« Pure Emotives? | Main | Constructing Attitudes »



That's very interesting regarding the demon. I guess I'm just not convinced that you can't deceive an irrational person. That is, if by 'irrational' you simply mean 'believes on the basis of too little evidence' then I certainly think they can be deceived. However, if by 'irrational' you mean 'believes an explicit contradiction' then I'm not sure that they can be deceived since, epistemically, any belief in any proposition p follows from their contradiction.


I agree that the generalization may be hasty. But I just don't see what accounts for intuition in this case--it sort of surprised me.

I do think the following is true, or something like it. One can't be deceived about C unless one is able to use evidence (and not merely ersatz evidence) for the conclusion that C. So the onus here is really on using evidence for C rather then merely using something as evidence for C. So what is the distinction? I'm not sure, but I doubt it can be spelled out in a way that is independent of an appeal to what a rational agent would do. But it can't just be a simple claim like: e is evidence for C relative to epistemic circumstances s iff a cognitive agent could rationally use e to support C in s. This won't work unless one can spell out the cognitive circumstances just right (e.g., by excluding the possibility of including the background belief that black cats indicate alien abduction in the example.)

Another thought is that evidence gets cashed out in terms of meaning (either natural or non-natural). After all, the fact that rocks fall doesn't mean that the universe has a telos, though I think it is plausibly enough evidence for that claim. But if we try to modalize the claim in some way, we will be back to needing to rule out worlds in which black cats do mean alien abductions are occurring near Roswell.

I don't know. Maybe I am making too much of the example.


Thanks for the interesting post! Here's my 2 cents:

I agree that there's a conceptual connection between deceit and rationality; but I think you have deceived Mr. Z, because he is rational in the relevant sense – not simply minimally, in having beliefs at all, but, more importantly, in drawing the correct inference (albeit to a falsehood) given his beliefs. He believes, falsely (let's hope), that if BC then AA. When he sees a BC, then, he quite naturally infers AA, as would we all (again, given the belief). It's the clear rationality of that inference that justifies our attribution to him of the resulting belief that AA in the first place (presumably yelling "AA! AA!" is not enough). In fact he better infer that AA, given BC, if we are to continue to attribute belief in (if BC then AA) to him. (Yes this is Davidsonian talk, but bear with me.) What would you have him do in order to count as "rational" in the relevant sense? After all, he doesn't have enough evidence to give up (if BC then AA), because he can't observe (not AA, even though BC): he's not in Roswell to make that observation (hard to observe a negative anyway).

The only irrationality here is (presumably) in his previous acquisition of (if BC then AA), which you do not discuss (and he undoubtedly has a number of other such beliefs and quasi-doxastic states). But that irrationality has nothing to do with the present deceit. Given Z's apparent rationality (at least in the present instance), your deceit would work just as well, and in the same way, if (if BC then AA) had been rationally inferred, or for that matter if it were true (that is, if "natural" occurrences of BC, as opposed to your "articifial" ones, actually were evidence of AA).

You say: "I am feeding him experiences that I knew he will take to be evidence for abductions, but is not really evidence for abductions." Why is this not deceit? Aren't all claims of the form "I am feeding him experiences that I knew he will take to be X, but is not really X" claims to deception? You distinguish between this deception (that there is evidence for AA) and deception that AA, but I don't see how this distinction is warranted. After all, he believes that AA because of (indeed, because of a rational inference from observing) what you did to deceive him. (And I'm not sure what Grice's example does here; he's making a point about meaning. Your showing A the photo doesn't mean ("non-naturally," the way an utterance would) "B is messing around with your wife." Similarly, your kitty-liberating activity doesn't mean "there goes another abductee" – but that doesn't mean you're not deceiving Z into believing it.)

And besides, this is a funny way of putting it. Given his beliefs, Z's experiences of BC (which you have deceived him into believing to be naturally occurring) are indeed evidence for AA. I don't want to use the concept of evidence as you seem to be doing here: such that it is an "objective" matter what is evidence for what, independent of what any particular agent believes. After all (for us Davidsonians, again), the very content of those propositions (ie those between which we are discussing the evidentiary relation) is conceptually dependent on the agent's beliefs (as well as the interpreter's, and the community's), and the perceived evidentiary relation between them. That's the holism in Davidson's picture. (If I understand the terms, which I may not, this does not commit us to either "internalism" or "externalism" about epistemic justification; but that's another story...)

Based on (your second thought in) your follow-up, it looks like you agree, sort of. But I would go farther and say that this dooms the very idea of distinguishing between "evidence" and "ersatz evidence". After all, we still have the idea of insufficient evidence (and false belief) - these should be enough.

The newest volume of Davidson papers (Truth, Language, and History) has one of my faves, "Meaning, Truth, and Evidence," which may help here. I don't have it in front of me, but I remember him as criticizing Quine on something like this point. Maybe we can tempt Professor Ludwig to comment ... ?

I'm not sure how this affects the demon case, but again, the basic Davidsonian line (not like I'm orthodox or anything!) is that in order to attribute false belief (and therefore deception) at all, we must attribute a background of true belief (and rational links between them) as well. Interestingly, on my reading anyway, this does not rule out the Matrix possibility (if it's coherent to begin with, that is).

Whew! More than 2 cents after all. Thanks again for the interesting discussion!

conscientious objector

I'm having a hard time understanding the difference between evidence and "ersatz evidence." You're saying you're having a hard time, too, so I guess I can take comfort from that. So, let me ask two questions.

First, is it possible to tell the difference between evidence and ersatz evidence without know the truth of the proposition in support of which they are provided? For instance, are astronomical observations that support the existence of "dark matter" evidence or ersatz evidence?

Second, an intuition test. The Allies conducted intense disinformation campaigns prior to D-day with the intention of causing the Germans to believe that the Normandy landings were a feint and that the real invasion would take place at Calais. They did so by creating circumstances that they knew the Germans would take to be evidence for a Calais landing. Was the evidence presented by the Allies ersatz evidence? If it is not, how is that situation different from your play with poor Mr. Z. If it is ersatz evidence, are you really comfortable claiming that, whatever else the Allies did, they did not "deceive" the Germans?


First let me thank Duck for a thoughtful and interesting post to which I unfortunately won't respond in any detail. Thanks, as well, to CO.

I take that I framed this post incorrectly. What I thought was most interesting was the intuition that I haven't deceived Mr. Z. Those who I have consulted around here share that intuition. I am not sure from their posts whether or not CO and Duck don't have the intuition or have it but also have sophisticated theoretical reasons for discounted (or, ugh, don't regard intuitions as having evidential force). My positive comments were mostly off the cuff suggestions about how to account for the intuition. So the primary question is just what accounts for the intuition, assuming you have it?

I am adding a new entry on the positive proposal. Mostly because I need a new post, but also because I would like to keep the discussion of the ersatz evidence seperate from further suggestions about what accounts for the intuition, assuming you share it.

conscientious objector

For what it's worth, I don't have sophisticated theoretical reasons for discounting the thought experiment (though Duck does make my heart go pitter-pat when he utters the magic word "Davidson"). And no, I'm not discounting the value of intuition. FWIW, I don't share your intuition, though. I've always associated deceit with the intention to cause false beliefs. As you are intending to cause Mr. Z to come to a false belief, I'm inclined to say that you've deceived him.

That's not to say that I have the theory that "x deceives y iff x intends to cause y believe p, and x intentionally acts in a way that causes y to believe p, and x intended that act to be the means by which his intention to cause y to believe p was to be effected, and p is false" (or whatever). My momma didn't raise no stupid children (foolish, maybe, but not stupid), and I'm not going to paint a target on my forehead. I'm just telling you my intuition, which says to look at the perpetrator, not the victim, to determine whether the victims' delusions are the result of deceit.


I wanted to add that I do have the intuition that in the example I have tricked Mr. Z into believing certain things. So however I conceive of deception it is a "stronger" concept than merely tricking someone. Deception is, no doubt, a particular mode of tricking, but the two concepts are not in my view coextensive.

Is that right? If I trick Z into believing that p does this entail that I decieved Z? It seems to me not.

Incidentally, here is a kind of off-beat example, especially for you Davidson apostles ;). In flyfishing there are two sorts of flies, naturals and attractor patterns. The first are designed to closely mimic some sort of aquatic insect and are tpically used to "match the hatch" (i.e., when bugs of just that type are momentarily abundant). Attractor patterns, on the other hand, are designed to be flashy and catch the attention of a fish and cause them to strike at it (and it is questionable that they are striking at it because they take it to be food of some generic sort, rather than just reacting to its presence). Now (I have to confess that I find this weird), but I have the sense that when I am using naturals it is correct to say that I am decieving the trout. Contrariwise, when I am using an attractor, I do not think I am decieving the trout, but merely tricking them.

The comments to this entry are closed.

August 2011

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad