From the archives: Gary Gutting on the limits of philosophical argument

This is a post that’s sort of important to me in terms of my personal history, because it was written in December 2010, shortly after I had announced I was leaving the philosophy program at Notre Dame. It was originally part one of a three-part series; let me know in the comments if you’d like to have the rest to discuss later.

In my post on leaving philosophy, I said that “I think philosophy gets even fewer real results than the meager results that philosophers have sometimes claimed,” linking to the Amazon page for What Philosophers Know, by Notre Dame professor Gary Gutting.

As an explanation for my comment, I’m going to do a three-part blog review of Gutting’s book. The first part of my review will correspond to part I of the book, which is subtitled “The limits of philosophical argument.” Gutting conveys the main point of the section when he says:

How often have we heard (or told others) that Quine refuted the analytic-synthetic distinction, that Kripke proved that there are necessary a priori truths, and that Gettier showed that knowledge cannot be defined as justified true belief? But, although I entirely agree that Quine, Kripke, and Gettier have achieved something of philosophical importance, a careful reading of their exemplary texts does not reveal any decisive arguments for the conclusions they are said to have established.

I think Gutting’s selection of the three claims (about Quine, Kripke, and Gettier) represent a nice variety, in terms of how plausible the claims are. I don’t think I’ve ever met anyone who thought Quine refuted the analytic-synthetic distinction, and based on the PhilPapers survey, the analytic-synthetic distinction seems to be alive and well. With Kripke, on the other hand, I do get the impression that a lot of philosophers think Kripke’s arguments are decisive, though I’ve personally never found them convincing. With Gettier, though, his critique of the JTB analysis of knowledge seems to me about as good a candidate for a conclusive philosophical argument as there could be. (Non-philosophers: don’t take my word for it. The Gettier paper is short and easy to read, so go read it!)

For this reason, I find the discussion of Gettier especially interesting, and I actually think Gutting makes a good point:

If I say that all women are under six feet tall, then the fact that a women is six-foot-four refutes my claim. But the force of the argument depends on the obviousness of the counterexample. In just what sense are the Gettier cases obvious? I suggest we can fruitfully judge the obviousness of a claim by the epistemic price of refusing to accept it. If I’m presented with a woman who certainly looks well over six feet and whose height has just been measured (by competent judges) as over six feet, denying her height requires me to deny the evidence of my own senses and the validity of a measurement I have every reason to think is reliable. To maintain my claim that there are no women over six feet will involve me in a cascade of epistemic absurdities that makes holding on to my claim a common-sense impossibility.

Gettier counterexamples do not have this sort of obviousness.

Gutting reinforces his point by giving examples of philosophers who’ve defended the justified true belief analysis of knowledge in spite of Gettier, and by noting an x-phi study showing that “non-philosophers are far from unanimous regarding Gettier intuitions.”

I think all of that is right, and I’d add that not only is Gettier’s argument not as decisive as our evidence that there are women over six feet tall, it’s hardly as decisive as the evidence for the major findings of 20th-century science. I say this in spite of the fact that it still seems to me that the Gettier cases are cases of justified true belief without knowledge. The distinction here is between what seems true and what’s genuinely obvious. I know in my past thinking about philosophy, I haven’t always paid enough attention to that distinction, and I suspect many philosophers are guilty of the same failing.

I find Gutting less convincing, though, when he claims that philosophical reflection on the work of Quine, Kripke, and Gettier has generated some knowledge other than what those philosophers claimed. Gutting claims, for example, that Quine has showed us that the analytic-synthetic distinction cannot be defined in non-modal terms, and that in most cases “justified true belief” is the correct definition of knowledge.

An obvious worry, which (unless I’ve missed something) Gutting doesn’t address is this: if Gettier’s arguments aren’t decisive, what are the chances that there will be better reasons for accepting claims like these? If we use the standard “to deny it involves as much absurdity as denying that there are women over six feet tall,” what are the chances that Gutting has reasons his claims that meets that standard?

Indeed, I can think of specific reasons to be skeptical. Maybe no one has given a good definition of the analytic-synthetic distinction in non-modal terms, but I think that’s at best weak evidence that no such definition could be given. And it seems confused to talk about a definition working in most cases: if you go around trying to determine whether things are human by checking whether they’re featherless bipeds, you’ll get the right answer most of the time, but it would be a mistake to say “the featherless biped definition of humanity works in most cases.”

Now, I don’t claim these criticisms are decisive rebuttals to what Gutting says about the analytic-synthetic distinction, or what he says about knowledge. I do think it’s clear, though, that Gutting’s reasons for his claims aren’t any more decisive than Gettier’s reasons for denying that justified true belief is knowledge. And I don’t think Gutting has shown that there are very many philosophical claims that are well-established in any sense, or that there are any philosophical claims with truly decisive arguments in their favor.

  • Kevin

    I have a question. WTF is wrong with philosophers thinking that choosing a definition is an objective process? It just seems so bizarre to me. Someone goes I am going to define X as meaning Y and then they go, well in this situation Y happens but I don’t think that the term applies…WTF?! That’s what a definition is, it specifies the conditions for when it applies or not and that case fits the set criteria so it does. It’s just an arbitrary label that we came up with so there is no truth value involved. You can see this happening when it comes to the Gettier cases, when someone is talking about morality, etc. It’s almost like philosophers don’t know how to define words.

    • http://homeschoolingphysicist.blogspot.com PhysicistDave

      Kevin,

      I think the answer is that we have a more-or-less clear idea of what we mean by “knowledge” which both involves denying that Gettier cases count as knowledge and also involves viewing knowledge as justified true belief. I.e., it is not so much an issue of the “right” definition as of showing that we have intuitions that turn out to be mutually contradictory.

      Contrary to what math textbooks and math teachers tend to teach, this process does actually occur in the development of mathematics: e.g., it took a long while to develop the “correct” definition of limit, where “correct” means capturing most of our intuitions, useful in proving interesting results, and, of course, non-contradictory.

      That being said, I’m not sure Gettier was actually useful in those ways – i.e., I’m inclined to think one can wiggle out of his paradox without great pains, and I’m not convinced that anything important follows even if he is right. But, it’s late, and perhaps I should elaborate on that in another post.

      Dave Miller in Sacramento

      • Kevin

        Dave,
        “I think the answer is that we have a more-or-less clear idea of what we mean by “knowledge” which both involves denying that Gettier cases count as knowledge and also involves viewing knowledge as justified true belief. I.e., it is not so much an issue of the “right” definition as of showing that we have intuitions that turn out to be mutually contradictory.”

        But the Gettier cases show that we don’t have a clear idea of what we mean by knowledge. That is the point. One person could say that the person(TE) had knowledge, while Gettier was trying to show that he didn’t. What I am criticizing is the philosopher’s tendency to say to the other person “No, that person(TE) didn’t have knowledge.” Well, we defined knowledge as justified true belief, the belief was true and justified, so he(TE) had knowledge. So that’s what has been defined as knowledge, and the philosopher’s response is “That’s not what knowledge really means,” as if finding the meaning of a word is an objective process.

        For example, when it comes to morality, some people say that morality is prescriptive, some people don’t, some people believe that morality entails ultimate justice, some people don’t, etc. So someone comes along and says that when they use the term morality, they mean X, and the other people dismiss their view is valid because they have defined the term differently than them saying “that’s not what morality means” or “why assume morality means X”, not simply because the term doesn’t lead to anything useful is or is contradictory.
        *(TE) = Person in Thought Experiment

        “Contrary to what math textbooks and math teachers tend to teach, this process does actually occur in the development of mathematics: e.g., it took a long while to develop the “correct” definition of limit, where “correct” means capturing most of our intuitions, useful in proving interesting results, and, of course, non-contradictory.”

        If the philosopher wants to say that he(TE) didn’t have “glitzpin” (h/t to Verbose Stoic), which means “true justified true belief” and that’s what’s really important, then fine, but they don’t do that. And I agree that some words here need to be defined more rigorously (e.g. justified), but that doesn’t mean that if someone defines it differently that they are somehow objectively wrong. It could just as well be the case that they are trying to capture a different phenomenon than you (as it is the case with morality, knowledge, etc.)

        • PhysicistDave

          Kevin,

          I see your point, but I think you are misunderstanding what Gettier and at least some philosophers are doing: I don’t think they are trying to point out that people have the “wrong” definition so much as saying, in effect, “Your idea of what what knowledge is and how it works is rather confused, and this example shows that, and this should puzzle you and cause you to think things through more carefully.”

          Where I disagree with many philosophers on Gettier cases is that I do not find them that puzzling. First, no one has trouble understanding the setup of Gettier cases: things like this can and do happen. Second, I do not think understanding these cases produces any insight about how humans learn things about the real world or exposes any confusion in what most of us think we understand about how people learn about the real world: i.e., mulling over Gettier cases does not seem to me productive in learning more about human beings or reality in general. And, finally, it seems to me easy enough to “escape” from Gettier paradoxes: you could say that the guy who thought he was “justified” in his belief turned out not to be justified when one understands the broader context, or you could say that “true knowledge” requires true beliefs about the justifying conditions, or whatever. Philosophers have of course considered such escape routes, but I think it is fruitless to worry over which is the “right” escape route: it does not really matter – it is all just verbiage, and nothing in reality seems to hinge on which you choose.

          On the other hand, there are issues raised by philosophers over the centuries – the mind-body problem, the problem of free will – that really do point to matters on which we truly are ignorant or confused: I know of no neuroscientist who sincerely claims that we know exactly how the brain produces consciousness, for example. Even on these issues, I am skeptical that philosophers have come up with convincing answers, but they surely have pointed to questions to which humans do not know the answers.

          Even on these issues, do we need philosophers to point this out? Well, I don’t know: competent scientists (or historians or mathematicians) are pretty good at saying “I don’t know the answer to this, but I’d really like to find out!” But, I suppose that you can make the case that sometimes outsides (i.e., philosophers) can see this better than insiders.

          Anyway, I think you are making a mistake on focusing on “definition”: perhaps philosophers say they are talking about “definitions,” but I think a charitable interpretation is that their real purpose is simply to point out how people are unclear, ignorant, or confused on some particular matter. You want to object that so many philosophers claim to be masters of clear thought and expression, so why do they say they are talking about “definitions” when they are really talking about something else? Well… claiming to be good at something that one is really horrible at is hardly a failing limited to philosophers!

          Dave

          • Kevin

            “I see your point, but I think you are misunderstanding what Gettier and at least some philosophers are doing”

            Sorry, I didn’t mean to imply that this is what Gettier was trying to accomplish. It would be with those who use Gettier cases in such a way that I was criticizing. I think we are mostly in agreement and I think you hit the nail on the head of where my criticism is when you said:

            “But, a fair number of philosophers, and, even more, guys on the Web who know very little philosophy but nonetheless view themselves as brave defenders of philosophy, insist that such conceptual nitpicking is of enormous importance, that it somehow has the authority to dictate to people who really do uncover truths about the real world, etc.” and “but I think it is fruitless to worry over which is the “right” escape route: it does not really matter – it is all just verbiage, and nothing in reality seems to hinge on which you choose.”

    • http://verbosestoic.wordpress.com Verbose Stoic

      Well, though, philosophers aren’t interested in “definitions” as you put it, but in concepts, and it is indeed the case that whether the concept “knowledge” is picked out by “justified true belief” is a statement with a truth value. Essentially, defining as you put it is like saying that 1 + 1 = 3 because you can get everyone to agree that that’s the case, while what philosophers are doing is saying that we know that 1 + 1 = 2 because that’s what it really means to be the numbers 1 and 2 and =, no matter what symbols you use to represent it and no matter what base you’re in. To me, defining as you put is basically saying that you can use the word “knowledge” or the word “glitzpin” to refer to the same concept and no one really will care, but the philosopher will point out that no matter what you’re using to denote it you’re still trying to point at the same thing, and we can figure out in various ways what that thing is, just like you can use the word “tree” and the word “pento” to refer to a tree (if you can get everyone to agree to use the second word, of course) which we can then figure out the various qualities of.

      Thus, the Gettier problems attack the thing pointed to by the word knowledge, which in this case is the concept of knowledge, and so you can’t simply dismiss them by definition, any more than you could dismiss someone saying that if you define a tree as a thing that has leaves we can point to bushes and say that they also have leaves but really don’t seem to be trees. Or that evergreen trees don’t have leaves and yet seem to be trees. Therefore, your definition doesn’t pick out the instances of the concept properly and so you’d need to change it.

      • Kevin

        Verbose Stoic,
        “Well, though, philosophers aren’t interested in “definitions” as you put it, but in concepts, and it is indeed the case that whether the concept “knowledge” is picked out by “justified true belief” is a statement with a truth value.”

        What do you mean by knowledge? That’s the whole point, it hinges on an arbitrary definition. The process of defining a word is subjective because we can each define it differently. You can cite that you want cases A, B, C to count as knowledge and X, Y, Z to not and I can agree with you that that’s what you want your term to cover and see if your description does so accurately, but it doesn’t make sense for me to object, as philosophers do, that that’s not what knowledge means by using different test cases for A,B, etc. So, you conveyed a certain concept using the test cases and then the philosopher objects by pointing to a different concept using different test cases, and then uses that to say you didn’t get the definition right. Why is that acceptable?

        • http://verbosestoic.wordpress.com Verbose Stoic

          So, you conveyed a certain concept using the test cases and then the philosopher objects by pointing to a different concept using different test cases, and then uses that to say you didn’t get the definition right. Why is that acceptable?

          Because in the cases where philosophers object using that, it’s because we are assuming that everyone was, in fact, talking about the same concept. To then decide that you’re just talking about different concepts isn’t at all helpful, and would also raise questions about whether you’re using the RIGHT concept in the discussion, or if you’re equivocating by talking about the issues as if you were using the concept that the philosopher is saying doesn’t pass these other test cases and then when challenged on that retreating to a claim that you were talking about a different concept.

          We can see this with morality. If you want to claim to have an objective morality, you can’t simply make it a matter of stipulation, at least not if you want to be able to criticize someone using a different definition for their actions based on their immorality. The choice you run into is: either you consider it objective and therefore you can criticize other conceptions, but then you have to be able to demonstrate that you can cover all reasonable cases, or else you claim that we can have all of our own conceptions, but then any criticisms you make can be dismissed as us, literally, talking about two different things.

          The same thing applies here with the Gettier problems. We used examples just like the ones that Gettier brings up to come up with the idea that knowledge was justified true belief, so it isn’t really kosher to say that in those cases we can just ignore it because we have the definition we like. But, as has been stated, there are a number of ways to deal with Gettier problems.

          • Kevin

            “Because in the cases where philosophers object using that, it’s because we are assuming that everyone was, in fact, talking about the same concept.”

            It’s fairly obvious we’re not talking about the same concept. For example, take the example of utilitarianism and a common objection: utility monsters. Whenever someone uses this objection, they are saying that the term moral can’t be applied to a system that would conclude that giving a greater amount of resources to utility monsters because that would go against our “intuitions”. However, as a utilitarian, I say that they do in fact exist to a certain extent and I already consider such acts as moral (think of medical care expenses that exceed the lifetime earnings of the patient). Clearly, we are talking about two different concepts and using this test case as an objection is just disingenuous.

            “To then decide that you’re just talking about different concepts isn’t at all helpful, and would also raise questions about whether you’re using the RIGHT concept in the discussion”

            I have no idea what you mean when you say the “RIGHT” concept. We’re just talking about how we are defining the terms in a particular question. What is the ‘moral’ action in X situation? Did the guy in the Gettier case have ‘knowledge’? Its obvious people define these terms differently (because they give different answers to different test cases) so they are talking about different concepts. My whole point is that there is no “RIGHT” concept.

            “The choice you run into is: either you consider it objective and therefore you can criticize other conceptions, but then you have to be able to demonstrate that you can cover all reasonable cases, or else you claim that we can have all of our own conceptions, but then any criticisms you make can be dismissed as us, literally, talking about two different things”

            First, something being objective doesn’t mean that it covers all of the test cases since the answers to the test cases are simply the result of intuitions that aren’t objective. This is the point; there is no RIGHT concept or answer when talking about these test cases. Second, I do claim that we all have our own conceptions (not necessarily unique of course) of the term “morality” (since definitions are arbitrary) but that doesn’t mean that all criticisms can be dismissed. For example, I don’t adhere to Divine Command Theory, but I can’t just dismiss their criticisms that I am disobeying God’s commands (if such a God exists). I would just say that I don’t care. Similarly, someone could say that they don’t care about the welfare of others. If this criticism is acceptable to you (though I doubt it would be to many), then you have no obligation to adhere to utilitarianism.

            “We used examples just like the ones that Gettier brings up to come up with the idea that knowledge was justified true belief, so it isn’t really kosher to say that in those cases we can just ignore it because we have the definition we like.”

            This isn’t what I’m criticizing since you’ve stipulated we’d evaluate the test case under a particular definition. I’m criticizing those who bring up Gettier type cases to say that knowledge isn’t justified true belief simply because our intuitions say that the guy in the hypothetical doesn’t have knowledge while he had a true justified belief. This is what I’m saying is not kosher.

      • PhysicistDave

        Hi, VS.

        One of the issues is when is it productive to try to carefully hone our concepts by considering weird and unusual cases? After all, Aristotle pointed out that one should not expect any more precision in an intellectual field than is inherent in that field: it makes sense to expect more precise concepts in, say, electrical engineering than in fashion design.

        I’ll give you one of my favorite cases from physics, where I myself lived through the transformation of the standard definition of a concept:

        After Newton, “mass” referred to the constant of proportionality between a force and the resultant acceleration: the important thing was that this is a constant for any particular body.

        Except Einstein came along: it turns out that, in relativity, the factor relating force to acceleration is not constant.

        Well, what to do?

        The initial decision was to note that the momentum and the velocity are in the same direction, but the factor relating them depends on the speed. So, people started talking about “relativistic mass,” which depends on the speed. That’s how things were generally described when I started learning relativity.

        Now, people eventually came to realize that, for pedagogical reasons and often as the most efficient route of solving problems, it was useful to deal with four-dimensional vectors: then energy and momentum form a four-vector proportional to another four-vector known as the four-velocity. And, the factor relating these two vectors does not depend on the speed: it is indeed the plain old Newtonian mass.

        So, most physicist nowadays have abandoned talking about the speed-dependent “relativistic mass”: it’s back to plain old constant mass.

        But nothing in our actual understanding of the physical world has changed: no one that I know of thinks this was an important development. We merely decided that this way of speaking was a bit more convenient and understandable. I, and any competent physicist, can translate back and forth between the old and new concepts with no loss of meaning. If we had continued using the old concept, it would not really matter.

        Of course, there are other cases in physics – e.g., the interpretation of heat as being microscopic random energy – where the change in concept was very important and resulted from a dramatic increase in our knowledge of the real world.

        You see my point? Exploring concepts in a way that does not point to a new discovery about reality, or at least to an important gap in our understanding of reality, is a pretty bush-league activity. Of course, when our understanding of reality does change in a fundamental way, then a change in our concepts often necessarily follows.

        I’m afraid that, to me, the Gettier stuff falls into the former class, as does also, I think, most of what philosophers have concerned themselves with in the last century.

        Why should I care? Well… I don’t, that much. But, a fair number of philosophers, and, even more, guys on the Web who know very little philosophy but nonetheless view themselves as brave defenders of philosophy, insist that such conceptual nitpicking is of enormous importance, that it somehow has the authority to dictate to people who really do uncover truths about the real world, etc.

        So, it seems to me worth pointing out that these grandiose claims have no basis.

        Dave

        • http://verbosestoic.wordpress.com Verbose Stoic

          You see my point? Exploring concepts in a way that does not point to a new discovery about reality, or at least to an important gap in our understanding of reality, is a pretty bush-league activity. Of course, when our understanding of reality does change in a fundamental way, then a change in our concepts often necessarily follows.

          Well, the issue is if philosophy, as I believe, is primarily about conceptual analysis then it always really cares about refining the concepts, even if it isn’t that useful for a field like science that I think cares about instances more than concepts. Sure, some of the conceptual issues we come up with won’t matter to the instances we have, but some of them will, just as some of the discoveries about the instances will impact the concepts and some won’t. But it’s certainly not “bush-league” for a field that cares about the concepts to care about issues that may not impact what another field cares about, just like certain things in mathematics are mathematically interesting even though they don’t really impact science, or that certain discoveries and questions in science are still scientifically interesting even if they end up not changing things at all in people’s every day lives. Sometimes, they do, but when they don’t it doesn’t mean that it’s uninteresting, just that it’s uninteresting to that other field.

          Now, of course, you get people going over the line on both sides. Sometimes, philosophers do indeed try to trump science with conceptual issues that don’t really matter to science, but by the same token sometimes scientists wander into philosophical discussions using the rather crude concepts that work for their field and try to tell philosophers that their conceptual analysis is wrong or overdone, despite there being tons of work showing that that crude concept doesn’t work. The key is knowing what you’re doing at the time, which isn’t easy and can be argued for.

          • PhysicistDave

            VS,

            Sorry for the delay in replying.

            You wrote:

            Well, the issue is if philosophy, as I believe, is primarily about conceptual analysis then it always really cares about refining the concept…

            Well… certainly a lot of philosophers historically had a very different conception of philosophy: Aristotle, for example, went out and looked at facts in the real world.

            VS also wrote:

            Sure, some of the conceptual issues we come up with won’t matter to the instances we have, but some of them will…

            Well, I guess there are two questions here: The first, raised by Kevin above, is whether there really is such a thing as “concepts” that one can investigate and try to clear up, or is it all just a matter of arbitrary definitions of one’s terms, a rather silly thing to argue about?

            As Kevin and I discussed, I think it is sometimes the one, sometimes the other.

            Which leaves the question: on those occasions when it is fruitful to investigate concepts, who seems to be good at doing this? Physicists certainly have fruitfully modified their concepts when impelled to do so by empirical discoveries, as I discussed above. To outsiders, twentieth-century mathematics often seems to be little but a continuous refining of concepts (I know it’s more than that, but it is often hard to get behind the careful refining of concepts to the meat of the math).

            And, then, philosophy… well, where have philosophers really refined concepts in a way that has given us actual knowledge? Kevin and I were arguing that Gettier cases do not seem to provide such an example. I know enough philosophy that if there are a huge number of examples, I should be able to come up with quite a few.

            I can’t.

            And, as far as I can tell, neither can philosophers: each philosopher seems to think that he and his friends and mentors have enormously clarified some set of concepts, but they seem unable to convince most of their fellow philosophers, who adhere to different schools, methods, and gurus.

            That seems to me to be a problem: the most convincing witnesses against the value of philosophers are other philosophers.

            I’m happy to be corrected here: can you (or anyone) give an example of such conceptual clarification on anything at all on which most philosophers agree?

            Dave

        • hf

          In principle, Gettier could have led to a result such as Bayesian epistemology. I don’t think it did. Certainly Laplace invented much of this for practical reasons, and many of the rediscoveries happened for practical reasons.

  • Pingback: From the archives: Gary Gutting on convictions and scientific beliefs

  • Pingback: yellow october


CLOSE | X

HIDE | X