Saturday, June 7, 2008

Hume's Value Naturalism

The Hume Society has chosen the topic of "naturalism" for the 2009 conference, and since a) I think I've got a line on naturalizing Hume's ethical theory, and especially since b) the conference is to be held in Halifax, Nova Scotia in August, my project this week is putting a submission together (not that that's an easy thing: if you check out their journal Hume Studies you'll see that these are indeed the preeminent Hume jocks, and a wonk I am not, although I've been a member for some years. But it's in Halifax so I have to give it a whack). So it's Saturday and I'm done with the kitchen and now I'll try blocking out the basic line here.
As most of my readers will know, Hume's best-known claim about ethical theory is that one cannot derive an "ought" from an "is." That is, descriptive propositions about the world do not as such contain any prescriptive meaning (it's important to remember that Hume is talking about propositions). Imperative prescriptions are motivated rather by "moral sentiments," non-cognitive feelings with which we have been endowed by nature, Hume knows not how. I think that Hume would be satisfied with Darwin's subsequent account of the etiology of ethical response. In fact Hume not only neatly flags the missing explanation, but asserts that the question is one for natural science, not philosophy (today I'm not going to provide any citations or passages, this is just off the top of my head). Part of Hume's bigger-picture agenda here is to debunk rationalist approaches to ethics, which the "schoolmen" took to involve ethics in transcendental metaphysics (Plato, by way of 17th century rationalism): Hume denies that ethical thinking is a branch of logic.
On my view, the most important thing to get about Hume is that he was not a skeptic, as is widely expounded. Rather Hume saw Cartesian skepticism as a pseudoproblem, and empiricism as the way out. Later the English phenomenalists of the early 20th century obscured this and now, with the aid of Wittgenstein, we are reconstructing Hume. Specifically Hume's is/ought distinction is frequently confounded with G. E. Moore's "naturalistic fallacy," which in fact is a very different sort of argument than Hume's. So if I can articulate the Humean critique of Cartesian skepticism, and show how Hume's is/ought distinction fits into that critique, this would help to nail down a persuasive account of Hume as a kind of naturalist about value.
Central to this is Hume's denial that it made any sense to speculate about an "external world" that might or might not be "corresponding" to the world of our perceptions. Experience cannot be distinguished from the world, and the world cannot be distinguished from experience. Thus there can be no question of "phenomenal properties," since any discussion of phenomenal properties is necessarily a discussion of experience, and any discussion of experience is necessarily a discussion of the world. There is no sense in which "my experience of blue" is different from "blue." To use Hume's language, mental contents consist of impressions and ideas, "impressions" being directly caused by interaction with the environment, "ideas" being fainter versions of impressions conjured by the mind during thought (e.g. memory).
The problem in interpretation of Hume's ethical theory is that often even people who grasp the anti-phenomenological import of Hume's empiricism don't interpret his ethical theory in a way that is consistent with his epistemology. When Hume says that there is no way to derive propositions about causality from propositions about correlation, or propositions about personal identity from propositions about self, he is saying no more nor no less than when he says this about propositions about value and propositions about fact. The is/ought distinction is simply another variation on the general theme that there is no need (no possibility) of "theory" over and above what is given by experience. And values (or rather, goodness and badness) are as much a given of experience as are causal relations and self-awareness. Thus Hume is not a "subjectivist" nor a "relativist" about ethics. In fact one of his targets is systematizers, for example of the religious variety, who claim that ethical responses are produced by embracing their systems, a form of cognitivism. The goodness and badness of experience is no more subjective or relative than is the blueness and redness of experience. On Hume's view there simply is no distinguishing between the "mental representation" and the "fact," and there are no exceptions to this general truth. Thus there is nothing to be said about "mental representation" nor about "facts," if either of these are taken to be somehow distinct from experience. Subjectivism is just as vacuous as realism on Hume's view.
Finally, notice that this is a general account of value, ethical and aesthetic. Arguing for the moral rightness of something is not different from arguing for the beauty of something. While we may feel queasy about this, realizing that the quality of experience is not always intersubjectively consistent, this is how it goes; there is no logical proof or refutation of the quality of experience. That does not mean that someone cannot learn to appreciate the goodness or the badness of something.

6 comments:

  1. a.b.: "As most of my readers will know, Hume's best-known claim about ethical theory is that one cannot derive an "ought" from an "is." That is, descriptive propositions about the world do not as such contain any prescriptive meaning (it's important to remember that Hume is talking about propositions)."

    The simplest of ostensive definitions, "That is a cat," relies upon the implicit prescription assumption "One should tell the truth when making ostensive definitions". For me, such can be "derived" from the former, since it is implicit in its use. The first "contains" the latter. Indeed the normatives of language games, and the public standing of language, make it such that prescription precedes, or at the very least is consubstantial with, description.

    The biggest problem that I have with Hume, in that I have one, is that he tended to think of epistemology in a self/world binary, neglecting a necessarily intersubjective basis, and knowledge which makes "experience" a source of objective knowledge in the first place, the bundling of beliefs which allow our beliefs to bundle.

    So when you say, : "Central to this is Hume's denial that it made any sense to speculate about an "external world" that might or might not be "corresponding" to the world of our perceptions. Experience cannot be distinguished from the world, and the world cannot be distinguished from experience."

    The purpose of distinguishing between our own experience and the external world is of course the means by which we check ourselves and others for the kinds of beliefs that are held. It is through our fundamental questioning of this relation/disjunction that we end up revising and attributing beliefs as true or false to others. We understand that the world is different than others are experiencing it, because we attribute to those persons erroneous beliefs which help explain their behavior in a rational interpretation. "He thinks that man is a woman" or even "I thought that man was a woman" is the very kind of world/experience disjunction which serves us all well.

    a.b.: "There is no sense in which "my experience of blue" is different from "blue." To use Hume's language, mental contents consist of impressions and ideas, "impressions" being directly caused by interaction with the environment, "ideas" being fainter versions of impressions conjured by the mind during thought (e.g. memory)."

    While I agree that "phenomenal properties" are not needed, what is needed is a rational conception of belief, and with it the attendant idea of false belief. If I really, really saw that the light was green, when it was red, doesn't so much matter in nearly every case. But in that I believed the light was green is a completely different matter. Of course these beliefs are understood to be caused by events in the world, but they are not traceable to any kind of "sense event" or "experience" which makes them true or false, real or unreal.


    Nice to read your thoughts. Good luck with your submission.

    ReplyDelete
  2. You have misconstrued my topic. I am not concerned at the moment with Hume's theory of truth. (As to that, Hume of course thinks that we are constrained in our beliefs by our interactions with the world. He denied completely the idea that our minds somehow create a picture of the world from some free-standing character of their own: he denies that "minds" in that Cartesian sense exist altogether). But my topic here is from philosophy of mind, not epistemology. The problem of phenomenology is a pseudoproblem, as phenomenal "properties" cannot be distinguished from properties of experience. Thus the Husserlian project is incoherent. When we see this we can see that goodness and badness, like blueness and redness,can no more be located "in the head" than they can be located "in the world." Your "self/world binary" appears to be a misinterpretation. Meanwhile, of course Hume understands that we might have wrong beliefs. That is a different issue. I cannot believe that I'm seeing blue when I'm seeing red. Values are like colors that way, not like states of affairs (Thought it was a man but it's a woman). But your comment is helpful in terms of showing where I need to do some sign-posting. By the way I'm impressed with how much work you're doing on your blog. Thanks AB

    ReplyDelete
  3. a.b.: "But my topic here is from philosophy of mind, not epistemology."

    It may very well be that I have misconstrued your topic, but I think that this is because Hume has misconstrued "is" and "ought". The idea that one could talk about "mind" without thinking about "knowing" is for me a profound breech. A theory of truth is implicitly in a theory of mind, and this is one thing in which, for me, Hume fails, (because he is an empiricist). Because the intersubjective makes up the capacity for the objective (and the subjective), there is no conceivable categorical exclusion of "ought" from "is". This would be the equivalent of a Private Language really. Oughts govern intersubjective relations, once your "is" statement is taken to be corrected by the views of others, and thus to acquire objective status, prescriptive enter the domain.

    I certain do agree that the phenomenological problem is a pseudoproblem. I am not arguing a Husserian point at all. It is just that what makes it a pseudoproblem is the very think that breaks down the is/ought categorical distinction. That is, the very possibility of "wrong beliefs" is founded on normative prescription; that is to say, there is a both a rational and a communitarian character to an "is" declaration.

    To put it another way, it is quite possible to hold an non-representationalist view of mind, and still have rational arguments for what is true, and what one ought to do.


    These are just my thoughts on your post, and could very well have nothing substantive to do with the non-representationalist point you are trying to put across in your paper. I find the is/ought distinction to be much abused and given more credit that it is worth. Again, thanks for your thoughts. One of the best thinking weblogs on the internet.

    ReplyDelete
  4. Just to "cherry pick" a couple of things here (more what I'm thinking about than a challenge to you), 1) Yes we will continue to have arguments about what is true and false, whether we are representationalists about mental processes or not. Following this line leads to the realization that a lot of resistance to eliminativism is motivated by mistaken presumptions. The most basic form of the mistake is to think that materialism leads to seeing people as "zombies." This is a version of the fallacy of begging the question. 2) Hume is talking, I think, about goodness and badness per se. So the argument is parallel to Frank Jackson's "Mary" argument: no amount of physical description/science, either about the world outside of my skin or the world inside of my skin, is ever going to include "what it's like to see blue." Same goes for "what it's like to feel bad." This is the same point (for Hume, not for Jackson or Nagel!) as Wittgenstein's when he says that there are no meaningful propositions about ethics (or aesthetics, or consciousness, or spirituality): he doesn't mean these are unimportant or nonexistent things. W. takes these to be among the very most important of topics in fact. Only there is no philosophical or scientific "explanation" of such matters. And that is Hume all over. Hume is much more Berkeley than he is Locke: he gets it that the distinction between physical and mental "properties" is unteneble.

    ReplyDelete
  5. I think that Kantian deontological arguments are reducible to arguments from utility: lying is bad because eventually someone feels bad (I admit that violating someone's autonomy by lying to them is wrong even if they never know the truth, and that that looks like a problem for my reductive view of Kant's ethics). But I suspect that we are just going to have to agree to disagree on that. I say: no beings with experiences that are good and bad, no value in the world. And beings with experiences that are good and bad are strictly contingent: they don't necessarily exist.

    ReplyDelete
  6. Sure. I do not ascribe to a Kantian notion of the ethical. But just because something is contingent and not necessary does not mean that it arbitrary either. A particular belief, for instance, ANY particular belief, might be false, but it is the nature of beliefs that they cannot all be false, for they would no longer operate as beliefs. This is not due to some transcendental fact, but is rather due to the nature language, communication and justification, the communally shared understanding of the world as causal of our mutual affects. This goes right to your concern with animal minds, for far below the propositional level we as animals know that the expressions of other beings informs us about the nature of the world. The prescriptions of this utility are of a kind that govern all kinds of animal behavior. In this way, all uses are contingent, not just ethical or aesthetic uses, but one of the things they are contingent on is agreement, and agreement buries itself down into the very roots of animal organization. (My dog and I can both "agree" that the intruder who is entering my house is a bad person, though my dog does not have language.)

    I don't know what a necessarily bad experience is, or a necessarily good experience is either. But I do know that if you do not obey the prescription that one should not lie when making ostensive definitions, you will find it very hard to teach a child how to use language. And if you do not operate under a charitable attribution of rationality to a speaking other, you will not be able to interpret what they are saying. As you say, this is a matter of utility, but a utility of how beliefs rationally hold together. There can be no rule-following (Wittgenstein), if there were no idea of getting-it-right or getting-it-wrong. The utility of getting-it-right includes the valuation that getting-it-right as a good thing, a thing worth doing. There is nothing transcendent about it.



    Of course, I am happy to agree to disagree.

    ReplyDelete