Publications

Abstract

We investigated neural mechanisms that support voice recognition in a training paradigm with fMRI. The same listeners were trained on different weeks to categorize the mid-regions of voice-morph continua as an individual's voice. Stimuli implicitly defined a voice-acoustics space, and training explicitly defined a voice-identity space. The predefined centre of the voice category was shifted from the acoustic centre each week in opposite directions, so the same stimuli had different training histories on different tests. Cortical sensitivity to voice similarity appeared over different time-scales and at different representational stages. First, there were short-term adaptation effects: Increasing acoustic similarity to the directly preceding stimulus led to haemodynamic response reduction in the middle/posterior STS and in right ventrolateral prefrontal regions. Second, there were longer-term effects: Response reduction was found in the orbital/insular cortex for stimuli that were most versus least similar to the acoustic mean of all preceding stimuli, and, in the anterior temporal pole, the deep posterior STS and the amygdala, for stimuli that were most versus least similar to the trained voice-identity category mean. These findings are interpreted as effects of neural sharpening of long-term stored typical acoustic and category-internal values. The analyses also reveal anatomically separable voice representations: one in a voice-acoustics space and one in a voice-identity space. Voice-identity representations flexibly followed the trained identity shift, and listeners with a greater identity effect were more accurate at recognizing familiar voices. Voice recognition is thus supported by neural voice spaces that are organized around flexible ‘mean voice’ representations.

Abstract

The possible-word constraint (PWC; Norris, McQueen, Cutler, & Butterfield, 1997) has been proposed as a language-universal segmentation principle: Lexical candidates are disfavoured if the resulting segmentation of continuous speech leads to vowelless residues in the input—for example, single consonants. Three word-spotting experiments investigated segmentation in Slovak, a language with single-consonant words and fixed stress. In Experiment 1, Slovak listeners detected real words such as ruka “hand” embedded in prepositional-consonant contexts (e.g., /gruka/) faster than those in nonprepositional-consonant contexts (e.g., /truka/) and slowest in syllable contexts (e.g., /dugruka/). The second experiment controlled for effects of stress. Responses were still fastest in prepositional-consonant contexts, but were now slowest in nonprepositional-consonant contexts. In Experiment 3, the lexical and syllabic status of the contexts was manipulated. Responses were again slowest in nonprepositional-consonant contexts but equally fast in prepositional-consonant, prepositional-vowel, and nonprepositional-vowel contexts. These results suggest that Slovak listeners use fixed stress and the PWC to segment speech, but that single consonants that can be words have a special status in Slovak segmentation. Knowledge about what constitutes a phonologically acceptable word in a given language therefore determines whether vowelless stretches of speech are or are not treated as acceptable parts of the lexical parse.

Abstract

Signed languages are articulated through simultaneous upper-body movements and are seen; spoken languages are articulated through sequential vocal-tract movements and are heard. But word recognition in both language modalities entails segmentation of a continuous input into discrete lexical units. According to the Possible Word Constraint (PWC), listeners segment speech so as to avoid impossible words in the input. We argue here that the PWC is a modality-general principle. Deaf signers of British Sign Language (BSL) spotted real BSL signs embedded in nonsense-sign contexts more easily when the nonsense signs were possible BSL signs than when they were not. A control experiment showed that there were no articulatory differences between the different contexts. A second control experiment on segmentation in spoken Dutch strengthened the claim that the main BSL result likely reflects the operation of a lexical-viability constraint. It appears that signed and spoken languages, in spite of radical input differences, are segmented so as to leave no residues of the input that cannot be words.

Abstract

Japanese listeners detected Japanese words embedded at the end of nonsense sequences (e.g., kaba 'hippopotamus' in gyachikaba). When the final portion of the preceding context together with the initial portion of the word (e.g., here, the sequence chika) was compatible with many lexical competitors, recognition of the embedded word was more difficult than when such a sequence was compatible with few competitors. This clear effect of competition, established here for preceding context in Japanese, joins similar demonstrations, in other languages and for following contexts, to underline that the functional architecture of the human spoken-word recognition system is a universal one.

Abstract

For optimal word recognition listeners should use all relevant acoustic information as soon as it comes available. Using printed-word eye-tracking we investigated when during word processing Dutch listeners use suprasegmental lexical stress information to recognize words. Fixations on targets such as 'OCtopus' (capitals indicate stress) were more frequent than fixations on segmentally overlapping but differently stressed competitors ('okTOber') before segmental information could disambiguate the words. Furthermore, prior to segmental disambiguation, initially stressed words were stronger lexical competitors than non-initially stressed words. Listeners recognize words by immediately using all relevant information in the speech signal.

Abstract

Four cross-modal repetition priming experiments examined whether consonant duration in Italian provides listeners with information not only for segmental identification ("what" information: whether the consonant is a geminate or a singleton) but also for lexical segmentation (“where” information: whether the consonant is in word-initial or word-medial position). Italian participants made visual lexical decisions to words containing geminates or singletons, preceded by spoken primes (whole words or fragments) containing either geminates or singletons. There were effects of segmental identity (geminates primed geminate recognition; singletons primed singleton recognition), and effects of consonant position (regression analyses revealed graded effects of geminate duration only for geminates which can vary in position, and mixed-effect modeling revealed a positional effect for singletons only in low-frequency words). Durational information appeared to be more important for segmental identification than for lexical segmentation. These findings nevertheless indicate that the same kind of information can serve both "what" and "where" functions in speech comprehension, and that the perceptual processes underlying those functions are interdependent.

Abstract

In foreign-accented speech, listeners have to handle noticeable deviations from the standard pronunciation of a target language. Three cross-modal priming experiments investigated how short- and long-term experiences with a foreign accent influence word recognition by native listeners. In experiment 1, German-accented words were presented to Dutch listeners who had either extensive or limited prior experience with German-accented Dutch. Accented words either contained a diphthong substitution that deviated acoustically quite largely from the canonical form (huis [hys], "house", pronounced as [hoys]), or that deviated acoustically to a lesser extent (lijst [lst], "list", pronounced as [lst]). The mispronunciations never created lexical ambiguity in Dutch. While long-term experience facilitated word recognition for both types of substitutions, limited experience facilitated recognition only of words with acoustically smaller deviations. In experiment 2, Dutch listeners with limited experience listened to the German speaker for 4 min before participating in the cross-modal priming experiment. The results showed that speaker-specific learning effects for acoustically large deviations can be obtained already after a brief exposure, as long as the exposure contains evidence of the deviations. Experiment 3 investigates whether these short-term adaptation effects for foreign-accented speech are speaker-independent.

Abstract

Prosodic influences on phonetic realizations of four Dutch consonants (/t d s z/) were examined. Sentences were constructed containing these consonants in word-initial position; the factors lexical stress, phrasal accent and prosodic boundary were manipulated between sentences. Eleven Dutch speakers read these sentences aloud. The patterns found in acoustic measurements of these utterances (e.g., voice onset time (VOT), consonant duration, voicing during closure, spectral center of gravity, burst energy) indicate that the low-level phonetic implementation of all four consonants is modulated by prosodic structure. Boundary effects on domain-initial segments were observed in stressed and unstressed syllables, extending previous findings which have been on stressed syllables alone. Three aspects of the data are highlighted. First, shorter VOTs were found for /t/ in prosodically stronger locations (stressed, accented and domain-initial), as opposed to longer VOTs in these positions in English. This suggests that prosodically driven phonetic realization is bounded by language-specific constraints on how phonetic features are specified with phonetic content: Shortened VOT in Dutch reflects enhancement of the phonetic feature {−spread glottis}, while lengthened VOT in English reflects enhancement of {+spread glottis}. Prosodic strengthening therefore appears to operate primarily at the phonetic level, such that prosodically driven enhancement of phonological contrast is determined by phonetic implementation of these (language-specific) phonetic features. Second, an accent effect was observed in stressed and unstressed syllables, and was independent of prosodic boundary size. The domain of accentuation in Dutch is thus larger than the foot. Third, within a prosodic category consisting of those utterances with a boundary tone but no pause, tokens with syntactically defined Phonological Phrase boundaries could be differentiated from the other tokens. This syntactic influence on prosodic phrasing implies the existence of an intermediate-level phrase in the prosodic hierarchy of Dutch.

Abstract

We conducted four experiments to investigate the specificity of perceptual adjustments made to unusual speech sounds. Dutch listeners heard a female talker produce an ambiguous fricative [?] (between [f] and [s]) in [f]- or [s]-biased lexical contexts. Listeners with [f]-biased exposure (e.g., [witlo?]; from witlof, “chicory”; witlos is meaningless) subsequently categorized more sounds on an [εf]–[εs] continuum as [f] than did listeners with [s]-biased exposure. This occurred when the continuum was based on the exposure talker's speech (Experiment 1), and when the same test fricatives appeared after vowels spoken by novel female and male talkers (Experiments 1 and 2). When the continuum was made entirely from a novel talker's speech, there was no exposure effect (Experiment 3) unless fricatives from that talker had been spliced into the exposure talker's speech during exposure (Experiment 4). We conclude that perceptual learning about idiosyncratic speech is applied at a segmental level and is, under these exposure conditions, talker specific.

Abstract

In a phonological priming experiment using spoken Dutch words, Dutch listeners were taught varying expectancies and relatedness relations about the phonological form of target words, given particular primes. They learned to expect that, after a particular prime, if the target was a word, it would be from a specific phonological category. The expectancy either involved phonological overlap (e.g., honk-vonk, “base-spark”; expected related) or did not (e.g., nest-galm, “nest-boom”; expected unrelated, where the learned expectation after hearing nest was a word rhyming in -alm). Targets were occasionally inconsistent with expectations. In these inconsistent expectancy trials, targets were either unrelated (e.g., honk-mest, “base-manure”; unexpected unrelated), where the listener was expecting a related target, or related (e.g., nest-pest, “nest-plague”; unexpected related), where the listener was expecting an unrelated target. Participant expectations and phonological relatedness were thus manipulated factorially for three types of phonological overlap (rhyme, one onset phoneme, and three onset phonemes) at three interstimulus intervals (ISIs; 50, 500, and 2,000 msec). Lexical decisions to targets revealed evidence of expectancy-based strategies for all three types of overlap (e.g., faster responses to expected than to unexpected targets, irrespective of phonological relatedness) and evidence of automatic phonological processes, but only for the rhyme and three-phoneme onset overlap conditions and, most strongly, at the shortest ISI (e.g., faster responses to related than to unrelated targets, irrespective of expectations). Although phonological priming thus has both automatic and strategic components, it is possible to cleave them apart.

Abstract

Although researchers studying human speech recognition (HSR) and automatic speech recognition (ASR) share a common interest in how information processing systems (human or machine) recognize spoken language, there is little communication between the two disciplines. We suggest that this lack of communication follows largely from the fact that research in these related fields has focused on the mechanics of how speech can be recognized. In Marr's (1982) terms, emphasis has been on the algorithmic and implementational levels rather than on the computational level. In this article, we provide a computational-level analysis of the task of speech recognition, which reveals the close parallels between research concerned with HSR and ASR. We illustrate this relation by presenting a new computational model of human spoken-word recognition, built using techniques from the field of ASR that, in contrast to current existing models of HSR, recognizes words from real speech input.

Abstract

We report detailed analyses of a very large database on timing of speech perception collected by Smits et al. (Smits, R., Warner, N., McQueen, J.M., Cutler, A., 2003. Unfolding of phonetic information over time: A database of Dutch diphone perception. J. Acoust. Soc. Am. 113, 563–574). Eighteen listeners heard all possible diphones of Dutch, gated in portions of varying size and presented without background noise. The present report analyzes listeners’ responses across gates in terms of phonological features (voicing, place, and manner for consonants; height, backness, and length for vowels). The resulting patterns for feature perception differ from patterns reported when speech is presented in noise. The data are also analyzed for effects of stress and of phonological context (neighboring vowel vs. consonant); effects of these factors are observed to be surprisingly limited. Finally, statistical effects, such as overall phoneme frequency and transitional probabilities, along with response biases, are examined; these too exercise only limited effects on response patterns. The results suggest highly accurate speech perception on the basis of acoustic information alone.

Abstract

Magnuson, McMurray, Tanenhaus, and Aslin [Cogn. Sci. 27 (2003) 285] suggest that they have evidence of lexical feedback in speech perception, and that this evidence thus challenges the purely feedforward Merge model [Behav. Brain Sci. 23 (2000) 299]. This evidence is open to an alternative explanation, however, one which preserves the assumption in Merge that there is no lexical-prelexical feedback during on-line speech processing. This explanation invokes the distinction between perceptual processing that occurs in the short term, as an utterance is heard, and processing that occurs over the longer term, for perceptual learning.

Abstract

Spoken word recognition consists of two major component processes. First, at the prelexical stage, an abstract description of the utterance is generated from the information in the speech signal. Second, at the lexical stage, this description is used to activate all the words stored in the mental lexicon which match the input. These multiple candidate words then compete with each other. We review evidence which suggests that positive (match) and negative (mismatch) information of both a segmental and a suprasegmental nature is used to constrain this activation and competition process. We then ask whether, in addition to the necessary influence of the prelexical stage on the lexical stage, there is also feedback from the lexicon to the prelexical level. In two phonetic categorization experiments, Dutch listeners were asked to label both syllable-initial and syllable-final ambiguous fricatives (e.g., sounds ranging from [f] to [s]) in the word–nonword series maf–mas, and the nonword–word series jaf–jas. They tended to label the sounds in a lexically consistent manner (i.e., consistent with the word endpoints of the series). These lexical effects became smaller in listeners’ slower responses, even when the listeners were put under pressure to respond as fast as possible. Our results challenge models of spoken word recognition in which feedback modulates the prelexical analysis of the component sounds of a word whenever that word is heard

Abstract

This study demonstrates that listeners use lexical knowledge in perceptual learning of speech sounds. Dutch listeners first made lexical decisions on Dutch words and nonwords. The final fricative of 20 critical words had been replaced by an ambiguous sound, between [f] and [s]. One group of listeners heard ambiguous [f]-final words (e.g., [WI tlo?], from witlof, chicory) and unambiguous [s]-final words (e.g., naaldbos, pine forest). Another group heard the reverse (e.g., ambiguous [na:ldbo?], unambiguous witlof). Listeners who had heard [?] in [f]-final words were subsequently more likely to categorize ambiguous sounds on an [f]–[s] continuum as [f] than those who heard [?] in [s]-final words. Control conditions ruled out alternative explanations based on selective adaptation and contrast. Lexical information can thus be used to train categorization of speech. This use of lexical information differs from the on-line lexical feedback embodied in interactive models of speech perception. In contrast to on-line feedback, lexical feedback for learning is of benefit to spoken word recognition (e.g., in adapting to a newly encountered dialect).

Abstract

Participants' eye movements were monitored as they heard sentences and saw four pictured objects on a computer screen. Participants were instructed to click on the object mentioned in the sentence. There were more transitory fixations to pictures representing monosyllabic words (e.g. ham) when the first syllable of the target word (e.g. hamster) had been replaced by a recording of the monosyllabic word than when it came from a different recording of the target word. This demonstrates that a phonemically identical sequence can contain cues that modulate its lexical interpretation. This effect was governed by the duration of the sequence, rather than by its origin (i.e. which type of word it came from). The longer the sequence, the more monosyllabic-word interpretations it generated. We argue that cues to lexical-embedding disambiguation, such as segmental lengthening, result from the realization of a prosodic boundary that often but not always follows monosyllabic words, and that lexical candidates whose word boundaries are aligned with prosodic boundaries are favored in the word-recognition process.

Abstract

We have recently developed a new model of human speech recognition, based on automatic speech recognition techniques [1]. The present paper has two goals. First, we show that the new model performs well in the recognition of lexically ambiguous input. These demonstrations suggest that the model is able to operate in the same optimal way as human listeners. Second, we discuss how to relate the behaviour of a recogniser, designed to discover the optimum path through a word lattice, to data from human listening experiments. We argue that this requires a metric that combines both path-based and word-based measures of recognition performance. The combined metric varies continuously as the input speech signal unfolds over time.

Abstract

We present the results of a large-scale study on speech perception, assessing the number and type of perceptual hypotheses which listeners entertain about possible phoneme sequences in their language. Dutch listeners were asked to identify gated fragments of all 1179 diphones of Dutch, providing a total of 488 520 phoneme categorizations. The results manifest orderly uptake of acoustic information in the signal. Differences across phonemes in the rate at which fully correct recognition was achieved arose as a result of whether or not potential confusions could occur with other phonemes of the language ~long with short vowels, affricates with their initial components, etc.!. These data can be used to improve models of how acoustic phonetic information is mapped onto the mental lexicon during speech comprehension.