Variation and standardisation : the case of Afrikaans (1880-1922)

Bibliography: p. 335-366.Following the general model outlined in Weinreich, Labov & Herzog (1968), this study is a contribution to the historiography of Afrikaans from a variationist perspective, investigating the patterns of linguistic variability in the context of the early standardisation of Afrikaans. The work is based on a newly collected historical corpus of private documents which includes letters and diaries from 136 individuals (written between 1880 and 1922), and can be said to represent acrolectal and mesolectal usage. Several morphosyntactic, morpholexical and syntactic variables were investigated: loss of person and number distinctions in the present tense paradigm, loss of the infinitive, regularisation of the past participle, loss of the preterite, loss of gender, the emergence of a new system of adjective inflection and of a new pronoun system, the so-called 'double' negation, infinitive clauses, the use of objective vir, and the periphrastic possessive with se. The quantitative analysis of these variables makes use of a variety of methods: descriptive techniques such as distribution analysis, implicational scaling, cluster analysis, multidimensional scaling and principal components analysis, as well as inferential statistics such as the chi-square test. Variation is furthermore described from a code-switching perspective. As a study in historical sociolinguistics this dissertation is also concerned with the epistemological aspects of socio-historical research, in particular the role of speaker agency in historical explanations, the 'measurement' of the extralinguistic variables in sociolinguistic research, the nature of the relationship between sociolinguistic and social theory, and in general the ontological status of our explanatory and descriptive concepts and taxonomies. While traditionally historiographers of Afrikaans have argued that there existed a sharp linguistic and functional distinction between Afrikaans and Dutch from the mid 18th century, the patterns of variability described for the corpus indicate the existence of a complex dialect continuum (rather than diglossia) until the early 20th century. The results of the quantitative analysis suggest furthermore that the process of linguistic change was slower than hitherto assumed, and variation patterns described for the late 18th century were still found to exist in the corpus. Such continuities challenge the conventional dating of the emergence of Afrikaans as a new language or dialect (characterised by almost complete morphological regularisation and a cluster of innovative syntactic features) to around 1800. As regards the standardisation of Afrikaans the study shows that from the 1850s a relatively uniform model of what constituted the 'vernacular' (or ‘Afrikaans, as it came to be known) existed as a well-defined entity in the popular consciousness, while the actual language use of many speakers remained rather more variable. Increasingly, linguistic practices which were not in line with the propagated model of 'Afrikaans' were identified by the contemporary metalinguistic discourse (which was strongly marked by 19th century cultural nationalism) as unauthentic and thus undesirable. The diffusion of the new standard is shown to have followed the path typical for modern standard languages, i.e. via the socially (and geographically) mobile professional class or intelligentsia. After about 1914 the new standard was widely diffused, and had replaced other Netherlandic varieties in many private documents