Defeasible Inference and Defeasible Reference

In this paper, we elaborate on some of Mikael Karlsson’s ideas on non-deductive inference and show their possible applications to philosophy of science and philosophy of language. We refer to two papers by Mikael that have been very influential on us: the first cites some history concerning the idea of non-deductive inference and its application to normative reasoning (Karlsson 1995), and the second suggests a new perspective on induction as defeasible, formally characterizable inference (Karlsson 2001). We will discuss what Mikael calls the “peculiar looseness of non-deductive inference”, that is, a looseness grounded on “default” or defeasible assumptions. He discusses this problem in the context of metaethics and topics in philosophy of law, as well as in natural science; we will discuss the same problem in relation to two sides of communicative competence: the ability to make inferences and the ability to refer. We aim to show that a fundamental feature of our communicative competence is the peculiar looseness of two aspects that we might call “defeasible inference” and “defeasible reference”. Concerning the first aspect, we will briefly present a classical problem in philosophy of science, namely, the way we revise our beliefs in the face of a contradiction; concerning the second aspect, we will hint at a classical problem in philosophy of language, namely, the way we are able to refer using inaccurate descriptions.