[Note: MM makes the same point, but more succinctly.]
Darwall thinks that Nagel’s argument fails because (B’) begs the question against egoism. He is correct, and for the right reasons, but I find his presentation a bit dense (possibly because I am the same). Here’s how Darwall puts it:
Since (A’) is true, what I judge when I judge that I have reason to do [x] is identical with what you judge if you judge that Darwall has reason to do [x]; in that sense, our judgments are the same. But in this sense, there may well be no judgment which has motivational content. That is, there may be no judgment which is such that any person’s accepting that judgment is capable of justifying a desire to promote what that judgment recommends … For if you accept that I (Darwall) have reason to do [x], that will only justify your desiring to promote my doing [x] on the assumption that … egoism is false.
As I understand it, the crux of Darwall’s objection is the italicized “any” but I find the force of the objection easier to grasp if it is unfolded slightly differently. This different angle also helps make sense of why one might think that the egoist must reject (A’).
Recall that egoism, on the conception that avoids the objection from GE Moore with which this post begins, is the doctrine that each person’s happiness [etc.] is the sole good for him. A consequence of egoism and (B’), therefore, is that if I accept that doing x is good for me that is sufficient to justify my desiring to do x. Or, alternatively, if I accept that I have a reason to do x in C then that is sufficient to justify my doing it. So far so good – there’s nothing objectionable to the egoist qua egoist here.
The problem comes with what (B’) does to the impersonal practical judgment. (B’) has it that if B accepts that A has reason to do x in C then (assuming (A’)) that is sufficient to justify B’s desiring that A does x. What’s fishy here is that (B’), which seems to capture an important conceptual truth in the personal case, has substantive first-order ethical implications in the impersonal case.
Here’s one account of what has happened. (B’) derives part of its plausibility from stating a conceptual truth about the connection of practical judgment and practical justification. The conceptual truth it states is unobjectionable to the egoist qua egoist, and actually yields the intuitive result for him when applied to personal practical judgments. But when applied to impersonal practical judgments it demolishes his position. Thus, one might suspect, what the egoist must deny is (A’), i.e. the principle that licenses the move from the personal to impersonal and so makes manifest (B’)’s anti-egoistical implications.
But this is to mistake expository proximity for argumentative proximity. For (as Darwall recognizes) all (A’) does is to make manifest work that (B’) has already done. This is so because (B’) doesn’t exclusively state the unobjectionable conceptual truth that judgment and justification are connected. In addition, it question-beggingly asserts that normative judgments give whoever make them reasons to act. (This is the effect of (B’)’s “a person’s acceptance of the judgment is sufficient to … justify his desiring what it recommends,” and it is the effect Darwall calls attention to when he emphasizes “any.”)
In short, in addition to a conceptual truth about the connection of judgment and justification, (B’) states the substantive proposition that if one person has a reason, everybody does. But the substantive claim just is a denial of egoism.