Abstract

The term “phoneme boundary effect” refers to the observation that synthetic speech stimuli which lie across a boundary between phoneme categories are discriminated more accurately than comparable stimulus differences which lie within a phoneme category. At the last meeting of the Society [C. C. Wood, J. Acoust. Soc. Am. 58, S58(A) (1975)], I described experiments which investigated the voiced‐voiceless boundary effect in discrimination of differences in voice onset time (VOT). The present experiments investigated discrimination of stop consonants differing in place of articulation as cued by differences in the direction and extent of second and third formant transitions. The principal results were directly comparable to those of the VOT experiments: (a) there were both clear increases in discriminability and marked changes in response bias near the [b]‐[d] and [d]‐[g] boundaries and (b) increases in discriminability at the phoneme boundaries were also obtained when the formant transitions were removed from syllable context and presented in isolation. Unlike the VOT results, however, the increases in discriminability for isolated formant transitions were smaller than for the full‐syllable stimuli and were also influenced by presentation order of the discrimination tasks. These latter results suggest that subjects' biases and expectancies regarding the synthetic stimuli may influence their discrimination of “isolated” phonetic cues. This possibility is currently being investigated.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call