Implications For Speech Perception Research Articles

Siri, you've changed! Acoustic properties and racialized judgments of voice assistants

As speech technology is increasingly integrated into modern American society, voice assistants are a more significant part of our everyday lives. According to Apple, Siri fulfills 25 billion requests each month. As part of a software update in April 2021, users in the U.S. were presented with a choice of 4 Siris. While in beta testing, users on Twitter began to comment that they felt that some of the voices had racial identities, noting in particular that Voice 2 and Voice 3 “sounded black.” This study tests whether listeners indeed hear the different Siri voices as sounding like speakers from different groups, as well as examines voice quality features that may trigger these judgments. In order to test evaluations of the four voices, 485 American English listeners heard each Siri voice reading the Rainbow Passage, via online survey conducted on Qualtrics. Following each clip, listeners responded to questions about the speaker's demographic characteristics and personal traits. An LMER model of normalized ratings assessed the interaction of voice and race judgment revealed that indeed, Voice 2 and Voice 3 were significantly more likely to be rated as belonging to a Black speaker than Voices 1 and 4 (p < 0.001). Per-trait logistic regression models and chi-square tests examining ratings revealed Voice 3, the male voice rated as Black, was judged less competent (X2 = 108.99, x < 0.001), less professional (X2 = 90.97, p < 0.001), and funniest (X2 = 123.39, x < 0.001). Following analysis of listener judgments of voices, I conducted post-hoc analysis comparing voice quality (VQ) features to examine which may trigger the listener judgments of race. Using PraatSauce, I employed scripts to extract VQ measures previously hypothesized to pattern differently in African American English vs. Mainstream American English. VQ measures that significantly affected listener ratings of the voices are mean F0 and H1–A3c, which correlate with perceptions of pitch and breathiness. These results reveal listeners attribute human-like demographic and personal characteristics to synthesized voices. A more comprehensive understanding of social judgments of digitized voices may help us to understand how listeners evaluate human voices, with implications for speech perception and discrimination as well as recognition and synthesis.

Speaking Rate, Oro-Laryngeal Timing, and Place of Articulation Effects on Burst Amplitude: Evidence From English and Tamil.

The relationship between speaking rate and burst amplitude was investigated in plosives with differing oro-laryngeal timing: long-lag voice-onset time (VOT) (North American English) and short-lag VOT (Indian Tamil). Burst amplitude (reflecting both intraoral pressure and flow geometry of the oral channel) was hypothesized to decrease in pre-vocalic plosive syllables with the increase in speaking rate, which imposes temporal constraints on both intraoral pressure buildup behind the oral occlusion and respiratory air flow. The results showed that decreased vowel duration (which is associated with increased speaking rate) led to decreased burst amplitude in both short- and long-lag plosives. Aggregate models of bilabial and velar plosives (found in both languages) suggested lower burst amplitudes in short-lag stops. Place-of-articulation effects in both languages were consistent with models of stop consonant acoustics, and place interactions with vowel duration were most apparent with long-lag English stops. The results are discussed in terms of speaking rate and language-internal forces, contributing to burst amplitude variation and their implications for speech perception and potential to affect lenition phenomena.

Inferior Frontal Cortex Contributions to the Recognition of Spoken Words and Their Constituent Speech Sounds.

Speech perception and comprehension are often challenged by the need to recognize speech sounds that are degraded or ambiguous. Here, we explore the cognitive and neural mechanisms involved in resolving ambiguity in the identity of speech sounds using syllables that contain ambiguous phonetic segments (e.g., intermediate sounds between /b/ and /g/ as in "blade" and "glade"). We used an audio-morphing procedure to create a large set of natural sounding minimal pairs that contain phonetically ambiguous onset or offset consonants (differing in place, manner, or voicing). These ambiguous segments occurred in different lexical contexts (i.e., in words or pseudowords, such as blade-glade or blem-glem) and in different phonological environments (i.e., with neighboring syllables that differed in lexical status, such as blouse-glouse). These stimuli allowed us to explore the impact of phonetic ambiguity on the speed and accuracy of lexical decision responses (Experiment 1), semantic categorization responses (Experiment 2), and the magnitude of BOLD fMRI responses during attentive comprehension (Experiment 3). For both behavioral and neural measures, observed effects of phonetic ambiguity were influenced by lexical context leading to slower responses and increased activity in the left inferior frontal gyrus for high-ambiguity syllables that distinguish pairs of words, but not for equivalent pseudowords. These findings suggest lexical involvement in the resolution of phonetic ambiguity. Implications for speech perception and the role of inferior frontal regions are discussed.

Revisiting Neil Armstrongs Moon-Landing Quote: Implications for Speech Perception, Function Word Reduction, and Acoustic Ambiguity.

Neil Armstrong insisted that his quote upon landing on the moon was misheard, and that he had said one small step for a man, instead of one small step for man. What he said is unclear in part because function words like a can be reduced and spectrally indistinguishable from the preceding context. Therefore, their presence can be ambiguous, and they may disappear perceptually depending on the rate of surrounding speech. Two experiments are presented examining production and perception of reduced tokens of for and for a in spontaneous speech. Experiment 1 investigates the distributions of several acoustic features of for and for a. The results suggest that the distributions of for and for a overlap substantially, both in terms of temporal and spectral characteristics. Experiment 2 examines perception of these same tokens when the context speaking rate differs. The perceptibility of the function word a varies as a function of this context speaking rate. These results demonstrate that substantial ambiguity exists in the original quote from Armstrong, and that this ambiguity may be understood through context speaking rate.

Differential effects of noise trauma and diminished endocochlear potential on neural temporal coding of complex sounds: Implications for speech perception

Communication problems due to cochlear hearing loss are pervasive in today’s society, even with amplification from digital hearing aids. Threshold elevation may arise from noise-induced trauma to hair cells or, as an alternative mechanism, through age-related depletion of the endocochlear potential (EP; the power supply driving hair-cell function). While both noise trauma and diminished EP may frequently underlie hearing loss in older people, it is unclear whether these pathologies have different effects on neural coding of complex sounds, and ultimately, on speech perception. Here, we describe several recent neurophysiological studies of this question based on Wiener-kernel analyses of chinchilla auditory-nerve fiber responses to Gaussian noise. We find that noise trauma causes pronounced distortions in tonotopic coding of temporal fine structure and slower envelope cues, occurring with as little as 20 dB threshold elevation. Similar changes in neural coding occur with diminished EP, but only when thresh...

Neural correlates of individual differences in processing of rising tones in Cantonese: Implications for speech perception and production

Event Abstract Back to Event Neural correlates of individual differences in processing of rising tones in Cantonese: Implications for speech perception and production Sampo Law1* and Jinghua Ou1 1 University of Hong Kong, Division of Speech and Hearing Sciences, Hong Kong, SAR China Introduction Two aspects of the F0 - the F0 level (high, middle, low) and the F0 contour (static, rising, falling) – are generally considered the perceptual correlates of lexical tones in tone languages, including Mandarin Chinese (Gandour, 1983), Cantonese (Khouw & Ciocca, 2007), and Thai (Gandour, Potisuk, & Dechongkit, 1994). Besides the dominant role of spectral information, much attention has recently been paid to the importance of temporal information in parsing the acoustic signal into relevant segments for decoding during auditory/speech processing (Luo & Poeppel, 2012). Acoustic cues from the temporal waveform envelope have also been shown to successfully cue tone perception in Mandarin Chinese (Whalen & Xu, 1992) as well as Cantonese (Zhou, 2012). Of the various cues to amplitude envelope, rise time, defined as the time taken for a sound to reach its maximum amplitude (Rosen, 1992), is proposed to be an important perceptual cue for the representation of amplitude envelope. The amplitude rise time has been found to be important in facilitating prosodic and syllable segmentation processes in children (Carpenter & Shahi, 2013; Leong, Hämäläinen, Soltesz, & Goswami, 2011), which are arguably critical for the formation of well-specified phonological representations (Goswami, 2011). Hence, one may question whether the rise time of sound amplitude envelope may likewise play a role in processing lexical tones. In other words, to process tones efficiently may entail the encoding of both spectral and temporal cues present in the speech signal to derive tone representations. The present study is the first examination of neural processes underlying the discrimination of the high rising and low rising tones T2/T5 in Hong Kong Cantonese (HKC) from two groups of typically-developed native speakers of HKC with comparable language and musical backgrounds. The participant groups represented, respectively, the pattern of good perception and good production of all Cantonese tones [+Per+Pro], and that of good perception of all tones but poor production of specifically the T2/T5 distinction [+Per-Pro]. Electrophysiological responses to the contrasts of pitch and amplitude envelope between T2 and T5 were measured to allow us to assess the timing and strength of neural activities associated with the auditory stimuli unfolding over time. Any difference in neural response between the two groups would shed light on how the acoustic cues of pitch and amplitude envelope are differentially represented in their auditory memory, and enable us to consider the relationship between perception and production. Method A total of 138 native speakers of Cantonese, all born and raised in Hong Kong, were recruited. No speaker reported a history of hearing abnormalities. According to the Edinburgh Handedness Inventory (Oldfield, 1971), they were all right-handers. They first participated in a tone perception and a tone production task. On the basis of their performance on these tasks, 41 participants were invited back to carry out a passive oddball task. The participants were classified into two groups, i.e. [+Per+Pro] (N = 20, female = 8) and [+Per-Pro] (N = 21, female = 13). The EEG experiment employed the passive oddball paradigm and was conducted in a sound-attenuated electrically shielded booth. Three syllables /fu1/, /fu2/ and /fu5/ were used. The experiment consisted of four oddball conditions of different Standard/Deviant pairs, including T2/T5 and T5/T2 as two experimental conditions, and two control conditions by pairing T2 and T5 with T1 as the common standard, i.e. T1/T2 and T1/T5. For the control conditions, the divergence point was at the vowel onset, where pitch height of the two stimuli begin to deviate. For the experimental conditions, as T2 resembled T5 in the early part of the pitch contour, the two began to diverge at 250 post stimulus onset. Additionally, in the period of 100 to 250 ms where the pitch contours of T2 and T5 fully overlapped, the amplitude rise time, computed as the duration between the vowel onset and amplitude peak during the overlapping pitch period (Tarr, 2013), differed between them. The rise time was 120 ms for T2 and 70 ms for T5. In each condition, the standard stimuli were presented in 85% of the trials, and the deviant occurred on 15% (or 80) of the trials in a quasi-random sequence. The sequence of blocks was rotated across participants. The pre-processed EEG data were analyzed in two ways. Statistical differences between the true and dummy waves were evaluated by a non-parametric cluster-based random permutation approach implemented in Fieldtrip (see Maris & Oostenveld, 2007). The conventional analysis was also performed to examine whether the two groups differed in the ERPs to rise time, the magnitude and latency of the MMN and P3a to pitch level/contour. To explore the relationship between perception and production, correlations between the T2-T5 production accuracy and the perceptual responses were computed, including the behavioral response latency to trials involving T2 and T5 in the tone discrimination task, as well as the neural correlates to rise time and pitch height/contour between T2 and T5. Results and Discussion Behavioral results Results of the tone discrimination task showed that the [+Per-Pro] group had significantly longer response time (RT) of trials involving T2 and T5 than the [+Per+Pro] group, [M[+Per+Pro] = 1046.18 ms, SD = 80.19; M[+Per-Pro] = 1204.54 ms, SD = 177.51; t(39)= -3.57, p = .001, Cohen's d = 1.14], although both groups achieved high accuracies (above 98%). ERP results The results of the cluster-level permutation test revealed several significant clusters in different conditions in the two participant groups (see Figure 1). For clusters that were considered MMNs, the [+Per+Pro] group exhibited the component in the conditions of T1/T5, T1/T2, and T2/T5 -- between 100 and 166 ms (post-divergence point unless specified otherwise) for T1/T2 (p < .001), between 100 and 166 ms for T1/T5 (p = .006), and between 150 and 200 ms for T2/T5 (p = .015). The [+Per-Pro] group showed MMNs in the T1/T2 (110 – 166 ms, p = .005), T1/T5 (104 – 154 ms, p = .024) and T2/T5 (150 – 200 ms, p = .015) conditions. No significant negative cluster was observed in the T5/T2 condition for either group. For P3a, only the T1/T2 condition elicited a significant positive cluster immediately following the MMN for both groups, in the time window of 300 to 400 ms for [+Per+Pro] (p = .025) and 342 to 404 ms for [+Per-Pro] (p = .039). For brain responses to rise time, both participant groups exhibited an early positive-going cluster in the T2/T5 condition in the time window between 62 and 154 ms for [+Per+Pro] (p = .015) and between 64 and 144 ms for [+Per-Pro] (p = .022). For the T5/T2 condition, an early negative-going component was observed only in the [+Per+Pro] group in the time window of 36 to 176 ms (p = .039). T-tests and mixed model ANOVAs of neural responses at Fz revealed that the [+Per+Pro] group showed a shorter MMN latency than the [+Per-Pro] group [t(39) = -2.305, p < .027, Cohen's d = - .74] in the T1/T2 condition. For the MMNs and P3a in the experimental conditions of T2/T5 and T5/T2, significant main effects of condition were found for the MMN mean amplitude [F(2, 39) = 5.85, p = .020, η2 = .13] and the MMN peak latency [F(2, 39) = 10.83, p = .002, η2 = .22], with stronger MMN responses to T2/T5 than to T5/T2 but longer latency to T2/T5 relatively to T5/T2. For rise time, results of a mixed ANOVA of the average amplitudes showed main effects of tone condition [F = (2, 39) = 47.18, p < .001, η2 = .55] and group [F = (2, 39) = 75.89, p = .017, η2 = .14], with T5 eliciting more positive responses than T2, and stronger responses from the [+Per+Pro] than [+Per-Pro] group. Correlations between production accuracy of the two rising tones and perceptual measures found that the averaged production accuracy was negatively correlated with the discrimination RT (r = -.502, p = .001), with shorter discrimination RTs associated with higher production accuracy. In addition, the production accuracy was positively correlated with the mean amplitude of brain responses to rise time of T5 (r = .421, p = .006), the larger the response, the higher the production accuracy. In summary, the present study demonstrated that tone perception is highly dynamic and exploits different acoustic cues at different stages of processing – rise time at the sensory/perceptual level and pitch feature at the cognitive level, as the auditory signal unfolds over time. Moreover, our findings revealed differential sensitivities between individuals with and without distinctive production of the two rising tones as evidenced by the differences in discrimination latency of the two tones and magnitude of brain response to short rise time. The individual differences found in production are proposed to have a perceptual origin, in that less defined phonological representations lead to less distinctive production. Figure 1 Acknowledgements This research was supported by a Small Project Fund at the University of Hong Kong (Project titled “Neural correlates and cognitive capability associated with individual variations in tone perception and production in Cantonese – An event-related potential (ERP) study”). We are grateful to all the participants in this study for their time and effort. References Gandour, J. (1983). Tone perception in far eastern-languages. Journal of Phonetics, 11(2), 149-175. Gandour, J., Potisuk, S., & Dechongkit, S. (1994). Tonal coarticulation in Thai. Journal of Phonetics, 22(4), 477-492. Khouw, E., & Ciocca, V. (2007). Perceptual correlates of Cantonese tones. Journal of Phonetics, 35(1), 104-117. Leong, V., Hämäläinen, J., Soltész, F., & Goswami, U. (2011). Rise time perception and detection of syllable stress in adults with developmental dyslexia. Journal of Memory and Language, 64(1), 59-73. Luo, H., & Poeppel, D. (2012). Cortical oscillations in auditory perception and speech: evidence for two temporal windows in human auditory cortex. Frontiers in Psychology, 3, 170. Keywords: amplitude rise time, lexical tone, mismatch negativity (MMN), ERPs (Event-Related Potentials), individual differences, speech perception and production Conference: Academy of Aphasia 53rd Annual Meeting, Tucson, United States, 18 Oct - 20 Oct, 2015. Presentation Type: Poster Topic: Not student first author Citation: Law S and Ou J (2015). Neural correlates of individual differences in processing of rising tones in Cantonese: Implications for speech perception and production. Front. Psychol. Conference Abstract: Academy of Aphasia 53rd Annual Meeting. doi: 10.3389/conf.fpsyg.2015.65.00048 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 10 Apr 2015; Published Online: 24 Sep 2015. * Correspondence: Dr. Sampo Law, University of Hong Kong, Division of Speech and Hearing Sciences, Hong Kong, Hong Kong, SAR China, splaw@hku.hk Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract The Authors in Frontiers Sampo Law Jinghua Ou Google Sampo Law Jinghua Ou Google Scholar Sampo Law Jinghua Ou PubMed Sampo Law Jinghua Ou Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.

Efficient coding of multiple nonorthogonal redundancies between acoustic dimensions in novel complex sounds

Stilp and colleagues (Proc. Natl. Acad. Sci. [2010]; JASA [2011]; PLoS One [2012]) provided perceptual evidence for efficient coding of robust covariance between acoustic dimensions in novel complex sounds. Discrimination of sounds that directly violated this redundancy (Orthogonal condition) was initially inferior to that of sounds obeying overall redundancy (Consistent condition). Performance was consistently predicted by principal components analysis (PCA), as experimental conditions aligned with statistical dimensions in PCA. Stilp and colleagues suggested efficient coding may contribute to perceptual organization for speech, but two aspects of their experimental designs qualify this extension: robust evidence for only one statistical regularity between acoustic dimensions was tested while speech possesses many; and, all statistical structure was mutually orthogonal which is often not true of speech sounds. Here, listeners discriminated sounds supporting two concurrent, nonorthogonal regularities (patterns of covariance between acoustic dimensions: attack/decay and spectral shape.) Despite nonorthogonality, these concurrent statistical regularities were efficiently coded, as discrimination of Consistent sound pairs was initially superior to that of Orthogonal sound pairs. Performance did not adhere to predictions made by PCA. Implications for speech perception and auditory ‘category’ acquisition will be discussed. [Supported by NIDCD.]

Acoustic correlates of narrow focus in Turkish

Acoustic correlates of narrow focus in Turkish focus is known to affect the acoustic properties of the lexical element under focus. Its effect on the pre-focus and post-focus domains has attracted much less attention. This study aims to investigate the acoustic changes as a function of focus in the pre-focus and post-focus domains as well as in the on-focus domain in Turkish and listeners' sensitivity to those changes in retrieving information from the acoustic signal. For this purpose, a production and a perception experiments have been conducted. For the production study, speakers read sentences in which the location of focus was manipulated via wh-questions preceding those sentences. For the perception study, listeners heard selected sentences recorded for the production study and were asked to judge the prominent word in the sentence. Results showed that focused words had increased duration and intensity, and post-focus words had reduced F0, duration and intensity. Interestingly, pre-focus words were found to have increased F0 and duration. Listeners' identification of focus was positively affected by the presence of acoustic changes especially in the post-focus domain. These findings have implications for speech perception and modelling prosody.

Speaker identification in hybrid synthesis: Implications for speech perception

As part of an evaluation of hybrid synthesis [Hertz, Proc. IEEE Workshop on Speech Synthesis (2002)], perceptual experiments were conducted that tested the hypothesis that stressed vowels are the primary cues to speaker identification. Hybrid sentences were constructed for eight voices, including child and adult and male and female, in which stressed vowels were taken from a single human speaker, but other segments were replaced by surrogates from different sources. Some surrogates were natural speech segments; others were formant synthesized. Some matched the age or gender of the stressed vowel speaker; others did not. After being trained on six human target voices, listeners were asked to identify the hybrid stimuli, and also fully synthetic and natural stimuli (for target and nontarget voices), in terms of age, gender, and whether and how much they matched a target voice. For all categories, hybrid stimuli, in contrast to synthetic ones, were identified as accurately as natural speech, both by listeners familiar with a target voice (e.g., family members) and those unfamiliar. The experiment demonstrated how little attention is paid to consonants and reduced vowels in determining speaker identity, and supported our hybrid approach to modeling voices. [Work supported by NIH Grant 1 R43 DC006761-01.]

Effects of Backward Speech and Speaker Variability in Language Discrimination by Rats.

Human infants use prosodic cues present in speech to extract language regularities, and it has been suggested that this capacity is anchored in more general mechanisms that are shared across mammals. This study explores the extent to which rats can generalize prosodic cues that have been extracted from a training corpus to new sentences and how this discrimination process is affected by the normalization of the sentences when multiple speakers are introduced. Conditions 1 and 2 show rats' abilities to use prosodic cues present in speech, allowing them to discriminate between sentences not previously heard. But this discrimination is not possible when sentences are played backward. Conditions 3 and 4 show that language discrimination by rats is also taxed by the process of speaker normalization. These findings have remarkable parallels with data from human adults, human newborns, and cotton-top tamarins. Implications for speech perception by humans are discussed.

Auditory discontinuities interact with categorization: Implications for speech perception

Behavioral experiments with infants, adults, and nonhuman animals converge with neurophysiological findings to suggest that there is a discontinuity in auditory processing of stimulus components differing in onset time by about 20 ms. This discontinuity has been implicated as a basis for boundaries between speech categories distinguished by voice onset time (VOT). Here, it is investigated how this discontinuity interacts with the learning of novel perceptual categories. Adult listeners were trained to categorize nonspeech stimuli that mimicked certain temporal properties of VOT stimuli. One group of listeners learned categories with a boundary coincident with the perceptual discontinuity. Another group learned categories defined such that the perceptual discontinuity fell within a category. Listeners in the latter group required significantly more experience to reach criterion categorization performance. Evidence of interactions between the perceptual discontinuity and the learned categories extended to generalization tests as well. It has been hypothesized that languages make use of perceptual discontinuities to promote distinctiveness among sounds within a language inventory. The present data suggest that discontinuities interact with category learning. As such, "learnability" may play a predictive role in selection of language sound inventories.

Perceptual discontinuities and categorization: Implications for speech perception

Behavioral experiments with infants, adults and nonhuman animals converge with neurophysiological findings to suggest that there is a discontinuity in auditory processing of stimulus components differing in onset time by about 20 ms. This discontinuity has been implicated as a basis for boundaries between speech categories distinguished by VOT. Here, we investigate how this discontinuity interacts with the learning of novel perceptual categories. Adult listeners were trained to categorize a nonspeech acoustic cue that mimics the temporal distinction of VOT. One group of listeners learned categories with a boundary coincident with the perceptual discontinuity. Another group learned categories defined such that the perceptual discontinuity fell within a category. Listeners in the latter group required significantly more experience to reach criterion categorization performance. The evidence of interactions between the perceptual discontinuity and the learned categories extended to generalization tests as well. It has been hypothesized that languages make use of perceptual discontinuities to promote perceptual distinctiveness among sounds within a language inventory. The present data suggest that these influences interact with category learning. As such, learnability may play a predictive role in selection of language sound inventories. Moreover, it may be possible to observe predictable learning effects in infant speech perception.

Maturation of human auditory cortex: implications for speech perception.

This project traced the maturation of the human auditory cortex from midgestation to young adulthood, using immunostaining of axonal neurofilaments to determine the time of onset of rapid conduction. The study identified 3 developmental periods, each characterized by maturation of a different axonal system. During the perinatal period (3rd trimester to 4th postnatal month), neurofilament expression occurs only in axons of the marginal layer. These axons drive the structural and functional development of cells in the deeper cortical layers, but do not relay external stimuli. In early childhood (6 months to 5 years), maturing thalamocortical afferents to the deeper cortical layers are the first source of input to the auditory cortex from lower levels of the auditory system. During later childhood (5 to 12 years), maturation of commissural and association axons in the superficial cortical layers allows communication between different subdivisions of the auditory cortex, thus forming a basis for more complex cortical processing of auditory stimuli.

Age-Related Changes in Temporal Processing: Implications for Speech Perception

Age-Related Changes in Temporal Processing: Implications for Speech Perception

The effects of speaking rate and stimulus variability on the perception of spoken words by young and elderly subjects

This study compared spoken word recognition in young and elderly normal-hearing listeners as a function of speaking rate and stimulus variability. Monosyllabic words were presented in quiet at one of three speaking rates (fast, medium, or slow). In one condition, all stimulus items were presented at the same speaking rate while in a second condition speaking rate was varied from trial to trial. In control conditions, items were presented at either fixed or varying overall amplitudes. For young listeners, spoken word identification did not differ as a function of speaking rate. Older subjects, in contrast, demonstrated similar identification scores for slow- and medium-rate items but had significantly reduced identification performance for words presented at fast speaking rates. Identification accuracy was also reduced for older listeners in the mixed—as opposed to single—rate condition but only for those items presented at the fast speaking rate. Young listeners were not affected by variability due to speaking rate and neither group showed reduced identification as a result of variability in overall amplitude. Implications for speech perception in older listeners are discussed in terms of current views about perceptual normalization for different sources of variability. [Work supported by NIH.]

Detection of silent temporal gaps between narrow-band noise makers having second-formantlike properties of voiceless stop/vowel combinations.

Temporal gap detection thresholds were measured in narrow-band noise-burst markers having acoustic characteristics representative of isolated steady-state second-formant (F2) properties for/p,t,k/paired separately with/i,ae,u,o/. The results revealed that gap detection threshold increased systematically as the difference was increased between the simulated stop and vowel F2 frequencies. A strong positive correlation (r = 0.87) between gap detection threshold and linear marker center frequency difference was highly significant (p < 0.001). Differences in other stimulus features had little influence on gap detection performance. Implications for speech perception are discussed.

“Temporal masking” of tone glides in hearing-impaired ears

It has been reported previously [M. J. Collins, J. Acoust. Soc. Am. Suppl. 1 69, S22 (1981)] that normal hearers' ability to discriminate the rate of frequency change of a gliding portion of a signal is affected by the duration of an adjacent fixed-frequency segment of the signal. This “temporal masking” is found to be greater when the glide precedes the fixed-frequency segment (backward masking) than when the glide follows the fixed-frequency segment (forward masking). The present paper reports some effects of sensorineural hearing loss on difference limens for glide rate as a function of the duration of an adjacent fixed-frequency segment. Using an adaptive procedure, data were collected on two ears of one, well-trained, hearing-impaired subject. The results show a trend for neither forward nor backward masking to occur when the discrimination is of some glide versus no glide. When the glide segment changes frequency at a rate of 17 Hz/ms, forward and backward masking is observed, with the magnitude of the effect appearing to be independent of degree of loss. The ear with the more severe loss shows the more normal result (greater backward than forward masking), while the ear with the less severe loss shows mixed effects. The results are discussed in relation in implications for speech perception. [Work supported by NINCDS.]

Isoresponse temporal masking patterns in sensorineural listeners

Pure‐tone maskers (200 ms) were adjusted to intensities that just masked short (20 ms) pure‐tone probe signals at low and at moderate sensation levels. (The probe is constant in both frequency and level during a masked threshold determination.) The delays between maskers and probes were varied to obtain what are called isoresponse temporal masking patterns. Results in normal‐hearing listeners are consistent with classical isointensity temporal masking patterns that were obtained by previous investigators, and with isoresponse temporal masking patterns that have been obtained more recently [L. L. Vogten, J. Acoust. Soc. Am. 63, 1520–1527 (1978)1. Some sensorineural listeners required considerably less masker level for temporal masking (interference) to occur than did normal‐hearing listeners, even when masker and probe were in a frequency region of relatively normal sensitivity. This result was obtained whether comparisons were made with probe signals at the same sound pressure levels or at the same sensation levels that were used for normal‐hearing listeners. Temporal masking in some sensorineural listeners also extended over a greater delay time between masker and probe than for normal‐hearing listeners. These findings have implications for current forward‐masking techniques that obtain estimates of frequency selectivity, they also have implications for speech perception by some sensorineural listeners in difficult listening situations. [Research supported in part by NIH Grants Nos. NS10561 and NS12125.]

Perceptual analysis of the amplitude envelope of AM signals

The manner in which the auditory system encodes the “shape” of the amplitude envelope of AM signals was examined. On the basis of modulation frequency tuned channels previously demonstrated for FM signals [R. H. Kay and D. R. Matthews, J. Physiol. (Lond.) 225, 657–677 (1972)] it was thought the auditory system might be responding to the power spectrum of the amplitude envelope. A 500‐Hz carrier was amplitude modulated at 2, 4, or 8‐Hz with either a sine wave or square‐wave signal. The Method of Adjustment was used to determine the “modulation thresholds” for the 6 frequency by waveform conditions. At each of the modulation frequencies, subjects' thresholds were lower for square wave than for sine wave modulation. The mean ratio between the thresholds for the two wave forms was 1.32 with a SD of 0.22, which approximates 4/π (1.27) as would be predicted on the assumption that the auditory system performs a Fourier analysis on the amplitude envelope of the modulated 500‐Hz tone. Additional data suggest that the two waveforms are indiscriminable until the third harmonic of the square wave reaches its independent threshold. Some implications for speech perception are considered. [Work supported by Penn. State Department of Psychology for partial fulfillment of master's degree.]