It is characteristic of realists to separate ontology from epistemology and of idealists to mix the two things up. By ‘idealists’ here I am mainly referring to the British neo-Hegelians (‘objective idealists’) but the charge of mixing up ontology and epistemology can be made against at least one ‘subjective idealist’, namely Bishop Berkeley, as his wellknown dictum ‘esse ispercipi’ testifies. The objective idealists rejected the correspondence theory of truth and on the whole accepted a coherence theory. The qualification is needed here because H. H. Joachim, in The Nature of Truth, found the coherence theory unable to deal with the problem of error.

This paper is a defence of the combination of three theses. First, selfknowledge is a requirement of good lives, but it must include knowledge of our limitations. Second, these limitations are set by certain unavoidable conventions whose observance is another requirement of good lives. Third, Sophocles's Oedipus the King is a moral text of the highest significance, for it contains a profound account of why self-knowledge must include knowledge of unavoidable conventions limiting human aspirations.

For a long time theories of history of the speculative sort have been out of favour. Accounts of the whole sweep of human history, like Hegel's, or even of more limited historical cycles, like Spengler's or Toynbee's, have been found much too grand for the workaday historian and have smacked too much of apriorism for post-positivist philosophy. Consequently, few take them seriously or treat them as more than fanciful aberrations which may serve as useful examples of how not to proceed in history or philosophy.

[Colour is king in our innate quality space, but undistinguished in cosmic circles.] Most philosophers would agree with at least the second half of Quine's dictum. It is indeed on the general view wrong to believe that, as qualities, colours are extra-mentally actual in even the humblest role. Mind-independent material things have on the general view powers to cause sensations of red or blue, but if, in [sensations of red or blue], [red] and [blue] name qualities, we are not to believe that these qualities are possessed by things causing the sensations. My first thesis, defended in section 2, is that partly because we do count colours as eminent among qualities, we would on reflection want it to be true that some things have such qualities when they are not perceived. It would therefore be sad subsequently to discover the wrongness of believing that this is how things are. My second thesis, defended in sections 3 and 4, is that there is in fact no danger as yet of this kind of disappointment. So far, the philosophers have not shown that, if we believe that colour qualities exist as contents of experience, we ought not also to believe that things have these qualities when they are not perceived. One might of course deny that colour qualities exist even as contents of experience, so that the desire for them to be mind-independently exemplified evaporates on the realization that it lacks an intelligible object. Our pre-scientific concept of red, according to Armstrong, is, apart from being the concept of something falling under a determinable, [all blank or gap].

The Humean legacy in moral philosophy has given rise to the following familiar divergence of views:

1. Moral questions are questions of feeling. Rational discussion can occur given the acceptance of unreasoned commitments to values or principles, but it must not be thought that these values or principles can themselves be justified rationally. Moral disagreements may be resolved through persuasion, but it is appropriate for rational persons to be permanently irreconcilable in their moral views.

2. Moral questions are questions of fact and/or logic. They are therefore in principle resolvable through rational discussion, like any other questions of fact and/or logic. The idea that rational persons should be permanently unable to agree on moral questions, despite indefinitely lengthy discussion and enquiry, is unintelligible.

In discussing Disembodied Persons we need to confront two problems:

A. Under what conditions would we consider that a person was present in the absence of the normal bodily cues?

B. Could such circumstances arise?

The first question may be regarded as epistemic and the second as metaphysical.