Haskins Publications

Number	480
Year	1984
Drawer	1984
Entry Date	11/19/1999
Authors	Fowler, C. A.
Contact
Publication	Perception & Psychophysics, 36, 359-368.
url	http://www.haskins.yale.edu/Reprints/HL0480.pdf
Abstract	Investigated how 36 college students segmented an acoustic speech signal into phonetic segments, using identification, choice RT, and 4IAX discrimination tests. Two manners of segmentation were contrasted: In one, Subjects segmented the signal into temporally discrete, context-sensitive segments; in the other, which may be consistent with the talker's production of the segments, Subjects partitioned the signal into separate but overlapping segments freed of their contextual influences. Two complementary predictions were tested: (1) Subjects would use anticipatory coarticulatory information for a segment as information for the forthcoming segment, (2) Subjects would not hear anticipatory coarticulatory information as part of the phonetic segment with which it co-occurs in time. The 1st hypothesis was supported by findings on the choice RT procedure; the 2nd was supported by findings on the 4IAX discrimination test. It is suggested that talkers produce phonetic segments in overlapping time frames that preserve the coherence of the temporally extended parts of an individual phonetic segment and the separation of distinct segments.
Notes