RESUMO 21 | COMUNICAÇÕES

Tolerance is not Charity: The Differences Between Carnap’s Metaontological Stance and Quantifier Variantism

CORRÊA, Cleber de Souza (USP), Brasil

Many views within contemporary metaontological debates are heirs to Carnap’s “Empiricism, Semantics, and Ontology”. One of these views is Eli Hirsch’s Quantifier Variantism (QV). According to QV, there is no single best language with which to describe the world. Quantifier variantists interpret (some) ontological disputes as involving parties whose languages employ different concepts of “there exists something”. In one language, the mereological nihilist can truly say “tables do not exist”, and, in another language, the mereological universalist can truly say “tables do exist”. For QV, the ontological dispute between these two parties is merely verbal, in the sense that both languages are equally suitable to describe the world. A fundamental driving principle in QV diagnosis of first-order ontological disputes is the principle of charity: each party in an ontological dispute should be interpreted as speaking truly in their own idiolects.

Carnap’s metaontological stance has some similarities with QV in the sense that both views take some first-order ontological disputes as matters to be solved by paying attention to language, and not by some sort of philosophical inquiry into the nature of the world. For Carnap, questions concerning the existence of a kind of entity (mereological sums, say, or numbers) can only be meaningfully settled by looking at how the linguistic framework deals with the entity in question (how its vocabulary work, what its rules of inference allows one to assert, etc.). In Carnap’s parlance, such questions are internal to frameworks. The answers to these first-order ontological questions are uncomplicated in the sense of requiring no proper ontological investigation.

Some philosophers (Ted Sider is the main contender here) argue that Carnap was committed to QV in his (Carnap’s) diagnosis of first-order ontological disputes. Some others (Amy Thomasson, for instance) deny this. In this presentation, I concur with Thomasson’s conclusion, but I focus on aspects of the differences between QV and Carnap’s metaontology that she does not touch upon (or touch upon only insufficiently) in her disagreement with Sider. The main points I deal with are the following. First, Carnap’s tolerance towards frameworks is nothing like the principle of charity that Hirsch employs in order to deflate first-order ontological debates as merely verbal. Carnap recommends tolerance towards framework proposals as a way of preventing that frameworks that can be useful in domains of theoretical inquiry be outlawed on ontological grounds (like, for instance, Quine was eager to do with relation to the framework of abstract entities in his nominalist period). Carnap was not concerned with charitably interpreting apparently diverging ontological parties. Second, Carnap disagrees (or he would, I argue) with a fundamental tenet of QV, namely, the idea that there is no single language with which to best describe the world. But Carnap would reject this idea not because he thought there is a single best language to describe the world; rather, he would repudiate the tenet because the view that the world can be correctly described at all (either in one or more than one language) would be considered philosophically misguided from his metaontological perspective. This tenet is of a piece with the kind of metaphysical concerns Carnap was at pains to avoid from interfering with the engineering and improving of linguistic frameworks.