Are all models wrong?

[uwot]

My only reservation would be that I don't understand what a 'fully determined' model - like a complete description - could possibly comprise - so the idea of underdetermination seems problematic. I'd like to pursue this, if you can help.

I'll do my best. The way I understand it is that the problem is pretty straight forward. In my view, a 'fully determined model' of the universe we find ourselves in, or any aspect of it, can only be judged fully determined when everything that could happen has actually happened, and no exceptions to the model have been found. So it will take until the end of the universe, to decide whether any given model is consistent with every demonstrable fact. But, between now and the end of the universe, I'm sure it will be possible to come up with more than one model that is not refuted by the facts. So even with every fact that ever happens in this universe, there are conceivable rival models, with no way to distinguish between them. That to me is underdetermination. In the meantime, it remains entirely possible that some of our current models are right.

[Peter Holmes]

My only reservation would be that I don't understand what a 'fully determined' model - like a complete description - could possibly comprise - so the idea of underdetermination seems problematic. I'd like to pursue this, if you can help.

I'll do my best. The way I understand it is that the problem is pretty straight forward. In my view, a 'fully determined model' of the universe we find ourselves in, or any aspect of it, can only be judged fully determined when everything that could happen has actually happened, and no exceptions to the model have been found. So it will take until the end of the universe, to decide whether any given model is consistent with every demonstrable fact. But, between now and the end of the universe, I'm sure it will be possible to come up with more than one model that is not refuted by the facts. So even with every fact that ever happens in this universe, there are conceivable rival models, with no way to distinguish between them. That to me is underdetermination. In the meantime, it remains entirely possible that some of our current models are right.

Thanks again. That makes a lot of sense, and you explain it very clearly. I'd like to pick up some points - for example about what constitutes a 'fact' and a 'model' here, and how it relates to what's been called physical determinism - so I hope to respond fully when I have time. Cheers for now.

[Sculptor]

George Box claimed that ‘All models are wrong but some are useful’. But if that claim is true, then at least one model is not wrong – in which case, the claim is false.

You have started with a falsehood
If you are trying to suggest that "All models are wrong but some are useful" is a model?
You are just playing the Russell game is a set of all things a set of itself.
You'd do better making a clear distinction between "a model" and "a fact".

If by 'a fact' you mean 'a feature of reality', then it has no truth value, and it clearly isn't a model. But if by 'a fact', you mean 'a description of a feature of reality', then it can exist only within a model - in this case a linguistic model. And if a factual assertion - here a linguistic expression - is true, then its generative model can't be 'wrong'.

Since no model can fully represent that which it models, it cannot be wholly true.

Not so. We've discussed this fully already. 'Incomplete' doesn't mean 'wrong'. And 'full representation' is incoherent anyway.

Who the fuck is "we"? Is that the Royal "we"?
BTW you have just contradicted yourself.
Think about it!

[Peter Holmes]

You have started with a falsehood
If you are trying to suggest that "All models are wrong but some are useful" is a model?
You are just playing the Russell game is a set of all things a set of itself.
You'd do better making a clear distinction between "a model" and "a fact".

If by 'a fact' you mean 'a feature of reality', then it has no truth value, and it clearly isn't a model. But if by 'a fact', you mean 'a description of a feature of reality', then it can exist only within a model - in this case a linguistic model. And if a factual assertion - here a linguistic expression - is true, then its generative model can't be 'wrong'.

Since no model can fully represent that which it models, it cannot be wholly true.

Not so. We've discussed this fully already. 'Incomplete' doesn't mean 'wrong'. And 'full representation' is incoherent anyway.

Who the fuck is "we"? Is that the Royal "we"?
BTW you have just contradicted yourself.
Think about it!

Thunked. Can't see it. Happy to be corrected.

[uwot]

I'd like to pick up some points - for example about what constitutes a 'fact' and a 'model' here, and how it relates to what's been called physical determinism...

What I understand as a fact is any phenomenal datum. A simple example being if you drop something and it falls to the floor. 'Model' is a bit more complicated. There's the inductive model - 'heavier than air things always fall to the floor', the physical model - 'there is a force called gravity', the mathematical model - Einstein's field equations for example, and the philosophical model - 'it's because there's a substance called spacetime that is warped by matter'. No doubt there are other types of model and there are certainly alternatives to the examples above, but what distinguishes them from facts is the wiggle room. Any one of them might be true, and in that sense they may be factual, not wrong in other words, but they are all subject to investigation and possible revision, whereas if you drop something, it either falls or it doesn't.
As for physical determinism, it seems entirely plausible to me that the facts about the behaviour of physical objects are determined by the physical environment they find themselves in. But however we model that physical environment makes bugger all difference to the facts.

[Peter Holmes]

I'd like to pick up some points - for example about what constitutes a 'fact' and a 'model' here, and how it relates to what's been called physical determinism...

What I understand as a fact is any phenomenal datum. A simple example being if you drop something and it falls to the floor. 'Model' is a bit more complicated. There's the inductive model - 'heavier than air things always fall to the floor', the physical model - 'there is a force called gravity', the mathematical model - Einstein's field equations for example, and the philosophical model - 'it's because there's a substance called spacetime that is warped by matter'. No doubt there are other types of model and there are certainly alternatives to the examples above, but what distinguishes them from facts is the wiggle room. Any one of them might be true, and in that sense they may be factual, not wrong in other words, but they are all subject to investigation and possible revision, whereas if you drop something, it either falls or it doesn't.
As for physical determinism, it seems entirely plausible to me that the facts about the behaviour of physical objects are determined by the physical environment they find themselves in. But however we model that physical environment makes bugger all difference to the facts.

Thanks again. This is very useful - and I think you've got it right.

The way I'd put it is that we use the word 'fact' in two ways, to mean 'a feature of reality' (for example, things dropping) or 'a description of a feature of reality' (what you call a model). To forget the difference and conflate the two meanings of 'fact' is to mistake what we say about things for things themselves - a description for the described. It's the origin of the myth of propositions, the truth-condition in the JTB definition of knowledge - S knows that p iff p is true - and correspondence theories of truth.

I think your point about different descriptions, causal explanations, or 'models' being correct is spot on. A description is not the thing described, and things can be described in different ways for different purposes - which means there can be many (limitless?) true factual assertions about the same feature of reality which, as you say, is what it is or was, how ever it's named or described.

I'm still puzzled by the idea of underdetermination, because, like the idea that all models are wrong, it seems to at least flirt with the possibility of a full or complete or 'right' description - and I still think that's an incoherent idea. Even at the end of the universe, when all events have happened, there could still be no complete description of the universe, because there's always another way to describe things. More to do on this, I know.

Many thanks for your thoughts. Been a pleasure.

[uwot]

I'm still puzzled by the idea of underdetermination...

Essentially it's just another way of saying that JTB is a myth. Professor Bloggs might be justified in believing something that, as it happens, is true - it's a fact, to use one of the meanings - but unless he or she can rule out every possible alternative, they cannot know it is true.

...it seems to at least flirt with the possibility of a full or complete or 'right' description - and I still think that's an incoherent idea.

Well, in fairness it does. There are, for instance, roughly a dozen competing mathematical and/or philosophical models that account for gravity which have yet to be proven 'wrong'. They all accept the inductive and physical models that, with a few caveats, have been the inspiration for mathematical and philosophical models since at least Aristotle. Underdetermination is simply the fact that until shown otherwise, any of the competing theories could actually be right, but there's no way of knowing which, if any, is the 'right' one.

Even at the end of the universe, when all events have happened, there could still be no complete description of the universe, because there's always another way to describe things.

Absolutely; which is why theories will always be underdetermined.

[Peter Holmes]

I'm still puzzled by the idea of underdetermination...

Essentially it's just another way of saying that JTB is a myth. Professor Bloggs might be justified in believing something that, as it happens, is true - it's a fact, to use one of the meanings - but unless he or she can rule out every possible alternative, they cannot know it is true.

Forgive me for picking away at this. I've been mulling over your explanation of 'underdetermination', and something still puzzles me. Completely understood if you want to leave it with what you've said.

I go back to the tripartite model I'm proposing: features of reality; what we believe and know about them; and what we say about them - the point being that only what we say about things can be true or false. Only our models can be right or wrong.

Now, the purpose of this three-way distinction is to clarify the absolute separation and difference between the three elements. So, for example, believing a feature of reality is or isn't the case has absolutely nothing to do with truth or falsehood - because truth-value is an attribute of language - what we say - not of belief or knowledge.

Features of reality have no truth-value - they just are or were, neither true nor false - and neither does what we believe or know, or claim to know. To put it another way, truth and falsehood are not epistemological matters. Believing (accepting) or knowing that a feature of reality is the case - that I'm using a laptop atm - has nothing to do with language.

This may seem obvious, but go back to Professor Bloggs. What she supposedly can or can't know is that a feature of reality is the case. And that's an epistemological matter, nothing to do with truth and falsehood - the right or wrongness of a model. The epistemological so-called problem of induction is about available evidence - which is why, I think, you mentioned the end of the universe - the end of events (evidence) that can affirm or disconfirm a model - a factual, causal description or explanation of how the universe works.

...it seems to at least flirt with the possibility of a full or complete or 'right' description - and I still think that's an incoherent idea.

Well, in fairness it does. There are, for instance, roughly a dozen competing mathematical and/or philosophical models that account for gravity which have yet to be proven 'wrong'. They all accept the inductive and physical models that, with a few caveats, have been the inspiration for mathematical and philosophical models since at least Aristotle. Underdetermination is simply the fact that until shown otherwise, any of the competing theories could actually be right, but there's no way of knowing which, if any, is the 'right' one.

The question that's nagging me is this: is underdetermination - as it seems from the way you describe it here - purely an epistemological matter?

Even at the end of the universe, when all events have happened, there could still be no complete description of the universe, because there's always another way to describe things.

Absolutely; which is why theories will always be underdetermined.

There's the rub. If the problem is epistemological (to do with induction), then it will be solved at the end of the universe. But why then would our model(s) still be underdetermined? (In other words, I think underdetermination isn't an epistemological problem - to do with what we know or can know - but rather the result of a confusion between the model and what is modelled - the map and the terrain.)

[uwot]

I go back to the tripartite model I'm proposing: features of reality; what we believe and know about them; and what we say about them - the point being that only what we say about things can be true or false.

Perhaps my problem is that I take epistemology too seriously, in that only a sentence that cannot be said without being self refuted counts as 'knowledge' in some absolute sense. In two and a half thousand years, only two such sentences have been discovered: Parmenides' 'There is not nothing' (Hang on, that's confusing. It's essentially what Parmenides actually said, but the thing that cannot be said without it being obviously wrong is 'There is nothing'.) and a slightly tweaked Descartes' 'There is thinking'. Everything else is theory laden.
I have yet to see an argument that persuades me that Kant was wrong to divide 'features of reality' into phenomena and noumena. Then, as you say, there are the models we create to account for the phenomena which, very broadly, are philosophical accounts of noumena, and the causal accounts that physicists invent.
Although, as this thread demonstrates, people don't always say what they mean, I'm not convinced the distinction between what we believe and what we say is significant. The thing is, however right or wrong our models may be, in order to express or rationalise them, and certainly if we wish to communicate them, we are restricted to some form of language; English or mathematics, for example. And if the word, symbol or equation doesn't exist to express our beliefs, we make them up.

Only our models can be right or wrong.

What can they be right or wrong about?

The epistemological so-called problem of induction is about available evidence - which is why, I think, you mentioned the end of the universe - the end of events (evidence) that can affirm or disconfirm a model - a factual, causal description or explanation of how the universe works.

The universe may have been such that by the end of it every swan was in fact white, so in that sense the problem of induction would no longer be a problem.

There's the rub. If the problem is epistemological (to do with induction), then it will be solved at the end of the universe. But why then would our model(s) still be underdetermined?

Because if there are only two sentient beings left, the chances are they will disagree why all swans were white.

[Peter Holmes]

I go back to the tripartite model I'm proposing: features of reality; what we believe and know about them; and what we say about them - the point being that only what we say about things can be true or false.

Perhaps my problem is that I take epistemology too seriously, in that only a sentence that cannot be said without being self refuted counts as 'knowledge' in some absolute sense.

As you know, I think there are two difficulties here. First, what counts as knowledge is what we call knowledge - for example, what we mean when we say we know something. And, as Wittgenstein kept pointing out, the facts are all out in the open. The idea that knowledge is really something different seems to me a metaphysical delusion; what and where is that thing?

And second, there's no reason to identify knowledge - knowing things - with sentences, self-refuting or otherwise. We can express what we know by means of sentences, but these are different and separate things - which is why something that's known can be expressed by means of different sentences. How many true factual assertions 'correspond' with a feature of reality that we 'know'?

In two and a half thousand years, only two such sentences have been discovered: Parmenides' 'There is not nothing' (Hang on, that's confusing. It's essentially what Parmenides actually said, but the thing that cannot be said without it being obviously wrong is 'There is nothing'.) and a slightly tweaked Descartes' 'There is thinking'. Everything else is theory laden.

I'm lost here. Are you saying that 'there is nothing' and 'there is thinking' are self-refuting claims, but that only those claims amount to knowledge in some absolute sense? (Sorry - I can't make sense of this. Why are these two sentences not theory-laden? Can you spell it out?)

I have yet to see an argument that persuades me that Kant was wrong to divide 'features of reality' into phenomena and noumena. Then, as you say, there are the models we create to account for the phenomena which, very broadly, are philosophical accounts of noumena, and the causal accounts that physicists invent.

What and where are noumena, and why are they different from phenomena? I don't think Kant ever demonstrated their existence as features of reality - he just argued them into existence. Do you have the killer fact that establishes their existence? Why is this not just old-school substance dualism re-packaged for the enlightenment?

Although, as this thread demonstrates, people don't always say what they mean, I'm not convinced the distinction between what we believe and what we say is significant. The thing is, however right or wrong our models may be, in order to express or rationalise them, and certainly if we wish to communicate them, we are restricted to some form of language; English or mathematics, for example. And if the word, symbol or equation doesn't exist to express our beliefs, we make them up.

I think this conflation is at the heart of what we're discussing. If you hear the doorbell and think or believe someone's there, nothing linguistic has occurred, just as it hasn't if a dog hears it and 'believes' it's its owner come home. We can express (verbalise) the belief, but that's a different and separate operation. And the dog can't verbalise it at all, though it may 'express' it by barking, and so on. Language was a primitive reaction and tool that our species developed way beyond what other species do, to the point where, because it's so potent and pervasive in our lives, we've forgotten what it is and how it works - for example how we use it to describe features of reality and our reactions to them.

Only our models can be right or wrong.

What can they be right or wrong about?

Features of reality, which just are or were, neither right nor wrong, true nor false.

The epistemological so-called problem of induction is about available evidence - which is why, I think, you mentioned the end of the universe - the end of events (evidence) that can affirm or disconfirm a model - a factual, causal description or explanation of how the universe works.

The universe may have been such that by the end of it every swan was in fact white, so in that sense the problem of induction would no longer be a problem.

Agreed. So why then would our model be underdetermined?

There's the rub. If the problem is epistemological (to do with induction), then it will be solved at the end of the universe. But why then would our model(s) still be underdetermined?

Because if there are only two sentient beings left, the chances are they will disagree why all swans were white.

But that they are all white would be fully determined. And there's no reason to think the why couldn't also be fully determined. One of the beings left could be shown to be wrong.

[uwot]

...what counts as knowledge is what we call knowledge...

Fair enough. I have to concede that what I count as knowledge is quite restricted

- for example, what we mean when we say we know something. And, as Wittgenstein kept pointing out, the facts are all out in the open.

Well yes, the world is the totality of facts, not of things, according to the opening lines of the Tractatus. I'm not an expert on Wittgenstein, but whatever else changed, I think that basic principle stayed with him. The way I interpret that is the way that coincidently (!) best fits my own argument. As I said above:

What I understand as a fact is any phenomenal datum. A simple example being if you drop something and it falls to the floor.

A 'thing' by contrast, is the 'model' we create to account for the 'facts'. A simple example of that being an apple. There are several facts associated with apples: colour, shape, smell, taste, texture and so on, from which we create a model of this 'thing' an apple.

The idea that knowledge is really something different seems to me a metaphysical delusion; what and where is that thing?

If you mean knowledge of the apple, I agree, and I suspect Wittgenstein would too.

And second, there's no reason to identify knowledge - knowing things - with sentences, self-refuting or otherwise. We can express what we know by means of sentences, but these are different and separate things -

Absolutely, if you are eating an apple, you add nothing to your knowledge by saying 'I am eating an apple.'

I'm lost here.

Yeah, me too. Sorry, I made a right pig's ear out of that. Start again. The sentence 'Something exists' is true. We know it has to be true, because its negation 'Nothing exists' is self-refuting. Similarly, the thought 'There is thinking' must be true, because no one can think 'There is no thinking' without thinking it.

What and where are noumena, and why are they different from phenomena? I don't think Kant ever demonstrated their existence as features of reality - he just argued them into existence. Do you have the killer fact that establishes their existence? Why is this not just old-school substance dualism re-packaged for the enlightenment?

The noumena are 'things', the phenomena 'facts'. So for instance if you have the experience of tasting an apple, then it is a fact that you have that experience. It doesn't follow that you are tasting an actual apple, as in 'thing'. And no, I don't have a killer fact, I just personally believe that the most plausible, or at least aesthetically pleasing, theory for why I taste an apple is overwhelmingly because I am eating an apple.

Only our models can be right or wrong.

What can they be right or wrong about?

Features of reality, which just are or were, neither right nor wrong, true nor false.

I have absolutely no problem believing that noumenal things called apples are features of reality, and are responsible for the bulk of experiences of tasting apples.

The epistemological so-called problem of induction is about available evidence - which is why, I think, you mentioned the end of the universe - the end of events (evidence) that can affirm or disconfirm a model - a factual, causal description or explanation of how the universe works.

The universe may have been such that by the end of it every swan was in fact white, so in that sense the problem of induction would no longer be a problem.

Agreed. So why then would our model be underdetermined?

Well, it wouldn't be a model anymore; it would be a fact.

But that they are all white would be fully determined. And there's no reason to think the why couldn't also be fully determined. One of the beings left could be shown to be wrong.

How? There's no swans left to examine?
And for the record, it is a pleasure to talk about real philosophy on a philosophy forum, of all places. Cheers.

[Peter Holmes]

... for the record, it is a pleasure to talk about real philosophy on a philosophy forum, of all places. Cheers.

Couldn't agree more. I need more time to respond to what you say - all of which is interesting. I think I've been misconstruing your terminology, so it's useful to know how you use words like 'thing'. I'll get back asap. Cheers.

[AlexW]

A 'thing' by contrast, is the 'model' we create to account for the 'facts'. A simple example of that being an apple. There are several facts associated with apples: colour, shape, smell, taste, texture and so on, from which we create a model of this 'thing' an apple.

I am interested in this discussion and I thus hope you don't mind me expressing my opinion:

I fully agree with you in that: A "'thing' is the 'model' we create to account for the 'facts'"
It follows that "things" are conceptual entities, not facts and, to me, not "real".

I also agree with:

the world is the totality of facts, not of things

To be more general we could replace the word "world" with "reality": Reality is the totality of facts.

Coming back to your example of eating an apple: A fact would be the direct experience of eating the apple (e.g.: the fact that we label "taste of apple").
There is no question, the experience is a fact - what is questionable is the interpretation: there is a "me", doing the "eating" of a thing called "apple". This interpretation is not a fact, and thus it is nothing more than another "thing" (a model) - it is not reality, but rather an idea, a conceptual representation of an extract (a recognised pattern) of reality.

The question arises: How do we separate one "fact" from other "facts"?

It is easy to separate one "thing" from other things, they have qualities, attributes, well defined borders... but what about facts?
Can we separate one fact (e.g. the direct experience of "eating an apple") from another fact without reverting to interpretation and conceptualisation (and thus, essentially, to using "things" to dissect reality)?
What if one fact can not actually be separated from (apparently) different facts without using a "thing" (a conceptual interpretation) as the separating divider?

Wouldn't that point to the "fact" that facts are not actually separate at all?
That reality is only (apparently) divided into facts because of our belief in "things" existing in their own right?

[uwot]

I fully agree with you in that: A "'thing' is the 'model' we create to account for the 'facts'"
It follows that "things" are conceptual entities, not facts and, to me, not "real".

Blimey. Maybe I'd better start with some sort of lexicon.

Fact empirical - any experience.
Fact contingent - the cause of an experience.
Real ontological - independent of experience.
Real conceptual - a product of experience.
Thing conceptual - a model we create from empirical facts.
Thing actual - the putative cause of empirical facts.

The above is far from exhaustive, nor do I claim it is authoritative, it's just how I kinda navigate this stuff when it comes up. So plonking in those definitions as I see appropriate, your sentence reads something like this:

A 'conceptual thing' is the model we create to account for the 'empirical facts'.
It follows that 'conceptual things' are conceptual entities, not contingent facts and, to you, not ontologically real.

I suspect that even at this point, there is stuff that needs thrashing out, but assuming I haven't got completely the wrong end of the stick:

There is no question, the experience is a fact - what is questionable is the interpretation: there is a "me", doing the "eating" of a thing called "apple".

Right, so an experience as empirical fact is the only thing we are certain of. The 'me', 'eating' and 'apple', if they are facts at all, are contingent facts - they just happen to be the case.

This interpretation is not a fact...

Well that's the thing, it could well be a contingent fact.

...and thus it is nothing more than another "thing" (a model)...

I think you are making an invalid leap from 'Not an empirical fact' to 'Not a contingent fact'.

...it is not reality, but rather an idea, a conceptual representation of an extract (a recognised pattern) of reality.

Well yes, it is conceivable that the 'me' eating an apple is just "a recognised pattern"; a more or less coherent string of sensations that includes previous episodes of apple eating sensations, that conform to some vague and private recognised pattern of conceptual apple model thing.

The question arises: How do we separate one "fact" from other "facts"?

It is easy to separate one "thing" from other things, they have qualities, attributes, well defined borders...

It is easy if you are sticking to facts being direct experience/empirical facts. The experience of eating an apple is different to the experience of sitting on a drawing pin, for example. On the other hand, if you take it that 'me' is a connected series of sensations, if you separate them, you lose any sense of 'me' and have to come up with an alternative conceptual thing to account for your own private 'me', or accept that it isn't real ontologically.

...but what about facts?
Can we separate one fact (e.g. the direct experience of "eating an apple") from another fact without reverting to interpretation and conceptualisation (and thus, essentially, to using "things" to dissect reality)?

Personally, I think the obvious place to start is the common sense/naïve view that actual things are real and cause the empirical facts we associate with them. I think we should keep shaking that theory until it is utterly broken; some people think it already is, but I am not persuaded by their evidence or arguments and still think the idea that the world/reality is made of some ontologically real stuff, that has broadly mechanical properties, has legs.

What if one fact can not actually be separated from (apparently) different facts without using a "thing" (a conceptual interpretation) as the separating divider? Wouldn't that point to the "fact" that facts are not actually separate at all?
That reality is only (apparently) divided into facts because of our belief in "things" existing in their own right?

Well yeah, it could be that empirical facts are all that are ontologically real and our projections are only conceptual things, and not actual things.

[AlexW]

Maybe I'd better start with some sort of lexicon.

Fact empirical - any experience.
Fact contingent - the cause of an experience.
Real ontological - independent of experience.
Real conceptual - a product of experience.
Thing conceptual - a model we create from empirical facts.
Thing actual - the putative cause of empirical facts.

Sounds very complicated to me...
As I see it, things are not that difficult, its all pretty simple:

1) Direct experience (what you call "Fact empirical") is really all that you can and will ever know directly.

2) A "Fact contingent - the cause of an experience" has to be again a direct experience, right? That it is a cause of another experience is a deduction, but not a direct experience.

3) Real ontological - independent of experience: What would that be? I don't think you can actually know anything directly that is "independent of experience"? Can you? Do you have an example of what you are referring to?

4 & 5) Real conceptual - a product of experience & Thing conceptual - a model we create from empirical facts: Aren't these two the same? Both are concepts, right? Both are interpretations of "empirical facts". To me, a concept is a concept, one is not "more real" than another - it would be like saying that red is more real than green... You might think that the concept "apple" is more real than "unicorn", you can see the apple, the unicorn you don't, but when actually analysing direct experience all the way to its source one will find that "apple" is just as much "thought up" as is "unicorn".

6) Thing actual - the putative cause of empirical facts.
I think number 6 emerges from this perspective:

Personally, I think the obvious place to start is the common sense/naïve view that actual things are real and cause the empirical facts

...and yes, I agree, the personal view (common sense) states that this has to be the case. There is an apple and the apple's separate existence makes it possible for me to experience it as an empirical fact.
But if you actually investigate direct experience you will find that this is actually not the case...
The question is thus: Do we subscribe to common sense - the naïve view - or do we actually want to figure out the truth of what experience actually is, of what "things" are and of what I am?

It is of course not an easy task to throw all old/conditioned knowledge over board - ignore common sense (which we can do at least for the time of our investigation) - and look in an honest way at what is actually happening here/now...
I thought that is what this is all about...
"Epistemology is the study of the nature of knowledge, justification, and the rationality of belief" - it would thus only make sense to question all and everything, suspend all established belief and lexical definitions and simply look at what is here/now. Where else should one find truth? In a lexicon? In acquired concepts and beliefs or in directly experienced reality/life?

To me, there are only two "levels" of reality - level one is real, direct experience, level two is the conceptual wrapper we have learned to wrap around it.
Its equivalent to the territory that is described and navigated using a map - the map shows borders, markers, elevations, distance and whatever else we might find useful, but the territory itself knows nothing of all of that. Problems arise once we believe the map is actually real and we forget about the terrain - we still live in it, but being lost in conceptual thought we only see the things that we have drawn onto maps (and start to believe they are actually real)... meanwhile life happens here/now, but we are actually to distracted to really see it...