Özet:

Our approach of the intonation of spoken (read and spontaneous) Turkish is based on P. Martin’s model ‘Stockage-Concaténation Incrémental’ (SCI). Our claim is that the minimal ‘prosodic words’ are hierarchically organised according to several levels but, contrarily to the metrical autosegmental model, this hierarchy is based on mechanisms which involve not only boundary tones but also the ‘prosodic contours’ on stressed syllables. This implies of course to get the right rules of lexical and grammatical units in spoken Turkish in order to define the relevant prosodic words and determine stressed syllables characteristics. In the SCI model we claim that these melodic contours are not distributed at random but that they are indeed the very basis of the specific mechanism which indicates the prosodic hierarchy. These contours contrast syntagmatically according to FO and vocalic length variations which are to be measured. Thanks to them the hearer can reconstruct the prosodic structure which was intended to by the speaker. In this process the global prosodic structure is understood dynamically as successive melodic contours are perceived

Anahtar Kelimeler: