I've been working on epistemology, via Wittgenstein, for some time, and the following is my take on epistemology using Wittgensteinian methods. I believe traditional JTB excels within these parameters.

Knowledge as Justified True Belief—Situated in Language and Life
(Merging with Wittgenstein)

I want to defend a classical claim of JTB with a contemporary twist. I hold that knowledge is justified true belief. However, I will claim that JTB only becomes stable when we embed it in Wittgenstein’s later philosophy: language-games, meaning as use, family resemblances, rule-following, the beetle analogy, the private-language argument, and the hinge propositions from On Certainty. Add to this an analogy I draw from Gödel’s incompleteness results: just as formal systems require unprovable truths to function, our epistemic practices require certainties that are not proved from within the practice but make proof possible. That, in a sentence, is my picture.

Let me restate JTB simply. To know that P is to believe P, to have P be true, and to have genuine justification for P. I emphasize genuine. Much of the literature treats Gettier as a mortal wound to JTB. I don’t. Gettier cases work only if we confuse seeming justified with being justified. If the support for a true belief essentially depends on a false ground, the belief fails the J-condition, full stop. I mark this with an anti-false-grounds constraint: justification must not essentially rely on falsehood. That preserves the classical core without endless epicycles.

Now, what do I mean by justification? In practice, we justify in five primary ways (the five ways are not exhaustive), each with its own standards:

1. Testimony—we lean on credible sources, expertise, honesty, and corroboration.
2. Logic—deductive validity and inductive strength.
3. Sensory experience—trained observation, intersubjective checks, calibrated instruments.
4. Linguistic training—competent grasp of concepts and criteria; knowing how a term applies.
5. Pure logic/tautology—truths true in virtue of form.

These are not abstract algorithms floating above life. They are language-games, rule-governed practices situated in our forms of life. “What counts as a good reason here?” is answered inside the practice: replication and statistics in science, chain-of-custody and cross-examination in law, careful term use in conceptual analysis. Meaning is use: we learn what “justify,” “know,” and “evidence” do by watching how they function in these games. And because concepts work by family resemblance, I don’t hunt for one essence of justification; I look for overlapping patterns that guide our reasoning.

Two Wittgensteinian reminders guard the gates. First, the beetle in the box: meaning isn’t secured by pointing to a private inner object, “my beetle;” it’s secured by public rules of use. Second, theprivate-language argument: without public criteria for correctness, the distinction between using a word correctly and thinking I’m using it correctly evaporates. The epistemic parallel is straightforward. Being justified is not the same as feeling justified. That distinction only survives if standards are publicly learned, shared, and, crucially, open to correction.

Related to this is a small but important grammatical point. We use “I know” in two ways. There is an epistemic use, truth, belief, and justification that meets public criteria, and there is a convictional use, an expression of inner assurance. Conflating them produces two classic mistakes: dogmatism (conviction masquerading as knowledge) and hyper-skepticism (demanding maximal proof where our practices don’t require it). Keeping the senses apart does quiet philosophical work.

So far, I’ve kept to the foreground of justification. Now the background. Following On Certainty, I hold that our practices of giving and asking for reasons presuppose hinge certainties, things that “stand fast” so doubt and proof can get traction. Hinges are not items of evidence. They are the conditions under which evidence counts as evidence.

I distinguish two kinds. Linguistic hinges are certainties embedded in how we talk and understand: that words generally retain their meanings; that others understand me; that our talk of the world hooks onto a world. Non-linguistic foundational beliefs are certainties carried in stable patterns of action, pre- or non-verbal, but still beliefs in my sense, acquired and held within a form of life. For example, our practiced confidence in a stable, manipulable environment, the way ordinary engagement presupposes a world with enduring objects and reliable regularities. We do not typically state these as propositions; they are expressed in what we unhesitatingly do.

Hinges also come in layers. At the base are bedrock hinges, largely immutable without collapsing inquiry: there is an external world; other minds exist; meanings are generally stable. Remove those and you haven’t revised a theory; you’ve disabled the language-games that make reasons possible. Above bedrock are cultural–historical hinges that can shift, think of the movement from geocentrism to heliocentrism, without destroying our capacity to inquire. And then there are personal–experiential hinges, what stands fast for a particular life, which can shape what one finds immediately plausible. One more point here: testimony functions as a social hinge at the general level. If we tried to doubt testimony wholesale, science, history, education, and ordinary life would unravel. We can and should scrutinize particular reports; but general trust stands fast.

This is where my Gödel analogy enters. Gödel showed that sufficiently strong formal systems contain true statements unprovable within the system. My claim is that epistemic practice works the same way. It contains indispensable certainties unprovable within the practice. These are not defects. They are structural necessities. Both systems contain an “outside” that enables the “inside.” Hinges sit outside the inferential game yet make the game possible. In this picture, the unprovability of hinges is not a concession to dogma; it’s a sober description of how our games of reason actually function.

Let me turn this into a method you can apply.

1. Fix the language-game. What practice are we in, experimental science, legal reasoning, ordinary observation, mathematics, or conceptual analysis?

2. State the proposition precisely and register the belief state.

3. Assemble the reasons via the relevant methods: testimony, logic, experience, linguistic competence, pure logic.

4. Apply the public criteria that govern the game: credibility rules, inference standards, observational protocols, and correct concept use.

5. Screen for defeaters and enforce anti-false-grounds: if a key ground is false, justification fails.

6. Identify the operative hinges. Does the claim tacitly require denying bedrock? Which cultural or personal hinges are active? Is general trust in testimony doing enabling work here?

7. Disambiguate “I know.” Reserve the epistemic use for claims that meet public criteria; mark convictional uses honestly as conviction.

On this method, the verdict “S knows that P” states something robust: P is true; S believes P; S’s justification meets the public standards of the operative language-game; it does not essentially rely on falsehood; and it rests on, without trying to prove, the relevant hinges.

Objections come predictably. “Isn’t this relativism?” No. Truth remains mind-independent. Practices differ in criteria, but those criteria evolved to track reality, replication, cross-examination, calibration, and proof. “Isn’t this dogmatic?” No. Bedrock hinges are not arbitrary commitments; they are pragmatic necessities. Deny them, and you do not tidy up your theory; you lose the very game of giving and asking for reasons. “Isn’t it circular to presuppose what you need?” No. I do not deduce hinges; I show their role. That is a grammatical elucidation in Wittgenstein’s sense, showing the conditions under which our concepts do what we use them to do.

Why prefer this to bare JTB? Because it clarifies what “justification” amounts to in real life, public, trained, practice-relative reasons, while hardening justification against Gettier with the anti-false-grounds constraint. It explains why justification can’t be private. It ends regress without skepticism by acknowledging hinges where they belong. And it integrates language and life: knowledge isn’t a disembodied logical state, nor a private feeling; it is belief and truth joined by reasons that count within the living practices of a community, standing on certainties that enable those practices to work.

This is the theory I’m offering: classical JTB disciplined by Wittgenstein, undergirded by hinge certainties, and illuminated by a Gödelian insight. It’s not an abandonment of rigor; it is a clearer picture of where rigor lives.

The problem with JTB boils down to the definitions.

True - What is truth? If knowledge requires truth, then don't we run into a problem that we also have to know if something is true?

Justification - What is justification? Is it proof that something is true? Do we know what is justified? In which case, how do we justify our justification?

Belief is at least straight forward. "A claim that X is a real state." We can of course add "possibly real", but the "possibly" must be believed to be a real state vs "impossibly".

So how did Wittenstein tackle this? "Knowledge is a web of beliefs."

Justification - A whole bunch of people hold this web of beliefs.
Truth - Its true that this web of beliefs exists.
Belief - The web. And yes technically we've just create a whole other belief that's not really proven but it WORKS. And that's the key. As long as it works like science giving us thing we can use correctly in reality, we're all good.

Honestly...kind of a dodge around the issue with fancy wordplay as a lot of philosophers attempt to do. Wittenstein abandoned any formal analysis of knowledge and tried a descriptive approach. The problem with a descriptive approach is that it describes how we get along, and that getting along works, but avoids the larger question about why it works. Wittenstien wouldn't have an answer to a society that based their entire culture around a sky God that made it rain. This society would be largely functional and have enough common predictability with a scientific culture, so people would get what they needed out of it. Wittenstein wouldn't be able to answer why this culture was inferior to another if the outcomes were the same or even better, even though its clear this system lacks real 'true' knowledge by relying on a belief system of a sky God.

Its a very, "Live and let live" approach to knowledge and demands further questions and follow ups. I have my own theory of knowledge if you're interested Sam. Its basically a break down of knowledge into contextual deduction, and from there a way to rationally analyze induction through a hierarchy. I know I've critiqued your works in the past, feel free to do so in return if you're interested. Its intended to be approachable to even people without a philosophy background, but there is a summary below from the first poster that I approved if you need to organize your thoughts. https://thephilosophyforum.com/discussion/14044/knowledge-and-induction-within-your-self-context/p1

Calling this “fancy wordplay” misunderstands what I’m doing. I’m not juggling synonyms; I’m tightening the spec for when “I know” is actually warranted. Disambiguating terms isn’t a rhetorical flourish; it’s how you remove failure.

When I separate epistemic “I know” from convictional “I know,” I’m not being cute with language; I’m blocking a well-known error path where confidence is mistaken for knowledge. That has operational consequences (e.g., whether we act on a claim, fund it, publish it, or treat it as settled).

When I distinguish justification from proof, I’m not hedging; I’m aligning the concept with real practices (science, law, measurement, math) where we need defeater management and anti-luck conditions, not impossibly global proofs.

When I talk about hinges, I’m not inventing mysticism; I’m displaying the enabling conditions of the game, exactly the places where attempts at global doubt collapse the practice that doubt presupposes.

Philosophy earns its keep when it sharpens the rules of the game so that good methods win and bad ones wash out. That’s what I’m doing: specifying the norm-set that separates knowledge from confidence. If someone’s theory survives these gates, public criteria, anti-false-grounds, practice-safety, defeater screening, I’m happy to call it knowledge. If it needs us to blur meanings, relax standards, or ignore failure modes, then yes, I’ll call that wordplay.

On this method, the verdict “S knows that P” states something robust: P is true; S believes P; S’s justification meets the public standards of the operative language-game — Sam26

I admire the clarity of this position -- many thanks.

As you no doubt know, there is a question about JTB concerning whether "true" and "(genuinely) justified" are independent criteria. How would your Wittgensteinian version of JTB respond to this?

Great question, and thanks.

Short answer: on my view, truth and genuine justification are conceptually independent but methodologically coupled. “True” says how the world is; “(genuinely) justified” says your reasons meet the public criteria of the relevant practice. They are not the same property, and neither reduces to the other, but my Wittgensteinian add-ons (public criteria, anti-false-grounds, practice-safety, and hinges) are precisely there to lock justification onto truth-tracking without collapsing them into each other.

Knowledge on my account = truth + belief + this kind of disciplined justification (public-criteria, no false grounds, practice-safe), all within the hinge-enabled framework. That gives you a clean separation of properties and a principled explanation of why good justification is not mere social agreement but a world-tracking practice aimed at truth.

Calling this “fancy wordplay” misunderstands what I’m doing. — Sam26

I didn't mean you, I was commenting on Wittgenstein.

on my view, truth and genuine justification are conceptually independent but methodologically coupled — Sam26

Good. So what we want to know is, does the coupling of the methodologies for determining what is true and what is genuinely justified result in a vicious circle?

That it is a circle seems clear, but that may not be a problem. We might start by asking, is it possible to determine what is true without using the methods that "lock justification onto truth-tracking"? -- that is, without engaging in justification?

I'm guessing not, but then how do we respond to the objection that we have "collapsed into each other" the criteria for truth and justification? Note that this objection doesn't depend on claiming that justification has been reduced to "social agreement," opening the door for some invidious form of relativism. The criteria for both truth and justification can be as "objective" as you please, but we still have the problem of whether they are indeed two separate legs of the tripod.

What do you think?

I would respond the same way, whether Witt or myself.

Good. So what we want to know is, does the coupling of the methodologies for determining what is true and what is genuinely justified result in a vicious circle?

That it is a circle seems clear, but that may not be a problem. We might start by asking, is it possible to determine what is true without using the methods that "lock justification onto truth-tracking"? -- that is, without engaging in justification?

I'm guessing not, but then how do we respond to the objection that we have "collapsed into each other" the criteria for truth and justification? Note that this objection doesn't depend on claiming that justification has been reduced to "social agreement," opening the door for some invidious form of relativism. The criteria for both truth and justification can be as "objective" as you please, but we still have the problem of whether they are indeed two separate legs of the tripod. — J

Great, this is exactly the pressure point to push on, and here’s my view.

There is a circle here, but it is a benign, hinge-supported feedback loop, not a vicious circle and not a collapse. Truth and justification remain conceptually distinct, two different “grammars” in Wittgenstein’s sense, even though, in practice, our only route to truth runs through justificatory methods.

Truth: how the world is (a world constraint on speech acts).

Justified: whether one’s reasons meet the public standards of the operative language-game (science, law, everyday perception, math).

We don’t reduce truth to justification, and we don’t pretend justification is free of truth. We couple them so that justification tracks truth (anti-false-grounds + practice-safety), and Wittgensteinian hinges stop the regress (and circularity) that would make any coupling impossible.

Why this isn’t a collapse (three quick tests)

I can be genuinely justified yet false. Example: at T1, I check the official schedule, the app, and the platform announcement, public criteria are satisfied. At T2, a last-minute disruption cancels the stop. Justification stands (by the practice’s rules at T1), truth doesn’t. They come apart.

Luck test (Gettier without false grounds; environmental luck).

I can be true yet unjustified: I guess, and I’m right. Or I form the belief by a method that would easily be wrong in nearby cases (Fake-Barn County). That’s truth without knowledge because it flunks practice-safety. Again, they come apart.

Methods that systematically miss reality (bad calibration, poor replication, unchecked bias) are revised or rejected. That asymmetry, the world’s recalcitrance, shows that justification is answerable to truth, not identical with it.

Conclusion: Different roles, different failure modes, no collapse.

Why the “circle” is benign, not vicious

You asked: “Can we determine what is true without the very methods that ‘lock justification onto truth-tracking’?” In practice, no, we have no non-method magic eye for truth. But that does not mean identity or vicious circularity.

It’s like a clear window and a landscape. You only see the landscape through the window, but the window isn’t the landscape. You can replace or polish the window; the hills don’t move with your squeegee.

The loop is externally constrained: prediction, intervention, replication, calibration, cross-examination, and defeater management are world-sensitive tests. When they fail, the world corrects us.

Hinges, bedrock certainties like “there is an external world,” “other minds exist,” “meanings are generally stable,” are not evidence; they are the conditions for evidence to count. They prevent regress without pretending to “prove” themselves from within. My Gödel analogy captures this logic: unprovable statements that enable the system are a feature, not a bug.

Why this isn’t “objective-in-name-only”

You’re right that objectivity matters, and I’m not smuggling in “mere social agreement.” Public criteria in this framework are geared to track reality:

Anti-false-grounds: your case for P may not essentially run through a false premise.

Practice-safety: given the domain’s recognized error profile, the same method wouldn’t easily have delivered a false belief in nearby cases.

Error-controls: replication, calibration, pre-registration, blinding, proper scoring (e.g., Brier), independent testimony, valid inference, and correct concept use.

These are norms with teeth. They preserve the independence of truth while bending methods toward it.

The tripod stays a tripod

Truth = world-constraint (not defined by our procedures).

Belief = our doxastic stance.

Justification = public rule-governed standing (anti-false-grounds; practice-safe).

Yes, the only epistemic way to reach truth is via methods, but method-dependence of access does not entail identity of property. The two legs, truth and justification, remain distinct, even as they are methodologically coupled and mutually calibrated by the way the world pushes back.

That’s my Wittgensteinian JTB: no collapse, no vicious circle, just a disciplined coupling inside hinge-enabled practices that keeps knowledge world-answerable.

Great, this is exactly the pressure point to push on, — Sam26

Glad you agree.

method-dependence of access does not entail identity of property. — Sam26

And this is the result we want. You make a strong case, which also has the advantage of replicating very closely what we actually do when trying to assess the validity of what we think we know.

I'll devote more time to this soon -- real life calls -- and focus on what may be the weak point: a somewhat cavalier acceptance of a "world" that is supposed to remain constant across multiple conceptions of justification. Is the "window/landscape" analogy good enough here? We can see the landscape, but not "the world" -- that might be an objection. But I need to think more about it. TBC.

The analogy is weak.

Non-linguistic foundational beliefs are certainties carried in stable patterns of action, pre- or non-verbal, but still beliefs in my sense, acquired and held within a form of life. For example, our practiced confidence in a stable, manipulable environment, the way ordinary engagement presupposes a world with enduring objects and reliable regularities. We do not typically state these as propositions; they are expressed in what we unhesitatingly do. — Sam26

I would emphasize the ‘how’ more than the ’what’ in forms of life. Not just that the world has stable, reliable patterns. After all, all forms of life open up stable, patterned ways of engaging with the world. What is intrinsic to any particular form of life is how it opens up such a stable comportment. What is the qualitative nature of the way these patterns are organized, and when a qualitative pattern is transformed as one form of life becomes another, how does this change the way the world appears? It also seems to me that what is most significant about justified true beliefs for Wittgenstein is that securing the validity of a belief is not the reference
to a pre-existing fact, rule, picture, criterion or norm.
In his discussion of rule-following , Wittgenstein indicates that this is not enough for knowing what is true. Justification requires a creative, inituitive use of criteria, norms and facts that modifies them for the contingencies of actual situations.

My stance in brief...if I understand you correctly.

I agree that forms of life are about the how of engagement, not only the what of a stable world. But I reject two further moves in your reply:

1) that “securing the validity of a belief is not the reference to facts/rules/criteria,” and

2) that rule-following requires a “creative, intuitive” modification of norms to count as knowledge.

On my account, facts still bite, and public criteria remain the arbiters of epistemic “I know.” There is skilled judgment in application, yes, but it’s judgment inside guardrails, not free-form creativity. That is a core difference.

Where I agree, and how I build it in...

Forms of life: the how.

Non-linguistic foundational beliefs aren’t just tacit endorsements of what the world is like; they show up in how we move, measure, compare, and correct. My framework already captures that “how” at the method level with practice-safety: a method counts as justificatory only if, given the domain’s known hazards, using the same method in nearby cases would not easily lead you to false. That is precisely about how we proceed in a form of life.

Transformations across forms of life.

When a qualitative pattern of practice changes (say, pre- to post-Copernican astronomy; pre- to post-germ theory), some cultural–historical hinges and method-norms shift. My layered-hinges view predicts that: bedrock hinges (external world, other minds, stability of meaning) remain; practice-level norms adjust; what counts as a good reason evolves publicly, not privately.

Where I disagree

1) “Securing validity is not a reference to facts/rules/criteria”

Wittgenstein’s later work denies that mere citation of a rule or fact settles anything, but it does not follow that justification floats free of facts/rules/criteria. His point is that application is shown in training, correction, and agreement in judgments, i.e., in the public criteria of the practice. So...

Truth remains a world-constraint (thin correspondence). If your model predicts rain and it doesn’t, the world corrects you.

Justification is publicly rule-governed: replication, calibration, valid inference, cross-examination, correct concept use, and defeater management.

Hinges are the enabling backdrop; they are not extra evidence but the conditions for evidence to count.

So, no: I don’t replace reference to facts/rules/criteria with “what works.” I discipline appeal to facts/rules/criteria by showing how they have epistemic force only in use, in the practice that allows us to teach, check, and correct.

2) “Creative, intuitive” modification of norms

What application requires is disciplined judgment, not creative norm-making. Yes, rules are “open-textured;" there is no decision-procedure that eliminates judgment. But that judgment is trained and answerable to public standards. If “creative” means improvisational within the practice (e.g., a physician integrating atypical signs without violating diagnostic criteria), I agree. If it means license to bend criteria ad hoc, I reject it: that collapses the difference between seeming justified and being justified. My anti-false-grounds and practice-safety constraints exist to prevent precisely that slide.

Bottom line...

I can happily emphasize the how of forms of life, indeed, I already do via practice-safety and my method-first account of justification. What I won’t concede is that justification floats free of facts, rules, criteria, or that it needs “creative, intuitive” norm-shifting to count as knowledge. In a Wittgensteinian JTB, facts still constrain, criteria still govern, and judgment still answers to the community that can teach and correct. That’s how we keep knowledge distinct from confidence while explaining how real inquiry actually moves.

I reject two further moves in your reply:

1) that “securing the validity of a belief is not the reference to facts/rules/criteria,” and

2) that rule-following requires a “creative, intuitive” modification of norms to count as knowledge.

On my account, facts still bite, and public criteria remain the arbiters of epistemic “I know.” There is skilled judgment in application, yes, but it’s judgment inside guardrails, not free-form creativity. That is a core difference. — Sam26

The question is whether justification is fundamentally a matter of rule-following within stable criteria, or of creative, situated responsiveness. You interpret “creative, intuitive” norm-use as “free-form ad hoc improvisation” unless tightly constrained. But I mean something more like Gadamer’s phronesis — a context-sensitive application of rules that inevitably alters their force.

As Joseph Rouse interprets Wittgenstein:

We can’t appeal to social regularities or collectively presupposed norms within a practice: there are no such things, but more important, if there were they would not thereby legitimately bind us. Any regularities in what practitioners have previously done does not thereby have any authority to bind subsequent performances to the same regularities. The familiar Wittgensteinian paradoxes about rule following similarly block any institution of norms merely by invocation of a rule, since no rule can specify its correct application to future instances Practices should instead be understood as comprising performances that are mutually interactive in partially shared circumstances

The intelligibility of performances within a practice then depends upon the anticipation and partial achievement of appropriate alignment with others' performances and their circumstances,

You emphasize that “facts still bite” and “truth as thin correspondence.” This is consistent with a realist bent, but I don’t see Wittgenstein as a realist. Wittgenstein sees “fact” not as a metaphysical anchor but as a role within a language-game. This is how “facts” remain practice-constituted and reality-constrained.

When a qualitative pattern of practice changes (say, pre- to post-Copernican astronomy; pre- to post-germ theory), some cultural–historical hinges and method-norms shift. My layered-hinges view predicts that: bedrock hinges (external world, other minds, stability of meaning) remain; practice-level norms adjust; what counts as a good reason evolves publicly, not privately. — Sam26

Wittgenstein stresses that criteria gain their force through use, not through a fixed “guardrail” independent of practice. Are your guardrails themselves subject to evolution within forms of life, or do they function as transhistorical constraints? What does it mean to assert that some meta-level bedrock hinge remains? Why should it? Your inclusion of the concept of other minds and the external world as transhistorical reminds me that these are the very concepts that Husserl bracketed as part of his method of phenomenologically reducing presuppositions. I think Wittgenstein would be sympathetic to Husserl’s aim here. All hinges are ultimately contingent, because they are formed within ongoing historical processes of discursive interaction.

Wittgenstein stresses that criteria gain their force through use, not through a fixed “guardrail” independent of practice. Are your guardrails themselves subject to evolution within forms of life, or do they function as transhistorical constraints? What does it mean to assert that some meta-level bedrock hinge remains? Why should it? Your inclusion of the concept of other minds and the external world as transhistorical reminds me that these are the very concepts that Husserl bracketed as part of his method of phenomenologically reducing presuppositions. I think Wittgenstein would be sympathetic to Husserl’s aim here. All hinges are ultimately contingent, because they are formed within ongoing historical processes of discursive interaction. — Joshs

Here's a short answer:
Methods evolve; no, the core “guardrails” aren’t optional. The standards by which we employ justification (replication, calibration, Bayes, preregistration, etc.) are historical. However, the meta-level constraints that make any justificatory practice possible , publicity of criteria, other minds, an external world, and sufficient stability of meaning to teach/correct, are not optional add-ons. They’re what Wittgenstein would call hinges: not evidences, but conditions of sense for giving and asking for reasons.

However, the meta-level constraints that make any justificatory practice possible , publicity of criteria, other minds, an external world, and sufficient stability of meaning to teach/correct, are not optional add-ons. They’re what Wittgenstein would call hinges: not evidences, but conditions of sense for giving and asking for reasons. — Sam26

I think what you’re looking for is not some meta-level hinge, but the interwoven threads (family resemblance) that run though all hinges, not as a general overarching category (Wittgenstein never placed the general above the particular) but as that which emerges always in its own particular and unique way in actual use. Put differently, it seems to me what you’re talking about here is not itself a hinge, a belief, whether certain or not, but the condition of possibility for any hinge, any belief. For instance, the structure of temporality is a pre-condition for hinge beliefs, but we dont have to believe in time in order for it to make hinges possible.

Some further reflections on keeping truth and justification separate. . .

When we say, “The world pushes back,” what are we describing?

"Truth remains a world-constraint (thin correspondence). If your model predicts rain and it doesn’t, the world corrects you.” — Sam

What if we changed the example from a prediction to an observation? I see a moist situation outside my window that appears to be rain; I offer genuine justifications for my belief, “It’s raining”; but when I go outside I discover that actually it had stopped raining quite a while ago, and what I saw from my window was the rainwater continuing to fall from the high trees in my yard.

What has happened here? Can I say that my belief in the rain was justified, but untrue? Shouldn’t I have taken into account the possibility of rainwater falling from the trees – a phenomenon I’ve seen many times before – when I provided my public-standard reasons for my belief? (the “defeater” criterion). That would make my belief unjustified; at best, I should have said that it was quite likely to be raining. But then again, my “JB” -- my assertion of belief plus justifications – was not offered as a piece of knowledge. Not if I believe in the JTB theory, anyway.

So -- at what point is justification only “genuine” if it indeed tracks the truth? You say, “I can be genuinely justified yet false,” and give the example of the cancelled train stop. Yes, it appears that you were justified in believing the train would come . . . but isn’t another analysis possible? Couldn’t we say, “We all know that train stops can be cancelled. You can affirm your belief that the train will come; you can give your genuine justifications for thinking so; but only the fact (truth) of the train’s arrival will turn this into a JTB, a piece of knowledge.”

I guess I’m asking how we should characterize a “JB” -- a belief that is genuinely justified, according to your criteria, but whose truth is still undetermined. Does a person who asserts a JB assert that they know it? Only the “know of conviction,” perhaps. If JTB is meant to be the definition of knowledge, we can’t say “I know X” until we discover whether X is true – we need all three legs of the tripod. So again, how should we describe “genuine justification” in a way that preserves some daylight between that concept and “true”? How carefully must we consider every conceivable defeater before saying that our justification is genuine?

These are reservations and puzzlements about the JTB concept in general, I think. I want to turn to your more focused version, with its use of Wittgensteinian hinge propositions, especially the idea that hinge propositions “stop the regress (and circularity) that would make any coupling [of methods for determining truth and justification] impossible.” But that will wait for a subsequent post.

BTW – I don’t think you and @Joshs have a serious disagreement about “creative, intuitive modification of norms.” Josh says:

I mean something more like Gadamer’s phronesis — a context-sensitive application of rules that inevitably alters their force. — Joshs

As a third party following along, this seems to me quite compatible with:

Yes, rules are “open-textured;" there is no decision-procedure that eliminates judgment. But that judgment is trained and answerable to public standards. If “creative” means improvisational within the practice (e.g., a physician integrating atypical signs without violating diagnostic criteria), I agree. If it means license to bend criteria ad hoc, I reject it. — Sam26

Gadamer’s phronesis is not at all ad hoc, and I’m pretty sure Josh wouldn’t recommend that.

Amusingly, this is a case of not having rules for knowing when and how to apply rules! And as we know, the lack of “rules for rules” doesn’t make everything ad hoc and chaotic.

Gadamer’s phronesis is not at all ad hoc, and I’m pretty sure Josh wouldn’t recommend that.

Amusingly, this is a case of not having rules for knowing when and how to apply rules! And as we know, the lack of “rules for rules” doesn’t make everything ad hoc and chaotic. — J

:100: :up: