Langeafstandsverplaatsing in het Nederlands, Engels en Duits: De sandwich ontleed

In this article, we present corpus data from Dutch and English on long-distance movement and discuss its diachronic development in Dutch, English and German. Long-distance movement is the displacement phenomenon characterized by the appearance of a part of a dependent clause in a higher clause (e.g. What crimes did the FBI discover he had committed?). It has played a central role within generative grammar over the past few decades. The picture that emerges is that long-distance movement appears to be currently most productive in English and least productive in German, whereas Dutch occupies an in-between position. As we will argue, the productivity of long-distance movement is strongly tied to the availability of functional alternatives. German has at least three of such alternatives that are fully productive, whereas Dutch has one particularly productive one. The alternative constructions do not involve long-distance movement: the dependency between the constituent in the matrix clause and the position in the embedded clause where it is interpreted is formed indirectly, in the semantics, and not via syntactic movement. In English, long-distance movement is most productive when the complementizer is deleted. This is not just the case for subject movement but also for non-subject movement. Special attention is paid to the so-called that-trace effect and its alleged absence in German and Dutch. The general conclusion is that long-distance movement is possible in all languages under consideration, but more restricted than commonly assumed.