Listening to a second language through the ears of a first

The processes by which listeners recognize spoken language are highly language-specific. Listenersââ"šÂ¬Ã¢"žÂ¢ expectations of how meaning is expressed in words and sentences are formed by the lexicon and grammar of the native language; but the phonology plays an even more immediate role. Thus the native phoneme repertoire constrains listenersââ"šÂ¬Ã¢"žÂ¢ ability to discriminate phonetic contrasts; and a further area in which such constraints arise is the segmentation of continuous speech into its component words. A large body of research is summarised here, motivating three conclusions: (1) In segmenting speech, speakers of different languages apply different heuristic procedures, efficiently exploiting the specific phonological structure of their various languages. (2) These procedures have become part of the listenersââ"šÂ¬Ã¢"žÂ¢ processing system, to an extent that they are also applied when listening to nonnative languages, even though this may lead to inefficiency. (3) It may be impossible to acquire the use of multiple procedures of this kind; but it is possible to inhibit the misapplication of native procedures to other languages for which they are inefficient.