If every text has its natural history, the one before you is no different. Indeed, the various chapters are the culmination of a long interactional and redactional process, which began in 1986 when the editors organized a working group around the theme of "Text and Social Action." In its first phase, the group met over a two-year period, at first on a monthly basis and later quarterly.

It would be wrong to think of this work as one in which complete agreement has been achieved. It would be better to think of it as dialogically constructed, with the dialogue as often agonistic as cooperative. We began with a common problematic framed by the editors having to do with different views of "text," as a structure of discourse nevertheless embedded in richly contingent social action versus something formal, along the lines of text grammar, in which a text is seen as a larger-than-sentence-level output conformable to something like a grammar.

An important first breakthrough on our way to Natural Histories of Discourse was the realization that the two views of text were not incompatible alternatives, but linked to distinct phases in the "entextualization" process, as we came to call it, the (at best) intersubjective and sufficient determination of a "text"—though divergent actor-perspective entextualizations also inhabit interactional realtime. We began also to draw out the dialectical relationship of entextualization to its complementary phase, "contextualization," the ever-changing "appropriate-to/effective-in" relationships that concurrently develop between text and nontext.

We began by exploring the ways in which social actors seem to lift texts from contexts, the transformations that even "the same" discourse undergoes in the process, and the social conditions making possible such entextualization in the first place. This is the observed or patent "natural history" of textuality. Analyzing such history in terms of decentering/recentering, we produced and discussed an initial set of papers, many of which matured into contributions to Natural Histories of Discourse.

A second major breakthrough occurred with the realization that even personal deictics, such as English I and you, and so on, which seemed to have fixed, understandable meanings as elementary shifters, in fact do so only through an intertextual projection at the metadiscursive plane. Recall that personal deictics present their referents as the individual(s) presupposed to be inhabiting a particular interactional role, like "speaker" or "addressee(s)." They describe their referent(s) as the role inhabitant(s) they index, and therefore have metapragmatic denotational backing; they are, in short, explicitly metadiscursive. But in what culturally specific ways does an individual come to inhabit such an interactional role? Are such interactional roles always the same? We realized in our discussions that simple pronominal forms are, in fact, complex in semiosis. Because the pronouns are found across distinct texts, the absolutely unique, contingent specificity of the history by which individuals come to inhabit relevant roles in each instance seems to give way to an abstract essentialization, which can then be read into any given text. The attempt to fathom the meaning of that projected essence results in the realization that there are many (perhaps infinitely many) distinctions that can be regarded as features of the essentialized categories denoted as "first" and "second person." These tend to cumulate in discursive historical processes which we read as tropes such as renvoi, reference, reanimation, re-presentation, and so on. With these insights, we began to explore how such projections or essentializations took place in actual empirical cases, and we looked beyond the realm of elementary personal deixis.

A first attempt at public presentation was a panel at the annual meeting of the American Anthropological Association held in Chicago in autumn of 1987. The papers from the panel were further discussed at subsequent meetings of the group, which included new participants as well as new papers. These old and new papers were revised, presented, discussed, and debated at subsequent meetings through the fall of 1988, when the group formally disbanded, although it met in different guises and at different times for two more years. It was agreed that the editors should assemble the papers into the present volume (14-16).