In the first conference situation students from two different campuses participated in the same conference as they set about the task of investigating the new world of a computer conference.

The assignment presented the original newspaper reports of the 1960s Sit-Ins as online text and stipulated a minimum number of responses. In their writing in the conference, the students used repetition and emulation to establish tacit conventions governing their exchange of ideas, opinions,and feelings, to confer authority upon themselves and display their recognition of authority in writings by their peers, and to enculturate each other to electronic discourse conventions in their discussions of socially sensitive issues. Elsewhere, we discuss the impact of same-campus and linked-campus audience for these conferences in detail (Davis & Brewer, in press 1997); here, we look at ways students signalled identity by varying modality within individual responses to suggest confidence and stance, and by individual patterns of shifts, graphically represented by type-token ratio analyses of student writings.

The second set of conferences was situated on a single campus within a linguistics class combining graduate and undergraduate L1 (American English) and L2 (Chinese, Japanese, French and Arabic) speaker-writers. Their cultural and linguistic repertoires presented differing cues for politeness, particularly those affecting question-answer routines, compliments, hedges, and single- or multi-party disagreements in public situations.

To become a discourse community, however ephemeral, the group needed shared tasks, shared sociorhetorical goals (Swales 1990) and shared knowledge. A sequenced set of conferences required them first, to post summaries of classic articles and respond to any two postings, and second, to participate in a staged debate. They participated in this conference as dyads, selecting a chosen text from a list of current, controversial studies.

Each dyad devised a format in which to post both support and refutation for the study's design, findings, and underlying assumptions, after which each individual would respond to postings by other partnerships. Here, we look at how the L1 and L2 student cohorts used modals to shade their expression across different text-types, and how one Mandarin speaker modified the use of compliments in her individual rhetorical pattern over the timespan of the two conferences.

Each set of conferences presented possiblities for monologue, dialogue and multilogue; their structure enabled participants to move from emulation and repetition of features from text as authority to those found in conversational exchange between and among each other.

In brief, the two conferences presented student participants with two different notions of textual authority. The Sit-Ins conference presented online texts of primary-source material, the chronologically-arrayed newspaper stories from the first week of the Sit-Ins movement.

These texts served as a kind of Constructed Authority. In their individual responses to the authority of those texts and to the writings of their peers, the student writers conferred authority upon themselves, presenting and guarding their identities of knowers by virtue of ability to understand, appropriate and manipulate the authority of text or by virtue of ability to remember, recognize, or project empathetic responses to situations of racial conflict.

The Language-Learning conferences required students to read authorities offline; their summaries and debate-position writings in the conferences became a sort of Reconstructed Authority. Their individual responses, directed to other students as knowers of text through their acts of reconstruction, also included their own self-presentation as knowers capable of manipulating authority of text and of recognizing and appropriating issues and situations in daily life outside of texts.

We tagged the electronic texts at the word-level and created a concordance in order to identify lexical variation characterizing individual and group patternings and to index and chart discourse features that suggested how students were employing what Bell calls "audience design" (Bell 1984).

In the next section, we present an overview of the repetition and emulation that characterises the rapidly-composed, extemporaneously keyboarded electronic discourse presented in the student responses from both conferences.