Wednesday, September 22, 2010

Desires based on false beliefs

Philosophers often claim that 'actual desire' theories (of welfare or reasons) are refuted by various cases involving ignorance or false beliefs, so that we need to move to an 'informed desire' view instead. This has always struck me as a mistake. (We might need to idealize in some more radical way, but that's another matter.) It's easy enough to show how the standard cases are accommodated by actual desire views: ignorantly ingesting a poison would have the effect of thwarting many of your actual desires, and pursuing some instrumentally ineffective means would fail to achieve your actual (ultimate) goal, etc. Parfit seems to agree with this much, but then he presents a trickier case:
In some cases, however, our telic [i.e. ultimate] desires or aims depend on false beliefs. I might want to hurt you, for example, because I falsely believe that you deserve to suffer, or because I want to avenge some injury that I falsely believe you have done me.

This sounds to me like a misdescription.* Ordinary vengeful desires are conditional in form. I may have a desire that you suffer presupposing that you injured me. This presupposition is built into the desire: if it is not satisfied, then neither is the desire (it would then be cancelled, rather than either satisfied or frustrated). Imagine asking the vengeful guy: "Do you unconditionally want this other fellow to suffer, even on the hypothesis that it wasn't he who injured you?" Presumably not.

When evaluating desire theories, we really need to pay attention to the sorts of subtle distinctions discussed in my old post, 'Structures of Dynamic Desire'. In particular, it's not enough to just observe that a desire is 'based on a false belief' in some sense or other, because there are importantly different ways that this could go. We need to consider whether a false belief is (i) part of the content of what's desired, such that the desire is thwarted by its falsity; or (ii) a presupposition for the applicability of the desire, so that the desire is cancelled; or (iii) a merely causal influence on the genesis, or perhaps even the continued retention, of the desire -- in which case the falsity of the belief makes no difference to the satisfaction-status of the desire.

We can imagine a case of the latter sort. This would involve an agent who strikes the rest of us as crazy and incomprehensible: he unconditionally wants Joe to suffer, even though he only acquired this desire in the first place because he thought that Joe did him an injury. Still, he now intrinsically wants Joe to suffer even if innocent. This is very strange, like the agent who intrinsically desires to count blades of grass. In such cases, the desire theorist is presumably committed to just biting the bullet and accepting the desires in question as genuinely reason-giving for the agent. (The desire theorist may say things like, "We must defer to their normative perspectives, however alien it seems to us. Who are we to tell them that they care about the wrong things?") Note, in particular, that the mere causal genesis of the desire - as having originally formed in response to a false belief - does not seem to be of the faintest relevance.

So this creates a kind of dilemma for the argument that we need to shift to an informed desire view. Either the false belief features in the desire such as to influence its satisfaction conditions in one of the two ways described above, or it is merely an external causal influence. In the former case we have already secured the result that harming Joe would not fulfill the desire. In the latter case, it is unclear why any such merely causal fact should be thought to have any rational significance. It is as if (say) I had a brain implant -- which would be activated by learning new information about plant biology -- which upon activation would rewire my brain, causing me to acquire a brute desire to eat grass. In such a case, my ordinary dietary habits in some sense "depend" on my ignorance of plant biology (i.e. to avoid triggering the implant), but given that the dependence is merely causal, it clearly has no rational significance for what I should now be eating. Likewise, the subjectivist should say, for the person who unconditionally desires to harm Joe -- even if as a matter of brute causal fact, he would lose this desire upon acquiring new information about Joe. So either way, cases of "desires based on false beliefs" provide no reason to move to 'informed desire' theories.

* Parfit has a footnote where he mentions conditional desires as "another way to describe such cases." I think it's important that they are a more accurate description of the ordinary case. The unconditional desire is an importantly different -- and more alien -- case.

1 comment:

  1. Hi,

    I'm in complete agreement. Chris Heathwood defends actual-desire-satisfaction theories against this sort of objection in his : "The Problem of Defective Desires." It's available here:


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)