Abi Rajeev

An epistemically self-aware agent should hold a view about how reliable its own judgment is, and should discount its judgments accordingly. But the agent’s assessment of its own reliability is itself one of its judgments, and so must be discounted by exactly the reliability it purports to establish. This paper formalizes the resulting self-referential structure. Modeling self-assessment as a self-trust function mapping candidate reliability estimates to the credence they warrant, I prove three things: (i) coherent self-assessments always exist — they are fixed points of the self-trust function, guaranteed by elementary topology; (ii) they need not be unique — natural self-trust functions admit multiple equilibria, including a “confident” and a “skeptical” fixed point; and (iii) distinct equilibria can be internally indistinguishable: two agents with identical evidence, identical reasoning, and identical experiences may occupy different self-trust equilibria, and no reflection either can perform will reveal which one it occupies. The trap, then, is not incoherence — Humean skepticism about reason turns out to be too strong — but radical underdetermination: self-calibration has solutions, but choosing among them is not an epistemic act the agent’s own faculties can underwrite. I distinguish the trap from nearby results on higher-order evidence and anti-expertise, examine four escape routes (track records, other minds, pragmatic vindication, and default entitlement), argue that each at best relocates the fixed point rather than verifying it, and draw consequences for epistemic humility and for the self-evaluation of artificial reasoners.

The metaphysics of dispositions has been built almost entirely on examples — fragility, solubility, conductivity — in which the disposition is treated as a standing property that persists unchanged through its manifestations, or is destroyed all at once. But a vast class of real capacities fits neither profile: they are worn by their own exercise. A blade dulls with each cut, a battery discharges with each use, attention fatigues with each demand, trust erodes with each test, a ligament weakens with each load. This paper argues that such consumable capacities constitute a distinct and philosophically neglected category, and that the standard machinery of the dispositions literature — conditional analyses and their fink- and mask-repaired descendants — delivers systematically wrong verdicts about them. I develop a positive framework, the Wear Calculus, on which a capacity is a family of dispositions indexed by a wear state, exercise is a transition on wear states, and familiar notions (exhaustion, fatigue, regeneration, maintenance, overuse) receive precise definitions. The framework yields a five-fold taxonomy of capacities by wear profile, a sustainability condition relating wear to regeneration, and a reframing of maintenance and repair as metaphysically substantive operations rather than mere causal background — extending the small but growing literature that treats repair, not creation, as the normal mode of persistence for both artifacts and organisms.