Toegepaste Taalwetenschap in Artikelen - Volume 54, Issue 1, 1996

In this article a description is given of the language production process based on Levelťs 'Speaking'-model. Using this type of model will help us to understand better the subprocesses in language production and the kind of knowledge and procedures needed to produce language. A full definition of what knowledge is needed will also help us to model the different stages a language learner goes through.In the second part of this article recent developments with respect to the role of output in SLA are discussed from the perspective of the model presented. It is concluded that in language acquisition both input and output are needed and that in fact output may have as an important function the generation of very specific input that will support the acquisitional process.

New data on Turkish-Dutch codeswitching (CS) have uncovered patterns not previously attested in this language pair. It turns out that three different generations within the immigrant community switch in three different ways. For the first generation, CS mostly concerns the occasional insertion of Dutch nouns in Turkish sentences. The intermediate generation switches much more frequently and in a much more varied way. The second generation also practices a lot of CS, but the type of switching is virtually always of an alternational, not insertional, nature. Insertional CS further seems to lend support to a model of speech production in which some elements are uttered with greater awareness than others. High awareness stimulates CS. Various aspects of linguistic elements, as well as of the act of speaking, are involved in determining an element's level of awareness.

A story teller may tell a story by simply putting together subsequent events and situations. However, he may also point forward to things that are about to come in the story (foreshadowing), or point backward to reveal the relevance of what is said here to what has been said before, often in the form of wrapping things up.In my article I will first discuss the question how children foreshadow and wrap up information in two different story telling tasks: a story retelling of a picture story and a personal story, about events that the child has experienced itself. Secondly, I will address the question whether the amount of foreshadowing and wrapping up in the two story tasks is comparable.>In the story retellings, foreshadowing and wrapping up were established by a global framework at the beginning of the story, local and global motivating states in the beginning and middle of the story and quotes, parallels and contrasts at the end of the story. In the personal stories abstracts, settings, motivating states and codas were used as devices to foreshadow and wrap up events and situations in the narrative.The amount of foreshadowing and wrapping up in the story retellings and personal stories turned out to be not comparable.

Do men and women make use of different strategies in communication? According to the theory of gender subcultures of Maltz and Borker, differences refer to relational orientation in women and task-orientedness in men (affiliation vs control). Verbatim transcripts of 85 medical interviews were analysed, using Stiles' Verbal Response Mode coding systemm (VRM), and notions of topic shift. In agreement with the hypothesis male doctors were more imposing and presump-tuous (more often giving advice and interpretations). Female doctors were more attentive and non-directive (giving more subjective and objective information and backchannels). The interviews of female doctors took more time than the interviews of their male colleagues. The topic shifts were more coherent among the female doctors.The results support the idea of the existence, to a certain degree, of different subcultures for men and women.

Male/female communication training offered by enterprises to their company executives starts mostly from common sense ideas about sex-based differences in language use. The corresponding guidelines for women as newcomers in the public sphere as well as for men, often derived from one or more out of three well known theories, are shown to necessarily lead to deceptions because of their gender blindness.A gender approach to differences in language use between men and women cannot produce direct guidelines for language users but may help them to benefit from the fact that all language users, newcomers included, are norm subjects in interaction.

One way of internationalizing education is introducing an international language - such as English - as the medium of instruction. But for Dutch lecturers English is a foreign language, of which they problably have a less adequate command. This may raise the issue whether the quality of their instruction is affected by a change of instructional language.Research conducted at Delft University of Technology reveals that Dutch lecturers experience language-related limitations because of the switch from Dutch to English. These limitations pertain to their vocabulary, the redundancy of their subject matter presentations, and their clarity and accuracy of expression. Other findings reveal a decrease in lecturers' speech rate, their expressiveness and their improvising skills. Lecturers' reduced use of these teaching behaviours may decrease student learning.Developing teaching-performance-based language tests and language courses focusing on 'teaching in English' are one way of reducing this potential loss of educational quality.

For the purpose of developing a curriculum for the teaching of speaking and listening, seven case studies have been conducted into the practice of teaching presentation, debate or discussion. This article addresses two of the problems encountered in the course of this research.The first problem concerns the fact that teachers do not succeed in observing and afterwards discussing the speaking and listening assignments performed in such a way that this, too, contributes to the pupils' learning process. A solution is offered by way of the approach of learning through the observation of model, in which the the primary aim of the observer is to enable pupils to improve their own performances.The second problem is that in teaching speaking and listening skills, teachers do not provide special treatment for those pupils whose first language is not Dutch. In this article it is argued that it would help these pupils if they were offered 'language devices' which would reduce the pressure on their working memory, thus increasing their fluency. An example is given of the survey of language devices developed for presentation and discussion, as well as reactions of teachers to these devices.

It is assumed that language education can gain considerable profit when more attention is given to the process of language learning and processing. In other words, it is not enough to look at what pupils are doing, but also at how they are doing it. More attention to the process means a shift towards a strategic approach of language education which tries to enlarge the strategic competence of students: they must become aware of language strategies (directed at the communication process) and learning strategies (directed at the learning process) which can be applied for problem solving. Within such a strategic approach proper exercises must comply with at least three important features: an educational goal in which the function of the task is indicated, assignments for monitoring the problem solving process, and explicit criteria for evaluating task output. In this article we report the results of an analysis of exercises -in particular exercises for the oral skills- in Dutch textbooks for mother tongue and foreign language education pointed to the presence of these relevant features.

For optimal second language acquisition (SLA) classroom interactions have to allow normative speaking children to participate actively, because such initiative-rich participation enhances the children's development of linguistic and communicative competence. This article focuses on turn-taking initiatives. Turn-taking initiatives vary in their initiative strength. Restricting analysis exclusively to strong initiatives yields an incomplete picture of the contribution of a conversation to SLA. Less strong initiatives, too, must be taken into account. To analyze each utterance for its initiative value, a turn-taking taxonomy was developed, leading to turn-taking indices that represent the mean initiative value of all utterances by a speaker. Language data from kindergarten conversations show that only results on strong initiatives differ from results on turn-taking indices. It is argued that in order to assess the value of a conversation for SLA both kinds of measures, strong initiatives as well as indices, are necessary.

In this article the 'Dinocroc method', the rationale behind the method and the principles on which the method has been constructed are described in some detail and a clear picture is given of how the method is used in the classroom. With this method a real effort is made to integrate natural first language acquisition processes into a second language teaching method.An explanation is given of the research set-up concerning the effect study of the 'Dinocroc method' when compared to more regular second language teaching methods. At this stage of the research it is not yet possible to provide the reader with the conclusions of this study, only a very general and cautious remark can be made about the effects of the method.

If anyone suggested ten years ago that computers could play an important part in the teaching of productive oral skills or pronunciation, most listeners would have reacted with amusement or disbelief. Computers were business machines capable of processing text and data efficiently and no more than that. As a result of recent technological developments computers are now capable of much more; technologically the manipulation of sound and data is no longer a problem. Unfortunately the development of software and courseware has not kept pace with the technological innovation. Nevertheless the use of computers for the training of oral skills and pronunciation (through automatic feedback on the student's utterances) lies already within the reach of 'ordinary' students and teachers.

Recently new objectives have been defined for foreign languages in Dutch upper secondary education. These objectives have served as a starting point for curriculum development at SLO (Dutch National Institute for Curriculum Develop-ment). Within the framework of a task-based syllabus as defined by Breen (1987) a communicative activities programme, mainly focussing on speaking, has been designed by SLO and tried out at several schools. Four criteria for grading and sequencing activities were used: level of control in activities, familiarity with theme, linguistic complexity, and complexity of classroom management.The development of tests on the basis of these new objectives will be a challenging task. Care must be taken that such tests are valid, i.e. that they contain a fair representation of the objectives. CITO (Dutch National Institute for Educational Measurement) has called for a refinement of the criteria on which the assessment must be based.

In 1988 CITO started research into the feasibility of valid and reliable oral proficiency tests based on communicative principles. This was to meet the demand for a communicative speech test to be used in school based examinations in secondary education. Using the Test of Spoken English as a guideline, tests for French, German and English were developed. Simultaneous research into the reliability and validity of the tests led to various adaptations of the original model. From 1992 onwards oral proficiency tests for each of the three languages in question have been published at levels VBO/MAVO, HAVO and VWO (approxi-mately: vocational, secondary modern and grammar school). The results of a user inquiry held in 1994 led to a number of further changes to improve the user-friend-liness of the tests. Early in 1996 a new research project concerning the reliability and validity of the tests was started. The results will be published in the autumn of 1996.

Rating of speaking ability in educational contexts is often confronted with the problem of selecting the most relevant criteria. Researchers as well as teachers are forced to design a rating scheme for each setting on an ad hoc basis. The most important characteristics of the setting are briefly discussed: purpose of rating, theoretical relevance of criteria, task definition, level of speaking ability and raters' working method. I will give examples of some rating schemes that were designed in different settings, including native, second language and foreign language speakers, in order to demonstrate the problems of selecting and defining rating criteria. Two perspectives for defining criteria are being distinguished: 1) a behavioral (using aspects of speaking behavior as basis), 2) a functional (using functions that speakers must fulfill as basis). Subsequently, I will give some suggestions for a theoretically founded scheme that can be used as a basis for selecting and defining rating criteria for several kinds of setting. The scheme is restricted to situations in which communication is the main purpose and speakers are evaluated on an individual basis. In these settings a functional perspective is regarded as more valid and efficient than a behavioral perspective in defining rating criteria. In conclusion an overview is presented of four rating dimensions from a functional point of view.