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

Remez, Robert E.

doi:10.3758/bf03199900

Cited by 13 publications

(8 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Remez (1979), for example, has shown that even a continuum in which formant bandwidths are widened in steps to create a series from a vowel to a nonspeech "buzz" is equally susceptible to adaptation by both endpoint sounds. Identical results were found for a series varying from [ba] to nonspeech "buzz" in Remez (1980), providing further evidence that, even at the auditory level, any use of the term "feature" to describe the basis of adaptive effect with speech sounds must be so qualified that it can hardly have any straightforward phonetic interpretation.…”

Section: Discussionmentioning

confidence: 72%

Comparison of consonantal and vocalic cues in selective adaptation

Godfrey

1980

Perception & Psychophysics

View full text Add to dashboard Cite

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 [sr] to [Ed] and another varying from [E] to [El] were synthesized; endpoint stimuli of both series used as adaptors caused identification boundaries to be shifted. Cross-series adaptation was also attempted on the [E:-E:l] stimuli. using I:>], [:>Gl]. and [at]. Only [at] 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.

show abstract

Section: Discussionmentioning

confidence: 72%

Comparison of consonantal and vocalic cues in selective adaptation

Godfrey

1980

Perception & Psychophysics

View full text Add to dashboard Cite

show abstract

“…To the difficulties exposed by these authors, we may add the following: (1)Detectors conceived phonetically fail to predict the occasions on which adaptation does and does not occur; (2) detectors conceived auditorily have so proliferated as to eliminate the appealing simplicity of the original model--to derive perceptual categories in the auditory version of the model requires cumbersome and implausible detector interactions; and (3) the adaptation test itself may not be appropriate if the aim is to catalogue fixed preferences and sensitivities of analytic channels (see Diehl, Elman, & McCusker, 1978;Remez, 1979;Simon & Studdert-Kennedy, 1978). Whether the theory is largely false (as claimed by Diehl et al, 1978;Remez, 1980;Simon & StuddertKennedy, 1978) or is minimally erroneous (as claimed by Cooper, 1979;Eimas & Miller, 1978;Samuel & Newport, 1979) is a matter that has received much attention lately. However, it is obvious that with no explicit use for distinctive feature theory possible, and with no manageable set of auditory templates in hand, the adaptation studies of phonetic stimuli can have little to say about the classic problems in speech perception (i.e., recognition, segmentation, normalization, etc.).…”

Section: Discussionmentioning

confidence: 99%

Cross-series adaptation using song and string

Remez

Cutting

Studdert‐Kennedy

1980

Perception & Psychophysics

Self Cite

View full text Add to dashboard Cite

The acousticrauditory feature "risetime" has been claimed to underlie both the phonetic affricate-fricative distinction and the nonphonetic plucked-string/bowed-string distinction. We used the perceptual adaptation technique to determine whether the risetime differences of the [d$a]-[ba ] distinction would therefore be registered by the same mechanism that mediates risetime differences for the plucked-bowed distinction. Two continua were used, one of digitally modified natural speech and one of synthetic violin sounds, in which the risetime was varied across each set of tokens from 0 msec to 80 msec in steps of 10 msec. The speech was sung and the violin notes were synthesized with the same fundamental frequency, 294 Hz. Adaptation of the category boundaries was observed only when speech adaptors were tested with the speech continuum and when violin adaptors were tested with the violin continuum. When crossseries tests were performed (violin adaptors tested with the speech series, and speech adaptors tested with the violin series), no effect of adaptation was observed. This finding indicates that these speech and violin sounds, despite obvious acoustic similarities, do not share the same feature detectors.

show abstract

“…A further difficulty with the fixed singularity hypothesis is the finding that the location of the category boundary can vary under the influence of different experimental conditions (Repp, 1984; Repp & Liberman, 1987). Adaptation can affect speech-sound identification, as shown by a shift in the location of the psychometric function (e.g., Ades, 1974a, 1974b; Cole, Cooper, Singer, & Allard, 1975; Eimas, Cooper, & Corbit, 1973; Eimas & Corbit, 1973; Kat & Samuel, 1984; Remez, 1980; Sawusch, 1976) and, by implication, a shift in the criterion. Many studies have shown sequential effects consequent on preceding stimuli or responses (Diehl & Kluender, 1987; Repp, 1983; Repp & Liberman, 1987; Sawusch & Nusbaum, 1979; Treisman et al, 1995).…”

Section: How Does the Hypersigmoid Function Arise?mentioning

confidence: 99%

There are two types of psychometric function: A theory of cue combination in the processing of complex stimuli with implications for categorical perception.

Treisman¹

1999

Journal of Experimental Psychology: General

View full text Add to dashboard Cite

Psychometric functions are customarily fitted using normal distribution functions. However, categorical perception identification tasks give functions that are more sigmoid than the normal distribution (M. Treisman, A. Faulkner, P. L. N. Naish, & B. S. Rosner, 1995). This can be explained by a theory of discrimination, cue combination theory (CCT), presented here. This theory assumes that complex stimuli (e.g., speech sounds) are composed of components or cues that vary in informativeness. The decision mechanisms sample individual cue dimensions and combine the information from each. The model generates a normal psychometric function if all cues sampled are informative. If not, the function becomes more sigmoid. CCT offers accounts of discrimination (categorical and continuous) and sensory reaction times, adaptation, sequential effects, contextual effects, trading relations, speed and accuracy, perceptual learning, and the separable-integral distinction.

show abstract

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

Cited by 13 publications

References 33 publications

Comparison of consonantal and vocalic cues in selective adaptation

Comparison of consonantal and vocalic cues in selective adaptation

Cross-series adaptation using song and string

There are two types of psychometric function: A theory of cue combination in the processing of complex stimuli with implications for categorical perception.

Contact Info

Product

Resources

About