by the P
Phil Benson University of Hong Kong [email protected]
This conclusion picks up the three main themes that have emerged out of theAnthology for me— defining autonomy, relationships between theory andpractice, and relationships between research and writing in ourfield. I connect these themes to an argument for more empirical research that I made in previously published work and end up in a state of self-critical uncertainty that is only made tolerable by copious quantities of metaphorical Bacardi. At theend of thechapter, I suggest thatwhat I have most got outof the Anthology is a better understanding of what empirical research really means whenautonomyis its object. m%x\zz\0mirmzmLx&tf%m*tzf--^£m%.o.^n\z, saniro^a. mmtm
momm, %tz^(omf\zm%mntn\m» mtzmmtz&±
#, r^|Efi<]j t\tmM\zfa*1kt<,
A u t o n o m y You Aski
Just as I was beginning to write this chapter, a title for it arrived in my e-mail: 'A Bacardi by the Pool.' It came in a provisional table of contents sent by the editors, and I quickly saw how it connected with the titles for the three main sections of Autonomy You Ask!: 'Charting the Waters,''Swimming through the Currents,' and 'Exploring the Deep.' Now, I do not know whether the editors seriously intended that I should use this title or not, but I decided that I liked it. All I had to do was to work out how it could possibly be related to what I wanted to say in my chapter! Over the next week or so, however, I realized that I had another problem: What did I want to say in conclusion to this book? LikeTim Murphey (Chapter 1), I was interested in expressing some of the 'resonant interconnections' that I saw between this Anthology and my own work. But how exactly would I do that? Would I review and comment upon each chapter? Should I react to the points that I had highlighted with marginal question and exclamation marks while reading the chapters? Should I be critical or supportive, critically supportive, supportively critical? Should I attempt to synthesize the main points of each chapter into one single compelling argument on the 'way forward for autonomy'? None of these options seemed quite right because, I realized, they all implied some kind of evaluation. Just as Tim, for example, begins from his recent work on group dynamics and looks forward to the ways in which themes related to it emerge as the Anthology progresses, I could look backwards at AutonomyYou Ask! from the perspective of arguments presented in my own recently published book, Teaching and Researching Autonomy in Language Learning (Benson, 2001). But the distinction between looking forwards in an introduction and looking backwards in a conclusion presented me with a problem. How could I avoid the trap of presenting my own previously published arguments as 'fixed positions' from which I now evaluate everything else that I read? It was in the midst of this conclusion-writer's angst, then, that the title I had received in my e-mail came to the rescue. This title, I thought, could have less to do with the content of the chapter and more to do with a particular style of writing. I would try to write this conclusion in a Bacardi-by-the-pool kind of way Writing in a Bacardi-by-the-pool kind of way could, of course, mean more or less anything! But at that particular moment in time it suggested to me a way of writing that would allow the ideas that had occurred to me as I read the manuscript to flow more freely on to the page—away of concluding the Anthology by,perhaps, telling the story of my reading and how it has influenced my thoughts about certain aspects of research on autonomy. In particular, I hoped that it would be a way of writing that would put what
Teaching andResearching Autonomy in Language Learning has to sayabout research'into motion.'
In contrast to Tim's opening chapter, then, which identifies resonances mainly in terms of the focus of research into autonomy (especially in its individual and social aspects), this
concluding chapter will mainly be concerned with what I can best describe as the 'nature' of research into autonomy. This Anthology is a collection of action research reports, but, unlike other collections of this kind that I have come across, this one constantly questions and subverts the 'action research paradigm' in ways that suggest a fundamental tension between conventional stylesof action research and our particular object of research, autonomy. So, with a metaphorical Bacardi in hand and the sun setting overthe pool, I have decided to begin by reflecting on what I wrote about action research in Teaching andResearching Autonomy in Language Learning. I will then pick up themes relatedto the nature of research on autonomy that emerged from each of the three sections of Autonomy You Ask! and explore my thinking about them in the light of what I have just read. These themes are: (a) defining autonomy (should we or shouldn't we?), (b) putting theory into practice (or practice into theory?), and (c) researching (or writing?) autonomy. — LD SIG 2003 A n t h o l o g y —
Bacardi by t h e P o o l
Point of Departure
I want to begin with Teaching and Researching Autonomy in Language Learning because it was intended to be a book about 'action research' and, second, because my thoughts on some of the issues raised in the book have changed considerably since I wrote it. I hope that this opening reflection will give readers some idea of how they have changed, and, therefore, of where my comments in this chapter are coming from. I also hope that they will help to establish the three themes that I want to look at later as areas of considerable ambiguity within the field. First, a little background. Teaching and Researching Autonomy in Language Learning was written for a series called Applied Linguistics in Action at the invitation of the series editors, Chris Candlin and David Hall, who consistently, and helpfully, emphasized the importance of maintaining a degree of critical distance from my subject matter. I make this point because there is often an expectation that books about autonomy will advocate autonomy. Chris and David, however, did not exactly want a book that advocated autonomy. Instead, they wanted a book that would evaluate what we knew about autonomy on the basis of research—one that would, in a sense, draw up a 'balance sheet' that would be useful in future research.
Now, in order to do this, I had to read and re-read the literature on autonomy in a particular way. Where I had previously been interested in what I could learn from this literature, without much regard to its status as 'research,' I was now evaluating it as research. The results of this evaluation were summed up towards the end of the book: In its broadest sense, research is a process of inquiry in which answers are sought to questions of interest to the researcher. These answers may be sought through reflection, logical reasoning or analysis of data. A great deal of the research on autonomy to date has been based on reflection and reasoning. Often, researchers draw conclusions about the nature of autonomy and the practices associated with it from reflection on their own and others' experiences of fostering autonomy. Far less research has been based on systematic analysis of data. (Benson, 2001, pp. 181-2)
The implication of this was, of course, that we need more data-based research, and this, in part, forms the point of departure for my interest in this Anthology. If I were reviewing Autonomy You Ask!, I would probably want to say that it makes exactly the kind of contribution to research on autonomy that I was arguing for. But this is not the kind of thing that you say while drinking a Bacardi by the pool. What I would rather say is that the Anthology makes a much broader contribution to research and one that in many ways undermines the 'certainties' of the argument quoted above.
Now,the following point is going to appear to be critical of some of the contributions to the Anthology,but it is not reallyintended as such. In severalchapters, I have seen the author(s) doubt whether their chapters are in fact action research reports, and in two cases I have even seen them jettison their data altogether in the face of the problems of analysingit! I have also read a greatdeal relatedto the difficulties of authors 'finding a voice' or saying exactly what they want to say about autonomy within the constraints of the action research paradigm. I have been aware, in particular, of a tension between two apparently contradictory feelings: the feeling that we need more data-based research and the feeling that, at the end of the day,data-based research is not goingto be enough to tell us quite what wewant to know about autonomy. What I am describing here is, of course, a tension that I am currently feeling myself—one that has in a certain sense been 'validated' by what I have read in Autonomy You Ask!.
In the light of these feelings, then, I find that the passage from Teaching andResearching Autonomy in Language Learning that I have quoted above does not exactly express what I — Autonomy You
would now want to say about research into autonomy. Instead, I feel that it represents a view of research on autonomy that expresses a certain kind of truth, but conceals others that may be more important. In a field where we do tend to advocate our object of research (autonomy), standards of evidence and argumentation are inevitably at issue, because of the risk of finding what we want to find. The point I wanted to make was simply that we should not plead a special case for research on autonomy within the field of second language research. But I also see how the argument that research involves 'reflection, logical reasoning or analysis of data' already moves beyond conventional norms of second language learning research, which tend to require the collection and analysis of empirical data. I now also want to ask whysecond language learning research must necessarily involve empirical data. Why can't it be based solely on reflection and logical reasoning? I also want to ask whyreflection should necessarily involve logical reasoning. Conventional assumptions about the norms of second language learning research, it now seems to me, should not be taken for granted. Nor should we necessarily assume that research on autonomy should follow these norms. So this is what writing in a Bacardi-by-the-pool kind of way means to me: tolerating a good deal more uncertainty than I am accustomed to. The point is that although I am able to pose these whyquestions, I am not really able to answer them. I hope that I am going to find some answers—albeit ambiguous ones—by writing my way through this Anthology
Defining Autonomy (Should We or Shouldn't We?) The first section of Autonomy You Ask! is explicitlyconcerned with definitions of autonomy, so I do not feel that I need to argue for my first theme here. In fact, this theme continues throughout the collection, and I want to start off by noticing Naoko Aoki's (Chapter 14) comment that there is no single authoritative definition of autonomy, and that we should not see this as a problem. If I am getting her right, Naoko is saying that teachers' knowledge is contained within personal theories grounded in personal experience,and this means that we are each likelyto have our own unique 'definition' of autonomy which will change and fluctuate over time.
Well, I largely agree with Naoko on this point, but I have to say that the author of Teaching and Researching Autonomy in Language Learning would not. He would say that there is a single authoritative definition of autonomy in the context of language learning, and it goes something like this: 'Autonomy is the capacity to control your own learning.' If you like, you can substitute 'ability' for 'capacity' and 'take charge of or 'take responsibility for' for 'control'—these are matters of wording only—but no other definition quite captures what most people mean by autonomy. This is not to say that the definition is unproblematic. But the problem lies not in the definition itself, but in the fact that it is so broad. What exactly does 'taking control of your learning' mean? Clearly it can mean many different things to different people, and this is where Naoko's point comes into play. But is Naoko right to saythat this is not a problem? The author of Teaching andResearching Autonomy in Language Learning is not so sure about this either. He has just come across a few new and interesting potential definitions of autonomy. MiyukiUsuki (Chapter 2), for example, suggests that we should look for autonomy in the 'learner's internal functioning.' Eric Skier and Stacey Vye (Chapter 3) suggest that we shouldlook at learners' out-of-class experiences and talk about 'time, shame, and maturity' as factors in the development of autonomy. Tim Stewart (Chapter 4) talks about the tension between autonomyas'a set of studyskills' and as'a wayof being in the world,' while AndyBarfield (Chapter5) describes its development as'a dynamic interplaybetween the routine and the reflective.' And then wehave Michael Carroll and Ellen Head (Chapter 6) trying to pin down autonomyin a quite different wayin the complexspaces of'the curriculum.' What exactlyis happening here, then? First of all, aren't these potential — LD SIG 2003 A n t h o l o g y —
Bacardi by t h e P o o l
definitions all related in some wayto the ideaof control overlearning? And if they were not related to this idea, wouldn't that be a reason to reject them? (Imagine, for example, that I choose to define autonomy as'the ability to obeyinstructions'!) More importantly,aren't these potential definitions beingadvanced because their authors would like us to considerwhether the particular aspect of control over learning they identifymay be more important to autonomy than others? But having got that off my chest, there is a very important sense in which I now think Naoko is right. After reading through 'Charting the Waters,' I have a strong feeling that there are many ways of defining autonomy and that—as long as they all somehow link to the idea of control over learning—there is no real way of deciding which will be the 'right' definition. I also have a strong feeling that there are many ways for an individual to beautonomous. It all depends on the cultural context, the particular situation, the stage of learning, and, ultimately, the individual and his or her experiences. From this perspective, debating definitions of autonomy may have the particular purpose of highlighting the importance of these variations. The point is, however, that we should perhaps stop short of insisting upon the importance of one or another aspect of autonomy for all learners in all contexts or imposing one definition on a concept whose strength may well be its inherent ambiguity and multiplicity of meaning.
Putting Theory into Practice (or Practice into Theory?) In the second section of this book we shift from a focus on theory to a focus on practice. This shift in focus is at the root of my second theme—the relationship between theory and practice. This theme first occurred to me when I realized that the 'theoretical' chapters in the first section actually had a great deal to say about practice and that the 'practical' chapters in the second section had just as much to say about theory. What, then, was the difference between the two sections? Was it simply that contributors to the first section wanted to talk about theory while those in the second section wanted to talk about practice? Was I looking quite literallyat a difference in 'focus'—at the same set of problems viewed through different 'lenses'?
I find these questions interesting because they imply a quite different understanding of the relationship between theory and practice than we tend to find in other branches of second
languagelearning research. If we viewpracticeas theory translated into pedagogical action, as many second language researchers do, we might wellexpect the second section of this book to show how the theory of the first section can be 'put into practice.' But this is not what happens at all, first because the theory of the first section is rather diffuse and, second, because the contributors to the secondsection do not generally adopt a 'theory-driven'viewof their practice. Instead,they tend to identify contextual factors as the most important driving force behind their decisions on how to help learners become more autonomous. From these chapters I also get a strong senseof theory following, rather than preceding, practice. As a community of researchers, we seemto be just as likely to 'put practice into theory' as weare to put theory into practice.
Now, I hasten to add that although I do not mean this to be a criticism of the chapters in 'Swimming through the Currents,' the author of Teaching and Researching Autonomy in Language Learning probably would! When I talked about the need for more empirical research in that book, what I was really talking about was the need for the kind of connection between theory and practice that we find in other branches of second language learning research. This
also carriedover into the two major reasons that I gave for the needto define autonomyat the theoretical level:
Firstly, construct validity is an important precondition for effective research. In order for a construct such as autonomy to be researchable, it must be describable in terms of observable behaviours. Secondly, programmes or innovations designed to foster —
autonomy are likely to be more effective if they are based on a clear understanding of the behavioural changes they aim to foster. Put simply,whether we are concerned with research or with practice, it is important that we know and are able to state what we mean when we talk about autonomy. This is not to say, however, that we all should necessarily mean the same thing. Autonomy may be recognized in a variety of forms, but it is important that we are able to identify the form in which we choose to recognize it in the contexts of our own research and practice. (Benson, 2001, pp. 47-48) The important point in this passage was, I now think, contained in its last sentence. When we say that we are trying to help our students to be more autonomous (i.e., when we talk about practice), we often mean that we are trying to help them to do something quite specific —something that is connected with, but at the same time less than what we understand autonomy to be. And it seems to me that, when we come to evaluate the success of our efforts, it is important that we state exactly what that 'something' is.
Now, in 'Swimming through the Currents,' I did in fact find several clear examples of this principle in action. To take Emika Abe's contribution (Chapter 7) as one of those examples, I saw that Emika had a very specific aim in her project—she wanted to help her students speak more fluently and more confidently in English. I saw how she connected this aim to autonomy: Autonomy involves an ability to take advantage of opportunities to create and negotiate meaning in authentic contexts of interaction, which presupposes fluency and confidence. I also saw that, from Emika's point of view, the immediate goal of the project was something less than the more distant goal of autonomy. And all of this allowed me to see how the practice she adopted could, in spite of the fact that it was teacher-directed and highly structured, help the students become more autonomous. But I also found much more in this section to challenge the assumptions underlying my argument in the passage above. The main assumption underlying this argument is that research into the practice of autonomy is more or less equivalent to the evaluation of the success of its various forms. If we are able to clearly identify the goals of different forms of practice and their relationship to some overarching concept of autonomy, I thought, we will ultimately be able to evaluate their success and perhaps see how they might fit together. This argument is linked to a further assumption that we choose to engage in particular forms of practice because we believe that they will be more successful than others. I was assuming, in other words, that we are (or would ideally be) collectively engaged in a 'rational-progressive' enterprise of experimentation that would ultimately lead to an understanding of the best way of helping learners become autonomous. In the light of the accounts of practice, I have read in this book, it is this assumption in particular that seems to lack validity. These accounts suggest to me that the argument tends to fall down, in fact, in face of the fact that we do notgenerally choose our practices, because we believe they willbe successful in terms of autonomy. Jodie Stephenson and Miki Kohyama (Chapter 8), for example, explain how Miki started using languagelearning projects, because there wasa limit to what she could do in class to accommodate the various interests of her students. They also explain how she had onlytwo 90-minute classes per week, in whichshefelt that she had to satisfy the needs of the majority, and how she introduced the out-of-class projects in an attempt to utilize individual students' ideas. Lastly, they explain that Miki had heard about language learning projects from a colleague who had tried them out a few years earlier and had shown Mikiher old notebooks. The choice of languagelearning projects as a particular form of practice here appears, then, to have been determined by three major contextual factors: Miki's beliefthat meeting learners' individual interests is important in teaching, the constraints of her contact time with them and
— LD SIG 2003 A n t h o l o g y
by t h e
what she felt obliged to do with it, and the knowledge that she had of languagelearning projects from her colleague. And in that contextI find their concluding comment thought-provoking: "This research helped us to see exactlyhow the projects enabled students to become more autonomous." It seems that Jodie and Miki did not start out from a particular conception of what autonomy was (theory) and then work out a wayof working towards that conception (practice). Instead, they began from the practice, which perhaps seemed 'right' for that situation at that time, and only connected it to particular kinds of autonomy through the process of research. In their different ways, Joyce Cunningham and Wade Carlton's class newsletters (Chapter 9), Akiko Tagaki's Co-operative Learning project (Chapter 10), Peter Mizuki's presentation work (Chapter 11), Heidi Evans Nachi's investigations of self-assessment (Chapter 12), and Etsuko Shimo's portfolios (Chapter 13) all 'put practice into theory' in essentially the same way. Their contributions also point to a very different way of carrying out research into the practice of autonomy than the one I outlined in Teaching and Researching Autonomy in Language Learning—one in which we are less concerned with comparing the effectiveness of different forms of practice, and more concerned with the factors involved in the practical choices that we make and the ways in which those choices influence our varied conceptions of what autonomy is/could be. It also suggests the possibility that we might view the multiple meanings of autonomy as not just a contextual matter, but also as a matter of the 'expansion' of the concept through its theorization out of diverse forms of practice. We are, it seems, constantly looking for new ways of implementing the broad idea of autonomy, and each new way appears to add a little more to the meaning of the idea itself.
Researching (or Writing?) Autonomy Before I plunge into the 'depths' of the third section,I need to pour myselfanother metaphorical Bacardi and take stock of where this conclusionis taking me, which is turning out to be a quite different place than the one I originally expected to be at! What I expected to do wasto set positions that I encounteredin this book against positionsthat I adopted in Teaching andResearching Autonomy inLanguage Learning. What I did not expectwasto discoverthat my own positions were so much less stable than I thought they were. In particular,I am beginning to have serious doubts about the ontology, or 'existence,' of autonomy. In fact, I am beginningto wonder whether autonomy is not much more than a 'space'for the development of theory and practice with the field of second language learning—an object of research that isposited as a goal, because commitment to a goal can serve as a kind of'glue' to hold a variety of otherwise unrelatedpractices together.
The chaptersin 'Exploring the Deep,' of course, do littleto help me resolve these uncertainties. On the contrary, they accentuate them! Mike Nix (Chapter 15),however, does much to help me understand wherethey are comingfrom: the process of writing about autonomy itself. As I understandit, whatMike isboth arguing and exemplifying is that in order to research autonomy we have to write aboutit.And in writing about autonomy, we also write autonomy—abstracting it as a construct from the complex realities of the practice of teaching and learning. We couldsay, in other words, that we do not seek answers to questions about 'autonomy' in the same sensethat we seek answers to questions about 'listening,' 'vocabulary' or 'learning strategies.' Instead we seek answers to questions about teaching and learningthrough something that we callautonomy—a construct that is very much a product of the process of writing about it. The writing process is, at any rate,verymuch to the fore of'Exploring the Deep,' and Mike, Cath Malone (Chapter 16), Stephen Davies (Chapter 17) and Steve Brown(Chapter 18) each havetheir own particular wayof going about that process. What interests me most about their chapters though is the way in which their experimentations in writing (and I can add
— Autonomy You
BE N S O N
here that I wasjust a little miffed that the bottle of Bacardi had alreadybeen opened before I arrived!) implies experimentation with the process of research. Mike and Cathy, for example, constantly weave around their data without ever actually analysing it in the conventional sense of the word, while Stephen Davies discards his data altogether and Steve Brown even goes so far as to invent a little of his! Their discomfort with conventional forms of research writing, in other words,implies discomfort with the idea of empirical research as the only means through which we should seek understanding of autonomy. It is as if the possibility of multiple ways of representing data undermines the force of purely data-based argument, and, after reading Autonomy You Ask!, this is a sense of discomfort that I share. Conclusion
T o sum up my response to this Anthology, I would say that it has caused me to think hard about the argument for more empirical research in the field of autonomy that I developed in Teaching and Researching Autonomy in Language Learning. To put it another way, I have had to think hard about that argument in order to respond. A metaphorical Bacardi or two by the pool has made the state of self-critical uncertainty I have fallen into a little more tolerable, but it is a state of uncertainty even so. But I should also say that these uncertainties did not exactly begin when I began writing this conclusion. They have been with me for a long time, and there are other arguments in Teaching and Researching Autonomy in Language Learning—arguments about the difficulty of'measuring' gains in autonomy, for example—that tend to subvert its overall argument for more empirical research. Nevertheless, I must admit that when I first heard of the proposal for this Anthology a year or so ago, I was interested in it because it seemed that it would be based on the kind of empirical research I was talking about. For me, then, the significance of the Anthology as the product of that proposal lies mainly in the waysthat the contributors have confronted problems in the treatment and representation of empirical data that they have encountered along the way In particular, I am beginning to understand that my uncertainties really revolve about the meaning of the word 'empirical' in the phrase 'empirical research.'
My argument for more empirical data was,I think, based on a feeling that it is too easy to distance oneself from the realities of teaching and learning when writing about autonomy. Mike Nix expresses the same feeling, I think, in Chapter 15. But, at the same time, I may have too readilyaccepted the idea that the solution to this problem lay in a move towards the kind of empirical research that we find in second languagelearning research more generally. Now, in my view, the problem with a lot of second languagelearning research is that it tends to equate empirical research with the measurement of causes and effects. If research is a matter of 'seeking answers to questions,' then second language learning research always seems to frame those questions in terms of the influence of some aspectof the teaching and learning process (cause) upon proficiency outcomes (effect). It is, however,rather difficult to define autonomy as either a 'cause' or an 'effect' —and this is to a large extentwhere the problemsdiscussed in this chapter arise— problemsof definition,of the relationship between theory and practice, and of writing about data. Perhapswhat we need, then, is a viewof empirical research that does not necessarily involve 'answering questions'—one that is more suited to concept of autonomy as an aspectof the teaching and learning process that is ambiguous in terms of cause and effect. Research is a matter of gaining a better understanding of (or theorizing?) some process or situation (or practice?) through engagement with data arising from it. Justa suggestion, but it seems to fit the contributions to Autonomy You Ask! rather well. And now for a real Bacardi...
LD SIG 2003 A n t h o l o g y —