Wie oft schreibt man das zusammen? The Puzzle of Why some Separable Verbs in German are More Separable than Others

In this work we address the question why different German particle verbs tend to occur with different frequency proportions in syntactically separated vs. non-separated forms. The problem has been studied from a theoretical point of view and the syntactic conditions that determine particle verb realization in separated and nonseparated paradigms are quite clear. But, to the best of our knowledge, the question of why there is a variation among particle verbs with respect to how often they appear in different paradigms has never been addressed empirically so far. In this paper we present a corpus-based study which tackles this question. We formulate various morphological, semantic and pragmatic hypotheses which might explain the variation and we test them with clustering and linear regression techniques.