Support for the frame model in Turkish and in English

Abstract

The look‐ahead and the frame model of anticipatory coarticulation make different predictions about lip movement in a word such as /ustu/. The look‐ahead model predicts that rounding from the second vowel will spread into the intervocalic consonants, producing a single movement with a long sustained phase; the frame model predicts two distinct gestures for the rounded vowels whose tails may overlap somewhat, producing two peaks with a “trough” between. In earlier work, it was shown that for nonsense words, such as /kuktluk/, English speakers produce troughs, while Turkish speakers show plateaulike patterns. Does this mean that Turkish and English speakers employ, respectively, look‐ahead and frame strategies for coarticulalion? Alternatively, the frame model holds for Turkish, but Turkish rounding gestures are larger and longer, overlapping for a greater proportion of their tails. An additive model of overlap could then produce a plateaulike pattern. Speakers' movement traces from /kiktlik/ and /kuktluk/ w...