“The search for knowledge is like the search for true love. We live in a web of relationships, be it of propositions or people. Sometimes we are in a skeptical mood and we grasp for a solid base—a belief that we’re sure of or a friend or lover we can trust completely—but experience seems to admonish us, ‘all are fickle’; at such times the web can seem inscrutable. Then at other times we’re completely in the moment and the web is worldwide, and we’re secure in its interrelationships, confident in our position. We’re in a web all right—just sometimes we see ourselves as the spider other times as the fly.” – Raam Gokhale

“Truth may have been found but might never be known.” – Kedar Joshi

---

Scene: Kedar’s flat in Pune, India.

Players: Ram, an older philosopher, and Kedar, a younger philosopher (for other philosophical adventures of this duo, go to slumdogphilosopher.blogspot.com).

Ram: You know the quote from Kipling, “East is East and West is West and never the twain shall meet”? Do you think it applies to philosophy?

Kedar: It applies there more than anywhere else. Western philosophy demands the rigors of sound arguments. Eastern philosophy is virtually indistinguishable from religion.

Ram: But “NEVER the twain shall meet” is pretty radical. Don’t you think they must meet in some sense if both are to be labeled as philosophy? To compare them, to use the same word, ‘philosophy’, to describe them, they must have something in common. It’s not like say comparing farming with the Dewey-decimal system, for example.

Kedar: You seem to have a common ground in mind?

Ram: I do. I think it’s contained in the inscription at the Temple of the Delphic oracle, namely ‘Know Thyself’.

Kedar: Hmm. I think you’re right. Western philosophy exhorts ‘Know Thyself’ in order to know all other things. Eastern philosophy exhorts ‘Know Thyself’ in order to forget about all other things.

Ram: As usual, you’ve put things very pithily. In western philosophy, whether you’re dealing with ideas that have an external existence as in Plato or Berkeley or ideas only present in the mind as in Descartes, Locke, or Hume, ideas are known first and are the basis of knowing everything else. Know thyself in order to know all other things.

Kedar: How about your favorite philosopher, Wittgenstein? Doesn’t he argue that even self-knowledge is only possible through outward criteria, that there is no such thing as a private language? I for one would be only too happy to dismiss Wittgenstein as not a true philosopher.

Ram: Ah so that’s how philosophers trade barbs huh? They don’t directly insult each other; they insult each other’s intellectual heroes. I agree Wittgenstein doesn’t fit the paradigm too neatly but you must admit, even his linguistic analysis can be described as an exercise in ‘Knowing Thyself’—except in his case the thyself that you’re exhorted to know is the linguistic practices of your “form of life”, your community of speakers. Not ‘Know Thyself in order to know all other things’ but ‘know the practices of thy community in order not to muck up the enterprise of knowing other things’.

Kedar: At any rate, the later Wittgenstein is understood partly as attacking the phenomenalists who definitely fit into the ‘Know Thyself in order to know all other things’ camp. So I guess he’s a philosopher in the same sense as the statement, ‘Philosophy is crap’ is itself taken to express a philosophical position.

Ram: You put it more crudely than I would but I think we agree that in western philosophy, ‘Knowing Thyself’ is generally necessary in order to know other things. Wittgenstein might be an exception, but if he is, he is the exception that proves the rule.

Kedar: OK…perhaps. Now how about eastern philosophy? I take it you have in mind ‘Atman is Brahman’ and the doctrine of Maya from the Vedas when you say eastern philosophy exhorts, ‘Know Thyself in order to forget about all other things’.

Ram: Actually you said eastern philosophy exhorts ‘Know Thyself in order to forget about all other things’. I recently read The Tao of Physics and I would have to redescribe eastern philosophy as exhorting, ‘Know Thyself in order to know the subjective nature of all other things’. I think that’s a more accurate formulation of eastern philosophy.

Kedar: Interesting reformulation. It still fits the ‘Atman is Brahman’ and Maya doctrines of Hindu philosophy. And I can see how one might draw parallels with quantum mechanics where the observer plays an inseparable role in the process of observing. But can it really be taken to represent all eastern philosophy? I mean I don’t know much about Chinese and Japanese philosophy for instance.

Ram: Actually it better fits non-Hindu eastern philosophy, especially Taoism, much better because, unlike Hinduism, Taoism, in exhorting us to see the subjective nature of all other things, actively encourages us to understand all other things. Each separate thing is a Tao, composed of opposites like the yin-yang, with each piece containing the seeds of its opposite. The parallels with modern physics are clear: matter is energy; particles are waves. In Hinduism, a genuine interest in understanding individual things outside the self is sometimes lacking. This is certainly true in Hinduism’s emphasis on the big picture, of transcending the veil of Maya in order for the atman or self to merge with the Brahman or God. But even Hinduism holds that the duality of Shiva and Shakti, the male and female elements, like yin and yang from Taoism, or matter and energy, particles and waves from modern physics, can be seen in everything.

Kedar: OK it seems ‘the twain do meet’ in the exhortation, ‘Know Thyself’. That’s how we can recognize eastern philosophy and western philosophy, as different as they are, as philosophy. But they clearly differ in their recommendation of how best we can know ourselves.

Ram: That’s right and the methods they recommend are suited to their widely divergent views about the nature of reality—the ‘twains’ do meet but from then on steam in opposite directions. Western science has been making successively more accurate maps of the world but their first approximation was always common sense. Common sense is what all the theorizing must tie back to. And the methods of common sense--ordinary seeing is believing—are at the core of classical science just as ordinary introspection is at the heart of western philosophy. Eastern philosophy seeks to transcend common sense; common sense is Maya. The essence of reality is glimpsed in mystical visions. The extra-ordinary visions may explain the world of common sense but their basis is an ineffable contradiction: all things are a unity of opposites, opposites like the tendency to rest/to move, to sometimes exhibit ‘male’ sometimes ‘female’ characteristics, etc. We see these opposites first in our innermost natures. ‘Know thyself in order to know the subjective nature of all other things’.

Kedar: And quantum mechanics…I guess it just so happens reality has dictated western maps have eastern legends?

Ram: I couldn’t have said it better myself.

Kedar: But like you I consider myself more of a western philosopher. Must the west be so short-changed?

Ram: Not at all. Western philosophy’s great contribution is a healthy skepticism. While eastern philosophy goes on to prescribe how to go about acquiring self-knowledge be it through yogic meditation or contemplation of Zen paradoxes, western philosophy, facing a much simpler task as far as acquiring self-knowledge is concerned, goes on to either explicate how we then acquire knowledge of external things or failing that raises doubts about whether such knowledge is even possible.

Kedar: Ahem, with all this talk of knowledge, don’t you think we should define knowledge first?

Ram: You’re right. Surprisingly, as different as they are, both eastern and western philosophy, I think, would agree, at least as a first approximation, that knowledge is justified, true belief…just the things they believe and their methods of justification would be as different as night and day.

Kedar: I know ‘justified, true belief’ has been a definition of knowledge at least as far back as Plato but I am not sure it’s correct.

Ram: Well a lot of contemporary epistemologists think that the traditional definition needs at least a fourth condition to provide necessary and sufficient conditions for knowledge. And even the ones that don’t think that the traditional 3-condition definition is deficient, think that justification has to be radically reinterpreted.

Kedar: I’m not referring to that. I think knowledge is just true belief not justified true belief. Yet even with the simpler definition, I don’t think we ever have knowledge because we’re never sufficiently confident in our beliefs. Indeed the more beliefs we have, the less sure we are of any of them. Our uncertainty is even warranted on purely probabilistic grounds.

Ram: You and I have had this conversation before. As I recall, you think justification is merely a tool that helps us to have greater conviction in our beliefs but that it itself isn’t really necessary for knowledge.

Kedar: That’s right.

Ram: I think you have the ‘man-on-the-street’ on your side. I think if we asked an ordinary person to define knowledge, he would say true belief. Justification is usually the kind of thing only philosophers worry about.

Kedar: Is this another way that philosophers trade insults, by saying the ‘man-on-the-street’ would agree with you? Philosophers are a passive-aggressive lot aren’t they? You know me: I usually walk on the opposite side of the street as the ‘man-on-the-street’. So I don’t exactly find comfort in having the ‘man-on-the-street’ on my side.

Ram: Philosophy makes strange bedfellows. I usually find comfort in using words as they are used in everyday speech and, as yours is the more ‘everyday’ definition, let me try to see things from your perspective. First let me try to put a philosopher in your corner so you feel a little more secure. The philosopher Robert Nozick, like you, drops justification from his basic account of knowledge though like you he also thinks justification has a very important role to play in how we acquire knowledge. According to Nozick, knowledge is true belief except the belief has to be so ‘secure’ that it would vary with the truth of what’s believed in ‘close’ counterfactual situations.

Kedar: Could you elaborate?

Ram: Sure. I believe I’m talking to you. And in fact I AM talking to you. But for my belief to count as knowledge, Nozick would say my belief would have to be such that if I weren’t talking to you, I wouldn’t believe I was talking to you and in other not-too-farfetched situations where I would be talking to you, I would continue to believe I’m talking to you. As Nozick says it, my belief has to ‘track’ the truth in close counterfactual situations in order to count as knowledge.

Kedar: I don’t think I walk on the same side of the street as Mr. Nozick either. If I weren’t talking to you, I might still believe that I was; and there may be situations where I would be talking to you but for whatever reason I wouldn’t believe I was talking to you. Still that doesn’t change the fact that I know here and now that I am talking to you.

Ram: Nozick tries to capture this intuition by restricting his tracking conditions to close counterfactual situations. For example if I were drugged I could believe I’m talking to you even if I wasn’t talking to you. But Nozick would say that is not a close possible world. In ‘normal’ situations where I wouldn’t be talking to you, I would (undrugged) simply be talking to someone else or to no one at all. And in such circumstances, it might seem reasonable to require that I wouldn’t continue to believe I’m talking to you if my belief is to count as knowledge.

Kedar: But even if in close-counterfactual situations my belief doesn’t track the truth, even then I think I could still be said to know. What does it matter if I’m drugged and would think I’m talking to you even when I’m not. If I believe I’m talking to you when I AM talking to you, and my belief is caused by the usual perceptual cues, and not the drug in my system, then in those circumstances I’m right and have knowledge. For instance, I can give an accurate report of our conversation. It doesn’t matter that the drug would make me believe I’m talking to you even when I’m not. In that case I don’t have knowledge but that shouldn’t infect the case where the belief is properly caused.

Ram: I think I agree with you, though other people, in particular some epistemologists, might have different intuitions so we have to make the situation more precise. Imagine the drug that’s in my system only works when I don’t hear anything for a length of time. We may imagine the doctor has prescribed it to alleviate my desperate fear of being alone. Then in close counterfactual situations when I’m not talking with you, I would still believe I’m talking with you. My intuition is that, despite this weird drug, when I’m talking with you, I know I’m talking with you. We could imagine a doctor saying the drug only affects my judgment when I don’t hear any sounds for a length of time.

Nozick with his observations about ‘keeping the method fixed’ has some wiggle-room to deal with our intuitions, but ultimately I think he fails. But let’s not lose sight of what his tracking conditions were really meant to do. They were intended to rule out lucky guesses as not instances of knowledge; the intuition he was trying to capture is that if your belief is only accidentally true, it shouldn’t count as knowledge. The trouble is in our example, the belief I am talking to you is NOT accidentally true—it’s properly caused by the fact that I am talking to you; it’s just that even if it weren’t true, there would be another cause—namely the drug—that would make me believe I was talking to you; but the existence of this other cause doesn’t make my belief accidentally true when the right cause is the one that’s operating at the present time.

Kedar: My view of knowledge, as you know, permits even accidentally true beliefs or lucky guesses to count, provided they are firmly believed.

Ram: I was wondering when you were going to say that. I happen to think lucky guesses should be excluded, though not as Nozick does, so let me try to dissuade you with the following example: suppose someone has a dream he is going to win the lottery; the dream firmly convinces him it’s going to happen and so he buys a ticket; the ticket wins. You would say that though he wasn’t justified in believing he would win, he nevertheless knew he would win?

Kedar: Sure I would. Wouldn’t we in such circumstances say he JUST KNEW he was going to win?

Ram: Watch it! You’re in danger of appealing to the man-on-the-street again. The man-on-the-street’s intuitions can be notoriously slippery. Though in the lottery example, we could picture him agreeing to the claim, ‘the dreamer just knew he was going to win’, if we pressed him, the man-on-the-street could equally be brought to say the dreamer didn’t really know he was going to win, he just got lucky.

We’ve had this type of conversation before. The other day when we were talking about necessity in the Deccan Dugout, you stated there can be multiple, sometimes conflicting intuitions about the proper meaning of common words. But you said the philosopher has the right, the obligation, to select or define a technical meaning to suit his purpose. I submit to you, justified, true belief or some account of knowledge that precludes accidentally true beliefs is a more technical, philosophically more interesting definition of knowledge than just true belief, and so should be chosen by the philosopher over the latter.

Kedar: Well I have to admit just true belief as a definition of knowledge is not very philosophically interesting. For example, as you pointed out before, a man could be said to acquire a lot of knowledge on this definition simply by believing each pair of a series of contradictory statements. One of each pair has to be true and if he ‘hedges his bets’ by believing both he’s guaranteed to have knowledge. This is clearly a pretty ridiculous way of acquiring knowledge. Perhaps when I proposed the ‘true-belief’ definition, I was guilty of appealing to the man-on-the-street’s lesser intuitions.

So it seems beliefs have to be non-accidentally true in order to count as knowledge. How do we flush out exactly the conditions necessary to rule out accidentally true beliefs?

Ram: Certainly if someone as illustrious as Robert Nozick, the winner of the Ralph Waldo Emerson Prize, has failed, we should be on guard.

Kedar: I’m not sure whether you’re being facetious or passive-aggressive again, but we definitely should be on guard. For example, your justification condition—we would want to say most people have knowledge about the content of their perceptual beliefs without having a justification for them. We know the causes that make most perceptual beliefs true—i.e. their reliability in conditions of good lighting, the perceiver not being under the influence of drugs, etc.—so we would say such people are justified but these people may not be aware all the premises that go into their justification.

Ram: You bring up an interesting point. Could someone be justified in his knowledge claim only from the outside and still be said to know?

Kedar: What do you mean ‘from the outside’?

Ram: I mean inaccessible to the putative knower. This is the internalism/externalism distinction one encounters in contemporary epistemology. We from the outside, external perspective know why the man-on-the-street’s perceptual beliefs are justified but from the internal perspective, based solely on what the man is aware of, the man might have no justification—he simply believes what he sees. The question is do we want to require the man be aware of the full justification of his beliefs in order to know them or do we allow external justifications to support his knowledge claims?

Kedar: I’ve re-girded myself against the man-on-the-street. My intuitions about justifications are that it’s an internal thing, so I would say if the man-on-the-street doesn’t have the full justifications for his perceptual beliefs internally, so much the worse for him—he doesn’t know.

Ram: Now we have to be careful what we require as full justification. For consider the following example. A man in an empty public square sees a prominent clock in a country known for punctuality and precision…

Kedar: Not India, I take it.

Ram: Ha, Ha, OK not India. The man notices that the clock reads 5:00 pm. He thinks to himself, ‘clocks around here generally tell the correct time’, looks at the position of the sun in the sky and judges that it’s about 5:00 pm and takes similar other measures to justify his belief that it is in fact 5:00 pm. Suppose it is in fact 5:00 pm but the clock he is looking at stopped working at 5:00 am. A philosopher named Edmund Gettier used examples like this to argue that knowledge can’t simply be justified, true belief because the man seems to meet all three conditions but we don’t want to say he knows it’s 5:00 pm.

Kedar: Interesting example. My first inclination is to maintain that knowledge can be justified, true belief. It’s just that the man’s justification is not a full, complete justification.

Ram: I had the same reaction when I first heard this example. But remember: you’re committed to justification being internal, that the man be fully aware of each premise in his justification. What would you have him do? Wait a few minutes to see if the clock is running? We could suppose the clock did coincidentally restart at 5:00 pm so it seems to be running. Still we don’t want to say that the man who relies on such an on-off clock knows what time it is. Do we want to further require he examine the inner workings of the clock? Again since the clock has restarted, inner workings might yield no clue that in fact it’s reading 5:00 am and that therefore the man’s belief that it’s 5:00 pm is only accidentally true.

In general any belief inductively justified can be false. That’s the nature of inductive justification. All we have to do to cook up a Gettieresque example is to imagine a scenario in which the inductively justified belief would be false but suppose the belief coincidentally is true. Then you have truth, belief, justification—just not knowledge.

Kedar: Extending your example to the n-th degree, we could imagine a full justification would require a justification for induction, which the history of philosophy has taught us is a losing battle.

Ram: Fortunately it’s not as bad as that. In the case of inductive justification you could argue that a justification only has to make it likely that the conclusion-belief is true. Against Gettieresque examples, the only thing that’s required is that there be no genuine defeaters of that justification as there are in the clock example.

Kedar: What’s a genuine defeater?

Ram: Well, a defeater of a person’s justification for a belief is a true proposition such that if the person believed this proposition, he would no longer be justified in holding the belief in question. The proposition ‘the clock stopped working at 5:00 am’ is a defeater, a genuine defeater, in the earlier example because if the man believed it, he would no longer be justified in believing it’s 5:00 pm.

Kedar: I guess to understand why you say genuine defeater, I have to know what’s a non-genuine defeater.

Ram: A non-genuine defeater, or a misleading defeater as it’s known in the literature, is a proposition that is a defeater in the sense that were the putative knower to believe it, he would no longer be justified. But it is a misleading defeater in the sense that its power to defeat is dependent on a false proposition. That at least is how my graduate school professor, Peter Klein, a defeasibility theorist, put it.

Kedar: Can you give an example?

Ram: We already had one. The drug that caused conversations to be imagined when no conversation was going on is an example of a misleading defeater. Imagine that I was unknowingly given this drug. This proposition is a defeater of my justification for believing that I’m talking to you. It is a defeater because if I believed it, I would no longer be justified in believing I was talking to you. Yet to some it seems like a misleading defeater because its power to defeat is dependent on the false proposition that the drug in my system is active right now. Remember we said the drug doesn’t operate so long as I hear sounds.

Kedar: OK—I think I have some sense of defeasibility theories of knowledge. And in Nozick I got some flavor of a different type of theory. I know we couldn’t have covered every epistemological theory but did we at least hit all the major classifications?

Ram: Let’s see…Nozick was an example of reliabilism. We talked about defeasibility. I guess the only other major strand is the causal theory of knowledge.

Kedar: That sounds interesting…maybe just the kind of theory I can espouse. I don’t know exactly what a causal theory of knowledge is but it sounds like it could fit our intuitions in the drug case. There we said I know I am talking to you because my belief is caused by the facts that make it true.

Ram: A causal theory has strong intuitive appeal. We want to say what’s true caused us to know it’s true. It’s the basis of our most common beliefs, namely perceptual beliefs. Seeing is believing because the thing we believe plays a causal role in forming our belief.

Kedar: And a causal theory would handle our Gettieresque clock example. The man doesn’t know it’s 5:00 pm because there is a causal disconnect between it actually being 5:00 pm and his belief that it’s 5:00 pm.

Ram: That’s right, but…

Kedar: I just knew there would be a but. Did you ever think, philosophy is crap because it’s full of butts?

Ram: Very funny. Seriously though, a causal theory of knowledge faces some challenges distinguishing causal connections of the right sort from causal connections that don’t result in knowledge.

Kedar: Could you give an example of the wrong sort of causal connection?

Ram: Certainly. Let’s modify our drugged conversation example. Suppose the drug administered unknown to me makes me imagine a conversation, perhaps a very flattering conversation, when someone is talking to me. I think we would say I don’t know you’re talking to me though my belief that you’re talking to me is caused by your talking to me.

Kedar: It’s funny how our strongest intuitions fall prey to such easy to dream up counterexamples.

Ram: They’re not that easy to dream up—I’ve just heard them or their kind before. For example another criticism of the causal theory I’ve heard is that it’s unable to handle deductive knowledge like our knowledge of the truths of mathematics. Numbers don’t cause anything because they’re simply logical constructs.

Kedar: Remember, I happen to be a Platonist about truths of mathematics so I can believe that somehow they cause our beliefs. They cause our beliefs because our mind directly ‘grasps’ them.

Ram: Well let’s just say numbers pose difficulties for most ‘normal’ causal theorists of knowledge.

Kedar: As usual you’re being very cagey and passive-aggressively insulting. We’ve surveyed the major epistemological theories but you still haven’t said which one you prefer.

Ram: If I am being cagey it’s not from any deviousness. Remember, like Wittgenstein, I’m an ordinary language philosopher. To borrow Wittgenstein’s metaphor, maybe knowledge is like a family resemblance: we see a family’s photographs and the faces all seem to resemble one another but we may be hard-pressed to find a single feature that is common to them all; knowledge may be like that, a concept that has many intuitions running through it without one being common to them all. For example, for knowledge we have intuitions that it must be reliable, non-accidental, caused by the thing known, be supported by a justification. It wouldn’t surprise me if all these intuitions couldn’t be brought together under one rubric. Maybe all philosophically interesting concepts are like that, making our attempts to find necessary and sufficient conditions for them ultimately doomed to failure.

Kedar: Don’t be so pessimistic. After all, as different as eastern and western philosophies are, you found a common thread running through all of them—know thyself.

Ram: I think I got lucky. But you may have a point. Maybe I got lucky because philosophy is a more technical word than knowledge. Maybe philosophers should only focus on words that have already been lifted out of the confused din of common discourse. Maybe epistemologists would do better to focus on concepts like evidence or justification which are already more technical than knowledge.

Though, wouldn’t it be funny if the word philosophy was the only philosophically interesting word that permits a philosophically satisfying definition?

Kedar: I think that would be funny only to a philosopher. I choose to believe that concepts like justification and evidence would permit philosophically interesting definitions. Let’s talk about them next.

Ram: Wait a minute. I see by the clock outside it’s 5:00 pm. I think I better be getting back. We can talk about justification and evidence next time.

Kedar: OK. Meanwhile I’ll take the imagined conversation drug and think about what you and I might say.