Introduction

Although great strides have been made recently in the phonetic analysis of prosodic features as a result of technological developments, the advances in phonological analysis have arguably been less impressive, at least with respect to the phonology of intonation systems. This is evident if one considers a traditional concern of phonologists: the comparison of related languages or dialects. A great deal of work has been done on comparative phonology at the segmental level, and a good deal also on comparing lexical tone, stress and accent, yet there is strikingly little published research that systematically compares the intonational systems of related languages, dialects or accents. For example, in John Wells' compendious study of English accents, intonation is given little space (Wells 1982). Cruttenden (1986) notes that, although there have been scattered individual studies of different accents, these have not been used as a basis for systematic comparative work. This is somewhat surprising, given the increasing recognition, within the world of speech technology as well as more traditional applications of linguistics, that prosodic features of intonation are centrally involved in the comprehension of speech and the management of conversational interaction. Surely the ability to handle dialectal prosodic variation will be a sine qua non of a successful speech recognition system, and, one would hope, of an acceptable speech synthesis system too.

This omission is not happenstance, but derives from a general unwillingness on the part of linguists to treat ‘intonation’ with due phonological respect. In particular, there has been a failure to recognize the importance of warranting, from the observable behaviour of naive native speakers, the functional categories that intonational features are held to be exponents of.