Reframing the Problem of Epistemic Peer Disagreement