ABSTRACT
The dy<1d consisting of a vowel and a following obstruent is thus a fairly stable production unit, more so than its component parts, and, just like the syllable within the foot, it changes proportionately within the syJlahle under the influence of prosodic factors such as tempo, final lengthening, and intonation (Fitch, 1981). The "Seide" /"Seite" data of Kohler, Kriitzmann, Reetz & Timmermann (1982) provide the linear regressions, summarized in Table 13.8, of phone and dyad on syJlable as well as of phone on dyad durations. The proportion of the dyad in the ,syllable stays the same across the dilrerent intonation patterns and sentence positions used in Kohler, et al. (1982), and it is practica\ly also identical for the sequences of [l)i] and Ienis or fortis stop closures. The global opening-closing-opening gestures from the beginning of the vowel to the end of the consonant are the same in the two cases. It is the internal microstructures of the movements that differ: There is a fast closing movement in "Seite," a slow one in "Seide," captured by the steeper slope of the diphthong/dyad regression in the latter. There seem to he two structural parts within the syllable, viz. the initial consonant (cluster) and the vowel +following consonant, the dyad being in turn broken up into its peak and coda {Hockett, 1955).
