As stated above,  the distancing effects of writing have constructed a (somewhat artificial) binary in the theory, that of “audience addressed/audience invoked.”  While many have sought to complicate this binary theoretically, CMC technologies further complicate it by reintroducing the common discursive space of oral rhetoric, while disrupting traditional notions of audience by the nature of that space.  As we will see, writers face challenges both in addressing the “actual” audience and in constructing an audience and asking the participants to play those roles.  The virtual nature of the space makes analyzing the readership difficult, while the dialogic nature of the space allows participants to actively resist the roles required of them by the prose. 

If “empirical descriptions of discourse communities are difficult at best” (Rafoth 140), analysis and description of a virtual discourse community is even more complicated.  Discussions of the ways in which writing, and specifically the distancing effects of writing, complicate notions of audience often speak of the presence and visibility of the listeners (Ong, Kirsch and Roen, Nystrand).  This visibility is key in two ways: first, in simply keeping the very notion of audience present in the mind of the rhetor (Nystrand), and also in providing instant feedback in the form of nonverbal cues (Willard and Brown).  However, visibility becomes more complicated in AOD spaces: while some spaces (usually web forums such as About.com) provide some visibility in the form of visitor lists (or subscriber lists in the case of mailing lists), the visibility of the participants is generally indicated by a posted response.  CMC theorists such as Laurie George and Gail Hawisher and Charles Moran have explored the lack of nonverbal cues on-line and its possible effects on the writing practices; one prominent theory is that this lack contributes to a lack of audience awareness, which in turn contributes to anti-social behaviors like “flaming.”  One cannot, however, accurately say that, like Nystrand’s writers, AOD participants “must often be reminded of their readers” (4), for they often are reminded of their readers in the form of response. 

Within the composing of an individual utterance, however, these “real readers” often must be conceptualized in the mind of the writer, for the fluidity of AOD space makes any attempt to analyze the concrete audience problematic.  While a mailing list such as Politics, or even a web forum like USPCE, might be held to have a relatively stable set of participants, the arrival (invisible to the general group) of a new participant can quickly disrupt what the writer believes to be her or his audience. 

Mailing lists like Politics, with a relatively small number of active participants (37 during the period of observation), often have a somewhat stable membership.  In a discussion labeled “once again, the social nature of this list,” several participants indicated that the list had a fairly stable set of regulars, and discussed whether the list itself constituted a community.  One person pointed out that “some of us go to the parties, and get to know each other. others of us can't,” and also indicates the fairly stable membership by pointing out that, “people (including me) go away, yelling frantically that they'll never post here again. a few months later, they're back...” Given this relative stability of membership, posters might assume a relatively constant audience, one whose reactions and predilections might be somewhat anticipated.  The posters would seem, therefore, to be able to “address” a real audience which is a known entity.  Moreover, that the subject would become a fairly active thread indicates at least some awareness of the issue – that is, participants not only assume a relatively stable audience, but are conscious of this assumption and some of the possible effects on their discourse.  Knowing who the participants are (or are likely to be) does not always mean knowing precisely how they will react: as in any arena, people are unpredictable, and can disrupt expectations.  However, on a list which tends to involve recurrent political themes and ideological questions, the patterns of behavior can be at least somewhat predicted. 

The sudden presence-through-posting of a new arrival might therefore pose a disruption in this stability, leading to a variety of reactions.  On May 18, 2001, a new participant (“Philip”) posted on Politics for the first time.  Of course, given the nature of visibility in AOD spaces, the length of time he had been lurking is unknown.  His first post was a brief response in a longer thread about Attorney General John Ashcroft’s daily “prayer meetings,” a response that went without reply.  The lack of reply could be attributed to a number of factors, but the pattern of response (and lack thereof) to Philip’s initial posts could be seen to indicate a reluctance on the part of other participants to engage an unknown audience.  While 14 of the 61 posts in this thread also went without reply, his was the only “first tier” response (a reply to the initial post) that went without reply.  The other 13 posts which did not receive replies were substantially further down in the thread; 5 were off-topic jokes about what color “Republican-Colored” glasses might be.  The lack of response might have been due to the content (“Hold that thought for about 3 years, then let's test it for veracity.” followed by a complete reproduction of the post being replied to) or the format (one line over a reproduction of the post, rather than replies below the relevant sections).  However, similar one-line replies did receive response, some of which (“Didn't "the people of Missouri" pretty much tell Ashcroft to go take a hike?”) contained little more substance than Philip’s post. 

One ignored post would not be enough to conclude that the silence was connected to a disruption in audience expectation.  However, Philip’s second post, this time a deliberate challenge to another poster, was likewise not responded to, also in a highly responsive thread.  Later responses were characterized by a variety of factors.  His status as a new poster was overtly recognized (“Because I don't recognize your name, I assume you are new here”).  He was also warned about other posters (“In any case, I hope you have learned a lesson.  If you don't agree with Peter, and are trying to make a different point, that means you can't read”), which may or may not be read as attempts to determine his predilections or even influence where in the long-standing  rivalries he falls, thus making him part of the relatively stable audience.  Another participant implied that he might not be fully cognizant of the discussion (“Actually, I posted a long article that articulated just that.  Too bad you missed it.”); Finally, he was dismissed as an unworthy participant by another member (“I suspected you were ditzy and not worth reading, but now I know it for sure”).  Interestingly, this last could be construed as an indication that Philip has been accepted as part of the known audience for that poster; his allegiance, predilections, etc, are now understood. 

The ways in which participants conceptualize a familiar audience is further seen when a familiar poster begins posting in unfamiliar ways.  When “Peter” posted an analysis of the overall tone and content of the list, another established poster questioned “who [was] writing from Peter's account.”  This exchange also hints at ways in which identity, including authorial persona and ethos, is constructed by audience. 

The disruption of a new and unknown participant in conceptualizations of audience can in some ways anticipate the further difficulty in conceptualizing audience as a “created fiction” (Long 225) in the minds of the reader.  While proponents of the invoked notion of audience may speak of the roles the reader is called upon to play (Long, Ong), or even “forced into” (Willard and Brown 40), Tomlinson points out that real readers often resist those roles.  Ong’s analysis of Hemingway posits a reader who, for the sake of enjoying a novel, will adopt the role of friend-in-the-know to the writer; presumably, the reader who does not wish to play that role will either not read the novel, or not enjoy it to the same degree.  However,  Parks points out that readers may choose or reject those roles to varying degrees: a person reading a gardening magazine to alleviate boredom may choose not to adopt the role of gardening enthusiast called for by the text, choosing instead to be adopt the role of amused outsider (183).  Motivation for reading, social context, and situation likely play a strong role in this choice: a person reading the gardening magazine to alleviate boredom while waiting in a doctor’s office is less likely to be motivated to adopt the required role than a person reading it at a gardening show. 

This resistance to imposed roles can be seen in an exchange on USPCE.  In a thread titled “Inept Teacher Training” (begun as a critique of the education of future teachers in the United States), Getlost2002 posted of her  pursuit of a career as a teacher, and the challenges she knew she would face in the classroom, later reiterating her altruistic motives for this career choice.  In casting herself in the role of altruist facing long odds, she also invites the reader to adopt the role of sympathetic, perhaps even admiring supporter, only to have that role soundly rejected by another poster. 

Getlost’s initial assessment of a previous poster’s message as “Beautifuly put!”  establishes a non-agonistic tone, which allows statements which might otherwise be read as confrontational to be instead read as communal.  Her statement that “somedays [she is] not sure why” she is pursuing a career as a teacher likewise constructs a relationship with the reader in confessing her own doubts.  Thus, her early question of “Do you know what I face when I graduate?” can be read not as a challenge, but as an invitation to share these doubts and to empathize with her situation.  The reader is thus invited to share in her worry over what she might face in the classroom, while admiring her for wanting “to make a difference in at least one child's life.”   Her explanations of family life invite the reader to identify with her by identifying with her parents, in particular with her father who may embody the more conservative ideals (coaching a team to win, rewarding effort rather than bowing to pressure to have all children play) that she herself does not share. 

Theresa Enos argues that in ethical argument (argument based largely on ethos), “ultimate persuasion” is effected through “ultimate identification” (111) – that is, the reader becomes persuaded by identifying with the audience constructed by the writer.  In the case of Getlost2002’s post, the reader becomes persuaded by identifying with the sympathetic (even empathetic) friend of an altruistic teacher-to-be.  This reader accepts Getlost2002’s construction of herself and her situation. Indeed, the initial poster in the thread, who began by expressing cynicism about the state of teacher education (and about future teachers), chooses to adopt the role, wishing her “Good luck.” 

The reader may, however, choose to reject the role invoked; in the dialogic space of AOD, he or she can make that refusal known.  Tarheel1978 chooses to refuse the role imposed on her/him by the text,  instead saying that he/she “wouldn't want you teaching [his/her] children with that attitude.”  Instead of accepting the role of sympathetic/admiring supporter, Tarheel1978 adopts a more agonistic role; the hardships which Getlost2002 outlined as proof of her determination and altruism instead become proof of her “attitude” (presumably bad).