Verwerving Van Woordvolgorderegels In Het Nederlands Als Tweede Taal

This contribution discusses the research questions and preliminary results of a PhD-project carried out at the Free University of Amsterdam. The aim of the project is to gain more insight into the ways in which Turkish and Moroccan immigrants in the Netherlands learn word-order rules of Dutch. The main differen-ce between the languages spoken by the two groups of learners is that Turkish, like Dutch, is an SOV-language, whereas the Moroccan languages (Berber and Moroccan Arabic) are VSO/SVO languages. The central question to be answered is whether all learners, regardless of their mother tongue, go through three acquisitional stages, in which certain word-order rules are acquired as part of a cluster. The idea is that learners build an interlanguage system on the basis of structures they can process; once certain rules are acquired, restructuring of the interlanguage system takes place. This, in turn, may lead to the acquisition of another cluster of rules.In order to be able to answer the research questions, we decided to analyse spontaneous speech samples and to elicite lacking information by means of four experiments. Once the second part will be completed, we hope to be able to decide whether rules are learned as a cluster, which would indicate that restructuring of the interlanguage system has taken place. The first part of the project is mainly aimed at individual language development; this part has not yet been completed. The analysis of spontaneous speech samples was made possible by the Max-Planck Institut für Psycholinguistik in Nijmegen, which made available data of the European Science Foundation Second Language Databank; these data consisted of transcripts of speech samples of 4 Turks and 4 Moroccans, collected over a period of 30 months.The first part of the paper presents some of the theoretical notions employed in second language research, as well as the main rules of Dutch that have to be acquired. Also, the three languages involved in the project are compared. In the next sections the research questions, the design, the data analysis and some interim results are discussed. The main conclusion is that there are indications that certain rules are indeed clustered together, although the different stages are not strictly separated.