What this high-lights is that we rely on a whole heap of assumptions in deciding what is true, or may be true, or isn’t true. But when we are reminded of them, we pause and question them and this can change our minds. Gathering evidence for all of them is impractical, so it is hard to see that this can be avoided. One might draw two conclusions from this.
1. There is an element of luck in all our knowledge claims, which makes it more difficult to distinguish lucky guesses, which most people would classify as not knowledge, from knowledge. To be honest, I’m not sure what to say about this.
2. Knowledge can be fallible. Or rather, it is obvious that knowledge claims are fallible. The question is whether a knowledge claim that fails because it is wrong is still to be classified as knowledge or whether it should be re-classified as belief, (until and unless an infallible, conclusive claim is established). It may be that this is just a linguistic question and doesn’t really matter. But if one accepts that knowledge (as opposed to a claim to knowledge) is fallible, I don’t see what distinguishes it from belief, so I’m inclined to the latter option.
This case is is too simple to do more than start considering the problem. We should consider a variety of cases. For example, consider another risk that Al has not taken into account, that his car may be smashed by a falling meteorite. Does that mean he doesn’t know about his car? Again, suppose Al’s car has a sun-roof. Rain is not certain, but has been forecast. Does Al know his car is safe?
Each of these cases is different. Do the differences make any difference? I’m not sure, but we should consider whether the question “Does s/he know?” is appropriate in the sense that it can really be answered.
Vogel seems to think that there's a difference between Al's and Betty's epistemic circumstances. — Agent Smith
The problem is set by the similarity between the two, which suggests that their epistemic classification should be the same, and the difference in their circumstances (i.e. the fact that Betty’s car has been stolen), which strongly suggests that it is different. So the fact that Betty’s car has been stolen makes me classify her as believing and rather than knowing. Conclusion – knowledge is not just about the psychological state of the knower. Hence, for example, how confident they each are is irrelevant.
So, on the ground, as a person in the world of those scenarios, I wouldn't consider having said one or both of them knew where there car was as some big error should it turn out that the car is not there. Because when I say 'know' it doesn't mean 'cannot possibly be revised stuff'. — Bylaw
It is certainly true that philosophers often get very dogmatic, especially about knowledge. I deduce that you would say that both of them know, except that Al’s knowledge is correct (so far) and Betty’s knowledge is not correct and hence should be revised. I think that does reflect how we actually use the word. When she finds out her car has been stolen, she will know that. Would that be a fair summary?
May I ask what would you say to this possible outcome? Her car is stolen but only for a joy ride, and the thief, being cautious, took the trouble to return her car to its place and managed to do so before she came back to retrieve it. So she never knew her car was stolen and believed it had been safely parked all the time.
But knowledge about the world is never infallible. What we hold to as something known can change over time. — Fooloso4
You’re right. The parallel universe is a bit over the top. I based this on Jennifer Nagel’s version of it (I couldn’t find the original). You’re also right that we only know what the example chooses to tell us, and that is usually very limited, which can be frustrating.
Surely, we can sometimes be conclusively right. For example, one could say that Al and Betty both have fallible knowledge. But then, when they get back to the cars and drive back home, they will know conclusively, won’t they?
The situation would change if the crime rate was so high that it was reasonable to expect that one's car would be stolen, in which case one would presumably take extra precautions, or expect trouble. — unenlightened
The situation would change if the crime rate was so high that it was reasonable to expect that one's car would be stolen, in which case one would presumably take extra precautions, or expect trouble.
Your point that Al has no justification for believing that his car has been stolen is a good one. Until she discovers that her car has been stolen, the same is true of Betty, of course. That’s a key problem, of course. Justification can be less than conclusive.
An induction based on logic, reason, and memory, but an induction none-the less. — Philosophim
Quite right. I’m not sure whether you think that induction can never result in knowledge because it is always uncertain or not. Certainly, in this case, it is an induction that can be replaced by certainty – when they get back. Though actually, I would say that unless one embraces full-blown scepticism, induction can also justify certainty. What is less clear is when we reach that point. I suspect that philosophers would be much less optimistic than everyday people.
Pretty soon you'll get to the Gettier problem. — T Clark
That's unlikely. I've been there and done that. But epistemology is perhaps in an impoverished state because all the big questions have been spun off into distinct philosophical fields. If you're not interested, that's fair enough.
I have to say, the only alternative to the JTB that I've come across is the "knowledge first" idea. That might have something so recommend it, but I haven't caught up with it yet.