Voice Onset Time Discrimination Research Articles

Phonetic categorization and discrimination of voice onset time under divided attention.

Event durations are perceived to be shorter under divided attention. "Time shrinkage" is thought to be due to rapid attentional switches between tasks, leading to a loss of input samples, and hence, an under-estimation of duration. However, few studies have considered whether this phenomenon applies to durations relevant to time-based phonetic categorization. In this study, participants categorized auditory stimuli varying in voice onset time (VOT) as /ɡ/ or /k/. They did so under focused attention (auditory task alone) or while performing a low-level visual task at the same time (divided attention). Under divided attention, there was increased response imprecision but no bias toward hearing /ɡ/, the shorter-VOT sound. It is concluded that sample loss under divided attention does not apply to the perception of phonetic contrasts within the VOT range.

Bilingual exposure influences infant VOT perception

Linking the discrimination of voice onset time (VOT) in infancy with infant language background, we examine the perceptual changes of two VOT contrasts (/b/-/p/ and /ph/-/p/) by Dutch monolingual and bilingual infants from 8 to 15 months of age. Results showed that language exposure and language dominance had a strong impact on monolingual and bilingual infant VOT perceptual patterns. In addition, perceptual turbulence was found at 8–9 months for bilingual infants, and stabilized perception was presented for all infants from 11 months onwards. We thus report a general input-driven developmental VOT perception in both monolingual and bilingual infants, with perceptual turbulence for bilinguals in the second half of the first year of life.

The development of voicing categories: A quantitative review of over 40 years of infant speech perception research

Most research on infant speech categories has relied on measures of discrimination. Such work often employs categorical perception as a linking hypothesis to enable inferences about categorization on the basis of discrimination measures. However, a large number of studies with adults challenge the utility of categorical perception in describing adult speech perception, and this in turn calls into question how to interpret measures of infant speech discrimination. We propose here a parallel channels model of discrimination (built on Pisoni and Tash Perception & Psychophysics, 15(2), 285-290, 1974), which posits that both a noncategorical or veridical encoding of speech cues and category representations can simultaneously contribute to discrimination. This can thus produce categorical perception effects without positing any warping of the acoustic signal, but it also reframes how we think about infant discrimination and development. We test this model by conducting a quantitative review of 20 studies examining infants' discrimination of voice onset time contrasts. This review suggests that within-category discrimination is surprisingly prevalent even in classic studies and that, averaging across studies, discrimination is related to continuous acoustic distance. It also identifies several methodological factors that may mask our ability to see this. Finally, it suggests that infant discrimination may improve over development, contrary to commonly held notion of perceptual narrowing. These results are discussed in terms of theories of speech development that may require such continuous sensitivity.

Timing is everything: Neural response dynamics during syllable processing and its relation to higher-order cognition in schizophrenia and healthy comparison subjects

Successful linguistic processing requires efficient encoding of successively-occurring auditory input in a time-constrained manner, especially under noisy conditions. In this study we examined the early neural response dynamics to rapidly-presented successive syllables in schizophrenia participants and healthy comparison subjects, and investigated the effects of noise on these responses. We used magnetoencephalography (MEG) to reveal the time-course of stimulus-locked activity over bilateral auditory cortices during discrimination of syllable pairs that differed either in voice onset time (VOT) or place of articulation (POA), in the presence or absence of noise. We also examined the association of these early neural response patterns to higher-order cognitive functions. The M100 response, arising from auditory cortex and its immediate environs, showed less attenuation to the second syllable in patients with schizophrenia than healthy comparison subjects during VOT-based discrimination in noise. M100 response amplitudes were similar between groups for the first syllable during all three discrimination conditions, and for the second syllable during VOT-based discrimination in quiet and POA-based discrimination in noise. Across subjects, the lack of M100 attenuation to the second syllable during VOT-based discrimination in noise was associated with poorer task accuracy, lower education and IQ, and lower scores on measures of Verbal Learning and Memory and Global Cognition. Because the neural response to the first syllable was not significantly different between groups, nor was a schizophrenia-related difference obtained in all discrimination tasks, early linguistic processing dysfunction in schizophrenia does not appear to be due to general sensory input problems. Rather, data suggest that faulty temporal integration occurs during successive syllable processing when the signal-to-noise ratio is low. Further, the neural mechanism by which the second syllable is suppressed during noise-challenged VOT discrimination appears to be important for higher-order cognition and provides a promising target for neuroscience-guided cognitive training approaches to schizophrenia.

Auditory discrimination of voice-onset time and its relationship with reading ability

The perception of voice-onset time (VOT) during dichotic listening provides unique insight regarding auditory discrimination processes and, as such, an opportunity to learn more about individual differences in reading ability. We analysed the responses elicited by four VOT conditions: short–long pairs (SL), where a syllable with a short VOT was presented to the left ear and a syllable with a long VOT was presented to the right ear, as well as long–short (LS), short–short (SS), and long–long (LL) pairs. Stimuli were presented in three attention conditions, where participants were instructed to attend to either the left or right ear, or received no instruction. By around 9.5 years of age children perform similarly to adults in terms of the size and relative magnitude of the right ear advantage (REA) elicited by each of the four VOT conditions. Overall, SL pairs elicited the largest REA and LS pairs elicited a left ear advantage (LEA), reflecting stimulus-driven bottom-up processes. However, children were less able to modulate their responses according to attention condition, reflecting a lack of top-down control. Effective direction of attention to one ear or the other was related to measures of reading accuracy and comprehension, indicating that reading skill is associated with top-down control of bottom-up perceptual processes.

Language background influences the emergence of voice onset time production and perception.

This study investigates the production and perception of voice onset time (VOT) among preschool children who acquired English and French simultaneously compared to monolingual peers of these two languages. Children participated in this study at the age of 18 months and returned at 24 months for a second session. During each session the children took part in a visual habituation procedure to evaluate discrimination of VOT using edited natural /b/+ vowel and /p/ + vowel syllables in a task format developed by Houston et al. (2007) to assess individual performance. In this task, three VOT values (produced by several talkers) were presented to the children: one within adult norms of the language for voiced bilabial consonants; one within the adult norms of the language for voiceless bilabial consonants; and one between the means for the voiced and voiceless consonants. The children also participated in a structured play session, and their spontaneous productions of word initial stops were acoustically analyzed to measure VOT. Preliminary analysis indicates that accuracy in discriminating VOT precedes production of voicing contrasts for all groups, and monolinguals appear to have developed adultlike perception and production of voicing at a younger age than their bilingual peers.

Lexical effects on speech perception in individuals with “autistic” traits

It has been claimed that Autism Spectrum Disorder (ASD) is characterized by a limited ability to process perceptual stimuli in reference to the contextual information of the percept. Such a connection between a nonholistic processing style and behavioral traits associated with ASD is thought to exist also within the neurotypical population albeit in a more subtle way. We examined this hypothesis with respect to auditory speech perception, by testing whether the extent to which phonetic categorization shifts to make the percept a known word (i.e., the ‘Ganong effect’) is weakened as a function of autistic traits in neurotypicals. Fifty-five university students were given the Autism-Spectrum Quotient (AQ) and a segment identification test using two word-to-nonword Voice Onset Time (VOT) continua (kiss-giss and gift-kift). A significant negative correlation was found between the total AQ score and the identification shift that occurred between the continua. The AQ score did not correlate with scores on separately administered VOT discrimination, auditory lexical decision, or verbal IQ, thus ruling out enhanced auditory sensitivity, slower lexical access or verbal intelligence as explanations of the AQ-related shift in phonetic categorization.

Objective phonological and subjective perceptual characteristics of syllables modulate spatiotemporal patterns of superior temporal gyrus activity

Natural consonant–vowel syllables are reliably classified by most listeners as voiced or voiceless. However, our previous research [Liederman, J., Frye, R., Fisher, J.M., Greenwood, K., Alexander, R., 2005. A temporally dynamic context effect that disrupts voice onset time discrimination of rapidly successive stimuli. Psychon Bull Rev. 12, 380–386] suggests that among synthetic stimuli varying systematically in voice onset time (VOT), syllables that are classified reliably as voiceless are nonetheless perceived differently within and between listeners. This perceptual ambiguity was measured by variation in the accuracy of matching two identical stimuli presented in rapid succession. In the current experiment, we used magnetoencephalography (MEG) to examine the differential contribution of objective (i.e., VOT) and subjective (i.e., perceptual ambiguity) acoustic features on speech processing. Distributed source models estimated cortical activation within two regions of interest in the superior temporal gyrus (STG) and one in the inferior frontal gyrus. These regions were differentially modulated by VOT and perceptual ambiguity. Ambiguity strongly influenced lateralization of activation; however, the influence on lateralization was different in the anterior and middle/posterior portions of the STG. The influence of ambiguity on the relative amplitude of activity in the right and left anterior STG activity depended on VOT, whereas that of middle/posterior portions of the STG did not. These data support the idea that early cortical responses are bilaterally distributed whereas late processes are lateralized to the dominant hemisphere and support a “how/what” dual-stream auditory model. This study helps to clarify the role of the anterior STG, especially in the right hemisphere, in syllable perception. Moreover, our results demonstrate that both objective phonological and subjective perceptual characteristics of syllables independently modulate spatiotemporal patterns of cortical activation.

An event-related fMRI investigation of voice-onset time discrimination

The discrimination of voice-onset time, an acoustic–phonetic cue to voicing in stop consonants, was investigated to explore the neural systems underlying the perception of a rapid temporal speech parameter. Pairs of synthetic stimuli taken from a [da] to [ta] continuum varying in voice-onset time (VOT) were presented for discrimination judgments. Participants exhibited categorical perception, discriminating 15-ms and 30-ms between-category comparisons and failing to discriminate 15-ms within-category comparisons. Contrastive analysis with a tone discrimination task demonstrated left superior temporal gyrus activation in all three VOT conditions with recruitment of additional regions, particularly the right inferior frontal gyrus and middle frontal gyrus for the 15-ms between-category stimuli. Hemispheric differences using anatomically defined regions of interest showed two distinct patterns with anterior regions showing more activation in the right hemisphere relative to the left hemisphere and temporal regions demonstrating greater activation in the left hemisphere relative to the right hemisphere. Activation in the temporal regions appears to reflect initial acoustic–perceptual analysis of VOT. Greater activation in the right hemisphere anterior regions may reflect increased processing demands, suggesting involvement of the right hemisphere when the acoustic distance between the stimuli are reduced and when the discrimination judgment becomes more difficult.

Linear Coding of Voice Onset Time

Voice onset time (VOT) provides an important auditory cue for recognizing spoken consonant-vowel syllables. Although changes in the neuromagnetic response to consonant-vowel syllables with different VOT have been examined, such experiments have only manipulated VOT with respect to voicing. We utilized the characteristics of a previously developed asymmetric VOT continuum [Liederman, J., Frye, R. E., McGraw Fisher, J., Greenwood, K., & Alexander, R. A temporally dynamic contextual effect that disrupts voice onset time discrimination of rapidly successive stimuli. Psychonomic Bulletin and Review, 12, 380-386, 2005] to determine if changes in the prominent M100 neuromagnetic response were linearly modulated by VOT. Eight right-handed, English-speaking, normally developing participants performed a VOT discrimination task during a whole-head neuromagnetic recording. The M100 was identified in the gradiometers overlying the right and left temporal cortices and single dipoles were fit to each M100 waveform. A repeated measures analysis of variance with post hoc contrast test for linear trend was used to determine whether characteristics of the M100 were linearly modulated by VOT. The morphology of the M100 gradiometer waveform and the peak latency of the dipole waveform were linearly modulated by VOT. This modulation was much greater in the left, as compared to the right, hemisphere. The M100 dipole moved in a linear fashion as VOT increased in both hemispheres, but along different axes in each hemisphere. This study suggests that VOT may linearly modulate characteristics of the M100, predominately in the left hemisphere, and suggests that the VOT of consonant-vowel syllables, instead of, or in addition to, voicing, should be examined in future experiments.

Cognitive influences in infant speech perception

Speech perception during the first year reflects increasing attunement to native language phonetic features, but the mechanisms underlying this development are not well understood. Reductions in non-native phonetic discrimination have been linked to improvement in native phonetic discrimination and later vocabulary growth (Kuhl et al., 2005), and performance on nonlinguistic tasks (Lalonde and Werker, 1995). The present study examined links between native and non-native voice onset time discrimination, receptive vocabulary (MacArthur-Bates CDI), and cognitive control abilities at 11 months. Infants (n=18) completed a double-target conditioned head turn task and two nonlinguistic tasks requiring attentional control and resistance to irrelevant cues (means-end and detour-reaching object retrieval). Infants with CDI scores above the median showed higher native discrimination scores, t(16)=2.15 0.05, but no group differences for the nonnative contrast. Infants with scores above the median on either cognitive task showed worse discrimination of the nonnative contrast (means-end, t(15)=2.27p0.04; detour-reaching, t(15)=3.49, p.01), but no group differences for the native contrast. These results suggest that cognitive control plays a role in infants’ ability to ignore acoustic cues that are irrelevant to their native languages phonemic categories. [Work supported by NICHD (Grant HD37954) and a UW NSF Science of Learning Center Grant (LIFE).]

A temporally dynamic context effect that disrupts voice onset time discrimination of rapidly successive stimuli

Across three experiments, voice onset time discrimination along a/ba/-/pa/ continuum was found to be influenced by the order of presentation of rapidly successive stimuli. Specifically, discrimination was disrupted when a relatively unambiguous /pa/ syllable was presented before, rather than after, a more ambiguous /pa/ or/ba/ syllable. In Experiments 1 and 2, for between-category discrimination, this order effect was significant at interstimulus intervals (ISIs) below 250 msec, but not at 250 or 1,000 msec. In Experiments 2 and 3, the order effect was also significant for within-category discrimination at ISIs below 250 msec. In addition, in Experiment 3 this order effect was not diminished by provision of performance feedback across eight testing sessions. These findings reveal a particular vulnerability of phonological processing in response to rapidly successive stimuli and may have implications for mathematical and neural models of speech processing of normal and impaired populations.

Discrimination of multichannel-compressed stop consonants

The effect of full-range multichannel compression (MCC) on stop consonant discrimination was studied. For normal-hearing subjects MCC was fit to three hypothetical flat losses, with thresholds ranging from 70 to 90 dB SPL, and one hypothetical sloped loss, with thresholds normal at 500 Hz and 90 dB SPL at 4 kHz. Hearing-impaired subjects were tested with the flat MCC as well as MCC fit to the subject’s hearing loss. The MCC had compression ratios from 1.75 to 7.00, numbers of channels from 2 to 31, and Robinson–Huntington compression [C. E. Robinson and D. A. Huntington, J. Acoust. Soc. Am. 54, 314 (1973)] with 10-ms time windows. Unprocessed stimuli and frequency-equalized linear amplification were control conditions. Deleterious effects of MCC on discrimination of voice onset time (VOT) were minimal. Normal-hearing subjects had difficulty with stop-consonant place discrimination in the most severe compression conditions. Hearing-impaired subjects needed either fitted MCC or frequency-equalized linear amplification to perform consistent place discriminations on these synthetic vowel–consonant–vowel (VCV) syllables. Complete results and their implications for the application of MCC in hearing aids will be discussed. [Work supported by Department of Veterans Affairs.]

Longitudinal changes in syllable discrimination among normal and handicapped school children

Cross‐sectional age‐related differences in discrimination of voice onset time (VOT) [L. L. Elliott, J. Acoust. Soc. Am. 80, 1250–1255 (1986)] and the place‐of‐articulation (POA) feature [L. L. Elliott et al., J. Acoust. Soc. Am. 70, 669–677 (1981)] when just‐noticeable differences among five‐formant synthesized CV syllables are measured using adaptive test procedures were previously described. Results are reported here for children tested longitudinally for 3 consecutive years. Discrimination of temporal information (VOT) showed larger changes over time than discrimination of frequency‐based information (POA). Also, larger performance differences occurred between normally progressing children and those with learning disabilities for VOT than for POA. Children with learning disabilities generally exhibited poorer auditory discrimination than normals; however, they demonstrated as much longitudinal change in performances as did the normally progressing children (i.e., there were no significant interactions between year of test and normal versus handicapped status). [Supported, in part, by NINCDS and NIH.]

The role of spectral cues in discrimination of voice onset time differences.

The influence of spectral cues on discrimination peaks in the region of the phonetic voicing boundary was examined. The discriminability of voice onset time (VOT) differences of the same temporal magnitude was assessed using stimuli from labial and velar consonant-vowel VOT continua that differed in the timing of spectral changes associated with the first formant (F1) transition, and in the location of the phonetic boundary. Subjects were initially given labeling tests and fixed-standard AX and all-step discrimination tests on both series. Half the subjects then received all-step discrimination training on one series and half received training on the other series. Finally, all subjects were again given the labeling and discrimination tests on both series. Just noticeable differences (jnds) in VOT were estimated from the all-step functions before and after training. Initial jnds showed that VOT discrimination was most accurate around the voicing boundary on the two continua, where differences in F1 onset frequency accompany variations in VOT. jnds on both series decreased significantly after training, although these regions of greater sensitivity remained. No evidence was seen of increased sensitivity around +/- 20-ms VOT, as expected if auditory processing constraints were influencing temporal order judgments. Comparisons of post-training jnds within and across series indicated that spectral components of VOT, primarily F1 onset frequency differences, exert a substantial influence on discrimination, and, along with other spectral cues provided by source differences at stimulus onset, can account for the discontinuities in discrimination often reported in research with VOT continua. Large phonetic effects also were seen in the initial performance of all subjects: jnds decreased consistently as standards drew nearer the voicing boundary. However, these effects were absent in the final jnds for most subjects. Implications of these findings for the understanding of basic auditory and attentional processes in speech perception are discussed.

Discrimination of Voice Onset Time by Human Infants: New Findings and Implications for the Effects of Early Experience

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 by Human Infants: New Findings and Implications for the Effects of Early Experience

Discrimination of Voice Onset Time by Human Infants: New Findings and Implications for the Effects of Early Experience

Discrimination of voice onset time in infancy.

In an effort to determine whether infants can discriminate speech sounds on the basis of a single acoustic cue, timing onset of periodic voicing, two experiments were conducted employing synthetic speech sounds. Naturally produced syllable pairs were also used for comparison. In the first experiment infants evidenced discrimination of a naturally produced /ba/ versus /pa/ pair and a naturally /du/ versus /tu/ pair. In addition, infants discriminated a synthetic /ba/ versus /pa/ contrast that was cued by several acoustic differences in addition to timing onset of periodic voicing but failed to evidence an ability to discriminate a synthetic /du/ versus /tu/ contrast that contained flat first formants and differed only in timing onset of periodic voicing. A second experiment was conducted in which infants once again evidenced discrimination of naturally produced /du/ versus /tu/ stimuli but not of synthetic /du/ versus /tu/ stimuli containing slight first-formant transitions. These results suggest that timing onset of periodic voicing alone may not be a sufficient cue for infant discrimination of English voicing contrasts.

Identification and discrimination of VOT by listeners with moderate, severe and profound sensorineurai hearing loss

To determine whether sensorineural hearing loss affects listeners' ability to detect the voice onset time (VOT) of stop consonants, listeners with moderate, severe, and profound hearing loss were presented stimuli from along a VOT continuum in both identification and discrimination tasks. The location of the phoneme boundary did not differ among the normal listeners, the listeners with moderate loss, and most listeners with severe loss. Some listeners with severe loss, however, evidenced longer boundaries while others could not identify the signals at all. A similar outcome was obtained for the listeners with profound loss. For the most part discrimination data mirrored identification data. However, in several instances listeners who were unable to identify voiced and voiceless stops were able to discriminate between them in a normal manner. Since identification and discrimination tasks assess different aspects of information processing, an evaluation of sensory capabilities in the hearing impaired should include both tasks.

A temporal order effect in voice onset time discrimination.

In two experiments ten adults were required to discriminate between synthetic speech stimuli varying in voice onset time. The stimuli included labial and velar stops. The data, analysed, using ANOVA and post hoc comparison techniques, indicated that while adult phoneme perception was categorical, it was also asymmetrical. Discrimination of VOT changes from voiceless to voiced stimuli was significantly better than detection of voiced to voiceless changes. The data are interpreted as supporting a theory that VOT perception is mediated by a dual feature detector mechanism.