Clause union and verb raising phenomena in German

Abstract

In this paper we discuss a class of constructions in German syntax which have been known as coherent infinitive, clause union or verb raising constructions. These data run against the predictions of strictly configurational theories by apparently having a syntactic structure where the subcategorization frames of two or more verbal heads are merged into one. Thus, in addition to a fully bi-clausal structure with two clearly separated verbal heads, we also have to envisage the case where a verb is apparently raised from an embedded to form a verb cluster together with its governing verb, while the sets of their arguments are merged into a single set, representing the case of clause union. In addition, there are constructions where there is no evidence for clause union, but where one could nevertheless argue for the formation of a verb cluster. We investigate these data by looking at a series of constructions which bear evidence on the issue. Among these are extraposition, which appears a reliable test for nonobligatory verb raising; subjectless constructions, which are possible only as the complements of so-called raising verbs but not of control verbs; S-pronominalization, which seems to be limited to equi-verbs; scrambling and long reflexivization, which we can take as evidence for clause union; the scope of adjuncts and negation which argues in favour of verb raising, but does not necessarily presuppose clause union; and finally certain topicalization phenomena which appear to violate almost any of the generalizations set up so far by configurational theories.