Reviewed by Jane Vinther, e-mail: jpv@get2net.dk

Angela Chambers begins this collection by describing that the majority of the authors of the chapters in this book originally came together as a result of projects initiated and/or funded by the European Commission such as the SOCRATES Thematic Network Project (TNP), coordinated by Wolfgang Mackiewics, Freie Universität Berlin. Also mentioned are projects such as the International E-Mail Tandem Network, the RAPIDO Project (1997) and projects under the aegis of the European Language Council. This explains the book's emphasis on the European dimension in a field which is global in nature. In the political context of European integration, ICT and language learning serve the aim of the European Commission to create a virtual European education area. Chambers points out that in this field there is a danger of letting the latest technological innovation determine research approaches, and she emphasizes the importance of theory- and pedagogy-driven research.

In chapter one, "New technologies: A suitable subject for research?," Graham Davies explores the question of what constitutes acceptable research especially in the light of the distinction between development and research. CALL is a field in which it is easy to fall between these two chairs which may cause problems within academic departments and with regard to funding of research projects. The organisations EUROCALL, CALICO, and IALL have played an important role in establishing the research field, and in a joint policy document issued after the Essen colloquium in 1999, CALL theorists, researchers and practitioners from all over the world laid down guidelines for CALL research and its evaluation. The statement is included in an appendix.

CALL research only dates back to the beginning of the 1980s, and Davies outlines how research topics have developed from a focus on technology and efficacy of outcomes to learner characteristics and strategies to a focus on the psycholinguistic processes. The chapter gives a good overall view of CALL as a developing research field, and the general discussion of what principles should apply in the future is central, but the chapter does not bear completely convincing evidence to the author's conclusion that "the future of CALL and CALL research looks rosy" (p.23) in that the chapter rightly points to the persisting difficulties with funding and recognition.

This theme of research is continued in the next two chapters which focus on learner autonomy and self-directed learning. Chapter 2, "Learner autonomy and the challenge of tandem learning via the Internet" by David Little, and chapter 3, "Learner autonomy, self-instruction and new technologies in language learning: Current theory and practice in higher education in Europe," by Jeannette Littlemore, both take David Little's (1991) definition of learner autonomy as their starting point, and both investigate the question of how new technologies can enhance learner autonomy. Despite their common starting point these two chapters supplement each other well, and together they cover a broad field of inquiry. Little and Littlemore both devote much space to the definition of learner autonomy, what it is and what it is not, arguing that responsibility for one's own learning should not be taken to mean being left alone to one's own devices, and it should certainly not be regarded as a cost-cutting measure.

In chapter 2 Little defines the autonomous learner as one who is capable of managing, monitoring and evaluating his/her own learning. The process of creating autonomous learners is a psychological, taking place in the individual learner. Learner autonomy is not merely a matter of organising self-access systems and self-study facilities. Little takes care to point out that 'self-access' and 'autonomy' are not interchangeable terms or concepts. Self-access systems can help the learner develop autonomy in that such systems make it easier for the learner to become self-managing, but supervision, advice and support are still necessary. Learner autonomy is not a concept exclusive to language learning, or to new technologies for that matter, but the new technologies in language learning do offer unique opportunities for development in this highly important area. Tandem learning is one way that technology and autonomy can benefit from each other. David Little (1999) reports from a European project which students at Trinity College, Dublin were enrolled in The International E-Mail Tandem Network, which was coordinated by Helmut Brammerts at the Ruhr-Universität Bochum. This project started as a purely e-mail based project, but later Klaus Schwienhorst, Trinity College, Dublin, developed a virtual language centre at Diversity University which enabled students to have synchronous communication in MOOs (Multi-user domains, Object Oriented). E-tandem learning solves some of the problems that are involved in face-to-face tandem learning but creates new ones of its own. General problems with tandem learning are accessibility, compatibility and the fact that pedagogy which encourages autonomy seems to be in short supply in many universities.

The latter aspect is one which Jeannette Littlemore investigates in the second half of chapter 3, in which she reports from a survey of 40 institutions of higher education, mostly universities, from across Europe. Language centre directors from these 40 higher education institutions were interviewed as well as asked to fill in questionnaires about their experiences with and attitudes to learner autonomy, self-direction and new technologies. She found that three groups emerged and only about a quarter of the respondents had any clear idea of what learner autonomy is. The remaining respondents either did not see learner autonomy, self-direction or new technologies as relevant to their situation (just over one fifth), or they found it difficult to motivate their students to be autonomous (over half). The group of respondents that did have a clear strategy and vision with regard to learner autonomy strongly encouraged their students to use new technology, and many universities in this group encouraged student participation in the development of self-study facilities as well as in their maintenance and support services. These universities expressed the view that students need systematic, explicit training in language learning strategies and that this training should be embedded in the curriculum in order to make students become truly responsible for their own learning. Littlemore also makes it clear that independence and autonomy are necessary ingredients in self-directed learning. She adopts Holec's (1988) distinction between the term 'autonomous', which should only be applied to a person, and the expression 'self-directed', which describes the learning process. It is difficult to disagree with Little and Littlemore's shared conclusion that there is too little learner autonomy at most universities, and that it would be desirable to encourage and cultivate student independence, not to compensate for limited resources but as an end in itself.

Chapters 4, 5, 6 deal with different aspects of Authoring Tools (AT). In chapter 4, "Criteria for the evaluation of Authoring Tools in language education," David Bickerton, Tony Stenton and Martina Temmermann give a very useful exposition of what criteria to include in an evaluation of ATs. Their focus is the RAPIDO project (Rapid Authoring of Packages using Innovative Development Tools) in which 45 academic linguists examined and evaluated the feasibility of a variety of ATs. They use Authoring Tools as a generic term for applications which facilitate software production. The value of this chapter lies not only in the information given about named ATs, but primarily in the very comprehensive and thorough models of evaluation that the three writers present. They stress that qualitative evaluation criteria must be combined with the more pragmatic parameters. An easy way to get an idea of the potential of a given AT is to work out its range features. Such a taxonomic evaluation should, however, be complemented by an appraisal of the ease of use and costs. The chapter barely touches on the pedagogical issues, but one should bear in mind that that is not the purpose of this presentation. What they do is provide a tool which can be used by anyone wanting to make a qualified evaluation of the increasing number of products in this area, and the relatively simple system they present will be of use to both the "technophobic linguists" and the "linguaphobic technologists" (p.59), as they call the two groups at either end of the scale of users.

In chapter 5, "DISSEMINATE or not? Should we pursue a new direction : Looking for the "third way" in CALL development," Philippe Delcloque presents his ideas for the DISSEMINATE concept and outlines of the development of CAL(L) since the 1960s. This outline presents the various development phases of CAL(L), primarily with the outset in technology. In the second half of his chapter, Delcloque first gives his theoretical foundation for his DISSEMINATE ideas. The theoretical basis is Jean Gagnepain's (1982/1990) four levels of mediation of human organisation, abstraction and behaviour. Delcloque muses on whether Gagnepain's hypothesised "technological mediation" in the second level could explain why some people are what he calls techno-anxious/computer-anxious (p.73). These are particularly to be found among women, it seems. There is a majority of men in ICT and even among Internet users despite the fact mentioned by Delcloque that women genetically are better communicators. There is a point here. It is debatable, however, what it is in evidence of. There is no doubt, though, that theory-driven or pedagogically driven developments are much needed, just as it is true that authoring tools should be authorable and expandable. DISSEMINATE (D-Distributed, I-Integrated, S-Stable, S-Superimposed, E-Evolutionary, M-Modulaware, I-Interactive, N-Networked, A-Authorable, T-Tracking, E-Education/Edutainment) contains the sound principles for development in the digitalised future. The vision as described in the conclusion is a digital ringbinder, accessible and flexible.

In chapter 6, "CALL material and learner competence," the reader is made familiar with the results of a very thoughtful study investigating the pedagogical application of the authoring system Learning Labs. Jean-Claude Bertin, who is the author of this chapter, combines an explanation of his theoretical and didactic concerns with an outline of the practical organisation of linguistic material. Bertin's overall concern is to give priority to a theory-based application of the authoring system. He, too, advocates individualisation and learner autonomy, but he is also wary of the rudderless learning situation of total freedom, which may be the result of students let loose on the Internet without guidance. Bertin operates with a combination of three axes in the learning situation: the didactic axis (linear), the heuristic axis (hypertextual), and the referential axis (satellite tools). All three should be included, but the emphasis that each of these axes are given can be customised to the individual learner or the individual aim of the learning situation. The higher the level of competence the more freedom can be given to the learner, and the applications of the authoring system can be adapted accordingly. Bertin concludes that total autonomy is now superseded by guided autonomy, which incorporates differences in learning styles and strategies and varying levels of competence. He points out that students, no matter what their level of competence, still want to feel the teacher's presence, be it implicit or explicit. This need to individualize makes demands on the authoring systems, and Bertin voices the opinion that adapting the tool should take no longer than half an hour. Bertin's chapter provides invaluable guidance to anyone involved in authoring.

Chapters 7 and 8 have as their common topic language engineering. In chapter 7, "From gap-filling to filling the gap: A re-assessment of Natural Language Processing," Sake Jager examines the question of what NLP has to offer with respect to CALL and what language learning requires. Practical applications of NLP have found a commercial expression in for instance computerised travel information and directory services, and Jager sees the greatest potential for NLP in spoken language processing with its two dimensions of speech analysis and speech synthesis. NLP can be used in ways which there are already non-NLP ways of dealing with, such as gap-filling exercises, but Jager proposes that NLP be used primarily where the non-NLP solutions are non-existent or undesirable. In the CALL perspective a distinction is made between the mechanical aspects and the meaningful aspects of spoken language. Increasingly, the emphasis is on the meaningful aspects of spoken language as NLP applications in speech recognition can be a step towards a communicative- based practice with a design based on open-ended types of activities, even dialogues. The major obstacle is that it is difficult to build systems which will accept a wide variety of pronunciations. Jager concludes that language engineers and language teachers need to work closely together to ensure that the pedagogical objectives are incorporated in the technical applications. Just because something is technologically feasible does not necessarily mean that it is appropriate in CALL.

In chapter 8, "Human Language Technologies in Computer-Assisted Language Learning," Mathias Schultze outlines some of the aspects and challenges of human language processing (HLP) in relation to CALL. His outline includes extensive exemplification of HLP applications, beginning with the PLATO Project (Programmed Logic for Automatic Teaching Operations) which was initiated by the University of Illinois in 1960 and ending with the so-called Intelligent CALL systems. The development as Schultze sees it has been mainly technology-driven, as has the classification. He cites Wollf's ( 1993) five categories as an example of this classification: traditional computer assisted language learning applications, artificial intelligence applications, utility applications, multi-media applications and communication applications. In the latter half of the chapter Schultze focuses on Artifical Intelligence applications in general and parser-based CALL in particular. Some researchers have taken a dim view of the possibility of creating CALL tools which incorporate Natural Language Processing and until recently few applications were available on a commercial basis. Schultze mentions phonology as an area where integration has been successful (e.g. APron, Autosegmented Pronunciation Teaching). Other projects worth mentioning are the GLOSSER Project, which was a COPERNICUS project, and the parser-based projects at UMIST. The advantage of parser-based applications is that the students are given more freedom than in the traditional CALL applications which are based on recognition and pattern matching. The disadvantage is that the formal aspects of language learning with its emphasis on the syntax of the input come to the foreground, but this can be avoided by creating relevant and authentic communicative tasks for the students. Schulze devotes some thoughts to the nature of human-computer interaction, and he points out that in order to understand this interaction, it is important to understand that humans are holistic entities who create their own environment and can cope with inherent contradictions, all of which computers cannot. The intentionality of human communication cannot be imitated in computers, which usually carry out sequenced operations, albeit in rapid succession. He ends his chapter by the interesting observation that the frustrations felt by students when the computer for instance rejects a correct input or raises a false alarm may be due to the tendency to project the human intentionality of actions onto the computer.

In chapter 9, "Learning out of control: Some thoughts on the World Wide Web in learning and teaching foreign languages," Thomas Vogel, on the basis of some personal experiences, investigates the new dimensions that the WWW can bring to language learning. One principal question is whether the Web can constitute a new naturalistic setting for language learning. Vogel likens it to walking down the high street in a country whose language you're trying to learn. There is an overload of information, and the input is unstructured so the learner needs to be selective in what is paid attention to, but motivation and learner autonomy, which are presumed to have a positive effect on learning, are qualities which are supported by integrating the Web in the language classroom. This is more likely to be the case if the medium is used to create an authentic learning environment in relation to text as well as task and learner through a problem-solving approach. Vogel does point out, however, that some empirical research is needed to throw light on whether learning is actually taking place and whether the pace and sequence of learning can be influenced. His theoretical stance is that "language learning for the major part is an unconscious process which can hardly be influenced by formal structured teaching" (p.139). A successful integration of the Web in the language learning classroom demands that the teacher is willing to give up the classical role as the possessor of knowledge, from which follows that students must also be prepared to accept a new role for the teacher.

In chapter 10, "Concordances in the classroom: The evidence of the data," Joseph Rézeau advocates concordancing as a way of combining authenticity and autonomy in the language learning classroom. In this matter Rézeau supports Johns (1991) rather than Widdowson (1996). The increasing numbers of readily available concordancers and corpora could turn the so-called data-driven learning (DDL) into a valuable contribution towards introducing research methods into language learning at university level. Rézeau gives a number of intriguing examples of possible ways of utilizing DDL resources, all of which he himself has been involved in. He expresses the hope that DDL might lead to more inductive approaches to language learning and perhaps even give rise to a reassessment of the place of translation in language learning.

Chapter 11, "ReLaTe: A case study in videoconferencing for language teaching," by John Buckett and Gary Stringer, the last chapter in the book, is a case study of a joint project by the University of Exeter and University College, London. The aim of the ReLaTe project (Remote Language Teaching) was to demonstrate the feasibility of using multimedia conferencing to share language teaching resources. The project used the UK's JANET MBone service, at the time 10Mb/s bandwidth between Exeter and UCL. The MBone (Multicast Backbone) provides for video, audio and shared workspace. Text and graphics can be scanned in and worked with in the shared workspace and students as well as tutors can add comments, etc. The importance of this shared workspace was one of the surprising results of the trial classes. Students found the equal access to the whiteboard very liberating. The authors conclude that as access to sufficient bandwith become more widespread, and cheaper, videoconferencing will hold much potential for language learning classes, and since the Mbone is an international network, cross-border cooperation and projects hold prospects for the future.

Overall, this book brings together a rich diversity of perspectives and topics under the common umbrella of ICT. Many of the articles have been published before in slightly different versions, or been presented at conferences, so the contents may not be completely new to people in the field. However, the collection of these articles gives a newcomer to the field, or just a generally interested party, a very good presentation of the developing technology and pedagogy within ICT in the classroom. An extra bonus is the historical perspectives given in many of the articles. Despite the few years that ICT has been established as a research area in language learning much work has been done, but in the past efforts were concentrated on what is called the technology-driven approach. The pendulum is now swinging towards the pedagogy-driven approach, and the general conclusion is that empirical research is much needed.

References

Bickerton, D., Ginet, A., Stenton, T., Temmerman, M. & Vasankari, T (1997). Final report of the RAPIDO project. Plymouth, UK: University of Plymouth (SOCRATES Project TM-LD-1995-1-GB-58).

Gagnepain, J. (1982, reprinted 1990). Du voloir dire: traité d'épistémologie des sciences humaines. Tome I: Du signe, de l'outil. Paris: Livre et Communication.

Holec, H. (1988). Autonomy - self-directed learning: present fields of application. Strasbourg: Council of Europe.

Johns, T. (1991). Should you be persuaded: Two samples of data-driven learning. ELR Journal, 4, 1-16.

Little, D. (1991) Learner Autonomy 1: Definitions, issues and problems. Dublin: Authentik.

Little, D., Ushioda, E., Appel, M. C., Moran, J., O'Rourke, B. & Schwienhorst, K. (1999). Evaluating tandem language learning by e-mail: Report on a bilateral project. CLCS Occasional Paper No. 55. Dublin: Trinity College, Centre for Language and Communication Studies.

Widdowson, H. G. (1996). Authenticity and autonomy in ELT. ELT Journal, 50(1), 67-68.

Wolff, D. (1993). New technologies for foreign language teaching. In Foreign language learning and the use of new technologies (pp.17-27). Brussels: Bureau Lingua.