Mainstream analytic metaphysics is a priori metaphysics. It is hemmed in by basic assumptions that rest on no more than a priori intuitions. Jaegwon Kim's arguments about causation are a paradigm example of sophisticated arguments with little or no justification from the world as we know it. And Peter van Inwagen's arguments about material objects are motivated by a question that, I think, has no nontrivial answer: Under what conditions do some x's compose an object y? The trivial answers are (...) "always" and "never." Any other answer, including van Inwagen's, seems to me to be arbitrary—and again based on a priori intuitions. To cite two more examples that I find egregious: The assumption that there is a metaphysically important distinction between what is mind-independent and what is not, and the assumption that there is a metaphysically important distinction, among things that are real, between what is fundamental and what is not. A priori intuitions, especially when not shared by philosophers outside a small circle of specially trained people, seem a weak reed on which to hang metaphysical systems. (shrink)

We argue that dualities offer new possibilities for relating fundamentality, levels, and emergence. Namely, dualities often relate two theories whose hierarchies of levels are inverted relative to each other, and so allow for new fundamentality relations, as well as for epistemic emergence. We find that the direction of emergence typically found in these cases is opposite to the direction of emergence followed in the standard accounts. Namely, the standard emergence direction is that of decreasing fundamentality: there is emergence of less (...) fundamental, high-level entities, out of more fundamental, low-level entities. But in cases of duality, a more fundamental entity can emerge out of a less fundamental one. This possibility can be traced back to the existence of different classical limits in quantum field theories and string theories. (shrink)

The discovery of high-level causal relations seems a central activity of the special sciences. Those same sciences are less successful in formulating strict laws. If causation must be underwritten by strict laws, we are faced with a puzzle, which might be dubbed the 'no strict laws' problem for high-level causation. Attempts have been made to dissolve this problem by showing that leading theories of causation do not in fact require that causation be underwritten by strict laws. But this conclusion has (...) been too hastily drawn. Philosophers have tended to equate non-strict laws with ceteris paribus laws. I argue that there is another category of non-strict law that has often not been properly distinguished: namely, minutiae rectus laws. If, as it appears, many special science laws are minutiae rectus laws, then this poses a problem for their ability to underwrite causal relations in a way that their typically ceteris paribus nature does not. I argue that the best prospect for resolving the resurgent 'no strict laws' problem is to argue that special science laws are in fact typically probabilistic, rather than being minutiae rectus laws. (shrink)