The Culture of Education

Jerome Bruner

• Publisher: Harvard University Press
• Year: 1997
• ISBN:
  9780674251069
• Topic: Education
• Subtopic: Multicultural Education

What we don't know about learning could fill a book--and it might be a schoolbook. In a masterly commentary on the possibilities of education, the eminent psychologist Jerome Bruner reveals how education can usher children into their culture, though it often fails to do so. Applying the newly emerging "cultural psychology" to education, Bruner proposes that the mind reaches its full potential only through participation in the culture--not just its more formal arts and sciences, but its ways of perceiving, thinking, feeling, and carrying out discourse. By examining both educational practice and educational theory, Bruner explores new and rich ways of approaching many of the classical problems that perplex educators.Education, Bruner reminds us, cannot be reduced to mere information processing, sorting knowledge into categories. Its objective is to help learners construct meanings, not simply to manage information. Meaning making requires an understanding of the ways of one's culture--whether the subject in question is social studies, literature, or science. The Culture of Education makes a forceful case for the importance of narrative as an instrument of meaning making. An embodiment of culture, narrative permits us to understand the present, the past, and the humanly possible in a uniquely human way.Going well beyond his earlier acclaimed books on education, Bruner looks past the issue of achieving individual competence to the question of how education equips individuals to participate in the culture on which life and livelihood depend. Educators, psychologists, and students of mind and culture will find in this volume an unsettling criticism that challenges our current conventional practices--as well as a wise vision that charts a direction for the future.

1

CULTURE, MIND, AND EDUCATION

The essays in this volume are all products of the 1990s, expressions of the fundamental changes that have been altering conceptions about the nature of the human mind in the decades since the cognitive revolution. These changes, it now seems clear in retrospect, grew out of two strikingly divergent conceptions about how mind works. The first of these was the hypothesis that mind could be conceived as a computational device. This was not a new idea, but it had been powerfully reconceived in the newly advanced computational sciences. The other was the proposal that mind is both constituted by and realized in the use of human culture. The two views led to very different conceptions of the nature of mind itself, and of how mind should be cultivated. Each led its adherents to follow distinctively different strategies of inquiry about how mind functions and about how it might be improved through “education.”

The first or computational view is concerned with information processing: how finite, coded, unambiguous information about the world is inscribed, sorted, stored, collated, retrieved, and generally managed by a computational device. It takes information as its given, as something already settled in relation to some preexisting, rule-bound code that maps onto states of the world.¹ This so-called “well-formedness” is both its strength and its shortcoming, as we shall see. For the process of knowing is often messier, more fraught with ambiguity than such a view allows.

Computational science makes interesting general claims about the conduct of education,² though it is still unclear what specific lessons it has to teach the educator. There is a widespread and not unreasonable belief that we should be able to discover something about how to teach human beings more effectively from knowing how to program computers effectively. One can scarcely doubt, for example, that computers provide a learner with powerful aids in mastering bodies of knowledge, particularly if the knowledge in question is well defined. A well-programmed computer is especially useful for taking over tasks that, at last, can be declared “unfit for human production.” For computers are faster, more orderly, less fitful in remembering, and do not get bored. And of course, it is revealing of our own minds and our human situation to ask what things we do better or worse than our servant computer.

It is considerably more uncertain whether, in any deep sense, the tasks of a teacher can be “handed over” to a computer, even the most “responsive” one that can be theoretically envisioned. Which is not to say that a suitably programmed computer cannot lighten a teacher’s load by taking over some of the routines that clutter the process of instruction. But that is not the issue. After all, books came to serve such a function after Gutenberg’s discovery made them widely available.³

The issue, rather, is whether the computational view of mind itself offers an adequate enough view about how mind works to guide our efforts in trying to “educate” it. It is a subtle question. For in certain respects, “how the mind works” is itself dependent on the tools at its disposal. “How the hand works,” for example, cannot be fully appreciated unless one also takes into account whether it is equipped with a screwdriver, a pair of scissors, or a laser-beam gun. And by the same token, the systematic historian’s “mind” works differently from the mind of the classic “teller of tales” with his stock of combinable myth-like modules. So, in a sense, the mere existence of computational devices (and a theory of computation about their mode of operating) can (and doubtless will) change our minds about how “mind” works, just as the book did.⁴

This brings us directly to the second approach to the nature of mind—call it culturalism. It takes its inspiration from the evolutionary fact that mind could not exist save for culture. For the evolution of the hominid mind is linked to the development of a way of life where “reality” is represented by a symbolism shared by members of a cultural community in which a technical-social way of life is both organized and construed in terms of that symbolism. This symbolic mode is not only shared by a community, but conserved, elaborated, and passed on to succeeding generations who, by virtue of this transmission, continue to maintain the culture’s identity and way of life.

Culture in this sense is superorganic.⁵ But it shapes the minds of individuals as well. Its individual expression inheres in meaning making, assigning meanings to things in different settings on particular occasions. Meaning making involves situating encounters with the world in their appropriate cultural contexts in order to know “what they are about.” Although meanings are “in the mind,” they have their origins and their significance in the culture in which they are created. It is this cultural situatedness of meanings that assures their negotiability and, ultimately, their communicability. Whether “private meanings” exist is not the point; what is important is that meanings provide a basis for cultural exchange. On this view, knowing and communicating are in their nature highly interdependent, indeed virtually inseparable. For however much the individual may seem to operate on his or her own in carrying out the quest for meanings, nobody can do it unaided by the culture’s symbolic systems. It is culture that provides the tools for organizing and understanding our worlds in communicable ways. The distinctive feature of human evolution is that mind evolved in a fashion that enables human beings to utilize the tools of culture. Without those tools, whether symbolic or material, man is not a “naked ape” but an empty abstraction.

Culture, then, though itself man-made, both forms and makes possible the workings of a distinctively human mind. On this view, learning and thinking are always situated in a cultural setting and always dependent upon the utilization of cultural resources.⁶ Even individual variation in the nature and use of mind can be attributed to the varied opportunities that different cultural settings provide, though these are not the only source of variation in mental functioning.

Like its computational cousin, culturalism seeks to bring together insights from psychology, anthropology, linguistics, and the human sciences generally, in order to reformulate a model of mind. But the two do so for radically different purposes. Computationalism, to its great credit, is interested in any and all ways in which information is organized and used—information in the well-formed and finite sense mentioned earlier, regardless of the guise in which information processing is realized. In this broad sense, it recognizes no disciplinary boundaries, not even the boundary between human and non-human functioning. Culturalism, on the other hand, concentrates exclusively on how human beings in cultural communities create and transform meanings.

I want to set forth in this opening chapter some principal motifs of the cultural approach and explore how these relate to education. But before turning to that formidable task, I need first to dispel the shibboleth of a necessary contradiction between culturalism and computationalism. For I think the apparent contradiction is based on a misunderstanding, one that leads to gross and needless over-dramatization. Obviously the approaches are very different, and their ideological overspill may indeed overwhelm us if we do not take care to distinguish them clearly. For it surely matters ideologically what kind of “model” of the human mind one embraces.⁷ Indeed, the model of mind to which one adheres even shapes the “folk pedagogy” of schoolroom practice, as we shall see in the following chapter. Mind as equated to the power of association and habit formation privileges “drill” as the true pedagogy, while mind taken as the capacity for reflection and discourse on the nature of necessary truths favors the Socratic dialogue. And each of these is linked to our conception of the ideal society and the ideal citizen.

Yet in fact, neither computationalism nor culturalism is so linked to particular models of mind as to be shackled in particular pedagogies. Their difference is of quite a different kind. Let me try to sketch it.

The objective of computationalism is to devise a formal redescription of any and all functioning systems that manage the flow of well-formed information. It seeks to do so in a way that produces foreseeable, systematic outcomes. One such system is the human mind. But thoughtful computationalism does not propose that mind is like some particular “computer” that needs to be “programmed” in a particular way in order to operate systematically or “efficiently.” What it argues, rather, is that any and all systems that process information must be governed by specifiable “rules” or procedures that govern what to do with inputs. It matters not whether it is a nervous system or the genetic apparatus that takes instruction from DNA and then reproduces later generations, or whatever. This is the ideal of Artificial Intelligence, so-called. “Real minds” are describable in terms of the same AI generalization—systems governed by specifiable rules for managing the flow of coded information.

But, as already noted, the rules common to all information systems do not cover the messy, ambiguous, and context-sensitive processes of meaning making, a form of activity in which the construction of highly “fuzzy” and metaphoric category systems is just as notable as the use of specifiable categories for sorting inputs in a way to yield comprehensible outputs. Some computationalists, convinced a priori that even meaning making can be reduced to AI specifications, are perpetually at work trying to prove that the messiness of meaning making is not beyond their reach.⁸ The complex “universal models” they propose are sometimes half-jokingly referred to by them as “TOEs,” an acronym for “theories of everything.”⁹ But though they have not even come near to succeeding and, as many believe, will probably never in principle succeed, their efforts nonetheless are interesting for the light they shed on the divide between meaning making and information processing.

The difficulty these computationalists encounter inheres in the kinds of “rules” or operations that are possible in computation. All of them, as we know, must be specifiable in advance, must be free of ambiguity, and so on. They must, in their ensemble, also be computationally consistent, which means that while operations may alter with feedback from prior results, the alterations must also adhere to a consistent, prearranged systematicity. Computational rules may be contingent, but they cannot encompass unforeseeable contingencies. Thus Hamlet cannot (in AI) tease Polonius with ambiguous banter about “yonder cloud shaped like a camel, nay ’tis backed like a weasel,” in the hope that his banter might evoke guilt and some telltale knowledge about the death of Hamlet’s father.

It is precisely this clarity, this prefixedness of categories that imposes the most severe limit on computationalism as a medium in which to frame a model of mind. But once this limitation is recognized, the alleged death struggle between culturalism and computationalism evaporates. For the meaning making of the culturalist, unlike the information processing of the computationalist, is in principle interpretive, fraught with ambiguity, sensitive to the occasion, and often after the fact. Its “ill-formed procedures” are like “maxims” rather than like fully specifiable rules.¹⁰ But they are hardly unprincipled. Rather, they are the stuff of hermeneutics, an intellectual pursuit no less disciplined for its failure to produce the click-clear outputs of a computational exercise. Its model case is text interpretation. In interpreting a text, the meaning of a part depends upon a hypothesis about the meanings of the whole, whose meaning in turn is based upon one’s judgment of meanings of the parts that compose it. But, as we shall have many occasions to see in the following chapters, a wide swath of the human cultural enterprise depends upon it. Nor is it clear that the infamous “hermeneutic circle” deserves the knocks it gets from those in search of clarity and certainty. After all, it lies at the heart of meaning making.

Hermeneutic meaning making and well-formed information processing are incommensurate. Their incommensurability can be made evident even in a simple example. Any input to a computational system must, of course, be encoded in a specifiable way that leaves no room for ambiguity. What happens, then, if (as in human meaning making) an input needs to be encoded according to the context in which it is encountered? Let me give a homely example involving language, since so much of meaning making involves language. Say the input into the system is the word cloud. Shall it be taken in its “meteorological” sense, its “mental condition” sense, or in some other way? Now, it is easy (indeed necessary) to provide a computational device with a “look-up” lexicon that provides alternative senses of cloud. Any dictionary can do it. But to determine which sense is appropriate for a particular context, the computational device would also need a way of encoding and interpreting all contexts in which the word cloud might appear. That would then require the computer to have a look-up list for all possible contexts, a “contexticon.” But while there are a finite number of words, there are an infinite number of contexts in which particular words might appear. Encoding the context of Hamlet’s little riddle about “yonder cloud” would almost certainly escape the powers of the best “contexticon” one could imagine!

There is no decision procedure known that could resolve the question whether the incommensurability between culturalism’s meaning making and computationalism’s information processing could ever be overcome. Yet, for all that, the two have a kinship that is difficult to ignore. For once meanings are established, it is their formalization into a well-formed category system that can be managed by computational rules. Obviously one loses the subtlety of context dependency and metaphor in doing so: clouds would have to pass tests of truth functionality to get into the play. But then again, “formalization” in science consists of just such maneuvers: treating an array of formalized and operationalized meanings “as if” they were fit for computation. Eventually we come to believe that scientific terms actually were born and grew that way: decontextualized, disambiguated, totally “lookuppable.”

There is equally puzzling commerce in the other direction. For we are often forced to interpret the output of a computation in order to “make some sense” of it—that is, to figure out what it “means.” This “search for the meaning” of final outputs has always been customary in statistical procedures such as factor analysis where the association between different “variables,” discovered by statistical manipulation, needed to be interpreted hermeneutically in order to “make sense.” The same problem is encountered when investigators use the computational option of parallel processing to discover the association between a set of coded inputs. The final output of such parallel processing similarly needs interpretation to be rendered meaningful. So there is plainly some complementary relationship between what the computationalist is trying to explain and what the culturalist is trying to interpret, a relationship that has long puzzled students of epistemology.¹¹

I shall return to this puzzling problem in Chapter 5. For now it suffices to say that in an undertaking as inherently reflexive and complicated as characterizing “how our minds work” or how they might be made to work better, there is surely room for two perspectives on the nature of knowing.¹² Nor is there any demonstrable reason to suppose that without a single and legitimately “true” way of knowing the world, we could ...