“Co-perception” of speech segments

Abstract

A definition of “co-perception” is proposed, in analogy to the definition of coarticulation: Co-perception is said to exist whenever the perception of a (phonetically defined) speech segment is influenced by a preceding or following segment. The measure of co-perception is the reaction time (in “same-different” judgments, speeded classification, or monitoring) to a certain well-defined segment, while other segments vary systematically. Three hypotheses to explain co-perception are discussed: (1) A temporal integration period (or limited buffer memory) in speech perception; (2) genuine functional analogies to coarticulation; and (3) “phonetic coherence” factors, such as the primacy of the CV syllable. These hypotheses may be investigated by appropriate variation of the phonetic composition and the temporal parameters of the speech stimuli used. Some experimental paradigms are discussed, and preliminary data are reported that demonstrate co-perception of consonants and vowels in CV, VCV, and (perhaps) VC syllables.