Those committed to (G) think that if there were no such things as principles like (A) then moral thought and judgement couldn't get off the ground. This seems quite reasonable: all that principles do is tell us which features generate what reasons. If we were incapable of formulating any such principles then we would be incapable of telling what reasons there are; and if this is the case, then moral practice would break down would it not?
This, Dancy wants to claim, is not quite right. Principles such as (A) track features which generate a certain kind of reason, or, rather, track features with a certain kind of valency: an invariable valency. To say that a feature has an invariable valency is to say that if a generates a reason with a certain valency in one context, then in every other context in which it appears it must generate a reason with the same valency. That is, in so far as those committed to (G) take it that there moral thought requires the existence of princples like (A), they are committed to a certain thesis about the nature of reasons which Dancy calls atomism:
Atomism: A feature which generates a reason in one situation must generate a
reason with the same valency in any other situation in which it is present.
Atomism can seem quite intuitive. But Dancy thinks that it is false and as such principles like (A) cannot do what they are supposeded to do: track features with invairable valencies. Notice, however, that (G) could still be true. If it were true then Dancy will have put forward a sceptical argument the conclusion of which would be that moral practice is impossible. But, of course, Dancy isn't as sceptic. We will see in the next post that he actually wants to argue for a thesis opposed to atomism which he thinks we can use in an argument for (P), and, hence, show that (G) is false.