Suppose the following syllogism:

  1. It is impossible for anyone to get X without him/her doing Y.
  2. It is possible to get X (by doing Y).
  3. I want to get X.
  4. Therefore I ought to do Y.

There is, very likely, nothing wrong in this type of reasoning, even though it contains a shift from is to ought. Perhaps, we can even call this reasoning deductive reasoning, the one that Hume appraises so much (even though any its justification using itself will be circular reasoning).

However, Hume argues that the is-ought shift is not the one we usually use in reasoning. How can that be true? Does this argument discredits Hume's problem?

  • 1
    Your "syllogism" is ambiguous. The second premise is either an ought in the Humean sense (it includes goals, wants, values, etc.) or a statement of fact concerning you, an is. In the latter case the conclusion is also an is. Either way, there is no shift from is to ought. This sort of argument is well known as means-end reasoning, see also deontic logic. "Wars kill millions of people, therefore we ought to end them" does contain such a shift.
    – Conifold
    Sep 16 '18 at 21:23
  • @Conifold, then can we say that 1. Do not kill. 2. For violating the rule 1 you will be punished. is a perfectly fine normative theory that will work for those who do not want to be punished? I just feel that most normative (especially those regarding jurisprudence) theories use ought in this sense (therefore escaping the is-ought problem).
    – rus9384
    Sep 16 '18 at 23:20
  • One is perfectly fine as long as general ought premises are explicitly listed and all subsequent oughts are derived from them. The problem is that a "theory" has to justify such general oughts somehow, and attempting to derive them from "nature" (as in natural law), God's will, "common humanity", evolution, or some other such thing inevitably makes the shift at some point. The law will not get far with "do not kill" alone, and even it follows from no facts. And non-controversial systematic justification for general legal oughts is hard to come by.
    – Conifold
    Sep 16 '18 at 23:35
  • @Conifold, well, voluntary organizations have oughts derived from wants. If one does not follow them, [s]he gets banned (expelled). I think most systems of law do pretty much the same with the only huge difference: ban is equal to imprisonment or fine, not an extradition or cessation of protection from government. And extradition is impossible pretty much because there is no terra nullius (except Antarctica), therefore people do not have choice to not be in such organizations (which makes governments involuntary).
    – rus9384
    Sep 17 '18 at 0:06
  • Law has to apply to an entire society, which is not exactly a voluntary organization, and most moralists will reject basing morality or law on "wants". It is also pragmatically intractable considering that wants are feeble and fleeting, and people have incompatible ones, while laws have to apply universally and for a duration. Believe me, you won't come up with a workable theory in your head, there isn't any non-controversial one at all, despite long efforts by many people who spent their lives thinking about it.
    – Conifold
    Sep 17 '18 at 0:16

You are tacitly assuming that if you want X, you ought to take the means to obtain X. Whether or not Hume would allow the premise, if you do allow it, then "I want X" is a claim about an ought (about a value rather than a fact), so therefore your whole syllogism adheres to Hume's principle: reason can't get you a conclusion about ought from premises about is. Once you include at least one premise about ought, then you can deduce conclusions about ought (and with at least one premise about is, you can deduce conclusions about is). I believe Hume would in fact insist that claims like "I want X" are closely related to ought-claims, because the is/ought distinction depends on a more basic Humean distinction between beliefs and desires.

The question of the conceptual connection between instrumental imperatives ("You ought to do X if you want to get Y") and true obligations has been important in ethics since Kant's Groundwork of the Metaphysics of Morals and the strategy of developing instrumental "obligations" into an obligation to take other people's goals/desires into account was developed quite nicely by Nagel in The Possibility of Altruism. — The general insight you've had, that ought and is can easily be linked by modal considerations, is typically referred to as "ought implies can". — You may also be interested in GE Moore's "Open Question Argument", which relies on a more radical form of the is/ought distinction (and so may be close to the view you seem to attribute to Hume).

  • 1
    So, it is about the nature of the desire. Then I agree it is not always produced by reason, there are original desires produced not by reason and all other desires might be drawn by reason from these. However, the things are gonna be more interesting if we assume panpsychism. Regarding Moore's argument, I don't really see that his argument is valid.
    – rus9384
    Oct 18 '18 at 8:00
  • Yes, I agree that Moore's OQA involves a confusion, but neither is it easy to lay out why Moore's "naturalistic fallacy" isn't a fallacy if Hume's "is/ought" distinction is valid.
    – guest1806
    Oct 18 '18 at 16:10
  • Well, in one of my questions I showed that open question argument with one of the premises being inversed results in the opposite conclusion. And both premises are debatable.
    – rus9384
    Oct 18 '18 at 20:17

If he is consistent on this question, then yes, I believe he must. Setting aside whether he really is consistent, let's consider one of his most famous statements on the matter, in A Treatise of Human Nature (quoted here):

’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger. ‘Tis not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me. ‘Tis as little contrary to reason to prefer even my own acknowledg’d lesser good to my greater, and have a more ardent affection for the former than for the latter.

The first sentence is the most famous, but for our purposes, the last sentence is the most important. As I read it, it says that there is no rational cause to prefer a better outcome to a worse one, even if you rationally believe that the better is the better, and the worse is the worse.

This lines up pretty neatly with your proposed syllogism. Say you want X, but you prefer not to do Y. Say even that you desperately want X, and every life outcome that is possible for you would be better if X, and you fully know it. Here Hume directly says it is not contrary to reason to prefer not to do Y, in spite of all that.

Now the only thing that clearly differs (to me) between Hume's statement here, and your four statements, is that in the fourth statement you say that you "ought" to do Y, given 1, 2, and 3. So the main question is whether there is a clear difference in this situation between what you "ought" to do, and what you "prefer" to do. And I don't think there is, at least not necessarily, as a syllogism would require. In a very, very large number of cases, what you ought to do is just what you prefer to do.

To make this concrete, suppose you strongly prefer to sleep on your side instead of your back. It seems obvious that you ought to sleep in your side in that case. Now suppose you learn that sleeping on your back will extend your life expectancy. And suppose you really want that! Nonetheless, if wanting that doesn't change your fundamental preference for sleeping on your side, then it seems uncontroversially true that you ought to continue sleeping on your side.

  • 1
    4 is the conclusion from the first three, not the premise.
    – rus9384
    Sep 17 '18 at 16:02
  • "Here Hume directly says it is not contrary to reason to prefer not to do Y, in spite of all that." Did you mean X?
    – PyRulez
    Nov 20 '18 at 14:42
  • @PyRulez no, I didn't. X is the thing you want. Y is the thing you don't want to do.
    – senderle
    Nov 20 '18 at 16:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.