Aspects of language contact: A variationist perspective on codeswitching and borrowing in Igbo-English bilingual discourse.

This study is based on empirical data collected from bilingual speakers of Igbo (a Kwa language) and English, in an environment propitious to the use of both languages. The study examines two of the most widely discussed constraints on CS namely, Free Morpheme Constraint and Equivalence Constraint (Poplack 1980). The controversy surrounding these and other constraints on CS arise primarily from the problems of drawing a clear distinction between CS and borrowing. Distinguishing between these language contact phenomena has been particularly difficult with singly-occurring lexical items from one language incorporated into the discourse otherwise of the other. Our investigation begins by determining the status of lone English-origin items incorporated into otherwise Igbo discourse. In order to determine whether these are CS or borrowings, we use the principles of variation theory to make a detailed assessment of the behavior of these forms in the context of the entire bilingual system. Our method entails a systematic comparison of the lone items with: (a) unmixed stretches of Igbo; (b) unmixed stretches of English and (c) multiword fragments of English (unambiguous CS) juxtaposed to Igbo. Since CS items are qualitatively and quantitatively similar to their counterparts in the language which lexified them, while borrowings assume the behavior of their counterparts in the recipient language, our method effectively disambiguates the contentious lone English-origin items by comparing their patterns of behavior with respect to predetermined diagnostics, vis-a-vis their counterparts in the unmixed stretches of the two languages as well as unambiguous CS. If the lone English-origin items patterned like their counterparts in unmixed stretches of English and unambiguous CS, they would be classified as CS. If, on the other hand, they patterned like their counterparts in unmixed stretches of Igbo, there would be no doubt that they are borrowings into Igbo. As expected, our results produced conclusive evidence that these lone English-origin items are borrowings into Igbo. In all the examined criteria namely, vowel harmony and affixation, the lone English-origin verbs patterned like their counterparts in unmixed stretches of Igbo, but differed from unmixed English and unambiguous CS. The English-origin nouns on their part also behaved like their counterpart in unmixed stretches of English in such areas as determiner usage, the use of generic reference, the linear structure of NPs. The lone English-origin adjectives were incorporated into Igbo as adjectival nouns, the most productive adjectival category in Igbo. These lone English origin adjectives followed the copula di (BE) in the same proportion as their counterparts in unmixed Igbo. Once the borrowed items have been identified and separated from the bona fide CS, we found that, with very few exceptions, the switches between Igbo and English occurred at points where the structures of the two languages are linearly analogous. Thus, Igbo-English CS is constrained under equivalence. (Abstract shortened by UMI.