By Daniel A. Kaufman
On tap this week is one of the most widely read and influential essays in epistemology, written since the Second World War, Edmund Gettier’s “Is Justified True Belief Knowledge?”
In this remarkably short piece, Gettier succeeded in casting doubt on what had been one of the most widely accepted ideas in philosophy – that to know something is to have a belief that one can justify and which is true – a view that goes at least as far back as Plato’s Theaetetus and which has been called the “Tripartite Theory of Knowledge” (TTK).
Gettier’s case against TTK is made via counterexamples. Gettier presents us with cases, in which a person has a belief, P, for which he has justification and which is true, but which he nonetheless, does not know. His cases are ingenious and rely on just two assumptions, both of which seem uncontroversial: (a) one can be justified in believing something that turns out to be false; (b) justification is transitive, through valid inferences, such that if S is justified in believing P and Q follows logically from P, then S is justified in believing Q.
Gettier presents two cases which are needlessly complicated. As a result, others have come up with “Gettier-style” cases that are simpler and yet, serve as counterexamples in precisely the same way as the originals. You can easily check out the original Gettier cases in the article, linked above, but here are two simpler, Gettier-style cases that make the same point.
(I) Imagine that we are sitting together in a sports bar, watching the Wimbledon final between John McEnroe and Bjorn Borg (my age is showing), and suppose that we see John McEnroe win the match. As a result, we believe that John McEnroe has won Wimbledon and certainly are justified in that belief (if asked, we will say that we just watched him win, on TV). Also imagine, however, that this year, Wimbledon’s TV feed has gone down and that unbeknownst to us, we have been watching a rerun of last year’s match. Suppose also that John McEnroe has won this year’s Wimbledon as well. Under such circumstances it would seem that even though we have a true, justified belief that John McEnroe has won Wimbledon, it would be very strange to describe us as “knowing” it.
(II) Imagine that we are driving through the countryside and observe a lone sheep on a distant hilltop. As a result, we both believe that there is a sheep on the hilltop. But suppose that, in fact, what we have seen is not a real sheep, but rather, a mock-up of a sheep. Suppose, in addition, that there is another sheep on the hill, behind it, that we cannot see. Once again, we have a belief that we are justified in holding and which is true, but it would be awfully odd to say that we know there is a sheep on the hilltop.
These cases – as well as Gettier’s originals – reflect two very common and well-understood facts about us and our “epistemic condition”: that we may stumble upon the truth entirely by accident or even by way of misleading information; and that what may seem, by every measure, to be the best reason in the world for believing something, may in fact turn out to be no good reason at all.
As you might have expected, there was an explosion of activity in the epistemological community, in the wake of Gettier’s article. One prominent strategy in dealing with the Gettier cases is to add a defeasibility condition: S knows P, if S believes P, P is true, S is justified in believing P, and there are no defeating statements (i.e. no statements that would defeat S’s justification). (1) This would render Gettier style cases harmless for TTK, for in each case, there is some statement that – had the person known it at the time – would have defeated the justification for the relevant belief. The trouble with this is that it turns out that it is not as easy as one might think to define what counts as a defeating statement and furthermore, one can never be sure whether some further information might not come along that would defeat the defeater and render the original belief justified after all. (2) One realizes, in taking into account such considerations, that no belief could ever really be justified, on the defeasibility view, because no one could ever know whether there might be some potentially defeating statement of which one is not yet aware.
Another prominent strategy is to build a causal condition into our definition of knowledge, in the manner suggested by Alvin Goldman. (3) What this does is require that for S to know P, his belief that P must — in some “appropriate” sense — be causally connected to P. (p. 369) This proposal also avoids the Gettier counterexamples, because in each Gettier-style case, having the relevant true belief is lucky, rather than the result of being causally connected to the truth-maker. It’s not the sheep that makes me believe there is a sheep on the hill, but rather, a mock-up. That there is an actual sheep is sheer luck. It’s not McEnroe’s win this year that makes me believe that he won Wimbledon, but rather, a rerun of last year’s match, which – again, luckily — he also won.
The one downside – if it is a downside – to Goldman’s response is that it abandons a long and widely held view of justification: namely, that one’s justification for believing P ought to be some reason or reasoning that one can explain oneself. That is, justifications should consist of reasons in the believer’s head, so to speak. This raises a much larger issue regarding the very nature of justification itself, one that has become a major topic of epistemological investigation in its own right, under the name, “The Internalism/ Externalism” debate.
1. Keith Lehrer and Thomas Paxson, Jr., “Knowledge: Undefeated Justified True Belief,”The Journal of Philosophy, Vol. 66, No. 8 (Apr. 24, 1969), pp. 225-237.
2. This is the idea behind the “Tom Grabit” case, in Lehrer and Paxson (1969).
3. Alvin I. Goldman, “A Causal Theory of Knowing,” The Journal of Philosophy, Vol. 64, No. 12 (Jun 22, 1967), pp. 357-372.