The role of consonant-vowel transitions in the perception of the stop and nasal consonants.

The motor theory of speech perception revised

“…We now add that the articulatory variation due to context is pervasive: in the acoustic representation of every phonetic category yet studied there are context-conditioned portions that contribute to perception and that must, therefore, be taken into account by theory. Thus, for stops, nasals, fricatives, liquids, semivowels, and vowels, the always context-sensitive transitions are cues (Harris, 1958;Jenkins, Strange, & Edman, 1983;Liberman et al, 1954;O'Connor, Gerstman, Liberman, Delattre, & Cooper, 1957;Strange, Jenkins, & Johnson, 1983). For stops and fricatives, the noises that are produced at the point of constriction are also known to be cues, and, under some circumstances at least, these, too, vary with context (Dorman et al, 1977;Liberman et al, 1952;Whalen, 1981).…”

“…Such context-conditioned variation is most apparent, perhaps, in the transitions of the formants as the constriction is released. Thus, place information for a given consonant is carried by a rising transition in one vowel context and a falling transition in another (Liberman, Delattre, Cooper, & Gerstman, 1954). In isolation, these transitions sound like two different glissandi or chirps, which is just what everything we know about auditory perception leads us to expect (Mattingly, Liberman, Syrdal, & Halwes, 1971); they do not sound alike, and, just as important, neither sounds like speech.…”

“…For example, if the onset frequency of the transition of the second formant during a stop release is sufficiently low, relative to the frequency of the following steady state, the stop is perceived as labial; otherwise, as apical or dorsal (Liberman et al, 1954). A value along such an acoustic dimension that was optimal for a particular phonetic category, or, more loosely, the dimension itself, was termed an 'acoustic cue'.…”

A unified account of categorical effects in phonetic perception

“…Previous studies have found a large amount of restructuring for stop consonant and much less for steady-state vowels (Liberman et al 1967, Lisker & Abramson, 1964b, Liberman et al 1954). This led researchers to propose that the degree of restructuring correlated with categorical effects on perception, namely that those sounds that exhibited a large amount of restructuring were perceived categorically and those that did not are perceived continuously.…”

“…Stop consonants were found to exhibit a large amount of restructuring, changing how they appear acoustically even though they have the same underlying phonemic status. This was found in both correlates of place of articulation (Liberman et al, 1967) as well as manner and voicing (Lisker & Abramson, 1964b;Liberman et al, 1954). Steady state vowels, on the other hand, show no such restructuring, when accounting for speaker normalization and speaking rates (Liberman et al, 1967).…”