Perceptual features for the identification of Romance languages

This paper deals with perceptual identification and differentiation of five Romance languages, namely French, Italian, Spanish, Portuguese and Romanian. Previous studies have investigated human capability to identify spoken samples in unknown languages after a relatively brief exposure. Moreover, they have shown that subjects use perceptual categories and adapt foreign categories to their own categories in language differentiation. Accordingly, we conduct an analysis to determine which perceptual categories are salient in Romance languages identification and discrimination. Four different sets of listeners are tested. Each set consists in speakers of a different mother tongue (French, Romanian, Japanese and American English native speakers). Results reveal that identification scores are a function of the previous exposure of the listeners to the languages. Moreover, the strategies of discrimination among languages are mother tongue dependent and several potential features emerge that may be relevant in automatic language identification.

[1]  Détermination expérimentale d'indices linguistiques pour la discrimination des langues romanes , 2000 .

[2]  A. D. Dominicis,et al.  Intonation Systems: A Survey of Twenty Languages , 1999 .

[3]  Zinny S. Bond,et al.  Can children identify samples of foreign languages as same or different , 1994 .

[4]  A. Lawrence Spitz,et al.  Automatic language identification , 1997 .

[5]  Ronald A. Cole,et al.  Perceptual benchmarks for automatic language identification , 1994, Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing.

[6]  Zinny S. Bond,et al.  Perceptual features of unknown foreign languages as revealed by multi-dimensional scaling , 1996, Proceeding of Fourth International Conference on Spoken Language Processing. ICSLP '96.