Phonetic effects of focus and “tonal crowding” in intonation: Evidence from Greek polar questions

This paper deals with the intonation of polar (yes/no) questions in Greek. An experiment was devised which systematically manipulated the position of the focused word in the question (and therefore of the intonation nucleus) and the position of the last stressed syllable. Our results showed that all questions had a low level stretch associated with the focused word and a final rise–fall movement, the peak of which aligned in two different ways depending on the position of the nucleus: when the nucleus was on the final word, the peak of the rise fall co-occurred with the utterance-final vowel, irrespective of whether this vowel was stressed or not; when the nucleus was on an earlier word, the peak co-occurred with the stressed vowel of the last word. In addition, our results showed finely-tuned adjustments of tonal alignment and scaling that depended on the extent to which tones were “crowded” by surrounding tones in the various conditions we set up. These results can best be explained within a model of intonational phonology in which a tune consists of a string of sparse tones and their association to specific elements of the segmental string.