We typically take a theory’s predictive power to indicate representational success. In this essay I argue that this is unjustified, particularly in the case of physical theories. Looking to Gödel’s theorem as a guide, I show that it demonstrates how a theory can prove sentences that are true ‘merely within’ the theory - not true of the theory’s subject matter. Given this, I explain that in cases where we don’t have a clear pre-theoretic knowledge about a theory’s subject matter it can be difficult or impossible to know into which category a given sentence falls. Applying this to physical theories, I first explain how this implies certain limitations on what we can learn from classical and relativistic representations of physical systems and laws. I then look to quantum mechanics, where a complete obscurity of the theory’s subject matter makes it nearly impossible to learn anything from how it represents the universe.
Michael Dascal