In my view, all students should be given an initial opportunity to pursue the science track as far as it goes. But for those who quickly decide that track isn't for them, a different style of teaching is in order.

So this is my attempt to give a preliminary - probably far too crude - account of how philosophy by showing can really teach us. The attempts we make to work through problems by reasoning always presuppose starting points, and even the most self-critical philosophers adopt some of those starting points simply by picking them up from the social environments in which they grow up.

I'm a pluralist about perspectives on literature. There seem to me to be all sorts of illuminating ways of responding to major literary works, some of them paying considerable attention to context, others applying various theoretical ideas, yet others focusing on details of language, or linking the work to the author's life, or connecting it with other works.

So is fighting incompleteness the source of artistic neurosis? I doubt it. At most, this would apply to artists who deal with particular kinds of problems. I don't think we should think of Haydn or Mozart or Dickens or George Eliot in these terms.

To my mind, Death in Venice represents an enormous advance in Mann's literary development, not simply for the commonly appreciated reason that he crafted a superbly supple and elegant style, apparently well suited to the kind of prose Aschenbach is supposed to write.

When I try to outline the history of ethical life, it's sometimes possible to find evidence for a hypothesis about how important transitions actually went. Often, however, that isn't so. There are many facts about human life in the Paleolithic we're never likely to know.

It's not at all a bad idea for scientific questions to be chosen because a democratic deliberation would identify them as important for people's lives.