Increases in complexity as a result of language contact
The paper discusses the possibility of partially or entirely new linguistic patterns developing in a contact situation and how this influences the complexity of the language system, in particular the grammar. In earlier work, I have argued that such linguistic complexity often arises out of the competition between different patterns or by the incomplete expansion of one pattern into the territory of another. As an example, I discuss the principles governing the choice of “long” and “short” Russian adjectives. Language contact plays an obvious role in that the competition may be between an inherited pattern and a borrowed one. Thus, since a borrowed pattern does not always replace existing patterns completely, distinctions are created that did not exist in either of the contributing languages. There is also a geographical aspect to this in that the effects of language contact are not likely to be equally strong everywhere, with “truces” between old and new patterns being most likely in intermediate areas (“buffer zones”). I exemplify this by two examples from Scandinavian. One is the rise of a suppletive verb paradigm as the result of the competition between two verbs for “become” in Swedish: <i>varda</i> and <i>bli(va)</i>, where the latter is a loan from Low German. In parts of Sweden, forms of <i>varda</i> are retained in parts of the paradigm, notably in the past and the supine. The second example concerns the two definite articles in Scandinavian: the preposed free articles (<i>den, det, de</i>), similar to the ones of West Germanic, and thee suffixed articles (<i>-en, -et</i> etc.). In definite noun phrases with preposed modifiers, there are two major patterns: (i) only a preposed article is used (Danish and marginally Norwegian); (ii) both a preposed article and a suffixed one are used (Swedish and Norwegian). However, the actual picture shows considerable diversity both within and between varieties. There is a general tendency for the two articles to have different degrees of strength in the Scandinavian languages, with the preposed article being strongest in Danish and the postposed being strongest in Swedish. The different geographical distribution of the two articles suggests that two separate grammaticalization have been operating in partially overlapping geographical areas. In the “buffer zone” between these, both preposed and suffixed articles are in use, but a division of labour has been found, the details of which vary depending on the location.