Toegepaste Taalwetenschap in Artikelen - Volume 64, Issue 1, 2000

This paper presents the results of a recently completed longitudinal evaluation study of Dutch-English bilingual programmes in Dutch pre-university education. During several years the outcomes of these programmes were measured in order to find out whether pupils in bilingual education gain higher levels of foreign language proficiency without any detrimental effects on their mother tongue proficiency and their achievements in other school subjects. Beside this, lessons were observed in order to investigate to what extent language teachers and teachers of other subjects display behaviour that can be said to stimulate second language acquisition.

In Flemish secondary schools, non-indigenous pupils underachieve. This is pardy due to the way teachers use Dutch as a medium of instruction. In the past, various pedagogical solutions have been put forward to address this problem. However, none of these appears to be successful, basically because they leave the 'instruction paradigm' unchallenged. In this article, the perspective of'powerful learning environments' is suggested as a possible way out. Rather than optimizing the way teachers use language to give instruction (i.e. to pass on abstract knowledge), teachers should develop the proficiency to guide learners while engaging in active confrontations with the world and building up knowledge themselves. Powerful learning environments may be visualised as consisting of three concentric circles: a positive and safe learning climate (outer circle), a variety of meaningful and challenging problem tasks (mid circle) and interactive support by the teacher or other learners (inner circle).

This article describes the basic principles of interactive language teaching, as currently developed by the National Centre for Language Education (Expertisecentrum Nederlands). Based on a social constructivist view on learning and development, we argue that language education in the primary grades should aim at social, authentic, and strategic learning. We show how these principles are applied to the teaching o f oral/communicative language skills and vocabulary acquisition. A framework for literacy standards, and its consequences for assessment is also discussed.

This paper describes the dictoglos procedure in which learners try to reconstruct in small groups a text that has been read to them by the teacher. The procedure can be used for all kinds of learners irrespective of L1, L2 or acquisition level. The aim, theoretical assumptions and advantages of the procedure are discussed. In the second part of the paper, criteria for texts are mentioned that can be used for dictoglos. These criteria concern the contents, subject, genre and form of the text. Two examples of texts illustrate the use of these criteria.

The Common European Framework of Reference (CEFR) is supposed to be applicable for all learners or users of second and foreign languages and to enhance the comparability of language qualifications. In this article, it is argued that, owing to a confusion of declarative knowledge and language proficiency, this framework does not enable language learners with a low level of previous education are not enabled to show progress in language proficiency.Furthermore it is argued that the open and flexible nature of the framework does not contribute to a higher degree of comparability of results of language learning.

In the vocational education of Dutch as a Second Language (DSL) there is a growing need for oral proficiency tests that are efficient to use. In this article, we describe ways to make use of speech technology for rating speech quality. In two experiments, the rating of DSL-speech is compared to automated rating of several aspects of the acoustic signal. The correlations found between human and automatic ratings were relatively high. For that reason a third experiment has been set up using speech items that are representative of commonly used oral proficiency tests in DSL and that can be rated automatically by a Continuous Speech Recogniser system. The results obtained so far are discussed.

The assessment of text difficulty has played an important role in the development of the 'Elektronische tekstbegriptoets voor de brugklas', a tool for screening students in the first year of secondary education on their text comprehension skills. On the one hand, the investigation was intended to enable a well-considered selection of texts as a basis for the comprehension test. On the other hand, it provided the criteria needed to develop a reliable ability scale. A corpus was built of over a hundred texts taken from schoolbooks of different levels. The texts were analysed on three different linguistic properties: mean length of words, mean length of sentences, and coverage (percentage of words belonging to the 2000 most basic Dutch words). It was found that especially coverage is a strong indicator of text difficulty.

Most measures of lexical richness in spontaneous speech data, based on the distribution of, or the relation between the types and tokens, appear to be neither reliable nor valid. The article describes a semi-automatic computer program, MLR (Measure of Lexical Richness) that measures lexical richness on the basis of the degree of difficulty of the words used, as measured by their (levels of) frequency in daily language input. The MLR is meant for the analysis of texts of (students in) primary education, with a vocabulary size of up to about 25,000 different lemmas, and provides an answer to the following questions: 1) What is the difficulty of the various words in the text? 2) What is the relative proportion of the degrees of difficulty of words in the text? 3) What is the covering percentage of the text for a student with a certain vocabulary size? 4) What is the size of vocabulary of the student, on the basis of the spontaneous speech data?

In this article, we report on the construction of a test of deep word knowledge in French as a second language that can be used at advanced levels of proficiency.By 'deep word knowledge' we mean the knowledge of the semantic network of L2-words; our test focuses on the knowledge of the relations between L2-words and the words semantically related to each of these words; part of this is the knowledge of different meanings of polysemous words.In our test, each item consists of an L2-word, i.e. the item word, accompanied by six other L2-words that have or do not have a semantic relation with the item word. The testee has to mark the semantically related words. These relations may be paradigmatic (i.e. cases of synonymy, antonymy, or hyponymy) or syntagmatic (i.e. collocations, ?art-whole relations, or defining terms).The test was pretested on students of different university levels, i.e. advanced and near native students, and also on several reference groups of real native students. The results of the pretest showed the fairly good discriminatingpower of the test: differences between the mean scores of the groups of advanced, near native and native students proved to be significant. The test even discriminates between different levels of native proficiency (the french equivalents of ??, ?A and highschool students). The reliability (Cronbachs a) was estimated separately for the groups of advanced and near native students and proved satisfactory: .84 and .88 respectively.The data of the pretest administration was used in an item analysis procedure to modify and improve the test and to give it its final form.In the near future we will investigate the relationship between the deep word knowledge our test measures and scores on tests of broad, quantitative word knowledge; furthermore we will investigate the power of these two kinds of vocabulary scores to predict integrated language skills, in particular writing ability.

The Yes/No vocabulary test seems to be a promising tool for measuring the receptive vocabulary size of foreign language learners. The test consists of items that are either words or pseudowords and the subjects have to indicate whether or not they know the meaning of these items.This article reports on two experiments in which the Yes/No test was used with French speaking learners of Dutch. In a first experiment, the format was used as a placement test in order to measure the extent to which the subjects had mastered the basic vocabulary of Dutch. In a second experiment, the test was computerised and followed by a translation task in order to obtain a concurrent validation. Both experiments were characterised by a rather low reliability and the presence of a bias in the data. The second experiment also revealed an unconvincing correlation between the results obtained on the Yes/No test and those obtained on the translation test. The accurateness of the Yes/No vocabulary test for Dutch is thus questioned and suggestions made for further research in order to improve this test format.

On the basis of the results in spontaneous second language acquisition research it is argued that, in second language teaching, the learner's first language is used to a lesser extent than is useful for the adult learner who has no other language learning experience than learning his mother tongue. Examples are given of learners, who after having been exposed to Dutch for three years, still construct their L2 expressions on the basis of the grammatical structure of their L1. The problem is that (1) teachers do not recognize these L2 expressions as such, (ii) the course material is made from the perspective of the L2 and does not take into account the specific problems of a language group, and (iii) the instruction in the classroom is in a language that is insufficiently understood.

In this paper, we argue that variation in child bilingualism has received too little attention. The literature on monolingual acquisition shows that children vary in the routes they follow towards linguistic proficiency: children have preferences for the acquisition of certain word classes (Nelson 1973, Bates 1994) and they do not all follow the same route in developing more word utterances (Clark, 1993). This variation is hardly taken into account in the study of bilingual acquisition. The fact that children vary in the amount of language mixing in their utterances may be due to the input they are exposed to, but may just as well be a sign of individual preferences and choices. In the case of the French/Dutch bilingual child we study, the first stage of her utterances seems to show that the amount of language specific utterances is variable but decreases with time. We suggest that the child is late in developing the pragmatic competence necessary to develop two differentiated language systems.

What makes it possible for learners of English as a second language to learn how to spell this complex system, without any instruction about phoneme-grapheme correspondences or rules? The author shows that a graphotactic description of English orthography reveals many systematic visual patterns. L2 learners appear to be sensitive to these patterns. This may explain how they acquire the spelling system: by picking up visual patterns.

In studies of the moral development of children the focus obviously is on the interaction between children. However, research in which the language involved is systematically explored is scarce and fragmented. The main aim of the project is to fill this gap. The project analyses whether the model of Brown & Levinson (1987) of politeness strategies can be used as a broader framework. In this article the focus is on one component of this model relating to the language of togetherness. The research is a longitudinal case study of one child in his interactions in a day-care centre and an after school centre. The data in this report concern the whole group of children in the first and the last recordings.