Differences in American English, Spanish, and monkey perception of thesay-stay trading relation

Abstract

An interesting phenomenon in human speech perception is the trading relation, in which two different acoustic cues both signal the same phonetic percept. The present study compared American English, Spanish, and monkey listeners in their perception of the trading relation between gap duration and Fl transition onset frequency in a synthetic say-stay continuum. For all the subjects, increased gap duration caused perception to change from say to stay; however, subjects differed in the extent to which the Fl cue traded with gap duration. For American English listeners, a change from a low to a high Fl onset caused a phoneme boundary shift of 26 msec toward shorter gap durations, indicating a strong trading relation. For Spanish listeners, the shift was significantly smaller at 13.7 msec, indicating a weaker trading relation. For monkeys, there was no shift at all, indicating no trading relation. These results provide evidence that the say-stay trading relation is dependent on perceptual learning from linguistic exposure.There is currently much interest in determining the basis for trading relations in human speech perception. A trading relation is a phenomenon in which two different acoustic cues signal the same phonetic percept. Within limits, the cues trade with one another along a synthetic speech continuum-that is, the value ofone cue may be increased while the value of the other is decreased, yet the phonetic percept remains the same for the listener. For example, one ofthe first trading relations to be studied in detail was that between voice onset time (VaT, a temporal cue) and Fl onset frequency (a spectral cue) as perceived by American English (AE) listeners: As the Fl onset frequency increases (decreases) in a synthetic VaT continuum, more voiceless (voiced) stimuli are perceived, thus shifting the phoneme boundary toward lower (higher) values of VaT (Summerfield & Haggard, 1977).What causes a trading relation? The most parsimonious explanation is that the effect is due to generalized psychoacoustic mechanisms that are innate and immutable in all normal listeners and, thus, independent oflanguage learning (e.g., Kluender & Lotto, 1994). A second possibility is that trading relations are due to perceptual learning from linguistic exposure, including articulatory processes (e.g., Best, 1995 nisms are not necessarily mutually exclusive, however, and it is likely that trading relations, as well as other aspects of speech perception, implicate both innate and learned mechanisms that interact to varying degrees (see, e.g., Pisoni, Lively, & Logan, 1994).There are several ways in which the relative contributions of psychoacoustic versus perceptual learning processes in speech perception could be evaluated. First, we can compare the perception of native and nonnative listeners. Differences would obviously support a perceptuallearning component, because there is no reason to expect psychoacoustic differences in sensitivity between native and nonnative listeners. Second, we can compare the perception ofnative list...

Listeners recover /t/s that speakers reduce: Evidence from /t/-lenition in Dutch

“…That is, the effect of the preceding context was larger in the cases of more ambiguous signals-the intermediate conditions of our Coda Signal series-than in cases of clear signals-the endpoints of our Coda-Signal series. This mirrors other findings on context effects in speech perception, which have also shown maximal context effects for ambiguous signals (Allen & Miller, 2001;Gow, 2003;Fowler et al, 2000;Liberman, 1996;Lotto & Kluender, 1998;Mann, 1980;Massaro & Cohen, 1983;Nittrouer & StuddertKennedy, 1987;Sinnott & Saporita, 2000).…”

On the causes of compensation for coarticulation: Evidence for phonological mediation

“…It may, nevertheless, be possible to make a more specific statement about the causes of compensation. Two other context effects for which learning seems crucial are the cases of the perception of stop voicing influenced by the f 0 in the vowel (Holt et al, 2001) and the trading relation of silence and F1 onset in the perception of [ ] versus [ ] onsets (Sinnott & Saporita, 2000). Effects seem to be based on audition for liquid-voiced-stop sequences, fricative-unvoiced-stop sequences, and even liquid assimilation in Hungarian.…”

The faculty of language: what's special about it?

“…They fail to segregate the initial consonant from the vowel when compensating for syllable length in discriminating phonemes (Sinnott, Brown, & Borneman, 1998). They fail to trade off the duration of the silent gap with the formant transition in perceiving stop consonants within consonant clusters (Sinnott & Saporita, 2000). They fail to show the asymmetrical "magnet effect" that characterizes infants' discrimination of speech sounds varying in acoustic similarity to prototype vowels (Kuhl, 1991).…”