The Coherence Theory of Truth

[@KaneB]

Correspondence theory: kzbin.info/www/bejne/eYmueaSbgK5jgZY

[@THEMATT222]

First!

[@rydernia]

Would you be doing other theories as well? (Semantic, pragmatic, and so on)

[@KaneB]

​​@@ryderniaYes, the next video in the series will cover Tarski's semantic conception of truth.

[@Jason-o5s]

Cheer~~~the quality of being logical and consistent.😊

[@Allyouneedisablender]

Wake up babe, new Dr. Kane Baker vid just dropped

[@KaneB]

Hope you enjoy it!

[@myfhoood]

cannot wait for the consensus theory of truth video ;)

[@captainzork6109]

There's a video on pragmatist theories of truth, and the bit on Charles Peirce features a consensus theory

[@Lojak-exe]

The uploading of a new Kane B video will necessarily make someone happy. I think this is evidence in favor of moral realism guys.

[@KaneB]

>> necessarily Modality 😠 I don't want my videos involved in such filth!

[@matepenava5888]

I believe that it is neccesarily possible that this video will make somebody happy. However, I strongly disagree that thos would contribute anything to the case for moral realism. As morality is a concept invented by human beings, the answer to the question about morality has to be tied in with intersubjective agreement of those same hunan beings. Of course, such an agreement or consensus has to have some restrictions, it cannot be of an 'anything goes' sort.

[@vallewabbel9690]

​@@KaneBit's metaphysically possible that it is necessarily involved in such filth => it is involved in such filth. I'm sorry I dont make the rules :(

[@KaneB]

​@@vallewabbel9690 System S5? More like System SFAIL!

[@vallewabbel9690]

@@KaneB very true lol

[@JimFarrand]

Is it possible that the "Connections" criteria around 8:00 has been overstated? It's very hard to imagine a system of propositions in which every single proposition by itself implies every other, and I think such a set would have some quite strange properties. And I don't think Euclidean geometry even comes close to meeting this criterion, despite being given as an example. The axioms of Euclidean geometry don't imply each other, for example, nor are they implied by most of the theorems of Euclidean geometry. One of the strange property I refer to is as follows. Let's say our set of true statements is T, and it contains some proposition P. From this we know that P can prove any statement in T. Take some other statement Q, such that P implies Q, which is more specific or restricted that P. (Imagine a universe in which the only possible truths are about whether or not people are mortal. Perhaps P is "All men are mortal" and Q is "Socrates is mortal".) Since Q is less general than P, well often find cases like this example where Q cannot be used to prove every statement in T, and therefore according to this strong version of the criterion, we can't accept it as true. But this is very odd. P is true. P implies Q. But Q is not true. Have I made a mistake here, or does the definition of truth being given really disagree with modus ponens? That seems surprisingly... incoherent! Is it possible that "or singly" in "jointly or singly" doesn't actually mean every proposition in T by itself must prove every other by itself, but is actually just pointing out that you don't always need every proposition in T to prove some statement P? Sometimes you'll need all of them, sometimes a few, and sometimes just one.

[@gfasterOS]

"Every proposition would be entailed by the others jointly end even singly" sounds remarkably close to the prospect of deducing the answer to life, the universe, and everything from a slice of cake.

[@martinbennett2228]

Clearly coherence is a powerful tool and essential for scientifically based knowledge (in effect all academic knowledge). I am not so sure that the circularity and possibly the regress problems are a bug in the coherence approach; aren't they more of a feature? That the circularity is coherent, in that it does not give rise to contradictions, is surely part of the point. I am not sure I understood the regress argument, although formally OK, once the designated set is specified won't the coherence or incoherence be apparent? It also seems to me that coherence and correspondence are complementary, because the coherence needs to correspond to empirical reality.

[@sibanbgd100]

Coherence and correspondence are complementary in some regard, but coherence is motivated by avoiding the no-access problem, therefore they explicitly disagree on quite a lot of essential details. Circularity and regress are almost universally taken as terrible for any theory in philosophy since the ancient skeptics first defined them. Circular coherence is arbitrary, which is obvious in shorter circular thoughts like I am right because I can't be wrong. I can't be wrong because I'm right. Big theories aren't apparent in their circularity, but if their circularity gets proven or argumented strongly they are usually disregarded or modified. Any exceptions to this are usually some complicated philosophical theories that argue that circularity and regress are non-problematic. I can't remember an example from the top of my head now, but you can surely find some if you're motivated to dig enough.

[@martinbennett2228]

​@@sibanbgd100 I think problems of regress may be related to the problems of the infinite and infinitesimal. There are difficulties of interpretation, but nevertheless inevitable concepts. In this case I am not at all sure the regress is sustained once the designated set is specified. Circularity (particularly if it is a simple, limited loop) usually nullifies an argument or leads to a trivial a priori truism, but with coherence in some cases lack of circularity could imply incoherence. Geometry and mathematics can provide examples where a failure of complementary proofs to correspond would imply an error. Science assumes extensive or universal coherence, it is an ab initio assumed tenet of science. This coherence can be characterised as circularity; it raises the problem of induction, however it also implies that an incoherence would reveal an error. Examples would be the laws of thermodynamics and atomic theory: virtually any scientific endeavour assumes these theories while also in effect substantiating them. Circularity is there but only a problem in the face of extreme scepticism.

[@sibanbgd100]

​@@martinbennett2228 Science and math aren't circular, they are built upon presupposed axioms that we don't prove or question, like a=a. Everything we say and/or prove is either presupposed, circular, or regressing ad infinitum. It's a trilemma proposed by Agrippa the Skeptic.

[@KaneB]

So let's say we have a set of beliefs, S, and we want to know whether S is coherent. A necessary condition for S's coherence is that S does not entail a contradiction. The problem is that if I use inference rules R1, I won't be able to derive a contradiction; but if I use inference rules R2, I will be able to derive a contradiction. Suppose I claim that R1 contains the correct inference rules. For the coherentist, this is just to say that R1 coheres with the designated set. But whether R1 coheres with the designated set depends on the inference rules we use to test it... I do wonder though whether the coherentist has other ways of specifying what inference rules to use. One idea might be that the inference rules are already part of S. That is, our set of beliefs already includes beliefs about inferences: say, that conjunction elimination and modus ponens are legitimate, while rules like "from "if A then B" and "A", infer "not-B"" are not. So if S already includes R1, then R1 is what's used to judge the coherence of S. The fact that we can derive a contradiction in S given R2 is irrelevant, since R2 is not part of S. So yeah, in the end, I'm not sure the circularity is a problem. >> I am not sure I understood the regress argument, although formally OK, once the designated set is specified won't the coherence or incoherence be apparent? I suppose it will be apparent, but as I understand the regress argument, the question is more about how the coherentist can even assert coherence or incoherence. So we have two propositions: (M) "The moon is made of cheese" coheres with the designated. (M*) "The moon is made of cheese" does not cohere with the designated set. Both of these propositions are meaningful, but (M) is false while (M*) is true. But all "... is true" means for the coherentist is "... coheres with the designated set". For any proposition, there will be some higher-level proposition asserting that this proposition coheres with the designated set. So Kirkham thinks the coherentist will be forced, at some point, so say something like: it's a *fact* that this proposition coheres with the designated set, or this proposition *really* coheres with the designated set, where the force of the "really" is implying that it's a fact.

[@matepenava5888]

I will enjoy going through it, the first one was rather ok. I hope you go on to cover the full breadth of the topic.

[@krystofjakubek9376]

Oddly enough the only fully coherent sets I can think of are the empty set and any singleton set. Both of these comply because any statement about members of empty set is true (for the empty set this is the very first quantification for the singleton set this is the second, for all other members, quantification). It seems to me like no other options are even plausible (without proof). Also this still rests on the axioms of logic which are not implied by the rest of the system. I am fairly certain seeking such an ideal system *without* axioms is in direct contradiction with the incompleteness theorem.

[@justus4684]

WOOOOOOOOOOOOOOOOOO HAIL TO THE KANE

[@KaneB]

Thanks dawg

[@xdrowssap4456]

good video, as always! 👍🏻

[@KaneB]

Thank you!

[@horsymandias-ur]

Even in Euclidean geometry it isn’t clear (to me) that the axioms/postulates logically entail each other-they just seem non-mutually-exclusive

[@GottfriedLeibnizYT]

This is just me (and probably my failure to understand). The coherence theory of truth seems to be just the coherence theory of justification that asserts nothing about what truth is. The correspondence theory conflates truth with the world and pragmatism conflates truth with utility. What does the coherence theory conflate truth with?

[@matepenava5888]

Well, I guess mostly with justification or consensus. However, if there is anything to truth beyond consensus or utility, this has to be a very thin concept

[@KaneB]

The coherence theory of truth asserts that truth just is coherence. Whereas the coherence theory of justification says that justification consists in coherence, the coherence theory of truth says that truth consists in coherence. I wouldn't say that correspondence theory conflates truth with the world, because for the correspondence theorist, truth consists in the relation between a proposition and the world. If you claim that a true proposition just is the world, or just is some state of affairs in the world, then you hold an identity theory of truth, not a correspondence theory. On correspondence theory, there needs to be a mapping between two distinct things (say, a proposition or belief or sentence on the one hand, and a fact or state of affairs on the other).

[@dominiks5068]

Awesome, please also make a video on Deflationism

[@KaneB]

Yes, I plan to do deflationism soon.

[@jonathanbieler7939]

For the regress M2 is seems one can just go through the list of propositions in the set until they find something like "the moon is made of rock" and stop there, doesn't seem so problematic ?

[@resteric230]

@KaneB ty for the really nice video as always very interesting! I actually had a question not related to this topic but philosophy of science. I had a discussion with someone and they were kinda shitting on philosophy and philosophy of science, and when i asked them "do you know what philosophy of science is about?", they kinda read the first 2 sentences of the wiki page and said "the basic questions of the field are "what is science?" and "what should science do?". Im no expert in the field, but i do love reading the contemporary literature and watching your videos on it, and i think that is not really what the topic is about/its completely missing on so many of the important questions topics and discussions. There are so many topics unrelated to these questions that are being discussed in the field, so many questions and debates that dont fit under that label, that i think that characterisation of philosophy of science is not suffiecient, even if you want to make it really simple, i think it just doesnt capture the entirety of the field. Also in terms of relevancy in current papers and discussions, it seems to me, like there are other topics that are a lot more in the focus than these. Unfurtunately my friend is constanly citing the first two sentences of that wiki article so i wanted to ask an actual expert on the relevant field (aka you), what your opinion on this is. Maybe im entirely wrong here and thats really all philosphy of science is, and everything else can be fit into these questions, but i really dont think so... so, what would you say, as someone who has a PHD in the field? If you take the time to read this and actually answer, thank you a lot! Ive been watching your videos for a long time and the fact that you just put them out for free is incredible...!

[@KaneB]

I mean... they could just look a bit further down the wikipedia page, and they'll see a bunch of other topics. Not that wikipedia is the best source for philosophy anyway, but the philosophy of science page does cover more than those two questions. In fact, the opening sentences of that page cover more than those two questions! Anyway, some important questions in philosophy beyond those two are: Are we justified in believing that our best scientific theories are true? What is the role of values in scientific methodology? What are laws of nature? What is causality? What are natural kinds? Is there progress in science, and if so, what does this progress consist in? Are all fields of science reducible to physics, and what does "reduction" involve? What is the relationship between the scientific worldview and common sense (the "scientific image" and the "manifest image", as Wilfrid Sellars put it)? How do we establish scientific expertise, and when should we trust experts? These are all questions in general philosophy of science. You can also consider philosophy of specific sciences, say philosophy of biology. Broadly speaking, there tend to be two camps here: (1) there are philosophers who will use the science in question to deal with more traditional philosophical questions, and then (2) there are philosophers who will use philosophical tools to contribute to debates within the science itself. As an example of (1), again sticking with philosophy of biology, consider questions like: Is there such a thing as human nature? Are there innate ideas? What is morality, and are we justified in trusting our moral intuitions? Does natural selection provide a way to naturalize teleology, or has it eliminated teleology? As an example of (2), there are many philosophers who have written on topics currently disputed among biologists: such as the species problem, adaptationism, the concept of race, the foundations of evolutionary psychology, etc. I suppose if you squint, you can probably see all of these questions as forms of the two questions, "what is science?" and "what should science do?", but you'd have to be interpreting those questions in a very broad way for that to work.

[@resteric230]

@@KaneB Thanks you soooo much for the answer, really much appreciated!!! Already hyped for the next video!

[@captainzork6109]

Interesting question and answer (:

[@Johny_Locke]

I don't think it's that obvious that "Snow is white" doesn't logically imply "Kane is a human". It might be "true that universe being such that it produces snow" logically implies "at some point it produces a human that Kanezizes". The question boils down to how closely tied are objects properties to its relations. It'd read an article about that. Another angle to look at it, is whther there are contingent beings or is necessitarianism true, since necessitarianism implies every truth logically implies every other truth.

[@TheNaturalLawInstitute]

KaneB - In this current Truth series, can you please cover the topics "performative truth", "testifiability" (what humans CAN testify to), "decidability" (vs likelihood or choice), and the domain of (scope of) reducibility between sets (logic, language), mathematics(measurements, numbers), and operations (computation, or physical operations)? This sequence answers more hard questions than all others I know of. No one will do it better than you do (that I'm aware of) for public consumption. Thanks.

[@KaneB]

The plan at the moment is to cover the standard theories of truth... some of these topics are related but I'm not sure how much I will cover from them.

[@TheNaturalLawInstitute]

@@KaneB I understand. Thanks for the work you do. I'd just appreciate seeing someone other than myself covering the performative truth sequence because as far as I know it's the only epistemically possible truth we can knowingly claim to be truth. ;). And of course, you'd be the best at doing so. ;) Always a fan. -CD

[@jamescantrell2092]

Thank you

[@leonardosoutello8440]

Does anyone has the references for Blanshards quotes?

[@THEMATT222]

Interesting 🤔

[@MrCesarification]

I don't understand Blanshard's proposition. In my experience, that's not how logic works. You cite Euclidean geometry as an example, where axioms entail everything else. But nothing entails the axioms, they're axioms, they're the definition of the system. They're a completely arbitrary set of assertions, and by proving theorems _from_ them you seek to explore the space of deductive implications _from_ the axioms, That goes in one direction, nothing implies the axioms. As a matter of fact, in logic, if you do find a way to prove an axiom, you remove it from the definition, you want a minimal set. Additionally, even within the things that you do prove from the axioms, there's usually very little entailment among them. Going back to Euclidean geometry, Thales's theorerm would probably be completely useless in proving Pythagoras's theorem. That might be false, but it's just an example, you get my point. A set of formulas with complete direct deductive implication links between every single pair in the set seems so restrictive as to be completely useless for theoretical endeavors. Am I completely missing the mark?

[@KaneB]

Blanshard gives Euclidean geometry as the closest exemplar of what he has in mind, not a perfect exemplar. None of our actual beliefs meet his standard, but mathematical systems like Euclidean geometry come the closest.

[@MrCesarification]

@@KaneB Hi man! I'm so happy to get an answer from you, wow, thanks! I'm not assuming that you have time to explain the world to me so here goes nothing 😅. My question is aimed more towards why anyone would think something with those characteristics would be a useful standard for a definition of truth, even an idealized one. In what sense is Euclidean geometry, or FOPC, "close" to this space of formulas where every single pair is inferentially bound, in both directions? There would be no benefit, it would render them useless. It seems contrary to the very aim of mathematical logic. So this seems like a completely arbitrary standard, I don't understand the motivation for it, that's all. Anyway, thanks for all your work, it's wonderful stuff, big fan over here!

[@KaneB]

​@@MrCesarification I assume the reason why Blanshard views Euclidean as "close" to his ideally coherent system, or at least as the closest exemplar we have, is that in Euclidean geometry, various complex and surprising theorems are logically entailed by the axioms. Euclidean geometry is an example of a system where (1) propositions are connected to others by logical entailment, but also (2) it doesn't seem to be merely trivial; that is to say, learning what is entailed by the axioms of Euclidean geometry is not like the trivial inference from "Frank is a bachelor" to "Frank is unmarried" (of course, spelling out the trivial/substantive distinction might not be so easy...). As for why anybody would define truth as ideal coherence in Blanshard's sense... well, I'm not sure about this either. It seems to me that much of the motivation for coherence theory came from dissatisfaction with correspondence theory. There were also some metaphysical idealists who thought that coherence theory followed from their idealism, though this is probably just another manifestation of dissatisfaction with correspondence theory, since idealists will obviously be suspicious of "facts" or "states of affairs" as standardly conceived by correspondence theory. Unfortunately, positive arguments for coherence theory are few and far between. One thing that's worth bearing in mind, though, is that many of the philosophers talking about truth are not so interested in the *utility* of truth. They would see a distinction between the question, "what is truth?" and the question "what is the utility of truth?" It's not clear why a theory of truth has to be useful in any sense. Perhaps it would be helpful to think of this from the point of view of an omniscient being. God knows everything that there is to know. He knows everything that is entailed by the axioms of Euclidean geometry, and he has no need to engage in any kind of reasoning from these axioms, since his knowledge of the entailment relations is immediate. It doesn't seem that mathematical logic would be of any benefit to God. But we'd probably still say that he has true beliefs about it. If we were to achieve an ideally coherent set of beliefs, we would be omniscient. Mathematical logic would no longer be of benefit to us, but we would still have true beliefs about it.

[@СергейМакеев-ж2н]

@@MrCesarification If you make a set whose every member is a conjunction of the form "all of the axioms AND X", where X is some kind of a theorem, then they will all follow from each other. Maybe Blanshard had something like that in mind.

[@jurgiskarijotasnekrosius7968]

Good stuff

[@KaneB]

Thanks!

[@СергейМакеев-ж2н]

The "ideal test" argument at 28:00 sounds more like an argument FOR global skepticism! 1. Coherence is the best imaginable test for truth (because of internalism, we have no access to anything outside propositions). 2. But (the many systems argument) a coherent system can be false. 3. Therefore... there is nothing we can do to get to truth!

[@italogiardina8183]

Arguably the coherence theory of truth relies on apperception primordialism which seems to be a property of human evolution through a process of group A fortiori as a form of consensus coherence truth. The notion of leader member exchange plays out through group formation where the members who cohere to a functional prototype of behaviour form a coherent community of in-group that then polarises with a coherent out group which is ideologically distinct. A recent phenomenon is when fields of science being culturally assimilated as syncretic interpretation to align with religious sentiment. The Bauhaus design theory might well be a modernist's ideal coherence theory towards an aesthetic truth with similar religious sentiment to world religious beliefs but coherence based on geometry and colour design for it uplifting sense to the boxes which modernist's reside.

[@quantumfineartsandfossils2152]

ROLFMAO well you are on the golden median if you accept the truth you need art literacy for that

[@InventiveHarvest]

This seems to be more about knowledge than truth. The parts of the video that talks about method vs definition and "what we mean when we say true" seem to undermine coherence as a definition. The only method that can provide certainty about everything would be an omniscient being revealing the truth to us. And omniscience is impossible. Again, I do not see truth as a condition for knowledge. Knowledge is derived from pragmatic, progressive method. The fact that the lion share of propositions are under determined in terms of correspondence is a good thing. The journey of knowledge is never ending, always growing. The growth is what is important, not the consistency. Circles are fine as long as the circle is growing and a fine way to escape the regression problem. The external justification of a circle does not come from what went into the circle, but rather what comes out of it. When I get some $ and pay off some medical debt, I will probably be joining your Patreon. Please leave the video up for the future.

[@KaneB]

Yeah, I think coherence is much more appealing as a view about justification, rather than truth. When coherentists like Blanshard say that our actual beliefs can only have some degree of truth, and that the "pure truth" of the maximally coherent set is not even expressible in our language, it's no longer clear to me what the ideal of coherence is even supposed to be. Like, I can imagine making adjustments to my actual beliefs that make them increasingly coherent. I can imagine, for instance, detecting an inconsistency and removing it. But I have no grasp on what the maximally coherent set is supposed to look like. Even if an omniscient being were to reveal it to us, presumably we wouldn't be able to understand it, unless the omniscient being also supplied us with wholly different concepts.

[@InventiveHarvest]

@@KaneB yeah, it reminds me of what Lakatos said about changing definitions in order to save a theory from refutation. Also thinking of Lakatos, knowledge doesn't come from coherence. An ideal set would be stagnant if not downright degenerate. Knowledge comes from the dialogue between prover and refuter - disagreement is necessary.

[@MrGustavier]

a -J- a