for “virtue.”2 In the aboriginal tongues of the highlanders of Central India “there seem to be no expressions for abstract ideas, the few such which they possess being derived from the Hindí. … . The nomenclature of religious ceremony, of moral qualities, and of nearly all the arts of life they possess, are all Hindí.”3 On a strict examination of the language of the Tonga Islanders, Mariner could discover “no words essentially expressive of some of the higher qualities of human merit, as virtue, justice, humanity; nor of the contrary, as vice, injustice, cruelty, &c. They have indeed expressions for these ideas,” he adds, but these expressions “are equally applicable to other things. To express a virtuous or good man, they would say, tangata lillé, a good man, or tangata loto lillé, a man with a good mind; but the word lillé, good (unlike our word virtuous), is equally applicable to an axe, canoe, or anything else.”4 Of the Australian natives about Botany Bay and Port Jackson Collins wrote, “That they have ideas of a distinction between good and bad is evident from their having terms in their language significant of these qualities.” A fish of which they never ate, was wee-re, or bad, whereas the kangaroo was bood-yer-re, or good; and these expressions were used not only for qualities which they perceived by their senses, but for all kinds of badness and goodness, and were the only terms they had for wrong and right. “Their enemies were wee-re; their friends bood-yer-re. On our speaking of cannibalism, they expressed great horror at the mention, and said it was wee-re. On seeing any of our people punished or reproved for ill-treating them, they expressed their approbation, and said it was bood-yer-re, it was right.”5
1 Hyades and Deniker, Mission scientifique du Cap Horn, vii. 251.
2 Powers, Tribes of California, p. 22.
3 Forsyth, Highlands of Central India, p. 139.
4 Mariner, Natives of the Tonga Islands, ii. 147 sq.
5 Collins, English Colony in New South Wales, i. 548 sq.
Considering, moreover, that even the European languages make use of such general terms as “good” and “bad” for the purpose of expressing moral qualities, it seems likely that, originally, moral concepts were not clearly differentiated from other more comprehensive generalisations, and that they assumed a more definite shape only by slow degrees. At the same time we must not expect to find the beginning of this process reflected in the vocabularies of languages. There is every reason to believe that a savage practically distinguishes between the “badness” of a man and the “badness” of a piece of food, although he may form no clear idea of the distinction. As Professor Wundt observes, “the phenomena of language do not admit of direct translation back again into ethical processes: the ideas themselves are different from their vehicles of expression, and here as everywhere the external mark is later than the internal act for which it stands.”6 Language is a rough generaliser; even superficial resemblance between different phenomena often suffices to establish linguistic identity between them. Compare the rightness of a line with the rightness of conduct, the wrongness of an opinion with the wrongness of an act. And notice the different significations given to the verb “ought” in the following sentences:—“They ought to be in town by this time, as the train left Paris last night”; “If you wish to be healthy you ought to rise early”; “You ought always to speak the truth.” Though it may be shown that in these statements the predicate “ought” signifies something which they all have in common—the reference to a rule,7—we must by no means assume that this constitutes the essence of the moral “ought,” or gives us the clue to its origin.
6 Wundt, Ethik, p. 36 (English translation, p. 44).
7 Cf. Stephen, Liberty, Equality, Fraternity, p. 343 sq.
Discarding all questions of etymology as irrelevant to our subject,8 we shall, in our analysis of moral concepts, endeavour to fix the true import of each concept by examining how, and under what circumstances, the term expressing it is generally applied. We shall restrict ourselves to the principal, typical terms which are used as predicates in moral judgments. If we succeed in proving that they are all fundamentally derived from either moral indignation or moral approval, there can be no reasonable doubt as to the origin of the rest.
8 The attempt to apply the philological method to an examination of moral concepts has, in my opinion, proved a failure—which may be seen from Mr. Baynes’ book on The Idea of God and the Moral Sense in the Light of Language.
The tendency in a phenomenon to arouse moral indignation is directly expressed by the term bad, and a disposition of mind which is characterised by some special kind of badness is called vice. Closely allied to the term “bad” is the term wrong. But there is a difference in the use of these words. Whilst “bad” may be applied both to a person’s character and to his conduct, only his conduct may be said to be “wrong.” The reason for this is that the concept of moral wrongness is modelled on the idea of a moral law, the breach of which is regarded as "wrong.” And, by laying down a moral law, we only enjoin a certain mode of conduct; we do not command a person to have a certain character.
The moral law is expressed by the term ought, a term which, in modern ethics, generally occupies a central position among moral predicates. The notion which it embodies is frequently looked upon as ultimate and incapable of analysis—“too elementary” (to quote Professor Sidgwick) “to admit of any formal definition.”9 This view, I think, instead of simplifying the matter, has been one of the chief causes of the prevailing confusion in ethical thought.
9 Sidgwick, Methods of Ethics, p. 33.
Far from being a simple notion, “ought” appears to me clearly decomposable, even though it have a special flavour of its own. First of all, it expresses a conation. When I feel that I ought to do a thing, I experience an impulse to do it, even though some opposite impulse may finally determine my action. And when I say to another man, “You ought to do this, or that,” there is certainly implied a purpose to influence his action in a certain direction. In the notion of duty, the ethical import of which is identical with that of “ought,” this conative element is not so obvious.
Closely connected with the conative nature of “ought” is the imperative character it is apt to assume. But, though frequently used imperatively, “ought” is not necessarily and essentially imperative. Even if the “ought” which I address to myself, in a figurative sense, may be styled a command, it is hardly appropriate to speak of a present command with reference to past actions. The common phrase, “You ought to have done this, or that,” cannot be called a command.
The conation expressed in “ought” is determined by the idea that the mode of conduct which ought to be performed is not, or will possibly not be, performed. It is also this idea of its not being performed that determines the emotion which gives