| PHILOSOPHY OF EDUCATION 1995 |
Scylla, Charybdis, and Social Epistemology:
A Response to Alvin GoldmanD.C. Phillips
Stanford University
It is probably outside the bounds of a commentator's duties to speculate about the presenter's state of mind or deeper psychologial motivations; nevertheless I must point out that Alvin Goldman seemed ambivalent about accepting the accolade of "defender of epistemological family values." Presumably he realized he was damned if he did, and damned if he didn't. For to accept the title is to run the risk, in these days when society is bifurcated about family values, of being labeled as an old-fashioned conservative, or a friend of Newt's. But to decline, or to pause to ask the obvious question -- "the values of whose family?" -- is to court identification as a radical, and probably a relativist to boot. (These days we might even throw in the epithet "postmodernist," for this is a rapidly growing family, the members of which are clamoring for their epistemic values -- or lack thereof -- to be recognized.) Navigating between this dangerous Scylla and Charybdis, epistemic conservatism on one hand and radicalism or skepticism on the other, is the task that faces not only Professor Goldman but all of us who wish to take the social epistemology approach seriously. But more of this later.
AN OVERVIEW The first half of the paper outlined what Professor Goldman called a veritistic approach; the opening pages bear the reminder that both individuals and society have a strong interest in the acquisition of knowledge (a term used here in the weak sense of "true belief"). From this perspective, education, whatever else it may hope to accomplish, has as a fundamental aim, the providing of individuals with knowledge and the skills that are requisite in order to attain it. In the course of considering a few objections to this basic position, Professor Goldman offered an important caveat -- he did not consider global skepticism about truth and knowledge (this is too large a topic to confront in the limited confines of a conference paper). I shall come back to this in my later comments. Goldman then proceeded to discuss whether teachers can be regarded as epistemic experts; this was followed by a consideration of the issue of multiculturalism which, in essence, he construed as the issue of people in different cultures wanting different subsets of the large stock of true ideas to be taught to their students (a matter which he found to be relatively unproblematic and quite compatible with a veritistic approach).
It was only at this stage that Professor Goldman introduced his conception of social epistemology: the study of the social or interactive practices of multiple agents to determine how such practices encourage or obstruct knowledge acquisition. He focused upon whether teachers always have to give their students reasons for believing (a position enticingly advocated by Israel Scheffler, Harvey Siegel and others).1 Goldman suggested that the "giving of reasons" should be regarded as a form of argumentation, and, following the line developed in his paper in the Journal of Philosophy,2 he spelled out some of the criteria that should apply here. He also suggested that, seen from the perspective of social epistemics, a good argument must be sensitive to the members of the intended audience and their capacity for appreciating the considerations that are presented -- a point of obvious educational relevance. Overall, Professor Goldman stressed that the giving of reasons is a veritistic activity (a point that, from his perspective, Scheffler and Siegel do not stress sufficiently).
This leads to the final part of the paper where Professor Goldman offered what is essentially a challenge to the reasons approach: Must a speaker or teacher give reasons for everything he or she says? Cannot a hearer be justified in believing X, simply because the speaker asserts it? Goldman pointed out that there are traditions that stress, to different degrees, the autonomy of the student -- the view that the student has the right and the responsibility to be self-governing, and should never accept what someone says simply because he or she says it. Professor Goldman -- citing recent scholarship on testimony, and presumably motivated as well by what he considers to be the epistemic practice in most societies -- indicated that he was prepared to countenance a modification of the claims of autonomy here: He suggested that while it is not always appropriate to place final trust in the say-so of other people, the default position is that trust is warranted even when the hearer has no independent grounds for certifying the speaker's reliability. I think that part of Goldman's background concern here is the age-old problem that children are not born with sophisticated intellectual skills; they have to be taught these things, and it is difficult to see how education can get off the ground if the child is supposed to demand reasons for every belief, especially when the child has not yet acquired the necessary groundwork of beliefs and skills.
SOME FRIENDLY AMENDMENTS It is time to turn to some comments which -- rather than being offered as criticisms -- are advanced as issues that I think should be placed high on the agenda, as a person who, for several years, has been intrigued by the possibilities of Professor Goldman's social epistemology program. (I do not want to be so churlish as to criticize Professor Goldman for sins of omission as there is so much to be said on the topics he opened-up in his paper, and he has another book in progress on these matters.) The first comment is global in nature. Then I will turn to two important issues that arise at specific places in the paper.
It will be easier for me to make my global comment about the paper if I can draw on a classic distinction made by one of the founders of analytic philosophy of education, the British academic Paul Hirst,3 and by an educational theorist at Chicago, Joseph Schwab.4 Hirst's and Schwab's distinction, which I argued many years ago does not stand up to critical scrutiny if taken too seriously,5 is a useful heuristic device (it is very revealing, for example, even at university level, to analyze one's own lectures, or the assignments one sets for students, in Hirstian or Schwabian terms). In the context of discussing the disciplines or "forms of knowledge" typically considered to be part of liberal education, Hirst distinguished between (1) a field's concepts and the relations between them (what Schwab conveniently called the substantive structure); and (2) the tests against experience, the warranting criteria, the key techniques of inquiry (what Schwab termed the syntactical structure). More crudely, the substantive structure constitutes, or at least expresses, the substance of the discipline; the syntactical structure is the discipline's methodology, broadly conceived. It was part of both Hirst's and Schwab's contention that, to be educated in a discipline, the student must have an understanding of both the substance and the syntax of that field. This is why this distinction, despite its crudity, is so helpful in practice; in many fields -- maybe philosophy is an exception -- learners are given a great deal of exposure to, and are tested on, the substance; but the syntax -- an understanding of the truth tests, criteria and techniques that are used to build knowledge in that field -- is too often given short shrift. Yet one can ask whether students really understand science, or math, or history if they have no conception of how new knowledge gets established in these fields.
To return to Professor Goldman's paper: While he acknowledged in passing that one of the aims of education was to enable students to acquire intellectual skills, the bulk of his discussion focused upon knowledge, and upon the giving of reasons for acceptance of beliefs. As usually understood, both of these fall squarely within the sphere of "substantive structure." After all, in a veritistic approach, the reasons for belief in a substantive proposition are, loosely, part of the evidence for that proposition being true, and hence to a large degree, are part of the discipline's substance -- although, as always, methodological or syntactical matters lurk in the background. The point to which I have been leading is that in this substantive domain, the insights that are provided by the social epistemology approach are not radically different from those that come from more traditional sources -- certainly so far as educational ramifications are concerned.
The situation is quite different, I would argue, when we turn to Schwab's and Hirst's "syntactical structure." For, on those (unfortunately quite rare) occasions on which teachers, textbooks, or other experts attempt to lead students to a working knowledge of this methodological domain, the traditional account -- the one that predominates -- is markedly different from that which a social epistemologist would be inclined to give. The accounts that are given are invariably twenty years or more behind the times, and the picture painted is one that depicts the inquirer in these fields as a lone, heroic figure struggling against nature (or the past!), using arcane logic and methodological guidelines that come from goodness knows where! I know from personal experience working with prospective teachers, all with degrees from high-quality institutions, that their knowledge of, and interest in the methodology or syntax of the discipline they are teaching is minimal. In my experience this is true even for beginning teachers of literature whose understanding of the syntactical issue of the nature of the literary canon and how it was established is also minimal.
The social epistemologist has a lot to offer here, especially if he or she also takes a reliabilist approach towards epistemology. For, to a large degree, it is the syntactical structures of the disciplines that are responsible for the reliability of the knowledge claims, and the evidence for them, in those fields. Thus, there is far more to be said than Professor Goldman's suggestion that the skills of a good reason-giver or of a good listener are important, as also is a degree of epistemic trust. To stay with science as an example, it would seem that students, and of course teachers, ought to understand the significance of such things as the following: the communal nature of science; the importance of open-mindedness, and frankness about assumptions and embedded values; the importance of communal pressures and rules that encourage scientists to face the issues raised by their critics; the existence of journal refereeing practices; the fact that so-called scientific method has changed over the ages; the debates over such syntactical or methodological matters as the admissibility or otherwise of ad hoc hypotheses; and the constant communal debates over the admissibility of evidence, as in the recent efforts to establish whether or not a couple of scientists in Utah successfully conducted nuclear fusion in a pickle jar. Examples such as the one Professor Goldman mentions briefly (the "enormous theorem") would be worth discussing with students, as would research in particle physics (research at the Stanford Linear Accelerator, for example, involves approximately a thousand physicists and engineers); or the attempts to prove Fermat's last theorem -- attempts in which mathematicians rely upon extremely complex prior work, and where few people could keep the total "proof" clearly in mind, or even adequately and independently assess it.
It is for this reason that I was disappointed that Professor Goldman gave so much attention to the imparting of substantive beliefs in his paper -- where his account merges or overlaps with that of traditionalists; this is why I wanted him to turn more than he did to the far-too-often ignored syntactical realm, where his social epistemics is excitingly different, and where his position is chock-full of implications about the kinds of social attributes that students ought to acquire in the course of their education.
I turn now to two necessarily brief comments about specific passages in the paper.
(i) The issue of how (or even whether) adults are to respect the autonomy of children who have not yet reached the age of reason (whatever age that is) is a difficult one. Professor Goldman, you will recall, is tempted to adopt a neo-Reidian position according to which students can accept the testimony of others on trust. Another strategy would be to say straight out that young children need to be indoctrinated, and therefore not all forms of indoctrination are bad. A variant of this latter view was put forward a decade or more ago in the Journal of Philosophy by George Sher and William Bennett,6 who argued for "directive teaching" (which in a response I showed was tantamount to indoctrination, and would not work in the areas that Sher and Bennett most wanted to use it).7 Harvey Siegel makes a different suggestion in his book Educating Reason, and I think that Professor Goldman could take this as a friendly amendment: When the child is at the stage where he or she does not understand what reasons are, or what counts as good reasons, we are not indoctrinating if we inculcate beliefs without giving reasons -- provided that we do so in the spirit of fostering the development of rationality and what Siegel calls an "evidentiary style of belief."8 Because if we carry out our instruction in this manner, the child will eventually come to understand that all beliefs have reasons that support them, and he or she could -- if the situation arose -- examine beliefs that were inculcated at an earlier stage. A child who had been indoctrinated might not be able to do this. I suspect that Siegel would want to make a similar point about the acceptance of testimony that is unsupported by the presentation of reasons.
(ii) I do not believe that the radical advocates of multiculturalism in education would be left content by Professor Goldman's discussion of this topic, or by his remarks on the teaching of true beliefs earlier in his paper. For, in this paper at least, he does not address what in my experience many of these individuals would identify as the key issue, and it is an issue that links multiculturalism with skepticism about truth. These multiculturalists adopt a position that I once labeled, in a slightly different context, as "rampant Kuhnism" (not to mention "rampant Wittgensteinianism" and "rampant Winchism")9; they believe that different cultures are rather like Kuhnian paradigms or Wittgensteinian forms of life, and by a short argument that follows predictable lines they reach the conclusion that there are "multiple realities," and they argue that it cannot be said that the beliefs held within one culture are more or less valid than those in another. On this view, then, the truths that the Neur want to teach to their young are not part of the same set of truths that we want to teach ours; similarly the truths that Freudians and Skinnerians want to impart are not different selections from a greater, encompassing set of truths that both groups would recognize as true. Thus, on this view, "truth" becomes relativized to particular (socio-cultural) frames of reference. (I ignore, for the sake of simplicity, the issue of whether these differing sets of truths are incompatible or merely incommensurable; the supporters of the radical view I am discussing do not usually draw so fine a distinction.)
It is important to recognize, however, that the rampant Kuhnian/Wittgensteinian multiculturalists are not alone in holding these views -- they are shared, I think, by some of the more radical social constructionists, postmodernists, feminist epistemologists, and perhaps some of the "strong" sociologists of knowledge (and this is no surprise, for Kuhn is recognized as an ancestor of several of these groups as well.) And here lies an important challenge for Professor Goldman's social epistemics. For (wisely, in my view) he does not seem to want to allow the "social" part of his position to lead him too far in the radical constructionist direction; however, he also ought not to underplay the social aspects of social epistemology, for to do so will leave him standing with traditional and relatively a-social epistemologists. This brings us back to Scylla and Charybdis, which is a good place to end.
1. For example, see Harvey Siegel, Educating Reason (New York: Routledge, 1988).2. Alvin Goldman, "Argumentation and Social Epistemology," Journal of Philosophy 91 (1994): 27-49.
8. Siegel, Educating Reason, 89.
9. Phillips, Philosophy, Science, and Social Inquiry, ch. 8.