PH29A The legacy of Descartes
Knowledge and certainty
It is common to attribute the direction taken by much traditional epistemology as a consequence of a search for certainty. This is usually offered by way of diagnosis of mistakes that have been made in following this path. Popper has made much the same claim about what he regards as the common-sense view of knowledge: it is distorted by this same search.
Uses of "certain"
What exactly is the certainty in question? It is common to contrast A's feeling certain, a subjective, psychological issue, with an objective notion that some claim or other is certain. On the subjective side, one can agree with Popper (1972, pp. 78-79) that this sort of certainty is a situational and psychologically variable matter. It would seem to be the extreme end segment of a continuum of subjective attitudes to the reasonableness of claims that range from certain rejection to certain acceptance. (Using just one primitive epistemic relation - p is as well justified as q - Chisholm (1989, p. 15) offers 13 qualitatively distinguishable positions between the two, though these are not about subjective feelings but rather epistemic probabilities.) If you feel certain that p then you believe that p, but it is clear that we cannot go in the opposite direction: you may well believe that p without feeling certain that p. And, simply but crucially, however certain you feel, you may yet be wrong, so certainty cannot be all there is to knowledge.
The objective notion, "it is certain that p", is none too clear. It might be a claim that things could not possibly be otherwise than as p says (p is necessary in some sense), or that we could not possibly be in error with respect to p (the claim that p is infallible), or that our claim that p could not be corrected (the claim is incorrigible). Or various other possibilities that we will come to.
Knowing and being certain
[It will helpful to use some abbreviations: Kap for A knows that p; Cap for A is certain that p; Bap for A believes that p; p -> q for if p then q.]
We have seen that not(Bap -> Cap). The more contentious issue is whether your knowing that p requires you to be certain in this subjective sense. The case that is usually cited here is Woozley's unconfident examinee: in the anxiety of the examination room, she gets the answers right but is completely lacking in confidence in them. One might say afterwards "Look, you really did know them". This is supposed to show that she knew but was not certain. This sort of story needs care lest it spill over into what most people think of as the lot less plausible claim that you can know p without even believing it at all. Pojman discusses this variant of the argument (pp. 6-7) and offers the contrast between occurrent and dispositional belief to save the standard claim. But it does seem possible to tell a story in which a person believes p, is no longer confident that it is true, but can be said to know it.
Armstrong claims that while it is not the case that Kap -> Cap it is very odd for anyone to say "I know that p but am not certain that p". I have emphasized the say, since I have just claimed that it is quite possible for it to be the case that I know and am not particularly confident of what I know. This is a point to keep in mind, since there is a frequent temptation to argue about what we do or do not say in the hope that this shows directly what is the case. (One example might be Wittgenstein's thought that when his interlocutor keeps saying "I know I have a hand" he might remark to a visitor that "he's not mad, we're only doing philosophy" - it would be if the argument is: outside the odd context of doing philosophy, we do not say "I know I have a hand," so having a hand is not an item of our knowledge.) We owe to Grice an approach (by way of appeal to the conventions governing polite conversation) that explains away many of the things we do not say or the things we take something to mean. But if we agree with Armstrong that there is some oddity in asserting knowledge together with lack of certainty, we need some account of what the link between them is.
Armstrong's own view is that in the case of singular propositions to claim to know that p is to claim that one's belief that p is linked in a law-like and thus reliable way with p's being the case. If that is what one is claiming, then it would be irrational to doubt that p, i.e. to have any other attitude to one's belief that p other than being sure. This seems to collapse Chisholm's 13 steps rather too quickly into doubt/certainty. Dancy, in the context of a similar reliabilist account of knowledge, allows that what the diffident examinee lacks is confidence in the reliability of her truth-tracking. Her beliefs are in fact reliably linked to the world (so she knows), but she wonders whether this is so (so she doesn't claim to know).
Another diagnosis of the supposed link between Kap and Cap sees it as committing a simple logical fallacy - moving from the truth that necessarily (Kap -> p) to the falsehood that Kap -> necessarily p, or alternatively as Flew puts it, if A knows that p then it is impossible that A is mistaken but only in the innocuous sense that A happens to be right, not that he cannot be wrong.
Any link may be mainly in the first-person case. Fred knows that p doesn't look like it has any consequences about Fred's feelings of certainty or reasonable grounds for p. Hintikka once claimed that I know that p epistemically implies (a notion that he defined by reference to assertion) I know that I know that p, but this does not carry over to the third person. Here again we are dealing with issues due to a person's assertion of something rather than the claim asserted. But the issue of knowing that one knows, and how it is entangled with simply knowing, is also one to keep in mind.
One final thought from the reliabilist camp: Papineau argues that if we want to arrive at knowledge we want to use processes that are as reliable as possible. Certainty would be 100% reliability in all possible worlds. We can reject that and still seek 100% reliability in this world and those relevantly close to it (this is a way of embracing the obvious fact that we do not know precisely how the world will turn out, so any method we use must be sufficiently robust to cope with this ignorance). He then suggests that people have embraced certainty because "it would be easy to slide, from the thought that you need ideally to guard against any possibilities that your information leaves with any non-zero probability, to the thought that you need ideally to guard against any possibilities whatsoever." The trouble with this is that, once Descartes reminds you of what some of the possibilities might be, it ought no longer to be so easy.
The Cartesian Project
Bernard Williams argues that if an enquirer wishes to arrive at non-accidentally true beliefs in a non-magical world, and if he or she has no other diversions from this epistemic task (no considerations of other costs — time and effort, need to make a decision, etc.), then it is sensible for such an enquirer to search for a method which is effective and error-proof, which means that the beliefs generated by this method would be certain. One crucial link here is the notion of effectiveness: the enquirer must be able to tell that the beliefs and procedures he is dealing with are true or reliable.
What exactly does that require? One possibility is that the beliefs are incorrigible: p is incorrigible - necessarily, Bap -> p.
Cogito ergo sum is also self-verifying — p is not necessary and necessarily, if p is asserted then p. The cogito is itself a contingent proposition.
The cogito is also evident to A: p -> Bap.
Yet another idea Descartes works with is, in Williams' terminology, that p is irresistible — if A thinks that p, A believes that p. Irresistible propositions are not for that reason true.
These are clearly an exceptional collection of items. Most of our everyday knowledge is neither incorrigible, self-verifying, or of any of the other types. Descartes has a difficult problem of getting from his collection of weird items to the rest of our supposed knowledge.
Michael Williams comments on a crucial underhand move in Bernard Williams' portrayal of Descartes' project. It is not the purity of the project, its lack of resource restrictions etc., that give it the shape it reveals — one might pursue physics, say, by building enormous particle colliders that encircle the earth. Instead, one turns aside to worry about malign demons or brains in a vat because of a stipulation that error must be avoided, not in normal investigations but in a special philosophical path from experiential knowledge to knowledge of the external world. It is this taking for granted of the terms of the problem that gives sceptical hypotheses their bite — indeed, on those terms they are probably unanswerable since everything we have to appeal to is by hypothesis the same, demon or no demon.
Can we get certainty?
Leave aside a priori knowledge (logic, mathematics). We can see most self-verifying items as due to general features of assertion (the others are purely formal claims such as that something is true or something is possible) — so we do not get necessary truths, only items that we cannot coherently assert. There is a general reason to doubt that there are any incorrigible empirical items: X's being F and the judgment that X is F are always distinct. There is a logical gap between them. So there is a logical possibility that one exists without other. In the case of many first-person psychological reports it is certainly very difficult to see how one might reveal such a discrepancy, but that does not undermine the general logical point.
Looking more generally at items that people have taken to be certain, there is an impressive argument from past failure: science littered with error once taken as (sometimes certain) truth. But this is not the case with people in pain. One issue here: can we distinguish levels of knowledge in the appropriate way — commonsense vs. disciplinary? We can of course make a distinction, but is this any more than a cut along what in other respects is a continuum? One way this question has been debated relates to a possible parallel distinction between observational and theoretical terms. (There is also a need to observe where error actually arises in the history of science — we cannot assume that what goes for particle physics holds equally for chemistry [cf. Papineau].)
Traditional foundationalist epistemology has assumed a break between 'the given' and what we make of it in our scientific theorizing. It is likely to characterise the different levels rather differently — often it wants infallibility at the foundation, while it can allow that the superstructure may be fallible. Views that do not see a qualitative cut between ordinary experience and the further flights of scientific fancy are happier to relinquish aspirations to any form of certainty in the everyday.
Fallibilism
Diametrically opposed to the Cartesian pogramme is the general endorsement of fallibilism. But given what we have already seen about the logical interrelations of Kap and other notions, it may not come as a surprise that it is not particularly easy to present fallibilism straightforwardly. Popper, for instance, revels in the clashes with commonsense entailed by denying that we know the sun will rise tomorrow ("It is simply a fact that we are commonsensically certain that the sun will rise over London tomorrow. Yet we do not know it for certain. There are millions of possibilities which may prevent it. ...we all hope that the sun will continue to rise. ... But even a necessary hope is not objective knowledge, though it may dispose us to belief" (1972, pp. 97-8)).
One point that needs to be seen is that when we think of knowledge as justified belief we have to leave open the possibility that however well justified, however reliable the link between world and belief may be, it is possible that it is not knowledge after all because things are not as our best reasons give us to think them. And this possibility is not mere logical possibility; it is epistemic — for all we know, p might be false. This may seem to conflict with principle that if we are not sure of p, we shouldn't claim to know p (or maybe the stronger claim that if we are not sure that p then we don't know that p).
Here is one take on how best to see fallibilism (from Mark Owen Webb, Giving up on Certainty):
One picture of the function of epistemic principles is this: justification and knowledge are completely reducible to natural facts about knowers and their contexts, and so true epistemic principles are true descriptions of the conditions under which knowers are well-placed with respect to their subject matter. Here is another picture: justification and knowledge are, at least partly, irreducibly normative, and so epistemic principles are endorsements of strategies for acquiring information. To say that someone knows something, or is justified in believing something, is in part to endorse her cognitive state, or how she got into her cognitive state. The proper way to evaluate an epistemic principle, then, is to ask whether people should follow it. The principle of fallibilism is no different. This second picture of the function of epistemic principles, which we owe to Wittgenstein and Sellars, sees epistemic talk as a sub-practice of linguistic practice, concerned with the asking for, giving, and evaluating of reasons. The practice is a rule-governed language game. On this account of epistemic principles, for example, Cartesian foundationalist principles can be seen as a set of recommendations for which assertions to allow to be unchallenged and which must be defended if challenged. The claim that I exist is always immune to challenge, as are claims about the current contents of my consciousness. Further claims must be defended (if challenged) by being deductively derived from these immune claims. Further, any possible doubt, no matter how 'hyperbolic' counts as a legitimate challenge.
On the naturalistic picture of epistemic principles, fallibilism is probably untenable. But it fares much better on the normative understanding of epistemic principles; on that sort of view, fallibilism is best understood as a proposal for a normative propriety in epistemic language-games. The idea is that it is best if we treat each other in such a way that no statement is de jure immune from criticism; any claim, no matter whether logically necessary or obtained by privileged access, must be defended if it is met with a challenge that is itself within the rules of the game. We may be subjectively certain that Graham Priest is wrong when he issues a challenge to the law of non-contradiction, and the law of non-contradiction can be necessarily true (after all, it is), but we may not simply ignore what he has to say. The fact that he challenges the law of non-contradiction from premises that he himself is entitled to hold, by methods that we all endorse as reasonable, makes it the case that his challenge is a fair one we must be prepared to answer. It may be that there are claims we cannot be wrong about, but even about those it is best if we follow the fallibilist rule. The reason this rule binds us is not strictly entailed by any facts about us. The various attempts to make such underlying facts explicit have all collapsed into skepticism, or entailed that there are no necessary truths, or had some other bad consequence. Instead, we should realize that the fact that we are frequently wrong, and that we are sometimes wrong about the things about which we feel the most certainty, and that we are in general bad judges of the success of our own epistemic endeavors, is enough to show that we ought to be very cautious indeed. Just as a judge may not rule on a case in which he has an interest, because people in that situation are prone to bias, so we ought to behave as if we could be wrong in any particular case, because we are wrong in so many cases. We think it the best procedure in general if the particular judge is not permitted to preside in that case, even if he is in fact immune to bias. In the same way, we should think it right to treat every one of our beliefs as potentially in error, even if in fact we cannot be wrong. The question fallibilism answers is the question of what rules are best for knowledge-seeking as an endeavor. So fallibilism is best understood, not as a thesis about our abilities, but as a rule to guide our epistemic endeavors.
First-person approach.
Egocentric predicament derives from reflective and self-critical nature of the Cartesian programme: the enquirer's need to consider how beliefs are acquired.
But knowledge is a we-affair. Criticism likewise a collective product. Popper went to the other extreme: objective knowledge is an inhabitant of world 3. World 1 is the physical universe; world 2 the mental, psychological realm. World 3 is "the world of intelligibles, or of ideas in the objective sense; it is the world of possible objects of thoughts; the world of theories in themselves; and their logical relations; of arguments in themselves; and of problem situations in themselves" (1972, p. 154). We can speak of knowledge in the objective sense as knowledge without a knower, without a knowing subject. Mackie summarises Popper's view as involving
Mackie's clarifications (primarily his insistence on distinguishing between logical and historical denizens of world 3) leaves us with a "stress on understanding the logic of problem situations and on the importance of public, inter-personal, and multi-personal traditions of knowledge and critical enquiry" (p. 127), and so on public matters rather than what individuals happen to have in the heads.
We may note here that other people have explored non-first-person approaches to knowledge. E.J. Craig sees knowledge as "the state our informants need to be in, for us to avoid error, not the state that we ourselves need to be in" (Papineau). Papineau continues (reference omitted) "my objection to Craig's line would be that he is in effect focusing on the special case in which we succeed in avoiding error by acquiring beliefs from informants who succeed in avoiding error. I accept that this special case may well have been of primary significance in the historical development of the everyday concept of knowledge, in that worrying about your informants' reliability calls for rather less sophistication than worrying about your own reliability. But, even so, Craig's third-person focus seems to me to have the disadvantage of cutting the link with the traditional normative issue of what we should do to avoid error. That is, even if it is unfaithful to the history of the concept of knowledge to view the desire for good informants as a special case of a general desire to have good belief-forming processes, I would argue that the more general perspective I have adopted nevertheless has the advantage of showing how the concept of knowledge relates to familiar philosophical worries about knowledge."
Return to home page.
Ed Brandon, last revised 22 October 2001.