Flower-1994 Chapter 8 Metacognition PDF

8 Metacognition A Strategic Response to Thinking Is Metacognition Really Necessary? This book began with an image of literate action as a social cognitive process and as a theory of how writers construct negotiated meanings through acts of interpretation, negotiation, and reflection. However, the fo...

8 Metacognition A Strategic Response to Thinking Is Metacognition Really Necessary? This book began with an image of literate action as a social cognitive process and as a theory of how writers construct negotiated meanings through acts of interpretation, negotiation, and reflection. However, the following chapters argued, the logic of these socially situated constructive acts becomes apparent only when we recognize writer’s strategic knowl­ edge—their goals, strategies, and awareness. Strategies or heuristics have long been seen as central to the art of rhetoric, just as goals are the driving force in cognition and motivation. But the place of metacognitive awareness is less clear. Is metacognition really necessary to learning or meaning making? And if so, does it follow we should teach it? Consider this freshman in the Rcading-to-Write project (chap, i) as she describes in class what she discovered about herself from her own thinkaloud protocol. “Since 1 was talking out loud, 1 was very conscious of the fact that I was making connections to what everyone [in the source material] was saying. Using my own—using things that have happened to me to connect to the people talking... And J realized that 1 actually do have strate­ gies to read. I thought I didn’t. I thought 1 was some kind of odd person who didn’t have any strategies ever.” Observation-based self-reflection seems to be succeeding in convincing this young woman that she is indeed making strategic choices, where six weeks of class discussion had failed. Yet, for all the wonder of her discovery, she had done well enough without this knowledge to make it into the freshman class of a competitive college. Vince (“I’m only discussing this,” chaps. 6, 7) not only succeeded in high school, he was on the road to success in his freshman class, holding on to his well-developed knowledge-driven 223 224 The Construction of Negotiated Meaning strategies and not particularly eager to to call either his strategies or his assumptions about writing into question. There is a strong tradition in liberal education that values, even venerates, reflection and awareness—a set of assumptions I tend to share. But as an educator, one cannot merely assume that the discourse practice of reflection is really critical to learning, much less to all students. There are reasons to take a critical look. To begin with, reflection is not a mimetic act mirroring an inner reality; it does not function as a gesture of pure awareness. It is, after all, a historically shaped discourse practice with a long tradition that includes philosophical pensees, the “confessions” of an Augustine and Rous­ seau, the “apologia pro vita sua” of a Newman and the Puritan practice of religious self-examination that helped shape autobiographies of a Pepys, Rousseau, or Boswell, as well as the emerging form of the novel (Watt, 1967). It includes the guided journeys of psychotherapy in its many forms and the implicitly rule-governed practices of an “authentic” personal essay (Macrorie, 1970), the freewheeling, freewriting style encouraged in journal writing (Elbow, 1981), or the politically directed action/reflection designed to create critical consciousness (Freire, 1989; Giroux & McLaren, 1989). Reflection, my point is, exists not as a faithful mimesis of inner experience but as a set of codified discourse practices, a kind of thinking one can learn to do. To say this does not deconstruct the practice of reflection into a mere game or deny the possibility of its enormous personal and educational value, but it does ask us to treat the educational significance of any given reflective practice as an open question. When reflection is self-motivated and incidental—something good stu­ dents do on their own time—its value would probably be unquestioned. But consider the harder case when intentional, strategic reflection forms part of the content and the writing of a course. Then reflection as a literate practice, chosen in place of other practices, needs to be accountable. And at this point, other voices that also equate literate action with liberating education enter the discussion. I hear Delpit (1986) arguing that many liberal, white “process” teachers fail to listen to the black teachers when they assert that what their students need is not a place to express their voices but straightforward instruction in how to handle the conventions of power— the patterns and conventions of Standard English. And I hear Hull and Rose (1989) writing about underprepared students for whom the five-paragraph theme is both a revelation and a generative structure. And I hear Peck (1991) standing in the midst of an interracial community center, where people on the streets of the city’s North Side do not need to be “made aware” of their oppressed condition (even if the middle-class kids in nearby colleges do), but they do need to control the kinds of literacy that lets them act, the Metacognition 225 rhetorical strategies for making plans and arguments that can lead people in a neighborhood into concerted action. If we are going to listen to these voices, we have to be sure that acts of reflection are not just literate classroom performances that produce charming texts with the cadences of engagement and discovery that make teachers feel good. We have to ask, Is guided reflection a robust occasion for learning, or is it just frosting on the cake? It is in that spirit of skeptical commitment to the possibilities of reflection that this discussion is written. Let me clarify how the sometimes interchangeable terms of metacognition, awareness, and reflection are being used here. Metacognition, as the most inclusive concept, signifies thinking about thinking at many levels of aware­ ness. Metacognitive awareness, when it takes the force of a verb, signifies acts of observation, alertness, and noticing when a part of attention or working memory is devoted not just to action but to the recognition that I am performing this action. Awareness, as a noun, refers to the knowledge possessed as a result of this divided attention. However, as think-aloud protocols often demonstrate, such “knowing” can be a fleeting event, pushed out of memory by the press of business, as my colleague, Dick Hayes, once brilliantly but inadvertently illustrated. Giving a seminar on protocol analysis to a group of writers in Washington, he was demonstrating the process, composing aloud on a difficult topic and promptly running into a planning problem with which he struggled for some minutes. Later, when asked if he had had any difficulty planning, he blithely reported that it had gone quite well as he reiterated his final, triumphant plan. It was only when he was reminded in front of the group of the details of his struggle that awareness of the dilemma returned. Reflection, as 1 will use the term, is an intentional act of metacognition, an attempt to solve a problem or build awareness by “taking thought” of one’s own thinking (Flower, 1989b). In the collaborative planning study in chapters 6 and 7, awareness was pushed to more articulate reflection by a dialogue with a partner in which it was used to identify problems, pose alternatives, or justify decisions. It was a side effect of collaboration. In the chapters that follow, awareness is the result of an intentional and assigned act of reflection. I. The Debate over Metacognition Meta, as in metaphysics, metaphor, and metacognitive, suggests a level of awareness that floats somew'here above more concrete events. But the link between knowing and acting and the content of what people know can be highly variable. In practice, the different meanings of metacognition 226 The Construction of Negotiated Meaning cluster around two levels of metaknowing: the in-process acts of regulating cognition and those of reflecting on it. The first level of cognition about thinking is often relatively tacit, a process of regulating or controlling other cognitive acts. Readers engage in such metacognition when they interrupt fluent reading to monitor a failure in comprehension, to reread, to try to form a gist, to clarify a momentary source of confusion. This level of metacognition, which Baker and Brown (1984), describe as control strate­ gies, monitors cognition and redirects attention, putting on the brakes, calling up new processes, setting a new goal for the thinker. In writing, acts of planning, detecting, and diagnosing problems are all control strategies that let writers monitor and guide cognition. Experienced readers and writers tend to make a greater number and/or different versions of these metacognitive moves than less experienced writers (Bereiter & Scardamalia, 1987; Flower et al., 1992; Hayes et al., 1987). But in all these cases, metacognition as control can be a highly automated, extremely fast, and tacit process. A second level of metacognition refers to the knowledge one has about cognition—“to the stable, statable, often fallible, and often late developing information that human thinkers have about their own cognitive processes; traditionally this has been referred to as knowing that ” (A. Brown, 1987, pp. 67—68). Such knowing has a privileged status as Brown asserts: “Several theorists from quite disparate schools agree that the most stringent criteria for understanding involve the availability of knowledge to consciousness and reflection; thus permitting verbal reports” (p. 72). Metacognition is knowing that you know something and being able to inventory and talk about your own knowledge as content and as process, that is, being able to talk about what you know (and what you do not, how your knowledge is organized, etc.) and secondly, about how your thinking operates. The object of cognition about cognition, then, is not only the topic knowledge one possesses, but one’s own thinking processes and strategies, as well. However, are all forms of this statable metacognition equally useful to a learner? One form, which has not fared well in educational research, is that “knowledge of pertinent facts” that is often a “permanent part of one’s naive theory on the topic” (A. Brown, 1987, p. 68). It is the sort of informa­ tion tapped by questionnaires and clinical interviews that ask people to generalize: to estimate their memory span (Flavell & Wellman, 1977) or to describe their normal writing process (Rubin, 1983). Unfortunately, when asked to retrospect about their own thinking and decision process, people not only “tell more than they can know” but are often demonstrably wrong (Nisbett & Wilson, 1977). Or, as Rubin found, such knowing may simply reproduce the reigning conventional wisdom rather than describe real perfor­ mance. Or, like sentence diagraming or truisms about thinking about the Metacognition 227 reader, it may tap the “inert knowledge students can produce on request but have never learned to access when needed or to use for real purposes (Bereiter &L Scardamalia, 1985). In contrast to these generalized “naive theories” and retrospections, a much more limited form of statable metacognition is clearly tied to successful performance. It is often referred to as the ability to mention as well as use, to recognize, or to access one’s strategies. Such knowledge may be statable as no more than the unelaborated comment we see in think-aloud protocols (e.g., an in-process recognition that “I think I need an example here to make the point”). However, such recognition marks a great watershed in learning, as evidence that a student has achieved both access to and control of his or her own knowledge. As A. Brown says, “children who know perfectly well how to use a strategy or have the relevant prior knowledge, often fail to access it on appropriate occasions” (1987, p. 72). Many see this as the classic problem of learning—getting access to what you know (Prawat, 1989). It is only, A. Brown argues, when learners have conscious access to and control of their own knowing, when knowledge is no longer welded to the situation in which it was learned, that we achieve the flexibility that is the hallmark of higher intelligence (p. 71). There are also some strong educational implications from metacognitive research. On the down side, many metastatements are merely inert theory. However, as Vygotsky argued, it is clear that one form of metastatement, self-talk, helps. It is possible, for instance, to improve both learning and transfer by doing nothing more than asking problem solvers to state a reason for each move they make as they are thinking (Ericsson & Simon, 1985). Educators and therapists have seen lasting effects from cognitive therapy and self-regulated learning in which hyperactive, impulsive children learn to talk themselves through tasks, talking out their intentions, calmly noticing and recovering from errors, and congratulating themselves for trying. In other studies, college students who saw themselves as uncreative learned to alter their self-defeating stream of self-talk into supportive and strategic self­ instruction (Harris & Pressley, 1991; Meichenbaum, 1977; Meichenbaum, 1985). The techniques of self-instruction and self-regulated learning vary, but the three essential features include (1) modeling a repertoire of strategies, (2) teaching learners to monitor their own thinking through group or self­ talk, and (3) motivating this process by helping students recognize the links between outcomes and their own strategies (A. Brown, 1987). I he case for statable metacognition, for self-talk, is a strong one and helps explain how process and strategy teaching can intervene in the very act of writing. How­ ever, we must maintain our initial skepticism and recognize that this evidence for self-talk does not extrapolate into a mandate for reflection as a literate 2.2.8 The Construction of Negotiated Meaning practice. The form of metacognition this research on self-talk supports is a highly situated, unelaborated, goal-directed, awareness in action. In the next two chapters, I would like to follow the traces of another, less studied, form of metacognition found in students’ observation-based self reflections. This literate practice appears to stand somewhere between the specificity of in-process self-talk (i.e., mentioning and access) and the generalizing sweep of naive theories or retrospections. This form of metacog­ nition is especially relevant to writing because it comes from students’ databased self-observations and subsequent interpretations of their own writing and thinking. It is, on the one hand, a relatively articulated, written, even revised account with a concern for coherence and evidence—a personal theory. It is, however, based on self-observations, tapes, and transcripts of a process in action making it potentially closer to in-process reflections and critical incident accounts (Flanagan, 1954) than to retrospections or students’ conventional accounts of their own writing (Rubin, 1983). It is, we might say, a situated theory that tries to generate an elaborated under­ standing or hypothesis about an issue, but it does so on the basis of systematic rather than incidental observation and with the goal of understanding and acting in a specific situation (Flower et al., 1993). This form of metacognition is a complex, intentional, time-taking act of reflection. For instance, a student writer might review the state of her topic knowledge (e.g., What do I know for an assignment on the meaning of Moby Dick, 5 pages, due Tuesday?) and then go on to consider the discourse conventions or even ideology of communities who assign (and apparently like to read) such texts. Facing a new task, reflection might lead her to think about the assumptions, values, goals, and strategies that are informing her present act of composing. She might even pause to monitor how she is managing her own thinking (e.g., I’d like to write a critique of this inane assignment, but since I’m actually coming up with good ideas, I’ll keep going) or to note her own strategies (e.g., I’m getting bogged down in perfecting sentences again). To operate at this level of reflective metacogni­ tion calls for both a heightened awareness (beyond that needed to regulate cognition) and for a language (beyond self-talk prompts) to talk about, evaluate, and manipulate that knowledge. Reflection is clearly a constructive process, but the question remains: What sort of knowledge or meaning does such reflection create? And do verbal statements (including our assumptions, insights, fictions, and self­ observations) actually reflect, predict, or guide the other business of regulat­ ing cognition (e.g., planning, monitoring, managing one’s thinking)? We will consider two issues in the vigorous theoretical debate over metacognition Metacognition that have a special relevance for writers. One is the problematic epistemology of reflection, and the second is the charge that reflection is merely a luxury. Reflection: An Interpretive Act The epistemological status of reflection and verbal reports in general has always been problematic. Although reflection has a privileged status in the humanities, it has been treated as an artistic or philosophical form rather than as a significant part of what students do or how they learn. What exactly does metacognitive cognition know? In psychology, self-reports have had a checkered history. The introspectionist enterprise wrecked on the fact that even with hours of training, people were unable to report on elementary mental processes such as perception or memory search. In radical behavior­ ism, thoughts (and verbal reports) were banished altogether as epipheno­ mena rather than causal agents in learning or behavior (Skinner, 1963). Because words and actions were seen as separate systems of response, the relationship between them is arbitrary. One’s words (interpretation, insight, explanation) are not the product of insight into the real causes of action but the product of whichever verbal response (interpretive scheme, discourse, set of commonplaces) is being reinforced by one’s reading, teachers, culture. In clinical psychology, by contrast, self-awareness claims authority based on its power to change behavior or predict future actions. However, clinical interventions based on awareness of motives located in an unconscious mind have a rather depressing record of success: “It is not entirely surprising that self-insights are often unaccompanied by changes in behavior. It is easier to alter people’s beliefs about the causes of their behavior than to change how they behave. Alcoholics, for example, can be more easily persuaded that they drink because of a fixated oral dependency than they can be induced to forsake alcohol” (Bandura, 1986, pp. 4-5). More problematic is the possibility that gaining insight is “more like a belief conversion than a self-discovery process.” The therapist’s conceptual orientation is a strong indicator, for example, of “whether or not clients will find an unconscious mind and what will be in it.... Hence, one can predict better the types of insights and unconscious motivators that persons are apt to discover in themselves... from knowledge of the therapists’ conceptual belief system than from the clients’ actual psychological status.” (p. 5) In contrast to behaviorism and psychoanalytic inquiry, cognitive psychol­ ogy attempts to open the black box, placing thought at the center of its inquiry into how people learn, structure their knowledge, and solve prob­ lems. Because cognitive psychology operationalized thought as the pro- 230 The Construction of Negotiated Meaning cessing of information, it tended until recently to ignore social and emotional aspects of thinking. However, that perspective also led to an explosion of knowledge about thinking as a process of creating and transforming sym­ bolic representations and solving problems. And it led to an intense interest in the relationship between knowing, broadly conceived, and the articulated, structured, and linguistically coded knowledge represented in a verbal report. We know that access to our own thinking is limited in some predictable ways. Processes that are frequently used may become automated or tacit, slipping underground, out of focal attention, while many perceptual/motor processes, including memory search, are simply not open to introspection (Hayes, 1978). If that were not enough, assumptions, ideology, or suggestion can conspire to create reasonable fictions about cognition, lightly constrained by reality. Retrospections by famous writers, for instance, typically function more as theories or proclamations than observations (Emig, 1971). And then there is an awkward gap between those features of our experience to which we normally and selectively attend, and those that we could perceive with a different mind-set. In sum, the relationship between our regulatory metacognitive processes (planning, monitoring, switching strategies, etc.) and our reflective metacognitive statements about those acts of planning, monitoring and about our assumptions, goals, reasons, habits, and so on is neither dependable nor straightforward. As educators, we need to approach metacognition, especially intentional reflection, with our eyes open. One solution, argued by Paris (1988), is to restrict the discussion of metacognition to reflective statements themselves, to knowledge about men­ tal states and abilities that can be shared so that such knowing can indeed be measured, evaluated, and related to action. Since people’s reflective aware­ ness of their own regulative metacognitive processes (such as, monitoring and testing) is normally limited, much of what we now call metacognition would drop out of the picture. Paris’s conception narrows the focus to statements about knowledge and strategies that are teachable and measur­ able. Yet unlike many definitions, his does include attitudinal and motiva­ tional perceptions about cognition, such as, “I’m not good at writing.” Note that in this approach to metacognition, it would not matter whether such a perception is true of the writer’s actual skill. As Paris acknowledges, limiting inquiry to reflective talk does not resolve epistemological problems with the status of such knowledge. Self-perceptions of the sort he describes can be notoriously inaccurate descriptions of cognition even though such beliefs may well dictate a student’s behavior. Another response to this problem is to develop a theory that can account for what people can—and cannot—report about their cognition under what circumstances, that is, to plumb the process of how reflections are being Metacognition 231 constructed. Ericsson and Simon’s (1985) widely cited theory of verbal reports has laid a foundation. It defines two conditions under which people do not have access to their own cognition and so will be forced to draw on memory and, in many cases, to generalize across multiple, dissimilar in­ stances separated in time. One such condition is when people are asked to make predictions about how they would think (such as “How would you read this story if paragraph two had been omitted?”). The other is when people are asked to offer retrospections about past performance (such as “How did you [or do you “normally”] deal with conflict?”). Such questions are widely used in research and by teachers, but as Ericsson and Simon’s theory predicts, they are not reliable accounts. As many studies and critiques of such verbal reports have documented, it is possible to imagine and report processes one does not, in fact, use and at the same time, to use many strategies of which one has little awareness, such as the fleeting acts of comprehension monitoring seen in experienced readers. This is not to dismiss the potential interest of such statements but to treat them like other general­ izations or inferences for which we inquire politely about the evidence and reserve judgment when there is a sizable leap between data and claim. Ericsson and Simon’s third condition exists when people are asked to give a concurrent verbal report of thinking as they are carrying out a task. Such reports allow more direct access to cognition but with some important qualifications. The cognition people can reliably report will be knowledge and processes that are operating in focal attention (i.e., not automated processes or weakly activated knowledge) and secondly, knowledge and processes that are already verbally coded. (Think, for instance, of the nonver­ bal perceptual processes computer troubleshooters seem to possess as they race through and manipulate a person’s files looking for a problem. How­ ever, this form of knowledge may require considerable recoding when they are asked for a verbal explanation.) People’s access to their own cognition is also compromised when they are asked to report selectively (e.g., to attend to word choice or feelings about authority as they work). Such requests will change the process by literally changing part of the writing task and altering the writer’s focus of attention. Ericsson and Simon’s theory shows how verbalizing thinking while in the act of thinking can give us a better window on cognition, but like any window, it also has a frame that blocks out tacit and nonfocal information in some predictable ways. Likewise, retrospective reports based on cued recall and recent experience will be subject to less interpretive distortion than are broad generalizations covering a sweep of events or speculations about causes (of which the writer was unaware at the time). Most impor­ tantly, and this is the heart of the Ericsson and Simon argument, the data 232 The Construction of Negotiated Meaning of verbal reports are still and only that—the data of observation that, like all data, await interpretation. The meaning of the data is not self-evident; it lies in the way that data is used as evidence by students and researchers alike. In sum, reflection must be recognized as a highly interpretive process, influenced by the prior knowledge we bring and the way it is conducted. The value of observation in this process is its ability to challenge as well as to shape the ways we construct ourselves as writers. Reflection: A Guide to Action? A second issue in the metacognition debate is whether metacognition plays an important role in learning: Is it a necessity or a luxury, a deep discovery or an epiphenomen—and if each can be true at times, which is which? We know that some metatalk is just another name for inert knowl­ edge—the party line on paragraph structure or writing that many students could produce on demand along with the names of state capitals. Such declarative knowledge was learned and remains as a self-contained package, unconnected to actions, problems, or situations a student might encounter. Even when relevant, it is useless because it is neither linked to nor cued by situations of use. At the other end of the significance scale, A. Brown (1987) argues that sophisticated, intelligent systems, human and computer alike, are highly dependent on the feedback supplied by metacognitive functions like planning in action, troubleshooting, monitoring, revising, contingency planning. But just how much do learners need to think about thinking to develop these metamoves? We know that the first steps in acquiring control strategies, such as the comprehension monitoring good readers do at the end of a paragraph, may take place in a collaborative, verbalized social experience (a parent interrupts reading to ask, “What did the wolf think?”). Those steps may also depend on awareness, conscious attention, and declarative knowledge, even though such moves later become automated (Brown & Palincsar, 1989; Dansereau, 1985; Vygotsky, 1986). Likewise, our procedural knowledge for driving a car (Anderson, 1980) goes through an initial period in which learning to operate the clutch, the gear shift, and the accelerator takes full concentration and muttered self-commands. Later, awareness reappears when people en­ counter conflicts or new, ill-defined problems that invoke a problem-solving process. It is no surprise that problem-solving pedagogies in math, physics, design, reading, and writing focus on both expanding students’ repertoire of strategies and increasing their awareness of when to switch among them (Hayes, 1981). But how much is awareness necessary? Metacognition 233 For example, Brown and Palincsar’s (1989) reciprocal teaching helped children make great gains in reading comprehension by making these control strategies explicit in the form of questions and prompts young readers gave each other in a collaborative effort to comprehend the text. Because the ultimate goal is for learners to internalize and automate these reading strate­ gies, students take turns helping each other to state the gist, predict what will come next, clarify uncertainties, and predict comprehension questions. However, they spend little time at the next “meta” level, reflecting on the strategies or on themselves. This form of “direct instruction” turned four key control strategies into named, articulated actions that readers could model and learners could try out. And when they needed help shaping a good question or a prediction, as many young readers did initially, a more experienced reader could intervene in the process, helping to shape a better question. Such direct instruction stands in contrast to “blind training” in which students are swept up in activities that the teacher assumes will promote learning or performance. Although the teacher may intend that a given activity (e.g., group talk) will elicit certain strategies or reach certain goals, that information is not shared with students. A familiar form of blind training criticized by Applebee is the use of process activities in writing that are not anchored to a goal-directed image of the process—students are not let in on the goals of the activity, including what it is good for (and what it is not). When journal keeping, for instance, was taught as an end and value in itself, students came to see it not as a heuristic that might help them learn material, but as yet another exercise/assignment quite unrelated to the serious business of preparing for the test. Advocates of cognitive apprenticeship, direct instruction, and strategic teaching make a strong case for teachers who not only engage students in trying out heuristics but who show their hand and let students in on the purpose and the structure of the strategies (Brown & Palincsar, 1989; Collins et al., 1989; Jones, Palincsar, Ogle, & Carr, 1987). They make the case for a first level of metacognitive instruction. However, informed practice is not the same thing as cultivating a more complex reflective awareness of one’s own strategic knowledge. And, a skeptic might claim, it seems unlikely that such cultivation would have much effect on normal comprehension or the fluent generation of unproblematic prose. People possess strategies they cannot and do not need to talk about. We cannot assume that heightened, articulate awareness is always necessary beyond the level of metacognitive regulation and control that can be developed through direct instruction. And yet, thinking about thinking—metacognition as reflection—has been treated as a critical step by humanistic educators from Socrates to Dewey to the current rhetorical and critical thinking movements. If reflection is not 234 The Construction of Negotiated Meaning necessary for the fluent, unproblematic moments of composing, does it have a place in learning to write? Reflection: A Strategic Act The metacognitive debate makes it clear that reflection does not hold up the mirror of mimesis; it is a goal-directed, interpretive act that can let people understand, monitor, and guide cognition. However, an undiscriminating embrace of reflection per se makes as little sense as the uncritical, uncondi­ tional valorization of collaboration we discussed earlier. Reflection, like collaboration, is not only a powerful constructive process, it is a highly variable one. The educator’s question is, What kind of knowledge is being created through reflection (under a given set of conditions), and how is that knowledge going to guide action? To develop a theory of reflection as a situated, strategic act, one needs to ask: How is reflection embedded within its educational situation; what purposes do students perceive their reflection to be serving? What is the nature of students’ interpretive, constructive process; what is the nature of the knowledge they actually construct? How does reflection lead to action? The next two chapters try to answer these questions in the context of a specific educational experience in which metacognition was built on the foundation of collaborative planning. In the two college classes we will examine, students collected tapes and observations of themselves in order to examine their own planning, writing, and/or collaborative processes. Reflection took the form of articulate, considered, written reflections that were an important part of the course work and discussions. Just as collaborative planning talk has its own signature as a form of collaboration, reflection based on self-observation will shape the experience in distinctive ways. When reflection is built on collaborative planning, it takes on the character of a problem-posing strategic inquiry quite different from the “how I write” introspections in which writers seek global patterns and speculate on underlying causes. Although such introspective journeys can yield revealing statements of writers’ assumptions, they also bring to mind the “key to all mythologies” that Casaubon, the slowly withering scholar of Middlemarch, labored so long to create. Introspective thinking about thinking is more likely to satisfy us with its ability to create a comfort­ ing sense of coherence than with its power to surprise or guide. The signature Metacognition 235 of observation-based reflection (like the theory building described in chap. 1) is that it looks at cognition in context in response to a problem the writer poses, aided by an independent observational record that can both support and challenge interpretation. Observation could—in theory—play two important roles in these reflec­ tions. (What it actually does becomes an empirical question.) First, it could help students balance their assumptions, prior knowledge (and the informing concepts presented by the course) with the evidence of experience and the feedback it gives to those expectations. Observation urges the writer out of the comfortable chair of retrospection where insight can be formed with the plastic stuff of memory to fit a schema or tell a coherent story. When, on the other hand, students observe their own learning, their reflections can be forged, tested, and reconstructed in the resistant context of experience. Secondly, self-observation can also play a decisive role in learning when success in the activity depends on the fidelity, the consistency, and the immediacy of self-monitoring (as when athletes monitor their form, readers monitor their comprehension, or collaborators monitor the coherence of their plan or the escalating signals of a partner’s confusion) (Bandura, 1986). Observation, we must be clear, is never independent of interpretation; it is not a pure mimetic act. For example, prior self-conceptions as well as mood will influence what aspects of experience people perceive in themselves, how they are evaluated, how they are linked to other memories, and so on. Moreover, reliable self-monitoring takes sustained, focused effort. Observa­ tion is important not because it stands as a simple veridical account but because it functions as a resistant, balancing, mediating force that is capable of surprising us into the unanticipated. Bandura has described this situation in his theory of reciprocal causation in which cognition, context, and ongoing behavior interact. If he is right, then a student’s writing is unlikely to be shaped by any single cause. The logic of a given learner must be treated as an open question. If our description of negotiated construction is right, significant writing is likely to emerge out of conflicting goals, contradictory schemas, and multiple unreconciled alternatives. In the face of this complex­ ity, observation can help counter the tendency of memory and introspection to form a good gestalt and achieve the comfort of closure. Observation (based on tapes of planning sessions, process logs, periodic self-interviews, and so forth) can retain the traces of the roads not taken and of the goals and decisions that lie behind the current meaning. The type of metacognition we will explore, then, is a distinctive construc­ tion, shaped by observation and focused on the strategic knowledge students found themselves bringing to some specific rhetorical situations. Their reflec­ 236 The Construction of Negotiated Meaning tions represent what they selected as significant out of those experiences and the working theories they created to interpret them. They also let us address the epistemology and the luxury problems sketched above by asking: 1. When students construct an image of their writing process from observational data, what sort of knowledge are they creating (the issue on which the rest of this chapter will focus)? Does the represen­ tation they build differ, for instance, from the images found in their textbooks and in journal articles? Can these writers gain access to their own cognition in a way that might challenge private theories and/or process fictions with which they started? 2. Does the opportunity to observe, to articulate difficulties, to interpret and reflect on one’s experience have any affect on what happens next? Does metacognitive awareness bear any relationship to action (the issue on which the next chapter will focus)? II. Representing a Cognitive, Social, and Affective Process Placing Reflection in Context According to the old adage, the best way to learn something is to teach it. Studying one’s own learning has some of the same virtues if students begin to examine, appreciate, and question what they know and how they think. However, doing research on thinking is often a slow and uncertain process, while teaching must do the best job possible in 15 weeks. Making observation and reflection part of an educational process poses two problems for teachers. One is to help students collect manageable amounts of highquality observation (that is, data that is independent enough of their own preconceptions that it has the power to surprise). And the second is to make reflection contribute directly to the course. Over the last few years, I have tried to make observation and reflection a part of my courses and have learned a lot from my mistakes as well as the occasional successes.1 Let me place this examination of reflection in some of its contexts. In the Context of Classroom Inquiry In the next two chapters, I want to examine the role of reflection in the context of two classroom inquiry projects in which both my students and I (with our differing reasons) were looking into what reflection might offer. As the teacher in the middle of these stories, I was not an independent observer but a participant whose interpretation of events on Monday could return to shape the action on Wednesday. Therefore, this account needs to Metacognition 237 be read with a recognition of how classroom inquiry shapes what people learn. Compared to an experimental design, classroom inquiry is not the best way to test a technique or to more generally validate the success of one’s own teaching as a recommendation to others. It is, however, a good place for discovery—to see how students are using or interpreting your teaching, to bravely examine the fate of your hopes and expectations, and to discover something of what your students are in fact doing or learning. In this process of classroom inquiry, I was joined by the other high school and college teachers of the Making Thinking Visible project. Although we were all interested in how to make the practice of collaborative planning work in our classes, our inquiries focused on students and on that mysterious link between our expectations and their experience. As this collaborative inquiry made it clear, teaching is a theory-building process; that is, it is a hypothesis-creating, prediction-testing process that leads to the framing and reframing of action. Theory-building is an act of mind by which teachers construct an imagined frame for actual pedagogy.... As a teacher, starting to run this complex scenario through my mind, I am spinning out a model of my students, imagining the assumptions, habits, and strategies they bring, predicting a dynamic interaction be­ tween these students, the task, the ethos of the course, other students, and projecting the cognitive process and intellectual stance I hope to teach and support. (Flower, 1993c, pp. 3—4) Classroom inquiry was a place to articulate those dynamic situated theories that guide teaching, to find out where hypotheses were wrong and expecta­ tions unmet, and to reframe our situated theories of practice. Unlike some more independent observations, classroom inquiry also helps create the event it studies. As my class and I discovered what was happening, we often used that knowledge to reshape what would happen next. So from one perspective, the interest of these accounts lies less in our particular conclusions than in the inquiry process they document—the attempt to understand some of the inner voices students were listening to and the strategic knowledge they brought to negotiation in this class. On the other hand, I have placed these classroom observations at the center of a more general theoretical debate on metacognition because I think these situated accounts can tell us a lot about the kinds of meaning that reflection lets students construct; that is, the particular insights my students “discover” are not independent of my goals for them or my ongoing interpre­ tations of the experience with them. Reflection is not a direct line to broad personal or public truths. But in the case I hope to build here, reflection may be the road to an even more useful, situated meaning making. 2j8 The Construction of Negotiated Meaning In the Context Under Negotiation The context for any act has many layers, some of which we as observers see, some of which we theorize about. Consider for a moment just a few alternative versions of the “context for writing” in which we could imagine these writers to be situated: The sociocultural context of Western society, its macrostructure shaped by economic and power relations, maintained by ideology, and reflected in material realities and practices of daily life, including school (Giroux & McLaren, 1989; Lave, 1988) The exclusive context of a discourse community and its set of discur­ sive practices that locate a text within academic discourse and fresh­ man composition (Bartholomae, 1985; Flower et al., 1990)—a com­ munity that stands in contrast, for instance, to the 17th-century emerging community of Royal Society scientists or the established expectations of physicists reading each other (Bazerman, 1985), or the hybrid discourse of a graduate student moving between an expressivist and empirical mode of thinking and speaking (Berkenkotter et al., 1988) The more explicit context of specific textual and literate practices: record keeping (Scribner & Cole, 1981), the essay (Scollon & Scollon, 1981), the bedtime story (Heath, 1982), the placement essay (Bartho­ lomae, 1985), biologists’ proposals (Myers, 1985), and the scientific article (Bazerman, 1988) The localized community of people that shapes learning from the historically changing communities of the disciplines (Graff, 1987) to places called school (Goodlad, 1984); to the social, ethnic background that demands a student’s allegiance (Fordham, 1988) and goes with a person when he or she leaves (Rose, 1989); to neighborhood groups organized around collaborative literate action (Peck, 1991) The even more intimate context of interpersonal relations: conversa­ tion (Cazden, 1988; DiPardo & Freedman, 1988; Nystrand, 1986; Sperling, 1990); the collective of classmates that share reading and writing (Dyson, 1986); and the peers and partners in collaborative planning, cooperative learning, and peer editing groups (Freedman, 1987; Gere, 1987) Each of these versions of “the context for writing” not only draws atten­ tion to distinctive features, it brings a language that interprets those features in theoretically loaded ways. The potential context for the metacognition in the following study included all these dimensions postulated by educational Metacognition 239 theory and more. However, for our inquiry, the most relevant version of context for understanding these reflections is the more limited arena we could call the context under negotiation. In contrast to the social and discursive structures outside observers might perceive, these reflections are located in those aspects of the rhetorical situation that are present to these writers and actively impinging on what they do. This is not to say that other contexts, other unrecognized forces, cease to exist or cease to have an effect, but the context under negotiation is the operative context for reflection. Let me be clear: this operative, present context, as I read the data, is not a writer’s solipistic, self-referential, self-generated fiction, his or her imaginary relation to reality. Nor is it the pattern of events, expectations, and material realities an ethnographer could document or a social theorist might postulate. In­ stead, the context these reflections help us to understand is created by the relationship between a writer’s personal representation of the rhetorical situation and the events, situations, and representations of other people that the writer is bumping up against, resisting, responding to, and reshaping. Context is a point of interaction. The context under negotiation is an interesting place, in part because it allows us to compare the writer’s representation of the context with other elements in the drama from the wording of an assignment to the canons and conventions of a discourse that the writer may misread, resist, or reinter­ pret. We can, for instance, compare the writer’s image of the teacher as reader to what the teacher wrote in an assignment, said in an interview, or stated as the assumptions behind a course. We can compare the writer’s intended text to the text a reader then constructs as she or he draws infer­ ences, forms gists, departs from the writer’s expectations, attempts a sympa­ thetic reciprocity, and succeeds or fails (Haas & Hower, 1988; Nystrand, 1986; Schriver, 1992). The idea of a negotiable context—as a relationship, a place where repre­ sentations are in contact and sometimes conflict—is a conceptual tool for understanding learning. It downplays the vast penumbra of contextual forces that might shape learning, to focus on what the writer either cares about, understands, recognizes, or has been forced by circumstances to see.2 The particular course that shaped these student observations was not typical, if any course is. What makes these observations worth our attention as images of possibility is not their typicality or uniqueness but the way in which students used them to respond to multiple layers of contex within and beyond this course—to other classes, to the writer’s past history as a student and writer, to his or her self-image, and to the relations of authority and agency. 240 The Construction of Negotiated Meaning In the Context of a Course In the course we will look at first, reflection was a tool students used to examine their own writing strategies. The course, Cognition and Rhetoric, focused on ideas from two sources: 1) theoretical articles on “what it meant to be a rhetor from different cognitive and rhetorical perspectives (e.g., Bartholomae, 1985; Bransford, 1985; Farb, 1974; Flower, 1993b; Freire, 1989), and 2) students’ self-observations of themselves as “rhetors” based on a variety of process-tracking techniques. It was designed to let students use various cognitive and social perspectives on what readers and writers do as a lens with which to examine their own process and to speak back to the theory. The 12 students, primarily senior and first-semester M.A. students majoring in professional writing, saw themselves as aspiring writers. This discussion draws extensively on work done with Barbara Sitko who joined the class on days devoted to students’ observations and reflections and coauthored our analysis of the papers (Sitko & Flower, in press). Assignments. The first three assignments were two-page papers in which students were asked to read chapters in Farb’s book Word Play, which offers a readable introduction to sociolinguistic research and to problems in language use, such as linguistic chauvinism, power relations, and discourse strategies. Students were asked to "’interpret or apply this chapter to the task of understanding how academic discourse, as you see it, works, using the chapter to analyze and explain some problem students encounter in using or learning academic discourse.” (Many students found the task surprisingly difficult, since it was neither a straight analysis of the readings nor a direct personal response, but a request to apply theory or research to a problem that they had to define.) They were also asked to monitor their own planning process, using strategies that will be described shortly, as a basis for their later observations and writing. The longer midterm paper then asked students to write about their own writing process and the way they used “such strategies to solve your own rhetorical problem or to develop your own purpose.” This paper was to be based not on readings but on the self-observations students had been collecting.3 Instruction, classroom discussion, and writing then focused on three elements: on theory, on students’ self-observations, and on the application of both sorts of knowledge to practical writing problems. Observation. In this class, as students experimented with new writing strate­ gies, they used various process-tracking techniques (logs, self-interviews, collaboration, and private thinking aloud) to observe their own thinking Metacognition and to evaluate how different writing strategies worked (or did not work) for them as they were writing their papers. Collaborative planning (which was developed with the help of this class) quickly became the strategy of choice and the most useful method of observation.4 In addition to taping their collaborative sessions, students did a self-interview on their plans before writing and a private wrap-up after the session. Taping the 15- to 20-minute sessions was nonintrusive; the social nature of the collaboration often elicited the writers’ best thinking; and the compressed nature of the session revealed the writers’ priorities and threw problems into higher relief. When this record was combined with the writers’ own knowledge, triggered by listening to the tapes or reading a transcript, students were able to make data-based reflections that were both richly supported and often surprising to the writer and the group. Reflections: How did it go? Reflection in this class was of two kinds. The first depends on taking an open, experimental stance to the strategies one is teaching and creating a forum that validates students’ observations by treating them as part of the knowledge of the course. Each out-of-class collaborative session was followed by time devoted to in-class reflection on what each writer observed. They followed a pattern I have since seen repeated when students are authorized to speak from the data of observation as the authority on their own experience. The initial question is usually, “How did this session/this strategy work for me?” For every student who has discovered new ideas, resolved a problem, or reversed his or her skepticism about collaboration, there will be students who felt little benefit (“1 was unprepared”; “I was past planning and didn’t want to revise”; “I’m not used to explaining my ideas”; “We used the blackboards as a checklist and found the discussion superficial”; etc.) This can be one of the most useful teaching and turning points in the course: sessions that do not work are as important to understand and talk about as ones that do (even if the teacher advocates the process in principle). Strategies carry no guarantee, so the teacher’s role is not to defend her instruction but to ask the writer: How could you design this process so it will work for you? What will you do next time to adapt the session so it achieves your goals and makes the best possible use of your time? How will you take advantage of any potential it holds? The question here is really one of ownership for learning. As we saw with Vince (“I’m only discussing this”), a strategy perceived as a “teacher’s process” is fair game for evasion. Other students may not even realize they have trivialized a potentially powerful strategy or turned it into a mere 242 The Construction of Negotiated Meaning checklist until they hear other students reflect on their own experience and what it bought them. Observation puts these problems on the table as negotiable questions. The use of valuable class time validates the significance of what students (as opposed to teachers) observe and their independent evaluation of how their learning is going. Perhaps more importantly, these observation-based reflections can make it suddenly clear that the responsibil­ ity for learning (for using and adapting a strategy to make it work for you) is in the hands of the student, not the teacher.5 Reflections: What did you discover? A second kind of reflection shifts attention to the larger question: What were you able to discover about your writing strategies, your assumptions, and yourself as a writer? The written reflections on which we focus in this chapter were a central part of this course. But one cannot assume that “assigned” reflection on writing will work as a learning process, since it can also sink merely to mirroring and mouthing the party line of the course. These reflections are interesting in light of some of these possibilities. For instance, it would be easy for self­ reports to be merely a regurgitation of the readings: students could repeat the claims of theorists and describe their own process in terms of abstractions, conveniently supplied by Bartholomae’s talk of stylistic discourse communi­ ties, Freire’s sociopolitical vision, Bransford’s cognitive processing theory, or Flower’s problem-solving concepts. Alternatively, students might be less influenced by their readings and rest with a set of truisms untainted by theory and uncorroborated by experience. For other students, the request to reflect might be interpreted as an occasion to generate a stream of con­ sciousness narrative about a writing event—such reflection could be a genu­ ine personal response but an underexamined, undertested representation that ignored underlying assumptions and processes. Perhaps worst of all, students might reduce and shape their planning to fit the instructor’s per­ ceived expectations and use their observations to report the dutiful use of a given dialogue or rhetorical planning strategy. Although one wants students to “try on” new strategies and points of view, what one hopes to see is not a perfect reflection but the refraction and transformation that goes with learning. For instance, these accounts clearly reflect the course’s problem-solving stance and emphasis on rhetorical purposes. The Planner’s Blackboard prompt almost certainly influenced the features students noticed in their own planning or noted by their absence. However, we also see students taking an experimental, evaluative stance toward these strategies. For me, the most significant feature of these reflec­ tions will lie in how they differ, as working theories, from the ways writing is represented in the assigned readings. Metacognition 243 Students’ Images of Their Writing Processes At the time of this course in 1988, a growing number of composition theorists might have described writing as a social or a cognitive process, but the literature in the journals (and the students’ readings) had little to say about how these mighty forces interact (Flower, 1989a). At best, the literature might have suggested a pervasive but inarticulate relationship, a twilight zone in which context would cue cognition in its silent, inexorable ways while the mind would inevitably construct a rhetorical situation in its own image. In short, the images of writing offered in the course readings had little to say about this relationship and even less to say about how affect (feelings, attitudes, motivation, and attributions) might influence students’ writing (McLeod, 1987). By contrast, the images students constructed of their own writing pro­ cesses were sites of dilemma-driven action that were marked to a surprising degree by the interaction of cognition, context, and affect. Trying to enter the world of advanced academic discourse, these writers were trying out new strategies required to apply theory to a problem they had to define. As they tried to articulate the process presented to them in the memory-aided data of tapes and transcripts, they described an active, conscious attempt to negotiate a confusing rhetorical situation. The process they constructed involved not only acts of cognition but sharply drawn images of the contexts that were shaping their writing, such as past confrontations with a news editor, projections of their future professional roles, and conversations in the hall. These images also involved clearly articulated statements of affect, including self-doubts and fear of depending on collaboration, balanced by relief at seeing their own needs for social support legitimated but qualified by a nagging assumption that real writers should not need any help. Finally, the multidimensional picture they drew was far more interactive, far more strategic, than their readings (or current textbooks) did justice to. In order to explore this interaction, Sitko and I used the following, some­ what reductive formalism of figure 8 to help categorize the writers’ comments (Sitko Sc Flower, in press). In reality, many of the comments writers made reflect the way cognition is suffused with affect or defined as a response to a given context or situation. This theoretical separation is merely a tool that helps illuminate the distinctive contribution each way of representing writing was making to the students’ images. If one envisions these reflections as a map of someone’s writing process, this map recognizes three semioverlapping territories. Within the category of affect in figure 8, one hears students reflecting on their attitudes, motivation, emotions, self-image, and so on. The process of 244 The Construction of Negotiated Meaning AFFECT attitude motivation emotion self-image r A J< CONTEXT history assumptions rhetorical situation COGNITION goals strategies metacognitive awareness other actions N J X J Figure 8. A schematic view of a writer’s representation writing is envisioned as an affectively loaded experience over which writers can often feel little or no control. But affect here also includes aspirations to be a writer, surprise at themselves, and the less conscious patterns of attribution—the ways students attribute their success or failure to luck, time, or ability versus effort, strategies, or experience. Within the category labeled context, students represent their writing pro­ cess as an event within their personal histories as students; it is seen as an expression of assumptions or expectations triggered by the situation or carried as unseen baggage from previous experiences. Writing is also repre­ sented as a contextually defined act when students focus on their “reading” of the class, the roles they play, or the ones they feel expected to play. In comments like these, the process of writing is represented as an essentially social act, seen at times as an almost unmediated response to past, present, or imagined contexts. Although all of these observations are obviously acts of cognition, the category of cognition in figure 8 is used when students represent writing in terms of a thinking process subject to their intentions and decisions or as a problem they are wrestling with. In the reflections writers talk about their goals, their strategies, the process they go through, and the choices they make. Such comments portray writing as a tapestry of mental actions. Other comments foreground not just cognition but acts of metacognition when writers stand back to reflect on any or all of these elements—on the context they are in, the feelings they have, and the thinking process they orchestrate. To explore the nature of these representations, we will look first at four students who built some distinctive, personal maps. The representation that our first student, Janet, constructed was one in which all roads seemed to lead to issues of affect. Metacognition 245 Janet: How Do You Motivate Yourself? From her very first protocol, Janet heard herself struggling with a problem of motivation. This attitude, however, was not mere disinclination; it ap­ peared to reflect an active conflict in her own relation to school. Here was a hardworking, responsible, and quite successful student. But now in her last semester of a M.A. program in professional writing, she saw herself as professional writer, not as a member of an academic community. She repre­ sents writing the three Farb papers as both an affective and contextually defined event: “In writing [the three] papers, I had a problem in motivation. I wasn’t interested in writing about academic discourse and could see no practical purpose in doing so.... I’m more interested in applying what Farb says to the task of writing in a medical community or in the computer industry... for which I can see many ties.” The links between context and affect stand out even more clearly in the tape she quotes from her collabora­ tive planning session: I mean, part of it was a reaction just to the topic of academic discourse, but it was bigger than that really: it was that I didn’t want to do these assignments at all because I really couldn’t see the purpose.... I have a different reason for being here, and some types of writing assignments or whatever that I’ve had to do since I’ve been here seem just more as get­ ting in the way of what I’m really here to do. Even the idea of getting good grades, which used to be a source of motivation, no longer matters much. The problem, she says, is “finding (that) I’m not identifying with this assignment; I’m not identifying myself as being a college student.” Given this emotional and social context for writing, the strategies Janet found herself using come as little surprise. She describes herself as “satisfycing” on the second paper: “1 went and did a response paper with no real purpose at all in mind, you know, just to say something.” However, the plot thickens. When she received some positive feedback on this paper (despite her limited effort), the context appeared to be sending messages that were out of synch with her personal evaluation of her process. The situation seemed to her “completely contradictory.” To top it off, when she mentioned the problem of motivation to the instructor (one of those conversations walking down the hall), she was encouraged to shift the topic to nonacademic discourse because the real point of the assignment was for students to use Farb for a purpose of their own. The constraints upon her were in part her own making. Figure 9 captures some of this confusing array of attitudes, expectations, actions, and social interactions. 246 The Construction of Negotiated Meaning ( A ( AFFECT low motivation grades no help feedback confuses wants a purpose \ CONTEXT task is academic role is professional writer returning to school feedback encouraged to shift assignment, find purpose COGNITION goals are to satisfice write response options grades (weak) change topic (against rules) write for audience (only at work?) personal purpose (difficult but essential) can we massage assignments? do we impose constraints? N y Figure 9. Negotiating affect This abbreviated sketch of Janet’s representation illustrates an interesting complexity. In imposing our interpretive framework of cognition, context, and affect on this story, we inevitably reduced detail to highlight a more general structure. But even this abstract representation does not add up to a set of simple cause and effect relations with school as the boogey man or to a map in which one element single-handedly determines another. Instead, we see a tangled web of intentions and feelings. Despite this writer’s feelings of resistance, she does not wish to abandon the situation or her writing: she describes herself caught in a “vicious cycle” created by the problem of motivation followed by frustration at her inability to forge a meaningful purpose for herself. When Janet presented this view of her writing process to the class, fore­ grounding the contradictory rhetorical context she perceived, the frustration she felt, and the low-effort path her thinking had taken, her talk evoked sighs of sympathetic recognition in the class. It was a situation everyone had seen before, and the question as they put it was: How do you, as a writer, negotiate such situations when they occur? As Janet and other students put it, “How do you motivate yourself?” Janet’s midterm “reflection” paper and the collaborative planning session that led up to it offer some insight into that negotiation and into this writer’s process of constructing metacognitive awareness. In them, she examines both her feelings and some strategies she might have used to break the vicious cycle and motivate her writing. She considers four key options. One Metacognition 247 was grades, but they simply did not matter enough to work. A second option was to change the topic, and in fact, her voluminous planning notes show she wrestled with this possibility. Why was this option abandoned? Here Janet s observation shows how cognition, the strategies she used, was bound up with her image of the context of school writing: Ultimately I didn’t feel comfortable changing the assigned topic. I wasn’t confident that I could break the rules of the game without penalty and so I assumed a constraint [italics added] to write about academic dis­ course. In retrospect, I believe many of the points Farb was making about speech communities could be generalized to apply to any discourse community and I could have legitimately stated that a particular point in Farb could be applied to technical writing in the same way it could be applied to academic writing. Alternatively, Janet reflects that she could have tried to take a more personal perspective, to write about something more personally relevant. But as she notes in her paper, “Writing from a personal perspective is difficult to do well and not widely accepted [even though, she feels] it is an essential technique to learn.” So that option died, too. Janet’s reflection is the story of opportunities not taken, the sort of story that makes a teacher wince with, “If only I had realized sooner....” But more importantly, perhaps, it is also the story of a writer examining her own role in negotiating this link between affect and cognition. The arrow in figure 9 symbolizes this active negotiation. In the collaborative planning session that preceded her reflection paper, Janet and her partner begin to examine how they deal with motivation in other contexts.6 Janet: I mean, I’m thinking like [about ] working and writing things at work that you really don’t want to write, but you need to write. In some cases they’re easy enough that it’s not a big deal. I mean, writing a memo... you just do it... But even, say I’m writing a manual [....] I think there... linking yourself to the audience or reader can help moti­ vate you. Because you can say... “I don’t really want to do this. I don’t really have any interest. Supporter: I’m not interested in it. I’ll never use it. But somebody else... Janet: But how can I explain this really well? How can I write a really good manual that, you, that a reader is... Supporter: It can be a challenge in that sense [....] You know, pride comes into play. Later in the planning session, Janet’s collaborator adds another piece to this picture of negotiation. She begins in a tentative way: 248 The Construction of Negotiated Meaning Supporter: How could you if you were given an assigned topic that didn’t motivate you in and of itself, is there a way to kind of massage that topic to... to try to find a way to approach it that maybe is a personal approach, but still somehow tried to deal with the topic? Um, I guess what I mean is maybe is, are there ways to try um... to shape a topic to fit a purpose, somehow? And I’m not [....] I—I haven’t re­ ally thought about this before. Janet: Yeah, I think I know what you mean. Finally, Janet’s representation of options expands yet again when she begins to recognize her own place as the interpreter of her context who can, in fact, try to negotiate this link between affect and cognition. “But so, I think that is one thing you can do is write personally. And, in fact, even with that first assignment I toyed with not even addressing academic discourse. I toyed with completely ignoring that and doing something else. And yet, 1 felt that I really couldn’t maybe... which, I don’t know, I mean maybe that’s just something we impose.” Janet’s representation of her writing shows a process in which cognition, context, and affect are linked with multiple, intricate threads but one in which those links can also become the subject of reflective awareness and renegotiation. Terri’s representation provides an interesting contrast, for here, the path of negotiation goes between cognition and context. Terri: I Couldn't Believe It Was Happening to Me Terri’s first paper came back with clear signals that the instructor found it disorganized. It told an interesting story about a childhood experience, but the purpose of the story (and the paper) was not stated. Its point, which was hard to infer, did not appear to link the individual paragraphs together. The paper was not meeting the instructor’s expectations for academic dis­ course. “I realized that I had on my hands a very serious problem that I never had before: I simply could not focus. But what was worse was that I didn’t know why I couldn’t focus nor did I know what to do to learn to focus again.” For Terri, the fact that she had a problem at all was a source of shock and bewilderment, but if we look at how she represents her own history and the context of writing, we can understand where these feelings came from and why her strategies for organization took the form they did. Since I had been taught to write like a journalist since high school, I had integrated this sense of organization... into all of my writing without re­ alizing it. In college while my writing and journalism classmates were writing and rewriting stories, I was spewing them out with amazing sue- Metacognition 249 cess.... I was promoted to an editorial position because I know how to organize a story and could spot another writer’s disorganization and eas­ ily remedy it. So when I came to [the M.A. program] and acquired a seri­ ous writing problem I couldn’t believe it was actually happening to me.” Terri’s history brought with it not only feelings of confidence but a set of tried-and-true strategies for “spewing out” stories and creating a linear, associative development of ideas. Unfortunately, these strategies, apparently well adapted to one literate practice, were not helping her write analytic papers that required a more explicit rhetorical purpose. For Terri, as figure 10 shows, reflection led to active negotiation. First, it let her create a far more clearly articulated sense of her own history and assumptions; it let her see how history had translated itself into writing strategies. In creating this reflection, Terri’s feelings of insecurity did not magically vanish but were, in fact, foregrounded by the opportunity to express them. However, in taking thought, she translated this problem into a strategic problem, a question of learning a new language and style of writing. Her way of negotiating this situation was to figure out how she could reinvent a strategy. “I started studying the papers from all of my classes [looking for causes, symptoms, and solutions].... I started seeing white space and lots of little paragraphs in the successful papers and darkness and crowded sentences in the others.” Based on this observation, Terri began to experiment with a new strategy of starting with small paragraphs and then trying to connect them rather than trying to maintain the focus of those disastrous long paragraphs. Terri’s experiment did not produce miracles. The marvel here is the very fact of AFFECT has a problem cannot tell why had confidence had self-image as a A J CONTEXT A feedback past success as a story writer old and new texts COGNITION goal: learn to focus past strategies spew out lead reader write little U experimental strategy link little D Figure io. Negotiating context 25° The Construction of Negotiated Meaning the experiment. She dealt with the “shock” to her self-image (a significant affective problem writers may often face as they cross the boundaries of new literate practices) by trying to negotiate the relation between her writing strategies and her own history. In reflecting on her process, she came to see her history in terms of strategies (indicated by a link between context and cognition in fig. io), and she came to see her own cognition and emotionally laden self-image in terms of the transition into a new discourse community. The concept of changing discourse communities itself offered a new way to think about context that was an alternative to “past success as a writer.” The new representation Terri forged created a manageable problem. Her new strategy was not plucked out of a textbook but was a metacognitive invention built on the foundation of reflection. Responding to Reflections Reflections unveil difficulties that implicate the teacher as well as the learner. One of my responses to Janet and Terri as a teacher is to assume responsibility for the dilemma—“If only I had realized, been a more sensitive teacher, spelled out more options, softened my criticisms,” and so on. The inevitable “if only’s.” Another response that also takes the dilemma out of the students’ hands is to see it as yet another piece of evidence for the oppressive structures of school that make writing an inauthentic act for Janet and that fail to validate the personal narrative voice with which Terri feels most comfortable. With this thought, our gaze shifts from the students to a critique of the system. Both these responses recognize our responsibility to make school a better place, but they do not appreciate the significance of what these students themselves did as critical agents. As Janet and her partner point out to us, the thorny problem of “motivating oneself” is a necessary part of writing in many contexts. And Terri had chosen this program, not journalism, because she wanted capabilities she did not already have. Her problem was managing a rocky transition. In her own way, each of these women was confronting a genuine dilemma. Teachers cannot eliminate the dilemmas that accompany learning even if they can affect them. The real achievement here is that of the students, who not only confronted a genuine issue but used these reflections to engage in a sustained dialogue with it. Reflections like these underscore that learning and literacy are an encoun­ ter with a series of genuine dilemmas. In the face of conflict, reflection can become a form of significant action. In their different ways, Janet and Terri constructed images of a writing process in which the elements signified by Metacognition 251 the terms cognition, context, and affect are linked by multiple, contradictory, visible, and invisible lines of interaction. For them, observation-based reflec­ tion became an attempt to understand these elements of their own experience and to negotiate the lines of influence they saw. Their effort to track their own cognition became an act of strategic metacognition. The tangled webs of feeling, thinking, and social action they constructed did not become untangled. But with awareness seemed to come a sense of options and the opportunity to mediate both context and feelings and to translate under­ standing into action. In the final two cases, we will see students who cast themselves even more explicitly as problem solvers and used their reflections to sketch plans for strategic action. Marla: A Great First Sentence Marla looked on her writing process during the first eight weeks of the course as an experiment designed to alter a habitual strategy that had led to “outstanding first sentences and half-baked ideas.” By midterm, she had experimented with three new strategies, evaluated what worked for her and what did not, analyzed the results of her experiments, and formulated a new strategic goal: to develop a personal method of inner dialogue, “the kind of dialogue within myself that Freire talks about.” Tracing the evolution of her new awareness, Marla begins with a descrip­ tion of what she observed as an habitual writing strategy. It is comprised of three typical steps, one more inefficient than the other. When assigned a writing task, my first move has been to randomly flip through a text looking for a key point. Then, having had a moment of Eureka 1 flip through the text or texts in search of topic information, pulling out data that supports my thesis while ignoring everything else. Finally, because my key point usually occurs to me in the form of an introductory sen­ tence, I plunge into writing with nothing more substantial than a great first sentence and a tenacious hold on a vague idea. Rereading the transcription of her first protocol, Marla observes the precise way in which this strategy failed her, which she represents as a lack of planning: “At one point, after about ten minutes of trying to pull something out of the chapter, 1 found a single sentence that looked promising. My reaction to it was ‘Here’s something interesting; maybe I can use this to talk about a freshman learning to write in college.’ But because finding that single sentence was all the planning I could do, writing the paper was really difficult and I got caught in the middle with nothing else to say. The goal Marla 252 The Construction of Negotiated Meaning sets links the new strategies presented in the course to her own specific needs: “In learning three new strategies, I was looking for a way to solve my problem of substituting an idea for a plan by doing a different kind of thinking before writing.” In effect, she turns the instruction the course provided into a personal experiment. At first, she finds her old habits hard to break, recording that she modifies the strategies to fit her habitual pattern of closing off planning prematurely. Thus, in using the first strategy, dialoguing, she limits herself by working only at a local level, with “two or three sentences rather than the general argument of the chapter.” When such a piecemeal approach does not result in “something interesting,” she “resorts to old tactics,” returns to her habitual method of searching for a key sentence, and turns out what she perceives to be an unsatisfactory paper. The Planner’s Blackboard becomes the object of a second experiment. She finds that the method helps her to reach one of her goals, to produce a lot of ideas. However, she records encountering a new problem, one that she attributes to her inflexible use of the method. Because she permits herself to become “busy filling in one box after the other,” she misses opportunities to develop a single major point. In writing about this attempt, Marla sees an alternative—to use the blackboards as a source of direction, letting go when she is on the trail of a good point. At this point in the course, Marla has articulated for herself one of the key dilemmas of learning to write— strategies only become useful when one can use them to fit one’s own needs and purposes. The third strategy in Marla’s experimental tale is collaborative planning, which she evaluates as “the most successful way of improving my planning.” To begin with, it solves her earlier problem of providing direction. By providing focused questions, her collaborator helps her “to complete and explain my ideas by filling up all of the blackboards. My protocol for the last paper was much more focussed, without the mental wandering characteristic of the first two, because my collaborator was asking me ques­ tions and prompting me in ways that helped me direct my thoughts and elaborate on good ideas.” On the basis of this post hoc learner’s experiment, Marla sees her midterm paper as a place to “work all three strategies into something of my own.” Her method for designing this new personal integration is to begin with the most successful strategy, the collaboration, to examine these sessions (one of which was transcribed) for evidence of what makes the collaboration work and then to use her knowledge to design “something private,” a strategy usable in the absence of another person. This private strategy turns Metacognition *53 out to be a set of questions based on specific moves she observed her collaborator using: To get me to focus on a key purpose my collaborator would ask me if she had gotten the gist of what I was talking about by saying, “So what you’re saying is....” She helped me make transitions by asking, “Now does that fit into what you were saying before?” Similarly, she elimi­ nated excess baggage by saying, “That’s interesting, but I don’t see how it fits with the point you’re trying to make.” She also helped me think more deeply about my audience by offering different ways of interpreting points which I had thought perfectly clear. Finally, my collaborator pre­ vented me from plunging into writing with an incomplete plan by asking me what I was going to do with a idea, or how I was going to tie my thoughts in with the assignment. To test this construction, Marla actually carried out a kind of experiment. She reasons that if it is, indeed, the questions that help her to solve problems, perhaps she can use them as a kind of “private collaborator.” Her evaluation of this variation records her dissatisfaction, however. In attempting to reduce the collaborative method to the content of the questions, she feels that she has created a mere “checklist, helpful, but hardly the imaginary collaborator I expected it to be.” The problem is that the list does not of itself suggest alternatives, “coming up with different ways of thinking about a problem or getting rid of dead weight.” Marla ends her paper by speculating about a way she might solve this new problem by developing a living inner dialogue. Marla’s plan for an “inner dialogue” is an interesting example of metacognitive awareness, in that it leads her into thinking over not only how to use strategies but how to design, control, and test them. For instance, in assessing her use of the blackboards, she says, “I know I am on the right track if I can see links between them in my plan.” What limits her use of these strategies is her own inflexibility that makes them “uncomfortable because I have placed restrictive limits on ways to use them.” Marla’s inner dialogue is her step toward being a more adaptive planner. It is hard to say if Marla’s new, experimental strategy will do all she hopes. In fact, when Marla and another student presented their plans to become their own “collaborators” to the class, it allowed us to examine assumptions that “good” writers do not depend on others, that they learn to work alone, and that regular collaboration is cheating or at least unprofes­ sional. The success of collaboration had itself created a generative dilemma for some students. Whether Marla’s new plan solves the problem or not, the act of developing it called not only for awareness and control but for a critical questioning and a creative stance to the ideas taught in the course. 2-54 The Construction of Negotiated Meaning This writer searches the data she collected in a goal-directed way, trying, first, to discern how she typically goes about writing. Having characterized her habit of “outstanding first sentences and half-baked ideas,” she searches for evidence of whether the strategies introduced in the course are working to solve her individual problem. When they work, she specifies what makes them work, such as a collaborator’s offering alternative ideas. When they do not work, she analyzes why and catches a glimpse of her own self-imposed inflexibility. It is apparent throughout this appraisal that her purposes are both diagnostic and creative. She wants to analyze “inefficient steps” as well as “work all three strategies into something of my own.” Although the prose of problems, experiments, steps, and solutions reduces a complex experience to a manageable story, it also records a student grappling with a significant educational issue realized in the terms of a personal dilemma. Tanya: Awareness Makes Progress Possible Like Marla, Tanya reads her protocols and reviews her papers, looking for “patterns and similarities” that would let her define problems and their causes. Unlike Marla, she does not see the patterns in the protocols them­ selves, perhaps because of feeling “artificial” while recording her planning sessions. Rather, the protocols function to “trigger the memory of what I was thinking during the planning stages of the first three assignments.” As she reviews this data her “frustration” at not being able to diagnose the causes of her difficulties turns to surprise: she is “startled” to see that some strategies appear to work for her even when causes are unknown. With this, Tanya decides to turn her attention from unlocking the secret of causes to designing helpful strategies. The following description traces this odyssey. Tanya’s initial goal is to locate writing problems. She finds three. The first is “finding my purpose for writing.” In contrast to Marla who believes that she cuts the planning dialogue short, Tanya records that she prolongs dialogue with the material, paraphrasing and interpreting “long after I was fully comfortable with the material.” She speculates that she fills up the time dialoguing with the text because she does not have anything of her own to say. In the absence of an alternative strategy, but knowing that she should be doing something, she reiterates Farb’s ideas. Tanya’s second problem resembles the first. She records that having decided on a purpose, she writes down everything she “knew or thought about my purpose whether or not it was relevant to the assignment.” Her problem is that having committed ideas to paper, she finds it difficult to decide what to keep and what to discard. Once she has made these decisions, the next step of outlining comes easily to Tanya, but a new problem arises at the point where she has to “make the jump from an outline of sentence fragments to coherent prose.” Metacognition 25T Notice here that Tanya and Marla refer to a common problem—finding a focus and purpose—for which it would be easy to prescribe some standard generic solutions. But while one writer needs to produce more ideas before she leaps to text, the other needs to reduce and channel an uncritical, associative invention process. The solutions for the two women are different because they arise out of a distinctive individual matrix of strategies and assumptions, a matrix that only emerges when detailed observations like these take us below the surface of the generic problem of focusing. Tanya apparently spends a good deal of time attempting to understand the causes of these problems, an exercise that ends in the frustration. During her analytical review of protocols and papers to ferret out causes, however, she notices a puzzling phenomenon: that simply by using the blackboard strategy she appears to solve at least the last of the three problems: “It was while reviewing the tapes and notes that are relevant to my third trouble spot that I realized that this problem was not as severe in the second and third assignments. After we had introduced the blackboard strategies in class, I had an easier time moving from outline form to written prose. This surprised me.” Closer examination reveals that using this strategy leads her to write more complete sentences, chunks of text that are easily integrated into the final prose form: “I realize now that using the blackboard strategy allowed me to elaborate on the features of my outline before I felt pressured to compose good text. I began to form complete sentences while filling in the blackboards, so I did not encounter the abrupt transition that I faced in the first assignment.” “Re-searching” this surprising result produces a second realization when Tanya concludes that her initial frustration may have been self-imposed: Recognizing that the blackboard strategies can minimize one of my writ­ ing problems is significant because it illustrates how 1 imposed the obsta­ cles to progress myself. I had unrealistic expectations of my metacognitive powers, and this prevented my seeing the value of a strategy. In fact, 1 was doubting that any strategies could be valuable until I reached some imagined higher level of understanding. I realize that my first and second trouble spots are more severe than the third, and effective strategies will not be easy to find or employ. However, I recognize that such strategies could exist—or that I could design them myself—and this awareness makes progress possible where before it was not. Tanya appears to be aware not only of the power of a particular strategy but also of her ability to control her writing process by designing personal strategies. Further, she knows something of self-imposed limits. In looking at the representations these writers constructed (i.e., our inter- 256 The Construction of Negotiated Meaning pretive analysis of them), the last two illustrate students whose text focused primarily on the cognitive dimension of our map: goals, strategies, inten­ tions, and decisions. Both discover something surprising when reading the transcriptions of their protocols. Both report using, evaluating, and rede­ signing strategies to suit their needs. Both record insights about self-imposed limitations. Both engage in some theory building, attempting to link data about their processes by making and testing the connections they think they see. Only a longitudinal study could look at the durability of their learning, the long-term effects of their newly designed strategies, but even a cursory reading of their texts shows that these writers are self-reflective and obviously motivated to actively participate in their own learning. Although they focus on cognition, these writers build representations that include affect and context. Tanya is “discouraged and frustrated” trying to assign causes to her writing problems, while Marla counts on her “Eureka” experience for the momentum needed to begin writing. For both, this af­ fective language is a part of the problem they hope to change. For Marla, the “Eureka” feeling creates an illusion of readiness to write. If she succumbs to its temptation to begin text on the strength of this spurt of energy, she soon discovers that the impetus is a single introductory sentence containing one key point. The problem is not the feeling itself—she has no desire to get rid of her “Eurekas”—but its lack of staying power. Her solution is to design other ways to sustain the energy needed to produce a complete text. Similarly, Tanya’s frustration at her inability to attribute causes to her problems does not disappear, but she chooses to focus her attention differ­ ently. Instead of concentrating on what she cannot do—assign causes—she concentrates on what she has proved she can do—design strategies that work. A sense of context is also present in the way students represent writing as a social act and as an event within their personal histories as students. Marla and Tanya use reflection to talk about usual problems, not just the writing in this course. And in trying to understand her own cognition, Marla highlights her collaborator’s suggestions, and Tanya uses a discussion with her husband to affirm her insight that one can work on a problem without understanding its cause. While Janet and Terri emphasize the influences of affect and context and Marla and Tanya focus on cognition, all four create representations that link social histories and personal feelings with a sense of new control over their own cognition. Some Patterns of Strategic Knowledge in the Class How shall we characterize the knowledge these writers constructed? And how does it differ from the familiar discourse of academic theory and text­ Metacognition 257 books? The cases of Janet, Terri, Marla, and Tanya were selected to illustrate four representations of writing that focused in differing degrees on affect, context, and cognition, but they also illustrate some ways of thinking and talking about writing that are characteristic of the reflections as a whole. First, the knowledge that emerges from such reflection is strategic knowl­ edge. Students not only recognized their own strategies, maneuvers, tech­ niques, habits, they described them in terms of the goals that were driving such strategies. These goals ranged from old and problematic ones, such as quelling anxiety by getting something/anything down fast, to personal goals, such as Ann’s skeptical need to test my claim that the role of “intellectual,” could apply to her, a forty-year-old returning student. For her, the dialogue strategy was not a homework exercise but a private test of our conversation and of the unlikely notion that she might have something to say back to academic authorities. Unlike the incidental reflection and strategic knowl­ edge we observed in the transcripts of freshmen, there is a striking awareness with these junior, senior, and M.A. students of where one’s goals come from and why writers cling to them, of whether the instructor’s advice really works, and of where new options lie. Secondly, the strategic knowledge the writers built was clearly driven by dilemmas. It emerged out of a sense of problems and desires. Clearly, some of these were triggered by my critical comments (as on Terri’s anecdotal alternative to analysis). But, as 1 think these excerpts suggest, within this advanced writing class, these writers quickly assumed ownership of their own dilemmas. Their reflections were highly action-oriented, whether the goal was developing personal strategies or simply understanding oneself as a writer. To show how these four case studies both typify and differ from the class as a whole, we analyzed the midterm papers of the group in terms of the features highlighted in the case studies/ Table 5 shows a profile of these reflections along three key dimensions that Sitko and I defined as representation, problem awareness, and control. We chose to look for the presence of awareness and control because these two features figure promi­ nently in the literature on metacognition and in the performance of our case study writers. Representation. How is the writing process represented? In terms of cogni­ tion, context, affect, and/or the integration of the three? Problem Awareness. Does the writer’s metacognitive awareness take the form of detecting and defining problems, creating more elaborated images of those problems, and/or making causal attributions that inquire into the source of problems? z$8 The Construction of Negotiated Meaning Table 5. Representations of Metaknowledge in Midterm Papers Note. Aft = affect; Cog = cognition; Cxt = context, lnt= integrated; Dect = problem detection; Attr = attribution; Elab = elaboration; Altr = alternatives; Stra= strategies; Eval = evaluation. Control. Is metacognition operating as a source of control in the form of imagining alternatives, mapping out new strategies, or specifying a way to evaluate one’s own success? The following quotations from papers of three writers not previously discussed illustrate statements of each category. Metacognitive Awareness Problem detection. Problem detection is signaled by statements such as these: I was struck by how often I encounter difficulty in the middle of writing a paper as the result of insufficient planning at the start (Writer 2). This problem is giving a general overview of a topic instead of giving an analysis of an issue that directly relates to my purpose or key point (Writer 3). Problem elaboration. As we have seen in the research on revision, detecting a problem is a good beginning, but a more elaborated representation offers a stronger basis for action (Flower, Hayes, Carey, Schriver, & Stratman, 1986). The following elaborations come on the heels of detection statements: What particularly interested me was discovering that my planning tracker interviews revealed that I was caught by surprise when my initial plan proved so problematic (Writer 2). In the Blackboard Planner paper, Metacognition 259 I had numerous examples of an issue, but I never defined a problem that students have when writing for academic discourse (Writer 4). Causal Attribution. Some writers extend their understanding of a problem by attributing its cause to some particular force or event: When I asked myself why I am not able to give an analysis of one issue that directly relates to my purpose, I knew that my problem began with the fact that I take too many notes on both the text material and on my own thoughts and ideas (Writer 3). 1 saw that I was so pleased with my­ self for generating so many ideas, that I ignored that important clue (Writer 4). Metacognitive Control Statements indicating awareness are often accompanied by statements indicating control, which in this case took the form of comments about alternatives, strategies, and evaluation. Alternatives. Alternatives are indicated when a writer juxtaposes several ways of acting: These strategies now have names, and just that