Haktan Akcin

This thesis examines what form of naturalized metaphysics can remain defensible once both the authority of mature science and the disruptive force of scientific theory change are taken seriously. It rejects, on the one hand, analytic metaphysics insofar as it proceeds as a philosophy-first inquiry insulated from scientific development, and, on the other, the anti-metaphysical conclusion that abandoning such autonomy leaves no legitimate role for metaphysical reflection. The thesis first argues that metaphysics cannot responsibly proceed through intuition, common sense, or other philosophy-internal methods when investigating the structure of the actual world. It then turns to scientific realism, defending a qualified success-based realism while acknowledging the challenge posed by the history of theory change and the Pessimistic Meta-Induction. In response, it argues that realism must be historically disciplined and that Direct Structural Realism provides the most plausible basis for such a position. The dissertation’s constructive claim is that structural realism remains incomplete unless retained structure is understood modally. Structures preserved across theory change are not merely formal residues of discarded ontologies, but encode objective patterns of constraint, dependence, and possible transformation. On this basis, the thesis develops a modal version of Direct Structural Realism and argues that modality provides the most promising positive task for naturalized metaphysics. It concludes by defending a science-led, realist, structurally informed, modally articulated, and historically disciplined metaphysical naturalism, according to which metaphysical commitment should be directed toward historically retained modal structure rather than attached uncritically to the ontology of current science.

I argue that the central conflict between epistemic and ontic versions of structural realism concerns whether it is possible to read off ontology from current fundamental science. Even if we assume that structures are metaphysically superior to objects, the possibility of reading off ontology from current fundamental science remains unjustified. I show that the conclusion as regards to the reading off ontology in the ontic version is already assumed in one of the premises; hence the argument begs the question. As a result, the problem of ontological discontinuity implied in the pessimistic meta-induction argument remains intact in ontic structural realism.