Wortstellungsvariation im Deutschen: Psycholinguistische Untersuchungen zur Relativsatzposition

Word order variation is a typical phenomenon in German language. Especially the extraposition of relative clauses is frequently used. In this case the relative clause does not follow directly the noun that it specifies but is located after the verb at the end of the clause. It is a grammatically correct process in German but what makes the decision for or against extraposition? The most discussed factors in the past concerned rather the grammatical structure as the length of relative clause or the distance of extraposition, or the information structure as the focus of the noun or the relative clause. But empirical evidence mostly derives from offline measures as corpus analyses or acceptability judgements. Online studies, that can show the influence of the factors during the processing of the critical structure do almost not exist or give rise to different results and interpretations. This is where the present thesis starts. In three ERP-studies I investigated the influence and the interaction of the most frequently discussed factors relative clause length, extraposition distance and focus on the processing of German relative clauses in their different positions and completed the list of possible factors by another one, namely the grammatical function of the head noun. The first part of the thesis gives an extensive overview of the grammatical description of extraposed relative clauses and the psycholinguistic discussion of influence factors. It follows a detailed method chapter with the illustration of EEG-Data analysis by frequency based analyses and event-related potentials (ERP) and a review of previous relevant studies. The second part of the thesis finally describes the three ERP-experiments of the present study, including the acceptability judgements of these experiments and two additional behavioural experiments. Beginning with the acceptability data, previous studies could been completed with new findings: Concerning the well-known factors relative clause length and extraposition distance the grammatical function of the head noun also playes a significant role. If the relative clause refers to the subject it is the extraposition distance that critically influences the acceptability judgements. With a relative clause related to the object the influence of the extraposition distance decreases. The crucial factor then is the relation between distance and relative clause length. Further evidence came from the EEG-Data that gave a more concrete impression: First of all there was a clear distance effect for the long extraposition distance that arose for both subject and object head noun as a N400 located at the relative pronoun and the relative clause verb. The difference between the grammatical function of the head noun only appeared in a late positivity that additionally showed up at the relative clause verb for a long distance to the subject head noun. Also only for a subject head noun an influence of information structure on extraposed relative clauses could be indicated in the experimental design of the present study. But this influence was only existent as long as no other, stronger information structural influences appeared as in case of object scrambling. An effect of information structure was shown in an early N400 on the relative clause verb for an unfocused, given subject whereas a focused new subject let disappear the late positivity on the relative clause verb in the long extraposition distance condition. As a functional explanation of the observed negativities I proposed difficulties in relative clause integration at the moments of retrieval of the head noun whose activation level is reduced by the long extraposition distance or for an unfocused, given noun, that is therefore less salient. This influence of extraposition distance and information structure is initially independent. Only at the evaluation of the whole structure with the integrated relative clause all influence factors interact including the grammatical function of the head noun. This is shown in the differential appearance of the P600 according to the subject head noun. The studies show that the variation of relative clause position in German depends on more than one factor alone. It is the result of an interaction of several factors.