Syntactic variation in Afrikaans : an empirical study

Includes abstract. Includes bibliographical references (leaves 114-122).This dissertation presents a variationist analysis of syntactic variation and change in modern spoken Afrikaans. The Afrikaans language community is heterogeneous, and can be divided into different communication communities according to patterns of segregated residential settlement and limited social interaction (linked to South Africa's history of apartheid). The selection of a sample for the study is informed by these realities and the sample is kept deliberately homogenous (following Barbiers, Cornips and Van der Kleij, 2000): participants (N=34) are White middle-class speakers of Afrikaans who are under 36 years of age and have been residing in Cape Town for at least the past seven to ten years. In addition, all participants are bilingual in English (as established through an electronically administered language use survey).In order to combine formal theory (generative linguistics) withΓÇó empirical analysis (sociolinguistics), the methodology follows a bi-modal approach. Both performance and competence are considered, and arguments are based on two types of data: speech data (interviews, narrative picture descriptions) and grammaticality judgements (elicited by means of an oral questionnaire). Grosjean's (2001) language mode model assists in refining the methodology of the study, because it recognises the fact that a bilingual speaker is a unique speaker-hearer (Chomsky, 1965). The empirical data are elicited in near-monolingual Afrikaans language modes. The results are quantified according to token frequencies and analyzed in comparison to other studies; significance tests are carried out using Chi-square and Fisher's exact tests. From the literature, the consensus seems to be that the word order in Afrikaans (XV structure) is changing to resemble an English frame (VX structure) because of language contact (cf, inter alia, Conradie, 2004; Donaldson, 1991). Two syntactic variables are studied to investigate variation in word order and verb placement: firstly, changes from XV to VX in subordinate clauses are explored by looking at the use of specific types of subordinate clauses, and the impact of matrix clause bridge verbs on complementizer omission and dependent/independent word order in the speech corpora. Secondly, the study examines the use of direct linking verbs and the role that complex verb initials play in proliferating VX structures. The findings are as follows: with embedded clause word order, the corpus data provides evidence of the frequent use of complementizer-less VX subordinate clauses that were not formally elicited in the questionnaire. These clauses have an important impact on variation in Afrikaans word order, leading to the proliferation of VX embedded clauses. Contributing factors are the weakening of the complementizer's semantic strength, and the role of the bridge verb as quotative marker in the matrix clause. Furthermore, the corpus data shows high frequencies of non-standard complementizer-led VX clauses, especially when compared to their low meta-linguistic acceptance in the questionnaires. The data thus shows significant variation in embedded clause word order and suggests that we are witnessing a change in progress for this variable. With respect to complex verb initials, the study finds a high acceptance rate in the questionnaires, as well as regular use in the spoken language corpus. By comparing the use of complex verb initials to a previous study (Ponelis, 1993) the study establishes a change in real time where an increase in the use of complex verb initials promotes the suspension of the main verb in clause-final position (V-final). The study recommends that that the bi-modal approach of considering both performance and competence data should be applied to similar studies of other groups in the Afrikaans language community