. A qualitative examination of the language of users, particularly the language and lexicon used to describe strong emotion, will be useful in explaining the relation between identity construction and psychological well-being.
These and other questions will guide the beginning of this study, and it is anticipated that new questions will arise in the future. For these original questions, as well as for those that develop later, specific hypotheses will be arrived at so as to test the quantitative data. For the qualitative data, a journal will be kept with care pertaining to all interactions and impressions, as well as any pertinent transcripts from the e-mail or chat interactions. It is expected that this blend of methodologies will procure meaningful results from an intense examination of identities and the internet.
Previous Section - Social Networks
Back to Title Page