The appeal in question is explained by reference to how it is that certain types of claim fail to get dislodged by moral theories. This argument is specifically aimed against consequentialism but would effect other theories that, like the consequentialist one, are revisionist of settled moral convictions. The example given is that of making a promise. When a promise is made we assume that it has binding force independently of any notion of optimific outcomes. When we "reflect" on the notion of a promise we do not find any reference in our understanding of its obligatoriness to considerations of the type that would be expected if it were grounded ultimately on consequentialist considerations. This does not entail that we have to reject any appeal to consequence sensitivity but it does show that even if we decide not to follow the prima facie duty of promise keeping that this does not, in itself, indicate that the ground of the duty is in fact a consequentialist one.
Ross also summarises his understanding of the self-evidence that applies with ethics in terms of a remarkable statement that relates ethical statements to a type of accumulative sedimentation when he writes:
We have no more direct way of access to the facts about rightness and goodness and about what things are right or good, than by thinking about them; the moral convictions of thoughtful and well-education people are the data of ethics just as sense-perception are the data of natural science. Just as some of the latter have to be rejected as illusory, so have some of the former; but as the latter are rejected only when they are in conflict with other more accurate sense-perceptions, the former are rejected only when they are in conflict with other convictions which stand better the test of reflection. The existing body of moral convictions of the best people is the cumulative product of the moral reflection of many generations, which has developed an extremely delicate power of appreciation of moral distinctions; and this the theorist cannot afford to treat with anything other than the greatest respect. The verdicts of the moral consciousness of the best people are the foundation on which he must build; though he must first compare them with one another and eliminate any contradictions they may contain. (40-1.)
This remarkable passage has a number of intriguing elements in it. Firstly, and perhaps most overarchingly, it indicates a limitation on the persuasive power of revisionist theories of morality. These theories need to take seriously the sense that there already exists, prior to their being elaborated, a deep moral sense that has been historically generated. There is a similar point to this made in the first section of Kant's Groundwork where Kant likewise seeks to show that there is already contained in general moral consciousness sufficient for the theorist. Secondly, this limitation on revisionist theories, or certainly on their persuasiveness, is not merely stated in favour of conservatism although a first reading of this passage might suggest so. The point is not that existent moral claims be simply and wholly accepted as apt but rather that the aptness of these claims requires reflective consideration in relation to each other (much like in the case of Rawls' notion of "reflective equilibrium"). Thirdly, the revision allowed room by theory needs to be related to that which works through the consistency and coherency of the general body of convictions. So the revision does not merely succeed due to a general argument in its favour but also has to be stable in relation to the reflective self-criticism of the data of theory, namely, the settled convictions of the "best people". The oddity of the appeal finally resides however in this last point, namely, that there are some people ("the best" or those who are "thoughtful and well-educated") better able generically to carry out the reflective process in question. What enables this ability to reside in such people is not stated and surely it would here have been preferable to refer instead to a procedural mechanism rather than a settled group thought to already itself be a locus of wisdom?