Tussentaalregels, Strategieēn en Taalproductieraechanismen

For a correct understanding of the causes of errors made in the interlanguage (IL₂) system by a L₂ student, a knowledge of (contrastive) linguistics and language (learning) psychology is essential. Error analysis as an a posteriori approach of IL₂ has only a heuristic function. IL₂ expressions (correct and incorrect) can be viewed as the product of either IL₂ rules or IL₂ strategies. The application of IL₂ rules is an unconscious or automatic process; the use of strategies, conversely, is a conscious or controlled process, aimed at solving as efficiently as possible problems caused by TL₂ deficiences, and experienced as such. An explanation can be seen as a cause only when the alternative explanations have been disqualified by means of hypotheses-testing research. Since IL₂ errors are the result of a IL₂ production process, language production mechanisms should be taken into account in the explanation of IL₂ expressions as well. In the German (L₂) of Dutch and American students, errors in case marking occur often in passive sentences. A few examples: (1) Es wird bei ihnen einen Fall erwähnt, wo es kaum anzunehmen wäre. (2) Dies geschieht indem einen Glottisschlag gebildet wird. (3) Er werde ... wie einen Hirsch oder als Hirsch erlegt. There are three possible explanations (hypotheses) for these errors. They have been tested experimentally as rule-hypotheses. From this it became apparent that the case errors in the IL₂ expressions (1), (2) and (3) have to be explained on the basis of a language production mechanism which Kempen (1977) calls 'partial parallelism'. The errors in case marking - which appear to occur particularly often at the end of passive sentences - can only be explained when one assumes that case marking of the actual subject-NP at the end of the passive sentence is determined on the basis of the semantic relationship between this NP and the main verb. That the passive form of the sentence here has less influence on case marking is the result of the fact that the process of case marking on this NP, and the realisation of the passive form, do not run parallel. That there are no serious problems at the beginning of the passive sentence can be explained by the fact that subject selection and case marking can already take place before the choice of a predicate is determined. Here, the subject-NP determines (the form of) the predicate, while at the end of the passive sentence it is the semantic relationship NP-main verb which ensures that an accusative case is conferred on the actual subject-NP.