Thanks to the feedback I’ve gotten already on evidence. Here are three other questions that I’ve already got from looking at those suggestions, and thinking about related stuff.
1) In the literature on testimony, there are several writers who deny that testimony is a source of evidence. Rather, they say, testimony provides a kind of non-evidential warrant.
Now I’m rather sceptical that this is the right approach to testimony. But one thing that’s worth thinking about from a perspective of a theory of evidence is what it could mean for this to be true. On some theories of evidence (on E=K for instance) it’s hard to see how this could even be a coherent position. I suspect it is a coherent position, and that’s a reason to doubt E=K. But turning this into an argument requires saying more about what non-evidential warrant might be, and hopefully saying that will provide some clues as to what evidential warrant might be!
It might be helpful here to think about the kind of warrant we get from doing simple logical inferences. When we validly infer q from p, and thereby come to know q, presumably p is part of our evidence. But I don’t think it is true in general that the fact that p implies q is part of our evidence. If it is, then we should also say that the fact that p, and p implies q, together imply q is part of the evidence, and so on ad infinitum, and that strikes me as absurd. So in inference there are things that we know, and that provide warrant, but which are not part of our evidence, properly construed. So I am sympathetic to the idea of non-evidential warrant, though I’d like to know a lot more about how it is supposed to work.
2) It’s interesting to think about what evidence a person has. And it’s interesting to think about what evidence a person has for different claims H, where H is a claim that the person could be considering. Here’s a third question: Which of those two concepts is primary? I sometimes think E=K is best considered as a view where the concept of evidence for is primary, and evidence simpliciter is a derived notion.
What I mean by this is to say that E is part of S’s evidence just is to say that E is part of S’s evidence for some hypothesis, and the order of explanation/analysis here goes from right-to-left. One question is whether that’s the right way to take E=K. (I think it makes the best sense of some arguments for E=K.) The other question is whether it is the right way to think about the question more broadly. Should we analyse ‘evidence for’ by first offering a theory of evidence, and then asking what of that evidence is evidence for the hypothesis in question, or should we analyse evidence by first offering a theory of evidence for different hypotheses, and then asking what propositions are evidence for some propositions or other? This seems like a fairly deep question, and I don’t know how to attack it.
3) In law, evidence typically does not consist of propositions. In philosophy we’ve focussed our attention on propopsitional evidence. Or at least we have recently. When Quine said evidence was sensory irritations, he presumably was not saying that our evidence consists of facts about sensory irritations. We might have made a deep mistake here; maybe propositional evidence is just a special kind of evidence.