« Dowd Again | Main | Fantasy Basketball »
Justified True Belief
Nick Kristof in his quasi-blog thingy writes:
Incidentally, when the White House did once raise the issue of my Niger reporting with me, the senior official who complained did not argue that any of this was incorrect. Rather, he noted that Bush's reference in the State of the Union address was to "Africa" rather than to Niger -- and that even if the Niger connection was fraudulent, there were possible linkages to other African countries like Congo that could have made Bush's 16 words technically correct. That was a very flimsy branch, and the official gave up the argument pretty quickly.It's sort of too bad the White House didn't go with this story, because that would have brought the country close to a major national debate about the Gettier Problem:
You have a justified belief that someone in your office owns a Ford. And as it happens it's true that someone in your office owns a Ford. However, your evidence for your belief all concerns Nogot, who as it turns out owns no Ford. Your belief that someone in the office owns a Ford is true because someone else in the office owns a Ford. Call this guy Haveit. Since all your evidence concerns Nogot and not Haveit, it seems, intuitively, that you don't know that someone in your office owns a Ford. So you don't know, even though you have a justified belief that someone owns a Ford, and, as it turns out, this belief happens to be true.It's not quite the same, but it's pretty close.
November 4, 2005 | Permalink
TrackBack
TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345160fd69e200d83424d9f253ef
Listed below are links to weblogs that reference Justified True Belief:
» Jones owns some uranium from Unfogged
Yglesias makes the right call-- Kristof's account of an administration line, viz., even if the Niger connection was fraudulent, there were possible linkages to other African countries like Congo that could have made Bush's 16 words technically correct.... [Read More]
Tracked on Nov 4, 2005 5:19:23 PM
» Jones owns some uranium from Unfogged
Yglesias makes the right call-- Kristof's account of an administration line, viz., even if the Niger connection was fraudulent, there were possible linkages to other African countries like Congo that could have made Bush's 16 words technically correct.... [Read More]
Tracked on Nov 4, 2005 5:23:56 PM
Comments
Didn't you make a similar analogy way back when using barns in a field?
Posted by: Clark | Nov 4, 2005 12:19:56 PM
The probabilty of there being any co-worker who owns a Ford must be greater than the probability that a specific co-worker owns a Ford. Your evidence for Nogot's Ford ownership is also evidence that at least one employee owns a Ford. To whatever extent that you are confident that Nogot has a Ford, you must be MORE confident that ANY co-worker has a Ford.
But, that aside, I think we'd all rather be arguing the Gettier problem. After all, had Saddam actually been obtaining yellowcake from another African country, things would be much better.
Posted by: Silent E | Nov 4, 2005 1:03:44 PM
Hmmm. I'm not sure.
So, the old thinking was - you "know" P if P is true and you have a justified belief that P is true.
But the Gettier problem is: P is true if either A is true or B is true. You have a justified belief that A is true but you do not have a justified belief that B is true. But, in fact, A is not true, although B is true. This means that you did not "know" P.
Now fill in P, A, and B.
From the example:
P = "someone in your office owns a Ford"
A = "Nogot owns a Ford"
B = "Haveit owns a Ford"
But for the Bush/Kristof/Yellowcake problem?
P = "Saddam attempted to purchase yellow cake from Africa" (it can't be that "the British have learned...", right?)
A = "documents showed that Saddam attempted to purchase yellowcake"
B = "linkages to other African countries like Congo showed that Saddam attempted to purchase yellowcake"
But did the Administration ever have a justified belief that the documents showed Saddam attempted to purchase yellowcake? That is, even though it turns out A was not true (because the documents were forged), did the Administration even justifiably believe that A was, in fact, true?
Is there any evidence that "linkages to other African countries like Congo" is true (even if the Administration didn't have a justifiable belief about them)?
Isn't it really the case, contra Kristof, that B should actually be "British intelligence having nothing to do with the forgeries showed Saddam attempted to purchase yellowcake". (Call that B'.) And wasn't the Administration's justifiable belief that B' (i.e., the British intelligence claim) was true, not that A (the forged documents) were true?
So, really, this isn't a Gettier problem at all. In fact, the Administration justifiably believed that B' was true, not A. (The Butler report supports the assertion that the belief was justifiable.) What we don't know is whether B' is, in fact, true.
Posted by: Al | Nov 4, 2005 3:12:40 PM
The Haveit/Nogot example is poor because my statement "Someone in my office owns a Ford" implies the statement "and I know who it is." Try this one:
I see a set of keys that someone forgot in the men's room, and there is a Ford key among them. I conclude, Someone in my office owns a Ford. However, the keys belong to a visitor. However, as it happens, the office manager owns a Ford, the key to which is in her purse.
Posted by: JR | Nov 4, 2005 5:04:48 PM
I don't understand why the "Gettier problem" is considered important. You are using "know" to mean something like "have good reason to believe". Some of the things you have good reason to believe are not true and some are true but not for the reason you are primarily relying on. So what?
Physicists give "work" a technical meaning. They do not then waste time trying to reconcile this technical meaning with the ordinary language meaning. So why should anybody be worrying about the "Gettier Problem"?
Posted by: James B. Shearer | Nov 4, 2005 9:17:40 PM
Al = hairsplitting. Where I come from we call this "making shit up and avoiding eye contact."
Posted by: fnook | Nov 4, 2005 10:45:13 PM
James - I have kind of come to believe that that's basically what analytic philosophy is.
Posted by: Toadmonster | Nov 5, 2005 2:02:14 AM
The Gettier counter-examples are an intuition pump for why "justified true belief" is an inadequate definition of knowledge.
Posted by: washerdreyer | Nov 5, 2005 5:07:25 AM
And analytic epistemologists aren't using "know" to mean just "have good reason to believe"; we're trying to capture what we think is the ordinary use of "know," which requires that at least that the belief be true. To expand on what washerdreyer says, if you describe the Havit/Nogot story (maybe in JR's variant) to people, a lot of people will say, "Wow, you didn't really know that someone in the office owns a Ford, even though it's true and you had good reason to believe it."
And one reason that epistemologists spend time studying the ordinary concept of 'know', I think, is that we figure if this concept weren't useful we wouldn't still be using it. It would have died out like the concept 'witch' (not everywhere, sadly).
That said, I'm currently working on a paper that argues that James's instinct is right--epistemologists should worry about "has a good reason to believe" rather than the ordinary-language meaning of "know." Because basically, if you have a good reason to believe something, it's not important if it's true for a different reason. (This basic argument goes back to Mark Kaplan's 1985 paper, "It's Not What You Know That Counts," but I don't think Kaplan addressed the question of why ordinary people still bother with the word 'know'.)
Posted by: Matt Weiner | Nov 5, 2005 10:28:49 AM
Now that everyone's asleep, I can add that I think Al is right to say that the Administration never had a justified belief that the papers showed that Saddam had attempted to purchase uranium from Niger. But I don't think the British intel ever provided any decent evidence either--Josh Marshall here gives reason to think that the Butler report was politicized and here provides specific evidence that it was deceitful. So I think there was probably never any justification for the "sought uranium" claim.
Posted by: Matt Weiner | Nov 5, 2005 10:33:29 AM
I was taught the working definition of "to know" was:
Alfred knows X if and only if:
1. X is true
2. Alfred believes X
3. Alfred has valid evidence that X is true
The evidence has to be, in truth, valid, not merely justifying.
Of course the trouble with this definition is that "Alfred knows X" can be true but "Alfred knows that he knows X" can be false. For example:
1. There exists a material world.
2. Alfred believes there is a material world.
3. Alfred perceives objects in the material world through his senses, and it is indeed the case that his senses correspond to a material world.
1b. (part of 3 above) Sensory perceptions correspond to the material world.
2b. Alfred beleives his sensory perceptions correspond to the material world.
3c. However Alfred does not have any valid evidence that his perceptions correspond to the material world - they may instead be caused by a Cartesian demon.
Posted by: wml | Nov 5, 2005 2:22:22 PM
And one reason that epistemologists spend time studying the ordinary concept of 'know', I think, is that we figure if this concept weren't useful we wouldn't still be using it
I'll bet anything that this doesn't scratch James's itch. To me, it sounds like he is asking why we should expect a theoretical account of knowing (whatever it might be) to comport with folk intuitions about the meaning of the word "know." Presumably, he doesn't think that ordinary language meanings of "work" are useless for getting around in the world, but obviously that doesn't mean physics ought to bother with describing the logic of such meanings. Therefore, what's different about epistemology?
Posted by: spacetoast | Nov 5, 2005 5:16:08 PM
What spacetoast said.
Also a problem with studying the ordinary concept of "know" is that this concept is fuzzy. One meaning of the word "know" is just "strongly believe" without any requirement that the belief be true or sensible. Consider the saying "It's not what you don't know that will kill you, it's what you do know that isn't so.".
Posted by: James B. Shearer | Nov 5, 2005 8:18:36 PM
Well, as I said above I actually agree with James on this. One of my views is that the concept of 'knowledge' is a vague and in fact inconsistent concept that is useful only because it can stand in for a lot of other more precise concepts, and epistemologists should study the latter. My view is a bit eccentric among analytic epistemologists, though (at least I hope so, it's no fun agreeing with everybody).
Someone probably has an argument about why we shouldn't expect a term that's useful in ordinary life (e.g., 'solid') to line up directly with any fundamental physical concept, but we should expect that in philosophy. Maybe it's because they proceed differently; physics ultimately comes down to precise experiments, and philosophy doesn't work that way. Also, most analytic philosophers think that we shouldn't expect our ultimate technical concepts to line up exactly with our intuitions--the intuitions are at best the starting point.
There is also an argument that "it's what you do know that isn't so" doesn't show that 'know' can really be used of false beliefs--it's that here 'know' isn't being used literally. Of course it would be possible to prove anything if you could use the "it's not literal!" maneuver freely--one thing a lot of people try to do is come up with tests for when an expression is meant non-literally. (Of course I haven't given any such argument here.)
Posted by: Matt Weiner | Nov 5, 2005 9:27:15 PM
Someone probably has an argument about why we shouldn't expect a term that's useful in ordinary life (e.g., 'solid') to line up directly with any fundamental physical concept, but we should expect that in philosophy
I guess the question is: whether or not well-developed theories of x are more or less retentive of x-words we use in ordinary life, why should anyone care whether holes show up when we try to precisely define those words?
Also, don't most people's untutored physical intuitions notoriously "line up" with Aristotelian physics? When I took mechanics, that was certianly the case. At the outset of the class we did a test where everyone had to predict the trajectory of projectiles from diagrams, and so forth, and it amazingly demonstrated that result.
most analytic philosophers think that we shouldn't expect our ultimate technical concepts to line up exactly with our intuitions--the intuitions are at best the starting point.
Gee, that's heartening, because it would be a really dumbass view to expect them to line up exactly.
Posted by: spacetoast | Nov 6, 2005 1:19:41 AM
Well you could argue that "what you do know" should be "what you think you know" but in ordinary language people do not distinguish between what they know and what they think they know. In standard usage "know" includes the meaning "think you know" and hence can refer to false beliefs.
Posted by: James B. Shearer | Nov 6, 2005 2:23:16 AM
Gee, that's heartening, because it would be a really dumbass view to expect them to line up exactly.
Well, the alternative is more like "Our job is to explain our ordinary concepts, so any deviation from them is an indication of error." As I said, this is a minority view if anyone ever held it. But if you go too far away from ordinary concepts you run the risk of being accused of changing the subject--not talking about what we originally meant by "knowledge," but about a less interesting concept of your own invention.
In physics the new concepts we come up with are clearly better, in that they help us explain things rigorously (or something--philosophers of science would be able to explain this better, I hope). It's not obvious that that'll hold for new concepts that we invent in philosophy. I think it can be, and there are other philosophers who really (I think) see philosophy as a scientific enterprise that comes up with concepts that are just better than our folk concepts. But it's less obvious.
Posted by: Matt Weiner | Nov 6, 2005 10:57:59 AM
A lie is a moral construct. Much tougher than "truth table" constructs. "Who is Man that Thou art mindful of him?" is the prayer which asks God to look the other way once in awhile.
The WH didn't tell us (suppresio veri) that the CIA opposed the notion and (suggestio falsi) that they didn't have the strong belief in its accuracy. That they used language to avoid a perjury rap in the State of the Union address is disgusting.
Posted by: Jeffrey Davis | Nov 6, 2005 2:44:22 PM
Chekhov once observed that only life itself overtakes cynicism. I also think it overtakes all other so-called science.
I think also that most will agree that analytic philosophy, in its zeal to stamp out nonsense, has created some of the most egregious nonsense, to say nothing of its absurd notions of common sense. The true terror of philosophy is that subject and object are both ultimately reached by speculation. Philosophy never could build the bridge to fulfillment with words.
But cynicism can. In Kant's phrase: the purity of a cause often stands in inverse proportion to its truth. We are right to lie, cheat and steal for the good cause, rather than leave the advantage to the other side.
The master cynicism of Kant is radiantly apparent in 43 and his suite.
Posted by: tom | Nov 9, 2005 12:21:03 AM
Good heavens I'm late to this one. Statements about the truth of some proposition, Yellow Cake, we examine and find lacking. We discard them as unfit, insufficient, irrelevant, fradulent...balderdash.
The terrorizing philosophers might say 'False' or even 'premises not affirmed', but not us.
It might still be true that the propostion, Yelloe Cake is true. And it might be claimed or proclaimed, but we'd say (you and me -- fair to middling English speakers) they knew it like they knew 17 was going to show up on the roulette table. Not real deep knowledge. No, not the know but can't tell variety; not the probabilistic sort; not the neo-Kantian sort;
Just the plain fudge sort when your mom asks if you know what happened to all those cookies in the jar. You don't know. Not really.
Posted by: calmo | Nov 11, 2005 11:27:38 PM
The comments to this entry are closed.
