Korean-Swedish code-switching : Theoretical models and linguistic reality : teoretiska modeller och den språkliga verkligheten

Sammanfattning: This study deals with Korean-Swedish code-switching, i.e. the alternate use of Korean and Swedish in one and the same utterance. Its purpose is to give a linguistic description of the way in which Koreans combine Swedish with Korean, and to test the universal validitv of the two most influential grammatical models in the literature of code-switching, that is, Poplack's two-constraint model and Myers-Scotton's matrix language-frame model (MLF model), on Korean-Swedish code-switching. The data consist of tape recordings of spontaneous adult, child and family conversations among Koreans in Sweden.Linguistic analysis reveals that multi-word sequences (i.e. prototypical code-switches) andsingle words (i.e. prototypical borrowings) from the other language show the same morphosyntactic characteristics: multi-word sequences are morphologically and/or syntactically integrated intothe language of the sentence as if they were single words, and single words contain the functionwords `or morphemes and the morpheme orders from the original language as if they weremulti-word sequences. Several important points are derived from this fact. Firstly, the cross-linguistic validity of the two models is called into question. Secondly, the view that there is a distinction between multi-word sequences and single words appears unconvincing. Thirdly, the distinction between code-switching and borrowing cannot be made on the basis of morphosyntactic integration. Finally, the distance between prototypical switches and prototypical borrowings is not as great as many researchers have assumed.Examination of a possible correlation between extralinguistic factors and the switching types inthe Korean-Swedish community investigated shows that the choice of the type of code-switching is also determined by factors such as the speaker's bilingual proficiency and age as well as conversation type. It is argued that by taking not only linguistic but extralinguistic factors into consideration, we might get a more comprehensive and accurate picture of the code-switching phenomenon.