This book is concerned with two problems of moral philosophy. The first of these has of ten been treated as the problem of finding a common criterion for all right action; or again, as the relation between the right and the good. Considered a little more generally, it is the attempt to systematize or render coherent the content of moral judgements. In referring to earlier theories, I have dealt with this first problem largely as the battle-ground between utilitarianism and deontology. As between those two theories, I side with deontology; but I accept the criticism that it is unsatisfactory to leave principles of obligation in 'an unconnected heap', and I try to unify them by means of the Kantian principle of treating persons as ends-in-themselves. The attempt to do so requires an interpretation of that principle, and in my interpretation I reach a view of moral obligation which owes much to Hume and Adam Smith, and something to Martin Buber.
The second problem is the nature of moral judgement. This is the battle-ground between naturalism and nonnaturalism or ethical intuitionism. I use the term 'naturalism' to include any theory which explains the meaning and function of ethical words wholly by reference to human nature, and which denies that the facts to be taken into account include entities or characteristics transcending human thoughts, conations, and feelings. I call naturalistic not only a theory which says that ethical words describe human attitudes, but also a theory which says that the function of such words is to express or evoke human attitudes. Mr. R. M. Hare, in his recent book, The Language of Morals, confines the name to the first of these two kinds of theory, on the ground that only the first kind commits what Professor Moore called the naturalistic fallacy. Although the expression 'naturalistic fallacy' was introduced by Professor Moore in PrincipiaEthica . . .