Nor is it in popular thought and speech only that we see this. If not wholly and definitely, yet partially and by implication, the belief is expressed by moralists. In his _Methods of Ethics_ (1st Ed., p. 6) Mr. Sidgwick says: "That there is in any given circumstances some one thing which ought to be done and that this can be known, is a fundamental assumption, made not by philosophers only, but by all who perform any processes of moral reasoning."[K] In this sentence there is specifically asserted only the last of the above propositions; namely, that, in every case, what "ought to be done" "can be known." But though that "which ought to be done" is not distinctly identified with "the right," it may be inferred, in the absence of any indication to the contrary, that Mr. Sidgwick regards the two as identical; and doubtless, in so conceiving the postulates of moral science, he is at one with most, if not all, who have made it a subject of study. At first sight, indeed, nothing seems more obvious than that if actions are to be judged at all, these postulates must be accepted. Nevertheless they may both be called in question, and I think it may be shown that neither of them is tenable. Instead of admitting that there is in every case a right and a wrong, it may be contended that in multitudinous cases no right, properly so-called, can be alleged, but only a least wrong; and further, it may be contended that in many of these cases where there can be alleged only a least wrong, it is not possible to ascertain with any precision which is the least wrong. A great part of the perplexities in ethical speculation arise from neglect of this distinction between right and least wrong--between the absolutely right and the relatively right. And many further perplexities are due to the assumption that it can, in some way, be decided in every case which of two courses is morally obligatory. § 101. The law of absolute right can take no cognizance of pain, save the cognizance implied by negation. Pain is the correlative of some species of wrong--some kind of divergence from that course of action which perfectly fulfills all requirements. If, as was shown in an early chapter, the conception of good conduct always proves, when analyzed, to be the conception of a conduct which produces a surplus of pleasure somewhere; while, conversely, the conduct conceived as bad proves always to be that which inflicts somewhere a surplus of either positive or negative pain; then the absolutely good, the absolutely right, in conduct, can be that only which produces pure pleasure--pleasure unalloyed with pain anywhere. By implication, conduct which has any concomitant of pain, or any painful consequence, is partially wrong; and the highest claim to be made for such conduct is that it is the least wrong which, under the conditions, is possible--the relatively right.