More precisely, initially, the principle was articulated as “evidence that there is evidence for p is evidence for p”. The intended application of this initial principle (to be discussed below) suggests the following first-pass: (EEE1) If E supports the claim that (some) S possesses evidence which supports p, then E supports p. For the sake of initial discussion, I will make the following precise assumption about “evidential support”: (R) E (evidentially) supports p iff E is positively (epistemically) probabilistically relevant to p. Assuming (R), there are simple counterexamples to (EEE1). Example 1. A card c is sampled at random from a standard deck. Let: (E1) c is a black card, (E2) c is the ace of ♠, (p) c is an ace. And, assume that John knows (precisely) which card c is. This is a counterexample to (EEE1), given (R). . .