Many of these papers are not specifically on the creation of identity per se, but rather on some aspect of personal identity creation and use on the net, as in possible factors that influence its portrayal, or the effects of pseudonymous communication on language usage and/or volume. In these cases it is not what the paper discusses in particular, so much as what they infer or assume about the changing nature of identity, that is of interest.
Most frequently a single element or variable (for example, race or gender) will be examined for its effect on discourse and/or play on the net (Donath 1999, Geierman 1996, Rodino 1996, Jaffe, Lee, Huang, & Oshagan 1995, Nakamura 1995). Specifically, Donath examines the use of deception to enhance or harm pseudonymous reputation on Usenet newsgroups. Geierman (1996), Jaffe et al (1995), and Rodino (1996) examine the nature of 'gendered' discourse in various areas or communities of the net, and finally Nakamura (1995) examines LambdaMOO for replication of RL hierarchical structures through stereotypical uses of race. Research has also been done concerning the effects of pseudonymous communication on the dominant paradigm (Donath 1999, Chester & Gwynne 1998, MacKinnon 1998, Lee 1997, Bruckman 1993, and Meyer & Thomas 1990). The impact of pseudonymous communication to encourage 'protected' speech is examined by both Chester & Gwynne (1998) and by Lee (1997), while the possibility of redefining virtual terminology to defuse and demystify virtual (and possibly RL) attacks by the pseudonymous/anonymous is the subject of MacKinnon's (1998) research. Bruckman (1993) examines how 'gender-swapping' can impact not just work practice but also culture and values, and finally Meyer & Thomas (1990) examine how pseudonymous communication has (re)created and (in)formed the subculture of the 'Underground net.'
Female-'reading' names receive far more attention (which frequently contains implied assumptions of availability and heterosexuality) than male-reading names (Nakamura 1995, Geierman 1996, Rodino 1996). There is also strong reflection and enforcement of societal 'norms' of heterosexuality on both LambdaMOO and IRC chat channels (Nakamura 1995, Geierman 1996). Nakamura (1995), after several hours of observation on LambdaMOO (the MU* on which she chose to conduct her study), hypothesizes that the user-constructed social world of LambdaMOO creates and enforces a protected, privileged space for ethnic 'tourism' that replicates RL hierarchical paradigms and statuses. Geierman (1996) also spends many hours of observation on IRC chat channels. All three studies unfortunately assume in their papers that RL gender equivocates to the 'gendered' name choices of the net users. This is an interesting assumption on Geierman's (1996) part, since he is himself, in his study, playing with portrayals of gender identity that differ from the RL user's gender. Also, Geierman (1996) and Nakamura (1995) both assumed that their particular areas of exploration were generalizable for the net. However, I would suggest that they are both, in a sense, conducting inductive, unobtrusive, exploratory studies that unfortunately are, by their very nature, of a 'tourist' type mentality. In other words, both Geierman (1996) on IRC chat channels, and Nakamura (1995) on LambdaMOO, are short-term 'lurkers': net users that do not contribute to social world building but merely watch. As a consequence the most 'interesting,' 'exotic,' or 'deviant' behavior is what they will most carefully note, without also noting (or even perhaps even realizing) that such behavior is not necessarily the standard for the net subculture in which they are vicariously participating. Also, they may not realize that while it is the nature of role-playing to try on roles other than what is considered quotidian by the role-player, this does not assure any skill at doing so. If, as Nakamura (1995) states, attempting to portray someone of another race is merely tourism, what does that infer about, for example, Geierman's (1996) study, where he used female-reading names even though he himself was male? Nakamura (1995) seems to infer that racial or ethnic identity should not be explored except under certain stringent criterion, but does not answer the question of how one can learn about the life experiences of others without attempting some sort of participant-observation.
Rodino (1996) also observes the effects of identity portrayal on the net, through examination of 'gendered' discourse on Usenet mailing lists. Her study consisted of downloading 3 weeks of mailing list communication and examining it for latent content. Like Geierman (1996) and Nakamura (1995) she too assumes RL gender equivocates to name choice on the net. However, after completing her study she notes her facts indicate both the gender construction and the IRC utterances she observed were not only of a dualistic nature. She concludes that gender should be reconceptualized as 'performance,' constantly (re)created and (re)interpreted. She also notes (as Nakamura  and Geierman  do not) that current binary-based examinations of oppositionally defined 'gendered' discourse do not serve to disrupt or demystify the binary gender system that currently naturalizes RL societal patriarchy. Instead she suggests that reconceptualization of gender as performance would not only accomplish that goal, but also allow researchers to escape the trap of methodology.
Jaffe et al (1995) concur with many of Rodino's (1996) observations. However, their study approaches the issue from the 'other side,' as it were. They offered the group of volunteers the choice of whether or not to adopt pseudonyms on an on-line forum created specifically for the study. They then researched both the differing uses of discourse that pseudonym use offered, and which genders used which forms of discourse. Unsurprisingly (to me, at least) their study agreed with and expanded upon Rodino's (1996) conclusions. They laid out their theories clearly: first, they hypothesized that women will hide their gender more often than men do; and secondly, that pseudonyms will encourage more response volume, with less 'gender-specific' discourse usage. Their study's results supported their hypotheses. They conclude that binary gender differences in communication patterns may be mitigated through use of pseudonyms in CMC channels.
I found both this research and Rodino's (1996) fascinating. Both studies not only bore out my anecdotal net observations concerning gendered discourse, but also verified that net users are deliberately engaging in pseudonymous identity play, apparently both to regain agency and in order to explore (at the very least) the limitations on RL parameters of gender. While neither study clearly addressed my question, they both touched upon it interestingly.
An examination of pseudonymous identity and deception in the context of the Usenet newsgroups clearly represents the virtual community as a communication system, and its inhabitants as 'signalers' and 'receivers' (Donath 1999). The study is specifically ethnographic, according to Donath (1999), and while it is obvious she has examined many Usenet e-mail postings on several mailing lists, she unfortunately does not state precisely the parameters of her research. She attempts an interpretation of closely examined social discourse, metaphorically linking biological deception motivations to those of the pseudonymous communicators on the Usenet. She concludes that in a pseudonymous environment based on non-fictional exchange of information, one's reputation/identity is vital for others to determine the worth or veracity of one's postings. Unsurprisingly therefore, identity deception occurs to take advantage of the benefits of a good reputation.
The effect of net pseudonymity and anonymity on individual creativity, reputation, and expressiveness is further examined in Chester & Gwynne (1998) and Lee (1997). The use of pseudonyms encourages on-line collaboration and community building (Chester & Gwynne 1998), and truly anonymous messages are found to be valuable in allowing users greater access and ability to explore a variety of 'deviant' subjects (Lee 1997). In both studies the 'identities' of the users are completely of their own construction. Chester & Gwynne (1998) explore the nature of CMC as it applies to teaching. They research the environment of the net as a fresh experimental medium needing investigation, and make a determined effort not to broadly apply face to face (FTF) rubrics on education to CMC. Their study consists of a group of less than 30 volunteers who were initially unaware of the true nature of the 'class' they were purportedly attending. FTF communication (including phone conversations) was forbidden, and a net 'realm' consisting of both asynchronous (via e-mail) and synchronous (via 'created' public discussion space) communication methods was created for the students. The two professors note their study is merely an introduction to the subject, but hypothesize that the initial study suggests the net offers many possible benefits to teaching. They recount their observations regarding the use of pseudonyms as personal identity creation to encourage learning, through increased group salience, response participation and volume, and play and creativity. They also discuss "'depaysement,' meaning literally, to 'decountrify' oneself, ... defined as the experience of (re)seeing (Chester & Gwynne 1998 pg. 2)." This (re)formulation is applied by the paper's authors both to their teaching methodology and to the nature of identity creation and formulation on the net.
Lee (1997) does not clearly state where her research is conducted, but does inductively address the asynchronous communication of e-mail and whether anonymous messages should be forbidden. She hypothesizes that anonymity encourages greater freedom of speech and communication, then conducts a documentary examination of the historical RL and VL nature of anonymity and protected discourse. She concludes that the benefits of anonymity (e.g. reclamation of agency through protected discourse) outweigh its potential disadvantages. Unfortunately she also assumes rational perusal of anonymous remailers by the majority of net users, whereas to my experience they're just as prejudiced as public forums. As an example I've seen remailing servers get reputations (sometimes deserved, sometimes not) for hosting textually inflammatory people. Once that reputation is associated with a server, any e-mail received from that server is frequently automatically consigned to a user's 'kill file,' without regard for the pseudonymous identity of the sender.
In both papers (Chester & Gwynne , Lee ) the exploratory examination is of a small number of respondents or postings, and they are both self-described as initial, inductive studies only. Further, the authors note anonymity has both advantages and disadvantages; a lack of identifying social cues does allow anti-social behavior as well as any other kind. Overall, however, they find anonymity and deliberate identity creation ultimately beneficial to social world creation and communication.
In a more generalized vein, Meyer and Thomas (1990) hypothesize that the net 'underground' (in)forms a postmodern rubric. They explore the net underground via surveys and conversations with several noted hackers, and participation/observation on hacker bulletin boards. Their research reveals the net underground is currently being defined by the public in general as 'deviant,' yet its participants view and (re)define themselves both as a mutually supportive fringe subculture operating in a new, 'semi-private' sphere, and as an act of defiance. There is no attempt to fit within or be defined by public or mass culture. The authors conclude the deliberately created, symbolic identities of the net are excellent representations of postmodern behavior, rather than examples of deviance.
Finally MacKinnon (1997) and Bruckman (1993) hypothesize upon their observations of the social phenomena of virtual identity manipulation and its potential effects upon the mainstream society. MacKinnon (1997) explores 'virtual rape,' including a history of the social construction of rape. He notes that identity and behavior is personally constructed on the net, and that LambdaMOO, where the most infamous 'virtual rape' occurred, has as one of its technological constructs the inability to textually or physically force anyone else. He raises the fascinating point that if the user defines both self and behavior, then there is none better to deny the existence of virtual rape than the user -- rationally speaking, rape need not and should not exist on the net. Furthermore the nonconscious importation of the term 'rape,' with all its associated emotional baggage, creates unnecessarily complex problems. He hypothesizes that virtual (and possibly non-virtual?) rape needs to be reconstructed, so as to render it less harmful, or even irrelevant, to the intended 'victim.' Bruckman (1993) speculates regarding the societal repercussions of playing with the nature of identity on the MU*s, specifically through gender swapping. Her research is based on time spent on a MU*, and she reports on her observation, exploration, and participation in the medium through discourse and self-presentation. She and MacKinnon (1997) do not clearly state, but seem to strongly infer, that MU* identity is performance-based, thus conscious reflection upon (and construction of) identity on the part of the MU* players is of benefit in disrupting the current societal hierarchies and paradigms.
In conclusion, there are numerous investigations of identity-exploring and -creating behavior on the net. These range from the effects of pseudonyms and gender on discourse, through the import/exportation of meaning to and from RL as a means to disorder and (re)conceptualize identity, to the nature of individuals who use the net as their chosen medium for identity performance and play. However, as far as I can tell, currently there is no examination of why role-playing net participant/users do this, nor what other elements they might appropriate and 'play' with as they link VL self-construction of identity into RL reclamation of agency. Thus my question hopefully remains as a feasible research project, namely: Do MU* participants use their pseudonymous virtual personas while OOC to express aspects of their personalities they do not feel comfortable expressing in real life?
Last Updated: Tue, 2002-Mar-28