The meaning of "knowledge" [Peter]

Hypothetical:

In Town X, a burglary is reported to the police. They investigate, and their investigation leads to Bob, who has previous convictions for that sort of thing. They find the stolen goods at Bob's house, Bob is questioned, he confesses, it goes to trial, Bob pleads guilty, the jury is convinced by the evidence, and Bob is sentenced accordingly.

In fact, Bob had not committed the burglary. In fact, he had a provable alibi - he was out committing another burglary (of equivalent severity) elsewhere in town at the time. The original burglary was done by a friend of his, and he was just storing the goods at his house. Given that demonstrating his alibi wouldn't have saved him any prison time, and would have involved fingering his accomplices (which would have damaged his standing in the criminal underworld), he felt that the simplest thing to do was to take the rap for the original burglary.

This is all to illustrate a philosophical point about the word "know" (and "knowledge" etc). A poll:

(ETA Note that the previous convictions don't count when I'm asking whether the legal system knew Bob had committed burglary... I meant to ask "...committed burglary on that day" or something similar. This ETA is before anyone but me has answered this, so all the results reflect this ETA)

Poll #10225

Open to: Registered Users, detailed results viewable to: All, participants: 9

View Respondents

Did the legal system know that Bob had committed burglary?

View Answers

Yes
1 (12.5%)

No
7 (87.5%)

Was the outcome just?

View Answers

Yes - Bob committed a burglary and served a sentence for burglary, your point?
4 (57.1%)

No - Bob was punished for a crime he did not commit.
2 (28.6%)

I don't believe in punishing people for burglary anyway, so No
1 (14.3%)

Furthermore

View Answers

This is a completely ridiculous hypothetical and the legal system wouldn't have acted that way
0 (0.0%)

This is reasonably realistic, at least in terms of how the legal system acted
2 (25.0%)

Ummmm...
6 (75.0%)

March

Sun	Mon	Tue	Wed	Thu	Fri	Sat
						1
2	3	4	5	6 1	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30	31

Sun

Mon

Tue

Wed

Thu

Fri

Sat

Of course the legal system doesn't know that Bob committed this burglary. It thinks he did, but it has been misled, because he lied to it!

I want to say that the outcome was not just, but not for the reason written beside the 'No' radio button. It is unjust because Bob's friend who did commit this burglary has not gone down for it, and presumably will not do so in future because the police think the crime is solved. That seems to me a much more convincing argument for injustice than hairsplitting about whether it's just for Bob to serve the right length of sentence despite the wrong things being said about it in court.

(Meanwhile, the poor police are off on a wild-goose chase for the perpetrator of the burglary that Bob really committed, since the actual culprit now has what appears to be an ironclad alibi!)

Edited Date: 2012-04-20 09:45 am (UTC)

[link] [reply]

Note the wording: "committed burglary", not "committed this burglary".

[link] [parent] [reply]

Ah, I see. Your ETA didn't sufficiently clarify the issue in my head, perhaps due to the 'or some such'. So the question is whether the legal system knows that Bob committed some burglary on that day, and not either of whether he committed burglary at all (which it clearly does know due to his previous convictions) or whether he committed this specific burglary (which it clearly doesn't know).

I think my answer is still no, though. Suppose his alibi had been that he'd been out elsewhere committing murder. He might still have confessed to the burglary (on the basis that it carried a lighter sentence and would act as an alibi should he come under suspicion for the murder), and then it would clearly not be true that the legal system knew he had committed burglary on that day (because he didn't, and lied and said he did). No evidence relating to where Bob really was is involved in the legal system's justification for believing he committed burglary, so I think it's reasonable to say that the question of whether its belief constitutes knowledge ought not to depend on where he really was either.

This is one of those situations in which a belief is both justified (in that the legal system has good reason for thinking it), and true, but still not knowledge because the justification doesn't connect up with the truth – the justification is in error, but the erroneously derived fact turns out to be true anyway for a different reason. Like this incident.

The legal system incorrectly thinks Bob committed *this* burglary and also incorrectly thinks Bob did not commit the *other* burglary.

I think the thing that's wrong with your hypothetical is that if Bob's alibi were "I was committing another crime at the time", he wouldn't go to prison for that other crime, because you're only on trial for what you're actually on trial for, even if some other criminal activity comes up in the evidence. However I'm not completely certain this is the case, and Bob may well still care more about his standing in the criminal underworld than he does about staying out of prison.

Can his testimony in this trial count as evidence in the other trial (for the other burglary)?

Yeah, I do wonder if it's possible to give just enough information that it looks like you have a 50% probability of committing the other burglary -- enough to prevent reasonable doubt that he's guilty of the first burglary, but not enough to say beyond reasonable doubt he's guilty of the second. But I don't think it actually works.

It's not quite the same, but I did once try to figure out what happened if you were charged with two related crimes, and they could prove you were guilty of one or the other but not both. And there was an obscure case where someone was in fact found guilty of the lesser of them. But I don't know if there's an official answer to what's supposed to happen.

A defence of alibi needs other witnesses, or clear evidence, to speak to it. You can't just assert alibi, you must prove it to the court's satisfaction (it's a reversal of the usual burden of proof, albeit not to as high a standard).

Sufficient evidence to prove alibi would very likely be sufficient to convict him.

I guess it depends when he gives the alibi. If he gives it while being questioned by the police, then presumably he would have ended up going to trial for the correct burglary - the wording in the post suggests to me that Bob makes the choice not to give the alibi when he confesses to the police.

I'm not sure, but I think that if you admit to a different crime during a trial you are likely to be rearrested because of what you have said. But I can't remember whether I only think this through watching shows like Law and Order UK :)

I think the application of "know" to a system is a bit problematic. Not to mention that part of what the legal system does is make rulings as to what is and is not a "fact" for legal purposes.

(Is the underlying question about "knowledge" that isn't actually true and may be falsified, despite following best practice to be sure of that knowledge?)

Certainly the institutions and the legality add problems. The core of it is: there's an old idea in philosophy that knowledge is "justified true belief", and then someone called Gettier kept coming up with situations where there were true beliefs where something was up with the justification. However I think I prefer my hypotheticals to his.

There was an incident in an RPG I was once playing. In the Warhammer 40,000 universe, there are psykers - people with psi powers. We were playing a group working for the inquisition. As far as we were concerned, psykers came in two types; sanctioned and unsanctioned. Sanctioned psykers were OK, we had one in our party. Unsanctioned psykers... there's a procedure for sanctioning them, but for those without the authority to do so, it's kill on sight.

In our party, we had the Assassin, the Scum (basically a con-man character) and the Sanctioned Psyker. Previously there had been incidents that caused the Assassin to hate the Scum. Also the Scum got corrupted by Chaos, and mutated, and developed psyker abilities, which he kept secret. Then there was a fight with some space dwarves. The Sanctioned Psyker used an ability, and had a freak psychic accident that caused him to swap minds (well, bodies - the bodies stayed in place and the minds switched) with the Scum. The Sanctioned Psyker, in the Scum's body, then proceeded to use a highly visible psi power for self-protection. The Assassin saw this, shouted "Unsanctioned Psyker!" and proceeded to try to gun the Scum down. And I thought: I've read web pages on philosophy that deal with this sort of thing.

It seems realistic to me, but my knowledge of the legal system, wrt matters of this sort, is based more on TV than experience, so judge of that what you will :)

The reason I chose 'no' to the question about it being just is that Bob had a strategy, admitting to the crime he didn't commit in part to maintain his standing in the criminal underworld. Also, he deliberately prevented his friend from being arrested for a crime he committed (which might have been the result of further investigation if he had not confessed, even if he had not said 'My friend did it'), which is in itself a crime that he hasn't been put on trial for.

But as you can see, all my quibbles are about justness, not about knowledge, so I should stop there :)

I've heard brief mentions of the discussion of what it means to "know" something.

However, to me, the mistake seems to be assuming that because we have a word "know", that it maps cleanly to a relevant philosophical concept (that fulfils all the expectations of "something you can reason with", "something you have good reason to think true", etc, etc). It seems more likely that there is (at best) a number of overlapping concepts that "know" describes well when they coincide, but people would disagree about when they don't.

So I realise this is one of the edge cases, but I'm not sure the answer is "yes" or "no"... :)

In Spanish, you use a different verb for knowing a fact (saber) from what you use for knowing a person (conocer). So in some languages there is already a primary devision between concepts that we group together under the umbrella term 'knowing'.

(Obviously Spanish is not the only language in which this is the case, it's just the one that comes most easily to my mind because I've spent time studying it.)

P.S. This should probably spark a discussion of whether you can ever really know a person :)

True, but I think people accept that "know a person", "know a skill", "know a fact", etc are different. I don't think anyone would seriously suggest that the justice system would have made the acquaintance of Bob committing a burglary, or have the skill of Bob committing a burglary -- I think everyone is wondering whether it can be said to know the fact of Bob committing a burglary. I am curious if there's languages that break down "know a fact" further.

Yes, I was trying to think of further breakings down of the concept in other languages but didn't come up with anything useful :) There are some technical ways in which Greek philosophers talk about knowledge and truth and so forth using various technical vocabulary, but that's not quite the same.

I'm definitely not sure that I know any language where 'knowing something that really has happened because you have been convinced by evidence that it has happened' is distinguished from 'thinking you know that something has happened because you have been convinced by evidence that it has happened when in fact it hasn't happened but something else equivalent has, with that result that you sort of know something that has happened for which you didn't have any direct evidence, and where you will never realise that this isn't knowledge type A' :)

'thinking you know that something has happened because you have been convinced by evidence that it has happened when in fact it hasn't happened but something else equivalent has, with that result that you sort of know something that has happened for which you didn't have any direct evidence, and where you will never realise that this isn't knowledge type A' :)

:)

Yeah, not really expect that :) The closest I can think of is when people use varying emphases and modifiers, like "I know it, I just KNOW it" means "I have a strong feeling but no evidence" or "You can't KNOW that" means "it's pretty certain that it is, but it's always possible you've misinterpreted it"

Ah, yes, they're good e.g.s. Unfortunately, they also involve breaking linguistics :) When you talk about a language having different words for different concepts, you're looking at good old fashioned etymologies and traditional semantics and so on. But when you talk about people meaning something different by the way they emphasise a word, you're into the field of pragmatics (which is all about why things mean what they mean in context), which is a relatively new concept[1]. Obviously when talking about philosophical and to some extent psychological aspects, this is perfectly justified. It's just a by-product that my head responds by dividing the two concepts and keeping them separate :)

[1] A case in point re pragmatics: 'relatively new concept' means something a bit different to a classicist like me from what it might mean to e.g. a modern linguist ;)

Pragmatics: mid-20th century it seems.

Yes exactly, thereabouts - this is very late by my internal timeline :)

Have you ever come across evidential markers in Quechua? I once had a colleague who studied Quechua in Ecuador for a while, and gave a talk in Cambridge about the overt marking of one's degree of certainty in a proposition. We all thought it would be a good language to require politicians to use :-)

http://en.wikipedia.org/wiki/Quechua_languages#Evidentiality

Oh that's fascinating - thank you for that link. My experience of Quechua has been very fleeting (one MPhil seminar, and a snatch in Indy 4, etc!), and I didn't know about the evidentiality markers at all. What a wonderful system!

I'd heard of evidential markers; I hadn't connected it with Quechua though. I had some memory of something along that lines being part of some Indonesian language, although maybe that's me just getting confused.

German has a distinction between kennen and wissen; again, kennen is typically for people and wissen for facts, but a brief google suggests it's more complicated than that.

And French has connaître and savoir.

Was it just? I'd say, more or less. As in, in this isolated incident, a satisfactory outcome was achieved, which is just. But it's still indicative that the system as a whole isn't just, because there's evidence that more probably other, similar mistakes are made that include guilty people being free and innocent people being punished. But it doesn't seem urgent whether or not that counts as "just", only whether we should care if it happens or not.

P.S. In Greek, 'I know' is 'oida'. Although this is semantically present 'tense', its form is that of the perfect 'tense' of the verb 'to see'. So in Greek, 'I know' explicitly means 'I have seen'.

There you go, that was more fun for a discussion of knowledge as a concept :)

(Scare quotes for 'tense' because it's a complex concept!)

That's interesting, and possibly relevant, given that "justified true belief" is thought to have been one of Plato's ideas. Let's see... in the Theaetetus it seems that the first definition of knowledge discussed is perception. So maybe there was an argument from etymology in there somewhere.

It's ages since I read Theaetetus, I have to admit, but I think there might be a kind of implicit awareness of etymology in there... I forget the details, sorry.

It seems to me that when we combine your hypothetical here, my one (and the one in the comments) in the old LJ post I linked to, and Gettier's ones in the Wikipedia page you link to, the common thread is that all of them have a true belief of the form "One of the following set of possibilities is true", with a justification "Specifically, I have reason to believe it's this one". In fact it then turns out that the thing asserted in the justification is false in spite of looking plausible, but fortunately one of the other possibilities in the set is true, and thus the originally stated belief manages to be true despite the justification being wonky.

Are there any examples of this phenomenon which don't have that form?

In a comment above I had something from an RPG which doesn't obviously fit that form, although maybe it could be argued to have it. It's also complicated by questions of personal identity which are a whole different philosophical can of worms.

Hmmm. So in that situation, the Assassin has the belief "The Scum is an unsanctioned psyker", justified by the observations that (a) the Scum is not a sanctioned psyker, and (b) the Scum just used a psi ability. But in a world where mind-swapping is actually possible (and moreover can happen unexpectedly and without obvious external signs), we have to expand that reasoning a bit and expose the implied assumption, so that it goes more like: (a) the Scum's mind is not a sanctioned psyker, (b) the Scum's body just used a psi ability, (c) we presume that the Scum's body is currently occupied by the Scum's mind. Of course the error in the reasoning is (c), and for all that it's an understandable assumption if mindswaps are possible but very rare, it's still inaccurate.

(Also, the above reasoning only yields the belief that the Scum's mind is an unsanctioned psyker; that must be combined with point (c) a second time to reach the conclusion that it's legitimate to gun down the Scum's body right now. That latter conclusion is certainly not knowledge, since it's not even true. Only the intermediate conclusion about the Scum's mind constitutes the sort of quasi-knowledge we're discussing here.)

I think I'd agree that it's not clear that this fits into my suggested unifying pattern.

I'd say the basic fact is correct (Bob commited a burglary) but the causality chain is wrong.

Peter

The meaning of "knowledge"

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

(no subject)

March