Toegepaste Taalwetenschap in Artikelen - Volume 23, Issue 1, 1985

This article discusses the first results of a research project carried out in a number of first-year forms in Dutch schools for secondary education. By the side of the regular programme for French, this research project offered a light-weight extra listening programme, while for the sake of comparison other classes were offered a light-weight extra reading programme.After a few months in an interim test the pupils' reading ability, grammatical knowledge and vocabulary knowledge were evaluated.As a rough conclusion of this interim test it can be stated that the reading programme has yielded the most positive results at this level of teaching. The results of the listening-programme on the other hand are as yet hardly spectacular.At the end of the research project, which will be carried out during one whole school year, a final test will be given.

Naming and describing are important aspects of referential communication (RC). In an investigation of RC in a second language (L2), 36 secondary school students, divided into three groups of 12 with different proficiency levels, had to work in pairs on a description/identification task involving twelve nonrepresentative shapes. One task session consisted of six trials which were performed in English. The results of three different data analyses (numbers of words, naming and describing, and referential strategies) showed that RC in L2 differs from RC in L1 in three aspects: nonnative speakers (NNS) need longer references for the shapes; NNS' final references are structurally more complex; NNS with a lower L2 proficiency use their L1 more often than NNS with a higher L2 proficiency.

This paper first presents some general background information about Computer Assisted Language Learning (CALL), for the benefit of those readers who have little knowledge of the field. It then gives an extended example of an educational authoring system which is currently under development, describing how a teacher will use it to construct a lesson, and how a student will interact with the lessons that his teacher has constructed for him. The paper concludes with a list of some of the general design features incorporated in the authoring system.

This paper argues that a) much of what is presented as Krashen's new theory on foreign (or second) language acquisition is not new but given a new name (e.g. "the affective filter" for motivational and attitudinal variables) and b) that what is new is not adequately substantiated by empirical evidence or maintained in spite of 'contrary evidence. The plausibility of the new part of the theory is questioned on various grounds, one of them being that it does not sufficiently take into account the possible implications (i.e. for the way a foreign language is mastered) of the differences between the various foreign language acquisition contexts such as age, goal and aptitude of the learner, the kind and/or degree of L2 exposure, formal vs. naturalistic acquisition etc. An example of the kind of research needed is reported viz. the data resulting from an investigation into the errors against adverbial placement rules made by EFL students with different mother tongues. The aim of the study was to collect data relevant to the question whether/ to what extent errors would be "universal" (i.e. the same for all learners irrespective of their L1) or language (one) specific.Finnish, Polish, German, French and Dutch E.F.L. learners took 4 tests operationalising knowledge of English adverbial placement in 4 different ways: acceptability judgements, "scrambled" sentences, placing the same adverb in syntactically different contexts and placing different adverbs in syntactically identical contexts. Since the investigation was mainly concerned with error patterns especially the test performance of the low scoring subjects was further analysed to determine the degree of language specificity of their errors. This analysis strongly suggested either direct or indirect mother tongue influence.

An investigation of the interlanguage of Dutch learners of English revealed evidence of L1-L2 transfer and interference. 40 pairs of Dutch lexical items, in disambiguating context sentences, were rated on a 7-point scale of semantic similarity by 30 native speakers of Dutch. Next, the same pairs, in the same context sentences, were translated into English by a total of 90 Dutch students of English, subdivided into three proficiency levels. It was hypothesised that the similarity ratings could be used to predict translation performance: in the absence of knowledge, pairs that rated high on perceived similarity would be more likely to be translated by identical English tokens, whereas pairs with low perceived similarity were more likely to be translated by different English tokens. The results showed a strong correlation between perceived semantic similarity and observed translation errors, in the direction predicted by the hypothesis. There were also indications that the lower-proficiency informants were at a relative advantage where cognates were concerned, but that the more proficient informants were better equipped to deal with false friends.

In this article a model of foreign (L2) vocabulary learning is first developed in which the representation of the spelling, the pronunciation and the meaning(s) of a word are stored in their respective networks. Vocabulary learning in a paired associate learning task is then defined as the building of nodes in a network and as the establishing of an associative pathway between each new node (representing the spelling, pronunciation or meaning(s) of a newly learned L2 word) and the corresponding node for its L1 equivalent. In this model differences in spelling or pronunciation between L2 words and their L1 translations are expressed in terms of differences in length of their associative pathways.On the basis of the above distinctions a prediction was made about a difference in input speed and in the period of retrievability between two kinds of Arabic- Dutch word pairs. It was predicted that word pairs sharing some phonemic features would be learned sooner and remembered longer than those without any of these features in common.The above prediction was confirmed in a group experiment. Moreover, it showed that a resemblance between L2-L1 word pairs had a greater effect on retrievability than on input speed. Finally, the diverging results for one of the testwords were interpreted as indicating that also in associate learning of L2-L1 pairs the semantic category to which that word belonged had played its part. Its abstract meaning was assumed to have negatively affected the time needed to store the word in memory.

This paper deals with the procedures followed to identify the strategies in the Nijmegen project on the use of compensatory strategies by Dutch learners of English. The project aims to shed light on the relationship between foreign language proficiency level and compensatory strategy (CpS) use. For this purpose three groups of fifteen Dutch learners of English, of three different proficiency levels, were asked to carry out four tasks. Two of the tasks comprised rather formal elicitation techniques; the subjects had to describe concrete and abstract pictures. The two other tasks, a story retell task and an oral interview, were aimed at obtaining more natural language data.To identify the CpS we make use of problem indicators such as pauses, repetitions, intonation, sighs, laughs and ongoing commentary. In addition, it was decided to collect retrospective data from the subjects on the two more natural tasks, as in these tasks CpS are particularly hard to identify, due to the relative inaccessibility of the subject's intended meaning.In this paper criticism of the use of retrospective data is discussed, as well as the conditions (as specified by Ericsson & Simon, 1984) under which retrospective data can be reliably collected. A description of our procedures goes to show that, on the whole, these conditions are met in our case. The usefulness of retrospective data is then illustrated by means of a comparative study of two identification procedures, one with and one without use of retrospective data. The results indicate that the additional use of retrospective data increased the number of identifiable CpS by 33%; raised the degree of confidence which we can put in our identification in 64% of the cases; and enabled us to discover a small number of incorrectly identified CpS. It was therefore concluded that retrospective data can be considered a reliable and valuable resource in the analysis of CpS.

This paper deals with the loss of FL-vocabulary in French and German learned in formal secondary Dutch education ("VWO). Using a cross-sectional design, we investigated the development of FL-vocabulary from the end of grade 4 - the moment at which pupils can drop one or two FLs - up to the end of grade 6 - the moment when they take their final exams in the seven subjects they did choose. We compared pupils from grades 4, 5, and 6 who had either dropped the FL, or had chosen it as an exam subject. A French and a German translation task were administered in which pupils had to fill in the target word in a meaningful, but noncoercive context. The tests consisted of a productive part (translation into the FL) and a receptive part (translation from the FL). Either part consisted of 48 equivalent items, 24 high-frequent and 24 low-frequent target words.The main finding was that there was a loss of 16% for French in the second year of non-use, and a comparable loss for German in the first year of non-use. However, the loss for German was entirely redressed in the second year of non-use. It was concluded that this may have been due to the cross-sectional design, and that, therefore, longitudinal designs are to be preferred in this kind of research.

On the basis of a statistical analysis of the scores of 50 Dutch highschool students on a purpose-made test of English grammar, this article tries to establish a relationship between types of grammatical rules and grammatical subsystems on the one hand and achievement in English grammar on the other. It arrives at the conclusion that syntactic rules are significantly more difficult than both morphosyntactic and morphological rules and that open grammatical systems are significantly more difficult than mixed open/closed systems. In a concluding section recommendations are made as to how the findings of this investigation can be turned to good pedagogical account.