No evidence for language-wide enhancement of contrast from perceptual integrality

Perceptual integrality is found when perceivers do not judge differences on physically orthogonal dimensions independently [Garner, W. R. (1974). The Processing of Information and Structure. Lawrence Erlbaum Associates]. For example, F1 and nasality do not necessarily physically covary, but native English listeners judge them as if they covary negatively [Kingston, J. & Macmillan, N. (1995). Integrality of nasalization and F1 in vowels in isolation and before oral and nasal consonants: A detection-theoretic application of the Garner paradigm. JASA, 97 (2), 1261–1285.]. Kingston, J., & Diehl, R. L. [(1994). Phonetic knowledge. Language, 70 (3), 419–454.] posited that if integrality of two phonetic properties increases perceptual distance between categories differing in both, language users covary those properties in production, enhancing the contrast. We investigated whether native English speakers exploit integrality of F1 and nasality to enhance vowel height contrasts by covarying them in production. Perceptual and real-word production results paired within participants showed that F1-nasality covariation in production was predictable partly from how perceptually integrated F1 and nasality were and partly from lexical factors such as the presence of minimal pairs depending on vowel height. These results are inconsistent with enhancement of language-wide phonological contrasts, but compatible with enhancement of contrasts between lexical items.