Williams on Internal and External Reasons - Part 3

I am trying to work out a if Bernard Williams, in “Internal and External Reasons” (in Moral Luck, Cambridge: Cambridge University Press, 1981, 101–13), believed that beliefs provide reasons for action.

I would argue that they do not. If the agent has a desire that P, and if φ-ing will bring about P, then agent has a reason to φ. He may have more and stronger reasons not to φ, but he at least has a reason to φ.

The agent may know this to may be ignorant of the fact. She may falsely believe that she has a reason to φ when she does not. But the reason to φ is in all cases independent of what the agent believes.

Two passages so far suggest that Williams states that the reasons the agent has (as distinct from an awareness of the reasons an agent has) depends on beliefs.

The first of these is the passage I cited at the end of my previous post:

A member of S, D, will not give A a reason for φ-ing if either the existence of D is dependent on false belief, or A’s belief in the relevance of φ-ing is false.

This suggests that there are members of S dependent on true beliefs that do give reason for φ-ing, or the belief about φ-ing is true.

The second passage says:

A may be ignorant of some fact such that if he did know it he would, in virtue of some element in S, be disposed to φ: we can say that he has a reason to φ, though he does not know it. For it to be the case that he actually has such a reason, however, it seems that the relevance of the unknown fact to his actions has to be fairly close and immediate; otherwise one merely says that A would have a reason to φ if he knew the fact. 

When the unknown fact is "fairly close and immediate", Williams claims that the reason to φ is independent of belief. It exists, and true belief reveals it. On the other hand, when the unknown fact is relatively far from the action, Williams wants to say that the belief does not reveal a reason to φ but, instead, creates a reason to φ.

I cannot see any justification for this distinction. We (or some of us) may have a linguistic habit to speak this way of beliefs that are relatively near to an action and beliefs that are relatively far. However, there is nothing of substance to support this distinction. If he has such a reason when the belief is relatively near the action, then he also has such a reason if the belief is relatively distant from the action.

Let us go back to the first of these two passages:

A member of S, D, will not give A a reason for φ-ing if either the existence of D is dependent on false belief, or A’s belief in the relevance of φ-ing is false.

Whenever D is dependent on a belief, true or false, then D is a means, and never an end. As a means, the reasons for obtaining those means comes entirely from the ends they serve, and not from belief. Ultimately, there is no "member of S" dependent on belief, true or false. Correspondingly, there is no member of S dependent on A's belief in the relevance of φ-ing.

The reasons for φ-ing exist or not as a result of desires and facts. Beliefs reveal reasons for φ-ing, they do not create them.

This should be restated:

D is not a member of S (thus, will not give A a reason for φ-ing) if the existence of D is either dependent on belief or on A’s belief in the relevance of φ-ing.