This essay develops and defends a diagnosis of (Carroll in Mind 4( 14):278-80, 1895)'s regress of the premises according to which the moral of the regress is that arguments are constitutively presuppositional. It is argued that this diagnosis allows to vindicate the key insights of the rule-following account of the regress, while overcoming the main difficulties that the rule-following account faces.
IntroductionCarroll (1895)'s famous regress of the premises arises in the course of an argument between Achilles and Tortoise that has the following structure. Suppose p and if p then q. From that, Achilles would really want to conclude that q. Tortoise would not allow it: q can be concluded-she objects-only if it is true that if p and if p then q then q.Hence, Achilles is led to suppose, in addition, that it is true that if p and if p then q then q. From that together with the earlier premises, Achilles would want to conclude that q. Tortoise would not allow it: it can be concluded that q-she objects-only if it is true that if p and if p then q then q. No provision of further premises will convince Tortoise to accept the conclusion. An infinite regress ensues. I am grateful to Harold Hodes, Arc Kocurek, and Tim Williamson for helpful discussion about these topics, as well as to two anonymous reviewers for comments that have greatly contributed to improving this essay.