Abstrakt
This paper defends realism about beliefs by presenting the rational indispensability argument, which claims that beliefs are indispensable for explaining how norms of rationality shape human behaviour. I begin by examining the normative aspect of belief ascription and the essential connection between beliefs and rationality highlighted in logic and cognitive psychology. Mild anti-realism, as articulated by Posłajko, allows beliefs to exist for explanatory purposes but denies them natural properties, claiming they are not causally efficacious or strictly individuatable. Against this, I argue ? drawing on Wundt?s argument about the relationship between normative disciplines and descriptive sciences ? that if rationality norms influence human reasoning, then beliefs must be real in a robust, naturalistic sense: causally efficacious and individuatable. I adapt the Quine-Putnam indispensability argument from the philosophy of mathematics to argue that our best theories of rationality require ontological commitment to beliefs as real entities. I conclude, contra Posłajko, that recognising the explanatory role of rationality norms commits us to a realist metaphysics of belief.

