Is "Truth" really a valid requirement for Knowledge?

[Fcacciola]

Hi people,

I've been reading a lot on the so-called JTB account of knowledge (that knowledge is (i) a belief, (ii) which happens to be true and (iii) which we're justified in believing so).
However, I have a really hard time agreeing with the requirement of truth, which seems to be taken for granted over and over.

For instance, I see the Gettier problem as discovering that the connection between knowledge and actual truth is just accidental, and all solutions seems to me that are basically attempting to turn that accidental connection into an intentional one. Causality, Defeasability, Truth-tracking, all these seem to me to reconnect the accidental "P is true" with the "P is Justified".

But, isn't "P is true" the real root of the Gettier (and related) problem?
Or, going even further.. is "P is true" really a proper requirement for knowledge?

The skeptic observation is that there exist scenarios in which our perceptions, or even the very construction of propositions, can be manipulated such that "P is true" is actually false but we can't tell. Then, the conclusion is that in those scenarios, we wouldn't know anything (and skepticism, beyond the observation, is the position that we cannot tell whether we are actually in such scenarios so in fact we effectively cannot know anything.. but I don't take that position and I agree with the view that this possibility is actually so unlikely that is of no practical concern).

But, isn't the skeptic conclusion so only because knowledge is required to be truth? Wouldn't it be different if "T" in JTB were replaced by something else?

Consider the claim that:

(A) In every likely scenario, in every likely reality (unlike the skeptic twisted ones), there is absolutely no mechanism that can directly transfer the objetive truth value of any proposition over an external system (not the ones we humans create, such as mathematics, language, algorithms and computer programs, social institutions, etc...) into a human mind.

If that claim holds, how can truth be a *requirement* for (external) knowledge? How does "P is true" gets effectively satisfied?

I see the Gettier problems as cases that makes it crystal clear that "P is true" is not a condition that is genuinely satisfied by the *process* by which the proposition is constructed in the mind. But what if, in fact, that condition is never, ever, genuinely satisfied no matter what? In other words, what if the connection to the truth is always accidental, not just in the Gettier cases? (and the Gettier cases were just extremes that made this fact visible).

While claim (A) above does not hold for human-constructed systems (there is a mechanism to transfer the truth about 2+2=4, etc... into a mind), and, knowledge about such systems have no problem with the condition of truth (that I'm aware of), I do feel that knowledge about external systems can never truly satisfy the truth condition, and so, for these systems (like the reality we're in), knowledge is better defined differently.

Granted, merely removing "P is true" does not work, because then, we can't tell apart mere believes from knowledge. Knowledge does require some sort of connection with the truth. Is just that in my opinion, that connection cannot be super-over-simplified as "P is true".

Perhaps, it might be stated that "we DO know, even to mean we don't just believe, except, not necessarily the truth"

What am I missing?

NOTE: William Jame's pragmatic epistemology comes really close to my views here, I think. But it seems it haven't caught up, for whatever reason.

[Eduk]

I think there are levels of truth?
Everyday meaning of truth is covered by critical thinking and the scientific method. It's true that I can tell my position on the world via GPS because of relativity. It's true that I believe I am conscious. ETC. I assume under this definition that the condition P is true is ok?
Then there is the TRUTH. Which we can't prove and where much bad philosophy hides. I think I'm conscious, but can I prove it (even to myself). How do I know I'm not just programmed to think I'm conscious, or that consciousness isn't an illusion? I mean personally I don't think you can be programmed to think you are conscious but that's by the by, I can't prove that. How do I know I'm not just plugged into the matrix? Here P is true becomes an axiom which can't be proved.
I personally take a pragmatic view. What difference does it make if you can PROVE if P is true is an axiom or not? For my local purposes there is no difference between believing I am conscious and TRULY being conscious. For example can I PROVE gravity (as I understand it) will stop tomorrow. No I can't PROVE it. Should I make steps to protect against the eventuality of gravity stopping tomorrow? No.
I do believe there is one good takeaway from our disassociation from TRUTH though. In science theories aren't TRUE, they are models which represent reality (they are not reality themselves). In this way those theories always have the potential for progression and improvement (however incremental) and are open to contradicting evidence. For example my pen drops due to gravity at speeds which are incredibly well predicted. But no one knows what gravity is. We constantly improve our understanding. However it's worth noting that improved understanding does not change the speed at which my pen will fall (the whole of the theory of gravity is unlikely to be thrown out wholesale).

[Cuthbert]

"In other words, what if the connection to the truth is always accidental, not just in the Gettier cases? (and the Gettier cases were just extremes that made this fact visible)."

In that case, we don't know anything. In fake barn country, when I see something that happens to be a real barn and identify it correctly as a barn I still don't know that it's a barn because it's just as likely to be a fake. So (according to Gettier) I have a true justified belief - but not knowledge. And if the whole world is like fake barn country then I don't know anything.

Some people are ok with the conclusion that we don't know anything. Others think that we do know some things, and therefore not all the world is like fake barn country.

Truth is definitely not sufficient for knowledge. But it does seem to be necessary. Nobody has offered a persuasive example of S knows that p where p is false.

[Sage4557]

I think "truth" would need to be irrefutably defined so as to not undermine it's existence. But truth takes on many forms. Than one needs to ask what is the purpose of "truth" If someone declares that 2+2=4, does one just need to accept that within the working framework of their goals and needs. Such as Religion. Each Religion professes to be the "truth" as it applies to their societal structure and function. Yet many Religions are vastly different in origin but claim a God. So which professed Religion is truth. They all are, as they relate to their origin of creation and structure of their society. There are many truths depending on the needs of a society. 2+2=4 based on a mathematical system that has been widely accepted through out the world. Starting with four people. 2 apples + 2 apples equal 4 apples this allows one to know that there is enough for the four people. Or simply there are a group of apples that when handed out is enough for each of us, excluding the mathematical structure.

[Atreyu]

I agree with the general proposition Truth should not be a requirement for Knowledge, since we truly know very little. However, once something is known to be false, we can no longer call it "knowledge". Now, it's simply a lie....

[Lark_Truth]

I have two definitions for you:
Knowledge: facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject.
Truth: that which is true or in accordance with fact or reality.
And yes, these both came off of google, but they do illustrate my point that: knowledge is what you learn, it is your experience and has a lot to do with your perception, while truth is already there and is unchanging, despite one's perception.

[Fcacciola]

I think there are levels of truth?
Everyday meaning of truth is covered by critical thinking and the scientific method. It's true that I can tell my position on the world via GPS because of relativity. It's true that I believe I am conscious. ETC. I assume under this definition that the condition P is true is ok?
Then there is the TRUTH. Which we can't prove and where much bad philosophy hides. I think I'm conscious, but can I prove it (even to myself). How do I know I'm not just programmed to think I'm conscious, or that consciousness isn't an illusion? I mean personally I don't think you can be programmed to think you are conscious but that's by the by, I can't prove that. How do I know I'm not just plugged into the matrix? Here P is true becomes an axiom which can't be proved.
I personally take a pragmatic view. What difference does it make if you can PROVE if P is true is an axiom or not? For my local purposes there is no difference between believing I am conscious and TRULY being conscious. For example can I PROVE gravity (as I understand it) will stop tomorrow. No I can't PROVE it. Should I make steps to protect against the eventuality of gravity stopping tomorrow? No.
I do believe there is one good takeaway from our disassociation from TRUTH though. In science theories aren't TRUE, they are models which represent reality (they are not reality themselves). In this way those theories always have the potential for progression and improvement (however incremental) and are open to contradicting evidence. For example my pen drops due to gravity at speeds which are incredibly well predicted. But no one knows what gravity is. We constantly improve our understanding. However it's worth noting that improved understanding does not change the speed at which my pen will fall (the whole of the theory of gravity is unlikely to be thrown out wholesale).

Indeed. It's a good idea to think of levels of truth. And I also agree with the pragmatic approach to knowledge.
I fact, I'm reading (a lot) on the subject (and posted here some thoughts) because I'm exploring the idea that knowledge might be said to be "Dependable Certified Belief". That is:

(i) P is a Belief.
(ii) P is Certified as True, meaning that a certain group of subjects agree on the truth of the proposition, even if this attribute is actually undecidable. And like all certifications is context dependent and provisional.
(iii) P is Dependable, meaning that we can safely function under the assumption that the belief is in fact true.

Condition (iii) is pretty much what you said.
In this account, Justification (and Undefeatability as some philosophers proposed) would build in the dependability requirement, whereas Certification serves as "P is true" but in a more "realistic" way.
In my account, certification is central because it serves to explain how knowledge is effectively shared by a community in which *only* a very few are directly justified, while all the rest just follow. Think of Science: we're all *just* taught scientific theories but statistically none of us do (or care about) any verification whatsoever, because we trust that "the scientists" did the right thing and followed through the scientific method. I.e, scientific knowledge is *certified* and so we accept it
(and I put it like this completely on purpose to empathize the similarities with religious believes)


After posting this, I've been reading a lot more, and I learned that this ideas have a name and have been proposed before by some 20th century Philosophers: Fallibilism (can't post links, but easy to google). It goes hand in hand with Pragmatism of course.

[Fcacciola]

I agree with the general proposition Truth should not be a requirement for Knowledge, since we truly know very little. However, once something is known to be false, we can no longer call it "knowledge". Now, it's simply a lie....

Good point. If my post about "Knowledge -> Dependable Certified Belief" shows up (I'm still moderated), then you'll see that you are describing the case when the certification expired.

[Fcacciola]

I have two definitions for you:
Knowledge: facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject.
Truth: that which is true or in accordance with fact or reality.
And yes, these both came off of google, but they do illustrate my point that: knowledge is what you learn, it is your experience and has a lot to do with your perception, while truth is already there and is unchanging, despite one's perception.

Indeed. Interestingly, however, over and over, the idea that knowledge implies the truth is (mistakenly IMO) taken for granted.

Perhaps the issue is that empirical, factual knowledge (*a posteriori* as Kant puts it) is, or seems to be, different from mathematical, logical or semantical knowledge (*a priori*), in which the truth is always decidable. Maybe we sometimes forget the difference.

NOTE: Reading on "Fallabilism", some argue that even the truth about a-priori knowledge is undecidable, in which case, all forms of knowledge are inconclusive, not just the empirical subset. I'm not sure whether I agree or disagree with this (yet).

[Eduk]

Think of Science: we're all *just* taught scientific theories but statistically none of us do (or care about) any verification whatsoever, because we trust that "the scientists" did the right thing and followed through the scientific method. I.e, scientific knowledge is *certified* and so we accept it

This is not true. Everyone is constantly verifying scientific theories all the time, I'm doing so right now as I type on this wonder of scientific advancement that did not exist before the theories which support it.
In addition all scientific theories have a mechanism which can be explained (and hopefully was in your science class, but regrettably perhaps not very well). Anyone is in a position to contemplate the mechanism for inconsistencies or contradictions. Also theories are predictive. So maybe you don't understand F=Ma but most people can do the math (once they've been handed the equation) to work out the speed of a falling projectile. Something which can, and is, tested in science classes. And again science is a club, there is no science leader, when studies are confirmed independently in different countries (sometimes totally independently, as in they never heard of each others theories they just both had the same idea at the same time) then this points at a way of removing or limiting bias. As does the various accepted rules which need to be followed to carry out experimentation in the first place, ie double blinded studies. So as a whole science is not just blindly accepted, in my opinion.
Finally, and this is not far fetched, you can if you wish repeat experiments at home. There are many very complex theories which can be tested by ordinary people (we don't all need hadron colliders).

[Sage4557]

I have two definitions for you:
Knowledge: facts, information, and skills acquired by a person through experience or education; the theoretical or practical understanding of a subject.
Truth: that which is true or in accordance with fact or reality.
And yes, these both came off of google, but they do illustrate my point that: knowledge is what you learn, it is your experience and has a lot to do with your perception, while truth is already there and is unchanging, despite one's perception.

I view "reality" as subjective. So it holds a very lose meaning for me. You and I could witness the same event and still not see the same thing. Life experience will play a role in this. So in turn our interpretation of a witnessed event is true for each of us but equally untrue based on specific life experiences. I am referring to our everyday experience of interpreting our world. And how each interpretation can build on it's self. So if I say something is unsafe based on my experience. You may say it is safe based on your life experience. Neither is true nor false. Truth is most often perception based on each persons life experience. And this makes the meaning of truth (for me) to be subjective. Losing it's value in terms of proof

[Fooloso4]

Not truth but the presumption of truth.

[Fcacciola]

Think of Science: we're all *just* taught scientific theories but statistically none of us do (or care about) any verification whatsoever, because we trust that "the scientists" did the right thing and followed through the scientific method. I.e, scientific knowledge is *certified* and so we accept it

This is not true. Everyone is constantly verifying scientific theories all the time, I'm doing so right now as I type on this wonder of scientific advancement that did not exist before the theories which support it.

Everyone is constantly supporting or adding to the evidence, but is not verifying anything per-se. When you're typing, you're not intentionally matching the outcome with the theories that predict it. Your actions are not your verification, they just (potentially) verify in the heads of someone else (if your actions are examined)

In addition all scientific theories have a mechanism which can be explained (and hopefully was in your science class, but regrettably perhaps not very well).

Certainly.
However, even in an ideal science class with an ideal science teacher, these mechanisms, which are not yet themselves the verification but just a direction for the steps that need to be taken to verify it, are skipped or glossed over due to time. A physics major is usually obtained in 5 or 6 years but it would take 20 if students were to go over every little detail.
On top of that, even in an unconstrained ideal science class in which students were put through every step of the construction and verification of every concept, at least half the class wouldn't at all care.
When I was in College (Biochemistry), even three years down the road, still most students didn't care about science and just wanted something to give them a job.

Yet, all these students do know science, even if they are clueless about why is it that things are they way science says they are.

Anyone is in a position to contemplate the mechanism for inconsistencies or contradictions. Also theories are predictive. So maybe you don't understand F=Ma but most people can do the math (once they've been handed the equation) to work out the speed of a falling projectile. Something which can, and is, tested in science classes. And again science is a club, there is no science leader, when studies are confirmed independently in different countries (sometimes totally independently, as in they never heard of each others theories they just both had the same idea at the same time) then this points at a way of removing or limiting bias. As does the various accepted rules which need to be followed to carry out experimentation in the first place, ie double blinded studies. So as a whole science is not just blindly accepted, in my opinion.

Absolutely.

Ideally at least you might want to read on Bohr's veto to the publication of any interpretation of quantum mechanics other than the so-called Copenhagen interpretation for a vivid example of a somewhat despotic science leader. But let's pretend the scientific activity is as ideal as Science ought to be.

Finally, and this is not far fetched, you can if you wish repeat experiments at home. There are many very complex theories which can be tested by ordinary people (we don't all need hadron colliders).

Taking the current body of knowledge, is actually a very narrow subset of theories, but your point still stand, in principle.

Now, besides my minor comments, what you said is all correct, but is missing my point. I wasn't clear enough.

The idea that knowledge is about certification stands on the observation of the way people effectively take external believes and accept them. As I shown, or tried to, even for scientific knowledge, the actual mental process is to just to accept a value of truth, completely despite that fact that this certificate acceptance is actually non-mandatory because first-hand verification is possible, just not practical.
Even actual scientists just accept the (explicitly provisional) truth of "certified" theories to build their own on top, and this is the way it should be, for no human can replay hundred of years of history just to make sure every premise is solid.

By the way, computer systems, such as the Internet, use this process of Certification as a way to make trust (in the sites we visit online) practical. Even if the way to verify the legitimacy of a site is know and independently doable, if we where to do that manually, we would have never done anything online. That's what certification is for: so we know something is legit because a proper agent did the homework.

By drawing an analogy, I think the exact same thing happens with knowledge: we just trust something (whereas is a domestic claim, a religious tenet or a scientific fact) is true because, for all we know, some proper agent did the homework. That is, knowledge is certified by some authority.

From this point of view, the rationality level (or lack of) a knowledge token is not so much focused on whatever form of justification for the *construction* of the belief (such as evidence) was first used, but on the justification on the trust put on the certificate. In simple words, we trust Science (for example), and so we *know* Science, even if we don't really go through the evidence ourselves.

[Eduk]

The idea that knowledge is about certification stands on the observation of the way people effectively take external believes and accept them. As I shown, or tried to, even for scientific knowledge, the actual mental process is to just to accept a value of truth, completely despite that fact that this certificate acceptance is actually non-mandatory because first-hand verification is possible, just not practical.
Even actual scientists just accept the (explicitly provisional) truth of "certified" theories to build their own on top, and this is the way it should be, for no human can replay hundred of years of history just to make sure every premise is solid.

But this is demonstrably not true. Scientifically accepted theories have been shown, many times, to either have error or be incomplete. They have been shown to be this way by applying the scientific method. I understand your point and the expression "we stand on the shoulders of giants comes to mind". But science can and does change, it changes when new evidence, or new ways of testing, or new theories combine to slowly tweak what we understand about the world we live in. Perhaps 99.9% of the time your average scientist does accept the theories which they have been taught because they trust (with good reason) that the science that went before them is sound. There is no need to reinvent the wheel for example. But there is .1% of the time left.
You can for example say that you don't care about verification. You could maybe say that most people don't care about verification (although you would need evidence to prove this). You could also say that even if we accept that most don't care about verification are there a small number who, sometimes, do?

[Fcacciola]

Perhaps 99.9% of the time your average scientist does accept the theories which they have been taught because they trust (with good reason) that the science that went before them is sound. There is no need to reinvent the wheel for example.

Which is what I'm saying....
And I meant not just, and not as much, what they have been taught in school and college but what they learn routinely as they read on their colleagues work.

So, you must be disagreeing with the implications or intention of what I'm saying. In which case, I must still not be clear enough.

But there is .1% of the time left.
You can for example say that you don't care about verification.

Actually, I couldn't even size in words how much I care about verification. Is the reason for thinking all this. And I don't mean the somewhat simpler verification of straight up physical phenomena.

You could maybe say that most people don't care about verification (although you would need evidence to prove this). You could also say that even if we accept that most don't care about verification are there a small number who, sometimes, do?

It is indeed true that Certification in the way I'm putting it sort of replaces verification (and is probably why you reject it).
It is also true that I argue that this is so because most people, even within the scientific community, don't really care that much.

But then is also true that, for example in the case of science, verification is just around the corner and anyone can access it (and in the case of other areas of knowledge, is in a whole different continent, so to speak, but still accessible if one really wants it)

My account for knowledge, which attempts to better match the actual psychological process of knowledge construction, in which certification effectively replaces verification, does not intend to invalidate science. On the contrary, intends to let us better understand how human inquiry (of which science is a part) really works.