Let's go back to Dretske's Gadwall duck case. Suppose you're walking through a marsh that in fact contains a lot of look-alike grebes. But you don't know this; in fact you never heard of a grebe. You look off to your left and you see something that looks like a Gadwall duck, so you believe "That's a Gadwall duck." In fact, it is a Gadwall duck. It's one of the few ducks in the marsh. But you're just lucky. If you had looked at a grebe instead, you would have believed that it was a Gadwall duck.

Now in this case, it seems that your belief that the bird is a Gadwall duck is justified or reasonable. After all, it looks like a Gadwall duck; and you've never heard about these grebes. And your belief is also true. But we're reluctant to say that you know that the bird is a Gadwall duck.

Cases of this sort are known as Gettier cases, after the philosopher Edmund Gettier. Before Gettier, philosophers thought that knowledge was equivalent to justified true belief. That is:

You know that P iff:

(i) P is true,

(ii) you believe that P,

and (iii) you are justified in believing P (you have good evidence for P).

But then Gettier came along and presented examples in which the subject has a justified true belief which, intuitively, fails to count as knowledge. The Gadwall duck/grebe case we just discussed is one such example. Here's another:

You have a justified belief that someone in your office owns a Ford. And as it happens it's true that someone in your office owns a Ford. However, your evidence for your belief all concerns Nogot, who as it turns out owns no Ford. Your belief that someone in the office owns a Ford is true because someone else in the office owns a Ford. Call this guy Haveit. Since all your evidence concerns Nogot and not Haveit, it seems, intuitively, that you don't know that someone in your office owns a Ford. So you don't know, even though you have a justified belief that someone owns a Ford, and, as it turns out, this belief happens to be true.

These cases are counter-examples to the claim that justified true belief is sufficient for knowledge.

The Gettier Problem is to state what, in addition to or instead of justified true belief, is needed to have knowledge.

  1. Notice that the Gettier Problem only arises because we were trying to say that you could know that someone owns a Ford on the basis of evidence that falls short of certainty. If we instead said that knowledge requires absolutely certain or infallible evidence, then it would be clear why you're not in a position to know that someone owns a Ford. You don't have infallible evidence that someone owns a Ford.

    This solution would do the trick. It would make our account of knowledge immune to Gettier-type counterexamples. But it would also make it next to impossible to have knowledge. So it seems like a bit of overkill. Most philosophers who have attacked the Gettier Problem have tried to find some solution which still allows subjects to know things on the basis of fallible, defeasible evidence.

  2. One salient feature of the Nogot/Haveit Gettier case is that the reasoning that leads you to the belief that someone owns a Ford goes through a false step, namely the step where you believe that Nogot owns a Ford. So a possible solution to the Gettier Problem might be this: knowledge is justified true belief--where the reasoning your belief is based on doesn't proceed through any false steps.

    Philosophers initially thought this was a promising solution. But unfortunately, Rich Feldman described Gettier-like cases where your reasoning doesn't proceed through any false steps, but intuitively you still don't count as knowing. So the present solution doesn't get to the root of the problem.

    Feldman's case works like this:

    You see Nogot waxing a Ford, humming Ford ad jingles to himself, and so on. On that basis, you conclude that someone in your office is waxing a Ford, humming Ford ad jingles, and so on. This belief is true. On the basis of that belief, you conclude that someone in your office owns a Ford. As before, Haveit owns a Ford so this belief is true.

    Once again it seems like you have a justified true belief that someone owns a Ford, which fails to count as knowledge. In this case it does not look like your reasoning proceeded through any false steps.

    Another kind of Gettier case where you don't seem to reason through any false steps is discussed in the Rosenberg dialogue:

    You're in the meadow, and you see a rock which looks to you like a sheep. So you say to yourself "There's a sheep in the meadow." In fact there is a sheep in the meadow (behind the rock, where you can't see it).

    Here again you have a justified true belief that there is a sheep in the meadow, which fails to count as knowledge. In this case, your belief doesn't seem to be based on any reasoning at all.

    The problem in the sheep-in-the-meadow case is that you don't really see the sheep. You just think that you do. The fact that there really is a sheep in the meadow, which you don't see, seems to be just a gratuitous accident. It doesn't have anything to do with your belief or your evidence for your belief.

    Another example to the same effect:

    You look at a shelf. An evil neuroscientist has some electrodes wired up to your brain, and he causes you to have visual experiences as if there were a clock on the shelf. As it turns out, there really is a clock on the shelf. (But you would still be having clock-experiences even if there weren't.)

    In that case, it seems like you have a justified belief that there is a clock on the shelf, and it's true that there is a clock on the shelf, but you don't know that there is a clock there.

  3. Another popular solution to the Gettier Problem is to say that you know that P iff you truly believe that P and it's not an accident that you are right about P.

    The difficulty for this view is to explain what we mean by "it's being an accident that you are right about P." This is especially difficult because not every accident generates a Gettier case.

    Tommy comes across some evidence that his wife is sneaking around. She isn't at work when she's supposed to be, he finds some matches from a fancy nightclub in her car, and so on. When he asks his wife where she's been, she is evasive. This gives him some evidence for believing that his wife is having an affair. As a matter of fact, there is a simple explanation for all the evidence that Tommy has encountered: his wife is planning a surprise birthday party for him. But as it also turns out, his wife is having an affair, with her old boyfriend in Chicago. But she's very discreet about it, and so hasn't left any clues lying around.

    This is a case where Tommy has a true belief that his wife is unfaithful, and it's just an accident that his belief is true. So far, so good. According to the present proposal, Tommy doesn't know that his wife is unfaithful. And that seems intuitively to be the correct thing to say about this case.

    But now consider a second case.

    Tommy goes on a business trip to Chicago, checks into his hotel, and goes up to his room. By accident, he gets off the elevator on the wrong floor, and opens the wrong door. But the door does open, and there is his wife, in bed with her old boyfriend.

    In this case, too, Tommy has a true belief that his wife is being unfaithful to him. And in this case, too, it seems to be just an accident that Tommy has a true belief about this. But in this case, we do want to say that Tommy knows that his wife is being unfaithful.

    So the problem is to explain the difference between the kinds of "accidents" illustrated in the first case, which do block your true beliefs from counting as knowledge, and the kinds of "accidents"illustrated in the second case, which don't. This is difficult to do. (You may be able to explain the difference between these two particular examples that I've given. What is harder is to come up with an explanation of the differences between all the cases where your true beliefs don't count as knowledge, and all the cases where they do.)

  4. Another popular solution to the Gettier Problem is to say that you know that P iff you have a justified true belief that P, and there's no true information "out there" in the world that would defeat your justification for P, were you to learn of it.

    This sounds good at first. In the Nogot/Haveit case, there is the information that Nogot is merely pretending to own a Ford. If you learn that information, that would defeat your justification for believing that someone in the office owns a Ford. So according to the present proposal, even before you learn that information, the mere fact that it is "out there" in the world shows that your justified belief that someone owns a Ford can't count as knowledge (even though it happens to be true).

    In Rosenberg's sheep-in-the-meadow case, there is the information that what you're looking at is a rock. If you learn that information, that would defeat your justification for believing that there is a sheep in the meadow. So according to the present proposal, even before you learn that information, the mere fact that it is "out there" in the world shows that your justified belief that there is a sheep in the meadow can't count as knowledge (even though it happens to be true).

    In the Gadwall duck/grebe case we were discussing earlier, there is the information that the marsh you're walking through has a lot of grebes in it, and they look just like Gadwall ducks. If you learn that information, that would defeat your justification for believing that you're looking at a Gadwall duck. Hence, even if it's true that you're looking at a Gadwall duck, according to the present proposal you don't know that you are.

    So far, this all sounds good. The present proposal also explains why you can't have knowledge in another sort of case, first introduced by Gilbert Harman:

    Jill reads in the newspaper that the president of her country has been assassinated. In fact, this story is true. However, the president's associates have mounted a campaign to suppress the story, and they've been broadcasting false reports on all the television stations that the president is OK, the assassin actually only killed a bodyguard. Jill is blissfully unaware of all this misleading evidence. The newspaper she read happens to be the only news source that's reporting the true events. All of Jill's peers, on the other hand, have heard the misleading TV reports and aren't sure whether or not the president was really killed.

    Harman claims that this is another Gettier case: he says that Jill has a justified true belief that the president was assassinated, but she doesn't have knowledge, because there is all this misleading evidence abroad in her community, which she has only managed to avoid by sheer luck.

    So far, so good. However, the current proposal also runs into difficulties.

    Do we really want to say that if there's any potentially defeating piece of information out there, that's enough to block your justified true belief from counting as knowledge? Consider a variant of Harman's assassination case.

    The president's associates are sitting in the TV studio, saying into the microphone "No really the president is OK, it was somebody else who got killed."The fact that they are saying this is a potentially defeating piece of information. If you were to learn that the president's associates are saying this, it would defeat your justification for believing the president had been assassinated. But suppose that in this case, unlike the earlier case, the associates' speech never gets broadcast to the public. (Maybe the TV producers think it would be wrong to deceive the public in the way the associates want.) So all the newspapers and TV stations carry the correct report about the president's assassination.

    In this case, it seems like you should know. There is a potentially defeating piece of evidence out there (the fact that the associates are saying what they're saying). But it's so remote, nobody knows about it except the associates themselves and a few TV crew. Perhaps this case is hard to assess. But my own inclination is to say that in this case, you can know the president was assassinated. The information about the president's associates in this case should not block your justified true belief about the president from counting as knowledge.

    Here's another case, to make the same point.

    You see Tom Grabit hide a book underneath his jacket and sneak out of Widener Library. On the basis of this, you form the justified belief that Tom stole a library book. As it happens, your belief is true. However, unbeknownst to you, Tom's mother was going around today telling people that Tom was thousands of miles away, and that Tom's evil twin John was visiting Harvard. The fact that Tom's mother said this is a potentially defeating piece of evidence. If you were to learn of it, it would defeat your justification for believing that Tom stole the book. However, as it turns out, it really was Tom who stole the book. Tom has no twin brother and his mother is a compulsive liar.

    In this case, too, it seems like you should count as knowing that Tom stole the book. The testimony of a compulsive liar, locked up in an asylum somewhere, which you never hear, should not block your justified true belief about Tom from counting as knowledge.

    In both of the preceding two cases, it seemed intuitively like you should count as having knowledge. Yet in both cases, there was some potentially defeating piece of information "out there" in the world. So the present proposal would say that you don't have knowledge. That is a problem for this proposal.

    You may have noticed that, in the preceding two cases, although there is some potentially defeating piece of information "out there," you would only be misled by that information if you somehow managed to learn only part of the truth. The information about what the president's associates are saying would only defeat your justification if you were ignorant of the fact that they were doing a cover-up. The information about what Tom's mother said would only defeat your justification if you were ignorant of the fact that she was a compulsive liar. And so on. So you might be tempted to say:

    Hey, we can fix the present proposal. We'll say that you know that P iff you have a justified true belief that P, and there's no true information "out there" in the world that would defeat your justification for P, were you to learn of it--unless there's also some second piece of information that would counteract that defeater.

    If you thought of this, that's good, you're on your toes. That's very clever.

    But if you think harder, you'll realize it's not clever enough. For if we amend the proposal in that way, then the proposal becomes incapable of explaining any of the Gettier cases. For in a Gettier case, your belief always happens to be true, and that's a piece of information "out there" in the world that would counteract any evidence that tells against your belief.

    To illustrate this, consider the original Nogot/Haveit case. In that case, you have a justified belief that someone owns a Ford, and this belief happens to be true. Now, there is some information out there--the information that Nogot is only pretending to own a Ford--that would defeat your justification for believing that someone owns a Ford. But, since your belief that someone owns a Ford is true, there is also a second piece of information out there that would counteract that defeater. Namely, the information that Haveit does own a Ford. If you were to learn both pieces of information--that Nogot is only pretending but Haveit does own a Ford--then you would remain justified in believing that someone owns a Ford. So according to the amended proposal, you would count as having knowledge. But that's the wrong result. In the Nogot/Haveit case, intuitively, you don't know that someone owns a Ford.

    So it is difficult to explain why in some cases (the Nogot/Haveit case, the sheep-in-the-meadow case, the Gadwall duck/grebe case, Harman's original assassination case) the existence of some defeating evidence "out there" in the world blocks your justified belief from counting as knowledge, even though it is true; and why in other cases (the second version of the Harman case, the Tom Grabit case) it does not.

    Sigh. Back to the drawing board.

  5. Let's consider one more solution to the Gettier Problem. This solution says that you know that P iff you truly believe that P and you have evidence that P, and the fact that P is causally connected in the right way with your belief or your evidence. (Sometimes this solution omits the reference to "evidence" altogether, and just talks about causal connections between the fact that P and your belief.)

    This sounds promising. In the sheep-in-the-meadow case, the problem was that the real sheep played no role in causing your belief that there was a sheep in the meadow. That's why we want to say it's just an accident that you get things right. In the Nogot/Haveit case, Haveit's Ford-ownership likewise plays no role in causing your belief.

    When we consider questions of the form "Why is so-and-so's belief correct?" and questions of the form "Why did so-and-so act that way?" there are two kinds of answers we can give.

    One kind of answer is a causal answer. This is a case where there is some reason or causal explanation why the belief is true, or why the agent acted as he did, but these need not be reasons the agent is aware of.

    A second kind of answer is a rationalizing answer. This has to be some reason that the agent has for believing what he does, or acting in the way he does.

    To illustrate:

    (Sometimes when people talk about "rationalizations," they only mean to talk about good rationalizations. Here I'm using the notion of "rationalization" in a broader sense, to include both good and bad rationalizations.)

    The preceding two cases concerned rationalizations of someone's actions. Let's consider a case where something rationalizes someone's belief.

    As this last case brings out, one and the same thing might be both a cause of Kurt's belief and something that Kurt takes to be evidence for his belief. So there is no rule that if something is a causal reason why you believe what you do, it can't also play a role in rationalizing your belief.

    So far in this class we've been focusing on evidence the subject has, and how that evidence rationalizes the subject's belief, and the role that plays in enabling the subject to have knowledge.

    In response to the Gettier Problem, though, many philosophers started suggesting that we shift our attention to the things that cause our belief, instead of the evidence that rationalizes our belief. Maybe the causal factors are also important. (Some philosophers suggest that they're all that's important, and we can forget about the rationalizing factors.)

    The next few papers we look at in the class will be investigating this possibility.

[Theory of Knowledge] [Syllabus] [Notes and Handouts] [James Pryor] [Philosophy Links] [Philosophy Dept.]

URL: http://www.princeton.edu/~jimpryor/courses/epist/notes/gettier.html
Last updated: 10:09 AM Mon, Sep 30, 2002
Created and copyrighted by: James Pryor