The Horwood Language Centre at the University of Melbourne has recently installed a large Macintosh multimedia laboratory, dedicated to the learning of languages. This has proved an enormously popular facility, both for teachers and students. After only four months of operation, booking requests by teachers for the second semester of 1993 have exceeded capacity by one and a half times; students often queue to use the machines when the laboratory is open for self access.

When the Language Centre began assembling the original submission to the University for funding for the laboratory, Centre staff were still relatively inexperienced in the design and installation of CALL laboratories, and the decision was taken to make extensive use of experts in FT and multimedia presentation within Melbourne University. However, despite this expert assistance, a number of unexpected problems presented themselves, some of which were potentially disastrous to the project.

As a result, the Language Centre has learned a number of valuable lessons about CALL laboratory design. This paper presents some of the more important of these lessons, in the hope that they will be of assistance to any other institutions who may be contemplating the installation of a CALL laboratory in the near future.

The Jean de I'Ours project explores the use of pictures and authentic (non pedagogical) language in foreign language learning/teaching. This project is based on the assumption that pictures can be used in a cognitive way, that they are more than mere entertainment, and that pictures used for communication in language teaching/learning should be more than mere pictograms.

Computer conferencing is a mode of communication in which participants send text messages between each others' computers which are all linked to a host computer via telephone lines. It is different from email in that it uses software that allows genuine group interaction. It is used for administrative purposes, for information exchange, for informal group interaction, and for distance education in which it can be combined with other media, e.g. face-to-face, satellite broadcasting, videoconferencing, ordinary correspondence, telephone contact, etc. The great benefit of computer conferencing for distance education is that it can be accessed at any time that is convenient to participants, since it need not be used for synchronous conferencing and the relevant files remain open to users for 24 hours of every day. It thus brings learners into frequent contact with each other as well as with their tutors and obviates, to a large extent, the need for cumbersome distance materials.

Research articles on CALL frequently proceed from the general to the specific. The general can take the form of a survey of the literature, a theory or theories of language teaching, learning or acquisition, or a summarised appraisal of a number of pieces of software; the specific is usually a collection of examples from a variety of sources to illustrate a principle. The merits of such an approach are widely recognised and are in no way disputed by the present writer. It is easy, however, for lack of attention to detail to leave the reader with many unanswered questions: Wouldn't it be alright to include/omit a given feature if… or if…? How did they come up with that technique? Why did/didn't they present the information this way? Exactly who was this designed for? How and when was it intended to be used? What constraints were they working under? In this article we have chosen to proceed in the opposite direction and trace the processes involved in developing one piece of software.

This article reports on aspects of an investigation into learners' approaches to vocabulary learning that was carried out as part of an experimental project integrating CALL into an ESL (English as a Second Language)/computer training course. Data from two of the learners is presented, and contrasting ‘deep’ and ‘surface’ approaches are identified. The link between approach and learning outcome is demonstrated. The importance of qualitative analysis of learning processes, and the possible contribution of CALL to this kind of analysis, is discussed.

Many factors affect the learning of a foreign language. When designing computer assisted language learning software it is usually not sufficient to think only about creating an exercise in the language, but students should be modelled in order to allow the program to take account of individuals' beliefs and learning. However, student models are criticised for various reasons, the most common of which include: 1. Modelling the learner places a great burden on the system, as it has sole responsibility for the creation of an accurate student model. 2. Student models are inadequate because it is not possible to model all aspects of a student's knowledge and learning. This paper describes the student model of an intelligent computer assisted language learning (ICALL) system which strives to overcome problems of traditional student models by taking into account issues important in the field of second language acquisition, and research in collaborative approaches to learning.

The Scandinavian countries are generally considered to have rather high standards in modern language teaching and learning, and at the same time to be reasonably well equipped with computers, videos and other information technology tools. This is true to a certain extent. Moreover, the situation differs a lot, not just from country to country, but within the same country from region to region. The following is not a complete survey, presenting the ‘state-of-the-art’ in the Nordic countries, but more a report of a number of initiatives taking place at the moment in the various countries or in a collaboration among countries.