Abstract:
A kinematic analysis of midsagittal movement (EMMA) data was performed for upper lip protrusion from [i] to [u] in [iku], [ikku], and [iktku], produced by a speaker of American English. Utterances were embedded in a carrier sentence and spoken in four conditions: normal rate, normal rate and clear, fast rate, fast rate and clear. The results show several regularities across all conditions: (1) the delay between the protrusion onset and the acoustic onset of [u] varies linearly with the vowel-to-vowel duration (VVD), [i]-end to [u]-onset, with a slope of 0.7; (2) the overlap between the protrusion movement and vowel [i] decreases as VVD increases, with no significant differences between conditions for similar VVD values; (3) the relationship between the ratio of maximum velocity (V[inf max]) to protrusion amplitude (Amp) and the movement duration (T) is constant: V[inf max]/Amp=c/T; (4) the number of velocity peaks increases linearly with VVD. These regularities suggest that the same control strategy underlies the protrusion gesture, whatever the condition. However, for a given rate, clear tokens were produced with longer VVDs and larger protrusion amplitudes. For similar VVD values, fast clear tokens had larger values of V[inf max] than normal tokens. [Work supported by NIDCD and NATO.]