MULTIMEDIA PROGRAM FOR THE LEARNING OF NAVAL ENGLISH

The program consists of a series of units on different aspects of Ship Construction, accessible through the main menu. Each unit is made of several pictures and texts in English explaining them, and gives the possibility of listening to the written texts read by native English-speakers, with interactive control: the user can examine the pictures, listen to the explanations and read the written texts as many times as he/she wishes, moving backwards or forward at will. Furthermore, the program has the characteristics of a hypertext, since some words in the written texts (marked in a different colour) offer the user the possibility of obtaining additional information about the subject, or of entering the technical dictionary included in the program. The main aspects of Ship Construction covered are: - Different types of merchant ships - Parts of a modern shipyard - The hull structure - Compartments inside the vessel - Names of decks - Welding in shipbuilding - The propeller - Anchors and cables - Naval Dictionary (English-Spanish) The program, based on a textbook and a set of cassettes by the same authors, can be considered a useful tool for the learning of Maritime English, allowing the students to work individually and follow their own pace.

Back to previous page.

A MULTIMEDIA PICTURE DICTIONARY OF PORTUGUESE-L2

In this paper I will present a prototype of a multimedia picture dictionary of basic Portuguese for young learners of Portuguese as a foreign language: «O Meu Dicionário Interactivo». This prototype has been built in the context of my master’s dissertation in Educational Technology in the University of Minho. In this presentation of the prototype I will refer to: a) the programming platform used and the reasons for its choice; b) the lexical coverage of the dictionary and the theoretical criteria for its selection; c) the basic metaphor of the program and the assumed learning strategies in the acquisition of new L2 words; d) the semantic network that structures the lexical information of the dictionary and its role in supporting learning strategies; e) the informational multimedia nature of the lexical nodes; f) the modular organization of the application: picture dictionary, conventional dictionary, the lexical learning activities.

Back to previous page.

DESIGN AND TECHNICAL ISSUES IN THE PRODUCTION OF A MULTIMEDIA TRAINING COURSE: THE CASE OF AUTOHALL

Issues relating to pedagogical matters and the design of user interfaces in multimedia training courses, taking into account the technical possibilities and limitations of current standard technology, are becoming a well trodden path in the area of multimedia research and development. Nevertheless, case studies still provide the best route to understanding the nature of the problems involved, as well as providing potential solutions which may be used in other projects of a similar nature. Autohall has presented challenges with respect to technical budgetary constraints, linked directly and indirectly to the third factor, the linguistic form and content of the course. This paper will outline and discuss the interaction between the linguistic and the technical/design issues involved in the production of Autohall, and point to possible ways in which what has been learned can be generally applied to projects in the area of computer assisted language learning.

Back to previous page.

MULTIMEDIA ON A DOS PLATFORM. EXAMPLES OF PRACTICE AND USE OF THE AUTHORING PACKAGE QUESTION MARK PROFESSIONAL FOR ENGLISH, GERMAN, SPANISH AND DUTCH LANGUAGE TEACHING.

In late 1992, the Language Department of the Facultés Universitaires Catholiques de Mons (FUCaM) opened a free access CALL centre. Equipped with a network of 11 IBM compatible 386-machines with screen monitoring and sound facilities, this centre was meant to give the students ample remedial practice in the four languages taught at the university (English, German, Spanish and Dutch). We were aware from the outset that the success of such a centre hinged upon not only intensive daily management, but equally, and perhaps even more importantly, a well balanced choice of CALL software. Particularly crucial was the selection of products that would make for an attractive learning environment and use all the technical possibilities of our computer-room. In short, the software was to meet the following requirements: a) Fully activate the added value of the computer as a teaching tool, combining a varied language content with immediate feedback and the use of pictures and sound as additional stimuli; b) Supply a wide range of different learning activities; c) Be user-friendly; d) Preferably be «open», so that we could adapt the software to the contents of our language courses rather than the other way round; e) Be flexible, so that the software programs could be placed in whatever directory we liked and could be activated in whatever way we preferred (e.g. via a self-developed «click-the-button» interface); d) And, last but not least, offer a highly stimulating working environment for the learners. We ordered a number of «open» gap filling and text reconstruction packages (Wordstore, Gapmaster, Telex, Adam&Eve), as well as closed grammar drills all of which displayed possibilities of feedback and could be accessed in a user-friendly way via our self-developed end user interface. Still, we found that most of the aforementioned programs left much of our hardware's fascinating potential untouched. Since the CALL centre's computers were all equipped with Sound Blaster compatible voice cards and VGA colour screens, we were still looking for a program that would combine pictures and sound with interactive language content. Applying high-tech multimedia on a rather technically limited DOS platform became the real challenge, since our centre was equipped with slow and (by today's standards) old-fashioned 386-sx computers, having a maximum addressable RAM-capacity of 640 K. One of the only programs which met these requirements was an authoring tool initially not developed for language teaching purposes: Question Mark Professional, a package devised in England. As its name suggests, it was designed to ask questions on any kind of subject. Consequently, its main use was as a questionnaire in sociological research. In its University package form, the program comes with six runtime (or student) modules in English, Spanish, German, French, Dutch and Italian (interesting for the teacher who wants to expose his/her learners to a monolingual environment where the command and target language are the same) and allows nine different types of questions, ranging from explanation screens (no answers expected), via multiple-choice and gap-filling to open answer questions (any answer accepted), with ample possibilities of feedback and help hints. The program also has a snap facility which captures screens and transforms them into a 16 colour «paintbrush» format (pcx-extension) files which can be cut and pasted in any question. Together with Question Mark's multimedia editor –actually a «call» function which activates other DOS or Windows programs from within the Question Mark file–, this constitutes a powerful authoring tool enabling the user to combine pictures and sound with interactive questioning on any DOS machine. Moreover, the program remains flexible, i.e., does not impose any structure on the user and can easily be manipulated by anyone who would like to incorporate it into a DOS or Windows interface. In our demonstration we will give examples of real practice for English, Spanish, German and Dutch, showing the evolution from plain questionnaire to full multimedia sequences as they are currently used by FUCaM's Language Department. We will also concentrate on the possibilities of question banking where different sets of question files can be combined and randomly displayed on the screen. Finally, we will throw a glance at the powerful answer analysis device of this authoring package thus proving that multimedia significantly contribute towards creating a highly stimulating working environment for the learners.

Back to previous page.

PROPOSAL FOR THE DEVELOPMENT OF A SYSTEM FOR LANGUAGE TESTING WHICH COMBINES A SET OF PARSER/ANALYSERS WITH A MULTIMEDIA INTERFACE

This paper discusses a proposal for the development of a prototype of an integrated system that can be used for language testing. The system combines a set of automatically generated parser/analysers and uses a multimedia interface for user input. Prototypes of the various components of this system have already been developed and it is hoped that pilot tests of a system combining these components will soon be conducted. The parser/analysers are non-deterministic LALR automata which are augmented with so-called semantic actions. These automata are Prolog programs which are generated by another program which has already been implemented in C. The input to the generator is a two-level grammar. The base grammar is written in the form of phase structure rules and the second level extension is based on the unification of directed acyclic graphs. These graphs are essentially feature structures which can carry information such as agreement, semantic features and other context information. In variance with some grammar checkers, the parser/analyser does not check for errors which are variations from the target language. Instead the parser/analyser seeks to accept all possible user input. Essential to this is the assumption that, for a given context, the possible grammars used can be systematically characterised by one (ambiguous) grammar. Initial pilot studies have shown that, for a given context, the range of variation of user input is limited and manageable. Because of the nature of the Prolog interpreter, it is possible to accept more than one analysis of an input sentence; an expression of a learner's interlanguage might well result from a combination of factors. A functioning prototype of the interface has already been implemented using Director 4.0. The interface uses different media (video, pictures, sound and text) to restrict the class of possible user input. The user is given a certain situation, and is then asked to perform a certain language task. The input of the analysis is open in the sense that the user is «free» to type in whatever sentence he/she feels is appropriate for the given situation and task. The author feels that such a limitation of context is necessary in order to make an analysis feasible. Since the parser/analyser is a program that is generated and not written, it is possible to tailor a different grammar specification to the context of each given problem; thus a set of parser/analysers –one for each context– is created. In practice, the design of language tests for learners of a foreign/second language often involves a trade-off with respect to goals and resources. The direction of language teaching in recent years has seen a shift toward notional-functional syllabuses and the teaching of languages for special purposes. Instead of providing varied and direct performance-related tests, there is a tendency to model tests on the older, indirect model of psychometric testing. This is largely due to the expense of the direct kinds of tests. While the testing system proposed here does not entirely provide a direct measure of performance, the use of complete sentence input could allow it to be model for tests which are more direct and performance-related than for example, multiple choice or cloze tests. Once implemented, this testing system could offer economies of scale in comparison to direct, performance-related tests based on interviews or role plays.

Back to previous page.

D'UN DISPOSITIF CLASSIQUE A UN DISPOSITIF D'AUTOFORMATION GUIDEE EN ANGLAIS, DANS UNE FORMATION D'INGENIEURS A PARIS 6

Au sein de l'équipe ORDI, rattachée à l'UFR de Linguistique, université Paris 7, je participe à la conception de didacticiels d'anglais et à l'expertise de produits des nouvelles technologies de formation (NTF). C'est ainsi que je travaille, depuis quelque temps, à la conception et à la réalisation de modules de traduction-navigation du français vers l'anglais avec le système Bonaccord, en collaboration avec l'université d'Aberdeen, dans le cadre du projet OPAL-LINGUA, et que j'ai participé à différents travaux d'expertise (Musée du Louvre, Archives Nationales, Direction des Enseignements Supérieurs). En tant que professeur d'anglais à la Formation Permanente de l'université Paris 6, j'interviens, depuis sa création, à l'Institut de Formation d'Ingénieurs en Techniques Electroniques de Paris (IFITEP, géré en partenariat avec l'université et la Chambre de Commerce de Versailles –CFA) où j'ai été chargée de mettre en place la formation en anglais pour ces élèves-ingénieurs (nouvelles filières Decomps), qui passent alternativement une semaine à l'université et une semaine en entreprise . Je me propose de décrire les différentes étapes de l'intégration des nouvelles technologies de formation dans l'enseignement de l'anglais à l'IFITEP, tant du point de vue ingénierie de formation (montage de dispositifs, de la première à la quatrième année) que du point de vue du travail de concepteur-ensemblier (choix des supports) que cela suppose. Je ferai tout d'abord un historique rapide des premières étapes en évoquant, en fonction des besoins de ce public d'apprenants, des objectifs assignés à cette formation et des contraintes institutionnelles, les raisons qui m'ont conduite à choisir un dispositif pluri-média, utilisant des supports EAO et vidéo, en situation de cours traditionnelle, avec travail en groupe en présence d'un enseignant, dans une première étape, et dans une seconde étape en situation de cours, avec travail en groupe, pour le travail à partir de la vidéo, et en situation de travail autonome, sans enseignant, pour le travail sur matériaux EAO. Puis je présenterai le compte-rendu d'une année d'expérimentation du module d'autoformation guidé proposé aux élèves-ingénieurs en quatrième année, en abordant plus précisément les points suivants: les objectifs spécifiques de ce module (pratique autonome, individualisation et poursuite volontaire de l'autoformation en dernière année), les grands axes du dispositif (tutorat, sous forme de rendez-vous individuels, séances de travail autonome en salle multimédia et séances d'oral en petits groupes) et un premier bilan (utilisation des matériaux pédagogiques, suivi des plans de travail des étudiants, appréciation du module, problèmes rencontrés). Enfin, à la lumière de cette première expérimentation, j'évoquerai plus particulièrement les éléments qui semblent favoriser l'autonomie dans l'apprentissage et je mentionnerai les modifications apportées dans ce sens pour la deuxième année d'expérimentation de ce module d'autoformation.

Back to previous page.

ASSESSING CORPORATE FOREIGN LANGUAGE NEEDS: CREATING A COMPUTERISED LANGUAGE AUDITING TOOL KIT ADEQUATE TO MEET INTERNATIONAL NEEDS

There are two central reasons for conducting an audit of corporate foreign language needs: the first is to assess a company's current language capability matched against an analysis of its requirements. The second is to provide an informed basis for language training and recruitment of appropriately skilled staff. Auditing tools therefore have to assist analysis at three levels: the strategic, taking into account a company's forward plans, an operational, that investigates requirements and capability across the organisation, and individual at the level of the post holder. This paper discusses a suite of tools developed in the LINGUA Business Language Workbench Project (1994) involving partners in the United Kingdom, the Netherlands, Portugal and Greece and discusses the strengths and weaknesses of the tools with regard to their presentation in face-to-face, questionnaire or computerised modes. It also considers the challenge of creating a tool kit sufficiently flexible to cope with a range of companies from the smallest to the multinational, and considers the implications of incorporating such tools into distance learning language materials.

Back to previous page.

THE APPLICATION OF LEARNING STRATEGIES IN MULTIMEDIA COURSEWARE WITH THE FOCUS ON THE LISTENING SKILL

Many theories of second language acquisition have based their views on the fact that listening input is paramount to the development of linguistic knowledge. On the other hand, current views in applied linguistics have shifted the focus from teaching to learning and the mental processing involved in it. This change of stress on the learning process has brought about an increase of interest in learning strategies. Although the language teaching profession is already applying all these assumptions in the classroom context, the integration of learning strategies to foster the acquisition of the whole spectrum of the listening skills is under-researched in the CALL environment; despite the fact that researchers claim that awareness of learning strategies makes good language learners. This paper will analyse the ways in which learning strategies applied to the listening skill can be implemented in the development of CALL materials. For such purposes a theoretical framework will be provided to evaluate to which extent existing materials incorporate strategies that promote the learner's cognitive processes. The multimedia computer, as a tool that can handle individual language practice, has a lot of potential for this type of reflective learning.

Back to previous page.

FEEDBACK: THE NEED FOR A FLEXIBLE TOOL

Feedback in second language acquisition is understood as comments, usually by a teacher, to a person who has completed some learning task. Such feedback is called «positive» when the task was completed with success and «corrective» when it was not. Restricting ourselves to corrective feedback we might say that ideal teaching circumstances will make feedback superfluous, while the need for corrective feedback will increase with teaching being less successful. Indeed, it could be said that when no teaching is given or when teaching completely fails, the teaching task will be taken over by corrective feedback. In this view the fundamental difference between teaching and giving corrective feedback lies in the moment at which information is supplied: before completion of the task, or after it. On this view of feedback, all information necessary to carry out a task should be available to the student prior to carrying out the task. Furthermore, as we may not expect students to always know what information is important for them to carry out a task successfully, it should be clear to students where, when, and how to ask for information, depending on the task to be completed. This paper describes an authoring program that allows feedback on three different levels: 1) at the level of a full text or exercise; 2) at the level of a particular textual problem or item of an exercise; 3) at the level of an individual response given by the student. The amount of feedback that can be given by the program is only limited by available disk space. Feedback is made available to the student either automatically by the program or by giving access to function keys associated with information relevant for the problem on which she is working. The program, not the teacher who prepares the teaching materials, will decide what keys are used to make feedback available and will free the keys when feedback is no longer needed so that keys can be used for other information. The program also looks for an appropriate place on the screen where to put feedback, in order to prevent feedback from covering materials on which the student happens to be working. Finally, a separate index enables students to get access to whatever information is available in lesson files. Because of this design, the program gives the author of teaching materials great flexibility: authors may program an arbitrary number of info/feedback chunks on three different levels and of an arbitrary size. Function keys (or buttons in windows) are dynamically associated with info/feedback as needed by the student. And if info/feedback chunks are too big to be displayed on (part of) the screen, students can scroll through the information. The design also allows the author to prepare teaching materials in a very traditional way. Teaching materials do not need to be coded for matters like: where the students have to give answers, how much space is needed to give an answer, what keys will display what information, when feedback should be displayed, etc. Experience shows that even teachers without experience in using computers are able to prepare their own lessons with feedback after one or two hours of training.

Back to previous page.

STRATEGIES FOR DISTRIBUTING AND RUNNING MULTIMEDIA COURSEWARE ON NETWORKED MACHINES

Multimedia software is memory-intensive. A session of 1 hour's duration is unlikely to fit onto a floppy disk. The intensive use of graphics, sound and video can mean memory demands of 10 Megabytes per session. The hardware requirements and expense of pressing multiple CD-ROMs, not to mention the relative permanency of doing so, means that distributing multimedia courseware over a network is the only practicable method. This study will examine the three main strategies which can be employed to enable software to run on a number of computers simultaneously. It will show how user input can be logged so that teachers can be kept informed of software usage and the progress of their students. What information should be recorded and what is unnecessary? How do you assess what can amount to pages of data on class and individual usage of the courseware? Do the users need to know their progress is being logged and can the data be used for assessment? These questions have been investigated at the Hispanic Studies Department of Glasgow University where multimedia software has been combined with a new course book and the continued use of the classroom to provide an integrated approach to language learning. The software element of the course is being funded by Teaching with Independent Learning Technologies (TILT), a TLTP project designed to evaluate the impact of technology on Education. After intensive usage by both beginners and honours Portuguese students, much useful information has been accumulated on implementing courseware over a network and obtaining data on its usage.

Back to previous page.

TEXT REPRODUCTION IN A HYPERTEXT ENVIRONMENT

This paper reports on the first phase of a CALL project for German. The program endeavours to combine experience in the fields of traditional CALL and hypertext applications with the approach of the theory of formulating, in order to design a package that would enable students of German to develop their reading and writing skills, learn new vocabulary and syntactical structures; and at the same time enhance their knowledge and competence in the production of challenging sorts of text. The aim of the first phase was to produce a prototype program that comprises a reading comprehension module and a guided tour through the reproduction of a text. In phases 2 and 3 redundancies in the contextual help will be streamlined and more linguistic material (texts, glossaries, rules) will be incorporated. Phase 4 will focus on authoring tools (textual analysis, tagging, information supply) for the package.

Back to previous page.

COMMERCIALLY PRODUCED CORPORA IN THE ELT CLASSROOM

The English Department at Hong Kong Polytechnic University has established a project which is exploring the potential benefits of integrating CALL/TELL as part of its ELT service courses. The emphasis, initially, is to investigate the feasibility of exploiting commercially produced corpora (in the form of CD-ROMs such as multimedia encyclopaedias, dictionaries, newspapers, collections, etc.) on the ELT service courses taught at the University. This paper describes the opportunities for using commercially produced corpora as a means of exploring English language usage, and suggests teaching methods and materials which would enable students to access the corpora with suitable programmes.

Back to previous page.