||Cross-linguistic trends in the perception of place of articulation in stop consonants: A comparison between Hungarian and French
A basic question in the study of speech is to understand how phonetic features contribute to phoneme categories in a given language. This question has crucial importance both for linguistic development during infancy and for the genesis of cross-linguistic differences between phonological systems. Possible answers to this question in current theories of speech development are that phonological features are acquired: (1) through selection of those among the universal predispositions for feature perception which are relevant for perceiving categories in a given language; (2) by exposure to phoneme categories, without relationship with the predispositions; (3) through couplings between predispositions. Although models of cross-linguistic diversity are traditionally based on the optimization of distances between categories, hence sharing affinities with the environmental model of speech development, the coupling theory might also be used as systemic model. Here we present some new evidence in support of this theory. Perceptual data were collected in the framework of the Distinc-tive Region Model of speech production, which assigns the four potential place categories to four different regions in the F2-F3 transition onset space, with flat transitions corresponding to natural boundaries in the neutral vocoïd context. Different synthetic speech continua, varying along different directions in the F2-F3 transition space were constructed and perceptual data were collected in two different languages differing as to the number of place categories: French – a three-category language- vs. Hungarian - a four-category language. Results indicate that labelling boundaries occupy similar positions for both languages and that the alveolar/palatal Hungarian boundary is poorly represented in the F2-F3 transition space. This suggests that a third phonetic feature is necessary for perceiving the fourfold place distinctions in Hungarian. Other aspects of the data suggest that the boundaries are indeed generated by couplings between predispositions. However, the data do not allow to discard a possible complementary role of distance-based optimization processes in the build-up of phonological categories.