The theory of adaptive dispersion and acoustic-phonetic properties of cross-language lexical-tone systems

Lexical-tone languages use fundamental frequency (F0) to convey word-meaning. Nearly half of all languages use lexical tone [Maddieson (2008)], yet those systems are under-studied. To increase our understanding of speech-sound inventory organization, I extend to tone-systems a model of vowel-system organization, the theory of adaptive dispersion (TAD) [Liljencrants and Lindblom (1972)]. This is a cross-language investigation of whether and how tone-inventory size affects acoustic tone-space size. Five languages with different-sized tone-inventories were compared: Cantonese (six tones), Thai (five tones), Mandarin (four tones), Yoruba (three tones), and Igbo (two tones). Six native speakers (three female) of each language produced 18 CV syllables in isolation, with each of his/her language’s tones, six times. Tonal F0 (semitones) was measured at three equidistant points across the vowel. Each language’s tone-space was defined in two ways: (1) the F0 difference between its highest and lowest tones and (2) the configuration of tones in a 2-D space defined by onset F0 x offglide F0. Following the TAD, I predicted that languages with larger tone-inventories would have larger tone-spaces; this was not supported by (1). However, the dispersion of tones in (2) supports the TAD hypothesis that sound-categories will be well-dispersed across the space and highly contrastive.