Machery has posted an interesting paper over at the Experimental Philosopher's blog and I've banged out some comments in reply. Not terribly well organized, but my worry had to do with the conception of evidence that seemed to figure in his discussion. To bang a drum I've been banging a lot recently, I think we ought to say the following about evidence:
(i) Your evidence consists of facts (it's propositional and it's not evidence if it's not true);
(ii) Your evidence will include any fact you know non-inferentially.
Suppose you combine these two points about the constitution and possession of evidence with the further claims that Machery seemed to want to make about evidence:
(Mi) Your evidence is provided by your judgments about thought experiments.
(Mii) Those judgments are the evidence you have.
I think (Mii) is a category mistake, but I take it that the real idea is that facts about the judgments that you make are the evidence you have when evaluating the merits of, say, consequentialism in light of various thought experiments. So, rewrite (Mii) accordingly.
Some worries. First, do you have any reason for those judgments? If so, what are they? If not, what sort of evidence are they? If I believe that my shoes are filled with rats and I'm aware that I have no reason to think this, this is a reason to believe that it's a good idea to get professional help. No, not help from the exterminator. If you have reasons for these judgments and they're epistemic reasons, shouldn't these reasons be part of your evidence? I think so.
Second, if you combine (Mii) with (i) and (ii), you get all the skeptical conclusions you want without any empirical work at all. You don't need cross cultural surveys to establish skeptical views since it follows from these trio of claims that our non-inferential judgments about thought experiments never themselves constitute knowledge. The only knowledge you could have about the thought experiments would be inferential judgments and I worry that these wouldn't be good candidates for knowledge if none of our spontaneous judgments about cases could constitute knowledge. Happy result, of course, if you're pushing the skeptical line. Cheating, you might think, since the skeptical line is derived from a skeptical conclusion. Skeptical conclusions shouldn't be premises in arguments for skepticism, you might think.
Now, to be fair, the experimentalist skeptics who want to say that our judgments about thought experiments don't constitute knowledge do offer empirical evidence that they take to show that our judgments have what we might think of as epistemically problematic features. For example, our judgments are subject to order effects, there's widespread disagreement across cultures and within the philosophical community, etc... Fine. Why do we think these features are epistemically problematic in the sense that they prevent judgments from constituting knowledge? Take the disagreement issue. On the view I like, your evidence is what you know directly, we can have non-inferential knowledge about thought experiments, we ought to be conciliatory only if someone disagrees with us and their evidence is as good as ours. In the cases described if I'm "in the know", anyone who disagrees with me has evidence that differs in kind from the fabulous evidence I do, so I don't see why my learning of their existence defeats my knowledge or justification.
Alright, so that's not really my view. It's a view, however. Given that view, a reaction to the empirical evidence is that it's interesting but not epistemically interesting to me (i.e., it tells us little if anything about my privileged epistemic position). And what do they have to reject this view? Intuitions about epistemic propriety? Can the skeptics we're talking about appeal to such intuitions?