In the second place, the view in question unduly limits the province of ethics. When we are told that actions of certain kinds ought to be performed or avoided, as, for example, that we ought to speak the truth, or that we ought not to steal, we may always legitimately ask for a reason, and this reason will always be concerned, not only with the actions themselves, but also with the goodness or badness of the consequences likely to follow from such actions. We shall be told that truth-speaking generates mutual confidence, cements friendships, facilitates the dispatch of business, and hence increases the wealth of the society which practices it, and so on. If we ask why we should aim at increasing mutual confidence, or cementing friendships, we may be told that obviously these things are good, or that they lead to happiness, and happiness is good. If we still ask why, the plain man will probably feel irritation, and will reply that he does not know. His irritation is due to the conflict of two feelings—the one, that whatever is true must have a reason; the other, that the reason he has already given is so obvious that it is merely contentious to demand a reason for the reason. In the second of these feelings he may be right; in the first, he is certainly wrong. In ordinary life, people only ask why when they are unconvinced. If a reason is given which they do not doubt, they are satisfied. Hence, when they do ask why, they usually have a logical right to expect an answer, and they come to think that a belief for which no reason can be given is an unreasonable belief. But in this they are mistaken, as they would soon discover if their habit of asking why were more persistent. (§ 2 ¶ 1)

It is the business of the philosopher to ask for reasons as long as reasons can legitimately be demanded, and to register the propositions which give the most ultimate reasons that are attainable. Since a proposition can only be proved by means of other propositions, it is obvious that not all propositions can be proved, for proofs can only begin by assuming something. And since the consequences have no more certainty than their premises, the things that are proved are no more certain than the things that are accepted merely because they are obvious, and are then made the basis of our proofs. Thus in the case of ethics, we must ask why such and such actions ought to be performed, and continue our backward inquiry for reasons until we reach the kind of propositions of which proof is impossible, because it is so simple or so obvious that nothing more fundamental can be found from which to deduce it. (§ 2 ¶ 2)