8 - Agent Reliabilism and the Relevant Sense of “Relevant Possibility”

Summary

Recall the skeptical argument (D3) from Chapter 2. The major premise of that argument was that good evidence must discriminate among alternative possibilities. In other words, my evidence for a belief p can give me knowledge only if it would cause me to believe that p is true in cases where p is true, and also would not cause me to believe that p is true in cases where p is false and some alternative is true. But, the argument continues, since sensory appearances do not discriminate in this way between what we believe and various skeptical scenarios, sensory evidence cannot give rise to knowledge of the material world.

I suggested that a solution to (D3) is to deny the assumption that evidence must rule out all alternative possibilities to generate knowledge. The idea was that evidence must discriminate among relevant possibilities, but not all possibilities are relevant. In particular the possibilities that I am a brain in a vat or the victim of an evil demon do not seem to be relevant ones. Of course if I were a brain in a vat or the victim of a deceiving demon, then I would not know what I think I do, but so long as these remain mere possibilities they fail to undermine my knowledge.