The Major Phrase and Utterance

Abstract

Abstract Given that certain aspects of the prosody of a sentence reflect its gram matical structure, it might be thought that they could be accounted for in the simplest and most straightforward manner by computing meas ured intonational, durational and intensity relations directly from the surface syntactic structure. However, as has already been pointed out, there are good grounds for assuming that between syntax and the phonetics of prosody there lies a level of properly phonological prosodic organization. One of the arguments used to support such a level was the independence of the units of prosody from the syntactic constituents onto which they are mapped. For instance, in the artificially slow dicta tion of polysyllabic words, feet and even syllables within a foot can be pronounced with a continuation pitch rise and an intervening pause; that is, feet and syllables can be pronounced with the intonational and durational features normally associated with major phrases. Conversely, in breathless colloquial speech, the syntactic sentences “I’m sorry. I’m late. I got held up” can be pronounced as a single prosodic sentence, that is with the intonational and durational characteristics of one sentence (Crystal 1969). Within the spectrum of normal speech rates, too, the same sentence can often be pronounced equally naturally with and with out internal pauses and intonational phrase boundaries. The melodies and rhythms of phrases and sentences evidently have their own autono mous existence and structure. They are autonomous not only from the syntax, as just noted, but also from the segmental phonology: when words are exchanged in speech errors, the phrase stress stays put, as in “I kept it pinned on the room to my door” or the stranding exchange “You have to square it facely” (Garrett et al. 1980). The autonomy of intonational melodies is a logical prerequisite for exploiting the Greek musical records as evidence for the intonation of Greek speech.