Korsgaard 03: Judging Reasons

We are not animals.

Well, yes, of course we are animals, but we are different. We are the only animals with the capacity to judge the quality of our reasons for acting. We can say, "This is a good reason. That is a bad reason."

Christine Korsgaard, in The Normativity of Instrumental Reason seeks to investigate whether the instrumental principle, whereby A’s being a means to B gives an agent a reason to do A if B is one of A's ends. She objects that:

A practical reason must function both as a motive and as a guide, or a requirement. I will show that the empiricist account explains how instrumental reasons can motivate us, but at the price of making it impossible to see how they could function as requirements or guides.

She further asserts that:

Empiricists . . . suppose that the instrumental principle is either obviously normative or does not need to be normative because we are reliably motivated to take the means to our ends. Instrumental thoughts cause motives.

I will go with the "reliably motivated to take the means to our ends" option . . . or, what is actually the case, "reliably motivated to take what would have been the means to an end if the agent's relevant beliefs were true." Clearly, we sometimes fail to take the means to our ends when our beliefs are false, but that should not be taken as the focus of the debate.

Using the model of how desire work that I wrote up a little while ago, cognition and perception put actions at action gates and then desire opens the gate for its preferred option. The reward system is a black box, but we can have an idea of how it works from its effects, and "choosing the action that would fulfill the agent's desires if her beliefs were true" appears to be a reasonable hypothesis.

Korsgaard argues that, if this is the case, then the instrumental principle cannot be a guide for action. That is to say, our actions are not guided by a principle that states, "since action A will fulfill my desire that P, I ought to do A."

Well, that's fine. It is not a guide for action. We do not need it as a guide for action. It is a statement about how the black box of desire selects actions at the gate - a statement that allows us to explain and predict which actions the black box will open.

Then, how is irrationality possible? It would seem, in this case, that an agent always chooses the most rationally best action. Korsgaard brings up the case of Howard whose fear of needles prevents him from getting a treatment he would otherwise get. Korsgaard claims that, if we always chose the means to our ends, and one of those ends is to avoid needles, then Howard is acting rationally to avoid the needles.

[I]f we reject the claim that prudence is a rational requirement, we will say: fear determines what Howard's preferred end is, but there is no irrationality in the case, for reason has nothing to say about what ends we should prefer.

However, this does not follow. Korsgaard is confusing the reasons for action given the agent's desire, and the reasons to desire. Both claims can be true at the same time: Given Howard's fear, he has more reason to avoid the needle than given the treatment. Howard has many and strong reasons to rid himself of this fear.

Howard's fear does not give him a reason to fear. It gives him a reason to act. The desires that Howard would fulfill if he did not fear needles gives him reason to not fear - to do something that would get rid of his fear, so that it would quit motivating actions that thwart those desires. In other words, Howard has no reason to fear. He has reasons not to fear. So, he has more and stronger reasons not to fear than to fear. Avoiding needles is NOT an end he should prefer or even have.

This means that, if there is some course of action that would rid Howard of this fear . . . such as sensitivity training or cognitive therapy . . . then Howard has more reasons to go through this therapy than not to. This may not continue to be the case if we add in the opportunity cost of time and money, but in their absence we can still draw this conclusion from Korsgaard's example.

So, it is not the case that, if the empiricist account of instrumental rationality is true, we "have nothing to say about what ends that we should prefer." Ends are also, at the same time, means to the fulfillment (or thwarting) of other ends. In their role as means, we have reason to judge them as we judge any other means.

It is important here to note that the grounds for our judgment is based on desires. If the fear of needles did not thwart other desires - if it, instead, tended to fulfill other desires - then it would not be true that Howard has reasons to rid himself of this fear. He might - depending on the consequences of having this fear on the fulfillment of other desires - be something he has reason to cultivate. Once again, it is desire that determines the "ends we should prefer" - not any kind of appeal to pure reason.