A diagnostic component for natural language utterances has been devised which allows one to integrate (possibly inconsistent) evidence from a wide range of knowledge sources into a unique decision procedure. It is based on a procedure for the disambiguation of dependency structures using graded constraints. Different description levels like syntax, semantics, and others are treated by separate structural representations, which are disambiguated simultaneously. Due to the complete symmetry of the system architecture the approach allows for the successful handling of a wide variety or student errors.

There are many CALL resources available today but often they need to be adapted for level or culture. CALL practitioners would like to reuse currently existing material rather than reinvent the wheel, but often this is not possible. Thus, they end up building CALL material, both language content and software, from scratch. This is inefficient in terms of CALL practitioners’ time and as Felix (1999) points out, there is no point ‘doing badly’ what has already been done well. Why can’t we reuse what already exists? Often the language content is hard-wired in the software and cannot be modified or the CALL material comes as an executable which is hard or impossible to change. Authoring tools can provide a degree of flexibility, but often focus on particular parts of the language learning process (e.g. interactive exercises) and the associated language content cannot, in general, easily be exported into other formats and presentation software. One solution to this problem is provided by XML technologies. They provide a strict separation between data and processing. Thus, three types of reuse are possible: reuse of the data processing engine (i.e. the XSL processing files), reuse of the language content structure (i.e. the XML data files) and reuse the linguistic resources (i.e. the language content). In this paper, an example is given of a CALL template that has been developed using XML technologies. The template provides a structure into which the language content can be slotted and a processing engine to act upon the data to create CALL material. The template was developed for the production of CALL materials for Endangered Languages (ELs), but could be used for MCTLs (Most Commonly Taught Languages) and LCTLs (Less Commonly Taught Languages) also. It has been used to develop a language learning course for Nawat (a language of El Salvador), Akan (a Ghanaian language) and Irish, demonstrating the reusability of the language content structure as well as proving the reusability of the processing engine provided by the template. A further reusable feature is the ability to create courseware in different media (Internet, CD and print) from the same source language content.

This study examines the possible effects of a mailing list discussion on second/foreign language learning in the form of an explorative case study. Forty-six students in an elementary-level Japanese language class at a Canadian university participated. The study consists of three parts: interaction analysis, content analysis, and a student survey. The first two parts referenced the entire mailing list discussion archive. The number of the messages totaled 298. In order to analyze learner interaction, a map of interaction was designed and Levin, Kim and Riel’s (1990) Intermessage Reference Analysis (IRA) was applied. Content analysis was then carried out on the topics, context-type, and depth of learning process involved in each message. Lastly, a survey was distributed in order to discern participants’ perceptions towards the use of a mailing list for language learning. The results of the interaction and content analysis show how a mailing list discussion can provide a place to reflect on course content, enabling students to increase their linguistic knowledge through an exchange of ideas, thoughts, and opinions via student-centered interactions. The result of the participant survey shows that although the students’ participation in and perceptions towards the mailing discussion is not uniform, 35% of the students perceived the value of a mailing list discussion to be high. Through the examination of three different methods of analysis, the study concludes that there is a good potential for the use of mailing list discussions in second/foreign language learning. However, further research is necessary to determine which factors contribute to the successful use of this medium.

The present study aims to verify the impact of dynamic aids on learning L2 prepositions in relation to individual learner variables. Situated within the cognitive linguistics (CL) framework and differing from previous research, the present study hypothesizes that dynamic (animated) aids are not equally effective for all learners; rather, their effectiveness differs according to learners’ first languages (L1s) (Chinese or Japanese) and information-processing styles (verbalizers or imagers). To verify this hypothesis, we utilized learning materials comprised of static and dynamic images for three English spatial prepositions (above, on, over). After conducting a Style of Processing questionnaire, we administered three cloze tests (pretest, posttest, and delayed posttest) of target words to Taiwanese and Japanese participants (N = 109), whose L1s differed in terms of their linguistic proximity to English. Although no significant differences were found between the treatment groups in tests for all participants, the results were differentiated by individual factors. In results of a two-way ANOVA, Taiwanese participants showed significantly greater improvement from the pretest to posttest than Japanese participants when the participants used dynamic images, whereas the Japanese group made more learning gains from the posttest to the delayed posttest test. Moreover, imagers obtained more benefits from the visual aids, whether static or dynamic, than verbalizers. Our findings indicate that CL-based visual aids are beneficial and that individual factors, especially learners’ L1, may produce different learning effects, especially in multimedia environments.

The paper describes how networked self-access EAP materials have been developed at Warwick University since 1992. The current package of materials (The CELTE Self-Access Centre) can be freely accessed from the World Wide Web, and aims to provide some basic training in Information Technology alongside more conventional language and study skills activities. Problems of development and distribution are discussed, including the resistance of those EAP practitioners who have little experience of the Internet in an educational context, and the unwillingness of users to interact with unknown task setters.

For the last few years an experimental course in ab initio Italian Language for postgraduate students of Renaissance Studies has been run at the University of Warwick. This paper examines the role of concordance programs in assisting the design and implementation of the course. In the light of growing request for Languages for Specific Purposes (LSP), it is argued that the concordancer provides suitable answers to the need to introduce students to subject specific language from an early stage.

The Internet allows the language learner myriad opportunities to consult authentic sources and to practise using the target language with native speakers and other learners both synchronously and asynchronously. The current paper describes the design, development and implementation of language learning activities using text-based virtual reality or MOO (Multi-User Domain Object Oriented) in conjunction with the World Wide Web. The underlying purpose was to promote active, autonomous learning behaviours by enabling geographically distributed language learners to work with native- and non-native speakers of the target language on an engaging and motivating task with a concrete and meaningful outcome.

This article discusses some theoretical and methodological problems and issues in the study of computer-supported learning activities with special reference to a discourse perspective in research. Special emphasis is laid on finding ways to reach the participant perspective in computer-supported collaborative learning projects, which are by nature polycontextual. By this term we refer to the diverse scenes and situations of learning activities that computer-supported learning involves. We will also address the issue of making claims on effective pedagogies, leaning on the theory of discourse as situated and constructive, which again sets special requirements on what kinds of assumptions and interpretations can and should be derived through empirical analysis of data. Instead of rushed normative guidelines for instruction we should now take time and examine critically what is really happening in computer-supported environments in real-life situations. A look at the particular is needed. An important question to examine, for example, is how students and teachers use these tools in their everyday lives and how these processes are talked about. We will discuss through a number of examples the multimodal and polycontextual nature of meaning making in participant activities during three different university language courses between 1998–2002. This article considers relevant research approaches with the aid of theoretical discussion and examples of analysis.

This paper outlines the findings of a research project studying the effects of task modification on learner interaction when using email. Taking as its starting point interactionalist theories of SLA, it argues that if those interactional features characteristic of negotiation of meaning which have been identified as promoting SLA are to be preserved when tasks are transferred from a face-to-face spoken environment to a computer-mediated written asynchronous environment, then modifications to the tasks need to be made.

With more than 1,200 publications over the past two decades, experimental mobile-assisted language learning (MALL) studies targeting second/foreign language (L2) acquisition outcomes are certainly not lacking in quantity. Their research quality, on the other hand, has often been brought into question, most notably with regard to the adequacy of their assessment instruments and statistical analyses. Yet limiting the determination of research quality to the evaluation of testing procedures, and the statistical analysis of the results they produce, ignores the critical relevance of the underlying research parameters that generate the results in the first place. A comprehensive evaluation of quantitative experimental L2 acquisition MALL research quality, encompassing design as well as assessment instruments and statistical analysis, thus remains to be undertaken. The present investigation endeavors to do so based on an extensive compilation of 737 MALL studies published between 2000 and 2021. The research quality of these publications is evaluated according to four main parameters: language acquisition moderators, treatment intervention conditions, assessment instruments, and statistical analysis. These are applied according to a modified version of the Checklist for the Rigor of Education-Experiment Designs (CREED), which classifies research design quality into five levels: low, medium-low, medium, medium-high, high. With over three quarters of all studies falling within the low category, the result leaves much to be desired. Since the modified CREED algorithm developed here can equally be applied to studies from their inception, it offers a way forward to improve the research quality of future experimental MALL studies.

To transform pedagogically sound classroom tasks into effective CALL-ware, it is advantageous to have an approach that takes into account the general principles applicable to all human computer interaction (HCI). This paper outlines an attempt at such an approach. Teacher-designed, pedagogically sound software often fails in terms of usability, while highly usable, professionally made software is often based on ideas that are not in accord with those of a particular curriculum or teacher. The approach outlined below may assist in the process of transforming pedagogic materials and practices already successfully used in the classroom into easily usable software.

This paper presents the results of a study that demonstrates an effective use of the Web as a tool to increase motivation and thus promote reading and writing skills in Spanish as well as a deeper sense of the culture of the Spanish speaking world. In the study, thirty students of second year Spanish at the University of New South Wales were required to prepare an itinerary for a trip to a Spanish speaking country of their choice using the WWW as the only resource. In general our findings regarding improved language skills were consistent with the literature: an increase of vocabulary, more use of references, more student initiated interactions and greater interactivity in the classroom amongst students were observed. There was, however, one aspect, linguistic accuracy, in which improvement was not greatly noted. The data collected confirms that a task-oriented Web based course can increase the motivation of students, improve the scope of their reading, and enhance their perception of the target culture, all with a great effect on range of language explored, learned and re-processed, but a much lesser effect on the accuracy of written language produced.

This paper describes some of the pedagogical and technical issues involved in adopting a hybrid approach to CALL materials development. It illustrates some of these issues with reference to a vocational CALL project, LANCAM, which took such a hybrid approach. It describes some of the benefits and considerations involved in hybrid development and evaluates the commercial web development tool, ‘Dreamweaver’ and its extension application ‘Coursebuilder’, which was used to author the LANCAM materials. It argues for a reconsideration of generic web development tools in CALL authoring and also for language experts to be actively involved in the instructional design process.

RECALL aims at providing intelligent error feedback to learners' input in a role-play scenario. This is achieved by integrating three intelligent modules into into system: a Diagnosis Module which analyses the learners' input, a Learner Module which reocrds errors made by the learner, and a Tutoring Module which provides exercises and offers feedback. In this paper the functionality of the RECALL system is described and the approach to error classification as well as the actual implementation of this approach is discussed. A pedagogical perspective on the navigational and teaching capabilities of RECALL are considered.

This paper reports on the development of a prototype tool which shows how learners can be helped to reflect upon the accuracy of their writing. Analysis of samples of freely written texts by intermediate and advanced learners of English as a foreign language (EFL) showed evidence of weakness in the use of tense and aspect. Computational discourse modelling techniques were applied to the data to generate semantic models of fragments of the narratives with particular focus on their temporal structure. These models have been converted into dynamic graphical representations of the temporal relationships between discourse events as the narratives are written. The system also provides access to the ontology devised to model individual events and this offers learners insights into the events’ semantic properties. These techniques provide the basis for a stimulating learning tool capable of capturing key elements of written narratives, and prompting learners’ awareness of language use, particularly tense and aspect.

French learners at university meet difficulties in the comprehension of oral English. Being in a formal context of language learning, they need to develop language awareness to compensate for insufficient exposure to the English language.

To meet the students’ needs, data was collected in order to pinpoint the main errors made by French learners in listening tasks. The errors were then analyzed and put into perspective with the system of oral English; a connection clearly appeared between the errors and what is barely heard or cannot be heard in reference to written English, hinting at what could cause poor oral comprehension. Three areas of knowledge appeared to be missing in the students’ background: the links between morpho-syntax and phonology, the mastery of phonological data found in dictionaries and the possible recourse to strategies in order to compensate for what has not been heard properly. These issues were addressed in an on-line program designed for non-beginners of English at university.

This paper deals with the assessment of the progress made by users of the program in a formal learning situation. Two groups of learners were considered: students whose major is English, and students for whom the study of English is optional. Two series of tests were implemented, before and after the use of the program. The tests focused on the ability of learners to read IPA transcription, to count syllables in oral English, and to pronounce auxiliaries and prepositions in different contexts.

The results to be discussed establish that the two groups of learners significantly improved their knowledge of oral English. Of particular interest is the fact that, even if the two groups had significantly different knowledge of oral English before using the program, with non-specialists of English having poorer knowledge, the two groups obtain similar results on the post-test, showing greater progress on the part of the non-specialists. All learners appear to improve dramatically their knowledge of IPA and their ability to use it.

The progress measured by the tests was corroborated by other modes of assessment: a survey on the students’ judgment as regards the usefulness of the program, and individual interviews focusing on what the students recall from the content of the program. In the latter, the students used relevant meta-linguistic and meta-cognitive expressions, showing their ability for further progress in developing listening abilities in English as a Second Language (ESL).

When we decided to put this issue together, we called for contributions from CALL researchers and practitioners who were ‘‥concerned with the teaching of grammar in the post-communicative world…’. We were aware of the contentious nature of such a description – it suggests that the communicative approach is not, perhaps, a complete answer to the question of how to teach languages. Amongst the correspondence we entered into, subsequently, were some exchanges with colleagues who believed we were advocating a retrogressive step. It is only comparatively recently, they pointed out, that language learners have been freed from the shackles of piece-meal learning, drilling, correction and self-monitoring, and allowed to focus on ‘the challenge of communication’

Morphology consists of inflection and word formation. In foreign language teaching it occurs mainly in the form of inflectional paradigms. While this is certainly an important part of mastering a foreign language, an adequate use of morphology, both inflection and word formation, can facilitate the acquisition of foreign language vocabulary. By applying word formation rules as a way to structure their knowledge, advanced learners can multiply their active vocabulary. A practical problem in using inflectional classes and word formation processes in vocabulary teaching is that the information required to devise exercises is not readily available. In a dictionary, it is not possible to get an overview of word formation relationships or classes of words belonging to the same inflectional class. The nature of word formation processes even makes it impossible in principle to fully cover them in a printed dictionary without creating a lot of redundancy. An electronic dictionary does not automatically solve the problems involved, unless it is structured appropriately. In the Word Manager (WM) project morphological relationships have been taken as a starting point in the design of the lexicon. As a consequence, the lexicon is structured in terms of word formation and inflection rules, in such a way that complete and flexible access to the morphological processes and classes of a language is guaranteed. The full flexibility of WM lexicons requires that the entire system be installed locally. For a dedicated operational component, however, no such installation is necessary. It can be installed as an independently running program or be made available through the Internet. Two such applications (which are freely available) and their possible uses for teaching purposes are described.

People learning a new language have problems with various aspects of the language they want to learn – with its grammar, with its pronunciation, with its morphology... CALL tools have been provided for most of these linguistic levels, but comparatively little has been done about helping students with problems of meaning. We will argue that language technology is now mature enough to support tools that address at least some semantic issues, and we will describe a tool that attempts to provide appropriate support for a specific class of learners.

This paper discusses the role of customized learning environ1ments in the teaching of English to undergraduate science and technology students and focuses on a more subject-specific and skills-oriented approach in English for Specific Purposes (ESP). It may be argued that multimedia technology and the Web have a vital role to play in integrating often complex subject matter and language learning. As most university level students nowadays are mostly conversant with multimedia and Computer Assisted Language Learning (CALL) packages and feel at home before a web-based interface, creating teaching materials to cater to their needs requires a whole new set of skills, not least a knowledge of Web-design and Web-authoring programmes. Educators now need to create environments in which students can learn; they engineer new situations. Let us call it “WELL by design”. The focus is on motivating learning, a move towards constructing knowledge and involving the students in meaningful interactions. This paper focuses on how web-based materials are incorporated in day-to-day classroom teaching, how students are encouraged to interact with these materials, and how interaction is promoted and encouraged through the creation of pair and group activities based on the materials. Particular attention will be given to the potential of Web Enhanced Language Learning (WELL) in encouraging collaborative and constructive learning processes. The relationship between tutor and learner in a technology-enhanced environment will also be considered and discussed. A demonstration of specially devised link-mediated learning units within a web-enhanced course in ESP designed specifically for environmental science students will be made, together with an overview of resources, tools and observations within the context of the technology classroom. Particular emphasis will be placed on CALL and WELL as an essential support within a fully integrated language course for science students, as well as on the community-creating and connecting spirit of the Web in a virtual language learning environment.