Part III - Mapping the big tent

Summary

Now we've adopted a grammar (code) versus pragmatic (inference) theoretical distinction (Chapter 4), and made sure that there are appropriate theories which can account for pragmatic inferences (Chapter 5), it's time to put our practice where our theory is. It's time to “do pragmatics,” that is, to tease apart grammatical and pragmatic aspects. Unlike the canonized list-of-pragmatic-topics approach, the code/inference approach cannot presuppose ahead of time which phenomena are grammatical and which pragmatic (see also Wilson and Sperber, 1993). Linguistic phenomena do not present themselves to us already classified as to which “component” should account for them (Bach, 1999b; Fillmore, 1996; Green and Morgan, 1981; Hornstein, 1986; Nemo, 1999). Doing pragmatics, then, is not simply analyzing any one of a predetermined set of (pragmatic) topics, but rather, carving pragmatic (and grammatical) bits out of a totality of language use. A theoretical decision is involved every step of the way.

Here's an example from a presumably grammatical case. According to England (2007), certain (Mayan) Mam proclitics, all but one of them obligatory, are consistently associated with specific aspects, as well as tenses. Thus, n– is associated with ‘incompletive’ aspect, and ‘present progressive’ tense, o– is interpreted as ‘completive’ and ‘past tense,’ ok+ suffix indicates ‘potential (i.e., irrealis)’ aspect and future tense, and ma is used for ‘proximate (to some event or reference time)’ aspect and ‘recent past’ tense. These form–function associations seem impeccably grammatical.