Has a moral agent really done his duty when he has done what he wronglybelieves to be his duty? Is it right to act in accord with one's beliefs, even when they are mistaken? Or are we always obliged to perform that act which is objectively obligatory? In some such ways as these the problem as to whether one's ‘objective duty’ or one's ‘subjective duty’ is one's real duty has been posed. It might be argued that the objective view is plainly right and that there appears to be a problem only because moral language is here being misused; that the problem is seen to be no problem once we accurately and fully describe acts based on erroneous beliefs. But this is like arguing that there is no problem as to whether the killing of innocent persons is intrinsically wrong, since we can describe each act involving killing in a long-hand way which circumvents the problem concerning the principle.

An Important distinction between statements of fact and statements of value is widely recognised. Some philosophers are now saying that the distinction has been treated as more determinate than it is, but most philosophers would agree that the distinction is definite and important. The major contributions to Anglo-Saxon moral philosophy of this century have set out to illuminate the nature of this distinction. Ethical statements have been thevalue statements mainly at issue, but on the whole the aim has not been to show wherein they differ from other value statements, but to show what distinguishes them in common with other value statements from factual statements. The characterisations of ethical statements which have become famous areones which if they apply to ethical statements at all apply equally to many(or all) other value statements as well.

In A paper called ‘The Responsibility of Psychopaths’, I think I succeeded in establishing that we cannot rule out a priori the possibility that psychopaths may be shown to be lacking in responsibility. I also examined some arguments that try to show the psychopathto be lacking in responsibility, but I concluded that these arguments were not very successful. In this paper I intend to make and examine some more attempts at showing the psychopath to be lacking in responsibility. But before I do that there is one point to keep in mind.

Like his earlier Language and Philosophy (1949). and Problems of Analysis (1954), Models and Metaphors is a collection of Black's papers unified by the belief that linguistic considerations can play an important part in framing and solving philosophical problems. Broadly speaking, the linguistic approach takes two forms: (1) examining the uses of a word, or of a set of related words, frequently occurring in philosophical inquiries, either for the general purposes of clarification (e.g., of ‘model’, ‘rule’) or as a useful aid to the solution, or dissolution, of a specific philosophical problem (e.g., of ‘implies’ in connection with ‘What is the relationship between an assertion and its presuppositions ?’); (2) considering questions about language, meaning, and related subjects (e.g., ‘Can one draw valid inferences from premises about grammar to ontological conclusions?’, ‘Do the accusatives of meaning-formulae designate anything?’, or ‘Are necessary statements validated by appealing to linguistic rules?’). Obviously, given that the concepts involved are of language or meaning, the former approach can be readily incorporated in the latter.