Constructivism: Moving Beyond the Impasse

Nicholas C. Burbules

Department of Educational Policy Studies

University of Illinois, Urbana/Champaign

Published in Constructivism in Education, D.C.Phillips, ed. (University of Chicago Press, 2000)

In this essay I offer a re-examination of the philosophical and pedagogical merits of this thing (these things) called "constructivism," in light of the discussions and criticisms offered by the authors in this collection. As they make quite clear, "constructivism" refers to many ideas, joined by the merest thread of family resemblance and often expressing quite contradictory views. Whenever a philosopher encounters a term used so variously, especially when it conjures powerful allegiances and opponents under one or another of its guises, one’s conceptual warning lights start flashing. As a first question, an outsider to this dispute might observe, Where is all this passion coming from? What is at stake that so energizes – and polarizes – these debates? The very vehemence of these discussions frequently suggests that something deeper must be at stake than the relative merits of Locke versus Piaget, or Socrates versus Vygotsky.

To me this seems one of those situations in which the apparent terms of an argument stand in for what people are actually disagreeing about. The terrain of "constructivism," partly because of the multiplicity of meanings people attach to the concept, has become a battleground for people to argue over larger issues like the objectivity of science, the perils of relativism, the spread of "postmodernist" ideas (another one of those multifarious terms), the transition from positivist to postpositivist philosophies of science, the emptiness of identity politics (in which race, gender, and so on are regarded as "mere constructions" – see Hacking 1999), and so on. People argue over constructivism at least in part because they see within it the potential, or the danger, of promulgating other ideas to which it becomes attached.

In response, I hope to suggest a few ways of moving beyond the impasse between the pro- and anti-constructivists. First, I want to point out some of the ways in which the characterizations of what constructivism is, pro and con, contribute to misunderstandings and unfruitful dichotomies. Second, I will try to reframe the epistemological and metaphysical disputes that advocates and critics seem to think divide them, when in fact they are often differing only in the relative weight they give to certain factors that most of them agree must be given some weight in assessments of what is true. This section will conclude with a set of five propositions that I believe most participants to this debate, whether pro- or anti-constructivism, can subscribe to. Third, I want to move past the philosophical disputes – which I am suggesting are often exaggerated and self-perpetuating – to the pedagogical merits of constructivism as a view of learning and as a view of teaching (two significantly different things), and conclude by building upon some of the important pedagogical conclusions offered by other essays in this collection.

Misframing the debate

Phillips (1997, p. 9) notes that the term has "attained the status of political correctness," particularly in the field of education. When broad (and multifarious) ideas do become widely accepted, there is a tremendous incentive to attach one’s theoretical (and often political) agendas to that term. In seeking allies, one is halfway toward success because people already feel a passionate (if vague) commitment to the broader notion; and if one can assert persuasively that acceptance of that idea also commits one to other conclusions, many people will not scrutinize the logic that carries them from one thing to the other. Constructivism is hardly the only concept that has received such treatment, particularly in education: a similar analysis could be given of "choice," "democratic classrooms," "equal opportunity," "progressive education," and similar terms. As a result, fights over what "constructivism" really means (and commits one to) take on the vigor of a crusade, because other important allegiances are at stake.

In the same way, opponents of what "constructivism" has come to stand for often attribute to it assumptions that mischaracterize or oversimplify what the view commits one to (although it must be said that they are aided in this by the exaggerated claims made by some of its proponents, as I will discuss later). So we encounter characterizations like that of Hull, quoted by Phillips here: "The most extreme constructivists seem to hold that all of us, scientists included, are helpless victims in the maws of our societies." Well, who would want to believe that? It sounds like being devoured by a giant creature in a monster movie.

And here I want to suggest a cautionary tip to readers: always watch for the word "mere" (or similar terms). Often constructivists are excoriated for holding that knowledge is "merely constructed," or that personal identity is a "mere social construction," and so on. Notice that these criticisms actually do acknowledge, tacitly, that such matters are in some sense constructed, but not merely so. Recognizing how "mere" operates to exaggerate and oversimplify the opposing view allows one to shift the issue to different, and I think more productive questions: What elements of such beliefs are subject to construction, and which ones are not? If knowledge (or identity, etc.) isn’t merely constructed, how constructed is it? For example, there is certainly a sense in which categories of racial identity are a construction, and not merely a consequence of skin color; yet it is also true that skin color is a relatively objective factor in this process. The idea of a "quark" is a construction, since we can never observe such things; yet it bears upon evidence that can be observed and given a fairly widespread consensual interpretation. The appellation "mere" gives the impression that constructivists think there is nothing to say beyond affirming whatever individuals or social groups – any individual or social group – might happen to come up with as an account of their social or natural environments. But only some constructivists say such things, and criticisms of such views are better leveled against those writers than against "constructivism" per se, it seems to me. Some authors in this collection are careful to make such distinctions; others are less so. Indeed, as amply demonstrated in these essays, it is often other constructivists who argue most strongly against such extreme views.

The fight, by some advocates and critics alike, to identify the constructivism clearly flies in the face of the multiplicity of views actually associated with the term. I am a big fan of studies that try to understand and explore this sort of complexity, and do justice to it, rather than accepting a polarized characterization that, for some advocates, suggests that if one does not accept this particular version, one is not a "real" constructivist; or, for some critics, sets up a particularly outlandish set of beliefs to argue the dangers of the view generally (surprisingly, the work of von Glasersfeld seems to serve both sorts of purposes simultaneously).

Another aspect of the debate that has ill-served a thoughtful consideration of these complexities is, as Matthews says here, that "constructivism" is often treated as a package deal, by advocates and critics alike, where commitment to a certain approach to learning automatically commits one to a theory of teaching, a view of epistemology, a conception of reality, and so on. This can be seen as an expression of the phenomenon described earlier, that subscribing to certain "isms" is taken as an all-or-nothing affair. Yet, ironically, this is a very un-constructivist thing to claim. If one believes that theoretical understandings are to some degree constructed, then a corollary is that these constructions will vary, and that different people or groups will account for matters in different ways – and this must pertain to alternative versions of constructivist theory itself. The attempt to offer a systematic account of constructivism, and what certain beliefs necessarily commit one to, seems not only inconsistent, but suggests the desire to impose a set of views rather than to engage the very processes of construction argued for within the theory. To me this is rather akin to arguments for relativism, or pronouncements about the need for everyone to believe postmodernism: where do such arguments stand in relation to their own claims?

In addition, as Popkewitz (1998) shows, constructivism is a theory that has its own genealogy. When viewed as a historically unfolding and changing theory, constructivism appears as much less obviously wedded to the liberating, open-ended forms of pedagogy that its adherents tend to support. Without reviewing all of Popkewitz’s argument here, the point of his analysis is to suggest that a constructivist understanding of constructivism itself would have to regard its premises, and its implications, much less as ahistorical, essential logical connectives that stand or fall together, and more as socially mediated associations that are formulated, expressed, and accepted for various reasons by situated persons and groups, and that have identifiable consequences for larger patterns of social inclusion and exclusion.

At the start, then, I am suggesting that the debate over constructivism is often mischaracterized by both sides. On the one hand, the discussion is not advanced by those critics who use exaggerated characterizations of what constructivism is to avoid having to acknowledge the rather inescapable conclusion that all human knowledge is (in some sense at least) constructed by sentient actors situated in concrete circumstances of institutional identification, social relations of power, and personal interest. All the interesting questions follow from this acknowledgment: What are the constraints that limit just any belief from taking hold as true? What are the correctives built into social organizations of inquiry that question, test, and modify such beliefs? What are the conditions that allow for the comparison of beliefs, as a way of assessing alternative accounts of phenomena? My point, which I will develop below, is that there is no need to step outside a constructivist framework to give answers to such questions.

On the other hand, I have suggested, the discussion is also not advanced by advocates of constructivism who mischaracterize a broad range of opposing epistemological stances as all holding a naive "correspondence" view of truth, when the entire trend of postpositivist philosophies of science has been to concede a number of the specific claims of the constructivist account (for example, that one’s theoretical commitments influence the ways in which phenomena are perceived), while denying that this leads to the more extreme relativist conclusions that some constructivists still seem to maintain. I have tried to point out here how utterly inconsistent such a relativistic stance is for anyone who argues that constructivism just is the way people learn, or is the way societies create and negotiate identities, or is the way in which "truth" (their quotation marks) gets made.

To commentators on both sides I want to say, "You aren’t being constructivist enough."

Epistemological and metaphysical antinomies

The epistemological debate between constructivism and its critics seems to be primarily about whether there can be a warrant for knowledge that rests upon criteria that are not socially and culturally specific. The metaphysical debate between constructivism and its critics seems to be primarily about whether there is a reality that is external to our attempts to grasp it.

Both of these debates are, I want to suggest, misguided and – as framed here – irresolvable. They are instances of the problems that arise when the way in which an issue has been defined guarantees that the disputants will occupy positions that reinforce their stereotypes of one another and strengthen their resolve that the other view represents a narrow-minded and dangerous notion that must be resisted at all times.

On the issue of epistemology, constructivism is typically characterized as the view, as Howe and Berv put it here, that there are "no criteria [of truth] outside of what people say and do," or as McCarty and Schwandt put it, "no meanings or standards outside or alternative to those of the community to which community members can appeal." When the issue is characterized this way, constructivists come off (to their critics) as relativists. At the same time, the critics come off (to constructivists) as seeming to think that there are free-floating "criteria" out there that can always settle knowledge disputes in an objective, impartial manner; the constructivists wave their well-worn copies of Thomas Kuhn’s Structure of Scientific Revolutions, and both sides of the dispute settle down into another round of that old familiar waltz around the issue of paradigms and incommensurability.

Please, let’s move on.

The self-perpetuating character of this debate, in which the roles and arguments are so well known and rehearsed, reminds me of that story where the jokes told in a prison become so familiar that they are just identified by number ("14! the inmate yelled, and all the others laughed"). And as often happens when advocates "know" what their opponents are going to say, real listening and engagement cease to be possibilities.

I would like to suggest a different starting point, and hence a different way of framing the question. It seems quite apparent to me that all endeavors of human thought and language are socially situated. That means that any criteria we could ever have will be socially formulated, negotiated, interpreted, and applied. Any assessments of warrant will be socially formulated, negotiated, interpreted, and applied. Any methods of inquiry will be socially formulated, negotiated, interpreted, and applied. Any description of an external world will be socially formulated, negotiated, interpreted, and applied. Thus, any knowledge claim will be socially formulated, negotiated, interpreted, and applied. For that matter, any account of epistemology will be socially formulated, negotiated, interpreted, and applied.

The new question I wish to pose is, "socially formulated, negotiated, interpreted, and applied by whom?" because the actual dispute that seems to be operating at the ground level here is that one group disagrees with the criteria, the assessments of warrant, the methods of inquiry, the descriptions of an external world, and so on, of another group (from here on, I will refer to this cluster of interrelated ideas as "criteria, etc."). The critics of constructivists worry that they are insufficiently respectful of a set of criteria, etc., that they (the critics) value; they see in the apparent relativism of constructivism an assault, not on the possibility of criteria, etc., generally (although this is how the criticism is typically stated), but on their criteria, etc. What most bothers the critics of constructivism, I think, is that constructivism seems to deprive them of the basis for claiming that their criteria are not just the ones that happen to be recognized by their own social group, but ones that should be held by everyone.

But this is where my reformulation is meant to be helpful, for the critics need to acknowledge that their criteria, etc., are the ones that happen to be held by their social group – it could not be otherwise – but that this does not disqualify them from also having a value and weight that could be recommended to others. It simply deprives them of the argument that the reason why they have such a value and weight is that they are the real or true criteria, etc., which must be binding on all groups whether they realize it or not. Instead, they have to do what they always had to do anyway, which is to seek to persuade others that their criteria, etc., are better accounts. If they succeed in this, everyone is happy. If they fail in this, then the only purpose served by the claim that these really are the proper criteria, etc., and that others are just foolishly (or stubbornly) failing to recognize them, is to grant legitimacy to social policies and institutions that impose such criteria, etc., on persons and groups who for whatever reason are resisting them. Reframing the matter in social terms, as this is, reveals how disturbing an option such an outcome would be.

Far better, then, to explore with greater care the dynamics by which criteria, etc., come to be shared across different persons and groups. Constructivism does not have to deny the possibility of their being shared, or discount its desirability. Conversely, critics of constructivism can be asked to ante up their confidence that the criteria, etc., which they favor really are the better, more rational, more fruitful ones. If my argument is successful at all, that little word "outside" in the Howe and Berv and McCarty and Schwandt quotes, above, has been transformed: there are never any criteria, etc., outside the holdings of a particular group unless and until another group comes to hold them also. If the critics of constructivism believe that there "really are" such criteria, etc., outside of their particular shared beliefs, the only way to instantiate that claim is by persuading others to come to hold them also. No part of this analysis steps outside the social (constructed) realm. The only alternative here is the rather widely discredited philosophical notion that these criteria, etc., drift abstractly, timelessly, above and apart from human affairs (whether we have discovered them yet or not) and that we need philosophers and other wise men and women to glimpse them and explain them to the rest of us. This is "outside" in the sense that Plato’s figure had to go outside the Cave and then report back to the poor devils inside, arguing about "mere" shadows on the wall; and we all know that he was totally unsuccessful in changing anyone’s view of things there.

When Slezak argues, here, that constructivists believe that "knowledge is merely consensus upon arbitrary convention" (italics added), we see a fairly typical example of how the issues are being misframed. A set of conventions (criteria, etc.) that come to be socially shared are in no way arbitrary, even if one thinks they are mistaken; for one thing, not just any set of conventions can come to be commonly shared. Unless a social group is totally delusional, their criteria, etc., must have something to be said for them, must have some value and efficacy, it seems to me. Challenging or changing them will require something more than calling them "arbitrary." For another thing, we encounter here that word again, "merely," as if social consensus were not inherently part of the processes by which any system of belief needs to be formulated, negotiated, interpreted, and applied. And if any constructivists are so foolish as to assert that "knowledge is merely consensus upon arbitrary convention," then one need only reply that this position then gives no reason whatsoever to accept the constructivist account itself.

So, in the first instance, I have tried to change an apparent epistemological dispute about objectivism and relativism into a discussion of which social groups accept which criteria, assessments of warrant, methods of inquiry, etc., and how they might come to be shared across different groups. I have suggested that the only meaning of "outside" that does any worthwhile philosophical work here is one grounded in the idea of criteria, etc., that come to be shared through a process of engagement and persuasion; that if this process is in good faith there cannot be any presumption about whose criteria, etc., are destined to win over the other; that assertions that some are the "real and true" criteria, etc., which others are compelled to accept, is a rhetorical move, not a persuasive one; and that any other meaning of "outside" tends to impede such conversations by prejudging what the conclusion of the engagement must be, or imposing that outcome by philosophical fiat.

On the issue of metaphysics, we come to the timeless philosophical question about the independent existence of an external world. I have given long and careful thought to this problem, and I would like to propose a seriously considered response: Who cares?

The question of the existence of an external world is, I believe, one of those philosophical pseudo-questions that is not answerable in the terms given. It is a more sophisticated version of that sophomoric conundrum about trees falling in forests. It expresses an article of faith rather than a demonstrable philosophical postulate, and as with other articles of faith it is not interesting to ask whether it is true or not, but only why people believe it and what effects their belief has.

Constructivists err profoundly when they take on the metaphysical issue. As McCarty and Schwandt point out, both Gergen and von Glasersfeld claim to be "agnostic" on this question, and yet neither one can apparently resist the temptation to weigh in on the subject. Too bad: agnosticism is the better road for any constructivist, I believe. By denying the existence of an external world, one takes on the same unredeemable burden of proof as those who assert it. But for the constructivist, this question cannot and does not need to be answered either way; the question needs to be reframed as, "What difference would it make to the processes of constructing an understanding based on experience?" The faith in the existence of an independent, unchanging external world has two important effects: it wants to guarantee (a) the consistency of experience and (b) the ultimate decidability of disputes over particular truth claims.

But on the first issue one must simply say, either our experience is consistent or it is not (consistent for the individual, over time, and consistent across individuals). If it is, then why it is need not concern us; we go about the business of constructing a coherent account of things, given what we and others experience, and leave the rest as superfluous information. If it is not, then no coherent account of things is possible anyway, and embracing the postulate of an independent, unchanging external world will not make it any more so. All that we need in order for the pursuit of knowledge to go forward, individually and socially, is consistency of experience: under the same apparent circumstances the same experiences tend to be produced. This is often expressed as "the world is stubborn," for people who seem to think that one needs a referent – "the world" (or "reality") – to stand behind and confirm what they believe to be true. But if the reasons and evidence for that belief are sufficient for it to be called "true," it adds nothing to say, "…and it really is true, too!" or "…it is true because the world really is that way."

Phillips (1997) falls into this error, I think, while discussing the change in belief from the earth being flat, to the belief that it was round. He says, "It is the shape itself, and not only political and social factors at work in discourse communities, that influences the knowledge that we generate about that shape" (p. 189). But the knowledge that the earth "really is round" was an outcome of those changing beliefs. It only makes sense to say that its actual shape drove the process of belief change after that belief change has already happened. The problem is that tomorrow we might figure out that its shape is something else again (even widely held scientific beliefs sometimes turn out to be mistaken). If we say that "the shape itself" explains that new shift in belief, is this the same "shape itself" that explained the previous shift? Or was that previous "shape itself" mistaken (since it was the product of a theory which said that the shape was round)?

This is the problem that arises when people feel the need to supplement, "we believe this is true" with "we believe this is true, and that belief accords with reality." What the supplement comes to is really just a statement of strength of confidence; that a belief or hypothesis was previously based on certain partial experiences, and that later experiences were consistent with that belief and so tended to reinforce it (so that it increases the confidence that "reality" has been captured). In fact, the supplemental phrase cannot mean anything more than this, since anything that one might say about the relation of a belief to reality is either an expression of very strong confidence, based upon repeated, consistent experiences, or it is, as mentioned before, an expression of faith. Either way, it might turn out to be mistaken.

This expression of faith also has relevance to the second issue mentioned above: it tends to support a hope that multiple versions of the world will either converge, or that the best one will win out (because the stubborn world will adjudicate the dispute in favor of one or another). But this hope for decidability is primarily motivational; it does not have any persuasive force itself. Indeed, it can be counterproductive to promoting agreement when it expresses the attitude that "there is a way that the world is; we know it and you do not." Using allusions to "reality" in this way, as a trump, has the effect of blocking inquiries into the elements of social and cultural particularity in any understanding of the world, and of evading the responsibility to acknowledge that possibility and the limits of one’s ability to see the limitations, interests, and biases built into one’s version – limitations that can often only be seen by others who do not share it. If one only views this relation in terms of "we are right and you are wrong," or "this is the way the world is," a crucial opportunity for reciprocal enlightenment is missed.

In short, realists on this issue are more likely to think that knowledge can be objective, complete, and unchanging, and that knowledge disputes can often be settled by allusions to "the way the world is." Constructivists and others who reject that premise are more likely to think that knowledge will be partial, provisional, and imperfect, and that knowledge disputes will be more intractable because strongly held beliefs are intertwined with other social and cultural elements that groups may be reluctant to give up or change. Who is right on this issue? I hope to have made it clear that the question cannot be answered at that level; both positions tend to prejudge what sorts of social agreement will or will not be possible, and they use claims about "reality" (one way or another) to undergird what are really premises about the likelihood of being able to settle certain types of disagreement. Rather than ask who is right, I have suggested that a more fruitful way of framing the matter is to ask what the effects of these competing beliefs might be; and I have suggested that for both the pro- and anti-constructivists, their presuppositions obscure and might actually inhibit the possibilities of pursuing and achieving actual understandings and agreements in the matters under dispute.

As I have said, many constructivists have no one but themselves to blame in mishandling such epistemological and metaphysical questions. Where they should be agnostic or silent, they cannot resist sticking a thumb in the eye of conventional philosophical views. Probably the most egregious example of this habit is the work of Ernst von Glasersfeld, whose more extreme and often philosophically confused claims (as is amply documented by the criticisms posed against him in the essays here) are a kind of anchor on the credibility of constructivism generally. Von Glasersfeld has a persistent habit of "bait and switch" argument; using a term or concept one way in one context, then later using it in a very different way and for a very different purpose. For example, one of his primary analogies (von Glasersfeld, 1984) is that of "match" versus "fit" as two views of knowledge. The first view says that to be true a belief must match or correspond to an independent reality. On the second view, von Glasersfeld’s preferred alternative, a belief must "fit" (like a key in a lock); it must do some work for us. Speculating on the nature of the lock is pointless, he believes; all we can know is whether our "key" (our constructed belief) is successful for our purposes. Matching reality has nothing to do with it.

I do not think that von Glasersfeld realizes that this distinction misses the point he thinks he is establishing. The match versus fit views are not necessarily competing views about "reality"; the fit view posits an independent reality just as much as the match view does. We construct the key, but we do not (as I read von Glasersfeld here) construct whether it fits or not. Indeed, von Glasersfeld himself invokes the term "reality" in explaining this concept of fitness (p. 22). In this context, the match versus fit views are really just two alternative ways of conceiving what the relation between beliefs and reality is. Yet later in this same essay, von Glasersfeld says something very different: that it is the object of understanding that we construct, that "fits" (p. 36). This is a crucial, and highly misleading conflation. That we construct an understanding of the world (that must "fit" reality) is not a particularly radical notion at all; that we construct the objects of our understanding, partly to fit our beliefs, is not only a different view but a diametrically opposed one. It seems that here von Glasersfeld draws readers into accepting a relatively straightforward and commonsensical account of what it means to "fit," then changes his use of it to suggest that this acceptance commits one to other claims that are much more extreme in their epistemological and metaphysical implications.

We see numerous examples of this same tendency in other quotations from von Glasersfeld, many of them cited here. In Phillips (1997, p. 185), von Glasersfeld is quoted as saying that "’facts’ are not elements of an observer independent world but elements of an observer’s experience." This is misleading: a "fact" isn’t an element in either place; it is an assessment of a relation between an experience and the world (a relation like, say, fit). Von Glasersfeld seems to be saying something provocative here, but he is only contesting a superficial and sloppy way in which people commonly use words like "fact," as if facts were scattered around the world, like gemstones, waiting to be collected. He is right: no serious thinker would say such a thing. But rejecting that view does not necessarily throw one over into von Glasersfeld’s preferred alternative, that "facts" are just a certain category of experience. Phillips provides another good illustration of this tendency in this book: in his essay he quotes von Glasersfeld as saying, "knowledge, no matter how it is defined, is in the heads of persons…the thinking subject has no alternative but to construct what he or she knows on the basis of…experience." This is put rather awkwardly, I think, but it is pretty easy to grasp. But the consequence of this belief, von Glasersfeld says immediately after, is that "all kinds of experience are basically subjective," and that knowledge claims can’t be reliably compared between persons. Yet in another quote, from Howe and Berv’s essay here, von Glasersfeld says that "Where knowledge is concerned, the concepts, theories, beliefs, and other abstract structures which the individual subject has found to be viable, gain a higher degree of viability when successful predictions can be made by imputing the use of this knowledge to others. The additional viability can be interpreted as indicating intersubjectivity and constitutes the constructivist substitute for objectivity." Not only is the intersubjectivity account of viability in conflict with the "fit" account von Glasersfeld provided previously; it is inconsistent, as far as I can see, with his own claim in the second passage quoted by Phillips, above, about the subjectivity of experience, in which according to von Glasersfeld the most one can say is that "I may find reasons to believe that my experience may not be unlike yours." (Which, I must say, is some pretty slippery phraseology.) It certainly falls far short of an account of intersubjectivity (or, as he insists on putting it, what "can be interpreted as indicating intersubjectivity").

And so, I am suggesting, constructivists have often been their own worst enemy in misframing issues, or taking on issues, that a more thoroughgoing constructivist account ought to approach in a very different way. It seems clear, for example, that even a construction is constructed out of something (let’s call it "evidence"). One can then argue that evidence is constructed, but it must be constructed out of something (let’s call it "experience" or "perception"). One can then argue that experience or perception is constructed, but this too must be constructed out of something. At some point you get to something that is not "constructed" in the sense that this term is meant. This is not necessarily the level of "reality," but it is an element in our experience that is not susceptible to construction, or at least not in the same way that these other interpretations are. (Notice that even the attribution of construction to these stages is an assertion about something that is "really happening" in our mental processes, and in everyone else’s too.) Constructivists want to insist that individuals or groups construct different versions of the world, or make sense of the same information or experiences in different ways (and I think they are clearly correct in this); but many of them do not seem to realize that this very conclusion rests upon a presumption of potentially shared experiences that may or may not provide a basis for agreement – not only through "mere" social and political consensus, but because at some level they are referring to (what they come to recognize as) the "same" thing. Beyond this point, I am suggesting, it is an empty dispute to argue about whether that thing is caused by or corresponds to an external world. What matters here are social processes through and through: what are the grounds for establishing the generalizability of such knowledge claims; what are the grounds for testing them; and what are the grounds for adjudicating competing knowledge claims? I see no reasons why most constructivists and non-constructivists alike cannot address these sorts of questions and reach some agreement on them.

Five propositions

As a way of advancing that conversation, I want to suggest five propositions, or basic ideas, that potentially cut across the pro- and anti-constructivist positions. They are not universally shared views within this debate, but I do believe that they identify points that any account must acknowledge and respond to, even if they would be given varying relative weights of significance by different accounts. As I said at the beginning, I think many debates in this area primarily come down to the strength of such relative weightings, not to fundamental disagreements.

1. All understandings of the world partake of a social environment, even when they are formulated by individuals alone. The very idea of a "construction" means invoking ways of making sense of experiences that are learned. With the possible exception of some innate quasi-Kantian categories, we construct understandings the same ways that we undertake any other complex human practice – under the guidance of others or drawing from examples of how to do it that we have learned from others. Von Glasersfeld’s idea that construction comes first, and that even our social interactions, relations, and intersubjective agreements with others are merely apparent accounts of the world that we may or may not accept, is far too neat, and can give no account of why we commonly construct the world in certain ways rather than others. If construction in any interesting and rich sense is possible, it is because it is a social process. Yet at the same time, this social perspective commits one (though some constructivists do not seem to recognize this) to some implicit view of a shared reality, not only as an accidental artifact that we happen to come to share, but as a condition of sociality itself.

2. Language provides the conditions for both understanding and misunderstanding. Appropriations of ideas such as Wittgenstein’s notion of "language games" has led some theorists to the idea that such pockets of language use are by definition incommensurable (which is far from Wittgenstein’s meaning, but that is another issue). A great deal is made, especially in some postmodern sources, that language can express such profound conceptual differences that the pursuit of intersubjective understanding is futile. (Jean-François Lyotard calls them "islands of meaning.") Other authors, such as Hans-Georg Gadamer or Jürgen Habermas, are much more optimistic about the possibilities of understanding or agreement. What is at stake here is the degree to which the processes of construction, which are expressed and negotiated through the available language(s), can be expected to reconcile radically different versions of the world. The pessimists and the optimists here both miss the point, I would suggest. The issue is not with the nature of the language, but with the practice of communication. The available language both facilitates and constrains the possibilities of communication (it is not a remarkable observation to say this); but whether communication can generate shared constructions is not determined by the available language itself, so much as by the ways in which persons enact the communicative relation. In fact, the very language that makes understanding possible makes misunderstanding possible and vice versa.

In the present context, my point has been that questions such as "Is there an objective, commonly shared reality?" do not lead anywhere useful. Because differences in experiences and the ways in which we construct understandings of these experiences are almost always adjudicated in and through language, the more salient question is, "Does it help or hinder this process to make assertions like, ‘This is the way the world is’?" I have suggested that in most cases it adds nothing constructive, and may inhibit a more substantial engagement of views.

3. Our efforts at understanding the world always occur at a distinct time and place; under a set of circumstances that motivate and influence our choice of questions, methods, and reference groups for cross-checking our understandings. Clearly some theorists seems to read off this premise a host of relativistic conclusions, since every person or group is by definition situated differently from every other. This is an entirely unnecessary conclusion. It seems clear to me that the inevitability of social and political influences on our attempts to construct workable understandings of the world does not mean that these conclusions are merely social and political. Denying such influences is one sort of mistake, I believe, but giving them determinative weight and force is another.

This discussion opens up another question addressed in several of the essays here: Is science different? Some writers seem willing to acknowledge that "social construction" is not a very remarkable thesis when applied to the phenomena of the social world, which seem directly susceptible to our efforts to define and describe them; but that the world of natural or physical science is different, because it is about phenomena that have an independent existence, apart from our efforts to describe or characterize them. This is a mistake. The processes of constructing understanding, and the relation between our versions and the phenomena and regularities they are trying to account for, is not fundamentally different in these two realms, even if the two sides of that relation might be given different degrees of weight in the social and in the natural contexts (although even there I would say that the differences can be exaggerated).

What makes the physical or natural sciences different is primarily something else, not the nature of the phenomena they are dealing with. They have, for a variety of reasons, established a stronger set of common standards of practice, common vocabularies, and common techniques of inquiry. Together these sorts of factors have made the practice of science more orderly and overtly consensual, while other forms of inquiry appear less so. But this difference does not mean that one endeavor depends less on constructions than the other. The tendency to put certain sciences in a separate category, as if they were about the cleaner and more objective pursuit of truth, has made it a special point of emphasis among certain social constructivists to try to show how the same factors of peer approval, self-interest, winning over others to one’s point of view, rivalry, avoiding group ridicule, seeking acclaim or financial reward, and so on, operate within these scientific communities of inquiry as they do in any other. This does not automatically make the practice of science corrupt, but it is difficult to imagine that they do not affect the practice of science, and how its theories get constructed and negotiated, in some very important ways.

4. The underlying issue that divides the anti- and pro-constructivists is their attitudes toward difference and disagreement. I mean this as a rather surprising claim. It seems clear to me that underneath the disputes over relativism, the existence of an external world, and so on, that are sketched in this book, what really divides the advocates of constructivism and its critics is that one side wants to maintain a broader scope for idiosyncrasy and pluralism in unreconciled (and perhaps irreconcilable) alternative accounts of the world, while the other expects, and wants to encourage, a resolution of differences in terms of criteria, etc., that can come to be commonly shared and applied. I have already argued that the latter desire is a matter to be demonstrated in practice, rather than asserted on a priori grounds: if these common standards do have some objective advantage, then bringing others to share them is the only way of actually establishing their generalizability. At the same time, the view favoring, or at least being willing to tolerate, a greater degree of idiosyncrasy and pluralism in competing accounts of the world, regards the imposition of criteria, etc., that people do not in fact share as presumptuous at best, a kind of intellectual domination at worst. Constructivism in its contemporary form provides a rationale for the more pluralistic view: this is pretty clearly behind the social constructivist project, and it is even more apparent in the radically individualistic constructivism argued for by von Glasersfeld.

Certainly advocates for inquiry of all sorts generally favor the benefits of disagreement, even vigorous and fundamental disagreement, as part of the process that subjects beliefs to rigorous questioning and testing; and that generates alternative hypotheses and potential solutions to problems. But these disagreements remain disagreements within an implicit compact of values, assumptions, and procedures that are not questioned; hence, however vigorous the disagreement might be, it does not challenge the fundamental purpose and legitimacy of the endeavor itself. This view can be linked in some respects with the characteristics of what Kuhn calls "normal science." But the more extreme, and potentially irreconcilable, forms of disagreement – of alternative constructions, if you will – are not just differences against a background of shared values, assumptions, and procedures, but a questioning of these as well. They pose a very different sort of challenge to the attempt to foster understanding and agreement across these differences, and they face a much greater chance of failing to do so. Moreover, they often require ways of speaking and imagining across differences that do not fit within the conventional languages and methodologies of scientific practice. It is this threat, I have suggested, that drives the insistence that there must be a set of common criteria, etc., that can be relied upon to adjudicate such differences, and that there must be a common world against which to compare, and ultimately settle, these differences. I have also suggested that some constructivists have erred in the opposite direction by denying out of hand that such commonalities exist, or can be established.

Instead I have suggested a greater agnosticism; that the identification and justification of shared criteria, etc., is itself a form of inquiry, one that has to operate within and across the existing views, and differences, of the persons and groups involved, and one that has to proceed with few preconceptions about what the end result of that shared inquiry will generate.

5. Constructivism operates within a problem-based framework, in which one potential problem is always the status of one’s constructions themselves. I have said that a fault that can be attributed to constructivists and their critics alike is in not being constructivist enough. On the constructivist side, this sometimes takes the form of stipulating what views constructivists can or cannot subscribe to; attempting to erect a superstructure of theory that is an account of "pure" constructivism, or leaping from a set of premises to a series of conclusions that are not particularly constructivist at all. If anything is constructed in our understandings, then everything must be, even the objects of our understanding themselves – which, as I have said, is not a very constructivist way of arguing.

On the other side, the critics of constructivism sometimes minimize the extent to which "constructions" (imagined, creative attempts to go beyond a set of given data to build meaningful and plausible accounts of the world) are an inevitable dimension of human inquiry, not a remarkable, or radical, or relativistic challenge to it. In this essay I have suggested that this is not where the disputes between constructivists and their critics reside: it is in taking on, from both sides, a set of broader epistemological and metaphysical assertions that are neither necessary to the constructivist account nor, in my view, particularly helpful to it. What can provide a way beyond this impasse, I have suggested, is to bracket such disputes (which simply cannot be resolved, in the terms within which they are framed, and do not need to be), and to focus on trying to understand the practices and procedures by which constructions come to be created, adjudicated, and commonly shared. I have suggested, in fact, that the standard epistemological and metaphysical disputes have often served a counterproductive purpose; and that since these arguments cannot be resolved, the chief remaining question is how holding such beliefs, and arguing about them, itself affects the substantive processes of inquiry.

Instead of questions such as, Is generalizability possible? Can our knowledge be objectively (or intersubjectively) tested? Is there a reality that grounds and affirms our constructions of it? I have suggested a different set of issues, informed by these five propositions: What are the processes and practices through which the activities of generalizability, objectivity, intersubjectvity, testing, and adjudicating differences must proceed? How do they work? How do they affect, and how are they affected by, the social and institutional contexts in which they take place? What are the points over which they can break down? What are their reasonable end-points? As I have said, this seems to me a more thoroughgoingly constructivist approach, and nothing outside of a "constructivist" account needs to be added to it. Where disagreements, conflicts, errors, and failures do arise, there is nothing outside of these same processes that can redeem, reconcile, or correct them.

The educational benefits of constructivism

In the area of pedagogy, I have less to add to the sensible and useful observations made by several of the preceding essays. It does seem to me worth underlining, as Matthews, McCarty and Schwandt, and others say, that constructivist approaches to pedagogy would be generally better off if their advocates stayed out of the epistemological and metaphysical speculations that they seem unable to resist. First, these broader pronouncements are neither necessary for constructivist pedagogy, nor particularly helpful to it; they distract the discussion far afield from the concerns pedagogues actually need help with. Second, people who might have excellent contributions to make to the area of constructivist pedagogy may not have the background to appreciate the depth and complexity of the epistemological and metaphysical disputes they are taking on when they wade into those waters. The sorts of questions about how social construction takes place, I have suggested, offer a far better starting point for thinking about pedagogy; since the way persons and social groups generally construct understandings may be a very good guide for how these processes take place in classrooms.

It seems to me almost a truism to say that all learning is constructed: understanding and evaluating new ideas and skills, even those of the most apparently rote character, requires reinterpreting them in light of one’s existing understandings and abilities. A corollary of this assumption is that no two people ever learn the "same" material in precisely the same way. As is touched upon in some of the essays here, some educational writers seem to think that there is a fundamental dispute between more didactic versus more discovery-oriented approaches to teaching. But the way material is taught may not be the determining factor in how it is learned; even ideas and skills presented in the most structured and directive manner will still need to go through a process of filtering and reinterpretation as they are being learned. There may be other reasons to avoid such general approaches to teaching; they may for example tend to promote a "hidden curriculum" of deference to authority, an attitude toward knowledge as something given, not subject to question, and so on. Even so, there may be instances where certain students, and certain subject matters, are taught best and most easily by adopting such an approach. One of the most important points made in this book, it seems to me, is echoed by Howe and Berv, and by McCarty and Schwandt: that one of the most detrimental legacies of constructivism as a view of pedagogy is that this "ism," like so many others, has been promulgated as the One Best Way of teaching; and that almost with religious fervor the advocates of constructivist approaches believe that the entire system of education should be transformed around their principles. Admittedly, constructivism is by no means the first or the only approach to do this, and by and large its stance has been one of correcting or counterbalancing the overwhelming dominance in schools of more didactic, content-oriented, test driven methods. Yet the only intelligent approach to teaching is one that recognizes that a skilled teacher needs many resources in her bag of tricks, and that different situations, different students, and different subject matters require the ability to adopt and adapt multiple approaches if they are going to be able to succeed as teachers in the face of many learning styles and degrees of motivation found among students. Constructivist approaches are valuable, but not the only resources they will need.

The other major lesson to be drawn from this discussion, also made clear by preceding chapters, is that a theory of teaching and a theory of learning are not logically locked together. The belief that they are remains a broadly held assumption in nearly every school of education, district or state office, and school in the country. Ironically, a constructivist view of learning should be especially aware that how new ideas or skills are presented does not determine how they will be learned. There may be other reasons why constructivist approaches to teaching have appeal, and there may be important ways in which they are more motivating or facilitative for students’ learning. But a very "inquiry-oriented" lesson may still result in a student ending up with a rigid and inflexible belief; and a more didactic approach may actually trigger a very active and imaginative process of reflection in the learner. The error made here is a typical one in the field of education: adopting a means-ends attitude toward teaching, in the endless pursuit of the "methods" that will reliably yield the "results" sought. It may be especially ironic that constructivists turn out to be just as susceptible to this myth as have others. But it also should cause us to consider why it is that this aspiration has such a hold over educators, why it continually reappears with each new One Best Way of teaching, and why its hegemony is so powerful that it even absorbs into it views of teaching and learning that ought to be highly critical of it, as in the case of constructivism.

The principles of constructivist pedagogy, encouraging collaboration, promoting activity and exploration, respecting multiple points of view, emphasizing "authentic" problem-solving, and so on, have a number of benefits and among these may be that these approaches do facilitate a more creative, synthetic attitude toward learning. But I have suggested here that the primary reasons why we might favor such an approach, generally, have less to do with driving a particular learning process, and more to do with the ways in which this sort of classroom reflects other values we might hold, or the ways in which these types of activities and relationships tend to foster other lessons, apart from the subject matter itself, that we also value.

One example of this way of reframing the pedagogical issue concerns the ways in which science is taught. A constructivist approach to science teaching would tend to try to reproduce in the classroom the kinds of conditions that drive scientific exploration generally (Burbules and Linn, 1991; Linn and Burbules, 1993; Linn, 1997). This is not necessarily because children learn science in the same ways that experts in the field establish new knowledge, but because learning how science is done is itself an important educational goal, apart from learning scientific content itself. First, the skills and dispositions of inquiry and collaboration may have relevance to all sorts of other activities in later life. Second, situating the content of science in the context of narratives about how science gets done (Watson and Crick’s The Double Helix, for example), can be extremely motivating for students, regarding the process of scientific exploration more like a mystery, or a race, or a comedy of errors, than a desiccated, cold, impersonal endeavor. Third, this sort of understanding of science also tends to demystify the status of scientific knowledge, making it clear that beliefs change, that they don’t come from "nowhere," and that the kinds of factors that influence human practices generally are not absent from the scientific domain either; this seems to me a crucial, underestimated educational goal in itself. Constructivist accounts of the scientific disciplines have provided a valuable perspective in looking behind the closed doors of laboratories, I believe, even when the broader epistemological and metaphysical conclusions of the sociology of scientific knowledge may have been overdrawn.

This discussion reveals another of those instances, I believe, where reflection on an educational problem provides insight into rethinking a philosophical issue (although the line of influence is typically drawn in the other direction). Teaching and learning in a constructivist manner force us to narrow in on understanding the practices and procedures by which human actors, in specific social contexts, generate hypotheses, assess them, compare them with others, adjudicate disagreements or controversies over methods of inquiry, and so on. The irony is that debates over constructivist theory quickly lose sight of these factors and get embroiled in disputes that are almost entirely generated at that abstract level. As I have said, nothing could be less constructivist than that.

REFERENCES

Nicholas C. Burbules and Marcia C. Linn (1991). "Science education and philosophy of science: Congruence or contradiction?" International Journal of Science Education, Vol. 13 No. 3: 227-241.

Ian Hacking (1999). "Are you a social constructionist?" Lingua Franca (May/June 1999): 65-72.

Marcia C. Linn (1997). "The role of the laboratory in science learning." Elementary School Journal, Vol. 97 No. 4): 401-417.

Marcia C. Linn and Nicholas C. Burbules (1993). "Construction of knowledge and group learning." The Practice of Constructivism in Science Education, Kenneth Tobin, ed. (Washington, DC: American Association for the Advancement of Science): 91-119.

D.C. Phillips (1997). "How, why, what, when, and where: Perspectives on constructivism in psychology and education." Issues in Education: Contributions from Educational Psychology, Jerry Carlson, ed. (Stamford, Conn.: JAI Press), 151-194.

Thomas S. Popkewitz (1998). "Dewey, Vygotsky, and the social administration of the individual: Constructivist pedagogy as systems of ideas in historical spaces." American Educational Research Journal (Vol. 35 No. 4): 535-570.

Ernst von Glasersfeld, "An introduction to radical constructivism." The Invented Reality: How Do We Know What We Believe We Know? Contributions to Constructivism, Paul Watzlawick, ed. (New York: W.W. Norton), 17-40.