Abstract According to Paul Boghossian’s ‘Inference Problem’, rule-following generates a regress because it involves inference, which itself involves rule-following. Boghossian claims the Problem leads to an ‘antinomy of pure reason’, showing that ‘we both must—and cannot—make sense of the notion of someone’s following a rule’. I argue that there is no antinomy by proposing a solution to the Problem based on the idea that following a rule consists in acting on a commitment to exercising one’s knowledge of how to follow it.
Abstract This paper introduces a novel non-Humean bundle theory based onto dispositions, according to which material objects are constituted by structural dispositions—second-order, polyadic properties that relate various other first-order properties to a structured whole. I show that the theory effectively resolves Van Cleve’s problem of change in bundles (Van Cleve 1985) and contributes to the ongoing debate between modal dispositionalism and essentialism. The proposed theory serves as a powerful tool for dispositionalists seeking to argue that only dispositional facts are fundamental. Unlike existing accounts, and especially Vetter’s (2015, 2021), this approach does not rely on primitive objectual essences. Instead, dispositions depend on the structural disposition that constitutes the bundle of which they are part. A central implication is that modal dispositionalists can now account for metaphysical modalities, including difficult de re possibilities of nonexistence, without invoking objectual essences. This challenges the prevailing view that essences are more fundamental than dispositions and provides an alternative for the foundation of modality.
Abstract Can a computer programme like ChatGPT perform speech acts such as making a promise, offering a guarantee, or firing an employee? This is an increasingly pressing question as AI technologies supplement or replace human communication, especially in the workplace. Philosophers typically maintain that artificial speakers at best perform ‘proto’ or ‘quasi’ speech acts. In contrast, I argue that in some contexts artificial speakers can perform speech acts in precisely the same way as humans. This is because the rules and norms governing speech act performance (for AI and for us) are more context-variable than it is usually thought. This is most clear in highly regulated contexts, where speech act norms are often determined by legislation or policy, but is also a feature of ordinary communication. The question then becomes not simply whether artificial speakers can perform speech acts, but whether we should allow them to.
Abstract It is often argued that maximizing normative theories are inappropriate for bounded agents. I argue that leading motivations for theories of bounded rationality are substantially compatible with some traditional maximizing normative theories. This finding expands the audience for theories of bounded rationality by making room for advocates of maximizing normative theories, and has implications for knowledge-action principles, the relationship between descriptive and normative theorizing, and the definitions of central tenets of bounded rationality.
