Contact-induced typological change (original) (raw)

2001, Language Typology and Language Universals

Introduction. It is easy to show that contact-induced change can have a profound effect on the typological profile of the receiving language. Probably the most obvious examples, and also the ones that are easiest to find, are changes in basic sentential word order. These are especially striking because it is word order features that have attracted the most attention in the typological literature, starting with the famous 1963 article by Greenberg that moved typology into the mainstream of linguistic research. But word order features-or, more generally, morpheme order features-are by no means the only parts of language structure that have been transformed under the influence of other languages. Contact-induced changes have also affected the typological character of morphological, phonological, lexical semantic, and discourse systems in a wide variety of languages. This article surveys contact-induced typological changes and shows where and how different degrees of change correlate with different social conditions. After some introductory comments on the contact conditions in which typological change is likely to happen (§2) and on the problem of determining whether a given change is typologically significant or not (§3), I will discuss and exemplify immediate typological effects (§4) and delayed typological effects (§5). Finally, I will consider the question of whether or not language mixing constitutes contact-induced typological change (§6). The article ends with a brief conclusion (§7). 2. When should we expect typological change as a result of contact? The first step in discussing contact-induced typological change is to sketch the circumstances under which it is likely to occur. The crucial point-hardly an astonishing one-is that intense contact is more likely than casual contact to result in typological restructuring of the receiving language. The question of what counts as intense contact (an admittedly vague term) depends on the perspective: the expected results differ according to whether the situation is one of borrowing, where fluent speakers of the receiving language adopt features from the source language, or imperfect learning, where native speakers of the source language have learned the receiving language imperfectly and incorporate their learners' errors into their version of it. The latter type of situation usually, though not always, involves language shift; for convenience, I will refer to these as shift situations. The expected linguistic results are as different as the processes: in borrowing situations, the first interference features to turn up in the receiving language are loanwords, followed (if contact becomes intense enough) by structural features, especially in the phonology and syntax. In shift situations, by contrast, the first interference comprises phonological and syntactic features, and sometimes there are very few loanwords at any stage. Intensity of contact in a borrowing situation depends largely, though not entirely, on the level of bilingualism (number of bilinguals, degree of fluency) among borrowing-language speakers: more bilinguals, more intensity, more interference. If the level of bilingualism among borrowing-language speakers is low, then interference features are likely to be confined to loanwords and minor structural features that do not disrupt the typological patterns