THE PRAGMATICS OF CONSTRUCTIONS
Charles J. Fillmore
University of California, Berkeley
I have been interested for some time in the kinds of decisions linguists have made in drawing boundaries in and around linguistics, by which I mean both the lines that separate one subfield of linguistics from another, and those that separate linguistics proper from impinging disciplines ( Fillmore, 1984). Probably every linguist, while working through some puzzling collection of language phenomena, has had the experience of beginning with the assumption that a given problem will yield to the system of principles that characterize one particular field of linguistics, only to conclude in the end that the explanation really belongs elsewhere. What began looking like a morphology problem turned out to have a phonological solution; what started out as a problem in semantics received a pragmatics solution; or what began as a mystery of syntax proved in the end to be an instance of some semantic generalization. At times, of course, we find that the problem really belongs to cognitive psychology, or ethnography, or logic, rather than to linguistics. The problems that linguists deal with day-to-day don't always have labels on them telling us who owns them.
These surprises are usually not embarrassing, especially in the modern world where integrative results are praised and cross-disciplinary research is encouraged. But in those circles in which disciplinary boundaries are defined in dogmatic or dictatorial ways, one sometimes feels pressure to be sure in advance that the outcome of an inquiry is going to be of the right sort, in the fear that one might not be seen as really working in one's declared field. It's one thing to face the occasional charge that what we are doing isn't really semantics, or isn't really grammar, or isn't really linguistics, but it's awkward to____________________