Psychoacoustic measures of stop production in Cantonese, Greek, English, Japanese, and Korean.

Abstract

Spectral analyses of stop bursts have revealed that the place of articulation can be predicted based on both invariant and time-varying cues present within the acoustic signal (e.g., Stevens and Blumstein, 1978; Forrest et al., 1988). While prior studies have had some success in uncovering such cues for American English, it is not clear whether these parameters are equally pertinent in stoping consonant classification in other languages. Furthermore, one of the main limitations of a linear acoustic analysis is that it imposes different scales of loudness and frequency on the acoustic signal than does the human ear, thus generating power spectra with different frequency distributions than are produced by the auditory system (e.g., Zwicker 1961; Kewley-Port, 1983). In the present study of word-initial stops produced in Cantonese, English, Greek, Japanese, and Korean, measures derived from a psychoacoustic model of auditory perception were developed in order to more accurately isolate the front cavity resonances of the burst transient. For all languages, peak amplitude frequency and a peak compactness measure successfully distinguished alveolar from velar stops with at least 74% accuracy and were more accurate at stop classification than were spectral moments measures.

The Role of Experience in the Perception of Phonetic Detail in Children’s Speech: A Comparison Between Speech-Language Pathologists and Clinically Untrained Listeners

“…Multiple regressions predicting the average ratings for the inexperienced and experienced listeners' ratings from the peak ERB, compactness, and peak loudness of the stop bursts were conducted separately for front- and back-vowel contexts, as work by Arbisi-Kelm et al (2008) showed that different sets of psychoacoustic measures discriminated between /t/ and /k/ in different vowel contexts. The correlations between the acoustic measures and responses was so low that the analyses were not meaningful and are hence not reported here.…”

“…The stimuli were analyzed using a set of psychoacoustic measures, as described in Arbisi-Kelm, Beckman, Kong, and Edwards (2008), Arbisi-Kelm et al (2010), and Munson et al (2010). Briefly, these are measures of the spectra of the stop bursts and the intervals of frication that are based on models of human hearing, rather than on linear measures.…”

“…The same measurements were performed on the stops, expect the peak loudness (in sones) was used instead of total loudness. The results for stops are presented separately by front- and back-vowel contexts, as the parameters that differentiate place of articulation have been found to differ between front- and back-vowel contexts (Arbisi-Kelm, Beckman, Kong, & Edwards, 2008). …”

Deconstructing phonetic transcription: Covert contrast, perceptual bias, and an extraterrestrial view ofVox Humana

“…The Greek speakers, for example, rated the English velar tokens as more-front than the English-speaking listeners did. This cross-language asymmetry indicates that this perception is mediated by language-specificity in the way that phonological contrasts are expressed, a topic that we explore further in Arbisi-Kelm, Beckman, Kong, and Edwards (2008).…”

“…Additionally, measures of duration and of the second-formant frequency at vowel onset (in ERB) are reported in Table II. A defence of the psychophysical measures, as well as an illustration of their benefit over traditional linear measures, can be found in Arbisi-Kelm, Beckman, Kong, and Edwards (2008).…”