31 de agosto al 2 de septiembre de 2000
EUROCALL 2000 took place in Dundee, Scotland between 31 August and 2 September, with the subtitle Innovative language learning in the third millennium. This year's conference was dedicated to the memory of János Kohn, a pioneer of CALL in Hungary. János's untimely death occurred in March 1999. His memory lives on with the EUROCALL-sponsored Annual János Kohn Scholarship which enables a worthy Hungarian scholar to attend the organisation's annual conference.
There were various events before the conference, including the Instil symposium, organised by Philippe Delcloque, who also organised the conference itself. The Doing WELL workshop provided an introduction to web-based language learning from the point of view of language teachers. An introduction was given by Uschi Felix (Monash University, Australia) about good practice and possible drawbacks. She emphasised that teachers should always consider looking for existing pages instead of creating their own ones from scratch. There are plenty of valuable language teaching pages. In her book she gives a thorough overview of these grouped by task type and language. Then work went on in two smaller groups. One dealt with Web-CT, a powerful tool for organising distance learning sessions, the other learned about creating web pages and exploiting authoring tools for language teaching. At the end of the workshop the Scottish video conferencing network was introduced. Every partaking institution has its own studio. Videoconference sessions are organised through the headquarters in Edinburgh using an ATM network with dedicated bandwidth.
The Human Language Technology in CALL workshop dealt with the role of natural language processing in language learning. Human Language Technology (HLT) helps us to use language more effectively, communicate in new ways or with people we could not have communicated with before. Cornelia Tschichold (Université de Neuchâtel) explained the difficulties we face when we try to process the speech of language learners. Trude Heift (Simon Fraser University, Burnaby, Canada) introduced the special needs of designing web-based language teaching systems through the example of German Tutor, an existing program. Anne Vandeventer (Université de Genève) spoke about FIPSGram, a French parser and the associated LT tasks. Finally we could hear talks on the use of parsers in the language classroom.
In accordance with the conference subtitle, the keynote speakers evaluated the results of CALL research from the point of view of future challenges.
Stephen Heppel (ULTRALAB, Chelmsford) talked about the future of e-learning in the context of 'gadgets and gizmos' and highlighted in particular the learning potential of mobile phones. He divided the spread of e-learning into three stages: the stage of denial, of opportunity and of adoption. He cited the results of research carried out among British children to show that the use of technology has become mainstream nowadays. However, he pointed out that technology itself does not change learning, and emphasised that the assessment of learning should focus on the process, not the product.
Wendy Mackey (Université de Paris-Sud) put the use of computers into a wider context. She listed several milestones of using computers in education during the past few decades. Among future challenges she mentioned going beyond WIMP interfaces (windows, icons, menus, pointing). She also emphasised the role of situated programming and bringing users into the design process.
Ron Cole (Center for Spoken Language Understanding, University of Colorado, Boulder) briefly summarised past visions of intelligent animated agents. Then he introduced Baldi, the conversational agent used successfully at the Tucker Maxon Oral School in Portland. This agent is capable of showing speech organs during speech synthesis. Although the agent is very popular among students, Ron Cole warned that existing systems are immature. We do not know enough about kids' speech, more corpora are needed to study speech and train recognisers. The speech of present systems is intelligible but not natural or expressive. Another drawback is that these systems are 'blind', they do not exploit visual information. Their results are available for the research community at http://cslr.colorado.edu.
Ray Kurzweil explained the consequences of exponential growth. He examined the development of instruments related to information technology and found that exponential growth began much before integrated circuits. His views of future possibilities were quite shocking:
Carol A. Chapelle (Iowa State University) analysed the vision of innovative language learning. Speech recognition for native language and learner language differs. Our view of language learning depends on our choice between Crystal's two options: the English speaking world and linguistic diversity. In the latter case, machine translation seems to be available, though language learning will not become faster in the next 50 years - some comfort for language teachers! In designing language teaching tools we need a harmony of criteria, research and software. Thus she introduced criteria coming from SLA research, questions for the evaluations of SLA tasks and the functions and purpose of software tools.
The conference papers, Show and Tell and poster sessions were organised into seven main streams: authoring and delivery, better language acquisition, language centres, delivery on the web, evaluation and testing, feedback and human language technology.
The conference was accompanied by a much-praised exhibition of the History of CALL, assembled by the conference organiser and co-sponsored by EUROCALL and CALICO. Much of he material from the exhibition is also available on-line at http://www.history-of-call.org/.
University of Veszprem, Hungary
Holder of the János Kohn Scholarship, 2000
Felix, U. (1998) Virtual Language learning, Australia: Language Australia Ltd.