Abstract
The research program Maps and Grammar investigates what the geographic distribution of linguistic features can reveal about the mental representation of grammar. It does so by developing advanced tools for statistical analysis and visualisation and applying these to large dialect databases (project 1). The linguistic areas uncovered and defined by the application of these methods are then analysed from the perspective of modern theoretical linguistics, with the primary aim to test the structuralist/generativist hypothesis that a dialect is not a collection of individual linguistic features but a grammatical system consisting of a set of coherent, interacting grammatical rules or constraints. This hypothesis will be tested in three projects. Project 2 investigates geographic inclusion patterns correlating with gradual grammaticality judgements. Project 3 looks at transition zones, linguistically unstable and heterogeneous areas in between more homogeneous dialect areas that show peculiar linguistic features and optionality. The goal of project 4 is to develop grammar fragments, visualise these on maps and compare the resulting maps with maps based on individual linguistic features rather than grammar fragments. The insights resulting from this program are not only relevant for linguistics but also for other geographically-oriented disciplines such as political geography and for language policy.