Haruka Iikawa

Robert Brandom’s inferentialism notoriously entails meaning holism, which has often been seen as unacceptable because it seems to make communication impossible. This paper aims to improve Brandom’s conception of communication as “navigation-across-perspectives” to reconcile meaning holism and the possibility of communication. The conception proposed here entails keeping track of speakers’ own and the other’s scores of commitments and entitlements. I argue that the whole of commonly endorsed inferences in each practice should determine the contents of utterances and those of the commitments of the participants. The local norm defined by such inferences is holistic and can make sense of each other’s commitments. Contents or meaning can be primarily determined at the level of each practice, not that of an individual, the whole community, or the objective world.

Abstract‘The other’ is one of the mysteries of modern philosophy. Since the other is thought to be essentially different from the self, how we can understand each other is a difficult problem. In the first place, what does it mean to understand the other? I address this question from the perspective of normative inferentialism, by explicating what it means to understand the other's beliefs and actions. I propose that we should distinguish between attributional and fundamental understanding. While attributional understanding serves to specify the other's ways of thinking, fundamental understanding is an endorsement of them. Contrasting these two types of understanding gives us a rational picture of how people understand each other.

The traditional debate over theories of reference of natural kind terms faces a serious dilemma. On the one hand, although direct reference theory, or the causal–historical analysis of reference to natural kinds, is still highly influential in the philosophy of language, there is a notorious “qua” problem: direct reference theory cannot uniquely determine the referents of natural kind terms. On the other hand, the standard descriptivism does not accommodate our externalist intuition. We propose temporal externalist descriptivism, where relevant future theorists determine references by past speakers. This position makes the externalist intuition about reference compatible with descriptivism about natural kind terms by ensuring the possibility of mistakes in the past and present practice. It also provides a defense of the controversial claim of temporal externalism based on a descriptivist analysis of reference. While causal–historical analysis focuses on inheritance through history, a reference has both a past and a future. In this sense, reference to natural kinds is essentially a provisional phenomenon.

The rule-following paradox represents one of the most fundamental forms of skepticism, as it suggests the impossibility of possessing any intentional content. Despite numerous solutions proposed since Kripke’s interpretation of Wittgenstein, a definitive resolution remains elusive. This suggests that some commonsense notions regarding rules or semantic content may need to be reconsidered. Some commentators advocate abandoning the conception of absolute determinacy in meaning, which holds that every logically possible application is definitively right or wrong. There are countless possible instances of the application of rules, but as finite beings, we cannot classify every one of them as either right or wrong. Embracing this perspective, I defend a version of radical conventionalism, asserting that whenever practitioners agree on the specific steps for applying expressions, their agreement establishes both rules and semantic facts. A common objection to this idea is that it fails to account for the possibility of errors because we might agree on incorrect points or disagree on correct ones. To address this objection, I propose that semantic facts could be determined retrospectively; subsequent practices institute the semantic facts about their antecedent practices by refining and overriding the semantic facts established in the past. Semantic facts are essentially indeterminate because the application of expressions without agreement in practice is neither correct nor incorrect; however, they are historically determinable.