Understanding PP placement in written Dutch - A corpus-based multifactorial investigation of the principal syntactic, semantic and discursive determinants

Abstract

The present paper aims at refining current knowledge about the driving forces in Dutch constituent ordering of non-predicate prepositional phrases (PPs) as well as re-evaluating the common assumption in traditional Dutch reference grammars that the middle field position is the standard slot. Building on journalistic data in the Dutch Parallel Corpus, it is first shown that non-predicate PPs are significantly more often placed in postfield position (the structural position after the final verb cluster) than in middle field position (the position before the final verb cluster), which indicates that the postfield position rather than the middle field position should be considered the standard slot for PPs in written Dutch. Second, a binary logistic regression model was fitted with PP placement as a function of several syntactic, semantic and discursive predictor variables. This model is able to describe, explain and predict more than 80% of the variation in the data set, leading to a much better understanding of the mechanisms underlying PP placement in written journalistic Dutch. On the basis of our findings, we furthermore propose a refined and more fine-grained version of the theoretical framework in which PP placement in Dutch is traditionally described and understood.