ABSTRACT This project continues and elaborates research begun under NSF grants BNS 8812552 and BNS 9009232 from September, 1988, through August, 1991. The research conducted during that period looks at the role of urbanization in language change. In addition to providing substantive conclusions about urbanization and language change and the "divergence hypothesis," that work develops innovative techniques for gathering and linking synchronic and diachronic data and explores several analytical techniques, such as cluster analysis and ARCINFO computer mapping, that are new to linguistics. The new work builds on those techniques, on spectrographic study, and on diffusion models developed in cultural geography to expand the analysis begun in earlier phases and to apply the results to some fundamental questions about language change and the role of social context in variation and change. Specifically, the work addresses these questions. First, do grammatical changes originate and spread in ways similar to phonological ones? Second, are there some general principles of linguistic diffusion that account for the spread of changes? Third, how do linguistic changes originate? Do they arise through language contact alone, or can they be internally motivated? Fourth, what is the relationship between internally (i.e., linguistically) motivated changes (which most of the ones we have studied seem to be) and the social factors that seem to trigger their spread? Finally, is there an overall direction to language change in vernacular dialects?
