Tim Maudlin - On the Methodology of Actual Physics
Physicists and philosophers often allow themselves the luxury of contemplating the methodology of a sort of idealized physicist. One such tempting model of how physicists make predictions is provided by Laplace's (or more accurately Bošković's) demon: the complete physical state of the universe at a moment is fed into some fundamental dynamical equation and then one calculates what will-or might-happen. Of course, everyone knows that this is an idealization. The requisite initial condition cannot, in fact, be known. And even if it were, the calculation could not be done. So arriving at actual predictions must involve idealizations and simplifications. But the extent and nature of those idealizations and simplifications has not, I think, been properly acknowledged, especially in the context of quantum-mechanical predictions.
I will consider the problem at a somewhat abstract level, and then make specific remarks about predictions of arrival-place and arrival-time predictions that are based in quantum theory. There, the conceptual foundations of the predictive methods are more shaky and contestable than is generally recognized.
This is wonderful stuff so far, and I'm glad I caught its brief appearance on the front page. Thank you for posting it!