Abstract

Whenever a fact P grounds another fact Q, one may ask why that is so. Karen Bennett and Louis deRosset independently argue that grounding facts—such as the fact that P grounds Q—are always grounded in their grounds-part. Bennett calls this the view that grounding is superinternal. My aim in this paper is to argue that grounding is not superinternal. I will do so by showing that superinternality, together with some widely accepted formal features of grounding—namely, transitivity and necessitation—yield implausible claims about how necessities are explained. Then, I will discuss how my argument compares with Dasgupta's argument against superinternality.