Tuesday, March 19, 2013

Is, Ought, Can

It's a common dictum, derived from over-reading Hume, that 'is' does not imply 'ought'(INO). But it is also a common dictum, derived from Kant (I make no judgment here about whether it is Kant over-read), that 'ought' implies 'can'(OC). It's notable that these two dicta are inconsistent with each other:

(1) If X ought to do Y, X can do Y. [OC]
(2) If it is false that X can do Y, it is false that X ought to do Y. [(1), Contraposition]
(3) If it is false that X can do Y, X is lacking in the ability to do Y.
(4) If it is false that X ought to do Y, it is true that X ought to do something other than try to do Y [as if Y ought to be done].
(5) If X is lacking in the ability to do Y, X ought to do something other than try to do Y [as if Y ought to be done]. [(2), (3), (4)]

But (5) tells us that an 'is' implies an 'ought', which violates INO. But (3) is clearly a tautology, and (4) is clearly true for anything to which 'ought' can apply at all, assuming that you ought not to try to do things that you ought not to do; which is arguably a tautology itself.

This is a general issue. The reason someone like Hume can accept something like INO is that he has a very restricted and minimal account of modality (i.e., necessity, possibility, etc.). Anyone with a more robust account of modality than Hume has to allow for the possibility that they can, in fact, get an 'ought' from an 'is' that has these more robust modalities. At least it would have to be proven on a case-by-case basis. An 'ought' is just a strong modal operator. It can indeed be difficult to get from a situation that has no well-defined modal operator, or a distinct and isolated null modal operator (as 'true' sometimes is), to one that has a strong modal operator. But it is not generally a problem to get from one well-defined modal operator to another; it just depends on the modal operators in question and how they are related.

Incidentally, this kind of modal situation is not uncommon. Kant has a more robust account of modality than Hume, but almost all of Kant's more skeptical results follow simply from the fact that for Kant all modalities are in one way or another epistemic, having to do entirely with how we know things. If you have a less restrictive account of necessity and possibility than Kant does, you can no longer assume that Kant's conclusions follow from his principles -- it would have to be examined on a case-by-case basis.

[ADDED LATER: A phrase accidentally was dropped out that made (4) read oddly, and this mistake was propagated through the rest of the argument. I have fixed it.]


  1. Brigitte1:28 PM

    I apologize for my comment being two days late.
    Your poem draft, Burdens, was the one responsible for this delay: ... this ever-winding stair; and yet another count of failing,...struck me last night with great force. I had been contemplating , with some regret and shame, the fact that I had failed in my Lenten promise to myself. But then I decided to follow up on the admonition... to rise anew...
    Thank you!

  2. Vanitas2:41 PM

    "(4) is clearly true for anything to which 'ought' can apply at all,
    assuming that you ought not to try to do things that you ought not to
    do; which is arguably a tautology itself."

    This is not an adequate defense of (4). There is an open option: X can do nothing at all. The only way to block this option is to include "doing nothing" in the class of actions, in which case I think Hume would have been content to rest easy. If what we are "deriving" is that there is an unimaginably massive set of contrary actions, one of which we 'ought' to do, a set which contains countless contraries along with the option "do nothing at all", then the problem of deriving oughts from is-es seems as live as ever.

    Also, if you're the non-naturalist moral realist that you've claimed to be in the past, perhaps OIC ought to make you nervous? As I understand it, it is very difficult to combine the idea that natural facts (such as abilities) can constrian normative ones wih the idea that normative facts are wholly non-natural. Perhaps I've misread your position, though.

  3. branemrys6:48 PM

    A phrase dropped out of (4) by accident, so you're right that there's something odd about it (in particular, it could only apply to certain kinds of oughts as it originally stood). However, X doing nothing at all is not literally possible: when we talk about X doing nothing at all, we are relativizing X to some kind of action; as you say, it just means 'doing something contrary to Y'. But this doesn't keep "the problem of deriving oughts from is'es" alive -- if we get this far, we've already met this challenge completely. (I agree, however, that Hume would rest easy: Hume doesn't accept INO in its full form, although it's often misattributed to him.) To be sure, if that is all we can get (nothing in the argument requires that it be all) we're in the awkward position of having to get specific oughts by elimination, but this is no different than doing certain kinds of physics, where, out of all the infinite possible mathematical models that fit any finite set of data we're often stuck with simply narrowing down the regions and proposing a current model of best fit. But that's actually quite robust, limited though it may be.

    I'm not a naturalist, largely for epistemological reasons, but I don't think normative facts are non-natural. At least large parts of my moral realism are consistent with some forms of nonreductive naturalism. Actually, I think this will inevitably be true; normative facts about human beings will be facts that could be recognized as natural by at least some form of naturalism; and almost all the normative facts we know will be about human beings. Further, there are obvious normative facts that any non-stupid naturalist will consider natural: facts about what is healthy, facts about what is or is not against the law, facts about what is rational.

  4. branemrys6:57 PM

    Glad you liked it! If I counted the Lenten resolutions I've failed to keep, I don't know if I'd ever get to the end.

  5. branemrys7:25 PM

    I should add, incidentally, that even if this is true:

    it is very difficult to combine the idea that natural facts (such as
    abilities) can constrain normative ones with the idea that normative
    facts are wholly non-natural

    (whether it is would depend, I think, on exactly what it means to say that a normative fact about how a human being should act is "wholly" non-natural), 'ought implies can' doesn't entail that natural facts can constrain normative ones; it's consistent with it but, for instance, Kant himself wouldn't accept the latter. It could be (a la Leslie) that normative facts constitute natural facts, for instance; or it could be that the 'natural facts' are in fact just logical facts that happen to constrain both natural facts and normative facts together. (The way to think of it is this: "Ought implies can" only indicates a correlation; but correlations admit of many different causal or 'constraining' explanations.)

  6. catalexico2:16 PM

    If is does not imply ought, then what does, perhaps isn't ? Where do the oughts come from, noughts ?

  7. branemrys4:36 PM

    That definitely is the question. There are different positions, but I think the usual idea is that oughts derive only from other oughts, and that there are some self-evident oughts. (Hume was originally concerned only to eliminate the idea that oughts were relations perceived directly by reason. Kant uses a similar expression, but mostly to distinguish speculative reason from practical reason. Later thinkers seem to use it in all sorts of different ways.)


Please understand that this weblog runs on a third-party comment system, not on Blogger's comment system. If you have come by way of a mobile device and can see this message, you may have landed on the Blogger comment page, or the third party commenting system has not yet completely loaded; your comments will only be shown on this page and not on the page most people will see, and it is much more likely that your comment will be missed.