7.3 Metaphors We Teach By

Given this background, we are now in a position to present and justify some of the grounding assumptions of our version of constructivism. We have separately (e.g., Cunningham, Knight & Watson, 1994; Savery & Duffy, 1995; Duffy, 1995) and jointly (Cunningham, Duffy & Knuth, 1992) offered such assumptions before, but never within the context of a model of "mind as rhizome." This addition has helped clarify our own thinking and, we hope, the readers'.

7.3. 1 All Knowledge Is Constructed; All Learning Is a Process of Construction

In accord with the MAR metaphor, all knowledge is local, a slice through the rhizome. Since all connections are, in principle, possible, we must stress that we are not talking about a partial or incomplete version of the "truth," the world as it is unmediated by sensation, perception, or cognition. Elsewhere we have talked about the concept of umwelt (Cunningham, 1992), a term coined by Jacob von Uexkull (1957) and discussed in his brilliant paper "A Stroll Through the Worlds of Animals and Men." In brief, the term means phenomena] world or self-world, the worlds that organisms individually and collectively create and that then serve to mediate their experience in the world. It is these structures that determine a world view, the things we notice and ignore, the things that are important to us and not important, the means by which we organize our lives. This umwelt, determined jointly by species-specific factors, the sociocultural history of the community, and particular experiences of the organism in a given environment, characterizes that organism's behavior.

In humans, this process of construction (or semiosis, as we prefer to call it) is unique in the universe as we know it. Structures are created which go beyond the immediate experience of the cognisizing organism. Words, pictures, mathematics, bodily movements, and the like generate structures of knowledge and objects that need have no basis in the "real" world and which can be manipulated independent of any such world. According to Deely (1982), it is the intervention of language that allows humans to engage in this particular type of sentiosis. Through language, we create culture: institutions such as religions, governments, armies, schools, marriage rites, science, and so forth. Culture, in turn, impacts our lives by determining what is important and what is not, what makes sense and what does not. The culture then makes these constructions available to the young and to new initiates for appropriation and use in transforming their participation in that culture.

Learning, then, becomes a matter of changes in one's relation to the culture(s) to which one is connected--with the gradual transformation of one's means of constructing one's world as a function of the change in membership in that culture. Lave and Wenger (1991) discuss this in terms of legitimate peripheral participation: a transformation from newcomer to old timer. These cultures can be conceived at various levels (e.g., caregiver, family, school, church, ethnic community, vocation, nationality, etc.), activities (working, playing, talking, eating, etc.), tools (hammers, computers, televisions, cars, etc.) and signs (language, music, art, etc.). A complete explication of such a view goes wen beyond our purposes for this chapter, but it is necessary to stress again the constructedness of our knowledge and the need to provide experience to learners of that constructedness and the means by which they can participate in that process. This will be detailed in the sections to follow.

7.3.2 Many World Views Can Be Constructed; Hence There Will Be Multiple Perspectives

We and other constructivists are often accused of advocating a kind of naive relativism or constructivist extremism. Since we argue that all knowledge is constructed, it is assumed that we must therefore accept the claim that every individual constructs his or her own meaning, untroubled by the realities of the real world or contact with other individuals. For example, Schwen, Goodrum, and Dorsey (1993) propound that they "find the extreme view of all knowledge being relevant to personal experience, and therefore idiosyncratic, too impractical and anarchic to be useful" (p. 6). But such a fundamental misconstrual of our position precisely illustrates our second principle and is nicely captured by Eco (1984, p. 12): "A world view can conceive of anything except an alternative world view."

As the MAR metaphor makes clear, knowledge is a construction, not by an individual in some pristine, autistic isolation but by participants in a community that simultaneously transforms and is transformed by such participation. What we choose to call knowledge is a consensus of beliefs, a consensus open to continual negotiation (Rorty, 1991). Such a process does not mean that the community will inevitably and perpetually debate whether the sea is blue or green, whether the word dog will continue to refer to a four-legged, domesticated, carnivorous canine, whether the Earth orbits the sun, or whether God is dead. A pervasive and largely benign effect of the structure of knowledge that we construct from the rhizome is that we tend to use those structures through which the world makes good sense to us, that seem "right." And we tend to assume that others see things in roughly the same way we do, that our world view is constructed as largely invisible. Providing experience that elevates our world view to a conscious level typically entails bringing up alternative views for comparison, as when we study cultures different from our own.

In a classroom with which we have worked (Cunningham, 1994), the teacher is exchanging material (stories, letters, photographs, HyperCard stacks, etc.) with a similar classroom in Northern Ireland. The children in both cultures are constantly surprised by the differences that have been revealed, from simple things like the way a date is written or the likelihood that the family owns a car, to the extreme, as when the children in Northern Ireland talk about the "Troubles" (the sectarian violence). The children in both cultures are invited to put themselves in the perspective of the other and examine their own cultural practices based on this new perspective. What would it be like to live in a town where army patrols can be seen several times a day? To come upon a policeman with a cocked semiautomatic weapon? On the other hand, what is it like to live in a culture where person-on-person crime is common (s much crimes are rare in Northern Ireland in comparison to the U.S.)? Even a term like integrated school has fundamentally different meanings in the two communities.

The "reality" of multiple perspectives should be a cause for celebration and optimism, not for fear that we will sink into some kind of utter subjectivism. Those who hold MAC view of mind expect and encourage acceptance and closure of a world view, while the MAB and MAR metaphors anticipate and encourage debate. It is this engagement with others, this establishment of the need to continually expand our web of understanding, that creates the awareness of multiple perspectives.

7.3.3 Knowledge Is Context Dependent, So Learning Should Occur in Contexts to Which It Is Relevant

Speaking of debate, this is a principle about which the authors have debated long and enthusiastically. While we are in sympathy with views of our colleagues concerning the need to situate (e.g., Brown, Collins & Duguid, 1989) or anchor (CTGV, 1992) learning in authentic, relevant, and/or realistic contexts, we disagree about why this is important. One of us (DJC) draws inspiration from the systems theoretic view of Maturana and Varela (1992), who argue that to say that someone knows something is to make the claim that she is acting effectively in a particular context. Thus to claim that Stephen W. Hawking knows science is to assert as valid that he behaves effectively in the domains of action that are accepted by the scientific community, a community that he in fact helps create. Knowledge is effective action. The contexts within which one can act effectively is an empirical matter. (See also Lave & Wenger, 1991.)

It is on the point of that empirical matter and its practical implications that the authors part company. The other author (TMD) agrees that to know something means that the individual can act effectively in a particular context. The concern is not with knowledge but with learning and, perhaps most importantly, with the issue of transfer. That is, if I want to prepare myself to be a scientist, what sort of learning activity must I engage in and in what sort of environment? It would seem that the general statement of this question--i.e., how do we prepare ourselves to act effectively in particular contexts?--is central to our development as individuals and as a society. It is certainly central to the instructional design community.

Thus the question of context is really a question about what aspects of the context must be represented if the learning (knowledge) is to be used (elicited?) in other contexts. TMD discusses this issue at length elsewhere (Honebein, Duffy & Fishman, 1993). In brief, the focus is on the qualitative character of the metacognitive and cognitive processing and the skills required. The physical character of the environment is relevant only to the extent it impacts the character of the "thinking" and skill requirements.

This entire issue of learning and transfer raises a problem for both authors and for most constructivist theory. It has been labeled the "learning paradox" by Fodor (1980) and Bereiter (1985). The MAR view, and more generally the sociocultural constructivist's view, stresses the distribution of cognition in the environment. While Vygotsky discusses the internalization of social experience (with the implication that knowledge is internalized and hence stored), more recent sociocultural theorists have suggested that a better translation of the original Russian is "appropriation" rather than "internalization" (Rogoff, 1990). In this way, the concept of distributed cognition (the rhizome distributed across minds and cultural artifacts) would seem to be preserved. However, as Ann Brown, Ash, Rutherford, etc. (1993) have noted, this shift in terminology does not resolve the learning paradox.

We certainly do not have the solution to the teaming paradox, and our own debate will continue. DJC worries that a view that posits abstract, generalizable operations, divorced from the contexts within which they were developed, will lead back to a MAC metaphor, while TMD worries that the focus on knowing fails to help us move forward on the critical issues of learning and the design of learning environments. Stay tuned.

7.3.4 Learning Is Mediated by Tools and Signs

In many ways, this assumption lies at the heart of constructivism as we view it. Wertsch (1994) agrees, asserting that any adequate theory of higher mental processes (i.e., beyond perception and involuntary attention) must be grounded in the notion of mediated action. Vygotsky (1978, p. 57) has argued that children's development proceeds on the basis of appropriating mediational means from the sociocultural milieu: "Every function in the child's cultural development appears twice: first, on the social level, and later, on the individual level; first, between people, then inside the child.... All the higher functions originate as actual relations between individuals." Vygotsky has proposed two mediational means: tools (technical tools) and signs (semiotic tools). The distinction is a slippery one, and particular examples often move back and forth between (even straddle) the two categories. But consider a hammer as a prototype example of a technical tool. How does the appropriation of this tool from the sociocultural milieu mediate action? As the needs of the culture encouraged the invention of a hammer as a more efficient means of driving posts into the ground or joining two boards, the hammer itself altered the very nature of carpentry itself. While it is true that the goal of driving a nail into a board is mediated by the use of a hammer, the invention of the hammer has radically altered the character of the structures we build (e.g., shelters built without the aid of a hammer tend to be less angular). Thus the invention of a tool and its use by members doesn't simply facilitate forms of action that would occur anyway; the tool changes the form, structure, and character of the activity.

If this is true for hammers, consider how substantial is the influence of more modem technological tools like automobiles, computers, video, etc. (see 20.4 for other examples). The word processor on which we write hasn't merely helped us to be more efficient in our professional writing as we did it 25 years ago. The nature of that writing process has changed radically. Culture creates the tool, but the tool changes the culture. Participants in the culture appropriate these tools from their culture to meet their goals and thereby transform their participation in the culture.

The computer is a good example of a mediational means that has aspects of both tool and sign. During the time and place where Vygotsky was writing, tools were used almost exclusively for physical labor, to manipulate physical objects in the environment. Signs, on the other hand, are mediational means used for cognitive functioning, and certainly word processors influence the writer as well as the written product. Language, of course, was the semiotic means about which Vygotsky wrote the most, but he also included numbers, algebraic notation, mnemonic techniques (his famous knot in a string to remember something), diagrams, maps, musical notation, etc. In fact these means are very reminiscent of the multiple intelligence proposed by Howard Gardner (e.g., 1993): linguistic, musical, logical?mathematical, spatial, bodily-kinesthetic, interpersonal and intrapersonal.All these are not simply alternative means of expressing some underlying meaning but rather semiautonomous systems for constructing meaning. They too have been invented by culture to address some need of the culture, but by their use actively transform the culture. It is the action produced by these mediational means that is crucial. Thus, humans "play an active role in using and transforming cultural tools and their associated meaning system." (Wertsch, 1994, p. 204). Wertsch goes on to argue that the "essence of mediated action is that it involves a kind of tension between the mediational means as provided in the sociocultural setting, and the unique contextualized use of these means in carrying out particular, concrete actions" (Wertsch, 1994, p. 205). In other words, all distinctly human instances of learning are constructions situated within a context that employs some form of mediational means, tools, and/or signs.

7.3.5 Learning Is an Inherently Social-Dialogical Activity

This assumption is actually a part of the previous one, but since Vygotsky and his followers have emphasized language as a mediational means above all others, we felt it warranted separate treatment. Certainly the central position of language and dialogue in human culture and cognition can hardly be overemphasized.

In many educational applications commonly characterized as constructivist (e.g., reciprocal teaching, problem-based learning, collaborative groups, etc.), one finds a strong emphasis on dyadic or group discussion: talk, talk, talk! Even in applications like hypermedia systems that are intended for single users, the interface often models a dialogic structure, as in querying a database to solve a problem (Knuth, 1992), or actually includes means for synchronous or asynchronous dialogue among users (Duffy, 1995; Duffy & Knuth, 1990). Why this emphasis on dialogue?

A child is borin into a sociocultural milieu that functions on the basis of some socially organized processes: operations, objects, and structures. As the child acts in this context, she is exposed to these means by which the community mediates its activities. Caregivers use language to interact with the child and intuitively coordinate these linguistic actions with the child's behavior. The child then appropriates this language tool to further influence and control her social interactions, but by adopting this mediational sign has transformed her ability to influence her own actions within her developing spheres of action. According to Wertsch: "The incorporation of mediational means does not simply facilitate action that could have occurred without them; instead, as Vygotsky (1981) noted, 'by being included in the process of behavior, the psychological tool alters the entire flow and structure of mental functions"' (1991, p. 137).

A primary way in which mental functions are altered by the mediation of language signs is that knowledge, and thereby learning, becomes a social, communicative, and discursive process, inexorably grounded in talk. James Wertsch (1991) has been particularly influential in arguing this case by presenting the views of a contemporary of Vygotsky, M. M. Bakhtin. Bakhtin focused his analysis on the utterance, or the shared activity of speech communication--that is, voice. In other words, he stresses the social functions of the linguistic sign, its use as a mediational means to express and share meanings within a social language community. Bakhtin has coined a wonderful term, ventriloquation, which is the process by means of which one individual (or voice) speaks through the voices or the language of a social community. In a very real sense, the way in which a student comes to manifest the effective behavior of a community (e.g., the community of scientists) is to speak with the voice of that community (e.g., to talk like a scientist). Paulo Freire's (1993) work has also stressed the importance of voice and dialogue as a means for action within a sociocultural context.

7.3.6 Learners Are Distributed, Multidimensional Participants in a Sociocultural Process

Perhaps the most "revolutionary" aspect of the MAR metaphor is the concept of a distributed mind and its corollary, a distributed self Displacing the individual from the central position in cognitive action is, we suspect, a shift on a par with displacing the earth from the center of the universe. And yet more and more books and articles with titles like Socially Shared Cognition (Resnick, Levine & Teasley, 199 1), Distributed Cognitions (Salomon, 1993), and Distributed Decision Making (Rasmussen, Brehmer & Leplat, 199 1) are appearing. Lave and Wenger (1991) use the term whole person to characterize this conception of self, where learning is not a matter of a person's internalizing knowledge but a matter of a person's transforming his participation in a social community. The whole person defines as well as is defined by this participation. Lave and Wenger (199 1, p. 53) describe identities as "long-term, living relations between persons and their place and participation in communities of practice."

Hutchins (1991) proposes a simple thought experiment to illustrate this idea. Look around where you are right now reading this and try to find something that "was not either produced or delivered to its present location by the cooperative efforts of humans working in socially organized groups" (p. 284). Unless your environment is strikingly different from ours, we think you will have difficulty identifying anything. Of course, your inclination is to declare those objects as different from you, as something other than self, but are they not really part and parcel of the means by which you participate in the communities that produced them? Isn't that your identity?

We won't belabor the point, but it should be clear that a distributed concept of self shifts the activity of learning to the connections one has with communities, to the patterns of participation, and away from efficient internalization of knowledge. Here then is another reason why so many constructivist applications employ discussion and dialogue in groups. Learning is not the lonely act of an individual, even when it is undertaken alone. It is a matter of being initiated into the practices of a community, of moving from legitimate peripheral participation to centripetal participation in the actions of a learning community (Lave & Wenger, 1991).

In anticipation of a common criticism, we would like to stress that the notion of distributed self does not remove self-agency from the learning process. It is sometimes argued that in models like this, the needs of the individual are sacrificed to the demands of the community. We admit that this is a danger, but no more so than the danger of indoctrination inherent in the process of internalization of knowledge transmitted from teacher or lesson to the learner. The important element missing in our model thus far, that of reflexivity, will reinforce the importance of self-agency (see 7.3.7).

7.3.7 Knowing How We Know Is the Ultimate Human Accomplishment

This last principle is the most important and probably the least controversial. We can't think of a single model of the teaching/learning process that would not stress the importance of self-awareness of learning and knowing. Certainly the extensive literature on metacognition (e.g., Flavell, 1979), thinking skills (e.g., Baron & Sternberg, 1987), theory of mind (e.g., Wellman, 1990), etc., within cognitive psychology are all pointed to the development of self-monitoring and self-control of the learning process. Even Skinner (1968, pp. 172-73) encourages learners to analyze the contingencies (see 2.2.1,3.1) that control their behavior and deliberately manipulate them so as to become self-reliant and self-managing (he even uses the word freedom to characterize this process!).

Where we differ, of course, is our account of the teaching/learning processes of which one should be aware! Many models of metacognition stress the development of strategies of efficient processing: primarily storage and retrieval (e.g., Bornstein, 1979). Programs in thinking skills frequently focus on the problem-solving process and train students to use systematic analytical procedures like Bransford and Stein's (1984) IDEAL problem solver or Sternberg's (1987) metacomponents. While we are unaware of any particular applications derived specifically from the MAB metaphor, we suspect that they would emphasize the process of perceptual tuning, perhaps in the sense of Donald Schon's (1987) reflective practitioner developing the ability to "see as."

We prefer the term reflexivity, which means directed, or turned back upon itself, or self-referential. To be reflexive about the principles cited above is to direct them back on your own efforts to learn, teach, and know. As specified in the principles above, we regard all learning as a social, dialogical process of construction by distributed, multidimensional selves using tools and signs within contexts created by the various communities with which they interact. This, we believe, is an entirely natural process of which we are ordinarily no more aware than we are of breathing or of our heartbeat. Our process of construction is directed toward creating a world that makes sense to us, one that is adequate for our everyday functioning. We are generally unaware of the beliefs we have adopted or created to live and teach by, but raising them to awareness can have salutary effects. Umberto Eco put the matter this way:

To speak about "speaking," to signify signification, or to communicate about communication cannot but influence the universe of speaking, signifying, and communicating (1976, p. 29).

How do beliefs change? How do we become aware of them? When we are confronted with some experience not accounted for by our existing beliefs, we invent a new set of beliefs or revise an existing one, a process we have elsewhere referred to as abduction (Cunningham, 1992). This new structure will provide a context within which the surprising experience is a matter of course (i.e., it makes sense). Abduction is instigated when we are in a condition of inadequacy or uncertainty that arises from experience; hence it is naturally embedded in a relevant context (is situated or anchored). Thus when we experience or are shown a situation where our existing beliefs are inadequate, our awareness of our own state of knowing is enhanced. This is the essence of reflexivity.

Further awareness of the cultural origin and mediated nature of our beliefs allows us to explore varieties of belief structures. A reflexive analysis of the metaphors by which we live and teach will allow us to reconsider them. If we are Dot satisfied with the metaphor of "school is work," let's try another: "school as consulting service." Under this model, the school might be seen as a community resource where the teachers, students, equipment, and facilities are placed at the service of members of the community who may bring problems and issues to be addressed. Teachers and more experienced students mentor the younger ones during problem-solving projects geared toward the betterment of the community of which all are a part. But if this metaphor proves unuseful, try another!

Finally, we believe that via reflexivity, and in a manner not possible in other models, learners have real control over and responsibility for their beliefs. An awareness of the principles of constructivism listed above demands a strong sense of responsibility for the state of the world in which we find ourselves. If many world views are possible, then our choice of participation in the community that holds a particular view requires both a commitment to and a responsibility to respect the views of others. We have within our capability the constant renewal of our world view. Human reflection is the key to understanding and creating anew a world in which we coexist with others. Someone else's world view, her belief structure, can be as legitimate as our own. To coexist, a broader perspective is necessary, one in which both parties cooperate to bring forth a common world where many perspectives are valid.