ITL - International Journal of Applied Linguistics - Volume 83, Issue 1, 1989

Scholarly publications about Dutch immigrant speech are practically non-existent -- excepting the work of Clyne (1967, 1972, 1985, 1987). This applies particularly to Dutch speakers who immigrated in the 19th and the early 20th Century. A 75-hour corpus of language data from Dutch immigrants to the United States is currently being investigated systematically at the Royal Netherlands Academy of Arts and Sciences' Institute for Dialectology. The general purpose of the project is to gain more insight in the nature of Dutch language and dialect maintenance or loss and of English influence on the language of these Dutch immigrants. This paper will focus specifically on English lexical elements in the speech of the first-generation immigrants (appr. 12,000 English elements in the Dutch of 97 speakers).

First a diagnostic set of characteristics for code-switching, borrowing and nonce borrowing will be presented, based in part on discussions in the recent literature on other-language lexical elements in speech in contact situations and in part on a thorough review of the Dutch-American data. These characteristics do not only concern the formal linguistic level, but also the discourse, processing and extralin-guistic levels. They form the basis of a systematic classification of the various types of English elements found in the data. Next, a number of problem cases will be discussed -- with examples -- which defy easy classification in one of the three categories of other-language elements. Finally, an attempt will be made to formulate a more generalized model for the analysis of other-language lexical elements, consisting of a diagnostic and an analytic layer. The diagnostic layer assigns an element to one of the three categories of other-language elements. The analytic layer is specifically concerned with the analysis of code-switching, comprising not only a categorization of formal linguistic phenomena, but also a categorization of functional sociolinguistic factors determining the use of switched elements. Given the variety of linguistic as well as extralinguistic aspects which apparently determine the use of switched elements, it seems unlikely that one single set of constraints could be refined enough to fully account for linguistic as well as extralinguistic aspects of code-switching. Therefore, a case will be made for the establishment of more than one set of constraining factors on the use of switched elements. Instead of limiting constraints on code-switching to the formal linguistic -- in particular the syntactic -- level, separate constraining factors for the discourse, processing, functional, and motivational level might be formulated, in order to gain full insight in the process of code-switching.