Constraining variation in decreolization

CONSTRAINING VARIATION IN DECREOLIZATION William Washabaugh University of Wisconsin, Milwaukee A replication of D. Bickerton's analysis of fi-tu variation in Caribbean English (1971) indicates that a number of overly-strong hypotheses about variation in decreolization must be modified. Variation in decreolization is notalways bidirectional, not completely directed by the pressure to acquire the standard language, and not entirely constrained by deep-structure categories. But this is not to deny that variation is systematic. Rather, it is directed by both social and linguistic pressures, and is constrained not only by purely linguistic factors, but also by lexical categories established by psychological processes.* 'Post-creole continuum' has become a conventional term among students of creóle languages, and the implicational analysis of post-creole continua is equally well-accepted as an analytic procedure. That analytic procedure has been refined by Bickerton 1971 and employed to describe numerous aspects of post-creole continua. The results of such implicational analyses of post-creole continua have been significant and exciting. Yet, exciting though B's model and its results may be, that model is the first step toward an understanding of decreolization, not the last. The first step in any scientific investigation of a new phenomenon is to postulate and support a set ofvery strong hypotheses. B's model ofdecreolization isjust such a first step toward understanding. As such, it contains some very strong hypotheses : (a) The motive for decreolization is the social pressure on speakers to acquire a standard language (acrolect); therefore speakers in a post-creole commum'ty vary their speech along the single dimension of standard (acrolect)/non-standard (basilect). (b) The constraints on, at least, the variation between the complementizers fi and tu in decreolization are all of a kind, and deal with the deep-structure configuration of the sentences in which./? or tu appear. These hypotheses are a first step toward an understanding of decreolization. The purpose of my investigation, however, is to undertake a second step, in which the hypotheses are put to an empirical test, to see ifthey should be weakened, or even discarded. Accordingly, I will deal with these questions: (i) Is the model proposed by B for describingfi-tu variation in Guyana applicable to other communities which employ Caribbean English? B implies that it should be; my evidence indicates that it is. (ii) Is variation in decreolization entirely unidimensional? My evidence says no; this strong hypothesis must be modified, (iii) Are constraints onfi-tu variation in decreolization all deep-structure constraints? My evidence again says no; this strong hypothesis must also be modified. * This study is based on data collected on Providence Island, Colombia, in 1972-73 and again in 1975. I gratefully acknowledge the financial support of the National Institute of Mental Health, the University of Wisconsin-Milwaukee Graduate School, and the American Philosophical Society, without which these data could not have been gathered. The arguments presented here are similar to those in Washabaugh 1974, 1976b, though considerable revising has been done to accommodate the data gathered in 1975. 1 am indebted to Jane Hill and to Fred Eckman for their comments on earlier drafts of these ideas. 329 330LANGUAGE, VOLUME 53, NUMBER 2 (1977) The result ofthis second step toward understanding decreolization will show that it is actually a much more complex process than B's model suggests. Rather than being subject to a single sort ofsocial pressure, speakers in a post-creole community are triply pressured : to avoid the basilect, to acquire the acrolect, and to vary the mesolect. Second, constraints on any variable in decreolization are both deepstructural and surface-structural; they are based both on linguistic categories and on categories developed through non-linguistic principles. 1. Bickerton on fi-tu variation. In this section I will outline B's analysis of the variation in the form ofthe pre-infinitival complementizer in Guyanese decreolization , in order to make clear the principles which he employs. I will follow the outline with an account of my own replication of B's analysis, using data from another and quite distant part of the Caribbean. The intent of the replication is to show that B's model is applicable to post-creole communities other...