JTB: the myth of propositions and the Gettier problem

[Terrapin Station]

Basically, on my view, the notion of "accidental" knowledge is incoherent, because knowledge is a type of belief, and we don't have "accidental" beliefs--that is, it can't turn out that we believed something that wasn't present in our mind in a particular way.

Smith can't turn out to have an "accidental" belief that Brown is in Barcelona. The only belief that Smith has is what's consciously present in his mind, and that happens to be "Jones owns a Ford (and you make a disjunction of that with any random garbage and it's still true, because Jones owns a Ford)."

When it turns out that Jones doesn't own a Ford and Brown is in Barcelona, that's not a belief that Smith had, even as a disjunction, so it's not a justified true belief on his part. He didn't have "accidental" knowledge.

[Consul]

I'm referring to my response to that, where I specifically addressed the original case presented by Gettier, here: viewtopic.php?p=387829#p387829

The case exemplified by Gettier clearly states that the proposition "the man who will get the job has ten coins in his pocket" is the conclusion that Smith reaches after processing some bits of information. The set of information is a reduced to a statement from his boss about a future action (hiring Jones), that Smith cannot control, and the account of money in Jones's pocket, an action that ocurred in the past. A simple look at these situations show that Smith never had a strong justification for believing what we're told he believed. First, a promise from his boss could never guarantee that Jones would be hired because 1) his boss could lie, 2) his boss could change his mind, 3) his boss could be forced by unexpected events to change his decision, 4) Jones could become unaivalable at the moment of hiring for many reasons. These are all possibilities that any rational person can take into account and most often lead to withhold judgement about the occurrence of future events, yet Gettier's example characterizes Smith as a naive fool that buys into any promise people make, basically hearsay, and this is supposed to be a justified true belief. Consider, on the other hand, if the type of information that Smith handled was the company's actual job contract with Jones or a company's memo stating such promise. In such cases there would be no possibility of this statement in Gettier's problem being true: "But imagine, further, that unknown to Smith, he himself, not Jones, will get the job." Or imagine that the information was not about possible future events, but about actual current properties or events witnessed by Smith and directly linked to Jones, in which case it would be fair to believe and state that the carrier of these properties (Jones) has ten coins in his pocket.

Let's look now at the other pieces of information: first, Smith has counted ten coins in Jones' pocket. The problem implies that this is all the information that Smith can have about money in people's pockets, which is evidently not true. Again, characterizing Smith as a naive fool, we're told he never actually checks the state of other people's pockets, including his own, yet infers that "there is a (exactly one) man who [...] has ten coins in his pockets", and this is supposed to be a justified true belief. Smith also never takes into account all the possible scenarios of money in people's pockets after he counted the money in Jones' pocket.

And so, we're told that these pieces of information can only take Smith to infer that the man who will get the job has ten coins in his pocket, which is evidently not JTB from the start. Conveniently for Gettier, he places Smith thinking of Jones when referring to the man that will get the job, but dissolves that instance into a generic "man", and implying that Smith has considered all the possible candidates that fall in that category, also evidently not true. Gettier has only played with ambiguities, making open undetermined scenarios to look as closed, determined scenarios.

In conclusion, Gettier's problem only works assuming a very stupid person making false inferences. Stupidity cannot be the basis for studying epistemology.

I concede that Gettier's original examples are pretty contrived, which is not to say they're totally incredible.

Do you deny the justifiedness or the truth (or both) of Smith's belief that the man who will get the job has ten coins in his pocket?

Anyway, as for semantics, from Smith's perspective the (intended) referent of "the man…" is Jones, whereas, unbeknown to Smith, the actual referent is Smith himself. So the definite noun phrase "the man…" is a "non-rigid designator" that isn't invariably, context-independently fixed to one particular person. What is presupposed is that whoever it is, there is a man who will get the job; and that's the case in Gettier's scenario, isn't it?
(Note that in "there is a man…", the indefinite noun phrase "a man" is used purely quantificationally, in the sense that it doesn't refer to any particular man.)

Now the question is whether the referential non-rigidity or ambiguity of "the man…" in the sentence "The man who will get the job has ten coins in his pocket" prevents it from being true. I don't think so, because, as I said, there is a man who will get the job; so it can be true that the man who will get the job has ten coins in his pocket, no matter whether that man is Jones or Smith, or someone else.

However, there is potential problem, precisely because the sentence "The man who will get the job has ten coins in his pocket" implies that somebody will get the job, which is about a contingent future state of affairs. It's contingent in the sense that it isn't logically or physically necessary that someone gets the job. Can statements such as "It will rain tomorrow" already be true today? If they can, what makes them true?

"Future contingents are contingent statements about the future — such as future events, actions, states etc. To qualify as contingent the predicted event, state, action or whatever is at stake must neither be impossible nor inevitable. Statements such as “My mother shall go to London” or “There will be a sea-battle tomorrow” could serve as standard examples. What could be called the problem of future contingents concerns how to ascribe truth-values to such statements. If there are several possible decisions out of which one is going to be made freely tomorrow, can there be a truth now about which one will be made? If “yes”, on what grounds could something which is still open, nevertheless be true already now? If “no”, can we in fact hold that all logically exclusive possibilities must be untrue without denying that one of the possible outcomes must turn out to be the chosen one?"

Future Contingents: https://plato.stanford.edu/entries/future-contingents/

As for the question of justification, Smith's belief that the man who will get the job has ten coins in his pocket is only fallibly, defeasibly justified. For instance, if Smith learned that his boss lied about Jones getting the job, he would certainly no longer be justified in believing that Jones will get the job. What justifies Smith's belief can be defeated or undermined by additional information.

[Consul]

The relevant belief is Bill's belief that Jim is in the restaurant. I stipulated in my imaginary example that this is Bill's justified true belief which doesn't seem to equal knowledge. (Bill is an unreal person, so we cannot discover what he really believes, since unreal persons don't really believe anything.)

There's no need to pick what's the "relevant belief". I have pointed out the two beliefs that Bill has: the first one: "the man I'm seeing now is Jim". This belief is the basis for his next belief: "Jim, the man I'm seeing now, is in the restaurant". As I explained, the first belief does not meet the requeriments of the Tripartite Analysis of Knowledge, or Justified True Belief analysis. Since the second belief is dependent on the first, it cannot meet the requirements either. The whole point of the JTB analysis is that if you believe what is true, then you have acquired knowledge. And the whole point Gettier wanted to make is that there are cases in which someone believes what is true (therefore, knowledge), but yet there's no sufficient and neccesary connection between the belief and the truth. Of course, Gettier failed miserably trying to achieve it.

If Smith believed only that "Jim, the man I'm seeing now, is in the restaurant" or, equivalently, "Jim is the man I'm seeing now, and he is in the restaurant", this belief would certainly be false and not be part of a genuine Gettier case of JTB, since Jim isn't the man in the restaurant he's seeing. But—for the sake of my Gettier argument—this isn't the Gettier-relevant true belief, which is only the belief "Jim is in the restaurant". So let's not get distracted!

1. Bill believes justifiedly (but falsely) that the man he sees in the restaurant is Jim (because that man looks exactly like Jim).

It is irrelevant for the JTB analysis if there are reasons to be fooled by appearances. The condition to call Bill's belief knowledge is that it is justified true belief. To be justified true belief, it has to meet the 3 conditions previously mentioned, that's why it is called Tripartite Analysis of Knowledge. If the belief is false, then it is not considered knowledge.

Right, but the belief in question—that Jim is in the restaurant—is true!

2. So Bill believes justifiedly that Jim is in the restaurant.

This is not justified true belief, as I explained above, and therefore not knowledge.
Bill only believes that Jim is in the restaurant because he is seeing a man who he confuses with Jim. His belief is inevitably tied to what he is seeing.

Right, but his seeing of someone in the restaurant who looks exactly like Jim gives him a good reason to believe that Jim is in the restaurant. So Bill does have perceptual evidence for his belief that justifies it!

3. Bill also believes truly that Jim is in the restaurant (because Jim is in the restaurant).

If that other man moves away from the restaurant, his belief immediately changes, so it is not related to the actual Jim in the restaurant.

Wrong, because when Bill starts to believe that Jim is in the restaurant, he in fact is; so Bill's belief is true, and it's "related to the actual Jim in the restaurant", because Jim's presence in the restaurant is what makes Bill's belief true. Of course, it remains true only as long as Jim is in fact in the restaurant.

4. Therefore, Bill believes both justifiedly and truly that Jim is in the restaurant, and yet he doesn't know he is.

What BIll believes is that he is seeing a man he knows as Jim in a restaurant, but that's not the case. Since it requires that this be true in order to constitute knowledge, it is not justified true belief. Bill does not know that the man he is seeing is not the real Jim, neither he knows that Jim is in the restaurant because of this.

Once again, the Gettier-relevant justified true belief that matters in my scenario is Bill's belief that Jim is in the restaurant. This true belief is derived from a false belief—Bill's belief "The man I see is Jim"—but it nonetheless happens to be true owing to Jim's presence. It is also justified, because the man Bill sees looks exactly like Jim.

Now it is arguable that Bill still doesn't know that Jim is in the restaurant, because his belief's truth is a matter of sheer luck, which consists in the fact that, coincidentally, both Jim and Tim are in the same restaurant at the same time, with Tim being seen by Bill (and being mistaken for Jim by Bill) and Jim being unseen by Bill.

[Consul]

…But 'knowing what's so' has nothing to do with language, and therefore nothing to do with the truth-value of a factual assertion. A state-of-affairs just is or was the case, neither true not false. The claim that truth and falsehood come into 'knowing' demonstrates the conflation of what's the case with what's said about what's the case. The supposed necessity of the truth condition is a mistake.
There's much more to say about what's wrong with the JTB theory - but this is the main problem, in my opinion.

Knowledge as a kind of mental state must have some representational content. You cannot know that a state of affairs is the case unless there's a mental representation of it in your mind. If the mentally represented state of affairs is the case, then the mental representation representing that fact is veridical or true (and so is the belief containing that true representation). Are all truth-apt mental representations ones requiring a natural language such as English? Most cognitive psychologists will say no, because they believe in a nonconscious "language of thought" that isn't one of the natural languages, and doesn't depend on any one of them. So nonhuman animals lacking a natural language can still have beliefs and knowledge, because they too have mental representations in the form of a special "language of thought", whose syntactic elements aren't natural-language words or sentences.

[Consul]

No. He doesn't have a belief that Jim is in the restaurant where that's not referring to the guy he sees. Again, this is just the point. We can't just ignore it and still be talking about Bill's belief.

You're wrong, simply because "Jim" in "Jim is in the restaurant" does refer to Jim.

[Consul]

Presumably, if we were to maintain that Bill's belief is simply "Jim is in the restaurant" then we would have to say that the belief is not justified.

No, we wouldn't, because Bill does have perceptual evidence for his belief: his seeing of someone who looks exactly like Jim!

[Terrapin Station]

No. He doesn't have a belief that Jim is in the restaurant where that's not referring to the guy he sees. Again, this is just the point. We can't just ignore it and still be talking about Bill's belief.

You're wrong, simply because "Jim" in "Jim is in the restaurant" does refer to Jim.

Part of this might be based on you believing that "Jim" can refer to something independently of how particular people are thinking about it. But that's not the case. Reference is always in some particular persons' head, re how that particular person is thinking about it.

[Consul]

You're wrong, simply because "Jim" in "Jim is in the restaurant" does refer to Jim.

Part of this might be based on you believing that "Jim" can refer to something independently of how particular people are thinking about it. But that's not the case. Reference is always in some particular persons' head, re how that particular person is thinking about it.

Nobody else but Jim is the intentional object of Bill's belief in his mind/brain.

[Terrapin Station]

You're wrong, simply because "Jim" in "Jim is in the restaurant" does refer to Jim.

Part of this might be based on you believing that "Jim" can refer to something independently of how particular people are thinking about it. But that's not the case. Reference is always in some particular persons' head, re how that particular person is thinking about it.

Nobody else but Jim is the intentional object of Bill's belief in his mind/brain.

Jesus, dude. That couldn't be more wrong. Bill's belief is about the guy he sees. At least attempt to argue why Bill's belief isn't specifically about the guy he sees.

[Consul]

Nobody else but Jim is the intentional object of Bill's belief in his mind/brain.

Jesus, dude. That couldn't be more wrong. Bill's belief is about the guy he sees. At least attempt to argue why Bill's belief isn't specifically about the guy he sees.

Bill has a belief about the guy he sees—that the guy he sees is Jim—, and he has another belief about Jim—that Jim is in the restaurant. Two different beliefs, and two different objects of belief!

[Count Lucanor]

Do you deny the justifiedness or the truth (or both) of Smith's belief that the man who will get the job has ten coins in his pocket?

I do, for sure. I explained why above.

Anyway, as for semantics, from Smith's perspective the (intended) referent of "the man…" is Jones, whereas, unbeknown to Smith, the actual referent is Smith himself. So the definite noun phrase "the man…" is a "non-rigid designator" that isn't invariably, context-independently fixed to one particular person. What is presupposed is that whoever it is, there is a man who will get the job; and that's the case in Gettier's scenario, isn't it?
(Note that in "there is a man…", the indefinite noun phrase "a man" is used purely quantificationally, in the sense that it doesn't refer to any particular man.)

Evidently that's the whole point in the fallacious trick of the Gettier scenarios: to replace rigid, unambiguous designators, with non-rigid ones, so one can play with the interpretations and reach odd conclusions. He could have used non-contingent, actual events that Smith could have witnessed and assimilate as knowledge, instead of relying on possibilities, but then Gettier wouldn't have scored points, would he?

As for the question of justification, Smith's belief that the man who will get the job has ten coins in his pocket is only fallibly, defeasibly justified. For instance, if Smith learned that his boss lied about Jones getting the job, he would certainly no longer be justified in believing that Jones will get the job. What justifies Smith's belief can be defeated or undermined by additional information.

Again, if we replace "the man who will get the job", which is ambiguous, for Jones or Smith, then Smith would have clear information to form a specific belief. And this belief would be either true or false. And then there would be a chance for meeting the 3 conditions for justified true belief as knowledge.

[Count Lucanor]

If Smith believed only that "Jim, the man I'm seeing now, is in the restaurant" or, equivalently, "Jim is the man I'm seeing now, and he is in the restaurant", this belief would certainly be false and not be part of a genuine Gettier case of JTB, since Jim isn't the man in the restaurant he's seeing. But—for the sake of my Gettier argument—this isn't the Gettier-relevant true belief, which is only the belief "Jim is in the restaurant". So let's not get distracted!

No. The belief that Jim is in the restaurant is completely accesory and dependent of the belief that Jim is the man being seen. Remove that belief and the belief that Jim is in the restaurant disappears also. What matters is Bill's account of the situation with regards to knowledge, which has to do with the connection between what he believes, how true it is and how justified he is in believing it.

Right, but the belief in question—that Jim is in the restaurant—is true!

No, that's not what Bill believes. One has to distinguish between the semantics of statements and the actual mental states. When Jim sees a man in the restaurant, he also gets information about the specific place where he sits, the way he's dressing, etc., all of which create the mental image of his friend Jim in the restaurant. This mental image does not equally represent the state of the actual Jim sitting in another part of the restaurant, so it cannot be said that this is what Bill believes. Only when you utter the statement "Jim is in the restaurant" the games of interpretation begin and allow for the confusion. Once you make more specific statements like "my friend Jim, that I'm seeing now, is sitting in a square table close to the entrance in that restaurant" it becomes impossible to link that belief to the actual Jim being in another part of the restaurant.

Right, but his seeing of someone in the restaurant who looks exactly like Jim gives him a good reason to believe that Jim is in the restaurant. So Bill does have perceptual evidence for his belief that justifies it!

So what? The JTB analysis requires that what he believes is true, not only that it seems justified, and it is obvious that he's not perceiving what he thinks he is perceiving.

3. Bill also believes truly that Jim is in the restaurant (because Jim is in the restaurant).

If that other man moves away from the restaurant, his belief immediately changes, so it is not related to the actual Jim in the restaurant.

Wrong, because when Bill starts to believe that Jim is in the restaurant, he in fact is; so Bill's belief is true, and it's "related to the actual Jim in the restaurant", because Jim's presence in the restaurant is what makes Bill's belief true. Of course, it remains true only as long as Jim is in fact in the restaurant.

As I explained above, that is not the case. It's pure semantics of how you phrase the statement of what Bill believes, but it is obviously stated in the more generic terms, so it doesn't reflect what is Bill's actual mental state, which involves more specific information.

[Peter Holmes]

…But 'knowing what's so' has nothing to do with language, and therefore nothing to do with the truth-value of a factual assertion. A state-of-affairs just is or was the case, neither true not false. The claim that truth and falsehood come into 'knowing' demonstrates the conflation of what's the case with what's said about what's the case. The supposed necessity of the truth condition is a mistake.
There's much more to say about what's wrong with the JTB theory - but this is the main problem, in my opinion.

Knowledge as a kind of mental state must have some representational content. You cannot know that a state of affairs is the case unless there's a mental representation of it in your mind.

You say 'must', but why? And what is mental representational content? Why does my knowing I have two hands mean that I must have a representation of my two hands in my mind? Sorry, but this is mentalist babble. The emperor is naked.

If the mentally represented state of affairs is the case, then the mental representation representing that fact is veridical or true (and so is the belief containing that true representation).

How does a belief 'contain' a representation? Mentalist talk is laced with such metaphors. They add up to a beguiling story that doesn't bear rigorous examination.

A real image (say, a painting) may look like the thing of which it's an image (say, a person). And we may call it a 'good' or 'accurate' or 'close' likeness. We may say it correctly depicts the person. That's all the word 'true' means in that context: a true likeness.

But the JTB theory says the truth of a proposition - not an image or representation - is a necessary condition for knowing something - and that's nonsense. And the idea that 'the mind' 'contains' more or less accurate images of features of reality is a mentalist fantasy.

Are all truth-apt mental representations ones requiring a natural language such as English? Most cognitive psychologists will say no, because they believe in a nonconscious "language of thought" that isn't one of the natural languages, and doesn't depend on any one of them. So nonhuman animals lacking a natural language can still have beliefs and knowledge, because they too have mental representations in the form of a special "language of thought", whose syntactic elements aren't natural-language words or sentences.

The delusion spirals. Why should a representation require any language, let alone a natural one? And what on earth is the language of thought? This is farcical rubbish - and it's time we called it out.

[Terrapin Station]

he has another belief about Jim—that Jim is in the restaurant.

No, he doesn't. His belief would be that Jim is in the restaurant BECAUSE that guy he's sees is Jim. The two are inseparable. Belief doesn't work so that someone has a logical consequence belief that becomes divorced from the reason that caused them to have the belief in the first place.

[Steve3007]

Presumably, if we were to maintain that Bill's belief is simply "Jim is in the restaurant" then we would have to say that the belief is not justified.

No, we wouldn't, because Bill does have perceptual evidence for his belief: his seeing of someone who looks exactly like Jim!

That perceptual evidence does not factor in the specific belief that I was talking about there, so it doesn't constitute justification for that belief. I discussed two possible different beliefs:

"Jim is in the restaurant" is true but not justified.

"This particular guy whom I see in the restaurant is Jim, therefore Jim is in the restaurant" is justified but not true.

So neither of them are both true and justified. Neither is a JTB.