At the pre-Central celebration of Jennifer’s book, listening to an interesting problem now being discussed by Lizzie Fricker. Here ’tis:
when you don’t fully trust a source, you tend to encode your source in the belief. Instead, of believing what they say, you believe something like: joe says (thinks) that p.
Now, take contextual terms. Lizzie’s example is ‘pricey’. Now when I fully trust you, I can’t just believe what you say, but have to encode your standards in some way. How? A clumsy way is to invent a term: ‘pricey-according-to-Joe’. I don’t think I do that, but something like this has to be done.
But then the handling of contextuality in testimony looks uncomfortably close to encoding a lack of full trust in one’s sources, and that would be a bad result. (It is also interesting to think about what happens to epistemic talk if one is a standard contextualist about ‘knows’.)
Maybe the discussion will show a way out of the problem.