Laws of Nature and Cosmic Coincidences

Abstract

In the past I have argued for the reality of the of physics follows. I have argued that the multifarious irregularities at the observational level can be explained by simple regularities at the theoretical level: it is too much to believe that the universe is just there were electrons, protons, etc. I have said that to believe this as if story would be to believe in a vast cosmic coincidence at the observational level. Appeal to simplicity of explanation suggests that we should believe full-bloodedly in the reality of the theoretical entities or (to cope with possible theory change in the future) something rather like them. (That is, the predicate '. . . is an electron' must be true of real entities, or at any rate approximately true of them.) My position contrasts with that of Bas van Fraassen, who has the idea that theories need not be true but need only be realisable in some model that ensures their empirical adequacy.2 Van Fraassen does not actually deny the reality of theoretical entities, but holds that his theory, which says that what is important in scientific theories is not truth but empirical adequacy, is truer to the actual procedures of scientists. He also holds that his view has an advantage in probability, the disjunction of (1) 'Electrons are real and the theory of them is empirically adequate', and (2) 'Electrons are not real but the theory of them is empirically adequate', cannot have less probability than the first disjunct (1) on its own. I shall attempt to answer this last probabilistic consideration later, but let me pose the main problem I wish to discuss in this paper by saying something else that I am strongly inclined to hold true. When I think of laws of nature I am strongly inclined to think of them extensional propositions, so that the difference between a law of nature and a mere accidental generalisation (e.g. that all the screws in my tool box in the garage are made of brass) is