For the sake of this discussion we must take some form of realism for granted. — Michael

That can be done (I don't consider myself anti-realist). It doesn't change the meaning of 'true' in JTB. I'm arguing that 'true' just means the same as 'justified belief' and so adds nothing. I'm not arguing that from an anti-realist position, just from a (roughly) Wittgensteinean approach to meaning.

The concept 'true' is an artefact of human language and it (mostly) means something like 'everyone clever enough would agree'. I argue it means this on the grounds that this is the use context in which we find the term.

The concept 'true' is an artefact of human language and it (mostly) means something like 'everyone clever enough would agree'. I argue it means this on the grounds that this is the use context in which we find the term. — Isaac

Interesting, considering it's a logical operator. Really, it's the only part of JTB that isn't dependant on the frame of reference. I always read it as justified, believed and also happened to be true.

I'm arguing that 'true' just means the same as 'justified belief' and so adds nothing. — Isaac

Long time! I think "justified" just means "with good reason", not "retrospectively justified". If my partner says my keys are on the coffee table and the kids confirm it, that justifies my belief that the keys are indeed on the coffee table, whether or not they are all mistaken.

It's supposed to distinguish from beliefs that are reached erroneously, but may also be true.

How can an expression convey a weather condition? — Isaac

By being about it? I honestly have no clue what you're trying to ask here.

Most of the time, it's used to get the listener to believe it's raining (by which I mean have a tendency to act as if it's raining - put a coat on, carry an umbrella, write a poem about it...). — Isaac

Most of the time, it's used to inform the listener that it's raining (by which I mean to convey information necessary for the listener to adapt to the rain - put a coat on to avoid getting wet, carry an umbrella to avoid getting wet, write a historically accurate poem about it...).

Rearranged for (hopefully) clarity:

We've gathered more justifications for believing it, but in JTB, we already have justifications and beliefs, the question is how to add the T. — Isaac

We don't add the T. The T is a relationship between the meaning of the claim and the state of affairs. The claim's meaning implies some truth conditions. The claim is true if the described state of affairs meet the truth conditions. A claim can be true even if nobody has any justifications for it.

Justifications are what you use to figure out what things are true.

But we haven't ascertained its veracity, you admit yourself, we could still be wrong. — Isaac

But "could be wrong" does not entail "being wrong". Assuming it's justified and believed, "being wrong" about its truth implies the claim is not true; that would make it a JFB. "Being right" implies the claim is true; that would make it a JTB.

You said
"I don't see what's stopping us from looking out windows. — InPitzotl"
in response to my reductio of "I know..." requiring the subject to be 'true'. — Isaac

More specifically I said that in response to this:

Either that or this ludicrous situation where a word refers to something we can't ever ascertain... — Isaac

...we can (aka "can ever") ascertain truth using justification.

We just gain more justification for our belief that it's raining. At no point do we find out that 'it's raining' is true, — Isaac

You apparently mean to talk about certainty (in a mathematical sense; philosophical sense?) that a thing is true, not "finding out". It's either raining or it's not raining. I "find out" whether it's raining or not raining by looking out the window.

Option (2) isn't about anything. It's part of a whole expression-act which is about the language game of quizzes. — Isaac

Isn't that a contradiction?

I propose a test. If JTB correctly describes knowledge then every permutation should correctly describe some type of thought one might have?
Unjustified False Disbelief - unwarranted skepticism
Unjustified True Disbelief - unwarranted yet accurate skepticism
etc.

If it is a system of measuring thoughts and not just an overfit method for the things believed to be knowledge.

It doesn't change the meaning of 'true' in JTB. I'm arguing that 'true' just means the same as 'justified belief' and so adds nothing. — Isaac

We don’t need to use the term “true”. We can say that:

John knows that it is raining iff:

1. John believes that it is raining,
2. John is justified in believing that it is raining, and
3. It is raining

Where 3. is to be understood as the propositional content of John’s belief, i.e what his belief is about. It’s the mind-independent state-of-affairs that our assertions refer to according to realism.

@Isaac

This is how I view the issue (from your above statements I think we agree so just letting you know you are not insane):

We don’t know or understand what quantum phenomenon is. We have abstracted knowledge about said phenomenon that can and does bear fruit.

For material objects, like keys and such, we do not have certainty as any term (like a key) that doesn’t possess universal quality (there are a plurality of keys not a singular universal ‘key’), cannot contain certainty and therefore is knowledge based on semantic interpretation.

‘U-Knowldge’ (universal knowledge) operates differently as it is complete within a set limit under set rules. ‘S-Knowledge’ (semantic knowledge) is open to some degree of interpretation. S-Knowledge is reality driven because we do not know everything about reality (U-Knowledge is only abstracted, bounded and operating under strict rules that are known and understood).

What JTB is is a formal set of rules set up in abstraction and then extended to ‘reality’. Such ‘knowledge’ is S-Knowledge only and cannot be confirmed as U-Knowledge.

What is True is used in formal logic (which is a universal abstract) yet when this is extended to human speech and action in the lived world there is U-Knowledge. The working principles of U-Knowledge can clearly be used well in reality (this is why I mentioned quantum phenomenon as a good example to show this) even though we have little or no understanding or knowledge of what is going on. The universal abstract of mathematics can be used to model and predict what we observe with quantum phenomenon to a practical end. The certainty lies in the mathematics not in reality because the rules and limits of the mathematics used are known explicitly.

Knowledge, such as historical knowledge -or experiential knowledge of whether it is raining or not - is ‘knowledge’ bound in lived-experience. All human experience is an artifice of some proposed reality. We can dream about the rain hitting our skin and ‘truly belief’ that it is raining when it is not raining in ‘reality’. This is precisely why I refer to this kind of S-Knowledge as being defined as ‘that which we are attending to’ (in phenomenological terms Intentionality).

It is my JTB that U-Knowledge can be, and is, applied to reality because science bears results. There is no JTB that it can be applied indefinitely (extended infinitely) as we are only able to apply it to limited data sets not all possible data sets - because our scope/capacity as humans is limited.

Just to go back to the ‘rain’ issue … the semantic problem is defining what is meant by ‘rain’. Again we find the same issue as with ‘key’. Rain is not a ‘universal term’ meaning when we say ‘rain’ it is not one explicit ‘rain’ understood by everyone as we can question it: How heavy? When? Where? We cannot question abstract universals and only abstract universals can be used to create definitive answers.

Does that all make sense or am I going insane? :D

I think "justified" just means "with good reason", not "retrospectively justified". — Kenosha Kid

Ah. I had it as both. I considered that to be the distinction between justifications and reasons, the latter can't be post hoc but the former can...

It's supposed to distinguish from beliefs that are reached erroneously, but may also be true. — Kenosha Kid

Indeed, but 'true' is the ultimate post hoc justification (at least, that's the case I'm arguing). To say something is 'true' is simply to say that it's about as well justified as we can get.

How can an expression convey a weather condition? — Isaac

By being about it? I honestly have no clue what you're trying to ask here. — InPitzotl

It coveys a belief about a weather condition, not the actual weather condition (which is composed of atmospheric molecules).

Most of the time, it's used to get the listener to believe it's raining (by which I mean have a tendency to act as if it's raining - put a coat on, carry an umbrella, write a poem about it...). — Isaac

Most of the time, it's used to inform the listener that it's raining — InPitzotl

You seem to have just repeated what I said. Does a listener, sucessfully informed that it's raining, not now believe that it's raining?

The T is a relationship between the meaning of the claim and the state of affairs. The claim's meaning implies some truth conditions. The claim is true if the described state of affairs meet the truth conditions. A claim can be true even if nobody has any justifications for it.

Justifications are what you use to figure out what things are true. — InPitzotl

I'm quite clear now on what it is you believe to be the case, repeating it isn't necessary. What I'm pursuing is why you believe it to be the case.

we can (aka "can ever") ascertain truth using justification. — InPitzotl

Great. How?

Option (2) isn't about anything. It's part of a whole expression-act which is about the language game of quizzes. — Isaac

Isn't that a contradiction? — InPitzotl

No. A football is part of a game, it's not itself a game.

We don’t need to use the term “true”. We can say that:

John knows that it is raining iff:

1. John believes that it is raining,
2. John is justified in believing that it is raining, and
3. It is raining — Michael

You saying (or writing) 'it is raining', in this context, is the same as saying that you believe it's raining.

Imagine the situation in the real world.

John says "I think it's raining" - that gives rise to (1).

John says "I can hear the patter on the roof, plus the forecast said it was going to rain" - that gives rise to (2).

You look out of the window, perhaps go outside and look at the sky, get wet etc. - that gives rise to (3).

But (3) is still just a justified belief. It's your justified belief, based on the justifications of your experience.

A third party might join the scene and say "yes, but you're on powerful hallucinogenic drugs, actually it's not raining at all. Their belief about (3) contradicts yours.

A fourth party might point out that the third party just wants to make you look bad, and it is, in fact raining...

(3) is only, and forever will only be, someone's belief.

3. is to to be understood as the propositional content of John’s belief, i.e what his belief is about. — Michael

But John's belief is about {the weather}. (3) is not {the weather}, it too is about {the weather}. (3) is a statement written by you about the weather. It can't be 'what his belief is about' because 'what his belief is about' is the actual weather and a proposition is not the weather.

just letting you know you are not insane) — I like sushi

I appreciate that, although, one of the useful quirks of writing here is to be so vehemently disagreed with. It's not something I get a lot of in real life. As I'm sure many here who've taught will know, one needs grounding every now and again to remind one that one's position as the 'arbiter of truth' in the lecture theatre does not extend into the rest of life!

What JTB is is a formal set of rules set up in abstraction and then extended to ‘reality’. Such ‘knowledge’ is S-Knowledge only and cannot be confirmed as U-Knowledge. — I like sushi

That's good way of putting it. It certainly covers the manner in which @Michael presented it in his last post. Statement 1 is referable to the real world - John says something like "I believe it's raining". Statement 2 likewise - John gives his reasons. Then Statement 3, he's asking us to treat differently - removed from the real world (where we can only ever form beliefs about whether it's raining), but somehow held in this abstract world where we have direct access to the states of affairs which cause our perceptions, access outside our Markov blanket.

It's not that I think we can't talk about hidden states - I'm doing so now and I'm understood (at least others I talk to about hidden states understand me). It's just that we need to be consistent about our models if we're to be understood. @Michael's statement above, to me, reads like...

"Harry Potter got up, went over to the window to see if it's raining, he couldn't quite tell so he asked JK Rowling who was writing this story whether she intended it to be raining or not"

Indeed, but 'true' is the ultimate post hoc justification (at least, that's the case I'm arguing). — Isaac

Oh definitely, but the usual demonstrations are about someone who has a justified belief that then proves true but for the wrong reason. I find it's this notion of untrue justification that's problematic.

the usual demonstrations are about someone who has a justified belief that then proves true but for the wrong reason — Kenosha Kid

Wouldn't it proving to be true but for the wrong reason just be better justifications?

For me, if someone has a belief which is justified by reasons which turn out to be insufficient (the broken clock for example), then it's an issue of the quality of justification. Gettier skirts the issue by saying that seeing the clock say 12 is a sufficient justification for believing it's 12 o'clock, then saying "what if the clock was broken at 12". Well if it's possible for a clock to be broken at 12, then simply seeing the clock at 12 is not sufficient justification (in some circumstances - say if nuclear war depended on it), but it's perfectly sufficient in others.

So, in my model, all we have is the perception of the clock at twelve, and the later perception of the clock's broken mechanism (and presumably some other clock, not at 12). A community of people with beliefs about what the time is, talking to each other, behaving differently according to those beliefs.

I have no other components to the model. There's no {what the time actually is} component. Not because I don't believe in an external reality where the time actually is something, but because I don't see a role for such a hidden state in any of our discourse, and "I know..." is a form of discourse.

Essentially, I don't see a justified belief which turns out to be true, but not for the reasons given, any more problematic that a belief which turns out to be untrue because of faulty justifications. all that matters is the best quality of justification we can muster (and at what threshold we get to call it 'knowledge') - the relevant part is the justifications, it's those we can improve. That they can sometime lead us astray despite being 'good enough' is just an occupational hazard of being inside a Markov blanket, I don't think we need to alter our language to accommodate it.

It can't be 'what his belief is about' because 'what his belief is about' is the actual weather and a proposition is not the weather. — Isaac

The third condition is saying that the actual weather has to be as the person believes it to be. If you accept that there is such a thing as the actual weather, distinct from what anyone believes it to be, then you can understand the third condition.

John knows what the weather is like iff it actually is as he justifiably believes it to be.

Nothing about the above has anything to do with what I or the language community believes about the weather. And the above isn’t the same as the below, which is false:

John knows what the weather is like iff he is justified in believing what he does about the weather.

The third condition is saying that the actual weather has to be as the person believes it to be. — Michael

Yes, I get that. But since people don't have direct access to 'the actual weather' yet still use the expression "I know what the weather is like", including it as a necessary condition for the proper use of that expression is flat out wrong.

the above isn’t the same as the below, which is false:

John knows what the weather is like iff he is justified in believing what he does about the weather. — Michael

It's not false. It's an accurate description of the conditions under which people generally use the expression "John knows what the weather is like", which is (so far) the only criteria that's been offered for judging what "John knows what the weather is like" means.

So you finally understand that the third condition isn’t “I believe that X is true” or “the language community believes that X is true”? You finally understand that the third condition “X is true” is just saying that the actual facts obtain?

“X is true” is just saying that the actual facts obtain? — Michael

I never didn't get what the expression was trying to say. The expression "x must be an odd multiple of two" is a perfectly understandable expression too, it's just a criteria that's impossible to meet.

Saying "X is true", however, is an action, it's a thing people do, and what it means is determined by the conditions in which they perform that action - namely things like when they believe it with strong justification, or they believe it and their epistemic peers do too, or they just really, really want you to believe it.

So you don’t understand the difference between these?

1. John knows what the weather is like iff it actually is as he justifiably believes it to be

2. John knows what the weather is like iff he is justified in believing what he does about the weather and I agree with him

No, they sound quite different to me, not sure how you got there from what I said.

Try this.

Jim: the weather in Barbados is sunny, it's always sunny in Barbados, I've been there.
John: I'm actually in Barbados right now and I can tell you, it's not sunny, it's raining
Jim: I don't believe you
Jack (to Jim); John knows what the weather's like, he's actually there!

Is anything about that conversation odd?

Does John or Jack have infallible direct access to the truth about the weather in Barbados? (ie can't be wrong)

I presume the answers are 'No' and 'No'. So the expression "John knows..." is being used on the grounds that John's evidence, his justification for his belief, is very good (he's actually there, looking at the sky, getting wet...). It's not being used by comparing John's belief to the actual weather - no-one has direct access to that, they only have access to their various beliefs about the weather. It's their beliefs about the weather they're using to decide whether to use the term "John knows..." or reach instead for something like "John believes..." or "John thinks..."

You could do a @Banno and say that John does have direct access to the actual weather, that looking at it is as good as direct access to it. That's fine, it's a model I've some sympathy with, but then we'd have to clarify why Jim's access isn't direct. What is it about John's access that's categorically better than Jim's? Once we have that criteria, we have a definition of 'direct', but it's still essentially the same as I've been arguing - namely that at some level of justification we can say "John knows...", the only difference being that we also label this level of justification 'direct' to distinguish it from other levels which we call 'indirect'.

No, they sound quite different to me, not sure how you got there from what I said. — Isaac

The fact that you’ve repeatedly said that you interpret “it is raining” as “I believe that it is raining.”

Applied to “John knows what the weather is like iff it actually is as he justifiably believes it to be”, and using rain as an example, we break it down to:

John knows that it is raining iff:

1. John believes that it is raining,
2. John is justified in believing that it is raining, and
3. It is raining

Applied to “John knows what the weather is like iff he is justified in believing what he does about the weather and I agree with him”, and using rain as an example, we break it down to:

John knows that it is raining iff:

1. John believes that it is raining,
2. John is justified in believing that it is raining, and
3. I believe that it is raining

So do you now understand the difference between “it is raining” and “I believe that it is raining”?

You're still ignoring context and trying to pin me down to one single meaning for expressions which clearly have different meanings in different contexts.

"It's raining", as a response to "will I need a coat" might well mean nothing more than "I believe it's raining".

"It's raining" on the end of

"John knows that it is raining iff:

1. John believes that it is raining,
2. John is justified in believing that it is raining, and
3. It is raining"

...might mean something more akin to "I believe it's raining, and I've good strong justifications for believing so"

Same expression, two contexts, two meanings.

So when you say "and I agree with him", this might be a casual concurrence with what we think is a good guess, or it might be a strong conviction that he has, in fact, reached the required threshold of justification.

Same expression, two contexts, two meanings.

You're still ignoring context and trying to pin me down to one single meaning for expressions which clearly have different meanings in different contexts. — Isaac

I’m trying to explain to you that your interpretation of “it is raining” as “I believe that it is raining” is misplaced in the context of the JTB definition of knowledge.

When the JTB definition of knowledge states that John knows that it is raining iff 1) he believes that it is raining and 2) he is justified in believing that it is raining and 3) it is raining, it is simply stating in specific terms the more general definition that John knows what the weather is like iff it actually is as he justifiably believes it to be.

Or to be even more general, S knows a fact iff the fact is as S justifiably believes it to be.

Nothing about the JTB definition of knowledge has anything to do with what I or the language community believes.

When the JTB definition of knowledge states that John knows that it is raining iff 1) he believes that it is raining and 2) he is justified in believing that it is raining and 3) it is raining, it is simply stating in specific terms the more general definition that John knows what the weather is like iff it actually is as he justifiably believes it to be. — Michael

Yep. And as such the JTB definition of knowledge is wrong, because that's not how anyone ever actually uses the word 'knowledge' in any actual context because in all actual contexts people replace "actually is" with their own strong belief that it actually is.

Yep. And as such the JTB definition of knowledge is wrong, because that's not how anyone ever actually uses the word 'knowledge' in any actual context because in all actual contexts people replace "actually is" with their own strong belief that it actually is. — Isaac

I clarified the mistake you're making here:

There's a difference between saying "a bachelor is an unmarried man because the language community uses the term 'bachelor' to refer to people they believe to be unmarried men" and saying "John is a bachelor because the language community believes that John is an unmarried man."

The former is true, the latter is not. The language community can be wrong about John.

And in the same vein, there's a difference between saying "things that are known are true and justified because the language community uses the term 'known' to refer to things they believe to be true and justified" and saying "X is known because the language community believes that X is true and justified."

The former is true, the latter is not. The language community can be wrong about X. — Michael

We use the term "bachelor" to refer to people we believe to be unmarried men, but that doesn't mean that "bachelor" means "someone I believe to be an unmarried man." The correct definition of "bachelor" is "unmarried man."

Sometimes I refer to people who aren't bachelors as being bachelors because I believe them to be, and I'm wrong.

We use the term "knowledge" to refer to beliefs that we believe to be true and justified, but that doesn't mean that "knowledge" means "a belief I believe to be true and justified." The correct definition of "knowledge" is "belief that is true and justified."

Sometimes I refer to beliefs which aren't knowledge as being knowledge because I believe them to be, and I'm wrong.

When the JTB definition of knowledge states that John knows that it is raining iff 1) he believes that it is raining and 2) he is justified in believing that it is raining and 3) it is raining, it is simply stating in specific terms the more general definition that John knows what the weather is like iff it actually is as he justifiably believes it to be. — Michael

I'll try again here.

Number 3 'it is raining' is a Fact by what judgement? Abstract judgement. See it yet? 'raining' is NOT a universally explicit term. JTB is an abstract notion used in 'reality' therefore it is reaching beyond its bounds. It is however useful YET has limitations because the limits of reality are not known to any of us.

I presupposed 'raining' concept the is absolutely defined without doubt is not applicable to reality if we can talk about different 'raining'.

Example: what rain? show me this 'rain'. Unlike what do you mean 'number one'? Which kind of number one are you talking about? The former clearly being a nonsense question are there are not different kinds of 'number one' anymore than there is a different kind of 'of' or 'or'.

This is basic stuff. If there are two apples we don't in reality have two identical apples (that is impossible). We cannot - for the same reason - have 'rain' as some universally applicable term when using abstract logic. It just doesn't hold up unless we are merely using the term as an arbitrary marker rather than S, X, P or whatever else we feel suits.

3) Would require godly intervention to know. We don't know what 'rain' is it is just a term used to refer to a phenomenon that can appear in multiple and constantly different ways - like water.

For a further example if there is water in a space station that is effected by gravity and falling is it 'rain'? If the space station is big enough how big is 'big enough' for there to be 'rain'? Where exactly do we draw the line between there being 'rain' and there not being 'rain'?

We cannot agree on these questions so we cannot then claim to apply the judgement that 'it is a fact that it is raining' other than in a colloquial sense.

The kind of knowledge you are claiming here to be JTB is not JTB because it is impossible to apply a+b=c to reality when we have no full underpinning to state clearly and absolutely what a, b or c are. We do often assume that there is an underlying law/limit to the universe and this bears fruit. That is evidence for the theory of applying abstraction to reality NOT a proof (as stated in the link from Stanford I posted previously).

Back to the apples. An apple and another apple certainly make two apples. But we have no strict line between when an apple is an apple and when an apple isn't an apple. In most day-to-day situations we don't need to ask for such definitions but when we are discussing more nuanced problems the personal and implicit subjective view of said items does not gel so readily in a common language/definition.

JTB is necessarily a limited definition of knowledge because it tries to over apply the abstract into the real without any justification other than piled up evidence that works in 'some' situations enough to warrant a belief in its universal application ... it is kind of ironic really don't you think?

Number 3 'it is raining' is a Fact by what judgement? Abstract judgement. — I like sushi

I have addressed this repeatedly. There are facts, independent of what anyone believes or judges. If these facts obtain then our beliefs are true. These facts must obtain for us to have knowledge, otherwise our beliefs are false.

Here's another example:

John will die if:

1) John drinks the potion, and
2) the potion is toxic

Do we interpret this claim as the below?

John will die if:

1) I believe that John drinks the potion, and
2) I believe that the potion is toxic

Of course not. That would be ridiculous. My beliefs will not kill John. The actual facts will kill John. The exact same principle applies to:

John knows that it is raining iff:

1) John believes that it is raining,
2) John is justified in believing that it is raining, and
3) it is raining

Wouldn't it proving to be true but for the wrong reason just be better justifications? — Isaac

Definitely. But in the meantime...

Let's say you know that the answer to some question is X not Y, i.e. you have whatever standard of truth is necessary, but I don't. I have to figure out the answer from clues, and come to the belief that the answer is X but I could be wrong, that is I might have made a mistake as I have only clues (a WYSIATI error or some such). I tell you what I believe the answer is and why. You agree that my reasoning is sound and that I hold a JTB.

But who can judge what is or isn't a Fact? I have asked the same thing repeatedly in many different ways too.

Here's another example:

John will die if:

1) John drinks the potion, and
2) the potion is toxic

Do we interpret this claim as the below?

John will die if:

1) I believe that John drinks the potion, and
2) I believe that the potion is toxic

Of course not. That would be ridiculous. My beliefs will not kill John. The actual facts will kill John. The exact same principle applies to:

John knows that it is raining iff:

1) John believes that it is raining,
2) John is justified in believing that it is raining, and
3) it is raining — Michael

These are abstractions painted as 'reality'. There is no universal 'poison' the term 'toxic' will vary from situation to situation (for the the same substance due to quantity and other non-explicit factors), John is who (?), and what the hell is 'rain' anyway and why do we believe/know that there is such a phenomenon as 'rain'.

Framing 'real' objects as 'abstract universals' is certainly useful. Where is the line though?

Surely you understand why I am disputing JTB as a good definition/theory of 'knowledge'. There are different kinds of knowledge under different circumstances prescribed by limits and rules (or lack there of).

I do not hold to the JTB and many others dispute it to as nothing other than a rule of thumb not to be taken too seriously.

I clarified the mistake you're making here:... — Michael

No. It's the same issue.

'Bachelor' is a term given to people who the user believes are unmarried and the who the user believes is a man. That is how 'bachelor's used. It is not reserved for use only when we have managed to obtain some sort of objective fact about a person's sex or marital status.

You're labelling it as a wrong use of the word, but I'm calling it a correct use of the word, just a wrong belief. It's correct to use the word 'bachelor' of someone you believe to be unmarried and believe to be a man, it's how everyone uses the word and it would be perverse to suggest it wasn't correct (ie everyone is wrong).

You might later come to believe that he is married, or a woman (or both), so now, believing this, it would no longer be correct to use the word 'bachelor'.

With the word 'knowledge' it's the same. It must be correct to use the word of something which you have strong justification to believe (particularly if that justification is the agreement of your epistemic peers) because that is how the language community uses the word, it would be perverse to say they're all wrong.

You may later come to believe that you justifications were not strong enough to warrant the term, but by the time you believe that, you no longer use the word.

So "I thought I knew the way to the pub but it turns out i didn't" means that you thought your justifications were strong enough, but it turns out they weren't, stronger ones (the evidence of your eyes) now lead to a belief that the pub is not where you thought it was.

Let's say you know that the answer to some question is X not Y, i.e. you have whatever standard of truth is necessary, but I don't. I have to figure out the answer from clues, and come to the belief that the answer is X but I could be wrong, that is I might have made a mistake as I have only clues (a WYSIATI error or some such). I tell you what I believe the answer is and why. You agree that my reasoning is sound and that I hold a JTB. — Kenosha Kid

So I can agree to that broadly (because you have 'truth' as being merely a sufficient threshold, a 'standard'), but where I still take issue is that the 'standard' can be no more than a set of justifications (usually something like - most of my epistemic peers agree, every time I act as if it's the case I get the expected results - these are the 'gold standard' justifications), so there's no difference in kind between your belief and mine, it's just that my justifications are better. One of the biggest complaints I have about JTB is that it makes a big deal of some unspecified threshold (truth vs other justifications), yet ignores what I think is the most important aspect of judging beliefs which is the relative quality of your justifications (the 'Gold Standard' I mentioned above, for example).