This paper explores the potential and limitations of Grammatik 5 as a tool for revising writing, particularly as a grammar checker and adviser. It will examine the proposition that writing strategies can be enhanced when the program is 'tuned' to suit a group of ESL writers, with grammar rules and advice written to meet the learner's specific needs.

The rules and advice designed by the authors for Grammatik 5 have been chosen and written on the basis of an analysis of errors in a corpus of learner scripts within the academic genre of essay writing. The corpus contains a set of 28 essays written by 4 students over a period of 14 weeks in a university course designed to develop ESL students' writing skills.

In the role of corrector the computer has to be unfailingly accurate. If, on the other hand, the computer is cast in the role of tool, to simply identify elements of the language to the learner, the shortcomings of the computer as infallible assessor can be avoided. This role is a more appropriate one given the complexities of evaluating written work and limitations in the capacity of the current generation of computer programs to correct reliably.

A basic assumption of the paper is that the goal in using a grammar checker in this way is to extend the learners' capacity to review their written drafts and develop a critical approach to the writing process. While grammar checkers and advisers are not infallible in their analyses of English sentences, good use can be made of commercial programs when they are adapted appropriately.

This paper explores the impact of multimedia audio delivery and recording as a tool in language learning, in comparison with traditional technology. It traces the development of the field, and the experience of MLS Ltd. Two general approaches have evolved: Language Practice and Interactive Communication. These are described, with examples and observations on each type. The paper concludes with feedback and lessons learned from field studies, and pointers to potential future developments.

Montevidisco is an interactive videodisc application intended for intermediate Spanish students. It combines language enrichment with elements of role-playing games as the student plays the role of someone visiting a hypothetical city in Mexico, interacting with native Spanish speakers in native situations via videodisc. The concept originated at Brigham Young University in the early eighties, with the video filming done at that time. More recently, the BYU Humanities Research Center has adapted the original material to run on standard microcomputer platforms.

There are three implementations of Montevidisco in various stages of development. The first is a side-by-side version that uses a HyperCard stack on a Macintosh to control a separate videodisc player and monitor. The second version uses a video overlay configuration with ToolBook and Windows on an IBM platform. Both of these versions are 'finished' to the extent that they are being used in intermediate Spanish classes at BYU and elsewhere. General marketing arrangements are expected to be finalized this summer. The most recent implementation uses QuickTime on a Macintosh to display digitized video directly from the computer's hard disk.

This paper begins with an explanation of the general features of Montevidisco that are common to all three implementations. It outlines some of the design considerations faced and explains how the choice of metaphor influenced those decisions. Some questions that have been raised during the testing and review process are also addressed. The paper then categorizes the significant differences between the three implementations and discusses the technological and pedagogical reasons for them. It concludes with a status report on the QuickTime version and gives an assessment of this new digital video technology in an instructional environment.

In this paper the application of the techniques of Computer Assisted Language Learning to aspects of pronunciation will be considered. The problems of such application will be discussed and will be seen to be serious; some products will be described; the prospects for future applications will be viewed and will be seen to be far from bleak. Although the paper addresses the situation in The Netherlands, it is hoped that it will be relevant for practitioners in other countries.

This paper reports action research into the use of electronic mail as a channel for giving feedback to English L2 writers in a sequence of writing process oriented technical communication courses. Four teachers used electronic mail feedback (EMF) regularly with 81 students on honours degree courses In information Systems and Computer Science. Electronic mail was used during planning and drafting stages in extended simulations of technical communication situations in concert with class meetings, workshop sessions, one-on-one and small group conference and with audiotaped summative feedback upon completion of major assignments.

Teachers and students reported in response to a questionnaire and in interviews that they found EMF significantly advantageous in that it supports prompt, complete, and instrumental commentary on student work, empowers students to use English for authentic communication, and helps build positive relationships between students and native speaking teachers. All participants agreed that the extra channel of communication contributed to students' motivation to use invention techniques freely, to take risks with language and style in drafting, and to make substantive changes to text in drafting. Students reported that the objectivity of EMF was face saving, allowing them to use critical commentary from their teachers more productively. They also said that they found their teachers' use of EMF was a powerful demonstration of their willingness to make themselves available as mentors to their students.

The article outlines the context of CALL development at the University of Ulster, and explains the basic concept of the Meta-Text program as an interactive notebook/workstation which is used for the teaching and learning of translation skills, but which may be linked to other exercises.

It gives an account of its various prototypes as it has developed over the last five years. Extensive details of the formal evaluation of prototype 3 (version 2.1) are included. A description of the functionality of the most recent version (3.0) is given and the planned future enhancements of the program are outlined.

On the basis of this incremental development, a set of software design principles are adduced, with particular emphasis being placed on the interaction of developers and users in that process. An assessment of the program is given, based on its formal and informal evaluation thus far. It is felt that the program gives instruction in key intellectual skills and, through its interactive and creative functionality, encourages the kind of judgement essential for specialist linguists. Its encouragement of the consultation of data and its capacity to accumulate linguistic material are also felt to be significant, as is its flexibility of use with different types of student. Although basically a selfstudy tool, it has considerable potential for various forms of tutorial use.

It is anticipated that further development and evaluation will bring to light other valuable uses for the program.