Unacceptably accented auxiliaries '

Wells and Local ( 1983) intend to propose a thoroughly pragmatics-based explanationfor the unacceptability ofcertain instances ofaccented injinitival to, have, and be in English, in reaction to Zwicky and Levin’s (1980) argumentsfor a constraint couched in syntactic terms. A comparison ofthe two accounts serves as the vehicle for a discussion ofseveral larger issues: choosing between theoretical frameworks on the basis ofanalyses formulated within them, metatheoretical considerations in deciding among competing analyses, the interplay between grammatical andfunctional accounts ofphenomena, and the needfor explicitness in analyses ofall types.