I think knowledge entails truth, but I like to keep an eye out for uses of ‘knows’ and its cognates that suggest otherwise.
This one just poppped up on Digg today.
First sentence of the article: “Contrary to what was known, all octopuses are venomous, a new study finds.”
What was previously known? I am inclined to posit an error theory here in one of two ways. Either (a) what was previously known is not in fact logically contrary to the findings (ie., some octopuses are venomous), or (b) there is a level confusion going on, for what was previously known was only believed to be known and what one believes one knows is not truth entailing.
My understanding is that the author of the article thinks that the following proposition was known.
(A) It’s not the case that all octopuses are venomous.
And that (A) is now not known, and not because it changed in truth value – but because of some new evidence we have concerning octopuses that were previously thought to be venomless.
So, if the first sentence were true, then when it was known, it would be a case of knowledge of a falsehood.
I’m actually inclined to think that this is not sufficiently weighty enough to conclude that knowledge doesn’t entail truth. Like you, I favor some explanation like your (b).
Yeah, I think error theory (b) is probably best. The author just wasn’t thinking clearly when writing that sentence.
Do you know this collection of non-factive uses of ‘knows’ ?
http://www.cassetteradio.com/hazlett/nonfactive.html
Cf. also :
http://www.cartoonbank.com/item/39782
h/t : Julien Dutant
http://julien.dutant.free.fr/blog/index.php?2007/08/26/222-savoir-est-il-factif