On Desire 2018. Part 58: A Desire that "P and a desire that not-P"

In our last exciting episode, I discussed Wall's argument that Moore's Paradox regarding beliefs lead to a "norm" with respect to beliefs that said that one ought not to have false beliefs. There was something wrong with false beliefs and, since Moore's Paradox concerned beliefs that, though they could be true, always contained a false component when believed, showed that there was a problem with believing that was false.

Wall, David, (2017), “Desiderative Inconsistency, Moore's Paradox, and Norms of Desire,” In Deonna J. & Lauria F. (eds). The Nature of Desire. Oxford University Press.

Wall's norm that we ought not to have false beliefs could be corrected in one of two ways - either by changing the belief to fit the world, or by changing the world to fit the belief.

The argument follows the same form as was used to argue that there is a norm against having a false belief. In that case, Wall noted that there was something odd in believing that "P and that I believe that not P". He deduced that this oddness was due to the fact that either "P and that I believe that not P" is false or "I believe that not P" is false. There is no way to get out of having a false belief, so there must be a norm against having a false belief.

The parallel argument against desires would begin by noting that there is something odd in desiring that "P and I desire that not P". This is wanting it to be the case that "P is true but I want P to be false". Wall interprets the oddness of this as the oddness of having an unfulfilled desire. It is just intrinsically wrong to want an unfulfilled desire. It violates a norm of desire.

It is important to not an implication that Wall draws from this norm. There are two ways of preventing an unfulfilled desire. One is by altering the world - make it the case that P is false. If P is false and I want P to be false then the world matches my desire, and all is good, right? The other option is to make it the case that I want P to be true. If P is true and I want P to be true, I have a fulfilled desire, and that is a good thing, right?

This, however, violates the "direction of fit" claim about desires - just as the analogous case does about beliefs. It suggests that there are two ways of creating a fit between a "desire that P" and the world. One is by changing the world (making or keeping P true). The other is by changing the desire so that the proposition that one desires is true, and thus the desire is automatically fulfilled. The same is the case with respect to belief. If a person believes that the lights are off, but discovers that the lights are on, he has two options. He can change his belief, or he can turn the light off. (Why he would turn the light off, if he believes the lights are already off, is a mystery to me.).

I do not understand why Wall picked this particular norm, given its conflict with the direction of fit, given that a different norm exists that matches the direction of fit. This is simply the norm, if your belief does not match the world then you ought to change your belief. If your desire does not match the world, then change the world.

As I have been describing belief and desire, a belief is the assignment of a value V(B) to a proposition where V(B) is the credence value of the proposition being true - the degree if certainty that the belief is true. A desire is an assignment of a value V(D) to a proposition where V(D) is the importance of the proposition being true.

In the case of V(D), it is simply the case that it makes no sense if one assigns a value V(D) to P being true that one would respond to V(D) by altering one's desire. What is important to the agent is that P be true, not that the agent has a fulfilled desire. The only way to make or keep P true is by altering the world. One cannot make or keep P true by altering one's desire to a desire that not-P. In fact, this would thwart the desire, as this would then motivate the agent to act in such a way that P is false.

Take, for example, an animal who is hungry. It has a desire to eat. There are two ways, according to Wall, that he can respond to this desire to eat. One is by eating (and thus sustaining its life and having offspring who also respond to its desire to eat by eating), or by changing its desires so that it no longer has a desire to eat (and thus starves to death and has no offspring who is so unfit for the world that it responds to desire in this way). A desire that P assigns a value V(D) to P being true. A desire to eat is a desire "that I eat" - which motivates the agent to make or keep the proposition "I am eating" true, until it has eaten enough, at which point the desire subsides.

Now, can one account for the oddness of the type of state that Wall is bringing up consistent with the thesis that a desire that P is the assignment of V(D) to "P" be being true?

Well, a person with a desire that "P and a desire that not-P" has a reason to act to bring about not-P. But not-P will thwart the desire that "P and the desire that not-P". It is important to remember that a desire that "P and a desire that not-P" is not the same as a "desire that P" and a "desire that not-P" in the same way that a desire for pumpkin pie with whipped cream is not the same as a desire for pumpkin pie and a desire for whipped cream.

So, still accepting the proposition that no desire can be evaluated except in terms of its relationship to another desire, the desire that "P and I desire that not-P" can be evaluated according to its relationship to the "desire that not-P", and the "desire that not-P" can be evaluated according to its relationship to the desire that P. We do not need a norm of not having conflicting desires. We simply need a norm of making or keeping true the propositions that are the objects of our desires.