The evolution of meaning and grammar
Central to Chomskyan language theory, often seen as a Kuhnian revolution, is the claim that universal grammar (UG) is innate and that it comprises the principles and parameters which constitute and constrain possible human language grammars. In this model, grammar is essentially a programmed fait accompli, and the causes of language change are sought principally in the reanalyses resulting from the language acquisition process. An alternate approach, grammaticalization, would appear to be potentially at odds with this view, depending on the degree of specificity assigned to UG. In studies of grammaticalization, distinct, restricted, and therefore virtually predictable pathways of change (as opposed to random parameter settings) and evolving (as opposed to innate) grammatical categories have been identified as the linguistic/pragmatic outcomes of children’s and adult speakers’ use of language. In the generative perspective, language change is seen as resulting directly from the acquisition process, rather than from language usage. Evidence from grammaticalization studies makes it clear that understanding the nature of language development and change requires that actual usage, performance, be taken into account– not just competence. In assessing whether these approaches can be brought into alignment, a variety of issues are examined: unidirectionality in grammaticalization, the linguistic views of Antoine Meillet and Michel Bréal, the poverty of the stimulus argument and UG, grammaticalization and uniformitarianism, the central role of pragmatics and discourse analysis, the relationship between UG and grammatical categories, evidence from pidgin and creole studies as a challenge to the bioprogram hypothesis, acquisition of ASL as evidence of UG being innate abstract patterning, the loss of expressivity as a factor in grammaticalization, and the emergence of apparently new linguistic categories and their relation to cognition.