Lisa Cassell, Dept. of Philosophy, University of Maryland, Baltimore County—It’s a widely held idea that logic is normative. But defending this idea is complicated. On a particularly strong view, to say that logic is normative is to claim that it can be analyzed in normative terms. MacFarlane (2004) famously argues, along such lines, that we can gain insight into the nature of logical consequence by exploring its relation to coherence constraints. However, skepticism about the normativity of coherence makes this strategy less promising than it might first have appeared.

This discussion appeals to the idea that we can understand formal systems as modeling epistemic situations—by analogy with the way models are used in the sciences—to provide the sort of account of the normativity of logic described above. My analysis highlights a form of arbitrariness our logics share with our coherence constraints. In so doing, it secures their epistemic credentials by establishing the legitimacy of normative modeling.