Abstract: Proponents of inductive risk make three claims: (1) When accepting or not-accepting a hypothesis, a scientist qua scientist must weigh the risks of error. (2) In weighing the risks of error, a scientist must resort to non-epistemic utilities. (3) Epistemically, these (non-epistemic) utilities ought to merely play an indirect role.
To make sense of (1)-(3), I will draw on recent discussions in epistemic utility theory concerning epistemic risk and Permissivism. Presupposing a Lockean Thesis, epistemic risk requires us to make a decision as to whether to believe, (suspend judgement on,) or disbelieve a proposition in light of the risks of error. These risks of error can be of two kinds: believing a false proposition or not believing a true one. I will introduce three types of Permissivism within this epistemic risk framework that look particularly promising as a model for inductive risk. The first concerns the permissiveness of different decision rules (risk inclined, risk neutral, risk averse) while the other two types “encode” Permissivism in the epistemic utilities (how much one disvalues one type of error over the other).

I will ultimately dismiss the decision-rule type of Permissivism as a suitable model of inductive risk, and argue that the remaining two types of Permissivism both are attractive options for the proponent of inductive risk to endorse.