Landesman's twelfth chapter nicely poses the impasse between Moore and the Skeptic. It becomes clear that the difficulty of adjudicating the dispute turns on the ambiguity of the ground-rules: On whom should the burden of proof fall? Must Moore prove that he knows his premise, here are two hands, or must the Skeptic prove that he does not know it?

By analogy, consider a court of law. A criminal defendant is not required to prove her innocence; rather the prosecution must show, beyond reasonable doubt, that the defendant is guilty, otherwise she goes free. But our problem is compounded. Not only must we determine in advance (and in some principled way) whether Moore or the Skeptic plays the role of prosecutor or defendant, we must determine whether this is a criminal, a civil, or some other kind of trial, and who precisely is qualified to sit on the jury. In a civil trial the prosecutor (or more properly plaintiff) must merely convince the jury that a preponderance of evidence, more than half of it, is against the defendant. The criminal standard is beyond reasonable doubt, meaning all but completely certain. Perhaps what we could call an epistemic trial demands a still higher standard, that of no possible doubt.

Perhaps we can stretch this metaphor, considering the alternative burdens and standards of proof, and find some clarity about the matter.