For many years there has been a consensus that early linguistic experience exerts a profound and often permanent effect on the perceptual abilities underlying the identification and discrimination of stop consonants. It has also been concluded that selective modification of the perception of stop consonants cannot be accomplished easily and quickly in the laboratory with simple discrimination training techniques. In the present article we report the results of three experiments that examined the perception of a three-way voicing contrast by naive monolingual speakers of English. Laboratory training procedures were implemented with a small computer in a real-time environment to examine the perception of voiced, voiceless unaspirated, and voiceless aspirated stops differing in voice onset time. Three perceptual categories were present for most subjects after only a few minutes of exposure to the novel contrast. Subsequent perceptual tests revealed reliable and consistent labeling and categorical-like discrimination functions for all three voicing categories, even though one of the contrasts is not phonologically distinctive in English. The present results demonstrate that the perceptual mechanisms used by adults in categorizing stop consonants can be modified easily with simple laboratory techniques in a short period of time.Over the last 15 years numerous studies employing synthetically produced speech stimuli have investigated voice-onset-time (VOT) perception in human adults, human infants, chinchillas, and monkeys. These developmental and cross-species comparisons have been undertaken to better understand the potential interactions between genetic predispositions and experiential factors in perceptual categorization of speech signals. The results of these diverse studies have shown the combined influence of two factors operating in speech perception. First, linguistic experience has been shown to have a substantial effect on speech perception, particularly in human adults exposed to different language-learning environments. Subjects identify and discriminate speech sounds with reference to the linguistic categories of their language. Second, basic sensory and psychophysical constraints on auditory system function seem to affect perception of speech and nonspeech control signals in similar ways. For example, the perception of voicing in stop consonants apparently requires the analysis of a temporal relation between laryngeal and supralaryngeal events. Basic constraints on auditory perception may play an important role in defining the inventory of acoustic correlates for distinctive features used in speech (Stevens, 1972). These considerations suggest that experiential and genetic factors are both implicated in the process by which speech signals are perceived by human listeners.The results of the earliest cross-language experiments on the perception and production of VOT by Abramson (1964, 1967) quantitatively confirmed that the linguistic environment exerts a profound influence on the ability to prod...

Discrimination of voice onset time (VOT) by 6--12-month-old infants was examined in 2 experiments. An operant head-turning technique assessed discrimination along a synthetic VOT continuum ranging from -70 msec to +70 msec. Infants from an English-speaking environment provided reliable within-subject evidence for discrimination of VOT contrasts located at both the plus and minus regions of the VOT continuum. These results provide strong evidence that infants from an English-speaking environment are capable of discriminating VOT contrasts that are not phonemic in English. Threshold delta VOT values indicated that the infants were more sensitive to VOT differences in the plus region of the VOT continuum than in the minus region. Threshold delta VOT values from English-speaking adults indicated greater sensitivity at every location along the VOT continuum. In addition, the adults showed heightened sensitivity to VOT differences near the voiced-voiceless boundary in the plus region of the VOT continuum, a finding that was not evident in the infants' data.

Discrimination of voice onset time (VOT) by 6-12-month-old infants was examined in 2 experiments. An operant head-turning technique assessed discrimination along a synthetic VOT continuum ranging from −70 msec to +70 msec. Infants from an English-speaking environment provided reliable within-subject evidence for discrimination of VOT contrasts located at both the plus and minus regions of the VOT continuum. These results provide strong evidence that infants from an English-speaking environment are capable of discriminating VOT contrasts that are not phonemic in English. Threshold ΔVOT values indicated that the infants were more sensitive to VOT differences in the plus region of the VOT continuum than in the minus region. Threshold ΔVOT values from English-speaking adults indicated greater sensitivity at every location along the VOT continuum. In addition, the adults showed heightened sensitivity to VOT differences near the voiced-voiceless boundary in the plus region of the VOT continuum, a finding that was not evident in the infants' data.Eimas, Siqueland, Jusczyk, and Vigorito (1971) demonstrated that young infants can discriminate small differences in voice onset time (VOT), a parameter that is sufficient to differentiate the phonetic categories of voiced (b, d, g) from voiceless (p, t, k) stop consonants. 1 They could discriminate VOT differences that straddled the voiced-voiceless boundary for English-speaking adults (approximately +25 msec), whereas comparable differences in VOT selected from within the phonemic categories of English were not discriminated. The finding that infants from an English-speaking environment do not discriminate VOT contrasts drawn from within English-adult phonemic categories is surprising because some of these same contrasts are discriminated by adults from other language communities (e.g., Thai and Spanish; see Abramson & Lisker [1970] and Williams [1977]). This failure to discriminate nonphonemic contrasts implies that either (1) phonemic categories are determined primarily by genetic factors (a view favored initially by Eimas) or (2) some (or all) phonemic categories are determined largely by early linguistic experience in a particular language environment. Address reprint requests to R. N. Aslin or D. B. Pisoni, Department of Psychology, Indiana University, Bloomington, Indiana 47405. Beth Hennessy is now at the Institute of Child Development, University of Minnesota. Reference Notes 1 Voice onset time (VOT) refers to the delay between release from stop closure and the onset of laryngeal pulsing (voicing). In synthetic consonant-vowel (CV) syllables, VOT is realized acoustically by varying the delay between the onset of energy in the first formant and higher formants and the presence of aspiration noise during this interval. In phonetic terms there are three categories of voicing in syllable-initial stop consonants: voiced, voiceless unaspirated, and voiceless aspirated. In English the phonetic distinction between voiced and voiceless unaspirated stops is not used p...

Discrimination of cookie-cutter patterns by active and passive touch was compared in humans under conditions which attempted to equate the con tactual movement of the stimulus over the skin surface in the two modes. We find that the lack of self-initiated movement made no difference in accurate identification of these patterns, contrary to the postulation by Gibson (1962). Further, the results agree with certain neurophysiological findings which suggest that the dorsal columns are critical for the tactile discrimination of spatiotemporal features of stimuli, rather than for mediation of active vs. passive touch per se.