This is a devilishly complicated chapter. This is what is going on.
Williams is talking about the attempt to say that morality comes from God. This sort of thinking about ethics was criticized by Immanuel Kant. Kant held that there are two possible reasons for obeying God’s wishes.
Kant thought the first alternative was fine but that it would not be compatible with basing ethics on God’s will. You would have to know that God is good and just before you could know that you should obey his will. If so, goodness and justice cannot come from God’s will.
Kant thought the second alternative was inappropriate for morality because it involves the wrong kind of motivation. People who comply with the moral rules out of fear of divine punishment do so for self-interested reasons and that, Kant thought, means that they are not truly acting morally.
It is this point about the motivation to behave morally that Williams spends the bulk of the chapter on.
By the end of the chapter, he will have argued that there is a third option, namely, that the reason to comply with God’s will is based on a relationship with God. You might think that you have to do what God wants because you do not want to be separated from God. That motivation is based on neither your ethical evaluation of God nor on your own interests.
Williams makes the case that for thinking that this is a recognizable motivation for a lot of what we commonly think of as moral behavior. He maintains that the task for religious ethics is not to show how compliance with God’s will can meet the standards for moral motivation. Rather, the problem is in explaining the nature of the relationship between finite mortals and an infinite supernatural being.
In the middle of the chapter, Williams compares Kant and Hume. As we know, Kant believed that moral motivation had to be pure: the person who acts morally is the one who is motivated by duty alone. For Hume, by contrast, moral behavior is motivated by sympathy and sympathy involves a mixture of concern for others and for yourself. You sympathize with others, for Hume, when you feel what they do and so when you act out of sympathy, it’s partly in order to experience the joys, or to relieve the pains, that others feel for yourself.
Williams makes a case for Hume’s side of this dispute, in part by noting some of the extreme positions that Kant is apparently driven to.
In our discussion, I think it would be worthwhile to reconstruct, in our own words, three of Williams’s points.
The discussion of the person who gives money to charity merely to improve his own reputation (Williams 1972, 66–67).
Why Williams thinks Kant is forced to adopt an “absurd apparatus” of “duties to oneself” and why he thinks it is absurd to suppose we have duties to ourselves (Williams 1972, 68–69).
What point is he making when he says that we can act in the interests of others because we love, admire, or respect them (Williams 1972, 70)?
Let’s go back to Kant’s criticisms of the religious moralist. Kant said there were two reasons for complying with God’s commands.
The religious moralist can add a third option: you have a relationship with God that explains why you should obey his wishes.
That said, Williams doubts this can be spelled out. But that is not because there would be a problem with understanding this as a kind of moral motivation. Rather, the problem is that it is impossible to spell out what the relationship is supposed to be like.
If God existed, there might well be special, and acceptable, reasons for subscribing to morality. The trouble is that the attempt to formulate those reasons in better than the crudest outline runs into the impossibility of thinking coherently about God. The trouble with religious morality comes not from morality’s being inescapably pure, but from religion’s being incurably unintelligible. (Williams 1972, 72)