Auditory and linguistic processes in speech perception: Inferences from six fusions in dichotic listening.

Abstract

A number of phenomena in speech perception have been called fusion, but little effort has been made to compare these phenomena in a systematic fashion. The present paper examines six of them. All can be exemplified using the syllable /da/ as in dot, and all occur during dichotic listening. In each type of fusion, the robustness of the fused percept is observed against variation in three parameters: the relative onset time of the two opposite-ear stimuli, their relative intensity, and their relative fundamental frequency. Patterns of results are used to confirm the arrangement of the six fusions in a hierarchy, and supporting data are summoned in an analysis of the mechanisms that underlie each with reference to speech. The six fusions are sound localization, psychoacoustic fusion, spectral fusion, spectral/temporal fusion, phonetic feature fusion, and phonological fusion. They occur at three, perhaps four, different levels of perceptual analysis. The first two levels are characterized by perceptual integration, the other (s) by perceptual disruption and recombination.

Auditory Illusions as Evidence for a Role of the Syllable in Adult Developmental Dyslexics

“…The paradigm has been transposed to the study of written language perception since the work by Prinzmetal (1981) and more recently with the study of letter migrations (Kolinsky, 1992;Mozer, 1983;McClelland & Mozer, 1986). Antecedents of the paradigm in the domain of spoken language processing include the study of fusions by Cutting (1976). Along these lines it was observed that dichotic presentation of a pair of spoken words may lead to the mistaken perception of a target (Kolinsky, Morais, & Cluytens, 1995).…”

Phonetic awareness and reading acquisition

“…One of the tasks could in principle provide evidence for:a stage of phonetic feature extraction. It has been shown that, when syllables beginning with a plosive are presented dichotically, most of the recognition errors consist apparently in blending one feature of one stimulus with another feature of the other stimulus, as in , Studdert-Kennedy & Shankweiler, 1970;Cutting, 1976). We inspected whether Portuguese illiterates and literates make feature blendings significantly above chance and in the same proportion or not, in a similar Situation (Morais, Castro, Scliar-Cabral, Kolinsky, & Content, in press).…”

Spectral discontinuities and the vowel length effect

“…Separation in space has also resulted in less integration in several studies employing speech (Cherry & Taylor, 1954;Huggins, 1964;Schubert & Schultz, 1962). However, integration of lateralized speech components, which are presumably perceived as arriving from separate spatial locations, has been demonstrated (Broadbent, 1955;Broadbent & Ladefoged, 1957;Cutting, 1976;Rand, 1974).…”

“…Many ofthe same attributes that have been demonstrated to be important for segregation of simple sounds also appear to playa role in segregation and grouping of speech sounds and are of especial utility in segregating the speech ofcontemporaneous speakers. Again, asynchrony ofonsets and inharmonicity have been found to promote segregation of the components (Cutting, 1976;Darwin, 1981;Darwin & Gardner, 1986;Scheffers, 1983). Spectral distance and trajectories of spectral change greatly influence speech source integration and segregation (Dorman, Cutting, & Raphael, 1975).…”

“…For example, Cutting (1976;Cutting & Day, 1975) has recommended that in order to account for results from dichotic listening experiments with speech signals, one may wish to incorporate putative processes such as phonetic feature fusion and phonological fusion in addition to sound localization, psychoacoustic fusion, spectral fusion, and spectral/temporal fusion. More recently, Remez and his colleagues (e.g., Remez, Rubin, Berns, Pardo, & Lang, 1994) have argued that phonetic perceptual organization also requires "specific sensitivity to the acoustic modulations characteristic of speech signals" (p. 129) in order to account for the performance of those subjects who succeed in rendering a phonetic interpretation of sine-wave replicas of spoken sentences.…”