Recently, several authors have claimed to have found graph-theoretic counterexamples to the Principle of the Identity of Indiscernibles (PII). In this paper, I argue that their counterexamples presuppose a certain view of what unlabeled graphs are, and that this view is optional at best.