Railton2: The Fully Informed Agent

Faithlessgod, in a comment concerning Railton's ethics, posted an account that describes it in the following terms:

there is the question over ethics serving to best fulfill the desires of Agent versus advised by Agent+. Railton and Griffin argue that ethics should only serve to fulfil the informed desires of Agent (that is as informed by Agent+) even if the Agent is, in fact, actually uninformed.

Agent+ is described as:

the Agent as full informed, having succeeded at a cognitive therapy etc. and so would know that certain actual desires-as-ends of the Agent are unfulfillable.

The question that I would then have is: What is the difference between the option that best fulfills the desires of Agent versus those advised by Agent+? Would Agent+ ever have any reason to recommend anything other than wht would best fulfill the desires of Agent?

We can draw a distinction between what uninformed Agent believes will best fulfill his desires, compared to what fully informed Agent+ believes will best fulfill the desires of Agent. Furthermore, it follows from the hypthesis that Agent+ only has true beliefs, while Agent is likely to have false beliefs.

However, it does not follow from this that there is a difference in the fact of the matter with respect to what will fulfill the desires of Agent.

This is true in the same case that Agent may have different beliefs about the chemicals that are found in an orange than a fully informed chemist. However, the fact that the uninformed Agent and the Chemist disagree, or the fact that the chemist knows the right answer, argues that the chemical composition of an orange is actually different depending on whether it is being referred to by Agent or by a chemist. There is still only one fact of the matter, and it is the same for both individuals.

We can see this in terms of the fact that Agent+ knows that certain desires are unfilfillable, while Agent is ignorant of this. The ignorant Agent may well waste energy trying to fulfill unfulfillable desires. However, the answer to the question, "What best fulfills the desires of Agent?" does not change.

The account above mentions cognitive therapy. We can describe cognitive therapy by looking at one of its specific applications. Cognitive therapy is used, in one set of cases, to deal with irrational fears by pointing out to the patient and getting the patient to fully realize that the fear is irrational. The patient is asked to fully visualize "the worst that could happen" from a situation and to realize - to truly know - that the worst that could happen is not really that bad.

So, let's assume that I have a fear of public speaking. The cognitive therapist helps me with this by helping me to realize that the consequences of public speaking are not as bad as I have imagine them to be. Once I realize this fact, I find it easier to get up in front of a crowd and give a speech.

Has the cognitive therapist actually changed my "desires-as-ends?" Or did the cognitive therapist actually only help me to realize that certain actions will not not, in fact, actually bring about the types of ends that I thought they would bring about - that is, brought me into a more rational appreciation of desires-as-means?

If it is the latter, then there is still no difference between what will best fulfill the desires of Agent, and Agent+ (Agent after having gone through cognitive therapy) would advise for Agent. The only difference is that Agent+ knows facts that Agent does not know. Desires-as-ends remain the same and, consequently, what best fulfills desires-as-ends remain the same.

If, on the other hand, the claim is that the cognitive therapist actually changes desires-as-ends, then we are no longer talking about "what would fulfill the desires of Agent as advised by Agent+". We are talking about "What would best fulfill the desires of Agent+, as advised by Agent+, whose desires-as-ends are different from those of Agent."

We are left with a series questions.

(1) What reason is there to believe that there is some type of inference from belief to desires-as-end that the cognitive therapist claims to exist? What is this inference? How can you derive a desires-as-end from a belief?

(2) Why should Agent act so as to best fulfill the desires of somebody other than himself - particularly somebody who doesn't exist?

In desire utilitarian terms, my fear of public speaking may be a bad desires-as-ends. It may well be a desire that tends to thwart my other desires. However, that merely implies that I have reasons-for-action (residing in the desires-as-ends being thwarted) to rid myself of this fear of public speaking.

Furthermore, one of the ways that I might accomplish this is by visualizing or imagining myself engaging in public speaking and realizing that the consequences are not that bad - making myself comfortable with the likely consequences through imagined and real-world practice.

However, the fact that I have a desires-as-ends that I would be better off without does not imply that the desires-as-ends is not an actual reason for action. I might be better off without the aversion to the pain that I would feel by having a dentist drill into the nerves of my tooth. However, that does not change the fact that the pain that I would feel gives me a valid reason-for-action to ask for some type of pain killer.