ASA 128th Meeting - Austin, Texas - 1994 Nov 28 .. Dec 02

5pSP1. The IPA problem.

Peter Ladefoged

Phonet. Lab., Linguist. Dept., UCLA, Los Angeles, CA 90024-1543

Every few years it is appropriate to review our knowledge of the phonetic structures of the world's languages and try to assess how many different speech sounds there are. Problems arise because it is not easy to say whether two sounds occurring in different languages are the same or not. Two sounds are definitely different if they distinguish words within a language. But if they occur in different languages this test cannot be used. This is the basis of the IPA problem. The International Phonetic Association tries to provide a way of symbolizing every distinct speech sound, but it cannot tell whether two sounds are potentially distinct when they have been observed only in different languages. The data from recent phonetic studies indicate that there are limitless ways in which the sounds of one language can differ from another, but that the number of parameters along which sounds can vary is fairly small. In most cases the variation in the sounds that occur in different languages is the result of the use of a different value of one of the parameters, rather than the use of some novel parameter that has not been observed in other languages.