A classroom study was designed to test the hypothesis that explicit knowledge is used by second-language (L2) learners in a free written response task if that knowledge is present. Eighty-one 12-18-year-old learners of Dutch as an L2 took part in a computerassisted language learning experiment receiving either explicit or implicit instruction about two grammar structures. The ability to use these structures was measured at three points in time by means of an untimed grammaticality judgment task and a free written response task. Explicit and implicit instruction promoted the use of the target structures in free response tasks equally effectively. However, for one structure, both facilitative and inhibitory effects of explicit instruction were observed if first language similarity was taken into consideration.Keywords explicit and implicit instruction; form-focused instruction; classroom research; L1 influence; explicit and implicit knowledge; second-language learning Within the second language acquisition (SLA) literature, form-focused instruction (FFI) research refers to the many studies that have investigated the effectiveness of different types of instruction in different circumstances of learning. R. Ellis (2001) defined FFI as "any planned or incidental instructional activityWe would like to thank Jan Hulstijn and Wander Lowie as well as the three anonymous Language Learning reviewers for their meticulous reading of earlier versions of this text and their valuable suggestions for improvement. All remaining issues are our own responsibility.

In the Netherlands, as in other Western European countries, many minority children are unsuccessful in their school careers. Their lack of achievement is associated with language and literacy problems. In secondary education they are faced with the task of text comprehension in the content areas, which is a reading task and a learning task at the same time.This study will report on the assessment of text comprehension in relation to vocabulary, grammar, non-verbal IQ and background knowledge about topics used in the text-comprehension test. Turkish and Dutch pupils took these tests during the first three years of their secondary schooling. In order to relate students' different reading styles to their linguistic background (L1 or L2), three different subtests of text comprehension were used, each referring to a certain discourse level.It was found that the L2-readers tend towards a top-down approach of reading, which implies a stronger reliance on conceptual guidance than in L1-reading. With respect to the childen's reading development there was a significant increase in their scores for Word Knowledge and Non Verbal IQ, but there was no progress in the scores for Text Comprehension. Especially, Turkish pupils in the lowest type of the Dutch secondary-school system showed no progress at all. RESUMEAux Pays-Bas, tout comme dans d'autres pays d'Europe de I'Ouest, beaucoup d'enfants d'immigrks ne rkussissent pas leur scolaritk. Dans le second degrk, ils sont sous-reprksentks dans les filikres qui permettent de poursuivre des etudes. Les Turcs et les Marocains, en particulier, sont sur-reprtsentts dans les voies professionnelles courtes, que l'on appelle Ibo/Lbo, ainsi que dans la partie inferieure des formations gknkrales, que l'on appelle Mavo. Leur faible rtussite va de pair avec un probl2me de langue t r b probablement lit A la maitrise de la lecture-kcriture.

Pupils of foreign origin need proficiency in both oral and literate skills to succeed in the Dutch secondary school system. In this article the development of literate skills in L-1 and L-2 is examined. The way in which a second language is learned is influenced by the level and type of mother tongue competence. In the case of older learners a higher proficiency in L-1 can facilitate the L-2-acquisition process. Especially language skills based upon cognitive development are expected to be transferred from L-1 to L-2 (Cummins, 1979,1980), which is the reason why reading was chosen as the object of this study. The study compares two groups of Turkish secondary school pupils, one group consisting of pupils who learned Dutch at a relatively early age, the other of pupils who had just arrived in the Nether-lands. The latter are expected to have a higher competence in Turkish than the former when starting to learn Dutch. For both groups a positive relation has been found between Turkish and Dutch for technical reading and reading comprehension. The findings suggest not only a positive transfer from L-1 to L-2 but also from L-2 to L-1 in the case of the group of pupils who had hardly been taught in L-1 and arrived in The Netherlands at an earlier age. Positive transfer seems to occur especially in reading comprehension provided that there is a certain level of competence in both languages.

The Dutch Text Comprehension Test was developed in order to help improve the education of minority students in the Netherlands. As it appeared that many second language (L2) readers have more difficulties with respect to the micro-processing of words and sentences while compensating on the macro level, there was a need for an instrument with diagnostic value so as to yield clues for further remedial teaching. The test proved reliable for both L1 students and L2 students. The test, with regard to diagnosing the type of reading problem, was validated by carrying out further observations by video. Although the classification results were not optimal, the test can be used for diagnostic purposes provided that there is a relatively large difference between sub-test scores on the micro and the macro levels.The results of a screening survey show that there are great differences in percentages of weak readers and distribution of reader types among the different secondary school types in the Netherlands. Unlike weak readers who attend the higher school types, weak readers at the lower educational school types ± among them many L2 readers ± are`problem readers' who do not compensate for their lack of word knowledge and sentence processing and who do not appear to be efficient`bottom-up readers.' Secondary schools are therefore encouraged to screen their first year students in order to gain insight into the problems that weaker readers ± including L2 readers ± may have, and to develop remedial teaching programmes tailored to the different needs of L1 and L2 students.

No abstract