A paradoxical situation exists in the study of Wittgenstein. There is a sharp disagreement in the interpretation of his thinking about the concept of following a rule. According to one group of philosophers Wittgenstein's position is that this concept presupposes a human community in which there is agreement as to whether doing such-and-such is or is not following a particular rule. A second group of philosophers hold that this interpretation of Wittgenstein is not merely wrong, but is even a caricature of Wittgenstein's thought: for when Wittgenstein says that following a rule is ‘a practice’ he does not mean a social practice, he does not invoke a community of rule-followers, but instead he emphasizes that following a rule presupposes a regularity, a repeated or recurring way of acting, which might be exemplified in the life of a solitary person. On the first interpretation it would have no sense to suppose that a human being who had grown up in complete isolation from the rest of mankind could be following rules. On the second interpretation such isolation would be irrelevant.

The other night I had a very strange, and strangely coherent, dream. Socrates and Meno appeared to be arguing with each other in my presence. They talked English, I suppose, since I clearly thought I followed them; but I seem to remember that Greek words occurred from time to time. When I woke it seemed to me that the dream had some bearing on disputed matters of Platonic interpretation, so I shall try to reconstruct it here. Meno speaks first:

Tell me, Socrates, do you think that the just and the admirable are the same thing?

I should hardly think so; for we call many things admirable—for example a young lad's cheek—which we would hardly call just; and perhaps vice versa. But I am only guessing, of course, because I cannot know whether A is the same as B unless I know what A is and what B is.

‘There just has to be some connection between thinking and what goes on in my head when I am thinking,’ said someone to me recently. ‘When I am thinking,’ he continued, ‘the thinking is going on in here’, at which he pointed with his forefinger at his head. I indicated mild dissent. ‘You seem to me to be overlooking plain facts,’ he retorted. ‘There are thoughts in my head. Nothing could make me doubt that. If I am given a complicated calculation to do, I might start off by trying to do it in my head (once again, pointing at his head) but, finding it too hard, I might continue the calculation I had begun in my head on paper, on the table in front of me (now pointing at an imaginary piece of paper in front of him). The calculation started off being done in my head and finished up being done on paper.’

The purpose of this paper is to examine some contentions advanced by B. A. O. Williams in his books Descartes: The Project of Pure Enquiry and Ethics and the Limits of Philosophy. In particular I shall be concerned with the claims he makes about the nature of ethics—namely that it cannot be ‘objective’ or ‘realistic’ and that we may not hope for rational convergence in ethical judgments. My claims will be that Williams's case on these matters is importantly unclear and incomplete and that we lack cogent reasons to accept the conclusions he offers us. But a good deal of scene-setting is necessary before we can pose the questions I wish to address.

My concern is to understand how it is that contemplation of the past— better, of this or that preferred past—evokes in some people an impression which is distinctively weird. It is unmistakable; and anyone who has felt it will soon know what I am talking about. What is the impression, and whence the impressionability?

To help identify my concern (and make it seem less eccentric) I shall let it emerge from some highly selective remarks about an issue in philosophy of history which is, by contrast, familiar and respectable: the debate between constructionists and realists. We cannot conceivably have direct acquaintance with, direct access to, the past; by their very nature, past events are over and done with and so unavailable for inspection. This much both camps agree on. However, they differ massively over what follows from this truth. For the constructionist concludes that what he calls the ‘real past’, what actually happened, can play no part whatsoever in historical thought. It is necessarily hidden, and we can have no inkling of it. What, and all, the historian can sensibly claim to know is the ‘historical past’, something which is constituted by and exists only in relation to his thought. Against this, the realist maintains that of course historians do not, necessarily or even typically, constitute the past; rather, they construct accounts of it which will be true if they conform to it as it actually was and false if they do not. And he charges his opponent with a number of fundamental confusions: mistaking accounts of historical events for the events themselves, confusing epistemological matters with ontological, and worst of all equating knowledge with direct perceptual awareness. Now, the realist is, in basics at least, fairly obviously right. And his criticisms are reinforced when we note that constructionists tend to combine with their vision of the impenetrability of the ‘real’ past the thesis that we undoubtedly know that there is a real past, with real people and real events. However, this piece of knowledge must for him, like that of an intelligible world for Kant, ever remain contentless, ‘factually vacuous’.

The existence of evil is compatible with the existence of God, most theists would claim, because evil either results from the activities of free agents, or it contributes in some way toward their moral development. According to the ‘free-will defence’, evil and suffering are necessary consequences of free-will. Proponents of the ‘soul-making argument’—a theodicy with a different emphasis—argue that a universe which is imperfect will nurture a whole range of virtues in a way impossible either in a perfect world, or in a totally evil one. The pain of animals is widely thought to constitute a major difficulty for both of these accounts, for if we ask whether the only evils present in the world result directly from the free actions of created agents, or contribute in some way to ‘soul-making’ of such agents, we are bound to admit that, on the face of it, much animal pain does not.