Richard on my Verificationism

Richard takes up my verificationism here. He proposes that it's principle V (scroll down a bit below here) is wrong because it's too strong. He says I require

he requires that the base facts themselves be epistemically accessible, at least in the limited sense that we can envisage possible evidence for and against them. But while this starts in the right place, I think the added restriction goes too far. We should merely require that the base facts be comprehensible, in the sense that someone could understand the difference between scenarios where they do or do not obtain.

However, it seems to me that "understanding" requires epistemic content. If you understand what a z is, then you are able to explain to someone who doesn't understand, and that explanation entails saying how to recognize when there is a z. "I understand what a z is, but I just can't tell when I've got one, that's all." That sentence is suspect, in my view. Consider:

A: "A square without corners or sides? I don't get it. How do you tell when you've got one?"
B: "Well, it can fit into a square hole, but not into a round hole of its size."
A: "But how can you tell that it can or can't fit?"
B: "Dunno. But you do understand me, right?"

Similarly:

C: "A zombie? I don't get it. How do you tell when you've got one?"
D: "Well, he acts like a conscious person, and with all the brain states, only he has no consciousness."
C: "Eh? How would you know that he isn't conscious? Even he can't tell."

I'm still suspicious of things masquerading as "understanding" that are nothing of the kind. I think the mind is very big and complicated and that in its spaces it can play tricks on itself that it itself cannot very well detect. It can seem to itself to understand undetectable ethers, undetectable separate universes, undetectable lacks of consciousness, etc., all the while understanding none of these things.

Richard needs to say what is the difference between "comprehensible" and "suspiciously incomprehensible" other than a feeling of "I've got it!" I don't trust that feeling. I need epistemic, evidential terms.

Two side points:

1. As for Richard's worry that his view of moral facts supervening on base facts is "slippery", I say don't worry, Richard. What needs to be added to actions (say, of your Hitler) is that they are performed in a society of people with dispositions to disapprove of those actions (sentiments of a certain kind). You have to include the subjective base facts about sentiment, as Hume suggested. On a planet of cruel nasties who kill for fun, but where no one has a nature such that he is ever even slightly disposed to have concern for others' welfare, none of the nasty things done is wrong. I'll pursue it this week on Philosoblog.

2. Please keep in mind that in order for a term to be meaningful on V, it only need be the case that one can say what the evidence for the term's correct application would be. One needn't actually be in possession of that evidence. You don't have to have evidence that there is a presently-causally-isolated universe for that term to be meaningful. You only need to have the physics (or whatever) that would at least roughly give an idea of what such evidence would be. Similarly for brains-in-vats (in a comment thread a few posts below.)

UPDATE: I wonder whether Richard's view entails that one can understand an incoherent term (such as "square with only three sides"). After all, "I know what it means, I understand it, but I just don't see how it could be rendered coherent" sounds Richardesque. I suppose he might say that the incoherence precludes comprehension. But that seems arbitrary. Why won't he also allow that evidential vacuity precludes comprehension, as well? Again I say that the mind is large and its imagination powerful. It can imagine a logically impossible thing (especially if it's in a sort of dreamy state: try it, you can do it. Or if the contradiction is buried deep enough, you can accomplish the task in a clear-headed state.) It can mistake this for comprehension. It can imagine the correct application of an evidentiallly vacuous term (e.g., "zombie") and mistake this for comprehension, too. How can Richard distinguish these two, such that in the latter case there is in fact no mistake?