(We'll set aside the question of why someone would want to develop a theory of knowledge separate from the sciences or how any theory of knowledge developed in isolation of the sciences could succeed in explaining anything. In my view, the whole epistemological enterprise began in speculation and broad strokes out of necessity, before there were any scientific methods to investigate what kinds of knowledge we have or how knowledge works. Truth be told, however, these days we can conceive of the methods but we might not possess them or minimally do not have the means to achieve what we want to achieve when finding out knowledge. But this is just a hunch, a provisional position, and maybe I would give it up after more consideration.)
Klein admits that any theory of knowledge that he or anybody would try to develop "cannot encompass all our uses of 'S knows that p,' simply because that expression functions in so many various ways." That is, to say a person knows something or any similar sentence, if given fuller definition, might not cover all cases. But Klein hopes that it will at least cover the kinds we'd be interested in for a theory of knowledge.
To deal with the Gettier cases, Klein proposes a principle that does not allow something to be called knowledge, and it is a principle he thinks we would all agree with. He calls it the felicitious-coincidence principle, and elucidates it thusly:
[I]f S's evidence for p and a description of some of the particular circumstances in which S believes that p are such that it would not be reasonable to expect that p is true (based upon S's evidence), even if p is true, So does not know p.I know that's a mouthful and maybe to someone not interested in philosophy or not used to reading philosophy will like seeing a sentence like that. (Heck, to be honest, I enjoy reading and studying philosophy, and I absolutely hate hate hate to see a sentence like that.) So I'll break it down and translate it like this.
My translation: Imagine that you believe something but the situation in which you find yourself or your evidence from that situation isn't quite adequate to know it. In that situation, even if what you believe then and there is true, and have good reasons for it, but there is somehow contrary evidence that the situation provides, then you don't really know what you thought you believed.
Think about this principle with respect to the Gettier cases mentioned in the last post. Gettier argued that his two cases show that justified true belief can't be what knowledge is because in his two cases people have justified true belief but not knowledge. In the first case, a man thinks the dude with two quarters in his pocket is going to get the job, and he knows that man with two quarters to be a man other than himself. But to his surprise, the man who has this belief gets this job and little did he know he had two quarters in his pocket.
In the other case, somebody makes a wild assertion like "My friend owns a Ford or else he's from Boston." Little did he know, despite having good reasons for believing his friend owned a Ford, he's friend does not in fact own a Ford, but his friend is from Boston. But the fact that the man was right about this is mere coincidence.
Both of these kinds of cases the felicitous-coincidence principle is supposed to block out of hand. Klein thinks that we can accept this principle and so not accept that these cases of justified true belief amount to knowledge.
So what is knowledge? Klein gives this answer. Knowledge, propositional knowledge at least, is when
(i) p is true;Basically, Klein maintains the first part, then. Knowledge is (i) true (iii) justified (ii) belief, and here I use in parentheses the lower-case numerals to denote how Klein's view matches up with the traditional view to the time of Plato. But then Klein adds that there ought not be any contrary evidence to somebody's belief because if there were, then that would provide reasons for someone to stop believing what they do and believe something else.
(ii) S believes p at t1;
(iii) p is evident to S at t1;
(iv) there is no true proposition such that if it became evident to S at t1 p would no longer be evident to S.
There are some other interesting things Klein writes in his article "A Proposed Definition of Propositional Knowledge," especially at the end regarding the definition's neutrality between Cartesian and anti-Cartesian sentiments toward knowledge. Klein writes:
Cartesian doubt, in its strong form, must grant that a certain proposition p is evident, given all the standard tests for p, but yet it must maintain that it remains possible to doubt that we know p. Whereas the anti-Cartesians seem to be maintaining that, if p is evident as a result of all the standard tests being applied to p, then it becomes gratuitous to doubt that we know p.I want to point out again my response to this Cartesian/anti-Cartesian debate, and insist that the very fact that both positions look both like polar opposites and attractive positions is sign that the debate/distinction is hopelessly confused. The big question is again "In what sense is 'knowledge' being used?' Surely we can all be anti-Cartesians in ordinary life where we use our standard measures for evidence. But in another sense we are very much Cartesians when science is pursued quite rigorously and it's assumed that the standard ways to measure or obtain evidence is admittedly limited, and so in a sense we are all fallibilists regarding what there is and what we can know about what there is.
This is just a part of the debate I'd like to stress and part of why I think that it's so hopeless in being fought along Cartesian or anti-Cartesian lines. So this is an interesting path here that Klein is tracing, putting knowledge out there as justified true belief provided there's no evidence to the contrary.
Do you think this gets us out of the Gettier mess? Tune in, same Bat Time, same Bat Channel.