Decision-Making in Language Learning: A Lens for Examining Learner Strategies

 

One of the defining features of the modern field of second language teaching is its cross-disciplinary nature. Growing out of linguistics and psychology, it has been stimulated by theoretical concepts developed in fields such as mother tongue education (process approaches to teaching), discourse analysis (genre theory), and artificial intelligence (schema theory). A rich area of research, one that is particularly relevant to issues of language learning and teaching, is that of cognitive science. This area has been linked to the study of language learner strategies, for example in Rubin's (1981) "The study of cognitive processes in second language learning." Another aspect of this field with potentially great relevance is the modelling of "expert systems" and processes of decision-making.

I would like to explore this relevance by combining these areas: looking at the topic of learner strategies through the lens of a decision-making model. This paper will be organized in the following way. First I will briefly summarize work on learner strategies as it has evolved over the past two decades. Then I will frame the learner strategy concept in terms of a decision-making model, relating it to work on teacher decision-making, pointing out some similarities between teacher and learner decision-making, and highlighting the negotiated nature of decision-making in classroom language learning.

Research On Learner Strategies

In the past two decades, the active role of the learner in the language learning process has been clearly acknowledged. The major emphasis in research on the role of the learner in the learning process is the work that has been done on learner strategies. The interest in learner strategy research was initiated in the 1970s and 1980s out of a communicative perspective on language teaching methods which emphasized learner involvement in the process. The concepts of needs analysis (learners making explicit their own goals) and learner autonomy (learners making explicit their preferred procedures for carrying out their learning) were central ones in discussions of communicative language teaching. In the on-going attempt to determine why some learners reach a higher level of proficiency in the second language than others, the question of the strategies learners use began to be pursued in earnest.

This research area first developed in the 1970s with investigations into the characteristics of successful language learners (Naiman et al., 1978; Rubin, 1975; Stern, 1975; Wong-Fillmore, 1976). The attempt was to determine by looking at individuals who had been successful at learning several languages to see what characteristics they had and what procedures they followed. The rationale behind such study was mainly pedagogical-to improve the learning of less successful learners. This practical goal hinged on the assumption that the strategies of the good language learners could be identified and made accessible to poorer language learners as a way to improve their learning. The practical nature of this work sparked a great deal of interest, and there is now an extensive literature in the field of second language teaching on learner strategies.

Learner strategies have been defined by Oxford (1989, p. 235) as "operations used by the learner to aid the acquisition, storage, or retrieval of information." She states that "language learning strategies are behaviours or actions which learners use to make language learning more successful, self-directed, and enjoyable." Rubin's (1981) definition is similar: "...learner strategies includes any set of operations, steps, plans, routines used by the learner to facilitate the obtaining, storage, retrieval and use of information, ... that is, what learners do to learn and do to regulate their learning" (p.19).

Results in this area of research were framed in terms of taxonomies by labelling the "behaviours or actions which learners use" and classifying them into different categories. For example, Rubin (1981) listed six strategies used by good learners: (a) clarification and verification; (b) monitoring; (c) memorization; (d) guessing, or deductive inference; (e) deductive reasoning; and (f) practice. Naiman et al. (1978) classified the strategies used by good learners into 5 groups: (a) active involvement in tasks; (b) realization of language as a system; (c) realization of language as a means of communication; (d) management of affective demands; and (e) monitoring performance as revising systems. Wong-Fillmore (1976, cited in Skehan, 1989) listed eight strategies, grouped into two main groups, cognitive strategies and social strategies. As noted by Skehan (1989), these classifications schemes were for the most part idiosyncratic, although there was some commonalities among the categories posited by the researchers.

This classifying actions into generalized groupings and labelling these groups represents a subtle change from the original definition of learner strategies: i.e., "what learners do to learn." As a result of the attempt to classify what learners do, the term "strategy" in the literature has come to refer not to exactly what learners do, but to the researchers' generalized categories or classes of things that they do. The expression "strategy" actually means "types of strategies" or "classes of strategies." This subtle difference is important in how we view, research, discuss and teach strategies and I will return to it below.

In the 1980s and 1990s, this trend of researching learner strategies continued with the main emphasis the same: elaborating a taxonomy of the classes of actions that learners use to learn. A number of distinctions emerged as being crucial to the task of the researchers. One initial distinction that developed was that between a learning strategy, a strategy intended to lead to learning some particular aspect of the language, and a communication strategy, a strategy intended to lead to effective communication (Tarone, 1981). It is not an unambiguous distinction, however, since the effective use of communication strategies may well lead to learning. In addition, language learners' intentions (i.e., for communication or for learning) are not always accessible for analysis, and there is no principled reason why an action cannot have two different motivating factors. A second important distinction was that made between personality characteristics (learner states which cannot be easily changed by instruction) and learning strategies (learner actions which can be more easily taught). Wong-Fillmore's distinction between cognitive and social strategies remained important. Within the cognitive category a further distinction was made between cognitive strategies and metacognitive strategies. Cognitive strategies involve actions directly operating on language input for learning (like repeating a word or finding a mnemonic device for remembering), while the metacognitive strategies involve planning and organizing future actions to allow the cognitive strategies to be more effective.

This period was marked by the development of the Strategy Inventory for Language Learning, or SILL (Oxford, 1985, 1990), a taxonomy of strategies organized into a set of classes and subclasses. The main distinction in this taxonomy is a slightly new twist on the above-mentioned categories-that between direct strategies (working with the language itself) and indirect strategies (general management of learning). Direct strategies are divided into three subclasses: memory strategies (strategies to remember aspects of the target language), cognitive strategies (strategies using the language and for figuring out how it works), and compensation strategies (strategies for when things don't work). Indirect strategies include metacognitive strategies (strategies for planning, organizing and evaluating learning), affective strategies (strategies for approaching the task positively), and social strategies (strategies for working with others to get input and practice). Each of these types of strategies was broken down into a larger set of more specific classes of actions. This comprehensive scheme also included a questionnaire used for determining those a learner uses (means to get this type of information), and pedagogical suggestions for teaching strategies. The notion of "learner training" developed-that learners could and should be taught how to learn language as well as being taught the language itself. Oxford (1990) is a good example of the application of language learning strategy research to the classroom.

In recent years, the complexity of strategy use has become more apparent. It has become clear that there are different strategies characteristic not only of different learners, but also of the same learner at different levels, with different language learning goals, engaged in the use of different skills, and so on. As a result of this growing realization, research began on the factors affecting the choice of strategies, rather than just the strategies themselves. This research is summarized by Oxford (1989). Strategy choice was found to be related to the language being learned, the learning goals, the level of learning (or proficiency of the learner), the learner's self-awareness, age, and sex. Affective variables have been found to play a role: the attitudes, motivational level, motivational orientation. Personality characteristics play a role as well-learning styles, learning experiences, and methods-as does cultural background (national origin or ethnicity). Another important factor was investigated by Wenden (1986) with the notion of learner beliefs. The connection was made between a learner's beliefs about language learning and the types of strategies that he or she uses. In addition to these more longer term and stable characteristics, there are also a number of short term factors which play a role in strategy choice: the requirements in the current communication or task in the specific situation of language use (and this of course includes the method by which the researcher attempts to elicit a learner's strategies).

This growing appreciation for the complexity of the issue of learning strategies and the factors that play a role in their use makes it increasingly apparent that a classification and taxonomy of strategies does not adequately represent the intricate ways in which strategies are chosen and used. One of the consequences, or costs, of developing a such a classification is that the learner strategies are removed from the context in which they occur. They are categorized and given labels which make sense to the researcher, i.e. explicitly relating strategies which are seen as falling into the same paradigmatic class or category. However such a categorization is not necessarily one which makes sense to the learner, nor one that is used by the learner for accessing possibilities or weighing and making choices about what to do. Such a classification does not give us a sense of why a learner uses a particular type of strategy at a particular time, how this strategies fits with other strategies that are being used, and how the strategy is related to the learner's evolving planning process and beliefs, and how it is evaluated and feeds into further strategy choices. Specifically, the strategies are removed from context.

To determine how and when the learner uses certain strategies, we need to look for relationships among them which are relevant to the learner: those strategies which are considered by the learner to be substrategies of others (i.e., the means to achieving others), and those which are related, from the learner's perspective, by patterns of sequencing (i.e., those which naturally stimulate or follow others).

Examining learner actions and behaviour in terms of a decision-making model provides an alternative for researching and reporting on learning processes. In my work on teacher decision-making (Woods, 1996), I have elaborated a number of concepts related to decision-making in the learning-teaching process from the perspective of the teacher. There is every reason to suspect that the other side of the coin-the learner's perspective on decision-making in the process-would provide interesting insights. In fact, there seem to be a number of parallels between learner decision-making and teacher decision-making that are worth exploring. The first consideration is that there is a great deal of similarity in the types of decisions made by teachers and those made by learners: many decisions could be made equally well by either party. For example, it may be a teacher who decides who to call on to answer a particular question, or a learner who decides to put up her or his hand to answer the question. It may be a teacher who assigns a particular exercise for homework of a learner who decides to do it for extra practice. Although there is an important difference-the teacher's decisions are ultimately geared towards actions by another person while the learner's decisions are ultimately geared towards her or his own actions (i.e., it is the learner who ultimately has to act for learning to take place)-the decisions can involve many of the same actions. In fact, there is often a negotiation that takes place with regard to whose responsibility is it to make which decisions. I will return to this point below, but first I would like to outline some of the concepts that were central to the research on teacher decision-making that I have carried out.

Teacher Strategies, Teacher Decision-Making And Event Cognition

The term "teacher strategies" has been used in the literature to describe what teachers do to implement their teaching in the classroom. What is included under this concept involves a complex set of factors-such things as a curriculum, a teacher's previous experiences (in this course, as a teacher, as a learner), other teachers' input, a previously made course plan and lesson plan, generalized views of learners characteristics, and a number of institutional factors related to the resources available and the constraints placed upon teachers. A productive way of looking at these intricate factors that I have chosen in my own research is a 'decision-making' model (Woods, 1996). In this research, I examined how teaching events play a crucial role in the decision-making processes-the cognition and practice-of ESL teachers. The notions of 'event' and 'event cognition' and 'event structure' are central concepts in this approach.

Event cognition refers to how we perceive, label, recall, and narrate the events that occur in our lives. The concept of event cognition was elaborated by McCabe and Bolzano (1986) in their book Event Cognition. The concept is not referred to by this label in two recent overviews of cognitive science, Cognitive science: an introduction (Stillings et al.,1995), and Foundations of Cognitive Science (Posner, 1989). However, it appears under the guise of episodic memory, which is "remembering a past occasion, which is accompanied by a distinct conscious awareness of a particular moment or event" (this is contrasted to semantic memory, which is "remembering a fact" in Stillings et al. (1995, p.111). Event cognition is also an underlying concept in the well-developed and well-publicized theories of background knowledge structures, schemas and scripts (Galambos et al., 1986; Schank and Abelson, 1977). In these approaches to understanding discourse, it is posited that we are aided by our abstract knowledge of particular types of events, and our expectations about what will happen as we participate in such an event. So, for example, when we enter a restaurant and are approached by a waiter who says "How many?", we have no trouble understanding that he is referring to the number of people in the party and not something else.

Similarly, it was found that teachers, in important ways, conceptualized their teaching as events and types of events. They interpreted prior classes as the occurrence of events that they describe and narrate to colleagues (and to me as a researcher), and they planned future classes in terms of events and they actions that they would carry out to produce the events that they wish to have happen.

The most important characteristic of the events they discussed is that they are structured. The events that make up the teaching-learning process are structured in the sense of (a) following a sequence (and the particular sequencing is seen as being an important aspect of the teaching), and (b) consisting of sub-events-smaller events that, when taken as a whole, constitute a larger event (Woods, 1989).

A second important characteristic of their discussion of the structure of events is that the teachers themselves play an important role in that structuring. It is in this structuring that the relationship between actions and events becomes evident. The teacher plans actions which, when carried out, will contribute to the events that ultimately occur. The teachers then interpret what happened as events, and assess these events as being more or less successful in leading towards their teaching goals. This assessment becomes information to be used in planning subsequent actions.

Planning in this view, then, is seen as a 'productive structuring' of events. Teachers make decisions (usually in conjunction with a pre-planned curriculum) about the conceptual organization of the macro-event of a course: how it will be subdivided into major units or lessons, how these units will be subdivided into activities, and ultimately how the activities will be subdivided into actions and utterances. This planning occurs before a course begins, and continues as on-going planning into the period during which the course is taught, and during that period it occurs between individual lessons, and also continues right into the lesson itself as on-going moment-to-moment planning.

The events are also structured in the sense that teachers described their planning of these subevents as being the means by which the more global events were to be accomplished (i.e., a vocabulary elicitation is seen as the first stage of carrying out a lesson in reading). Conversely, the accomplishment of the higher level events was often described as the purpose or goal for carrying out the subevents (i.e., the vocabulary activity is described as leading to the goals of allowing the students to read the passage and successfully complete the lesson in reading). In other words the higher levels of events have a certain type of inherent relationship to those at lower levels. These relationships among levels implies a kind of hierarchical activity structure in which at all levels there is an inherent connection between goals and means. (I should note in passing that this structure is not a hierarchy in the strict sense of the word, with each lower event [or node] linked only to a single higher event [or node]. An action [and the event that results] can often have more than one purpose at the same time, and so it would be more accurate to say that we have multiple hierarchies, or a 'tangled hierarchy' [Anderson, 1983], or to borrow a term from deMey (1977), a 'heterarchy.')

Teachers describe their retrospective interpretation of what happened in their classes in similar structured fashion. They interpret smaller events as making up larger events and being the means by which the larger events were accomplished (or not accomplished), and they judge success in these terms. Sometimes this structured perception of what happened corresponds to the planned structure, sometimes it is perceived as a different structure from what was planned or hoped for. This retrospective interpretation can be seen as a kind of 'receptive structuring' of the events. One of the key elements in this interpretive process is that it is influenced in crucial ways by the teacher's background knowledge structures (or as I termed it in the study, the teacher's beliefs, assumptions and knowledge). In this view, the teachers' beliefs about how things are and how they should be played an important role in how they interpreted what happened and whether they judged it as successful or unsuccessful in light of their goals (and, as a result of this interpretive process, two teachers-or a teacher and a student-might have quite different interpretations of what happened in a particular class).

The interaction of planning and interpretation results in a kind of 'event cycle' in which there is pre-event planning, then the undertaking of the actions and the occurrence of the resulting event, followed by the interpretation of what happened. This interpretation, and its success or lack of success in light of the higher level goals, feeds into the planning of subsequent actions and events, and with this updated information, the teacher repeats the cycle. Because of this hierarchy of higher and lower level events, this cycle is occurring simultaneously on many levels, spiralling down into subevents (how to achieve the goals) and back up to superevents (the goals themselves). This is the kind of means-ends analysis that is posited as characterizing human goal-directed behaviour.

The complexity of the process of teaching and learning is that it involves not just a single participant. In the early analyses of human problem-solving processes, the problems to be solved and goals to be achieved were static and self-contained (for example, a mathematical problem). In teaching, the 'problem' and goal of having learners learn the language is dynamic. The teacher is making decisions and carrying out actions that involve decisions and actions by the other participants in the process-the learners. The process is interactive.

As noted above, in such an interactive process the distinction between actions and events is highlighted. The teacher plans actions which are intended to control or influence the outcome of events. But the event that results is not entirely predictable because of the contribution of other participants-the learners. They may have their own agendas (i.e., plans and goals) that the teacher is not aware of and may take the initiative to control the events that occur, or they may react in ways which are not what the teacher expected, producing events that were different from the teacher's predictions or expectations. It is the co-occurrence of actions from different participants in a particular learning-teaching setting that produces an event, one that the participants later interpret and judge (for example, "Oh, that was a great class" or "Oh, that activity was boring"; of course the interpretations are often much more intricate).

In other words, when we look at the larger picture of the teaching-learning process, and the decision-making that goes into it, we see that some of the decisions which occur are taken by the learner. For example, it may be the learner who initially decides to take a course, it may be the learner who decides to put up her or his hand in class to answer a question, or to do the homework carefully. Learners may also make decisions not directly related to the teacher's course decisions but ones which nonetheless have an impact on their improvement in the language, for example, doing extra reading, or watching programs in the target language, or spending time with speakers of that language.

Learner Decision-Making and the Event Cognition

These kinds of issues are ones that can be explored in a decision-making model of language learning, where learner strategies are seen as being part of the learner's decision-making process. This perspective implies a more naturalistic accounting of the learner's thinking during the process of learning (or attempting to learn) another language. The event structure framework described above allows us a mechanism for examining these relationships, in particular the means-goals relationship as well as the interconnection between beliefs and choices of actions on the part of the learner.

In a study of learners perceptions about their classes (Allwright and Woods, 1992), we noted several characteristics about learner decision-making that were both similar to and different from teacher decision-making. The first is that, within the interview format that we used to collect data, we found that learners often had goals for their learning and were making decisions about how to achieve these goals; however, they were often much less active decision-makers than the teachers. This is of course not surprising given the usual social-educational norms of the teachers being responsible for the majority of the classroom-related decisions (it is the teacher who holds or announces the plan for the lesson). With the responsibility and the expectation about that responsibility on the shoulders of the teacher, it is natural that the teacher will put more time and effort into thinking through procedures and alternatives than will the learner. As a result, the pattern of decisions produced by the learner reveals that it is often not as complex and hierarchically organized as the teacher's. There are fewer levels of action and subactions distinguished and made explicit by the learner.

A second aspect is that students nonetheless have expectations about what will happen in the classroom, and how different things will be done, and who will do what. As with teachers, these expectations depend on previous experiences in language classrooms and on beliefs about language learning. These expectations can be considered a kind of 'implicit plan' that the learner holds for which no advance actions need to be explicitly decided upon until the expectations are not fulfilled, resulting in what Linde (1980, discussed in Woods, 1996), terms a 'hotspot.' At the point when the learner becomes aware of the discrepancy between her or his expectations and that is happening, she or he is more likely to begin to carry out some explicit compensatory planning (even if it is verbalizing a complaint to a classmate).

A number of case studies carried out by graduate students at Carleton have begun to examine some aspects of this process, incorporating the intricate interplay of beliefs and different types of motivation in the process. Valenzuela (in press) discusses some aspects of the decision-making process of a learner who found that the events making up his ESL class did not include what he would recognize as 'grammar teaching', and developed several compensatory strategies. Allen (in press) describes the evolution of a learner's beliefs about language learning-how they are affected by classroom activities and the teacher's verbalizations, and how they in turn affect the types of learning decisions or strategies chosen by the learner. Thomas (in press) describes how the learner's strategies are intimately linked to beliefs and intrinsic motivations, but that in the short run motivations can be affected by external factors, resulting in fluctuating strategy use. This article shows clearly that a one-shot elicitation of strategies can be very misleading, and that a list of the strategies used by a learner such as this one can hide a great deal of what is going on.

One of the most important issues that arises out of these studies, highlighted in the Valenzuela article, is the question of who is supposed to be doing what in the language teaching/learning decision-making process. The idea that this question does not have an unambiguous answer was the reason that Allwright (1981) coined the term 'management of language learning.' In this conception, the traditional view that the teacher 'teaches' (tells the learners what to do to learn) and the learners do the 'learning' is seen as a oversimplification of the process. In fact, he claims, the actual learning we cannot see. All we can observe is the actions designed to lead to learning, which can be decided upon by the learner or by the teacher. Whose roles it is to do what is potentially up for negotiation in every class. For example, the exhortation to teach language learning strategies carries with it the possibility that learners will have to learn (i.e., be convinced) that it is their role to make decisions about which actions they should take to learn, a job traditionally done by the teacher and one that they might well resist. The notion that decision-making processes are negotiated and shared brings us to a new conception of strategy research, one that focuses not on learner strategies but rather on learning strategies and the intricate interplay of learner and teacher in their determination.

References

• Allwright, D. (1981). What do we want teaching materials for? ELT Journal , 36, 5-18.

• Allwright, D. & Woods, D. (1992). Making sense of classroom observables: Learner, teacher, researcher interpretations. Poster session presented at TESOL Conference, Vancouver, B.C., March 1992.

• Anderson, J. (1983). The architecture of cognition. Cambridge MA: Harvard University Press.

• Galambos, J., Abelson, R., & Black, J. (1986). Knowledge structures. Hillsdale NJ: Lawrence Erlbaum Associates.

• Naiman, N., Froehlich, M. Stern, H., & Todesco, A. (1978). The good language learner. Research in education series 7. Toronto ON: Ontario Institute for Studies in Education.

• Oxford, R. (1985). A new taxonomy of second language learning strategies. Washington DC: ERIC Clearinghouse on Languages and Linguistics.

• Oxford, R. (1989). Use of language learning strategies: A synthesis of studies with implications for strategy training. System ,17(2).

• Oxford, R. (1990). Language learning strategies: what every teacher should know. New York: Newbury House.

• Posner, M. (1989). Foundations of cognitive science. Cambridge MA: The MIT Press.

• Rubin, J. (1975). What the "good language learner" can teach us. TESOL Quarterly , 9, 41-51.

• Rubin, J. (1981). Study of cognitive processes in second language learning. Applied Linguistics, 11, 188-131.

• Schank, R. & Abelson, R. (1977). Scripts, plans, goals and understanding. Hillsdale NJ: Lawrence Erlbaum Associates. Skehan, P. (1989). Individual Differences in Language Learning. London UK: Edward Arnold.

• Stern, H. (1975). What we can learn from the good language learner. Canadian Modern Language Journal, 31, 304-318.

• Stillings, N., Weisler, S., Chase, C., Feinstein, M., Garfield, J., & Rissland, E. (1995). Cognitive science: An introduction. Second Edition. Cambridge MA: The MIT Press.

• Tarone, E. (1983). Some thoughts on the notion of "communication strategy." In Faerch, C. and G. Kasper (Eds.), Strategies in interlanguage communication. London: Longman.

• Wenden, A. (1986). Helping language learners think about learning. ELT Journal, 40(1), 3-12.

• Wong-Fillmore, L. (1976). The second time around: Cognitive and social strategies in second language acquisition. Unpublished PhD dissertation. Stanford University.

• Woods, D. (1989). Studying ESL teachers' decision-making: Rationale, methodological issues and initial results. Carleton Papers in Applied Language Studies, 6, 107-123.

• Woods, D. (1996). Teacher cognition in language teaching: beliefs, decision-making and classroom practice. Cambridge UK: Cambridge University Press.