Identifying and processing topicalization in Danish

Laburpena

In Danish there is a surprising asymmetry between pronouns and full DPs: pronouns can always be fronted, whereas full DPs are subjugated to more restrictions. This puzzle has not been addressed in the literature, and I argue that the split-CP analysis as presented by Rizzi (1997) cannot distinguish between pronouns and full DPs when they are marked with a topic feature. The analysis I propose is based on the claim that in Danish the purpose of fronting a constituent is to mark it as the main topic, rather than the need or necessity to check a certain feature. It is furthermore based on the observation that the presence of an overt case marked pronoun, an adverb or a modal/auxiliary enables topicalization. In order to account for the data I propose a (PF) processing constraint on dislocation, which ensures that the derivation has an unambiguous interpretation.