ASA 129th Meeting - Washington, DC - 1995 May 30 .. Jun 06

4pSC6. Demarcating prosodic groups with articulation.

Cecile Fougeron

Phon. Lab, Dept. of Linguist., UCLA, Los Angeles CA 90095-1543

Inst. de Phonet., Paris Sorbonne Nouvelle, Paris, France

Patricia A. Keating

UCLA, Los Angeles, CA

It is hypothesized that phrasal position affects articulation, and that what looks like overall declination of articulatory displacement could arise from local enhancements or reductions at particular phrasal positions. To test this hypothesis, linguopalatal contact was measured for both the consonant /n/ and the vowel /o/ in reiterant speech. Utterances were based on arithmetic statements like ``(1+1)x(1+1)'' [reiterant ``(no no no) no (no no no)''], where stress and phrasing were varied by choice of numerals and location of parentheses. For the one speaker analyzed, there was no overall downtrend in contact of /n/ through any sentence type when all syllables---stressed, word-initial, or phrase-initial---were compared. Instead, word-, phrase-, and sentence-initial /n/'s have cumulatively more contact than medial and final ones. It is proposed that greater initial contact contributes to marking prosodic boundaries by enhancing the consonant--vowel contrast. Also enhancing the CV contrast is the more open position for phrase-final /o/'s. As a result, final edges of domains are marked by a greater articulatory distance from final /n/ to /o/, and initial edges of domains are marked by a greater articulatory distance from final /o/ to initial /n/.