Search

This paper deals with a dramatic change in the presentation of linguistic examples in the linguistics literature of the twentieth century, a change coinciding (not accidentally) with the introduction of transformational generative grammar (TGG) in the 1950s. Our investigation of this change and the circumstances that gave rise to it leads us to reconsider the question of continuity and discontinuity in the history of linguistics in this crucial period, focusing on the question of the audience to which the earliest publications in TGG were directed. We argue that this was an audience of nonlinguists (information theorists and mathematical logicians) and that transformations were introduced by Chomsky, the founder of TGG, as a way to preserve, within the new paradigm of formal grammar theory, established insights of a purely linguistic nature.

The Syntactic Atlas of Dutch Dialects (SAND) is a database of syntactic features observed in the language spoken by people from different dialect regions in The Netherlands and Flanders. We would like to know how specific syntactic features are for the different dialects. For this purpose we try to generate dialect boundaries from the syntactic data only. We show that a plausible binary division of the dialect can successfully be derived but that is more difficult to divide the data in three or more regions. We build on earlier work by Nerbonne, Heeringa en Kleiweg (1999), who performed this task for phonetic data, and on work by Spruit (2008), who also attempted to identify dialect regions from syntactic data.

Although sentence final verbal clusters in dialects of Dutch demonstrate a large amount of variation in the order of verbs, we argue that this is only apparently so. We take each dialect to allow just one order of verbs in three-verb clusters with a past participle. In the north of the Dutch language area, the order is descending (V3-V2-V1) and the rest of the dialects show an ascending order (V1-V2-V3). The large amount of apparent counterexamples will be explained by independently motivated, interfering properties. First, participles might be V-type or A-type. Only V-type participles occur in V-positions in the verbal cluster. Secondly, non-verbal elements (such as A-type participles) may interrupt a verbal cluster. We will show that the distribution of the different orders in dialects of Dutch strongly supports such a restrictive approach. We thus take this to be an argument that a structural approach to dialectology is required to gain insight in the properties of the formation of verbal clusters in Dutch.