Comparison of consonantal and vocalic cues in selective adaptation

The acoustic cues to the phonetic identity of diphthongs normally include both spectral quality and dynamic change. This fact was exploited in a series of selective adaptation experiments examining the possibility of mutual adaptive effects between these two types of acoustic cues. One continuum of syllables varying from [εi] to [εd] and another varying from [ε] to [εi] were synthesized; endpoint stimuli of both series used as adaptors caused identification boundaries to be shifted. Cross-series adaptation was also attempted on the [ε?εi] stimuli, using [?], [∞], and [ai]. Only [ai] proved effective as an adaptor, suggesting the mediation of a rather abstract auditory level of similarity. The results argue strongly against interpretations in terms of feature detectors, but appear compatible with an “auditory contrast” explanation, which might in turn be incorporated within adaptation level theory in the form recently discussed by Restle (1978). The cross-series results further suggest that selective adaptation might be used to quantify the perceptual distance between auditory cues in speech.     

The specificity of perceptual learning in speech processing

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 [epsilonf]-[epsilons] 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.     

Adaptation of the category boundary between speech and nonspeech: A case against feature detectors

Two experiments were performed employing acoustic continua which change from speech to nonspeech. The members of one continuum, synthesized on the Pattern Playback, varied in the bandwidths of the first three formants in equal steps of change, from the vowel /α/ to a nonspeech buzz. The other continuum, achieved through digital synthesis, varied in the bandwidths of the first five formants, from the vowel /æ/ to a buzz. Identification and discrimination tests were carried out to establish that these continua were perceived categorically. Perceptual adaptation of these continua revealed shifts in the category boundaries comparable to those previously reported for speech sounds. The results were interpreted as suggesting that neither phonetic nor auditory feature detectors are responsible for perceptual adaptation of speech sounds, and that feature detector accounts of speech perception should therefore be reconsidered.     

Selective adaptation of vowels

Two experiments investigating the selective adaptation of vowels examined changes in listeners’ identification functions for the vowel continuum [i-I-∈] as a function of the adapting stimulus. In Experiment I, the adapting stimuli were [i], [I], and [∈]. Both the [i] and [∈] stimuli produced significant shifts in the neighboringand distant phonetic boundaries, whereas [I] did not result in any adaptation effects. In order to explore the phonetic nature of feature adaptation in vowels, a second experiment was conducted using the adapting stimuli [gig] and [g ∈ g], which differed acoustically from the [i] and [∈] vowels on the identification continuum. Only [gig] yielded reliable adaptation effects. The results of these experiments were interpreted as suggesting arelative rather than a stableauditory mode of feature analysis in vowels and a possibly more complex auditory feature analysis for the vowel [i].     

辅音和元音知觉中的启动效应差异

采用启动范式,在两个实验中分别考察了辅音范畴和元音范畴知觉中的启动效应。启动音是纯音和目标范畴本身,目标音是辅音范畴和元音范畴连续体。结果发现辅音范畴连续体知觉的范畴反应百分比受到纯音和言语启动音影响,辅音范畴知觉的反应时只受言语启动音影响;元音范畴连续体知觉的范畴反应百分比不受两种启动音影响,但元音范畴知觉的反应时受到言语启动音影响。实验结果表明辅音范畴和元音范畴知觉中的启动效应存在差异,这为辅音和元音范畴内在加工机制的差异提供了新证据。     

Constraints on the processes responsible for the extrinsic normalization of vowels

Listeners tune in to talkers’ vowels through extrinsic normalization. We asked here whether this process could be based on compensation for the long-term average spectrum (LTAS) of preceding sounds and whether the mechanisms responsible for normalization are indifferent to the nature of those sounds. If so, normalization should apply to nonspeech stimuli. Previous findings were replicated with first-formant (F1) manipulations of speech. Targets on a [pt]–[p?t] (low–high F1) continuum were labeled as [pt] more after high-F1 than after low-F1 precursors. Spectrally rotated nonspeech versions of these materials produced similar normalization. None occurred, however, with nonspeech stimuli that were less speechlike, even though precursor–target LTAS relations were equivalent to those used earlier. Additional experiments investigated the roles of pitch movement, amplitude variation, formant location, and the stimuli's perceived similarity to speech. It appears that normalization is not restricted to speech but that the nature of the preceding sounds does matter. Extrinsic normalization of vowels is due, at least in part, to an auditory process that may require familiarity with the spectrotemporal characteristics of speech.     

Adaptation of speech by nonspeech: evidence for complex acoustic cue detectors

Three selective adaptation experiments were run, using nonspeech stimuli (music and noise) to adapt speech continua ([ba]-[wa] and [cha]-[sha]). The adaptors caused significant phoneme boundary shifts on the speech continua only when they matched in periodicity: Music stimuli adapted [ba]-[wa], whereas noise stimuli adapted [cha]-[sha]. However, such effects occurred even when the adaptors and test continua did not match in other simple acoustic cues (rise time or consonant duration). Spectral overlap of adaptors and test items was also found to be unnecessary for adaptation. The data support the existence of auditory processors sensitive to complex acoustic cues, as well as units that respond to more abstract properties. The latter are probably at a level previously thought to be phonetic. Asymmetrical adaptation was observed, arguing against an opponent-process arrangement of these units. A two-level acoustic model of the speech perception process is offered to account for the data.     

Perceptual learning in speech

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., [WItlo?], 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).     

Susceptibility of a stop consonant to adaptation on a speech-nonspeech continuum: Further evidence against feature detectors in speech perception

The present experiment uses the perceptual adaptation paradigm to establish the validity of a previous test of the feature detector model of speech perception. In the present study, a synthetic stimulus series varied from a CV syllable, [ba], to a nonspeech buzz. When the endpoint tokens were employed alternatively as adaptors, the category boundary was shifted relative to unadapted identification in each adaptor condition. This result suggests that a prior test which used a vowel as the speech endpoint was legitimate because a stop consonant, an exemplary speech sound, was also susceptible to perceptual adaptation in a speech-nonspeech context. Feature detector models predict, incorrectly, that this outcome is impossible. Therefore, this finding may be taken to undermine the interpretation of adaptation as fatigue in a set of detectors tuned to detect the distinctive features of linguistic analysis.     

Context effects as auditory contrast

Three experiments are reported that collectively show that listeners perceive speech sounds as contrasting auditorily with neighboring sounds. Experiment 1 replicates the well-established finding that listeners categorize more of a [d–g] continuum as [g] after [l] than after [r]. Experiments 2 and 3 show that listeners discriminate stimuli in which the energy concentrations differ in frequency between the spectra of neighboring sounds better than those in which they do not differ. In Experiment 2, [alga–arda] pairs, in which the energy concentrations in the liquid-stop sequences are H(igh) L(ow)–LH, were more discriminable than [alda–arga] pairs, in which they are HH–LL. In Experiment 3, [da] and [ga] syllables were more easily discriminated when they were preceded by lower and higher pure tones, respectively—that is, tones that differed from the stops’ higher and lower F3 onset frequencies—than when they were preceded by H and L pure tones with similar frequencies. These discrimination results show that contrast with the target’s context exaggerates its perceived value when energy concentrations differ in frequency between the target’s spectrum and its context’s spectrum. Because contrast with its context does more that merely shift the criterion for categorizing the target, it cannot be produced by neural adaptation. The finding that nonspeech contexts exaggerate the perceived values of speech targets also rules out compensation for coarticulation by showing that their values depend on the proximal auditory qualities evoked by the stimuli’s acoustic properties, rather than the distal articulatory gestures.     

A “labial” feature analyzer in speech perception

A selective adaptation experiment was conducted to determine the ability of various adapting stimuli to alter the perception of a series of 13 synthetic speech syllables. The synthetic test syllables, which varied acoustically in the starting frequency and direction of second- and third-formant transitions, included stop consonant distinctions ofplace of articulation for the syllable types [bae], [dae], and [gae]. A systematic adaptation effect was produced in the locus of the bae-bae phonetic boundary for these stimuli after repetitive listening to each of the following adapting syllables: [bae], [p^hae], [mae], and [vae], indicating that perception ofplace distinctions among the stop consonants can be altered even by repetitive listening to certain speech sounds not belonging to the stop-consonant class.     

Evidence for a central representation of instrument timbre

An information processing account of perception seeks to delineate the stages of processing through which a stimulus passes and determine the properties of the representation at each stage. Research in phonetic perception has identified two stages, the second of which is thought to encode abstract acoustic attributes of sounds. The present study provided a further test of this proposal by assessing whether nonphonetic stimuli could yield results similar to those obtained with phonetic stimuli. Five selective adaptation experiments were carried out with a trumpet—piano timbre continuum. Two manipulations were used to measure abstract encoding: cross-ear presentation of adaptor and test series, and the use of adaptors that were acoustically different from the continuum endpoints. The results provide evidence for an abstract representation of timbre. The similarity of the findings to those in the phonetic adaptation literature is discussed.     

Re-examining selective adaptation: Fatiguing feature detectors,or distributional learning?

When a listener hears many good examples of a /b/ in a row, they are less likely to classify other sounds on, e.g., a /b/-to-/d/ continuum as /b/. This phenomenon is known as selective adaptation and is a well-studied property of speech perception. Traditionally, selective adaptation is seen as a mechanistic property of the speech perception system, and attributed to fatigue in acoustic-phonetic feature detectors. However, recent developments in our understanding of non-linguistic sensory adaptation and higher-level adaptive plasticity in speech perception and language comprehension suggest that it is time to re-visit the phenomenon of selective adaptation. We argue that selective adaptation is better thought of as a computational property of the speech perception system. Drawing on a common thread in recent work on both non-linguistic sensory adaptation and plasticity in language comprehension, we furthermore propose that selective adaptation can be seen as a consequence of distributional learning across multiple levels of representation. This proposal opens up new questions for research on selective adaptation itself, and also suggests that selective adaptation can be an important bridge between work on adaptation in low-level sensory systems and the complicated plasticity of the adult language comprehension system.     

特征关联的类别间面孔适应

研究通过系列实验探讨了面孔适应不仅仅发生在形状选择性上, 也能发生在任务相关的特征上有内在关联的两个不同类别的物体间。实验1以带有明显性别倾向的物品图片作为适应刺激, 让被试对男女之间morphing程度不同的图片面孔进行性别辨别, 考察了不同适应刺激呈现时间的类别间面孔适应。结果表明适应刺激呈现时间大于50 ms时均存在类别间面孔适应效应。实验2评估了“性别”这一特征以及适应刺激形式在类别间面孔适应中所起的作用, 结果发现带有性别倾向的物品图片、相应的物品名称和性别文字(“男性”、“女性”) 3种适应刺激类型均能产生类别间适应。实验3通过操纵适应刺激上的注意负荷(高负荷、低负荷和无负荷), 探究了注意对类别间面孔适应的影响。结果表明随着注意负荷的增加, 类别间面孔适应效应减小。3个实验报告了一个新异的类别间适应后效, 证明了适应也能发生于在任务相关特征上有内在关联的两个不同类别的物体间。     

Functional equivalence of masking and cue reduction in perception of shape at a slant

The present study was undertaken to investigate the effects of syllabic stress and segment structure on selective adaptation in speech. To this end, a CV place of articulation test continuum was selectively adapted by seven different adapting stimuli; the monosyllables [ba] and [ga], two disyllabic stimuli containing equal stress on both syllables, [baga] and [gabal, and three disyllabic stimuli ([baga]) in which stress placement varied and was cued by the acoustic parameters of fundamental frequency and duration. Results for the two adapting stimuli demonstrated significant [b] adaptation for the stimulus [ba] and significant [g] adaptation for [gal. Of the five other adapting stimuli, only [g] adaptation for the stimulus [bagá] was found to be significant. These findings indicate that the operation of detector mechanisms susceptible to fatigue by an adapting stimulus are even more constrained than has heretofore been suggested. It appears that the adapting and test stimuli must not only have the same phonetic and syllable structure, but also the same syllabic organization.     

Phonetic recalibration only occurs in speech mode

Upon hearing an ambiguous speech sound dubbed onto lipread speech, listeners adjust their phonetic categories in accordance with the lipread information (recalibration) that tells what the phoneme should be. Here we used sine wave speech (SWS) to show that this tuning effect occurs if the SWS sounds are perceived as speech, but not if the sounds are perceived as non-speech. In contrast, selective speech adaptation occurred irrespective of whether listeners were in speech or non-speech mode. These results provide new evidence for the distinction between a speech and non-speech processing mode, and they demonstrate that different mechanisms underlie recalibration and selective speech adaptation.     

Are selective adaptation and contrast effects really distinct?

Although there is evidence that selective adaptation and contrast effects in speech perception are produced by the same mechanisms, Sawusch and Jusczyk (1981) reported a dissociation between the effects and concluded that adaptation and contrast occur at separate processing levels. They found that an ambiguous test stimulus was more likely to be labeled b following adaptation with [pha] and more likely to be labeled p following adaptation with [ba] or [spa] (the latter consisting of [ba] preceded by [s] noise). In the contrast session, where a single context stimulus occurred with a single test item, the [ba] and [pha] contexts had contrastive effects similar to those of the [ba] and [pha] adaptors, but the [spa] context produced an increase in b responses to the test stimulus, an effect opposite to that of the [spa] adaptor. One interpretation of this difference is that the rapid presentation of the [spa] adaptor gave rise to "streaming," whereby the [s] was perceptually segregated from the [ba]. In our experiment, we essentially replicated the results of Sawusch and Jusczyk (1981), using procedures similar to theirs. Next, we increased the interadaptor interval to remove the likelihood of stream segregation and found that the adaptation and contrast effects converged.     

Selective adaptation effects on end-point stimuli in a speech series

Recent experiments in speech perception using the selective adaptation paradigm have found that the phonetic boundary of a test series shifts following adaptation. However, no changes within the phonetic category have been found. In the present experiment, a series of voiced CV syllables which varied along the feature of place was used in a selective adaptation paradigm. The end-point stimuli trom the test series were used as adaptors. Subjects used a 6-point rating scale to respond to the stimuli instead of the usual two-category identification. The average rating for end-point stimuli from the same category as the adaptor, as well as the boundary stimuli, shifted as a function of adaptation. In all cases, the average rating response shifted toward that of the unadapted category. The average rating for stimuli in the opposite category from that of the adaptor remained relatively unchanged. These results indicate that the entire category of the adapting stimulus changes as a function of selective adaptation and that the effect is not confined to stimuli near the phonetic boundary.     

Vowel-contingent feature detection

Do vowel-contingent selective adaptation effects for place of articulation depend on vowel identity, or on the particular formant frequencies used? An experiment is reported here which tested the adaptation effects of consonants with exactly the same formant transitions before different diphthongs. In this experiment, the phonetic identity of the vowel and the formant frequencies of the consonant are not confounded as they have been in previous studies. In the contingent adaptation condition, no phoneme boundary shifts were observed, and this is interpreted along with previous evidencefor such phoneme boundary shifts when phonetic identity of the vowel and formant frequencies are confounded as indicating that adaptation operates on the spectral representation of the stimulus. Other evidence consistent with this conclusion is that adaptation with alternating adaptors from each end of the test series produced negligible shifts, and that a single adaptor from a diphthong series different from that of the test series produced phoneme boundary shifts as large as those with a single adaptor from the same test series.     

Phonetic prototypes

Speech perception may be viewed as a phonetic categorization task in which the listener assigns incoming sounds to various phonetic categories. The present experiment tests two classes of models of phonetic categorization: (1) models in which the listener has a threshold or boundary between alternative categories vs. (2) models in which the listener compares the input to prototypical representations of the alternative phones. In a pretest, listeners located on a VOT continuum the /ga/ that they thought was prototypical. Selective adaptation was then conducted using both the selected prototype and adaptors nearer and further from the phoneme boundary. The prototype adaptor produced more adaptation than the other adaptors. This result supports a prototype-based representation for phonetic categorization; several process models using such a representation are considered.