Sunday, October 17, 2021

The Limits of Knowledge



The Baron lifts himself out of the muck by his own hair...


Not long ago Dcleve and myself started a conversation about the limits of knowledge which I think might be interesting to some of our readers here, and thought this might be a good place to continue it.  It began with Dcleve claiming that there can be no "justified true belief" ( JTB ).  Though this was raised in conversation with someone else, I jumped in and D. provided several interesting comments which hopefully we can expand upon.

First, what is a "justified true belief"?  In a word, knowledge.  Knowledge means broadly that we have certainty of information, that we understand facts.  Here "having" and "understand" refer to the personal conviction we hold, which is a belief, and "facts" refers to truth - to a correlation between our belief and reality.  As we all know, correlation does not imply causation, and hence it is possible to have a belief which is true but for untrue or unsupportable reasons.  A stopped clock is correct twice per day, and all that.  So the final aspect of knowledge is that it is caused by reality, that the reasons we hold that belief to be true leave no room for equally explanatory alternatives.  This justifies the truth of our belief.  A justification in this sense is equivalent to the warrant offered by premises for an argument's conclusion.  Hence, all knowledge - and thus all facts - are the products of reason, of argument and debate, not of transcendent intuition.  Well, almost.

Here's the conversation we started about that:


D:  There can be NO "justified true beliefs".

T:  Well now that's an interesting statement. Presumably you agree it is not true, or not justified, or both. But then that statement of agreement must also be not true, not justified, or both. What to do?

Surely you mean to limit this statement to claims of empirical truth, since any abstracted definition serves as its own true justification. But since logic and truth states are also abstracted definitions, I must admit I'm a little confused as to what you mean, here.

Even suggesting that empirical beliefs are impossible to justify as truth seems dubious, however, since the sense experience from which they are drawn are immanent. If we agree to that, then we're right back at Descartes and Democritus. And if we don't, well that means you're unsure you're having the experience of reading what I've written, and what you might yet respond to.

I'm playing, of course, but I'm genuinely confused by your phrasing here.


D:  The ultimate answer to your question is that I have rejected analytic truth as unjustifiable under its own merits, and am attempting to make do with pragmatic truth, as a poor alternative. And I consider anyone who asserts analytic truth to be valid, to be mistaken.


Yes, if we could have direct apprehension of the world, that is an exception, although that would require a LOT MORE direct apprehension than i think is possible to plausibly argue for, before it could be significant in establishing much in the way of truths that matter. But I am suspicious of how much of experience really IS direct.

Daniel Dennett, among others, has put significant effort into identifying multitudes of cases where we are clearly confused about what we are perceiving. The blind spot in our vision, which we are entirely unaware of, is a remarkable example. The climbing spider experiment, in which a pin head is touched to three spots on a subject's arm, with a small timing gap, is nearly universally identified as a spider crawling up their arm with one gap in touches, while it is perceived as pin heads with other gaps. The confabulation that spit brain patients engage in, strongly suggests that much of our reasoning is rationalizations. Our perceptions are clearly constructed. The perceived burning from an ice cube touching one's skin, then a delayed conversion to coldness sense, shows that the quale for heat/burning appears to be constructed too. The phenomenologists of ~1900 found they could not agree on the base elements of phenomenology -- what is a quale. Something similar happened to memetics -- one simply cannot identify a base meme -- all memes appear to be either constructed from other memes, or intrinsically need to come in a micro memeplex. I strongly suspect that this inability to figure out a substrate for reduction for memes and qualia, hides some major flaw in the logic of those projects. And of relevance to this question, makes me believe that one cannot definitively draw the line between experience (direct realism), and perception (constructed mental artifact). Which brings the directness, and undeniability of quale/experiences into question.

Dennett argues that self, qualia, and consciousness are all delusions. I consider this to be a leap far beyond the data he cites. But his data -- achieves his lesser goal he wants to of subverting the undeniability of experience. If you have not encountered this family of thinkers, the delusionists, I strongly recommend reading at least a few of their works. The best single work I have found is Susan Blackmores's A Very Short Introduction to Consciousness. I review that here: https://www.amazon.com/gp/c.... Dennett himself is much more obscure and indirect than Blackmore, but if you want to read the obscure master rather than the clear acolyte, check out Consciousness Explained. Two other works I have encountered in the same family are David Eagleman's Incognito, and Wegner's Illusion of Conscious Will.

The suspicions that the delusionist data throws on direct realism, leaves really only selfhood as a plausible certainty. And Descartes found he was unable to construct anything useful from that alone.

And as for logic -- the presumption of JTB is that T is T in classical logic, and this is somehow primary, and real in the world. But the exploration of logic, from Kant through Today, and the multitudes of non-classical logics, showed that one can basically make any set of postulates one wants, so what is "true" in that abstraction is irrelevant to our world. There IS no singular logic. What logic one should use to get the most useful results in this world -- is an empirical question. And empirical questions only give pragmatic approximations to a "true" answer.

Meanwhile, all justifications run afoul of Munchausen's Trilemma. Nothing can be justified, other than pragmatically, with logically invalid justifications. Combine this with no logical truth, and we have no JTBs.

Munchausen's Trilemma states that if knowledge requires proof, and a proof is given, then a proof for that proof must be given, and so on in an infinite regress.  Hence there only three ways to knowledge: first, the infinite regress.  Second, circular reasoning, in which some proof is said to be proof of itself, and third is dogmatism, in which knowledge requires no proof.  All are, as you can see, problematic.  This problem has been known since at least as early as Aristotle, who described it in more technical terms as the Problem of First Principles.

This is not where to stop -- as reconstruction is a necessary step after deconstruction. I rely upon pragmatic approximations to truth, based on effectiveness. And pragmatic approximations to justification. But pragmatic empiricism is -- only supported by its empirical effectiveness, which is an explicit circular fallacy. That is the best we can do, so it is what I rely upon.

T: I love this post, thank you for taking the time to write it up.


Generally I think we agree on the big picture, though perhaps differ on some of the implications to draw from that. As I suspected, your refutation of JTBs is limited to empirical truths or, as you put it in more precise terms, of the ability to derive JTBs about realism directly from empirical experience. I think we're in broad agreement on this aspect.

I think we disagree somewhat about the import of the cogito and the reasons for Descartes inability to proceed from this discovery alone. But I am curious: do you feel that sense experience itself offers justification for the truth of that experience?

I don't mean we should expect that the feeling of pin pricks should prove the truth belief that a spider is crawling up one's arm, but rather that the sense of having been pricked is itself immanent. In the past I've used the metaphor of the cinema, where I argue it is impossible to be wrong about one's having viewed this stimulus, though it is impossible to derive JTBs from that viewing to the actuality of the reality it purports to depict.

We might disagree as well about our ability to derive JTBs about abstract concepts. You seem to be arguing that the existence of differing forms of logic leaves reason generally incapable of determination. Is that what you mean? I'm not sure that's true, but this again seems concerned with justifying ontological beliefs. I think, however, that the ability to create or define concepts demonstrates that JTBs are in fact possible outside of sense experience.

That's an important distinction, because it allows us to construct complex concepts which are founded on JTBs, and can be extended by JTBs. I think where Descartes goes off the rails is in his refusal to accept the fundamental ambiguity of reality. He not only wanted to offer certainty with regard to philosophy per se, but also essentially to defeat solipsism. That he failed in the latter doesn't mean he didn't succeed in the former. And the construction of complex concepts, with multiple layers of actual JTBs, seems to me necessary to our shared understanding of science and knowledge of nature broadly as a practical pursuit with a high degree of confidence, rather than a theoretically certain one.


D:

"do you feel that sense experience itself offers justification for the truth of that experience?

I don't mean we should expect that the feeling of pin pricks should prove the truth belief that a spider is crawling up one's arm, but rather that the sense of having been pricked is itself immanent. "

I am glad you recognize the issue is experience, not perception. Our perceptions are nearly 100% constructed. No, I do not. And this is because I have read and accepted the data that the delusionists I cited, use to argue in favor of delusion. I don't think their data does what they claim for delusion, but it does bring immanence into pretty serious doubt. I cited the burning from ice cubes specifically to illustrate this point. We sense SOMEHING, which we at first interpret as burning. But then realize over time, is freezing instead. Our unconsciousness uses Quales to communicate with us, but sometimes is it just a logic category, and our unconscious is lazy and miscategorizes, and the wrong quale gets picked out -- or at least we think so...

The process of seeing things is similarly semi-constructed. We can look at a photo of a stop-light, and think it is red, then look more closely and realize it is orange instead. We can be wrong about our own quales! We can look at a close-ups of a leaf-littered forest floor, with no other cues in the scene to give perspective, and take minutes to resolve it into a sensible image. What quale's did we have before we knew the context? Just jumbled!!! Look at a scene, or a picture, look away, then describe it. Or draw it. You will find that there is almost no detail. Look back, and hone in on a section of the view, and one can find lots of detail.

I can think about a stop sign - what quales do I have? For me, it is just initially the red and the white word. I can think further, then have the shape, the post, the reflective stainless back, the hardness of rapping on it, the rust of bolts, the oxidation of the paint surface (most of which I have never experienced in a live stop sign).

The degree of construction to quale's -- and our ability to "revise" our images, correcting colors, temperatures, context, and details, just do not give me any confidence in immanence. What can one say to a skeptic, who notes the errors about quales? The errors don't REALLY matter -the red quale was what one saw in the stoplight, until one saw the orange one? How about during the transition in one's perception? Immanence and logic categories are once more the kinds of absolutes that do not seem to apply to our contingent world.

I consider that we HAVE quales to be undeniable. And that the "cartesian theater" model, of our unconscious creating logic categories (the "focus here", and "the rest is this general category and unimportant" is a big feature of what we get presented, as well as the quales) partially clothed in quales for our conscious mind, (which is a VERY dualist model, very uncomfortably so for most materialists) , seems to match pretty well what we experience. And I consider WHAT quales we have, to be mostly reliable. But not quite immanent.

"You seem to be arguing that the existence of differing forms of logic leaves reason generally incapable of determination. Is that what you mean?"

Yes, that is what I mean. I believe the success to date in forming differing logics, and different math constructs, shows that "formal logic" is only one of a plausibly infinite set of logics. I cannot demonstrate this, as I am neither a mathematical theorist, nor a logic theorist. But I consider this to be beyond likely based on the success in speculative abstractions that those fields have developed to date.

I also believe that "true" would not even appear in many of them. I know it does not always or definitively appear in the 4-state logic that I believe propositions should be interpreted under (tends to be supported, tends to be refuted, irresolvable in principle, current uncertainty between categories). The tentativeness is not a definitive "true" or "false", and the other two categories are neither. With true and false getting iffy in even this small step away from formal logic -- think how likely they are to be valid categories in many of the extremely divergent logics?

And if true and false are discretionary, based on the logic one chooses, among an infinite set, how can one get to JTBs about logic propositions? You would have to prioritize ONE logic, one that has absolute true and false in it, and declare that to be the only valid metric for evaluating the others. But this is unjustifiable, hence cannot be a JTB, nor the source for JTBs.

What **I** think about this -- is that evolution has gifted us with an innate logic sense. Flawed, but innate. And apply that logic sense in self-critique, and one arrives at "formal logic". So "formal logic" is PRAGMATICALLY "justified" by its evolutionary utility. Hence, we could pragmatically choose to prioritize formal logic, and use it as the metric for the others. But then it [ formal or rather "classical" logic ] fails its own metric, as its "justification" suffers the circularity of empiricism, and the tentativeness of empiricism rather than the certainty of logic, AND -- empiricism need the four categories I listed, not the two of Formal logic -- so empirical justification of formal logic is partially self-refuted!

As a pragmatist, I can and will use formal logic as a mostly useful tool. But disagree that it is JTB.

"I think where Descartes goes off the rails is in his refusal to accept the fundamental ambiguity of reality."

It is not just Descartes who balks at uncertainty. Almost everyone who goes into philosophy, is trying to find "truth". And the pragmatic approximation to truth I think is all we can get, is a VERY unpopular conclusion among philosophers.

But be encouraged. Your tentative construct approach need not require JTBs, as the product is only tentative anyway. Try this reformulation:

"The construction of complex concepts, with multiple layers of [approximately justified beliefs], seems to me [sufficient] to [support] our shared understanding of science and knowledge of nature broadly as a practical pursuit with a high degree of confidence, rather than a theoretically certain one."

JTBs are not necessary for the tentative construct of science as a pragmatic activity.

For interest, I have an answer posted here in Philosophy Stack Exchange that you may find of interest. I extrapolated from how we need to accept indeterminate and unknown cases, in a developing and tentative pluralistic knowledge environment, and showed how this can be used to accept both libertarian free will and causation simultaneously, as local rather than global models. The subject itself may be of interest to you. But I was particularly thinking you might be interested in how this answer provides a general approach of dealing with incomplete knowledge, plus incompatible pluralisms as a pragmatist, to resolve apparently irresolvable questions. Take them as local working models, without universal generalizability. https://philosophy.stackexc...




And this is where the conversation stands.  I think it raises several important questions:


1. If the senses provide the raw data and our minds the analysis of it necessary to experiencing it, is it possible that our "experiences" do not match the data provided by the senses?  Is it possible our experiences do not match our own mental analysis?


2. There are multiple formal systems of logic, from classical logic to IFF, 4-state, and others.  Can one system of logic invalidate others?  Are there any interdependencies on which these systems rely?


3. Is certainty the same as confidence?  Can knowledge be less than 100% justified?  Can we assign degrees of confidence as minimum thresholds for certain types of decisions?


Thanks for reading, and I hope you find the topic as interesting as I do.  Or even half as much!  Thanks to Dcleve for the interesting and stimulating comments.  And remember: how one thinks is at least as important as what one thinks...

No comments:

Post a Comment