Interaural Switching Effects on the Recognition of Consonants

Abstract

Experiments in which subjects repeated back (that is, “shadowed”) speech that was systematically alternated between their two ears have led to the hypothesis that the place—with respect to certain linguistic boundaries—at which the alternation occurs may affect perception [G. W. Hughes et al., J. Acoust. Soc. Amer. 40, 1283 (A) (1966)]. This hypothesis was investigated in an experiment in which speech samples of the form /hə'CVt/—where C was one of 14 different consonants and V was one of three different vowels—began in one ear and were switched to the other ear at some desginated place in the waveform. Each stimulus item was mixed with noise and presented to 16 subjects who identified the C and the V in a forced-choice test format. The results showed that the place in the waveform where the switching occurred had very little effect on consonant recognition and suggest that any decrement in the shadowing scores when speech is alternated between the ears is better described in terms of syllables than in terms of phonemes. Possible ear-preference effects were investigated and will be discussed. [This research was supported in part by the Air Force Cambridge Research Laboratories under contract.]