It’s probably obvious to readers of this blog that I’m relatively unmoved by arguments on behalf of the JTB analysis of knowledge. In case it’s not, however, I hereby confirm the point. So what follows is slightly tongue-in-cheek, but not wholly. The idea is to see if the rejection of Gettier examples to the JTB account can be made more palatable, in the following way.
Let’s distinguish ordinary justification from settled justification. You have settled justification when your total evidence justifies your belief and also justifies you in thinking that your belief is justified, true, and ungettiered (i.e., your justification is not accidentally related to the truth of your belief in the way that leads to the intuition that you don’t know). If you have settled justification, your total body of evidence justifies for you that further inquiry into the matter would reveal, at most, misleading evidence against what your evidence justifies. (When I use ‘justified’ and its variants, I mean only ordinary justification.)
Suppose then that we interpret the JTB account of knowledge in terms of settled justification.
There will still be intuitions that true beliefs that are settled-ly justified are not knowledge, since one can have justification for thinking that a belief is ungettiered and be wrong. But if you’re swayed at all by Brian Weatherson’s arguments that we should accept a JTB theory in spite of contrary intuitions, then, I say, the JTB theory you should accept is the one that favors settled justification over ordinary justification. For there will be far fewer gettier intuitions to deny, and one won’t have to say that anyone knows their lottery ticket is a loser.
Even for those not especially swayed by Brian’s argument, and you think there is some fact of the matter that distinguishes knowledge from justified true belief even of the settled kind, there is a value argument for linguistic change here. True belief justified in the settled way gets us as valuable a state as we can have, and the difference between it and any concept of knowledge that is distinguished from it is a difference that makes no axiological difference. So let’s hang onto the honorific character of ‘knowledge’, using it to talk about what matters from an epistemic point of view with regard to individual propositions, and let the remainder fall where it may. Or, if you like me tend toward semantic skepticism, let’s adopt the JTB theory, and explain away the contrary intuitions in terms of historical accretions of connotation to a term that aren’t really part of the semantic content of it anyway. Either way, from now on, our slogan with be “knowledge just is JTB”.
OK, I admit that the above doesn’t even satisfy Kent Bach’s sincerity norm of assertion, but it might be interesting to see at what point we want to resist the above.