Woordproduktie bij Tweetaligen

An attempt is made to validate an adaptation of Levelťs (1989) Speaking model to bilingual speakers. It is concluded that the best empirical evidence available at the moment comes from research on speech errors and code switching. Data from the latter source suggests that in word retrieval, 'language' serves as one of the cues for the selection of a specific lemma. Such cues can have more or less weight, and their impact will vary accordingly. Language choice in code switching can be regarded as the result of an interaction between communicative intentions and the information that is available in time during the production process. In order to arrive at a full-fledged model, observational data will have to be supplemented with experimental research using real-time techniques.