É possível integrar o discreto e o contínuo em um modelo de produção do ritmo da fala?

Abstract

A model of speech rhythm production is proposed that is able to explain the differences found between the durational patterns of stress- and syllable-timed languages from strict considerations of timing. This is possible from a two-parameter characterization of Brazilian Portuguese V-to-V duration patterns within stress groups. The parameters are speech rate (measured in terms of V-to-V units per unit of time) and the coupling strength between two oscillators (syllabic and stress group oscillators), which act as a cognitive pacemaker whose pace is the vowel onset succession entrained by the beats of the stress group oscillator. Differences in amount of coupling strength seem to explain differences in rhythmic typology. A link with lexical processing is also suggested here.