ASA 127th Meeting M.I.T. 1994 June 6-10

2pSP21. Vocal effort as a cue for linguistic stress.

Agaath M. C. Sluijter

Holland Inst. of Linguist., Phon. Lab./Leiden Univ., Cleveringaplaats 1, 2311 BD Leiden, The Netherlands

Intensity differences as a function of stress are mainly located above 0.5 kHz [A. M. C. Sluijter and V. J. van Heuven, Proc. ESCA Workshop on Prosody, Lund, 246--249 (1993)]. Results of a perception experiment bear out that intensity manipulations in this region provide stronger stress cues than uniform intensity differences do, and are close in strength to duration differences. The differences in energy in the higher bands were arguably caused by a difference in the shape of the glottal-pulse waveform. However, energy differences were measured in separate filter bands, thereby not removing the effects of vocal tract and lip radiation. These latter effects could also be different for stressed and unstressed syllables. Now, instead of investigating the energy differences in separate filter bands, voice source and vocal tract were studied separately. Ten American subjects produced minimal stress pairs. The differences in the glottal-pulse shape with which the vowels were produced were determined using inverse filtering. By removing the vocal tract and the lip-radiation effects, the glottal-pulse waveform is revealed. The hypothesis was tested that the glottal-pulse waveform of a stressed vowel is more pulse-like than that of an unstressed vowel, which in turn has a more sinusoidal shape.