Numerous Things

Abstract

“We have defeasible reason to believe that the conceptual decisions of successful theories correspond to something real: reality’s structure.” T. Sider, Writing the Book of the World [25, pp. 12] This quote voices a very strong form of metaphysical realism. Too strong, I think. My goal in this essay is to make trouble for it. I begin by briefly situating the view historically. I then discuss the relevant notion of success and the relevant notion of a conceptual decision. I go on to point out a rather severe obstacle for the project of reading “structure” off of a successful theory. To this end, I sketch an account of theoretical equivalence, which gives conditions under which two theories are the same, even if they differ with respect to their language or their domain of quantification. From this perspective, a “successful” theory is always equivalent to a similarly “successful” theory that makes different “conceptual decisions.” It’s unclear how to decide which of the theories’ “conceptual decisions” are “structural,” and so it’s correspondingly unclear how to learn structure. Next, I directly question the reasons for postulating facts about “structure” in the first place. One reason is that such facts, if true, offer a tidy solution to a family of puzzles. I examine two of the more well-known puzzles: Goodman’s new riddle of induction and Putnam’s model-theoretic argument for anti-realism. If we resist the urge to dissolve the puzzles with speculative metaphysics, I argue, then we stand to learn some interesting lessons. In the case of the new riddle, we might rethink the view that takes inductive inference as a purely formal endeavor. In the case of the model-theoretic argument, we might rethink the view that takes the world as a domain of objects ready to be mapped onto.