When Edward P. J. Corbett was editor of College Composition and Communication, his fairly rigid standards for article length undoubtedly had the effect of forcing some loose thinking to a fairly sharp point. It also had the effect of pushing some discussions into an awesome degree of compression that made them less available to casual readers than they might have been otherwise. On the other hand, they offer a rewarding read now, if one is willing to commit the mental energy to put them together with the world. To my mind, a classic illustration of this sort of essay is George Yoos' An Identity of Roles in Writing and Reading from the fall, 1979 issue. In that paper, Yoos provided a model for reading and writing processes that finds reciprocity between writing and reading strategies at four different levels-at the level of objective expression or of content, at the level of face-adjustment or ethical appeal, at the level of audience, and at the level of logic or truth. Under this system, both writer and reader perform in roles defined by these four topics, and if one is generally accentuated in any specific situation, it is pretty clear that accommodation or sensitivity to all roles can provide a highly enriched perspective on writing. However, any conceptualizing like this, anchored in Collingwood, Croce, and George Herbert Meade (the names cited here) is probably going to seem rather alien and have some apparently rough points for present day readers. One of these is Yoos' flat assumption-deriving from Collingwood and Croce-that Kinneavy's effort to see expression as a mode of communication is wrong, and that the need to keep expression separate from communication is basic to an understanding of the writing process. Our present pedagogical tendency of using personal expression as a way to develop fluency and authenticity will tend to make readers unreceptive to the basic truth that writing will always be writing, that is, texts in which expression can be found, but which should never be confused with expression. To ignore this fact is to run a far graver risk of creating writing anxiety than would be possible by framing writing as an impersonal formalistic game. Another rough point would have to be Yoos' notion of the faceadjustment role, which he identifies with ethical appeal as a matter of clear about what one is doing. Yoos draws a clear distinction between this and the audience role which involves a strategic awareness and management of how different audiences will react, and, from the reading point of view, a reader's awareness of how these audiences are being managed. These are very subtle distinctions that take us quite a way back to a classical view of rhetorical operations (pace Knoblauch and Brannon). More generally, Yoos makes it clear that the relation between writing and reading is much deeper than writing scholars tend to acknowledge, in spite of the years of research into reading and writing connections. Certainly the kind of mirror-imaging that his essay provides-in which, say, an objective-expressive is one the writer plays by getting what he or she knows down into words, and that the reader reads for to see what the writer really knows, as a ground to be comprehended before processing rhetorical and logical acts-involves a complex conceptualizing of the communication process that promises a very rich critical