I have been reading Foley's paper "What is wrong with reliablism?" and it got me to wondering whether or not there is a conceptual connection between rationality and deceit. Consider a case in which a certain paranoid-schizophrenic Mr. Z believes that whenever a black cat crosses his path, an alien abduction will occur near Roswell, NM. For fun, I go out and get a bunch of black cats from the local shelter. Every few hours I let one go right in front of Mr. Z with the intention of getting him to believe that such an abduction is occurring every few hours (and, perhaps, that alien activity is on the rise). I succeed. Have I succeeded in deceiving Mr. Z?
My sense is that I have not. While I have led Mr. Z to have certain beliefs I have not deceived him. [It is useful to compare here Grice's discussion of showing Ms. A (a Canadian) a photograph of Mr. B showing undue familiarity with A's wife. Grice says that in such a case that, even though I have led A to believe something, I have not meant that thing by showing her the photograph.] I am not sure what would account for this if not for the fact that deceit, properly so-called, requires that the deceivee be rational. This makes some sense, since deceit is clearly an evidential concept: to deceive is to give misleading evidence for a certain conclusion.
What is going wrong in the case of Mr. Z is that I am not really giving Mr. Z evidence for abductions, but only as it were ersatz evidence. I am feeding him experiences that I knew he will take to be evidence for abductions, but is not really evidence for abductions.
If this is right, then we might strengthen Foley's conclusions: in the Evil Demon world, I am not only rational, but it is a conceptual necessity that I be so.
(Somebody has probably made this point somewhere. If anyone has a pointer, I'd appreciate it.)