Accommodation and foreigner talk in an experimental setting

Foreigner talk (FT), the modified language frequently directed toward nonnative speakers (NNSs), can be seen as a product of accommodation, the tendency of speakers to adjust their language based on their speech partner. This dissertation investigates FT within the framework of accommodation theory to determine what native speakers (NSs) react to when they produce different kinds of FT.;This work is based primarily on research conducted in Berlin and the surrounding area in June and July of 1994. Subjects were 160 German students who completed a questionnaire comprising four situations, each of which required subjects to imagine they were speaking to a particular person and to write down what they would say. Because this study was intended to investigate accommodation to another's language, subjects were given varying exposure to texts of their addressee's language.;Subject responses were analyzed according to the length and complexity of utterance and occurrence of typical FT characteristics, yielding the following main results. Subjects addressed an imagined NNS of their language in substantially different ways than they did an imagined NS, regardless of whether or not they had exposure to texts of nonnative language. Subjects who did have text exposure, however, produced quantitatively and qualitatively different language to the NNS than did subjects who lacked text exposure. Subjects who addressed an imagined NS also produced significantly shorter and less complex language when they had exposure to authentic texts than they did without such exposure.;These results have implications for FT and for accommodation theory. First, exposure to authentic language in written form had profound effects on the language subjects produced to an addressee, indicating that NSs are sensitive to the language of others and can accommodate toward it even after short and highly artificial exposure. Second, NSs apparently produce different kinds of language to a NNS when they have to rely on stereotype and experience than they do when presented with samples of that person's language. In this study, speakers made their major language adjustments (quantitative) based on their addressee's nonnativeness and then made fewer but often more dramatic adjustments (qualitative) after exposure to nonnative language.