Plosive consonant identification in ambiguous sentences

The hypothesis tested in this paper is that the acoustic–phonetic information at the sentence level (irrespective of the meaning of the sentence) makes it easier to interpret those cues to plosive consonant identity which are contained in vocalic transitions. Listeners were given a two‐way choice in meaningful Dutch sentences which allowed two different intervocalic plosive consonants in the same position, and from which various parts of the VCV segments (plosive burst, vocalic transitions) had been deleted. The results were compared with results from a condition in which the VCV segments were presented without the sentence context. The hypothesis was confirmed for CV transitions, except for /k/, but disconfirmed for nearly all VC transitions. This could mean that in experiments involving the perception of isolated VCV or CVC utterances, whether spoken in isolation or excised from longer utterances, the contribution of the CV transition to plosive consonant perception may be underestimated.