Is knowledge justified true belief?

For the serious discussion of weighty matters and worldly issues. No off-topic posts allowed.

Moderators: Azrael, Moderators General, Prelates

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Is knowledge justified true belief?

Postby TrlstanC » Tue Aug 16, 2011 3:40 pm UTC

The classic definition of Knowledge is that it's "Justified True Belief" - you believe something that's true, and you're justified in your belief. This definition works well for everyday usage, but there are examples of "Gettier Problems" that seem to show that the definition doesn't hold true 100% of the time. Here's the Wikipedia page with a good summary of different examples. Basically it seems like you can have someone with a justified true belief that isn't knowledge.

But it seems like these examples point out a relatively easy twerk to the original definition that would take care of all the Gettier Problems, and leave us with a nice consistent definition of knowledge. In the original the three conditions are given, but who is making the judgment isn't always explicitly stated, and I think this is the real problem that's being highlighted by the Gottier examples. If we spell out who's making the judgement for some piece of potential knowledge X, then we have:

1. I believe X - I think most people would assume I'm the only one who can make this judgment, and in common usage that's certainly true. I could imagine a case though where I don't actually believe X, but claim that I do, or maybe even think that I do? Also, I would imagine at some point we'd be able to put someone in fMRI and be able to tell if they actually believe something.

2. I'm justified in believing X - I should have good reason to believe it. Again, this judgment is being left up to me, I judge if the reason is "good enough."

3. X is true - This one I can't judge, and in fact it may be impossible for any person to judge accurately, there's just some knowledge that no human knows or will ever know. To be able to say if something is true we're assuming a kind of omniscient knowledge, a perfect 3rd party observer. This is a fundamentally different requirement than the first two judgements.

If we accept teh strong requirement for #3, that the fact must be true to be knowledge, than we have to accept that in at least some situations humans will never be able to judge if something is actually known or not since we'll never be able to judge if something is true or not.

Once we accept that limitation on #3, there's no reason not to extend the same limitation to #2 (and #1 too if we want). Gottier problems all apply to situations in which a person believes they're justified in believing something, they feel they have a good enough reason to believe, but in fact they're wrong, they've judged incorrectly. If we instead rely on a hypothetical omniscient 3rd party observer to make that judgment then the problems go away, while I may think I know X, someone with perfect knowledge would be able to judge that in fact I don't.

The real issue is that there are situations where we won't have enough information to judge if something is true, and may also not have information to judge if we have a good reason to believe it or not (and maybe even not be able to judge if we actually believe something in the first place). It's a limitation of the potential for humans to have 100% certainty, not a problem with the definition. In fact, we can split the definition in to a weak, common (how the term "knowledge" is common used) and strong form of knowledge like this:

Weak form - "I think I know X"
1. I believe X
2. I think I have a good reason to believe X

Common form - "We think he knows X"
1. I believe X
2. We (everyone with some information) think he has a good reason to believe X
3. We think X is true

Strong form - "He knows X"
1. He believes X
2. He has a good reason to believe X
3. X is true

This means that for philosophical problems we can assume the existance of a perfect 3rd party judge, and use the strong form of the definition, in everyday usage we can continue to use the common form of the definition, and if we want to be technically correct when we speak we can use the weak form.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Tue Aug 16, 2011 5:33 pm UTC

TrlstanC wrote:In the original the three conditions are given, but who is making the judgment isn't always explicitly stated, and I think this is the real problem that's being highlighted by the Gottier examples.

What judgment? You seem to be treating justification as a subjective thing, such that a belief p is justified, or at least justified for some subject S, if S thinks that p is justified. For example, you say "I judge if the reason is 'good enough'" and "Gottier problems all apply to situations in which a person believes they're justified in believing something." But nobody who holds justification to be a necessary condition for knowledge would agree with this characterization of justification, for the very reason for making justification a condition for knowledge is to avoid the problem of objectively faulty justification.

Consider a classic argument for adding justification to the true belief requirement for knowledge: Smith, who has no medical training, begins coughing and concludes that she has cancer. Smith, as it happens, has cancer. So, Smith believes that she has cancer, and this belief is true. However, Smith does not know that she has cancer, for her belief is only true by luck. To rule out luck, and therefore to acquire knowledge, she needs justification.

This argument could not function if justification is simply taken to be whatever someone believes is justified. Smith believes that her coughing justifies her conclusion that she has cancer. But, on any plausible justified true belief theory of knowledge, she is wrong.
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Tue Aug 16, 2011 7:37 pm UTC

TheGrammarBolshevik wrote:What judgment? You seem to be treating justification as a subjective thing


Exactly, justification by its very nature is based on a point of view, and is subjective because whether something is justified or not is based on what information is available to the person making the judgment. All justification ultimately relies on observation, either directly by the subject, or by an appeal to authority (which is also based on observation) and observation is fallible. This is in stark contrast to truth, which we accept as independent of observation. If something is true it doesn’t matter if anyone knows it or observes it to be true or not.

The classical definition of knowledge relies on these two separate requirements of what information needs to be available to make the judgment, and will run in to a problem when there’s a mismatch between the two.We want to be able to say in all situations that a person either knows X or doesn’t. But this is only possible if we have access to perfect information about whether X is true or not, as humans, this isn’t possible. We will never be able to say with certainty that X is absolutely true.

If however we accept the hypothetical situation in which we have perfect information available, than we can say that X is definitely true. If we accept that, we can also accept that we have perfect information about whether a subject is justified in believing X (thus moving the decision from the subject to a hypothetical omniscient judge). This definition is only useful in hypothetical situations, but it also resolves the Gettier problems in those situations.

If we want to actually apply the term “knowledge” to actual events in our lives, we have to give up the reliance on absolute truth, and we can also rely on everyday, less than perfect information, judgment of what’s justified. In this case, one person could claim to have knowledge, but if other people had additional information they could rightly claim that he in fact didn’t know what he thought he did. And then if we want to always be strictly correct, we should preface “I know X” with “I think”, at least that way if we’re being truthful we’ll always have a correct statement, even if we don’t have perfect knowledge.


TheGrammarBolshevik wrote:Consider a classic argument for adding justification to the true belief requirement for knowledge: Smith, who has no medical training, begins coughing and concludes that she has cancer. Smith, as it happens, has cancer. So, Smith believes that she has cancer, and this belief is true. However, Smith does not know that she has cancer, for her belief is only true by luck. To rule out luck, and therefore to acquire knowledge, she needs justification.

This argument could not function if justification is simply taken to be whatever someone believes is justified. Smith believes that her coughing justifies her conclusion that she has cancer. But, on any plausible justified true belief theory of knowledge, she is wrong.


But the arguement does work if someone else can judge that she’s wrong, someone with more information than her. That judge can either be another person, or if we want to resort to a hypothetical situation, someone with perfect information to judge. There’s no way to have a justification without appealing to someone’s observations, but it is important how much information that person has. For example, we could recreate the same example, but 2,000 years ago, and it would go something like this:

Smith begins coughing, concludes she has a disease of the lungs and will die. In fact, she has cancer, and will die.

This could be a justified true belief before widespread medical knowledge and treatments were available, Smith knew she was sick and was going to die. But now, it wouldn’t be considered knowledge, since we’d say she wasn’t justified. Of course a hypothetical omniscient judge could also correctly say that Smith never knew she was going to die since the cough wasn’t related to the cancer that would eventually kill her.

Of course, we could also ask who can judge if something is true or not, and depending on if we're asking philosophical questions, or using the term in everyday life, different answer will be useful. But if we're going to define the term "knowledge" we should be consistent across the different requirements, if we're going to require absolute information about "truth" we should also require absolute information about "justified."
Last edited by TrlstanC on Thu Aug 18, 2011 2:02 am UTC, edited 1 time in total.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Tue Aug 16, 2011 9:04 pm UTC

TrlstanC wrote:
TheGrammarBolshevik wrote:What judgment? You seem to be treating justification as a subjective thing


Exactly, justification by its very nature is based on a point of view, and is subjective because whether something is justified or not is based on what information is available to the person making the judgment.

Both halves of that sentence are important:
TheGrammarBolshevik wrote:You seem to be treating justification as a subjective thing, such that a belief p is justified, or at least justified for some subject S, if S thinks that p is justified.

The sense in which justification is uncontroversially subjective is not the sense in which you're treating it. Whether S's belief is justified depends on the evidence available to S, but it does not depend on whether S thinks that her belief is justified.

TrlstanC wrote:The classical definition of knowledge relies on these two separate requirements of what information needs to be available to make the judgment, and will run in to a problem when there’s a mismatch between the two.We want to be able to say in all situations that a person either knows X or doesn’t. But this is only possible if we have access to perfect information about whether X is true or not, as humans, this isn’t possible. We will never be able to say with certainty that X is absolutely true.

The analysis of knowledge isn't about being able to say with certainty whether S knows p. Rather, it's about establishing the necessary and sufficient conditions under which S knows p. Investigating whether those conditions are met in a given case are not really a matter for epistemologists, any more than the question of whether Lee Harvey Oswald shot JFK is a question for ethicists.

TrlstanC wrote:There’s no way to have a justification without appealing to someone’s observations, but it is important how much information that person has. For example, we could recreate the same example, but 2,000 years ago, and it would go something like this:

Smith begins coughing, concludes she has a disease of the lungs and will die. In fact, she has cancer, and will die.

This could be a justified true belief before widespread medical knowledge and treatments were available, Smith knew she was sick and was going to die. But now, it wouldn’t be considered knowledge, since we’d say she wasn’t justified.

The treatments available are irrelevant; if medical technology at the time were such that people with coughs could be expected to die, then Smith did indeed know that she would die. On the other hand, to point to the medical beliefs* of the time is simply to show that many medical beliefs 2,000 years ago were unjustified. You seem to be arguing that the fact that more people were wrong should make it more easy to be justified. This flies in the face of the uncontroversial claim that modern medical science is more justified than that of the past.

*To call such beliefs "medical knowledge" would be to beg the question.
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Wed Aug 17, 2011 12:17 am UTC

TheGrammarBolshevik wrote:You seem to be treating justification as a subjective thing, such that a belief p is justified, or at least justified for some subject S, if S thinks that p is justified.

...

The sense in which justification is uncontroversially subjective is not the sense in which you're treating it. Whether S's belief is justified depends on the evidence available to S, but it does not depend on whether S thinks that her belief is justified.

...

The analysis of knowledge isn't about being able to say with certainty whether S knows p. Rather, it's about establishing the necessary and sufficient conditions under which S knows p.



I probably wasn't very clear on this point, but I agree with you. Whether Smith thinks they know p isn't really the point, we're more interested in whether sufficient conditions are met that someone else could judge that Smith knows p. "Thinking you know something" is just called "Believing" and is the first condition for actually knowing something.* The question is, how much information is available to the person making the judgment? In the classical definition, the person making the judgment has perfect information about the world to judge if the belief is true, but is limited to Smith's information about the world to judge if it's justified. I believe that the Gettier problems highlight this problem, and that the solution is to have the subject who's judging whether "Smith knows p" to have a consistent level of information about both the truth of the belief and the justification for the belief. The level of information available to such a subject would be:

1. Perfect - They will always be able to correctly judge whether "Smith knows p" and Gettier problems won't be an issue.

2. Society - This would be the level of information available to the sum of any other person who has information. We can think of this as the level of detail that would be revealed in a perfect court case. Gettier problems might exists, but we wouldn't be able to tell if they exist or not.

3. Individual - This is basically just "believing something"


TheGrammarBolshevik wrote:You seem to be arguing that the fact that more people were wrong should make it more easy to be justified. This flies in the face of the uncontroversial claim that modern medical science is more justified than that of the past.


What I'm claiming is that using the traditional definition of knowledge, we could find ourselves in a situation where all rationale knowledgeable people could judge that "Smith knows p" and all of them would be wrong. There can be a disconnect between the definition of knowledge using perfect information (which is useful for philosophical questions), and the common usage of the term where we have limited information. If we try to use one definition to cover all situations than we'll run in to inconsistencies (the Gettier problems being a few examples).




----------------------------------------------------------------------

*There may be some very rare cases where someone claims to believe something, but doesn't actually, but I would say our current understanding of neurobiology probably isn't at the point where we could give it a proper treatment.
Last edited by TrlstanC on Thu Aug 18, 2011 2:03 am UTC, edited 1 time in total.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Wed Aug 17, 2011 3:23 am UTC

TrlstanC wrote:The question is, how much information is available to the person making the judgment? In the classical definition, the person making the judgment has perfect information about the world to judge if the belief is true, but is limited to Smith's information about the world to judge if it's justified.

Not at all. The judge can have all the information she wants; it's just that only Smith's information is relevant to determining justification. (This assumes internalism, but I take it that you're an internalist, since you agree with me that "Whether S's belief is justified depends on the evidence available to S." But your claim is also incompatible with externalism, since an externalist judge isn't limited to Smith's information about the world.)

TrlstanC wrote: I believe that the Goetier problems highlight this problem, and that the solution is to have the subject who's judging whether "Smith knows p" to have a consistent level of information about both the truth of the belief and the justification for the belief. The level of information available to such a subject would be:

1. Perfect - They will always be able to correctly judge whether "Smith knows p" and Goetier problems won't be an issue.

Why wouldn't Gettier problems be an issue? Gettier problems aren't about being able to judge whether S knows p; they're about understanding what conditions make that judgment accurate.

TrlstanC wrote:What I'm claiming is that using the traditional definition of knowledge, we could find ourselves in a situation where all rationale knowledgeable people could judge that "Smith knows p" and all of them would be wrong.

Only if they're missing some knowledge or some rationality. But is this anything special about "Smith knows p," or is it just a specific case of the general fact that knowledgeable, rational people can believe something but be wrong?
Nothing rhymes with orange,
Not even sporange.

klackity
Posts: 1
Joined: Wed Aug 17, 2011 12:11 am UTC

Re: Is knowledge justified true belief?

Postby klackity » Wed Aug 17, 2011 4:49 am UTC

We started with a naive definition of knowledge as "true belief". Then we realized a flaw in this definition: we can believe true things for stupid reasons. I might read a lousy horoscope and believe I'm destined for a car accident. If I have an accident later that day, it's not that I really had knowledge of the future. I just had an accidentally-true belief.

So we think, perhaps a true belief has to be justified to count as knowledge. Certainly a horoscope is not sufficient justification! But we have a new problem: we must define what "justified" means!

The Gettier problems all arise from a poor (weak) definition of "justified". Any proper definition of "justified" ought to imply truth! I mean, suppose I say I have a proper justification for believing something which later turns out to be false. You'd say to me, "Your justification mustn't have been good enough!"

Thus instead of "Justified True Belief", we ought to say "Justified Belief", because the former is deceptive. But people tend not to use the latter definition of knowledge, because it seems to be almost circular. That is, any sufficiently strong definition of "justified" will look a lot like "I know it's true".

But I think I have a sufficiently strong definition of "justified" which isn't that circular: A belief is justified if the cause of the belief necessitates the truth of the belief. This definition eliminates all Gettier problems, and moreover allows us to answer to questions about knowledge by talking about causation and necessity.

User avatar
mmmcannibalism
Posts: 2150
Joined: Tue Jun 30, 2009 6:16 am UTC

Re: Is knowledge justified true belief?

Postby mmmcannibalism » Wed Aug 17, 2011 5:15 am UTC

Not sure if this is a meanginful contribution; but I remember having to write a college essay with a prompt very similar to the thread title(more info on page describing essay obviously). So this may be homework
Izawwlgood wrote:I for one would happily live on an island as a fuzzy seal-human.

Oregonaut wrote:Damn fetuses and their terroist plots.

elasto
Posts: 3778
Joined: Mon May 10, 2010 1:53 am UTC

Re: Is knowledge justified true belief?

Postby elasto » Wed Aug 17, 2011 10:54 am UTC

I really don't get what this is about. Calling something knowledge doesn't mean it isn't or can't be false. If we go that far then not much outside of 'I think therefore I am' and '2+2=4' can be called 'knowledge'. It all has to be called 'belief'. That renders 'knowledge' as a not very useful concept in everyday parlance and we'd simply have to make up a new word to replace it. After all, 'I have a head' might be false if we're all simulations on God's PC.

I'd go with a definition of 'knowledge' simply as 'justified, unfalsified belief' therefore. If it happens to be true it'd be true knowledge, and if it happens to be false it'd be false knowledge (something can be false but unfalsified - and remembering that it's almost always unknowable whether anything's really true or false, as stated earlier).

As for coming up with a good definition of 'justified', I guess I go with something like 'A belief is justified if a rational person would have the same belief given the same evidence' but then you have to define what 'rational' means. It all smells a bit circular to me, but maybe I haven't put enough thought in to that part for now.

Again, it doesn't mean a belief wasn't justified just because it turns out to be false. Nor does it mean it wasn't justified if it turns out to be right but for the wrong reason. It was still justified given the evidence known at the time and that's all that matters. It's why examples such as this don't seem to demonstrate anything for me:

Gettier problem on Wikipedia wrote:"After arranging to meet with Mark for help with homework, Luke arrives at the appointed time and place. Walking into Mark's office Luke clearly sees Mark at his desk; Luke immediately forms the belief 'Mark is in the room. He can help me with my logic homework'. Luke is justified in his belief; he clearly sees Mark at his desk. In fact, it's not Mark that Luke saw; it was a marvelous hologram, perfect in every respect, giving the appearance of Mark diligently grading papers at his desk. Nevertheless, Mark is in the room; he is crouched under his desk reading Frege. Luke's belief that Mark is in the room is true (he is in the room, under his desk) and justified (Mark's hologram is giving the appearance of Mark hard at work)."

Again, it seems as though Luke does not "know" that Mark is in the room, even though it is claimed he has a justified true belief that Mark is in the room

Luke's belief was justifiably called knowledge to me (ie 'justified, unfalsified belief') - up until the point it was found out that there was a hologram in the room but Mark was hiding, whereupon it is still knowledge - just the nature of the justification has then changed. To not regard Luke's belief as 'knowledge' is running into the problem I described in my first paragraph: It ceases to be a useful concept if we go to that far an extreme.

Cres
Posts: 67
Joined: Tue Dec 22, 2009 2:14 pm UTC

Re: Is knowledge justified true belief?

Postby Cres » Wed Aug 17, 2011 11:44 am UTC

klackity wrote:The Gettier problems all arise from a poor (weak) definition of "justified". Any proper definition of "justified" ought to imply truth! I mean, suppose I say I have a proper justification for believing something which later turns out to be false. You'd say to me, "Your justification mustn't have been good enough!"
[/b] This definition eliminates all Gettier problems, and moreover allows us to answer to questions about knowledge by talking about causation and necessity.


This definition of justification is too strong, because surely we can have a justified false belief. I look across the room and see what looks, sounds etc. exactly like my dog and form the belief 'My dog is over there'. But in fact it's a robot placed there by mischievous aliens that looks just like my dog. Am I not justified in having the belief? Do I need to perform a thorough alien/robot/hologram etc. investigation before I can say 'I know my dog is over there'? On your definition of justification, even in cases where the dog is in the room, the bare logical possibility of its being a robot would mean that your belief about the dog (and indeed almost everything else) wouldn't constitute knowledge.

elasto wrote:Luke's belief was justifiably called knowledge to me (ie 'justified, unfalsified belief') - up until the point it was found out that there was a hologram in the room but Mark was hiding, whereupon it is still knowledge - just the nature of the justification has then changed. To not regard Luke's belief as 'knowledge' is running into the problem I described in my first paragraph: It ceases to be a useful concept if we go to that far an extreme.


The intuition the Gettier cases are getting at can be brought out best I think if you consider some counterfactuals around those cases. The problem with calling Luke's belief knowledge in the Wikipedia example is that if Mark were not in fact in the room, Luke's belief would still be the same (and false in this case) because of the hologram. The sense is that, for 'I believe that P', my belief that P and the truth of P must be linked in the right way. Gettier cases show that justified true belief isn't a sufficient condition for the existence of this sort of link. The existence of a causal chain between them is one influential suggestion for how this 'right way' might be analysed.

elasto
Posts: 3778
Joined: Mon May 10, 2010 1:53 am UTC

Re: Is knowledge justified true belief?

Postby elasto » Wed Aug 17, 2011 12:09 pm UTC

Cres wrote:The intuition the Gettier cases are getting at can be brought out best I think if you consider some counterfactuals around those cases. The problem with calling Luke's belief knowledge in the Wikipedia example is that if Mark were not in fact in the room, Luke's belief would still be the same (and false in this case) because of the hologram. The sense is that, for 'I believe that P', my belief that P and the truth of P must be linked in the right way. Gettier cases show that justified true belief isn't a sufficient condition for the existence of this sort of link. The existence of a causal chain between them is one influential suggestion for how this 'right way' might be analysed.

That's why I wouldn't go with 'justified true belief' as a good definition for knowledge; Rather I'd go with 'justified unfalsified belief'.

That means if something turns out to be false, the earlier belief was still knowledge because at that point it was not yet falsified.

To require that knowledge has to be true to be knowledge is too strong a requirement imo, because almost nothing that in common parlance is knowledge (eg I have a head) can really be viewed as knowledge, because we don't actually know the truth of it - we just think we do.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Wed Aug 17, 2011 1:45 pm UTC

I definitely agree with everyone who's saying that the whole issue hangs on the definition of "justified." I think we all have a basic understanding of what it means for something to be justified, that goes something like:

A person is justified to believe p if and only if, then can logically conclude p given all the information they have available at a certain point in time.

This probably isn't a perfect definition of “justified”, but it should be in the ballpark for our everyday understanding. Crucially, it specifies that the justification must be made with limited information - only what's available to the person at a certain point. I'm proposing that depending on how we define knowledge we should make use of a slightly different definition of justification. For example, we can call one new kind of justification "perfect justification" or "justification with complete information" or “perfect assumptions” (if we don’t even want to include the word justification in this new kind of judgment). This would be what a person could be justified in believing if they had access to all information - except what they’re trying to justify. Adding this additional definition should raise some questions:

    1. “Isn’t that just the same as knowing all true things?” Almost, by excluding the one piece of information that’s trying to be justified though, it’s slightly different. It’s a test to see if that one piece of information can be logically deduced from all other true information.

    2. “Why allow the use of all information, isn’t that unrealistic?” Absolutely, but in the classical definition of knowledge we’re allowed to judge if something is true. Knowing if something is true is only possible with perfect information, this just allows us to use that same perfect information to make the other judgments as well, instead of limiting it to just one.

    3. “What good is it?” This definition allows us to judge in hypothetical situations, with perfect certainty, whether someone knows something. Which was the goal of the classical definition of knowledge, it’s a tool for answering philosophical questions.

    4. “Does this fix Gettier problems?” Yes, the Gettier problems came up because the goal of the classical definition of knowledge was to judge with perfect certainty if someone knew something, and we allowed ourselves perfect information – but only to judge truth. For belief and justification we limited our information. Any time we’re making a judgment with limited information there’s the possibility for error. The only way to remove the possibility for error is to remove the limits on information in the other two judgments.

    5. “What about belief, is that limited too?” If we’re going for a perfect judgment we shouldn’t limit our information about whether someone believes something or not. Classically we limited ourselves in the same way as our judgment on justification, to only the information available to the person. But I think we should expect that as we learn more about how the human mind works that we will find situations where a person doesn’t have perfect information about what they believe. For example we might be able to put someone in an fMRI and be able to detect that they believe something before they realize they do. If we want to make perfect hypothetical judgments about knowledge we should allow perfect (and outside) information in judging if someone believes something or not too.
We can also ask “what about an everyday definition of knowledge?” In our common usage we don’t have access to perfect information, so we can’t judge with perfect certainty if something is true or not. We rely instead on what everyone can deduce to be true based on all information available. If we allow ourselves to apply this same level of information to the judgment of justification we end up with a 2nd new type of justification. We can call this “common information justification” or “all deducible true statements.” I think this is the definition of knowledge that we commonly use in everyday speech. It’s different that the “perfect” definition of knowledge in that it’s fallible, it’s possible to have situations where we think someone knows something, but that it turns out that they didn’t. I wouldn’t call this a problem with the definition, it’s just the way knowledge works in the real world, without perfect information we can’t judge perfectly.

If we limit ourselves to the standard definition of “justification”, to only the information that’s available to a single person at a specific point in time, then we can also extend this limitation to the judgment of “truth.” This basically collapses down to “what a rational person believes.” Or can be stated something like if with all the information available to me right now I can logically conclude p and I also believe p then I can say that I know p.

Summary: problems arose in the classical definition of knowledge because the three different requirements (belief, justification and truth) were making use of different levels of information. Using the same level of information for all three requirements (even if it requires us to rename some of them to avoid confusion) eliminates any problems. Depending on what level of information we use, we can have different definitions of knowledge, for philosophical questions where we want a perfect determination of knowledge in hypothetical situations we can allow ourselves perfect information for all three requirements, and this eliminates the Gettier problems.




-------------------------------------------------------------------------------------------------------------------------
Edit:
mmmcannibalism wrote:Not sure if this is a meanginful contribution; but I remember having to write a college essay with a prompt very similar to the thread title(more info on page describing essay obviously). So this may be homework


I first learned about the classical definition of knowledge and the Gottier problems back in Philosophy 101 too. I don't remember if I had to write an essay on it, but I would hope that I could write one now that would get a passing grade :)
Last edited by TrlstanC on Thu Aug 18, 2011 2:04 am UTC, edited 2 times in total.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Thu Aug 18, 2011 12:26 am UTC

TrlstanC wrote:4. “Does this fix Goetier problems?” Yes, the Goetier problems came up because the goal of the classical definition of knowledge was to judge with perfect certainty if someone knew something, and we allowed ourselves perfect information – but only to judge truth. For belief and justification we limited our information. Any time we’re making a judgment with limited information there’s the possibility for error. The only way to remove the possibility for error is to remove the limits on information in the other two judgments.

Why don't we apply this to an actual Gettier problem? The first one from Gettier's paper will do.

  1. With perfect knowledge, we know that the man who will get the job has ten coins in his pocket.
  2. With perfect knowledge, we know that Smith believes that the man who will get the job has ten coins in his pocket.
  3. With perfect knowledge, we know that Smith will get the job and that Smith has ten coins in his pocket. We therefore conclude that, given perfect knowledge, Smith's belief that the man who will get the job has ten coins in his pocket is justified.
But Smith still doesn't know — in any sense, philosophical, casual, whatever — that the man who will get the job has ten coins in his pocket.

Hell, you don't even need a Gettier problem to show that this is an inadequate definition of knowledge. Consider cancer-Smith. She has a true belief, and perfect information gives us all the medical evidence we need to conclude that she's right — "justification with complete information." In other words, your analysis of knowledge (once again) includes all true beliefs.

elasto wrote:nothing that in common parlance is knowledge … can really be viewed as knowledge, because we don't actually know the truth of it

This is flagrantly circular.
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Thu Aug 18, 2011 2:43 am UTC

TheGrammarBolshevik wrote:
  1. With perfect knowledge, we know that the man who will get the job has ten coins in his pocket.
  2. With perfect knowledge, we know that Smith believes that the man who will get the job has ten coins in his pocket.
  3. With perfect knowledge, we know that Smith will get the job and that Smith has ten coins in his pocket. We therefore conclude that, given perfect knowledge, Smith's belief that the man who will get the job has ten coins in his pocket is justified.
But Smith still doesn't know — in any sense, philosophical, casual, whatever — that the man who will get the job has ten coins in his pocket.


These are all judgments about the truth of the statements, which would all be exactly the same under the classical definition of knowledge. They skip looking at how perfect information would affect the judgment of justification, which is the only point that I believe needs to be addressed (except the third point which has the incorrect conclusion that having a true belief makes it a justified belief).

If instead we extend the ability to have perfect information to the judgement of justification (or call it something else that means that), then we can judge that while Smith believes that "the man who will get the job has ten coins in his pocket" he is mistaken in that belief - with perfect information he wouldn't be justified in that belief. Making that judgment correctly would lead to us correctly saying that Smith didn't have the knowledge, and avoiding the problem in the example. Allowing perfect information is the only way to avoid making mistaken judgements of belief, justification or truth, and that's the only way to avoid falsely concluding that Smith has knowledge when he doesn't.
Last edited by TrlstanC on Thu Aug 18, 2011 11:59 am UTC, edited 1 time in total.

User avatar
Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

Re: Is knowledge justified true belief?

Postby Charlie! » Thu Aug 18, 2011 3:14 am UTC

From a machine learning perspective, knowledge and evidence are nearly interchangeable. If I tell a computer to look at a field with a sheep in it from far away, it'll see a white shape. This is good evidence that there is something white in the field (probability 0.95), which is good evidence that there's a sheep in the field (probability 0.2). So the computer would not "know as qualitatively true" that there was a sheep in the field, "knowledge" here refers to the quantitative degree (about 0.2) to which it can predict the sheep.
Some people tell me I laugh too much. To them I say, "ha ha ha!"

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Thu Aug 18, 2011 3:33 am UTC

TrlstanC wrote:with perfect information he wouldn't be justified in that belief.

Why not?
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Thu Aug 18, 2011 1:35 pm UTC

TrlstanC wrote:with perfect information he wouldn't be justified in that belief.


Sorry, that really was a bit of an oversimplification of a complex statement. Obviously we can't assume that Smith has perfect information about everything, otherwise he would clearly know everything because he would be justified in believing all true statements. The full statement should've been something like "An outside observer with perfect information would judge that Smith could not logically conclude p." Which, admittedly is a bit of a mouthful. Although we don't actually need to assume that the "outside observer" needs to have perfect information about everything. In the classical definition we're already assuming that we can allow the judge (or "outside observer" or "us") perfect information to decide if the belief is true, and we're also allowing them to make perfect logical judgments about the justification as well - they can ignore Smith believing he's justified if he used faulty logic to get there. At this point the classical definition stops short and refuses to allow perfect information to make any other judgments. There have been many proposed changes that allow more information in the judgment of "justification" (such as causality or defensibility) to get around Gettier problems, but still stop short of perfect information. I think this is because of the commonly held understanding of the word "justification" as being limited to what a person knows (which is never perfect). But if we want a definition of knowledge that's infallible we have to ensure that all three judgments we use to get there are also infallible. If we insist on allowing false assumptions in to the judgment of justification then there will always be situations where that definition of knowledge will fail.

Fortunately we don't have to assume that the judge has perfect information about all facts, they only need to be able to judge perfectly the assumptions that are being made, and be perfectly logical. When someone is justifying a belief they are allowed to:

1. Make observations
2. Make an appeal to authority
3. Assume that their observations and the authority are correct
4. Make any logical conclusions from those assumptions

If we want a perfect definition of knowledge to use in hypothetical situations then when judging if a person has a justified belief, we have to allow ourselves the assumption of perfect information to judge #3. In that case we would only judge that a belief is justified (or call it "logically consistent" or anything besides justified if that causes confusion) if all the assumptions used to come to that conclusion are true. Making this change to the classical definition of knowledge - which already includes the assumption that we are perfectly logical to judge if #4 is true, and have perfect information to judge if the belief is true (in addition we should also allow ourselves perfect information to judge if Smith really believes p as well, although that would address a different set of problems.)

We can call this twerked definition of knowledge "Strong justified true belief" or "justified true belief with perfection assumptions" or "JTB with perfect information." Anything that indicates that while the classical definition allowed perfect information for some of the requirements it stopped short of allowing it for all the requirements, while this definition allows perfect information (or assumptions) for judging all the requirements. This would also differentiate it from our everyday use of the word "knowledge" which we accept can be used in error.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Fri Aug 19, 2011 12:28 pm UTC

…Is this just a roundabout way of stating the "no false premises" solution?
Nothing rhymes with orange,
Not even sporange.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7604
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Is knowledge justified true belief?

Postby Zamfir » Fri Aug 19, 2011 1:03 pm UTC

Charlie! wrote:From a machine learning perspective, knowledge and evidence are nearly interchangeable. If I tell a computer to look at a field with a sheep in it from far away, it'll see a white shape. This is good evidence that there is something white in the field (probability 0.95), which is good evidence that there's a sheep in the field (probability 0.2). So the computer would not "know as qualitatively true" that there was a sheep in the field, "knowledge" here refers to the quantitative degree (about 0.2) to which it can predict the sheep.

Ditching knowledge for evidence might be a good idea in lots of cases, but I am less sure if attaching a single number is a generally useful method to qualify evidence. It can work for styalized cases, but hardly in general.

Take your machine above: what do probabilities mean? It sort of assumes there is a set of representative test cases, so you can test the machine on the test cases to find out how reliably it recognizes white things. But how do you determine that a set is representative? For "recognizing white things" you might still be able to make a case, like a procedure to generate white patches randomly over the field of vision of the machine.

But how do you make a representative set of cases where a white patch is or is not a sheep? To do so, you'd have to know all the white things that might occur in fields that might look like sheep, and how likely those are in the situation the machine is actually in. That comes suspiciously close to the perfect information assumptions that haunt this thread.

Let's say there is a sheet factory next to the field. That would greatly increase the odds that a white thing is an escaped sheet instead of a sheep. The machine does not take this into account, so on seeing something white it overestimates the "probability" of seeing a sheep. But the factory is on strike today, so the machine is actually correct in not taking the factory into account...

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Fri Aug 19, 2011 1:26 pm UTC

TheGrammarBolshevik wrote:…Is this just a roundabout way of stating the "no false premises" solution?


It's very similar to the "no false premises" solution, which makes sense since the Gettier problems all arise because of the subject on at least one mistaken assumption. That mistaken assumption has to be addressed one way or another. I see the differences to traditional solutions as:

1. A slightly more formal definition (or re-definition) of "justified" since the traditional usage of that word has a pretty strict definition that allows for reliance of incorrect information.
2. It recognizes that we're not just using the information available to the subject (we're allowing perfect information in the judgment of truth) in the traditional definition of knowledge, but haven't allowed ourselves to use the same information in the judgment of the assumptions for the justification. The adjusted definition of justification (or definition of a new, but similar term) should allow this, and we're already allowing ourselves information which would only be available in a hypothetical situation.
3. It proposes different definitions of knowledge depending on how the term is being used. If we're using it hypothetically and want a perfect judgment, we have to rely on perfect information. There's also the "common" usage that relies on less than perfect information, and a "weak" version if we're judging whether we have knowledge just based on the information that's available to our self.

I think the traditional solutions to the Gettier problem were too ambitious; they tried to use the traditional definition of justified, even though the Gettier problems pointed out that the incorrect assumptions in the justification caused errors in judging knowledge. Some also seemed to want to be able to judge knowledge correctly in all cases, even though that would be impossible in every day usage since we'll never have perfect information.

User avatar
Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

Re: Is knowledge justified true belief?

Postby Charlie! » Fri Aug 19, 2011 4:52 pm UTC

Zamfir wrote:That comes suspiciously close to the perfect information assumptions that haunt this thread.
In fact, probabilities are tied up very closely with imperfect information. For example, say someone flips a coin fairly and slaps it onto the back of their hand. I would assign probabilities 0.5 to heads and tails. But when you think about it, that's really weird. Given that the coin has already been flipped and is just sitting there, it's already whatever it's going to be - if it's tails, it was already tails with probability ~1. By deviating from this "real" probability, my assignment of 0.5 to tails is already wrong. But at the same time, it's the best that I can do, because I have incomplete knowledge!

Similarly, there isn't 0.2 of a sheep in that field. By assigning a probability you're already not reflecting reality. There could be a sheet factory or a sheep herd or a big rock out there, and whatever is there is already there, and by assigning a probability you're already not reflecting reality. It's just that your information is incomplete, so p(sheep) = 0.2 is the best you can do given the state of your evidence - rather than reflecting reality (usually), probabilities reflect states of imperfect information.
Some people tell me I laugh too much. To them I say, "ha ha ha!"

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7604
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Is knowledge justified true belief?

Postby Zamfir » Fri Aug 19, 2011 5:35 pm UTC

But coin flipping is a horrible model for most situations of imperfect knowledge. It is a special case where we can use symmetry to deduce that there are two equal outcomes without hardly any necessary knowledge of the processes that actually govern the movement of the coin . On top of that, coins are extremely similar, so previous coins and coin flips serve as a reliable confirmation that coins do indeed behave that way and we didn't forget a factor.

So we can collapse everything we don't know to that simple 50-50.

But most situations are not like that at all. There is no a priori deduction that tells us how to assign probabilities to the 'white thing I saw was a sheep'. Depending on the context, it could be any number from 0 to 1. So experience from other contexts doesn't even translate to a new context without extra knowledge of the nrew situation.

Of course, scientists try to create artificial situations that are more like the coin toss. Experiments try to exclude factors until the situation is amenable to analysis, and at the very least repeatable. But even they often fail, let alone that ordinary situations fit.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Fri Aug 19, 2011 6:20 pm UTC

Coin flipping is an extremely simplified model. Which means it's not a great representation of real world probelms, but it can still be a useful tool to understand how we assign probablities to our chances of being wrong. Just like in flipping a coin, in more complicated scenarios I would say that when we're trying to determine the chance that we're correct, we have no choice but to rely on past experience. That doesn't mean that we always simple mindedly assume "I've only ever seen white swans, therefore there's a 100% chance that swans are white" - although some people have certainly made that mistake. We're capable of thinking "I've seen lots of sheep, and some small percentage have been black instead of white, there could be a similiar chance of seeing a black swan." Or even "If I thought I was 100% right before, and have turned out to be wrong about that 2% of the time, maybe I'm only 98% correct when I think I'm 100% correct."

Of course that doesn't mean that we'll assign the correct probability, just that we're capable of judging the chance of being wrong in new situations from past experience. Also, that we don't have a very good model of human inteligence, I don't think we could say with any certainty that we understand how we make judgments in complicated situations. That doesn't mean that we can't, just that we don't fully understand how we do it yet.

User avatar
Charlie!
Posts: 2035
Joined: Sat Jan 12, 2008 8:20 pm UTC

Re: Is knowledge justified true belief?

Postby Charlie! » Fri Aug 19, 2011 6:45 pm UTC

Zamfir wrote:But coin flipping is a horrible model for most situations of imperfect knowledge. It is a special case where we can use symmetry to deduce that there are two equal outcomes without hardly any necessary knowledge of the processes that actually govern the movement of the coin. [...]

But most situations are not like that at all. There is no a priori deduction that tells us how to assign probabilities to the 'white thing I saw was a sheep'. Depending on the context, it could be any number from 0 to 1. So experience from other contexts doesn't even translate to a new context without extra knowledge of the nrew situation.

The point is that once you see that probabilities are already wrong, you can be a little more comfortable with them not including all sorts of relevant information, like whether there's a herd of sheep over the next hill. You can only include the information you have, after all. And yes, that's wrong, but it's the best you can do.
Some people tell me I laugh too much. To them I say, "ha ha ha!"

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Sat Aug 20, 2011 2:31 am UTC

TrlstanC wrote:
TheGrammarBolshevik wrote:…Is this just a roundabout way of stating the "no false premises" solution?


It's very similar to the "no false premises" solution, which makes sense since the Gettier problems all arise because of the subject on at least one mistaken assumption.

How could you aware of the "no false premises" solution, but unaware that the last clause of that sentence is false? There are plenty of Gettier problems that don't involve false premises. Barn County, the robot dog that Crea mentioned, and so on. No strategy can be successful by pursuing this supposed feature of Gettier problems.

TrlstanC wrote:There have been many proposed changes that allow more information in the judgment of "justification" (such as causality or defensibility) to get around Gettier problems, but still stop short of perfect information.

It's "defeasibility," but this is still inaccurate. The arguments to which you refer don't modify the definition of "justification," but instead add an additional criterion to JTB.
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Sat Aug 20, 2011 1:55 pm UTC

Again, I think that the thought that barn country or the robot dog would lead to the judgment of false beliefs under my proposed definition comes up because “justification” has a common definition that’s incompatible with any kind of infallible judgment. Let’s assume that the definition of “justification” is (some people may disagree with this, but that’s ok, we’re going to be losing the word “justification” shortly.)

If S wants to justify believing in p they:

1. Can make use of their observations
2. Can make use of an appeal to authority
3. Assume that their observations and any authority are correct
4. Can make use of any conclusions which can be logically inferred from 1, 2 and 3.

If we want to replace the term “justified” in the classical definition of knowledge, we could instead use a definition of “knowledge is a logical true belief.” This would give us these requirements for perfectly judging knowledge, call it the Strong definition of knowledge:

If:
1. S believes p
2. p is true
3. Given S’s assumptions it's possible to logically conclude p
4. S’s assumptions are true
5. We have perfect information to judge if 1, 2, 3 and 4 are true

Then we can conclude with perfect certainty that “S knows p.”

If we don’t want to make the assumption of perfection information in #5 then we can use a weakened assumption instead, such as “Using information available to us we can judge if 1, 2, 3 and 4 are true” then we would have a “common definition of knowledge” which is pretty much how most people use the term most of the time. Of course it won’t always judge correctly, but that’s a limitation of weakening #5 to make it useable in everyday language. Or we could even just rely only on the information that S actually has and create an “individual definition of knowledge.” In this case we won’t need #4 and #5 at all, and in fact from S’s perspective #2 and #3 are functionally equivalent to #1 – if you think p is true, or you can logically conclude p than you believe p and you think “I know p.”

Edit, additional thoughts:
------------------------------------------------------------------
The goal of defining knowledge isn’t to be able to test it using purely hypothetical scenarios. We test possible definitions with hypothetical situations to make sure that we all agree that the definition comes to the correct conclusion in all cases, even very unrealistic ones. But it’s in particularly unrealistic cases that we have to be careful to clearly state all of our assumptions in great detail. This definition makes it clear that depending on the assumptions we make two similar situations could have opposite outcomes. For example in the “barn land” situation:

1. If Smith sees a lot of barns (most fake, one real) and assume that they’re all barns, and that this particular one is a barn, then we would judge he doesn’t know it’s a barn
2. However, if Smith is driving along, day dreaming, and not paying attention to the country side, stops next to the one real barn and judges that it’s a barn, we would have good reason to say that he knows it’s a barn.

Of course in both situations we’re still making a lot of assumptions that aren’t being spelled clearly and unambiguously, if we want to make a perfect judgment even in very unrealistic cases we must be perfect in defining the assumptions we’re using. And eventually as we try to be as clear as possible in what we’re assuming we’re going to run in to situations where we have to admit we don’t have any rational reason to be able to assume one thing over another. We might hit a point where we can’t answer the question “does someone assume X consciously or unconsciously” or “when exactly can we say someone believes X?” These are good questions because they point us in the direction of what we need to study to be able to answer them.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Mon Aug 22, 2011 2:47 am UTC

TrlstanC wrote:If we want to replace the term “justified” in the classical definition of knowledge, we could instead use a definition of “knowledge is a logical true belief.” This would give us these requirements for perfectly judging knowledge, call it the Strong definition of knowledge:

If:
1. S believes p
2. p is true
3. Given S’s assumptions it's possible to logically conclude p
4. S’s assumptions are true
5. We have perfect information to judge if 1, 2, 3 and 4 are true

Once again, with Barn County:

  1. Smith believes that the thing she's looking at is a barn.
  2. The thing she's looking at is a barn.
  3. Smith, assuming that our hypothetical account of how she forms her belief fits ordinary human behavior, is making an assumption that is essentially identical with her belief: She looks at the barn and says "That's a barn." Certainly, you can conclude that the thing she's looking at is a barn from this assumption.
  4. Smith's assumption is true.
  5. Obviously, we have perfect information, since we invented all relevant information ourselves.
Under your definition, then, Barn County is actually an example of knowledge, in flat contradiction of intuition. Now, in your edit you even acknowledge that this may be the case: Smith knows that she's looking at a barn if she just looks at one barn, but if she sees a lot of barns, then the occasion where she happens to be correct isn't knowledge. However, this misapplies your own standard. When Smith, if she behaves at all like a real person, looks at a series of barns, she doesn't assume that she's looking at a real barn in each case because she previously saw other examples of what she thought to be barns. Instead, she makes each judgment independently: That's a barn, that's a barn, and that over there is Cabot a barn. No matter how many similar, incorrect assumptions she has previously made, Smith's assumption on the occasion when she is actually looking at a barn is a correct one. Thus, your analysis of knowledge would ascribe knowledge to S even in a case where you yourself admit that she has no such thing.

An even clearer demonstration of the inadequacy of this standard is that someone can validly infer a true belief from a true, but unjustified, assumption. It's implausible to call such beliefs knowledge, but your standard does just that. For example, I might assume without justification that my friend is in Rio de Janeiro right now, and from this assumption conclude that she is in Brazil. If, by happenstance, she actually were in Rio de Janeiro, criteria one through four would be satisfied, and a perfectly-informed judge would conclude under your standard that my belief is knowledge. But it isn't, due to its shoddy justification.

What I find most perplexing about your analysis, however, is your attempt to work our information about the criteria for knowledge into the definition itself. This is akin to saying that a chess player is in checkmate if a) her king is in check, b) she cannot meet the threat, and c) we have perfect information to judge if (a) and (b) are true. Likewise, your "individual definition of knowledge" is akin to saying that there's a kind of checkmate that exists if a player thinks that (a) and (b) are true. This is absurd. Whether a player is in checkmate or not has nothing to do with what we think about it. All you're uncovering with your observation is that there's a difference between thinking X and X being true. But this isn't anything special about knowledge; it can fit literally any proposition.
Nothing rhymes with orange,
Not even sporange.

Greyarcher
Posts: 708
Joined: Thu Oct 18, 2007 3:03 am UTC

Re: Is knowledge justified true belief?

Postby Greyarcher » Mon Aug 22, 2011 4:27 am UTC

Hmmm. It seems like the conclusion on whether something is knowledge or not tends to be affected by "extraneous" factors (e.g. incidental coins in pockets, the presence/absence of fake barns). This seems a touch peculiar. Presumably, we may add/subtract these factors however we please in our hypotheticals; thus, the presence of knowledge in people fluctuate according to incidentals being plunked down in their hypothetical, none of which they are aware of?

If we may do things like undetectably alter T or the probability of T from scenario to scenario, we're like a Cartesian Demon of hypotheticals. It doesn't seem like there's a methodology that allows hypothetical worlders to avoid our treacherous changes--not unless they have perfect information, in which case the issue is trivial.

I've become a bit skeptical of the whole "inquiry about knowledge" business anyway. If it's not abstract frippery, but driven by a notion of some practical value, I wonder if there isn't a more practical mode of thought that bypasses the entire need for such an inquiry. And if it's merely about coherent describing an ideal for use as a reference point, then it seems like the "conclusion of perfect information, perfect logician" type scenario would be enough.
In serious discussion, I usually strive to post with clarity, thoroughness, and precision so that others will not misunderstand; I strive for dispassion and an open mind, the better to avoid error.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Mon Aug 22, 2011 2:36 pm UTC

TheGrammarBolshevik wrote:[*]Smith, assuming that our hypothetical account of how she forms her belief fits ordinary human behavior, is making an assumption that is essentially identical with her belief: She looks at the barn and says "That's a barn." Certainly, you can conclude that the thing she's looking at is a barn from this assumption.

This raises the question of what is “ordinary human behavior” and do we know enough about how our brains form assumptions, beliefs and logical conclusions to definitively say exactly what “ordinary human behavior” is? In a perfectly hypothetical situation we have to spell out exactly all the assumptions being made. Especially in such extreme examples like barn country because our everyday understanding of terms like “knowledge” can fail us when we’re dealing with very unordinary situations. If Smith looks at anything that looks like a barn, and without any other mental processes (either conscious or unconscious) believes “that’s a barn” than I don’t know if I would call that “ordinary human behavior, it would seem to be more like an automated image recognizing robot. But if we were to assume that that’s how humans think, that there’s only 1 step between visual input and the formation of a belief, than I think we would have to say that Smith doesn’t know it’s a barn. A barn isn’t defined by what it looks like, it’s defined by how it’s used. If Smith is completely ignoring any information about how barns are used then she will never be able to actually know if something is a barn or not. If we want to make the hypothetical situation slightly more realistic then we can say that Smith assumes that things that look like barns are used like barns. Although I would hope that the amount of prior observations and knowledge that’s used when humans judge “that’s a barn” is at least an order of magnitude larger than just 1 or 2 simple assumptions. But in a hypothetical barn country situation, we can come up with different assumptions about Smith’s mental process, and the assumptions she makes that will lead us to either conclude that she knows it’s a barn or not. Which naturally leads us to the question “what assumptions are we actually making, consciously or unconsciously, when we think something’s a barn?” I don’t think we can answer that question yet.



TheGrammarBolshevik wrote:An even clearer demonstration of the inadequacy of this standard is that someone can validly infer a true belief from a true, but unjustified, assumption.

This is a good point, we would need to change #4 from:

4. S’s assumptions are true

to

4. S’s uses only assumptions she knows.

This would incorporate the requirement for true assumptions, but also make sure that they can be logically concluded instead of just assuming facts that happen to be true. And pushes back the judgment of knowledge a step further, this ensures that ultimately we’ll only ever use information from observation or appeal to authority (which is just a different kind of observations really).


TheGrammarBolshevik wrote:"individual definition of knowledge" is akin to saying that there's a kind of checkmate that exists if a player thinks that (a) and (b) are true. This is absurd.

That’s not absurd, that’s called a mistake, and whole point of differentiating between a perfect judgment of knowledge, and common, or individual judgments of knowledge is to acknowledge that sometimes people will make mistakes because they don’t have access to perfect information. In this example, if two people were playing chess, and both agreed that one player was in checkmate, they would both say that they “knew” there was a checkmate, even if a valid move existed (that they didn’t see). They made a mistake, this is how our everyday judgment of knowledge works, sometimes we’re wrong. On the other hand if we want to make a perfect judgment of knowledge in hypothetical situations, we can assume we have perfect information and never make a mistake.
Last edited by TrlstanC on Mon Aug 22, 2011 6:51 pm UTC, edited 1 time in total.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7604
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Is knowledge justified true belief?

Postby Zamfir » Mon Aug 22, 2011 2:51 pm UTC

Greyarcher, there might be connection between the inadequacy of the "perfect everything" ideal, and the extraneous factors.

After all, extraneous factors are not special tricksiness of philosohical toy problems. The real world is crammed with extraneous factors. It's usually philosophical examples that are atypical in their absolute clarity of which factors are relevant and which can be ignored.

In the real world, "perfect information and perfect reasoning" is not only an unattainable ideal, it is also an ideal that cannot be approximated in a clear-cut way. To know when you are "close" or "close enough" to it, you need some extra concepts to guide you on which of the infinite amount of possible factors have to be included, which can be ignored, and how you know the relevant factors in the first place.

Basically, it would be nice if we could avoid too much reliance on "perfect information". After all, we do seem to be capable of making up heuristics on when to treat statements as facts or beliefs as knowledge, even in the absence of perfect information. But those tend to be context- and domain specific, and sometimes they fail. It can't hurt to think about general principles for this, to think about theoretical failures modes that heuristics are likely to encounter.

User avatar
TheGrammarBolshevik
Posts: 4878
Joined: Mon Jun 30, 2008 2:12 am UTC
Location: Going to and fro in the earth, and walking up and down in it.

Re: Is knowledge justified true belief?

Postby TheGrammarBolshevik » Mon Aug 22, 2011 8:28 pm UTC

TrlstanC wrote:But if we were to assume that that’s how humans think, that there’s only 1 step between visual input and the formation of a belief, than I think we would have to say that Smith doesn’t know it’s a barn. A barn isn’t defined by what it looks like, it’s defined by how it’s used. If Smith is completely ignoring any information about how barns are used then she will never be able to actually know if something is a barn or not. If we want to make the hypothetical situation slightly more realistic then we can say that Smith assumes that things that look like barns are used like barns.

Well, if she assumes this then she must be right or wrong. If she's right, your standard leads us to conclude that she knows she's looking at a barn, even though she doesn't. But, if she's wrong, then your standard leads us to conclude that she never knows when she's looking at a barn, even if she's worlds away from any fake barns.

TrlstanC wrote:
TheGrammarBolshevik wrote:An even clearer demonstration of the inadequacy of this standard is that someone can validly infer a true belief from a true, but unjustified, assumption.

This is a good point, we would need to change #4 from:

4. S’s assumptions are true

to

4. S’s uses only assumptions she knows.

This would incorporate the requirement for true assumptions, but also make sure that they can be logically concluded instead of just assuming facts that happen to be true. And pushes back the judgment of knowledge a step further, this ensures that ultimately we’ll only ever use information from observation or appeal to authority (which is just a different kind of observations really).

Why would you want to push the judgment of knowledge back if defining knowledge is exactly what you're trying to do? What you've done here is made your definition circular: To say whether something is knowledge, you already have to know whether the thing that it rests on is knowledge. To apply to my example again: Instead of asking whether my friend is in Rio de Janeiro (my assumption), we can ask whether I know she's in Rio de Janeiro. This means running through your criteria again: I believe she's in Rio de Janeiro, she is in Rio de Janeiro, this follows from my assumption that she's in Rio de Janeiro. So now we're back at 4, asking "Do I know my assumption (i.e., that my friend is in Rio de Janeiro)?" But that's precisely what these criteria are supposed to tell me!

TrlstanC wrote:In this example, if two people were playing chess, and both agreed that one player was in checkmate, they would both say that they “knew” there was a checkmate, even if a valid move existed (that they didn’t see).

Sure, but knowing about the valid move wouldn't help them if they didn't know the criteria for a checkmate. You seem to think that stipulating that we know all the details of any test case for knowledge gets us closer to knowing the criteria for knowledge, and it doesn't. I mean, look at your original "strong form" of the definition of knowledge, which is just the classic JTB definition:
TrlstanC wrote:Strong form - "He knows X"
1. He believes X
2. He has a good reason to believe X
3. X is true

I stress: This is the analysis of knowledge that Gettier's paper refutes. In Gettier's examples, lack of "perfect information" is not the problem. We have all the information we need to say that 1, 2, and 3 are satisfied, but there still isn't knowledge. Any additional, "perfect" information is superfluous.
Nothing rhymes with orange,
Not even sporange.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Mon Aug 22, 2011 9:57 pm UTC

I’ll try to go through and point out the ways in which I think this definition of knowledge (we can call it Logical True Belief) works in coming to the correct conclusion in each of the cases raised. I think we would all hope that there would be a straightforward and intuitive definition of knowledge to use in all cases, but if we’re going to test the limits of the definition in hypothetical situations we have to acknowledge that if we rely on false or limited information at any point than there’s the possibility for incorrect judgments. The additional criteria are to disallow any false or limited information in the requirements.

Hopefully this will illustrate the point that in the real world we’ll never be able to judge knowledge perfectly, we’ll have to rely on less than perfect information and we’ll occasionally think we know something when we don’t. However, if we want to be able to make perfect judgments of knowledge in hypothetical situations we have to assume we can use perfect information to judge all the requirements, and will therefore also have to perfectly describe all assumptions that are being made. This will also highlight the fact that attempting to perfectly describe all the assumptions that any human uses in their everyday decisions is currently limited by our knowledge of how our brains work, we simply don’t have the information to correctly model these assumptions in all cases.

To summarize the requirements I’m suggesting, in their latest, most clearly stated form, we can judge that “S knows p” if and only if:

1. S believes p
2. p is true
3. Given S’s assumptions it's possible to logically conclude p
4. S’s assumptions are logical true beliefs
5. We have perfect information to judge if 1, 2, 3 and 4 are true

(#5 isn’t strictly necessary for the definition, but it does make it clear the information we’re assuming, this could easily be changed to any other kind of available information for different scenarios.)

TheGrammarBolshevik wrote:
TrlstanC wrote:But if we were to assume that that’s how humans think, that there’s only 1 step between visual input and the formation of a belief, than I think we would have to say that Smith doesn’t know it’s a barn. A barn isn’t defined by what it looks like, it’s defined by how it’s used. If Smith is completely ignoring any information about how barns are used then she will never be able to actually know if something is a barn or not. If we want to make the hypothetical situation slightly more realistic then we can say that Smith assumes that things that look like barns are used like barns.

Well, if she assumes this then she must be right or wrong. If she's right, your standard leads us to conclude that she knows she's looking at a barn, even though she doesn't. But, if she's wrong, then your standard leads us to conclude that she never knows when she's looking at a barn, even if she's worlds away from any fake barns.

Correct, and Smith’s proximity to any fake barns has no bearing on whether she knows if something is a barn or not. If she has to rely on the incorrect assumption that “everything that looks like a barn is a barn” than we have to conclude that she never really knows if something is a barn. Using this definition allows us to correctly identify that the weak point in the hypothetical situation is ignoring that a barn is defined by it’s function, without knowing what a building is used for we can never know if it’s a barn or not. If we allow Smith to know what a building is used for, we should never come to a false conclusion of knowledge.

TheGrammarBolshevik wrote:Why would you want to push the judgment of knowledge back if defining knowledge is exactly what you're trying to do?

Because we have to recognize that knowledge doesn’t exist in a vacuum, it’s based on something. We say that something is known if it’s based on incorrect information. We acknowledge this limitation in the real world, and therefore should include it in the definition. Also, the definition is still useful in hypothetical situations where we can state definitively whether an assumption is known or not – we don’t have to go back any further, but we can if it’s useful for our hypothetical situation.

TheGrammarBolshevik wrote:What you've done here is made your definition circular: To say whether something is knowledge, you already have to know whether the thing that it rests on is knowledge.
That’s correct, and it's absolutely the point. And if we create a hypothetical situation to test if S knows p by assuming “S assumes they know p" we will end up with a circular situation where we can never conclude that S knows p. Since this fails one of the criteria, this seems like the correct result to me.


TheGrammarBolshevik wrote:
TrlstanC wrote:In this example, if two people were playing chess, and both agreed that one player was in checkmate, they would both say that they “knew” there was a checkmate, even if a valid move existed (that they didn’t see).

Sure, but knowing about the valid move wouldn't help them if they didn't know the criteria for a checkmate.

Correct, but if we assume that they know the criteria for checkmate, but don’t have perfect knowledge about the rest of the world, then it’s possible for them to make a mistake. Or they could misunderstand the criteria for checkmate. Either way, there’s the possibility that they will judge if they “know there’s a checkmate” incorrectly. But we can apply the definition to come to the correct conclusion in any case we want to specify.

TheGrammarBolshevik wrote:You seem to think that stipulating that we know all the details of any test case for knowledge gets us closer to knowing the criteria for knowledge, and it doesn't.
That’s correct, having perfect information isn’t a prerequisite for having a definition of knowledge. I believe that “justified true belief” will work almost all the time in everyday usage, however there are "edge" cases where it fails. If we instead use the definition “logical true belief” as I’ve laid out then I don’t think there are any cases where it will fail. Of course we have to recognize that it’s not possible to properly judge a definition by creating a hypothetical situation that’s vague or open to misinterpretation. In addition, if we want to be able to judge knowledge perfectly in every case then we have to accept that we’ll need perfect information to be able to judge it. We’ve accepted that we have to use perfect information to judge whether a belief is true, I don’t think there’s any argument on that point. If we simply extend the use of perfect information to the other criteria, and replace the requirements that are implied by “justified” with carefully spelled out requirements that acknowledge we can use perfect information in some hypothetical situations then we can use the definition, and not run in to any situations where it comes to a conclusion that is incompatible with our everyday understanding of what it means to know something.


----------------------------------------------------------------------------------
Fake edit: this is pretty off topic at this point, so I hid it in spoiler tags so it won't drag the discussion too far off, but it's still worth mentioning.
Spoiler:
Obviously at this point it should be clear that the original requirements I laid out weren’t sufficiently clear. The real point I was trying to make in the original post was that “justified” had implied requirements that weren’t compatible with a perfect judgment, and that we were already assuming perfect information in some of the requirements of JTB, so why not use it for all the requirements? “He has good reason to believe X” is not the same as “justified” in particular the change was meant to highlight that we’re relying on more information that just what’s available to the subject. Obviously more detail was needed to spell out exactly what I meant by “He has good reason to believe X.”

TheGrammarBolshevik wrote:
I mean, look at your original "strong form" of the definition of knowledge, which is just the classic JTB definition:
TrlstanC wrote:Strong form - "He knows X"
1. He believes X
2. He has a good reason to believe X
3. X is true

I stress: This is the analysis of knowledge that Gettier's paper refutes. In Gettier's examples, lack of "perfect information" is not the problem. We have all the information we need to say that 1, 2, and 3 are satisfied, but there still isn't knowledge. Any additional, "perfect" information is superfluous.

If we’re going to learn anything by discussing the topic we can’t go back to dredge up old points that have already been discussed and updated. We should be discussing this definition, or something functionally equivalent to it:

1. S believes p
2. p is true
3. Given S’s assumptions it's possible to logically conclude p
4. S’s assumptions are logical true beliefs
5. We have perfect information to judge if 1, 2, 3 and 4 are true

Greyarcher
Posts: 708
Joined: Thu Oct 18, 2007 3:03 am UTC

Re: Is knowledge justified true belief?

Postby Greyarcher » Mon Aug 22, 2011 11:23 pm UTC

Zamfir wrote:In the real world, "perfect information and perfect reasoning" is not only an unattainable ideal, it is also an ideal that cannot be approximated in a clear-cut way. To know when you are "close" or "close enough" to it, you need some extra concepts to guide you on which of the infinite amount of possible factors have to be included, which can be ignored, and how you know the relevant factors in the first place.

Basically, it would be nice if we could avoid too much reliance on "perfect information". After all, we do seem to be capable of making up heuristics on when to treat statements as facts or beliefs as knowledge, even in the absence of perfect information. But those tend to be context- and domain specific, and sometimes they fail. It can't hurt to think about general principles for this, to think about theoretical failures modes that heuristics are likely to encounter.
Mm, but if we're trying to acquire decent beliefs and aren't concerned about unattainable ideals, then wouldn't it make more sense to focus directly on the methodology of acquiring a quality belief? Rather than trying to describe what the highest quality of belief (e.g. knowledge) is, a shift in focus to methodology seems to handily avoid getting caught up on Fake Barn type scenarios while still serving the same pursuit of top quality belief.

For instance, I think there's little problem with someone thinking "That's a barn" in Fake Barn Country if they have no familiarity with regions full of fake barns, since the data available to them suggests that it is a barn. They just need to, say, be able to revise their belief in the face of new information, and/or be willing to investigate the barns further to test their belief from different angles. If we want to exert especial caution, we might insist that the proper thought is "That looks like a barn", and the conclusion "That is a barn" should only come after investigating further. Or such.
In serious discussion, I usually strive to post with clarity, thoroughness, and precision so that others will not misunderstand; I strive for dispassion and an open mind, the better to avoid error.

User avatar
Zamfir
I built a novelty castle, the irony was lost on some.
Posts: 7604
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: Is knowledge justified true belief?

Postby Zamfir » Tue Aug 23, 2011 6:15 am UTC

I'd say that Gettier problems are fairly illustrative in such a process. What they underline is that the process that leads people to knowledge is in important, that you looking at people's beliefs is not enough, that you need to look at least as carefully at the generating process. That is still true if your goal are less-then perfect levels of truth and knowledge. If someones belief meets a hypothetical "good enough", the process that generated the belief has to be good enough too, otherwise they are still just guessing lucky.

The barn problem does appear in reality, I think. Take the credit crisis: lots of people followed the rules and experience in judging investments that they had build up over their entire career, that were the best they had available, and then turned out wrong in a way they had expected. As if they were in barn country. But some people turned out right, they made their investments using seemingly the same methods as their colleagues, but found themselves on the safe side. That's pretty close to the real barn in barn country,

It seems relevant to have criteria, guidelines, to distinguish between people who knew what they were doing, and people who just happened to be right. Perhaos to reward people, to knwo who to trust in the future. But even if people retire, they might personally want to know whether they got lucky or or really knew something others didn't.

Doesn't mean philosophising is going to give clear answers, but it might be a starting point.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Tue Aug 23, 2011 1:06 pm UTC

Zamfir wrote:I'd say that Gettier problems are fairly illustrative in such a process. What they underline is that the process that leads people to knowledge is in important, that you looking at people's beliefs is not enough, that you need to look at least as carefully at the generating process.


I'd definitely agree with this, if there's anything that thinking about these problems has taught me it's that there's a huge gulf between judging knowledge in a hypothetical situation, and judging it in the real world. But still, looking at those hypothetical problems lets you get a better understanding of the process we're using everday. In particular, the importance of assumptions, without making some key assumptions, we'd never be able to know anything. I certainly don't make these assumptions consciously everytime I think or learn something (I'm not thinking to myself "I assume I can trust my observations to be true", ect.), but somewhere in our brain is something like this going on?

And as our understanding of what's going on in our brains improves, we may move away from definitions like JTB, to definitions that are a better reflection of the underlying physical process. For example, something more like "S knows p if there is a pattern of neural activity in S's brain that reflects the information contained in p, and this pattern was created by interacting with p's enviorment, or logically deduced." Obviously that's not a very rigourus definition, but it might be the direction we end up going in, especially as we improve not only our understanding of ourselves, but as we improve our ability to create machine learning.

Katrex
Posts: 7
Joined: Wed Nov 04, 2009 9:26 am UTC

Re: Is knowledge justified true belief?

Postby Katrex » Wed Aug 24, 2011 7:20 am UTC

Justified true belief is a nice easy definition, but that's not what knowledge is.

Knowledge doesn't have to be "true" the definition most people have of it anyway.To go with a pure science approach,

Knowledge is the belief that accurately explains all the evidence.
Truth requires no evidence. It is true that 2+2 is 4.

the only Truths with a capital T are provable.
Everything else is meaningless use of language and confusing ourselves.

The statements, this knowledge is true, this knowledge is false are meaningless, as knowledge is our best explanation. If it's not our best explanation, then it's just less justified belief.

Take the question "How can the universe have always existed? How can it have had a beginning" These questions are meaningless, You can't prove them mathematically and there is no evidence to use that could allow us to come up with theory to explain. They are meaningless abstract statements. All we know is the evidence suggests the universe had a beginning, that's it. It might be incomprehensible, but that is all it comes down to. It had a beginning end of.

You'll find half the questions in philosophy aren't provable and have no evidence either way. They are meaningless and that's why no one ever comes up with a real answer. People confuse matters by trying to find meaning. There's no such thing as meaning, it's a human invention so that the concept of human existence doesn't destroy us.

User avatar
TrlstanC
Flexo
Posts: 373
Joined: Thu Oct 08, 2009 5:16 pm UTC

Re: Is knowledge justified true belief?

Postby TrlstanC » Wed Aug 24, 2011 3:50 pm UTC

Katrex wrote:Justified true belief is a nice easy definition, but that's not what knowledge is.

...

You'll find half the questions in philosophy aren't provable and have no evidence either way. They are meaningless and that's why no one ever comes up with a real answer. People confuse matters by trying to find meaning. There's no such thing as meaning, it's a human invention so that the concept of human existence doesn't destroy us.



The purpose of this thread is to discuss the philosophical definition of knowledge, which is closely related to our everyday usage of the term. The question we're trying to answer is "we all use term 'to know' but what exactly does that mean?" Knowledge is justified true belief is a very good answer to that question, and in everyday usage will work as well as can be expected given our natural limitations. But will it work all the time? No, there are a number of cases (Gettier problems, and similar Gettier-style problems) where the JTB definition will give a different answer than our common understanding of the term. I believe we can use a refined definition of "logical true belief" and it will give us the same answer as we would expect from our common understanding of the term. The requirements I've laid out for "logical true belief" certainly are the simplest or most intuitive definition I've ever seen (and there's certainly the possibility that they can be improved to be easier to understand and apply), but they appear to work in all situations.

Does this help us at all when we're not posing hypothetical questions though? I think it does, the logical true belief definition makes it clear that we can't know anything without making some assumptions, at some point we won't be able to prove anything and we'll have to assume some facts are true to base the rest of our knowledge on (I believe, could it possible to prove anything without any assumptions?) The simplest assumption might be something like "my conscious observations are correct" or something similar. Are these assumptions hard coded in the way our brain works, if not, how do we learn to make the most basic assumptions we have? Also, are our brains truly logical? We certainly seem to be able to perfect execute logical deductions sometimes, but are our brains actually going through similar steps, or is there a similar process going on that arrives at the same results (at least most of the time)?

Philosophical questions are purposely unrealistic, we make simplifications and assumptions to allow us to think about complex problems that would otherwise be beyond our ability comprehend. But they also allow us to come to conclusions that we can apply to the real world, and they can show us what we still don't know, and what we need to research and study. We shouldn't expect a philosophical definition of knowledge to unlock the secrets of how humans actually learn and know things, but at the very least it should give us a tool to think about the world. The other options is to assume it's impossible to ever define what knowledge is, and therefore coming to conclusions like "there's no such thing as meaning."

User avatar
An Enraged Platypus
Posts: 293
Joined: Sat Mar 19, 2011 10:17 am UTC

Re: Is knowledge justified true belief?

Postby An Enraged Platypus » Wed Aug 24, 2011 4:22 pm UTC

Katrex wrote:Justified true belief is a nice easy definition, but that's not what knowledge is.

Knowledge doesn't have to be "true" the definition most people have of it anyway.To go with a pure science approach,

Knowledge is the belief that accurately explains all the evidence.
Truth requires no evidence. It is true that 2+2 is 4.

the only Truths with a capital T are provable.
Everything else is meaningless use of language and confusing ourselves.

The statements, this knowledge is true, this knowledge is false are meaningless, as knowledge is our best explanation. If it's not our best explanation, then it's just less justified belief.

Take the question "How can the universe have always existed? How can it have had a beginning" These questions are meaningless, You can't prove them mathematically and there is no evidence to use that could allow us to come up with theory to explain. They are meaningless abstract statements. All we know is the evidence suggests the universe had a beginning, that's it. It might be incomprehensible, but that is all it comes down to. It had a beginning end of.

You'll find half the questions in philosophy aren't provable and have no evidence either way. They are meaningless and that's why no one ever comes up with a real answer. People confuse matters by trying to find meaning. There's no such thing as meaning, it's a human invention so that the concept of human existence doesn't destroy us.


Citations needed. For everything.
We consider every day a plus/To spend it with a platypus/We're always so ecstatic/'Cause he's semi-aquatic!

- Phineas & Ferb

Katrex
Posts: 7
Joined: Wed Nov 04, 2009 9:26 am UTC

Re: Is knowledge justified true belief?

Postby Katrex » Wed Aug 24, 2011 8:16 pm UTC

If you can't understand the argument, and the simple logic of it there's no helping you. If you don't understand you wont agree, that's fine. If you're capable of understanding it you will agree.
I know that sounds arrogant but its just one of those things either you get it or you don't. You think just because it might be written in some book will make it more valid than if not?
Regardless it's self consistent, and no ones been able to find a fault in it yet. Having no citations makes no difference to it's validity. Look at the evidence, try to think up a scenarios where it's not true and decide for yourself, not because some overrated philosopher said it was the case.

Of course if you're religious and/or hold unjustified beliefs to be "TRUE" then will not fit in with your world theory, but if you "believe" in logic and evidence it is a very good definition on knowledge.

quantumcat42
Posts: 221
Joined: Mon Jun 28, 2010 6:06 pm UTC

Re: Is knowledge justified true belief?

Postby quantumcat42 » Wed Aug 24, 2011 8:21 pm UTC

Someone get Katrex the email for the President of Philosophy.


Return to “Serious Business”

Who is online

Users browsing this forum: No registered users and 6 guests