Problem of The Criterion
The Problem Of The Criterion is an epistemolgical puzzle that supposedly demonstrates the impossibility of knowledge.
[quote=Roderick Chisholm (1916 - 1999)][The Problem Of The Criterion is] one of the most important and one of the most difficult of all the problems of philosophy.[/quote]
Chisholm's version:
[b]1. What do we know? or What is the extent of our knowledge?
2. How do we know? or What is the criterion for deciding whether we have knowledge in any particular case?[/b]
The classic version of The Problem Of The Criterion claims that before we can answer question 1 we must answer question 2 BUT before we can answer question 2 we must answer question 1, effectively creating an infinite loop with no way out.
According to the Internet Encylopedia of Philosophy (link above), the general form of The Problem Of The Criterion is:
[b](1) Which propositions are true?
(2) How can we tell which propositions are true?[/b]
The difficulty is identical to Chisholm's interpretation - before we know which propositions are true, we must have a criterion to identify them but before we have a criterion, we must know, beforehand, which propositions are true.
There seems to be an embedded contradiction in The Problem Of The Criterion viz. that it claims, at one moment that
1. Propositions can't be true prior to the existence of a criterion (hence the need for a criterion)
and the next moment it claims that
2. Propositions have to be true prior to the existence of a criterion (hence the need for particulars)
Comments...
[quote=Roderick Chisholm (1916 - 1999)][The Problem Of The Criterion is] one of the most important and one of the most difficult of all the problems of philosophy.[/quote]
Chisholm's version:
[b]1. What do we know? or What is the extent of our knowledge?
2. How do we know? or What is the criterion for deciding whether we have knowledge in any particular case?[/b]
The classic version of The Problem Of The Criterion claims that before we can answer question 1 we must answer question 2 BUT before we can answer question 2 we must answer question 1, effectively creating an infinite loop with no way out.
According to the Internet Encylopedia of Philosophy (link above), the general form of The Problem Of The Criterion is:
[b](1) Which propositions are true?
(2) How can we tell which propositions are true?[/b]
The difficulty is identical to Chisholm's interpretation - before we know which propositions are true, we must have a criterion to identify them but before we have a criterion, we must know, beforehand, which propositions are true.
There seems to be an embedded contradiction in The Problem Of The Criterion viz. that it claims, at one moment that
1. Propositions can't be true prior to the existence of a criterion (hence the need for a criterion)
and the next moment it claims that
2. Propositions have to be true prior to the existence of a criterion (hence the need for particulars)
Comments...
Comments (73)
So knowledge does not require awareness of knowledge, Rather, successful action is an index of knowledge.
Quoting TheMadFool
A logical argument ends as merely a worthless sophism, when the means to create it necessarily presuppose the very impossibility it is meant to demonstrate. That is to say, on the one hand, if one believes there is a problem, the problem cannot be about knowledge, and on the other, if one knows there is a problem he contradicts himself by attempting to demonstrate knowledge is impossible.
Stereotypical human proclivity......use reason, the purpose of which is to alleviate confusion naturally, for the creation of it artificially.
This looks like a theory of truth, not a theory of how we know what is true.
Why? The other way around is obvious but this seems obviously not.
Suppose you study a messy but widely used concept like "fascism," and your research results in a formula, something like 'all fascist regimes have at least five of the following seven properties ...'; your formula selects the regimes we all agree are fascist, rejects the ones we all agree aren't, and gets you booked on cable news shows to tell us whether the borderline cases we're concerned about are or aren't, and how we'll know when they 'cross the line,' if they do.
This is just business as usual for determining a criterion: you take your data as already partially sorted since no one is claiming that maybe we don't know which regimes are 'really' fascist. We notice some natural variation and try to tighten things up a bit, and then we let "fascist Mk 2" deal with new data and corner cases. (If convincing objections are raised to how it handles these, it's on to Mk 3.)
Sadly this procedure has an obvious shortcoming as an approach to the theory of knowledge.
(If you don't like that answer, consider that you would have to know at least one thing before you deploy your criterion and start sorting statements into "true" and "false," namely what the criterion is.
In neither case do we need or expect a complete partition of statements, just some starting point, something already known.)
I would simplify it to the pragmatist formula of how to reply to scepticism. It is indeed a loop. Belief can ground doubt because doubt can ground belief.
We have to assert a belief - claim a hypothesis. And then it can be held as true to the degree it survives a countering act of doubt - a search for evidence of exceptions.
We thus need to believe to be able to doubt. We need to start by creating the possibility of having been wrong.
And then we must actually doubt until we are ready to believe. The belief has to pass the test of inductive confirmation.
Knowledge then becomes perfectly possible within that rational-empirical framework. It just carries a proviso of not being infallible knowledge. Instead it is knowledge adequately tested for its fallibility. It passes the actual test of being reasonable.
Science is part of the Methodist approach to the problem - answering the second question first i.e. developing a criterion for truth/knowledge. This isn't exactly a solution to the problem though - it actually ignores it altogether.
Quoting Pantagruel
There has to be a criterion for what it is to know before you can claim to know anything. You know that you can ride a bike because 1. you can ride a bike and 2. there's a criterion that helps you in establishing whether that (riding the bike) qualifies as knowledge.
Quoting Mww
:chin: Point! Does it make sense to say we know what The Problem Of The Criterion (PC) is if the skeptical take on it is that we can't know anything at all? If we claim that we know PC then that means we have a criterion but that's exactly what PC says is impossible. So, the option of taking a skeptical position (that we can't know anything) based on PC is self-contradictory. It seems then that the criterion for truth/knowledge that allows us to know what PC means is the same criterion that we can't know PC.
I wonder what this leads to? Any ideas?
See above
Quoting Srap Tasmaner
According to the Internet Encyclopedia Of Philosophy:
[quote=Internet Encyclopedia of Philosophy]Hence, perhaps the best way to formulate the Problem of the Criterion in its most general form is with the following pair of questions (Cling (1994) and McCain and Rowley (2014)):
(1) Which propositions are true?
(2) How can we tell which propositions are true?[/quote]
Quoting Pfhorrest
Well, there are two possibilities:
1. We need to know true propositions to help us decide the criterion. The criterion is basically a classificatory system and we need instances of true propositions to determine their essence as it were which will give us something to go on in developing the criterion
2. True propositions are part and parcel of the criterion itself. Suppose the criterion consists of statements A, B, and C. We have to know A, B, and C are true and unless the criterion is, if I may say so, self-referential i.e. can establish the truths of its own component statements, this will be an aspect of The Problem Of The Criterion.
What do you think?
So what do you mean here:
Quoting TheMadFool
Are you talking about what makes a proposition true, or about how we know that it is true?
Aren't they the same thing?
No.
Why?
Many truth claims. What criterion of truth are you using?
Pragmatic, ok, but you still need a criterion for truth/knowledge.
(1) 'Dewey defeated Truman' is true iff [ insert theory of truth ]
(2) Alice knows that Dewey defeated Truman iff [ insert theory of knowledge ]
(2a) Alice knows that 'Dewey defeated Truman' is true iff [ same kind of theory as in (2) not as in (1) ]
(1) is about what happened, maybe about what properties that sentence must have and what other properties the world must have for that sentence to be true; (2) is probably about whether Alice believes that it happened and whether we by-and-large approve of how she came to believe that, but maybe something else; (2a) is a recasting of (2) to emphasize that (this sort of) knowledge is propositional.
Roderick Chisholm's account of The Problem Of The Criterion
1. Go for Particularism i.e. answer question 1 first. We assume some propositions as true.
2. Choose the Methodist way i.e. answer question 2 first. We develop a criterion.
3. Opt Skepticism i.e. accept the The Problem Of The Criterion is unsolvable and ergo, conclude that nothing can be known.
@Mww's post seems to expose the heart of the matter. Basically, any criterion of knowledge/truth that allows us to know what The Problem Of The Criterion is is self-refuting because The Problem Of The Criterion asserts that we can't know anything. This means that The Problem Of The Criterion counts as something that if it can be known then it's something that can't be known.
Since the criterion we used to know The Problem Of The Criterion, whatever it is, is the exact same criterion we use to know other things apart from The Problem Of The Criterion, it follows that this criterion, again whatever it is, is self-contradictory. :chin:
Ok. I'd like you to have a look at the preceding post. I'll get back to the point made in your reply in a while.
Yes, truth is different to knowledge but the point is truth precedes knowledge in the sense that before one can claim to have knowledge, truths must exist. Ergo, I feel, the need to reframe The Problem Of The Criterion in more basic terms - propositions and the criterion that establishes their truth.
Well good luck to you.
Thanks but I suppose you were being sarcastic. :smile:
What are you talking about? That was it.
https://en.wikipedia.org/wiki/Criteria_of_truth#Pragmatic
People knew things long before there were criteria of knowledge, don't you think?
Ok. How do you know that this criterion for truth is correct? Methinks being pragmatic is just a cop out, another way of avoiding The Problem Of The Criterion.
Knowledge and truth are judgements - they need a criterion.
So you are saying no one knew anything until there was epistemology? That doesn't seem right......
Not so. People were logical before Aristotle developed formal logic. However, that doesn't mean the principles of logic were different before and after Aristotle.
Quoting apokrisis
The Problem Of The Criterion shows us that we can't develop a criterion for knowledge/truth without knowing some propositions as true but that's impossible to do without already having a criterion. Ergo, any attempt to propose a criterion, even pragmatic ones, is avoiding the issue.
That it is impossible to know it is true is what is accepted. And from there, we move on to a more achievable ambition. Why bang your head on a brick wall?
Do you know that "...it is impossible to know it is true"? Since you have what you call a pragmatic criterion for truth, you must realize that it's self-refuting.
Exactly. So having knowledge does not depend on having a criterion of knowledge. Knowledge must be self-validating.
A newborn infant cries as a reflex. At some point, however, it learns that crying summons its mother. Now it knows that crying equals summoning mother. So instead of just crying automatically, it can choose to cry. So what its knowledge has done is endow it with the power of choice, i.e. will. So the fundamental criterion of knowledge is its successful application. Which is essentially pragmatic I suppose. The infant does not need to know anything about its own knowledge in order to have that knowledge. That would be "meta-knowledge". Which is really what epistemology boils down to.
If you can show it is not impossible to know after all, then my position might be in trouble.
That's not what I said. There was, had to be, a criterion. How else would you know a proposition is true/false? We just didn't make that explicit for reasons that are obvious - nobody was bothered by it.
The Problem Of The Criterion does the exact opposite I'm afraid - show you that it's impossible to know. That's where things get a little hairy - for you can only know that nothing can be known only when you know what The Problem Of The Criterion is and what it entails. I can't know that I can't know anything. :chin:
Quoting TheMadFool
Time for you to decide which of these two statements you believe.
First things first. The three of us are operating under some criterion of knowledge/truth and I daresay it is exactly the same criterion as the vast majority of us are working under.
1. All three of us (even everybody) agree(s) that we [I]know[/i] what The Probelm Of The Criterion is and what it entails.
2. The Problem Of The Criterion entails that we can't know anything at all
Ergo (from 2)
3. We can't know The Problem Of The Criterion and what it entails
4. 1 and 3 contradict each other
5. The source of the contradiction in 4 is the criterion all of us, including the three of us, are using
6. This criterion, whatever it is, that we're using must be a flawed/faulty criterion because it leads to a contradiction
My question is, what's the criterion all three of us are using? There's something wrong with it. Why the contradiction?
See above.
Please read this
Propositional knowledge is a particular subset of knowledge and not its primary form for organic beings. All kinds of creatures "know" things. So saying that a special feature of propositional knowledge "knowing that it meets a criterion" is a limitation on knowledge per se is invalid. It is like saying that all matter must be wet because water is wet. The "criterion" of knowledge in its most general form is its successful application, as I suggested.
If A knows X then X has some practical ramifications, such that acting in concert with the knowledge X will have different (and intended) consequences versus acting without the knowledge X.
Criterion? Sounds like some super hero who instead of saving the town and beating up the bad guys just makes everyone take an annoying test instead. Criteria, basically? Standards for something ie. a logical/factual floor to stand on? A reference point that is "true" or rather more likely to be true than false and more likely to produce something useful (neither including nor excluding the idea of "pragmatism").
My question would be how sure are you that semantics don't play a role here? Do you believe every word open to interpretation in Chisholm statements were interpreted by you in the manner as they were written or intended? Knowledge may or may not be absolute. Basically, an "absolute truth" or something- anything- that is simply less false than another view can constitute knowledge.
Well, I'm only guessing here but if you must talk about animals, notice that what they know we can know in terms of propositions e.g. wildebeest know where a watering hole is and just like that animal knowing is expressible in propositional form.
This is beside the point thought. What I'm actually interested in is what the criterion for knowledge/truth we're using in this conversation is.
Quoting apokrisis
There is no conflation. To know The Problem Of The Criterion is to know the truth of the [i]propositions[/I] that constitute it or are entailed by it.
If you don't mind telling me, what criterion of truth are we using here to know what The Problem Of The Criterion is?
It's possible that I've misunderstood. I'm interested in knowing where exactly I or others could've misconstrued Chisholm?
Quoting TheMadFool
Quoting TheMadFool
Which parts of the above questions are likely to be misinterpreted?
Quoting TheMadFool
The latter would have been my idea as well, with the antecedent indicating criteria themselves imply a system in which they serve as operative conditions. Ultimately, such system must relieve the self-contradiction inherent in the PC, because without at least a logical proof that knowledge is indeed possible, we are left with nothing but mere sophisms which hold no profit whatsoever. Mathematics provides sufficient reason for claiming proof that knowledge is possible, and from that, a system arises in which the criteria for all judgements, both a priori and a posteriori, follow necessarily.
Nevertheless, it remains an unavoidable scandal for the human condition in general, that there is no absolute unconditional proof of anything, including knowledge itself, derived from a system conjured solely by the possessor of it, because that possessor is himself a condition for the system.
So......metaphysical reductionism asks, is the validity and purposefulness of knowledge, its reality being tacitly given, worth neglecting the intrinsic circularity involved in justifying its very possibility. If worth neglecting, the PC falls; if not worth neglecting, the PC may or may not fall but the system itself does. Hence, the inevitable and altogether irreconcilable sophisms.
Then a generalized criterion of validity for propositional knowledge would be that it is (potentially) capable of self-validation. So depending on the nature of the proposition, it would fit within a larger scientific-coherent framework, a la Karl Popper.
This is what's intriguing because in math there's complete freedom - there is no necessity except that of logic - and we may begin with axioms that contradict each other as long as they're put in differenf systems a la Euclid's parallel postulates.
This implies that the 1st question is not as important as the 2nd question in The Problem Of The Criterion. It doesn't matter what is true as much as it matters how/why they're true. This understanding is reflected in non-Euclidean geometry. Can we extend this to philosophical matters as well? Indeed we can - people do say things like, "suppose what you say is true".
It must be noted here that when we claim Euclid's parallel postulate is true in Euclidean geometry AND Euclid's parallel postulate is false in non-Euclidean geometry we're using two different [contradictory] propositions (what's true is not the same) but logic to decide truth (the criterion for truth is identical).
What say you?
Self-validation. Ok. I can go with that but what I want to know is does The Problem Of The Criterion make sense to you? It can only make sense to you if you know what it is but that's impossible because The Problem Of The Criterion says that you can't know anything at all, including The Problem Of The Criterion itself. So, if you know The Problem Of The Criterion then you can't know it - contradiction. What led to this contradiction? The Criterion which allowed us to make sense of (know) The Problem Of The Criterion. Something's off...
Know/knowledge and criterion. We all "know" (right?) what these words mean. But they can hold meanings some support and others think are either overly-complex or even oversimplified. Or you can just be lazy and call it splitting hairs.
Quoting TheMadFool
We "know" what we're told. Empiricism apparently. Fire- hot. Snow-cold. Pain-bad (at least for the individual) and so on and so forth. Let's call these common sense for now, from which I believe the term was derived from. The extent of our knowledge is simple- or rather can be determined simply. What you can and cannot answer and if answering can point to sufficient and logical enough reason. ...even if that "reason" is "cuz someone told me so" lol. Or I suppose the ultimate "I saw it (at least) once."
Yes, I think that the problem of the criterion arises from comparing knowledge in two different senses, what we know (which is always specific) and how we know it (which is a question about knowledge in general, at the meta-level). So the second question, "How am I capable of having knowledge at all" is really a red herring. I do have knowledge; you have knowledge; my do has knowledge.
Maybe we cannot account for how we know, any more than we can account for how we think. It is just a faculty. To me, it makes more sense to investigate the causes of error....
Yes. Go on...
I say....pretty much agreeable, with the exception that in math there is not complete freedom, insofar as any mathematical structure must adhere to the principles of universality and necessity. But I understand you to mean we are free in our development of different mathematical structures, consistent with the paradigms to which they might apply. The usefulness of Schrodinger's Equation is itself predicated on Nicomachus‘ arithmetic, among others of course.
————
To follow up:
Quoting TheMadFool
Experience a posteriori, understanding a priori.
Quoting TheMadFool
Judgement pursuant to the categories of modality. To know anything whatsoever, it must first be possible, then it must exist, and from those, the necessity of it is given.
Ehhhhhh........or not. Lotsa things can only be assumed in philosophy, right?
1. What is the criterion of knowledge/truth you/we are using here?
2. Does The Problem Of The Criterion make sense to you?
Let me have a go...
Setting aside the matter of truth, the criterion of knowledge at play here is justification as a must-have item. Why else are we arguing? We are arguing, right?
If so consider the argument contained in The Problem Of The Criterion. It entails, for reasons you already know, the fact that nothing can be known. Basically, The Problem Of The Criterion justifies the inadequacy of any and all logical justification i.e. knowledge is impossible but it all hinged on you having knowledge of The Problem Of Induction. In other words, logic isn't self-validating as you would've liked. In fact it's self-refuting in this context.
‘S-ok. I’m not here to teach, so......as long as I can’t be proven wrong, I’m happy enough.
Ok, every thing that thinks has some knowledge, right? Everyone on this board knows something. I know my name. If a being is able to survive, it must have knowledge. If a being is able to communicate, it must have knowledge. So knowledge is possible. Ergo the "problem of the criterion," whatever it does establish (if anything) does not refute the possibility of knowledge. Like I said, it's confusing knowledge simpliciter with knowledge about knowledge.
Poor choice of words on my part. Just bowing out of the discussion.
I think the Problem of the Criterion is slightly more interesting than the Church-Fitch paradox or the Münchhausen trilemma because it invites solution, and possible solutions have something interesting in common. Robert Stalnaker, at the beginning of his Locke lectures, put it this way: that the right place to begin philosophy is not at the beginning, but in the middle.
For your kind consideration:
The father of taxonomy Carl Linnaeus may have something to offer by way of a solution; after all the issue is about how to classify propositions - whether they're knowledge or not.
When Carl Linnaeus classifed animals into mammals, bird, reptiles, amphibians, etc. it wasn't the case that he knew, beforehand, what these various classes of animals were - he began by collecting specimens, studying them, looking at anatomical characteristics that were similar or dissimilar and these classes of animals emerged from that study. Carl Linnaeus didn't possess a criterion for the various classes of animals before he classified them - the criterion emerged from his studies of animals.
The Linnaean version of The Problem Of The Criterion:
1. What are mammals/birds/reptiles/amphibians?
2. How do we know what are mammals/birds/reptiles/amphibians?
The Linnaen Problem Of The Criterion would have us think that we can't know what are mammals/birds/reptiles/amphibians without having a criterion that allows us to make that classification but we can't have a criterion for doing that without knowing some mammals/birds/reptiles/amphibians beforehand.
Carl Linnaeus developed his criterion for the various classes of animals not because he knew beforehand what mammals or birds or reptiles or amphibians are but from his study of animals and noting similarities and dissimilarities among them. The classificatory criterion was then developed from these notes. This goes to show that having a classificatory system doesn't imply preknowledge of the classes.
Likewise, concerning the epistemic Problem Of The Criterion, constructing a criterion for truth/knowledge doesn't imply that we must already know which propositions are true/not. In Linnaean fashion, what was/is/should be done is get our hands on a list of propositions, note similarities and dissimilarities and, automatically, classes/categories will emerge from that list. One possible outcome from such a study will be, given any list of propositions, two categories of propositions: one category that consists of, say, justified, true beliefs and another category of propositions that lack any or all these features. We then decide to call the first category consisting of justified, true beliefs, knowledge and the second category made up of propositions lacking some or all these features, non-knowledge.
What's important to note here is that we didn't have a criterion when/before we sorted/classified the list of propositions - we simply classified them based on observed similarities and differences - and that the criterion was developed from/after the list was sorted. The Problem Of The Criterion is averted because we didn't use a criterion to sort/classify propositions. The propositions simply fell into different categories based on how they were alike and how they were unlike. This was followed by the development of a criterion based on the sorted/classified list of propositions. The sorted/classified list of what counts and doesn't count as knowledge didn't depend on there being a criterion beforehand although the criterion for knowledge is based on that list. :chin:
The criterion didn't emerge, the definition of each animal was expanded to include the species. How is this example different in principle from saying, for example, if X is red then X is coloured? If X is a bear, then X is a mammal? At best, I think you've injected the problem of the ontological status of universals (abstract categories) into what you've presented as an epistemological dilemma.
I stand by the transcendental argument that, since knowledge is self-evidently a reality, it cannot be impossible to achieve knowledge. Knowledge means you believe something and what you believe is true. If a belief is false, then it is refutable. If it is not false, then it is not refutable.
Nothing irreparably wrong with most of that, but the point is being overlooked, in that the original PC claims knowledge is impossible. Knowledge herein pertaining not of things, but knowledge itself. In order to refute the PC, the negation of it must be demonstrated. We don’t need to prove we know about things, we only need to prove we have the capacity for knowledge, the things being whatever they may. It follows that rudimentary mathematical concepts are sufficient to justify the possibility of knowledge, because it is we ourselves who create the predicates mathematics employs. The added bonus being, that experience serves as the apodeictic proof that these definitive inventions are sound, and thus it is that we can prove our capacity for knowledge beyond its mere possibility.
Still, the vagueness you mentioned arises, in that the conditions under which mathematical concepts themselves are possible, being given from the very same notion of categories already mentioned, yours of the empirical kind, mine of the rational, remains valid, but for all that, nonetheless theoretically plausible. And when facts are absent, as is always the case with epistemological speculation, all that’s left to work with, is theory.
You mean to say that Carl Linnaeus knew, beforehand, what mammals/birds/reptiles/amphibians are? But the characteristic defining qualities (the criterion) of what these various classes of animals are were developed after he took note of how these classes of animals were alike and unlike.
If knowledge is Description then it makes no sense to suggest we start off with "zero Description" or a blank slate.
Linnaeus did not create the taxonomic structure, he only described it. And he could be wrong. Alternate taxonomies may also apply.
I would say essentialism is like a minimal or necessary description. Necessary qualities. I say this because the platonic or aristotelian
essentialism is nonsense. Just positing unobservsble abstract entities.
You might be well acquainted with the problem without knowing the truth of it. A logical argument can be deemed valid and yet not validated.
Well, that's true about particular individuals but surely you won't deny that [I]knowing[/i] The Problem Of The Criterion involves, in terms of justified true belief theory of knowledge, the justification, the truth, and belief in re the propositions in The Problem Of The Criterion and the propositions that can be inferred from it.
Quoting Pantagruel
Of course he didn't create the taxonomic structure but the point is there were no compelling prior reasons, i.e. there was no pre-existing criterion, that guided Linnaeus in the classifcation scheme he developed. As far as I know, the terms mammals/reptiles/aves/amphibians didn't exist, as a criterion of categorization, before Linnaeus did his thing.
Again, if I don’t accept that criterion, the problem as stated doesn’t exist for me. Just because a paradox can be proposed and accepted as such doesn’t mean one is trapped. It means one is demonstrably better off considering the alternatives.
What about the contradiction if one opts for skepticism?
1. If I know The Problem Of The Criterion then I know I can know nothing (premise 1)
2. If I know I can know nothing then I know I can't know The Problem Of The Criterion (premise 2)
3. If I know The Problem Of The Criterion then I know I can't know The Problem Of The Criterion (1, 2 HS)
4. If I know I can't know The Problem Of The Criterion then I don't know The Problem Of The Criterion
5. I know The Problem Of The Criterion (assume)
6. I know I can't know the Problem Of The Criterion (3, 5 MP)
7. I don't know The Problem Of The Criterion (4, 6 MP)
8. I know The Problem Of The Criterion AND I don't know The Problem Of The Criterion (contradiction)
9. I don't know The Problem Of The Criterion (5 to 8 reductio ad absurdum)
Basically, The Problem Of The Criterion is unknowable :chin:
@apokrisis@Pantagruel
Precisely. This problem isn't for any kind of a pragmatic epistemology. :up: