You are here


In Theatetus, Plato sets the agenda for Western epistemology: What is knowledge? More exactly, What is it that distinguishes knowledge from mere true belief? What is this elusive quality or quantity enough of which, together with truth and belief, is sufficient for knowledge? Call that quantity, whatever it is, ‘warrant’, in Warrant: The Current Debate, the first volume in this series, I considered some of the main contemporary views of warrant. You may not be surprised to learn that I found them all wanting. Now it is time to take the next and more dangerous step, leaving the safety of the philosophical bunker from which one snipes at other views for the more risky business of exposing my even views on the subject. (My plan was to call this volume Warrant: the Sober Truth, but wiser heads prevailed.) Before taking that precarious step, however, I wish to recapitulate briefly the main themes of the first volume. Its principal aim was to take a careful and critical look at some of the main contemporary accounts of warrant, trying to understand them, noting where they go wrong or are inadequate, seeing what can be learned from them, and trying to figure out where to look for a better account.

In the first chapter of Warrant: The Current Debate I noted that twentieth-century British and American epistemology has been heavily internalist. Internalism is really a loosely related set of views about epistemic access—access to whatever it is that makes for warrant. According to the internalist, the knower can know, and know in some special way, that a certain proposition or belief has warrant or justification for her; alternatively, she can know that the condition constituting the ground of warrant is present, or that she has the property that makes it true that a belief has warrant for her, or something else in that neighborhood. But this epistemic access must be special in some way. I have access to the distance from Vienna to Prague (I own an atlas); but that kind of access doesn't count. The access in question must be privileged in some way: perhaps it is certain for the epistemic agent that the condition in question holds, or perhaps she can determine by reflection alone that it holds, or perhaps there is a certain kind of mistake she can't nonculpably make; however precisely she puts it, the internalist claims that a knower has special epistemic access to the conditions of warrant.

Internalism in epistemology, so I argued, goes hand in hand with the idea that warrant is really justification. More exactly, what it goes with is the thought that justification is necessary for warrant and nearly sufficient for it: what is required in addition is only a fillip to mollify Gettier, I noted that twentieth-century epistemology displays a vast and confusing diversity. The major movements, however, unite in declaring that there is an intimate connection between justification and warrant, between justification and internalist constraints, between justification and evidence, and between justification and the satisfaction of epistemic duty. Twentieth-century epistemology also displays, however, great diversity with respect to the question what, exactly, justification is. Some say it is intimately connected with duty and responsibility—a matter of aptness for epistemic duty fulfillment (Chisholm) or perhaps a matter of pursuing one's epistemic goals responsibly (Bonjour). Others say that justification crucially involves having evidence for a belief (Alston, Conee, Conee and Feldman, many others). Others see the degree of justification enjoyed by a belief as a measure of how well you have fulfilled your epistemic goals (Lehrer); still others see a belief's being justified as a matter of everything's going right with respect to the knower as knower; yet others say it is a matter of a belief's being produced by a reliable belief-producing mechanism.

I argued that this blooming buzzing confusion with respect to justification can be reduced to order by going back to the fountainheads of contemporary theory of knowledge, those twin towers of Western epistemology, Descartes and Locke. What is crucial to Descartes and (perhaps even more important) Locke, is epistemic deontologism: the view that there are epistemic duties and obligations. According to Descartes, the central epistemic duty is that of abstaining from any belief that isn't clear and distinct; according to Locke, it is that of proportioning degree of belief to degree of evidential support by what is certain. Justification itself, however, is just the condition of having done your duty and satisfied the requirements: the condition of doing no more than is permitted, going contrary to no duty or obligation. The main contemporary conceptions of justification can all be understood in terms of this deontological tradition—either as explicitly carrying it on, or as diverging from it to one or another degree, and thus using the term ‘justification’ in a sense analogically related to this aboriginal sense.

It is therefore easy to see the historical roots of contemporary concern with epistemic justification. Further, as I pointed out, deontology generates internalism; so it is also easy to see the roots of contemporary internalism. Finally, given Locke's view that the prime epistemic duty just is that of proportioning degree of belief to the evidence (evidence from what is certain for me), it is equally easy to see the origin of the contemporary stress on the importance of evidence with respect to justification.

I turned next to some prominent internalist construals of warrant. First there is the powerful and powerfully influential work of Roderick Chisholm. According to the classical Chisholm, warrant is aptness for epistemic duty fulfillment. We all have a duty to get into the right relation to the truth: the better you can fulfill that duty by believing a certain proposition, the more warrant that belief has for you. This is an attractive view (and for many years I followed Chisholm is accepting it); closer examination, however, shows that it can't possibly be correct. First, the very notions of duty and obligation don't apply at all directly to the formation and sustenance of belief; for the most part these things are not within our direct control. But concede as much control as Chisholm needs: it is still wholly obvious that epistemic dutifulness is nowhere nearly sufficient for warrant. I may be doing my level best, I may be trying my hardest to get into the right relation to the truth, but, by virtue of epistemic malfunction, may still fail miserably—and fail in such a way that my beliefs have little or no warrant. And therefore I may be magnificently dutiful in forming or maintaining a given belief, but still be such that it has no warrant for me. The moral is that justification strictly so-called is nowhere nearly sufficient for knowledge or warrant. I also argued, though less vociferously, that justification isn't necessary for warrant either.

Turning to the post-classical Chisholm, the Chisholm of “The Place of Epistemic Justification”1 and Chisholm's “Self-Profile”2 we find quite a different view. Here the idea is that warrant, for a belief, is a matter of a certain fittingness between that belief and a person's “evidence base,” that is, “the conjunction of all the purely psychological properties that that person has at that time.” (Purely psychological properties are “those properties to which we have privileged access. Every such property is necessarily such that, if a person has it and if he attributes it to himself, then his attribution is evident in the strongest sense of the term.”3) This is a much more general view than classical Chisholmianism; one gets different specifications of it by differently specifying that relation of fittingness. Taking the view broadly, it includes coherentism; taking it still more broadly, it might even encompass the externalist theory I mean to propose. Taken Chisholm's way, however, this view seems clearly mistaken, as I argue in chapter 3 of Warrant: The Current Debate.

I turned next to coherentism überhaupt (chapter 4), then to the coherentism of Laurence BonJour (chapter 5), and then to Bayesian coherentism (chapters 6 and 7) a recent and very interesting coherentist entry in the lists. Coherentism überhaupt is unsuccessful because it sees warrant as involving only the relation between beliefs; but the fact is the relation between experience and belief and between environment and belief is also crucial to warrant. Bonjour's articulate version of coherentism (so I argue, but BonJour might conceivably disagree) suffers from that problem, as well as some others specific to it. Bayesian coherentism isn't really and answer to the question of warrant (and if addressed to it would be a wholly unsatisfactory response); it is directed, instead, to the question of rationality. Rationality is protean and ‘rationality’ is multiply ambiguous; thinking about Bayesianism gives us a chance to disentangle some of the main varieties of the former and some of the main meanings of the latter. There is Aristotelian rationality, which goes with being a rational animal; there is means-end rationality and Foley rationality, an epistemic special case of it; there is the sort of rationality that amounts to sanity, to epistemic proper function; and there is the rationality that is really a matter of believing in accord with the dictates of reason. Satisfying the Bayesian constraints, I argue, is neither necessary nor sufficient for any of these kinds of rationality. Bayesianism, so I argue, really describes the mental life of a certain kind of ideally rational agent—one who, like us, holds beliefs, reasons, learns, and the like, but is also unlike us in certain crucial respects. As such, Bayesian constraints aren't really conditions on our rationality (so one who violates them isn't necessarily irrational); but in some special cases these constraints may function as ideals to which we can appropriately strive to conform.

So internalist accounts of warrant go awry. Moving toward externalist accounts, I turned next in chapter 8 to the work of John Pollock, who offers, so I claimed, a sort of uncomfortable halfway house, an uneasy compromise between externalism and internalism. Pollock argues, fundamentally, that a belief is justified for a person if she arrives at it in conformity with her own norms. His discussion of norms is subtle and penetrating, but (so far as I can see) it does not yield a successful account of warrant. The reason is that it seems perfectly possible for my epistemic norms to be incorrect norms, and incorrect in such a way that my carrying on my epistemic life in conformity to them is nowhere nearly either necessary or sufficient for warrant. My beliefs could therefore be Pollock-justified, but have no warrant; Pollock-justification, therefore, is deficient as an account of warrant.

I turned finally to reliabilist and paradigmatically externalist accounts of warrant; here I examined the proposals of Fred Dretske, William Alston and Alvin Goldman in chapter 9. Reliabilism marks a real advance—or better, it represents a fortunate retreat, a happy return to the externalist perspective occupied much earlier by Thomas Reid, and earlier yet by Aquinas and Aristotle. Still, reliabilism does not offer a correct account of warrant. The early Goldman offers a stylized and paradigmatic reliabilism: A belief has warrant if and only if it is the product of a reliable belief producing mechanism (or process or faculty). But there are two problems here: first how can the reliabilist account for the fact that warrant comes in degrees? Attempts to follow Goldman by doing this in terms of degree of reliability lead straight to the generality problem, the reef on which the early Goldman founders. And second, a belief may be the product of a reliable belief producing mechanism, but if the mechanism in question malfunctions (the agent is drunk, or ill, or under attack by attack by a shark) the resulting belief has little or no warrant, despite its respectable source. This problem bedevils both the early and the later Goldman. Reliabilists, as I see it, call attention to one of the four conditions characterizing paradigm cases of knowledge; reliabilism is therefore an approximation to the truth. But Reliabilists also neglect the other three traits of paradigm cases of knowledge; reliabilism is therefore (at best) a zeroeth approximation to the truth.

Can we do better? Indeed we can (so I claim, anyway). As I see it, a belief has warrant if it is produced by cognitive faculties functioning properly (subject to no malfunctioning) in a cognitive environment congenial for those faculties, according to a design plan successfully aimed at truth. In the first couple of chapters of the present volume I fill out, develop, qualify and defend this view, exploring along the way some of the convoluted contours of the notion of proper function. In the next seven chapters I consider how the proposed account works in the main areas of our cognitive design plan: memory, introspection, knowledge of other minds, testimony, perception, a priori belief, and probability. Then in chapter 10 I consider broader, structural questions of coherentism and foundationalism. My account of warrant meets the conditions for being a naturalistic account; but in chapters 11 and 12 I claim that natural him in epistemology flourishes best in the context of supernaturalism in metaphysics, for (as I argue in chapter 11) there appears to be no successful naturalistic account of the notion of proper function. Finally in chapter 12 I argue that metaphysical naturalism when combined with contemporary evolutionary accounts of the origin and provenance of human life is an irrational stance; it provides for itself an ultimately undefeated defeater.

Here I must acknowledge a complication with respect to my way of thinking of warrant. I aim at something in the neighborhood of an analysis of warrant: an account or exploration of our concept of warrant, a concept nearly all of us have and regularly employ. (As we all know, desperate difficulties beset any attempt to say precisely what analysis is.) Thus at the least I should be looking for necessary and sufficient conditions. But I very much doubt that there is any short and elegant list of conditions at once severally necessary and jointly sufficient for warrant. This is a way in which philosophy differs from mathematics; and epistemology differs more from mathematics, along these lines, than, for example, philosophy of logic or the metaphysics of modality. Our concept of warrant is too complex to yield to analysis by way of a couple of austerely elegant clauses. The structure of this concept, I believe, involves a central picture, a group of central paradigms—clear and unambiguous cases of knowledge—surrounded by a penumbral belt of analogically related concepts, concepts related by different analogies and standing in different degrees of closeness to the aboriginal paradigms. Between the central core area and this penumbral belt there is a more shadowy area of borderline possible cases, cases where it isn't really clear whether what we have is a case of warrant in the central sense, or a case of one of the analogically extended concepts, or neither of the above; and beyond the penumbral belt we have another area of borderline cases.

Hence perhaps a good way to characterize our system of analogically related concepts of warrant is to give first, the conditions necessary and sufficient for the central paradigmatic core. (Even here, as we shall see, there is no stylishly sparse set of necessary and sufficient conditions: various qualifications, additions and subtractions are necessary.) Second, what is needed is an exploration of some of the analogical extensions, with an explanation of the analogical bases of the extensions. This way of proceeding is less elegant and pleasing and more messy than the analysis we learned at our mother's knee: it is also more realistic.

In the preface to Warrant: The Current Debate I acknowledged my indebtedness to many (embarrassingly many) from whom I have learned and who have helped in various ways with these books. Here I want to single out for special thanks Bill Alston, Tom Senor, Dean Zimmerman, and Robin Collins. I should also point out that there are points of contact between the position I develop and the work of Ernest Sosa, Marshall Swain, and Hector Castañeda. Still another kind of intellectual debt: the position I shall develop is broadly Reidian; the global outline of Thomas Reid's epistemology seems to me to be largely correct.

Of course the fundamental style of Reidian epistemology with its externalist emphasis didn't originate with Reid; it can be traced back to Aquinas and indeed all the way back to Aristotle. A more proximate precursor is Claude Buffier (1661–1737),4 in his First Truths, and the Origin of Our Opinions, Explained: with an Inquiry into the Sentiments of Modern Philosophers, Relative to our Primary Ideas of Things (Paris, 1724). There is an anonymous English translation (1780) which adds as it were another subtitle, “To which is Prefixed A Detection of the Plagiarism, Concealment, and Ingratitude of the Doctors Reid, Beattie, and Oswald.” The author complains that “Of later years the Transtweedian regions have swarmed with a new species of men, different from their itinerant peddlers in the wares they sell, but familiar in the manner of packing them together from the labours of others …” (p. vi). These Transtweedians, furthermore, “persevere in collecting materials from other authors, and, industrious to conceal their plagiarism, compile and assume them as their own” (p. vii); they have “clandestinely taken the principles and opinions of Pere Buffier, converted them to their own purposes of acquiring fame, and concealed the theft by ungratefully unacknowledging the person to whom they are obliged …” (p. viii). The most dangerous (because ablest) of these Transtweedian ingrates is Reid himself: “Dr. Reid, in his Enquiry, has carefully avoided literally transcribing the passages relative to DesCartes, Malebranche, Locke and Berkeley, and the observations on them, which are to be found in Buffier; but he has with no less care adopted his sense, and modestly affirmed it as his own” (p. xi). Naturally I hope to avoid Reid's offenses. I therefore concede in advance that what follows has been influenced to any degree you please by Reid and hence, by the transitivity of influences, Buffier himself.

Notre Dame, Indiana

April 1992