The point of my paper (the paper this thread is based on) was to strengthen traditional JTB with Witt's later philosophy.

Thanks, this is very good. You've helped me understand better what's at stake in the particular, problematic "why?" When this level of "why?" is reached, the question is actually no longer about the original subject (in this case, having two hands). It morphs into a demand for justification of the entire conceptual apparatus. And since this must inevitably include the concept of "justification" itself ("borrowing the tool") . . . we have a problem.

The only place I'd put up a little flag would be when you speak about "the ordinary criteria as illegitimate in advance" as a sign of hinge-questioning doubt. This problem goes back to Descartes, and is outside the scope of your OP, but I would make the case that Cartesian methodical doubt doesn't actually posit anything as illegitimate in advance, and neither does the skeptic in our example. In both instances, the skeptic is really raising a question about certainty, not about some subject. No genuine doubt is being expressed -- existential doubt, I might call it -- concerning two-handedness. Doubt is wielded as a tool to sculpt certainty, to learn how far the whole method can be pushed before we have to cry "I can conceive of no further doubt!"

The only place I'd put up a little flag would be when you speak about "the ordinary criteria as illegitimate in advance" as a sign of hinge-questioning doubt. This problem goes back to Descartes, and is outside the scope of your OP, but I would make the case that Cartesian methodical doubt doesn't actually posit anything as illegitimate in advance, and neither does the skeptic in our example. In both instances, the skeptic is really raising a question about certainty, not about some subject. No genuine doubt is being expressed -- existential doubt, I might call it -- concerning two-handedness. Doubt is wielded as a tool to sculpt certainty, to learn how far the whole method can be pushed before we have to cry "I can conceive of no further doubt!" — J

From a Wittgensteinian standpoint, I still can’t make sense of the Cartesian maneuver as doubt. Doubt isn’t a free-floating posture you can apply to anything at will. It’s a move inside a practice, and it only makes sense where there are criteria for what would count as checking it, correcting it, or settling it. When you try to doubt everything at once, you don’t get a deeper form of doubt, you remove the background that gives “doubt,” “test,” and “justification” their role in the first place.

That’s also why it helps to keep my four senses of certainty explicit: subjective certainty (conviction), hinge certainty (what stands fast and makes inquiry possible), epistemic certainty (defeater-resistant stability in practice), and absolute certainty (logical or moral necessity). Ordinary, practice-governed justification aims at epistemic certainty. Cartesian doubt pushes toward absolute certainty, and it treats hinge certainties as if they were ordinary claims waiting for ordinary support. From the Wittgensteinian angle, that isn’t a legitimate extension of doubt, it’s a shift in the grammar of the activity.

The issue isn’t that the skeptic declares ordinary criteria illegitimate. The issue is that the exercise changes the kind of question being asked, and once it does that, it stops looking like genuine doubt within a practice and starts looking like a philosophical performance aimed at an impossible standard.

The “magically turns into not knowledge” worry comes from treating knowledge as if it had to be indefeasible. — Sam26

Which is exactly what JTB does. I understand you’re trying to modify it to address that issue, but I’d rather just toss the whole thing in the hopper.

We say, “I knew, given what I had,” and we also say, “I was wrong.” Those aren’t contradictions. They mark two different evaluations: what was justified at the time, and what we now know after a defeater has arrived. — Sam26

Those are fine things to say. So why do we need all the JTB trappings—with or without U. What I want to do is focus on the important part of the JTB formula—J. Adequate justification is what’s needed. It’s the best we can do. What does adequate mean? It depends mostly on the consequences of being wrong.

That's also why my guardrails matter. They're not demanding absolute certainty. They're making explicit the constraints we already use to separate knowledge from lucky success and from fragile support. Defeater screening, in particular, is not a demand to foresee every possible
counterexample. It's the ordinary discipline of not ignoring live alternatives and known failure modes. — Sam26

You talk about this with really different language than I do. That’s why I stopped participating in this discussion. As I said, I don’t want to try to make JTB work, I want to discard it entirely.

If you keep “adequate justification,” you haven’t really escaped JTB, you’ve just renamed it, and you’ve made key distinctions harder to state.

“Adequate justification” still presupposes a target. Adequate for action isn’t the same as adequate for knowledge. If you collapse knowledge into “what I’m willing to act on,” you turn epistemic standing into risk tolerance. Two people with the same evidence can differ just because they’re more cautious. That’s prudence, not knowledge.

The so called JTB trappings are the distinctions you still need. Truth is about how things are, belief is what the subject holds, justification is the belief’s standing. If you throw those out, you end up rebuilding them anyway to explain the difference between “I acted responsibly” and “I knew.”

Your engineering language already matches my guardrails. QA/QC is No False Grounds. Standards of practice are Practice Safety. “Don’t ignore known failure modes” is Defeater Screening. The real question isn’t JTB versus adequacy. It’s whether “adequate” stays vague, or whether you spell out the failure modes that make a belief look supported when it isn’t.

Discarding JTB doesn’t remove Gettier, it relocates it. If knowledge is “adequately justified true belief,” you still need to exclude lucky truths. Either you tighten adequacy to rule out luck, which is exactly what Practice Safety and defeater sensitivity do, or you let luck count as knowledge.

If you keep “adequate justification,” you haven’t really escaped JTB, you’ve just renamed it, and you’ve made key distinctions harder to state. — Sam26

No. It’s adequately justified belief. Truth isn’t in the equation.

Adequate justification” still presupposes a target. Adequate for action isn’t the same as adequate for knowledge. — Sam26

As I define it, adequate justification means sufficient to allow responsible decision making. So, yes. Adequate for action is the same as adequate for knowledge.

The real question isn’t JTB versus adequacy. It’s whether “adequate” stays vague, or whether you spell out the failure modes that make a belief look supported when it isn’t. — Sam26

You can make the standards for adequacy anything you want. It’s a question of risk management and liability.

Discarding JTB doesn’t remove Gettier, it relocates it. — Sam26

As I quipped previously, let’s not get into Gettier. I have strong negative feelings about the whole subject.

That said, I lean more toward the first analysis than the second. Is it possible to doubt whether I have two hands? Yes. Do we know the general sorts of things that justify our (comparative) certainty about two-handedness? Yes. — J

My analysis would actually be closer to the second than the first, and I largely agree with 's reply, though I framed it differently. The basic idea is that some "why?" questions misfire because they try to put into question that which makes questioning possible in the first place. Framed like this, it can be seen as a retorsion argument, or argument from pragmatic contradiction.

The basic idea is that some "why?" questions misfire because they try to put into question that which makes questioning possible in the first place. Framed like this, it can be seen as a retorsion argument, or argument from pragmatic contradiction. — Esse Quam Videri

Yes, I think we're all on the same page with this now. I raised a question about how the Cartesian method does or doesn't fit this conception, and that could be a good discussion too (I appreciate what @Sam26 is saying about it) but outside the OP.

Doubt isn’t a free-floating posture you can apply to anything at will. It’s a move inside a practice, and it only makes sense where there are criteria for what would count as checking it, correcting it, or settling it. — Sam26

Right, that would be the difference between a skepticism that is meaningful, versus one that merely capitalizes on our language's ability to frame "why?" questions. As above, I think Descartes stays within a recognizable practice using his method, but TBC elsewhere.

Back to Sam's questions:

Do you think Gettier cases still refute JTB even if we build in the guardrails and the “+U” clarification.

Is my diagnosis too dependent on relabeling the justification condition rather than answering the core intuition. — Sam26

Your analysis of JTB pretty well answers the first question, I think. You conclude, "Gettier is trading on an impoverished picture of justification," implying that Gettier cases never refuted JTB, properly understood, in the first place. Your guardrails and "+U" show why, and once again it's key to your concept that it is a showing, an explanation, not an added ingredient required to save JTB from the jaws of Gettier.

The second question is a little unclear to me. How would you lay out "the core intuition"?

You’re right that my view is that Gettier trades on an impoverished picture of justification, and that the guardrails and the “+U” are meant to show what our practice already treats as decisive, not to bolt on a fourth condition. I find it amazing that people find Gettier significant.

On your second question, by “the core intuition” I mean the feeling that makes Gettier cases grip us:

The belief is true.

The subject can qbite supporting considerations that look like justification.

Yet the truth shows up by luck, because the support is fragile, or dependent on a false ground, or insulated from the mistake-conditions the practice recognizes.

That’s why we resist calling it knowledge, even though it can look like JTB is met on the surface.

This is not rare. It happens a lot in ordinary life because most of what we call knowledge isn’t absolute certainty. We’re usually dealing with epistemic certainty, defeater-resistant stability in practice, and in many domains that stability is unavoidably probabilistic.

My diagnosis isn’t just relabeling. It’s an attempt to explain why the Gettier intuition arises so often: we mistake surface marks of support for justificatory standing. Once you make explicit the ordinary constraints that already govern standing, No False Grounds, Practice Safety, and Defeater Screening, the Gettier case stops looking like a refutation of JTB and starts looking like a case where the belief was true, but the route was lucky or too fragile to count as knowledge.

My analysis would actually be closer to the second than the first, and I largely agree with ↪Sam26's reply, though I framed it differently. The basic idea is that some "why?" questions misfire because they try to put into question that which makes questioning possible in the first place. Framed like this, it can be seen as a retorsion argument, or argument from pragmatic contradiction. — Esse Quam Videri

I agree. Some “why?” questions misfire because they try to question what makes questioning possible in the first place. In that sense the move is retorsion, a pragmatic contradiction: it borrows the norms of justification while attempting to place the background that makes those norms operative on trial. The key is to keep the point limited to that kind of global “why?,” and not treat it as a dismissal of ordinary, practice-governed activiity for reasons aimed at epistemic certainty.