So I was stuck on a bus from Ghent to Sopot for around 20 hours. I had on me my laptop and a bunch of papers to read, among them Leitgeb's What is a self-referential sentence? Critical remarks on the alleged (non-)circularity of Yablo's paradox (2002), Logique et Analyse 177-178, 3-14. It's a fun paper and I really got into it. To the extent that I decided to spend a few hours typing up a note about this stuff. Here is a fairly sketchy first draft. Abstract below. All comments welcome.
Abstract. Leitgeb (2002) objects against the clarity of the debate about the alleged (non-)circularity of Yablo's paradox, arguing that there are actually two notions of self-reference and circularity at play. One, on which Yablo's paradox is not circular, is defined via the reference of the constituents of a sentence, and another, on which the paradox is circular, is defined via syntactic mappings and fixed points. More importantly, Leitgeb argues that both definitions aren't satisfactory and that before we can undertake a serious debate about the circularity of Yablo's paradox we first need to clarify the notions involved. I will focus on Leitgeb's criticism of the first definition and will argue that the problems arise not as much on the level of our definition of circularity as on the level of our definition of reference of sentences (aboutness). Leitgeb's main worry is the failure of a requirement called `Equivalence Condition', which says that if a formula is self-referential, any formula logically equivalent to it should also be self-referential. I will argue preservation under logical equivalence is unreasonable with respect to self-reference, but is indeed needed with respect to aboutness. Since Leitgeb's own tentative notion of aboutness doesn't satisfy the requirement, I will suggest another approach which fixes this problem. I also explain why the intuitions that circularity should satisfy the equivalence condition are misled. Next, I argue that the new notion of aboutness is not susceptible to slingshot arguments. Finally, I compare it with Goodman's notion of absolute aboutness, emphasizing those features of Goodman's approach that make his notion inapplicable in the present discussion.