Methodologische en Praktische Moeilijkheden bij Het Onderzoek Naar De Verwerving Van Het Nederlands Door Volwassen Buitenlanders

In the first and second section of this paper some methodological topics in the study of interlanguage (IL) and error analysis are discussed. The author adheres to Adjemian's point of view (1976): any interlanguage is a natural language and contains a linguistic system. Its systematicity is therefore given a priori. However, not all utterances of an IL speaker might be consistent: some rules of the system might be unstable over a certain period of time. Research would have to show what performance strategies (like simplification) influence the stability of the system, in other words in what way the consistency of the system is permeable.In the third section a preliminary study is presented dealing with the topic of inconsistency in IL. The research question was: to what extent can the inconsistency in the learners' utterances be accounted for by distinguishing different kinds of performance? To nineteen adult foreigners in Holland, with 15 different mother tongues, all partici-pating in an intensive Dutch course for beginners, eight tests ^licitation procedures) were administered: two reproduction tests, four production tests and two "intuition tests" (see table I). Eight rules of Dutch syntax and morfology were tested, (table II). Results show that:(a) if someone makes one or more errors with a particular rule in a reproduction test, he will make one or more errors with the same rule in a production test and an intuition test as well: the reverse is not true;(b) if someone makes one or more errors with a particular rule in a production test, he will make one or more errors with the same rule in an intuition test as well; the reverse is not true;(c) the overlap in errors/correct forms was greatest with both repro-duction tests. Of the production tests translation and the anagram test were almost equally representative for the four tests. Finally, there was no congruence between both "intuition tests", the sentence correc-tion test yielding consistently more errors (for all eight rules) than the multiple choice test.