Learning to Learn Languages with ICT - But How?
The huge potential of ICT (Information and Communication Technology) in foreign language teaching and learning barely needs restating—indeed it has become almost ubiquitous in many situations. However, it is often promoted for the wrong reasons at all levels, from governments to institutions to teachers, and despite the plethora of research, results as far as learners are concerned are often disappointing: ICT does not necessarily increase motivation, and learners tend to lack sufficient autonomy to make the most of it without specific training. This paper retraces a number of our own experiences with ICT in a variety of contexts in order to examine what goes wrong when learners are left to their own devices. It then discusses possible ways forward in integrating ICT fully into language learning courses.
Keywords: advising, autonomy, CALL, ICT, language learning, learner training, learning to learn, isolation
From CALL to ICT, teachers and researchers have long been interested in uses of new technologies in foreign/second language teaching and learning, and not without reason. ICT is often argued to increase motivation (Chateau, 2005), and is frequently linked with learner autonomy (Duda, 2005)—autonomy defined by Holec (1981, p. 3) as "the ability to take charge of one's own learning". In other words, learners should be able to take all the decisions concerning their learning: determining targets and objectives, choosing contents and materials, selecting methods and techniques, organizing their learning, and assessing their progress. ICT thus seems to befit autonomy because it is claimed to provide greater freedom and flexibility to learn at one’s own pace and convenience, whether within the context of a language course or beyond. It can contribute to lifelong learning, perhaps most apparent in use of the internet, which represents a worldwide library and resource centre where information can be accessed fast and efficiently: learners can thus benefit from a vast range of authentic language materials and resources which can be selected according to learning needs, aims, styles, strategies and preferences (Richards, 2005). Furthermore, information about the target language and culture may favourably influence intercultural representations, while the interactive potential provides opportunities for flexible communication via e-mail, chats, forums, and so forth. This may be in "authentic" environments not primarily aimed at language learning, though clearly the implications for distance or blended language learning programmes are enormous, with opportunities for e learning and interaction between learners and also with teachers (Mangenot, 1998).
Our university, Nancy 2, like many other French higher education institutions, promotes the use of ICT in various areas, including language teaching. Among the tools and facilities we have recently been involved with are Langues-U, Etudes et Ressources Universitaires à Distance at Nancy 2 (henceforth ERUDI), and Ecouter pour Comprendre (EPCO):
These projects and many others have been granted considerable logistical and financial resources because of the perception that, as Chapelle (2005, p. 78) puts it, "technology should be able to afford the learner more appropriately individualized instruction than what can be achieved through classroom learning". But, and this is crucial, "should be able to" does not necessarily entail "actually does". In other words, ICT has the potential to improve the learning situation, but this needs careful thought in each case as there is no essential link with improved learning (Benson, 2001). Our purpose in this paper is thus to understand why learning languages with ICT is not always the panacea some thought it would become in the 1980s (Antoniadis, 2006). As has frequently been mentioned (e.g. Mangenot, 1998; McCarthy, 1999), ICT is not suitable for all purposes for all learners in all situations, and may require some considerable learner training for effective use. This paper, after a summary of the background of the research, examines possible reasons for these difficulties and discusses a number of potential ways forward.
With the encouragement of governments worldwide, more and more students are enrolling in courses which incorporate e-learning. However, "pure" e-learning courses are rare—an OECD report (2005) puts the figure at "well under 5%"—and in many institutions even blended learning using ICT is unusual. Furthermore, in many cases it remains an open question whether increased use of ICT has actually changed the way teachers teach and learners learn, or if it is "simply a case of students typing up their essays on computers and professors sending them course reading lists or work assignments by e-mail" (p. 3). The situation at Nancy 2 is perhaps fairly typical, as even many teachers are uncomfortable with more than the most basic uses of ICT, and the vast majority of students still have virtually all if not the entirety of their tuition in the classroom. Even where the most ardent teachers present ICT in a highly favourable light, the feeling persists that few students seem ready to work with ICT spontaneously.
While it is unlikely that disillusion is an entirely French phenomenon, there may be some aspects of ICT which do not easily lend themselves to French culture. In particular, the French school system focuses on language rather than learner, product rather than process, performance rather than mastery, extrinsic rather than intrinsic motivation (Bernardini, 2001). Language is either right or wrong and the teacher is the ultimate expert on this; her role is to dispense that knowledge, while the learner is a passive unit to be taught, to memorise and reproduce knowledge. This is of course a caricature, but there may be relatively little space in traditional French educational culture for focus on the learner as an individual, for autonomy, and for the construction of knowledge in a personalised way compared to some other cultures. At the end of the day, the only question is, "how can I pass?" The common perception that a university, department or teacher in France has to have a high failure rate to be taken seriously is totally at odds with the Anglo-Saxon model. Boulton (2006) presents a similar argument as to why data-driven learning has as yet made little impact in France, but the implications clearly apply to a wider range of practices in language learning.
The rest of this paper examines use of ICT in three different settings at our university. Even though the tools were designed explicitly to help students become autonomous and work on their own, and include means of communicating with tutors in order to solve possible problems, learners seem reluctant (or unable) to take full advantage of them. Our aim is thus to try and understand why, starting with following hypotheses:
It seems necessary then to start by determining learners' needs and objectives and identifying their learning strategies in order to select relevant ones and provide suitable materials and activities.
Before we could address its sources we needed to identify the problem itself in detail. To do this we studied three different populations of students enrolled at our university:
Through these large samples it is possible to obtain insights into three different populations, each with its own aims and needs—one group needing English for Specific Purposes, another being English majors, the third seeking only to improve their chances of passing their end-of-year competitive examination. At the same time, they represent a relatively homogeneous group in many respects: they are learning English in higher education, and nearly all have passed through the French education system and thus have comparable backgrounds in languages and ICT, not to mention similar cultural representations towards ICT, English, languages, and language learning. Our study is firmly based in the tradition of action-research (Chateau, 2003; Chateau & Georges, 2002; Riley, 1996), and relies on past and on-going studies carried out with these learners. The investigation methods are both qualitative and quantitative, involving notably questionnaires, interviews, surveys, teacher-learner feedback, and so on.
For the first time last year, MA psychology students at Nancy 2 who need to read specialized articles in English by the end of their first year were given the opportunity to experiment with a flexible system partly based on Langues-U during the first semester. Students also had to spend a minimum of one hour a week in the university's language resource centre (for a description of this centre and its possibilities for language learning, see Gremmo, 2000). The teachers in charge of the system were present in the centre regularly and the learners who wished to do so could come and ask them for advice. The aim of this system was to help them improve their skills in English in order for them to be able to follow a course in English for psychology in the second semester. The objective of the course (designed in collaboration with colleagues in the psychology department) being to improve production of specialized abstracts in their discipline, they clearly had to brush up their English before the start.5 The flexible system combining work on Langues-U and in the resource centre thus seemed the best way to guarantee that they had reached a sufficient level by the end of the first semester. Langues-U indeed offers a number of thematic files composed of authentic video, audio and written documents focusing on the same subject (e.g. education, the environment, technology, etc.), accompanied by pedagogical activities. More "traditional" activities such as vocabulary and grammar exercises are also present on the digital campus. Learners thus had the choice of a wide variety of subjects and tasks that could, we hoped, appeal to all. Although we had decided not to administer a formal levels test, the students had been told that both the number of hours they spent in the centre as well as the number of connections to Langues-U would be checked, the campus offering the possibility to access students' individual work files, that is, to keep track of the activities worked on. At the beginning of the second semester, after six months' work on the digital campus and at the resource centre, they were asked to fill in a questionnaire to provide feedback on their overall impressions. We also examined the way they used the digital campus and the type of activities they focused on.
The results of this study are detailed in another paper (Chateau & Nussbaumer, 2006), and reveal a number of interesting features:
The general impression emerging from this first population might seem to be rather negative; however, it is important to note that all the psychology students who had already worked on line before (nine) found the activities offered by Langues-U very varied and work on the campus useful. The frequent connections made by some students suggest that they appreciated the digital campus, with 31 of them connecting more than 30 times, and six more than 50 times (the record being 79). Furthermore, the number of students finding the activities very varied correlates with the number of students who contacted the tutors (41, see above). This is thus probably an indication that the difficulties encountered by some of our learners are due to their not asking for advice in spite of their lack of familiarity with ICT and with autonomous work, rather than to a flagrant misconception of the tool.
The typical profile of ERUDI students is perhaps rather atypical of many other student populations: 25 to 40 years old, female, working at least part time, and very often with a family to look after. These are intelligent adults who have missed out on the possibility of a full university education immediately after school for socio-economic or other reasons, but who later wish to resume their studies. For them to achieve this, our philosophy is to keep obligations to a minimum and to open up our courses to as wide a population as possible, but this is not without important implications, notably of geography, time, money, and ICT:
This brief context is necessary to underline fundamental differences with programmes such as at the Open University (OU). While our ultimate aims may be the same, the OU provides all materials, sets regular deadlines, imposes a fixed learning path, requires attendance, and is comparatively expensive; in other words, OU students are allowed little room for autonomy (Hurd et al., 2001).
Previous studies at ERUDI have highlighted a number of specific difficulties, not the least of which is lack of contact. We have already noted poor attendance at group meetings, but Boulton and Booth (2001) report that very few ERUDI students have any significant direct contact with teachers by whatever means, including e-mail, telephone, fax, post, forum, in office, and so on, despite regular encouragement to do so. For many of them, it seems, they access their courses, and then go it alone. Explanations for this behaviour remain speculative and indeed puzzling, as for 84% of them (Boulton, 2005a), the context of distance learning is a constraint dictated by their individual situation rather than a learning preference. There is a similar lack of interaction with peers at ERUDI, which has also been cited as a major disadvantage of computer-mediated learning (Felix, 2000), leading to a feeling of social isolation which is largely responsible for the high drop-out rates at university (May & Bousted, 2003), especially in distance education (Cookson, 1989).
Each course includes extensive discussion and advice presenting different learning techniques and encouraging learners to experiment to find strategies which correspond to their own individual needs, preferences and learning styles. But again, although the opportunity is there, it seems that very few students actually avail themselves of it: they are clearly impatient to get to grips with the content of their courses and feel that time spent reflecting on the process is time wasted (White, 1995). As Hurd et al. (2001, p. 342) point out, it is not enough merely to provide the support: "equally important is the ability each student has to make the most of that support."
These issues of isolation and learning-to-learn are reminiscent of Chamot and O’Malley's (1994) socio-affective and metacognitive skills, both crucial for effective learning especially in a distance context. Since, as we have seen, the courses themselves do not seem to fulfil our students' needs adequately, we have created a forum with the aim of creating a "virtual community" and encouraging exchanges of good practice. It is potentially a major tool in our students' armoury, and is far more widely used than in any other discipline at our university: of the seven forums on the university website in 2005-2006, ours accounted for nearly 95% of all messages, a first indicator of success.
A total of 512 threads were posted over the year on the ERUDI forum (excluding spam). On average, each thread received 3.3 responses, but 152 threads (nearly one third of the total) received no responses whatsoever. One exceptional thread received 81 answers, and it is perhaps worth noting that it had the general title of "let's get to know each other". Traffic peaked early in the year (October to December) and at the start of the second semester (January and February) when the first piece of marked homework was returned. After this the numbers trailed off, stabilising at around 30 new threads per month from March to July, although we might have expected the imminence of exams to provide substance for discussion. In all, the forum contained 1680 messages (threads plus responses). If this does not sound enormous for a student population of 690, it is clear that the site receives far more visitors than contributors: the total number of hits for the forum over this period was 274,419, an average of 535 per thread; the lowest number of hits for any thread was 110, while the highest was 7,290. A small number of these may be attributed to outside visitors, and each student may have been responsible for more than one hit; nevertheless, the figures support informal feedback that many students visit without contributing.
Although students were not required to identify themselves to use the forum, only 34 of the threads were initiated anonymously; 182 separate identities were responsible for the remaining 478, although clearly it cannot be ruled out that a single person used more than one identity. Of these, 112 (over 60%) initiated only a single thread, while the top 10 contributors accounted for 182—over a third of the total. These most prolific contributors included four of the five teachers on full-time posts—the fifth giving only a single announcement. All teachers combined posted 110 new threads, representing just over 20% of the total; allowing for the occasional administrative and visitor posting, a conservative estimate is that three quarters of all threads were student-initiated. However, it is clear that the vast majority of students did not post any new messages of their own: even if all the anonymous postings were from different individuals, this still leaves 474 of the 690 students not starting a single new thread, well over two thirds.
To complement these data gathered directly from the forum, a questionnaire was sent out to all students along with their exam enrolment form; disappointingly, only 24 were returned, so any conclusions can only be extremely tentative. However, this in itself underlines the difficulty of non-obligatory communication with distance students, especially at undergraduate levels: although the MA students make up only 12.5% of our students, they provided exactly half of the completed questionnaires. This might suggest that more advanced students are more at ease with ICT, feel less threatened by interaction with teachers, or have more to say, perhaps because they have the pedagogical maturity to think more carefully about their own learning. Indeed, the MA courses are more likely to demand higher levels of autonomy, reflection and ICT-awareness than at lower levels. Not for nothing is an ERUDI course requiring the production of a research paper in corpus linguistics reserved for MA students (described in Boulton, 2006; Boulton & Wilhelm, 2006).
Although the questionnaire was sent by post, all the respondents have frequent access to the internet, mostly at home, yet the majority receive their courses on paper, suggesting they are by no means entirely technophile. Nor are they all devout forum users: although they all claim to visit the forum at least occasionally (and half of them at least once a week), a fifth never start a new thread, and nearly a third never answer other people's messages (Table 1). Despite this, it is reassuring that 19 find the forum "helpful" in general, only 4 disagreeing.
|at least once a week||at least once a month||less than once a month||never|
|to post a message||0%||20%||60%||20%|
|to answer a message||5%||25%||40%||30%|
|to visit only||50%||29%||21%||0%|
Table 1: Forum Use
In a series of open questions, the majority of respondents claim the forum is most useful for study-related topics. Specific examples given include peer exchanges of good practices, past experiences, useful websites and resources, learning tips, homework feedback, discussion topics, and so on. In this, it seems that the aim of the forum to encourage reflection may be partly effective, at least with some students.
As regards the socio-affective aspect, most respondents find the forum enables a positive form of human contact, which can reduce the feeling of isolation and help them to feel part of a community and to realise that they are not the only ones experiencing particular difficulties, thus boosting motivation. They particularly appreciated the informal atmosphere of the forum both between students and with teachers, important for reducing anxiety and encouraging input (Kern, 2006, p. 18). Indeed, with regard to computer mediation, it seems that "the distance itself seeks its own antidote: the greater the distance, the more important it is to feel close" (Bourdet, 2006, p. 35). This may also partly explain why the vast majority of messages (over 80%) were in the students' mother tongue: it seems that writing in a foreign language is potentially threatening for many even at undergraduate level, and outweighs the advantage of language practice. Indeed, Boulton's earlier (2005a) study found only 4% of students claiming they made significant use of the internet to learn the language itself: overwhelmingly they used it as a source of information gathering.
The data so far suggest that the ERUDI forum can contribute to learning-to-learn in the form of peer exchanges as well as reducing feelings of social isolation. At the same time, these two main functions seem to be incompatible on occasion. For example, non-study related topics were most frequently cited as the least useful, especially in the form of personal exchanges from which individuals felt excluded. While these responses came from a minority of students, it does suggest that the socio-affective dimension of the forum is not fully appreciated by all. Similarly, there is a split between students who like the occasional light-hearted postings, and those who find them a waste of time. However, in a closed question asking if they liked the "mix of study and non-study related questions" on the forum, 17 agreed and only seven disagreed. Earlier attempts to separate the two functions proved unworkable, as a study-related thread would frequently elicit a number of non-study-related responses. Furthermore, students do not necessarily appreciate attempts at control of what is, after all, "their" forum. The only answer would seem to be to allow the mixture which they initiate themselves; those who object are under no obligation to read messages they find less relevant.
Students' own recommendations for ways to improve the forum largely revolve around increasing teacher involvement, especially as a source of study-related discussion issues. When asked explicitly if teacher contributions to the forum were "a good thing" or not, one student declined to answer, but the others were unanimous in approving. Another recommendation was to encourage more students to participate—ironic, perhaps, given these very students' limited active participation, the lack of uptake of other opportunities for contact (such as personal e-mail and the group meetings), not to mention the lack of return on this questionnaire.
All the first-year students at the IUFM de Lorraine are given the opportunity to prepare the CLES with Langues-U. Out of the 300 or so students, 51 subscribed to the site, but the majority used it less than half a dozen times, if at all. We sent an e-mail to these students asking them the reasons why they had given up on Langues-U. Only six students replied: five claimed they had not had the time to work on the CLES, and one explained that as she already had a degree in English, she felt her time was more fruitfully spent on her other subjects. Obviously, among other reasons, success of distance learning platforms depends on how much students feel they need what is on offer, especially when they have to cram in knowledge and skills in the numerous subjects that make up the curriculum for future primary school teachers.
The fifteen students who had actually used the site more than six times were also contacted by e-mail asking them if they would agree to an interview, or to answer a questionnaire. Only two answered, one accepting the interview and the other returning the questionnaire. Such a small number may be accounted for once again by the students' lack of time, but, as at ERUDI, it is likely that many found the idea potentially threatening—providing feedback to a teacher who might be part of the assessing team in the second year. In the light of the paucity of feedback, these two sets of answers are treated here as case studies. Even if the results cannot be generalised, we may see them as representative of at least some students.
The student interviewed (S1) logged on to the site 18 times, which is quite high for this population. She was happy to receive e-mails from the teacher, which she felt added a human dimension to a potentially impersonal learning scheme which might otherwise lead to a loss of motivation. This need for teacher-initiated contact seems to be even greater at the beginning of a course. S1 remarked on the difficulties students had using the computers at the IUFM, which could partly explain why many students opted out of the computer-assisted learning scheme. Fortunately for S1, she had the technological advantage of having her own computer at home, which is not the case for all.6 Nevertheless, her computer and methodological skills were shaky, making it difficult for her to navigate ("difficult to find the links", "it was complicated, I didn't really find my way around", "I'd have liked someone to explain to me how it worked; where to go, how to find the links"). Admittedly, the Langues-U interface might be partly to blame, as it is always possible to make software more user-friendly. Nonetheless, S1 expressed satisfaction with the learning activities in the software, which correspond to the skills that are assessed by the CLES; furthermore, she also saw language learning as something more than just grammar exercises and vocabulary lists. We may assume that with more time, support and a more user-friendly interface, she might have become a successful autonomous learner.
The forum itself was not a prime attraction, as the moderators were disappointed to find out: nobody used it. S1 did not even know it existed until a specific e-mail was sent out to all, but when she logged on she found it empty. Not daring to post the first thread, which would probably be most French students' attitude, she did not visit it again.
Fewer details are available from the student who filled in the questionnaire (S2), but her answers were overwhelmingly positive. On the whole, she seems to find computer-assisted language learning as efficient as, if not better than, teacher-directed lessons. Despite this, she stopped using the software after only 10 occasions. In the absence of an interview, we can only speculate on the reasons for this, but it seems likely that, as at ERUDI, lack of human contact is partly responsible, as well as the lack of learner training which students are rarely prepared to undertake voluntarily.
The global results at this stage allow us to draw the following tentative conclusions:
These results support the hypotheses on which our study was based. They also seem to indicate that distance learning programmes need to be as user-friendly as possible (Boulton & Booth, 2001). There is, indeed, little point in presenting learners with valuable resources if they cannot or will not use them.
These general observations indicate certain deficiencies in the tools themselves, but more importantly they underline the fact that ICT is not a panacea, is not suitable for all learners in all situations and for all purposes, and may require some considerable learner training for effective use. The freedom ICT provides may in fact be counterproductive as regards learners becoming autonomous. As Hurd et al. (2001, p. 343) remark:
Those unaccustomed to reflection in any aspect of their lives may find it difficult to accept this link between self-awareness, strategic competence and effective learning, and may well resist it if they are not convinced of the so called benefits and relevance to themselves as individual learners. Most language learners want to get on with it, to see rapid results. They are prepared to work hard, but may need some convincing that effective learning is not just about following instructions from teacher or book and doing exercises.
In other words, it seems that some form of constraint may be necessary before learners can actually become efficient autonomous learners.
To summarise the situation, we are faced with students who are reticent to use ICT because they do not know how to use it efficiently, and are reluctant to find out because they think the process is uninteresting, unimportant or time-consuming. This is even the case to an extent with EPCO, which was designed and marketed explicitly as a methodological tool for use without a teacher: the feedback we have had so far, however, shows that many users were impatient with the learning-to-learn approach on a DVD-ROM and would have liked more traditional content (Boulton, 2005b).
One partial response may be found at a nationwide level, with recent plans for all French schoolchildren to be trained in ICT (http://www.eduscol.education.fr/B2i). However, it will take a few years for the effects to be felt in higher education, and even then there is no guarantee that future students will be happy to work in this way: Boulton's previous (2005a) study at ERUDI showed that most students prefer to follow their courses on paper for a number of reasons which are unlikely to disappear with the new training. For the present, in any case, we need to devise new methods to improve the success of our distance, autonomous or blended learning programmes.
The solution is clearly not just to present large quantities of methodological advice—the problem with EPCO. Because of their representations of ICT, of languages and of language learning, many learners simply skip such methodological information: they know their destination, but are unprepared to consider novel ways of getting there (Boulton, 2006). Motivation is evidently a key area here, correlating strongly with language learning (Chateau, 2005; Dörnyei, 2001). Given this background, it is perhaps not surprising that a sudden plunge into autonomy is unlikely to succeed and might be perceived as a lack of interest on the part of the teachers.
However, we remain convinced that the solution lies in learner-constructed knowledge rather than teacher-transmitted knowledge (Esch, 2002). This implies greater flexibility for individual learning styles, preferences and so forth; a process of autonomisation rather than an assumption of autonomy; greater interactivity, especially in demonstration rather than explanation; and improved human contact (Holec, 1981; Little, 1991). This contact can be between learners who have much to share with each other, but of course the learner-teacher interaction is also essential for improved advising (Azzam-Hannachi, 2005; Chapelle, 2005). If "learner autonomy in institutional settings inevitably requires both teachers and learners to modify their representations of their respective roles" (Riley, 1989, p. 70), especially in letting learners decide individually how they work and what resources they use, such modifications may be more difficult in some cultures, such as in France, than others.
So how can we encourage our students to see the need for learning-to-learn through ICT? Perhaps one way would be to tempt them into spending time thinking about their own language learning, in other words by exploiting their own egocentric curiosity. If we begin by helping our students identify who they are as language learners, they might start thinking about language learning rather than just trying to learn the language. This has been done quite deliberately in a number of books promoting learner autonomy (Ellis & Sinclair, 1989; Narcy, 1991), and there seems little reason why the same could not be done using ICT. However, deciding to start with this would mean that the teacher or the on-line programme suggests a sequence, which might be perceived as going against the principles of learner autonomy. But if we see learner autonomy as an aim, anything that can lead to it ought to be acceptable as long as the learner does not have to submit blindly to authority. Once she knows what type of language learner she is, she might be helped to decide what she needs and how she can learn it. This is where human contact is most needed. Our various informants stressed how much they appreciated teacher-initiated contact. We therefore need to develop this and make sure that contact is made at regular intervals—but again, the question is how to achieve this without making it compulsory and penalising students who choose not to use it. It may be hoped that in the long run more students will feel that they can initiate contact freely, as a small number are already doing at ERUDI.
The form this contact might come in is not necessarily the same for everyone or for every institution. The SCELV and the IUFM have blended-learning programmes which leave room for one-to-one encounters or small group meetings to discuss learning processes. It has thus been decided, for example, that next year the psychology students will have to attend a compulsory meeting at the beginning of the university year. The teachers in charge will then present the flexible system and provide a live demonstration of Langues-U to explain all the possibilities of the digital campus, ensuring that everyone understands the way the tool functions and how they can obtain information if necessary. ERUDI, on the other hand, can only rely on distance communication. The forum at ERUDI has already taken off, mainly thanks to the efforts of the director M. Nussbaumer, and a small number of teachers who regularly respond to messages and add their own. Collaborative language learning can also be implemented within the three institutions: through synchronous chats, with or without a webcam, or asynchronous exchanges via e-mail, students can co-construct their language learning skills. In its new blended learning system, the IUFM has set up a tandem programme with the University of Minnesota7 which has enabled 47 students to interact with their foreign counterparts. This is considered a way to offer social interaction which leads to linguistic and cultural development but also, one may hope, metalinguistic development since the learners may discuss their various learning styles, preferences and strategies, just as on the ERUDI forum. However, as recordings of some of these exchanges show,8 without imposed topics the students often do not know what to talk about and find themselves trapped in an artificial exchange that does not last long. Collaborative tasks have to be set to help create relationships, but of course they might once again be seen as a form of constraint on learner autonomy. However, the student who feels she can learn better without these tasks could do so. Scaffolding (Vygotsky, 1978) must only be seen as an option for those who need it, but as our studies show, not many of our students are ready to direct their own learning unaided. Leaving them to study entirely on their own would surely go against the principles of autonomous learning, as it would create inequalities between those who can be autonomous and those who will always have to rely on someone else to direct them (Barbot, 2001). Our programmes must therefore be designed in such a way that they will gradually help the learner gain increased autonomy. Some learning styles might indeed be less likely to accept self-directed learning (see Willing, 1987, on authority-orientated learning styles), but we do not know for certain that they do not evolve.
The study shows that our learners are not convinced of the value of self-assessment, but strongly believe in being assessed by others; this is certainly a reality for many students. Once again, training is needed if they are to assess their own learning process and performance (in fact, it is not incompatible with external evaluation as, in the long run, it should help the learner attain imposed objectives). This might be done gradually, starting with dual assessment (conducted by the teacher and by the learner), or peer assessment (working in pairs or groups to assess each other's productions), so that little by little, the learner might realise that she is capable of evaluating how she learns, what she knows, and how successful she is with different techniques. This underlines once again the importance of contact between teachers and learners, which is why language-learning with ICT tools that do not allow for some sort of contact might only be of use to the already autonomous leaner. It is indeed the responsibility of universities and teacher-training colleges, as well as primary and secondary schools, to lead their students and pupils towards autonomy in order to prepare them for lifelong learning. This line of reasoning implies that teachers must learn to become facilitators and open their classrooms to ICT, but it certainly does not mean that we need to replace teachers by more economical machines. The teacher has the role of helping the learner to learn how to learn, and may be considered as a facilitator of learning (Ellis & Sinclair, 1990; Voller, 1997; Kohonen, 2003), guiding students with different learning styles towards their desired learning goals. The facilitator is then to develop the methodological rather than linguistic aspects of learning, and provide suggestions on how to improve vocabulary, grammar, skills, and so forth (Embi, 2004; Ramon Monferrer & Alvarez Plareto, 2006).
Nonetheless, whatever we suggest our students do will be doomed to failure if we keep on believing that learners will do it voluntarily and spontaneously. Like us, they have tight schedules and, like us, they do first what seems most urgent; if the target language is not perceived as urgent, it will always be put off until tomorrow. Keeping a trace of their learning activities and setting deadlines might be one way to increase the perception of urgency.
Although we seem to be suggesting here that autonomy be "imposed" on learners, we are quite aware that this in itself is an oxymoron (Boulton, 2005a), and only applies in very general terms. As discussed earlier, autonomy is linked with individualisation, which means that each individual has her own learning styles, preferences, strategies, motivations, attitudes and needs. Information, explicit training, consciousness-raising and self-assessment,9 in addition to an introductory needs analysis, could all serve to identify various paths which might be followed to develop learner autonomy and learning-to-learn ability as regards both languages and ICT. Toronto University has devised a distance learner-training programme which offers interactive tasks in both fields (Barbot, 2001). We in Nancy hope to work in this direction.
Alex Boulton is senior lecturer in English and linguistics at ERUDI Nancy Université. Current research interests at the CRAPEL focus on use of ICT in language learning and teaching; corpus linguistics and data-driven learning; distance learning, learning-to-learn and autonomisation in language learning.
Anne Chateau is senior lecturer in English for Specific Purposes Nancy Université, teaching non-specialist students at the SCELV. Besides ESP, her research interests at the CRAPEL include blended learning systems as well as language learning with ICT.
Myriam Pereiro is senior lecturer in English at Nancy Université and involved in the pre-service training of teachers of English at the Lorraine IUFM. Her current research interests at the CRAPEL are: teacher training for foreign language teachers, language learning with ICT, English as a lingua franca.
Radia Azzam-Hannachi is a senior lecturer in French as a second language at Bretagne-Sud Université. She formerly taught English as foreign language at the Lorraine IUFM and Nancy Université. Her research interests are teacher training for foreign language teachers and the use of ICT in language learning.