I believe that the classic argument that conceptual regress goes back forever may be wrong.

For instance, if I try to infinitely regress on concepts, I actually end up at a point where I can't go on much further, and I find what is irreducible or basic (or perhaps even indubitable?). I find that there is a start, or something which IS really basic or foundational.

I end up with things like the distinction between is and not, the distinction between infinite and finite, between concept and not a concept. Even the schema of infinite regress must be accepted first, as well as the inferences it draws from and leads to, and the fact that it negates its opposing viewpoint.

The infinite regress, being of no beginning or end, cannot actually be situated as anything. It is formless, and yet in doing so, we are still talking about it somehow.

Does this not prove that the infinite regress problem itself has foundational beliefs, and that the members of the infinite regress are somehow existing in what has no form at all? And, if it does not have a form, how can we even say it is formless? Surely, it cannot oppose or be like anything, even itself.

Even the most radical skeptic must see that they accept similarity and distinction as concepts. They clearly aren't saying their position is not different from a non-skeptic’s, for instance.

So they actually DO have certain things which are foundational to even their own formulation. One such primitive notion is that of NEGATION or NOT or whatever you want to call it. This is not reducible to anything else. IT has no other parts, and it just is what it is. How do we know what it means? This is a whole other question, which I fear may never be answered. Perhaps the next species after us might be able to.

I am reminded of Wittgenstein’s hinges, before anyone answers with this, and even this has some criticisms. For instance, he says hinges have certain properties in other writings, and that they are not true or false, propositions or not, conceptual or not, in others. In certain notes, he says hinges are actually necessary to even explicate the concept of hinges, and yet he NEVER addresses the simple fact that hinges are opposed to not-hinges. There is an opposing viewpoint, and one in which someone (a radical skeptic) could envision as being the correct explanation of a part of reality, as opposed to the one he propounded. He cannot escape saying something is something else, and yet he is trying to explain this in the first place.

The concept of a concept, the concept of 'itself', the concept of listing things out the way I am listing them out right now, these are all concepts. I have heard some saying that concepts are wholes with parts, and primitive concepts are unities without any parts, and which are indivisible. But even this description itself is a bunch of concepts, and even saying 'bunch', 'of', etc, are all concepts, and even saying that all of these are concepts one-by-one, like so, is itself a bigger framework, and is a concept nonetheless.

This begs the question: how do we know what 'concept' means? It doesn't seem to be divisible or indivisible, and yet, as a concept, it encompasses all of them, none of them, and some of them. Everything I have written right now, including even the question and answer framework, is a concept.

This so-called infinite regress cannot be, without members which get situated in this absolute nothingness that is supposed to be the infinite. And in the infinite regress it is odd that we should prescribe a direction to it, as well as others, about the way things should be justified, like a ladder or chain - a chain which would not exist without its individual links nonetheless, and which provides the reference for us to even know that it is before or after; that it is in fact a regress from previous belief periods.

  • 4
    I don't know if "solved" is the most apt word; "dissolved" might be more precise. We know now that there are more than the three horns of the trilemma, we appreciate the combinations of the horns, for example (e.g. Susan Haack's foundherentism). Some chains of knowledge seem to go back arbitrarily far, others seem to halt, some seem to loop; why assume that there is only one regress with only one ultimate solution? Jan 2 at 2:04
  • 1
    @JoWehler the analogy she uses is of a crossword puzzle: there are individual linear regresses that are only some number of steps long, and then these regresses are linked in a coherentist way when steps overlap. One can also envision intersecting circles, though, and generally any graph-theoretic arrangement one pleases as an image of interconnecting regresses. Jan 2 at 13:20
  • 1
    @JoWehler none of them are dead ends; they're all correct solutions to the problem. Then their combinations are also correct solutions, and the maximal solution is all three of them being combined, although foundherentism and infinitistic coherentism are more popular, I suppose. Jan 2 at 19:33
  • 2
    @KristianBerry The three horns of the Muenchhausen trilemma are correct solutions to the problem? To which problem? - None of the three horns is a solution to the problem of ultimate justification. In general there is no ultimate justification.
    – Jo Wehler
    Jan 2 at 19:54
  • 2
    @KristianBerry Hans Albert's Muenchhausen trilemma was the main argument on his side in the philosophical dispute with Karl-Otto Apel whether a philosophical ultimate justification of reasoning (German: "Letztbegruendung") is possible.
    – Jo Wehler
    Jan 2 at 20:29

4 Answers 4


The Muenchhausen Trilemma is a philosophical insight, not a philosophical problem. Hence there is no need for a solution, but the call for application.

The trilemma was formulated by Hans Albert. It states that any chain of reasoning ends in one of the following dead ends

  • Infinite regress
  • Circular argumentation
  • Break of reasoning

The consequence of the Muenchhausen Trilemma is to abandon the search for ultimate justification in philosophy or science. Instead one can make progress by eliminating errors following Popper’s method to improve hypotheses by falsification.

For an introduction, further references and criticism see Muenchhausen trilemma.

Note. The Muenchhausen trilemma is not about a conceptual regress but covers - among others - the regress of reasoning.

  • I like this insight, but I tacitly have to disagree with it to a certain extent. To not be able to ground knowledge or concepts or whathaveyou at all is nonsensical as well. We could not have even the idea of not having ideas at all! They are grounded in what is ultimately of no possible characterization or whathaveyou. It is grounded in a sort of groundless ground according to some I might add. But clearly, we have ideas and these ideas as such mean something. Jan 3 at 1:55
  • (continued...) And of course to say that a ground is grounded by another ground is a sort fo groundlessness too. For then it is turtles all the way down. In this way we see that there is still a chain of something of which is being talked about. To have no justification for knowledge at all, is itself to draw an inference and make a judgement. Jan 3 at 1:55
  • 1
    @LawrenceLee: I think that you would do well to look up deductive systems; there is inherent structure to our reasoning, and the trilemma is merely an observation about such systems. Another comment pointed out that this is graph-oriented, since deductions form a graph; to me, it's category-oriented, since every category is a deductive system. See this article for framing and details.
    – Corbin
    Jan 3 at 16:56
  • @LawrenceLee, how do you know that "To not be able to ground knowledge or concepts or whathaveyou at all is nonsensical"? Or that "ideas mean something"? The essence of the Trilemma is that I can ask the same question again about whatever justification you propose. And again. And again. There is no end (infinite regress) unless you return to a previous justification (circular argumentation) or rest your position on an assertion without any supporting justification (break of reasoning). Jan 3 at 22:50
  • @JohnB I understand what you are saying, but this is the point. to even say to 'even say' is itself something. the concept of 'meta' is itself a thing as well. we can always look at a system in a meta outside perspective, but it will be such that evenetually we realize that the thigns we use to say it is the only real thing. and even then, this is itself a statement of which lacks justification. why this over another? necessary truths? well these are certainly existent too. but are they not contingent on contingency? i don't speak here. yesterday

The unspellable trilemma cannot be solved on its own terms, because it rests on an artificial definition of 'certain' knowledge which fulfils the conditions for philosophers to pontificate about how unfathomable it all is. I am certain I am wearing trousers. I am certain I am sitting in my office typing on my laptop. I am certain my house was not blown away to Oz by a twister during the night. I am certain (to take a topical example that will be recognised by at least two well-known and popular visitors to this site) the River Nile is more than an inch long. If you are prepared to accept that degree of certainty, then the trilemma is an irrelevant absurdity. Clearly we do not go through life infinitely regressing in order to find our socks (although at my age it sometimes seems like it), so one might take the view that the claimed requirement for infinite regress is patently wrong.

Personal knowledge is a adaptive configuration of the mind designed by evolution to help us survive and procreate, one we have applied to all kinds of other purposes. Whether your knowledge is certain, in the kind of absolute sense insisted upon by those who say we cannot be sure the Nile is more than an inch long, might be considered a purely academic point. That said, we used to be so certain the Sun orbited the Earth that we would put to death those who challenged the view, so clearly evolution hasn't equipped us to be infallible.

  • At any rate, denial has to be deep enough for some people to be up to their necks in it, I assume they are more than an inch tall :-)
    – Scott Rowe
    Jan 2 at 15:14
  • 2
    Infinite Sock Regression is the new name of your band.
    – user121330
    Jan 3 at 7:33
  • Scepticism about we know is more than pontificating or a parlour game, for instance it's fundamental to progress in physics, where assumptions like geocentrism have kept getting in the way of good observations
    – CriglCragl
    Jan 13 at 11:56

First, you seem overly focused on infinite regress, when one of the other prongs of the trident is 'break in reasoning' or axioms: things consisered to be self-evidently true. These began as a formalised methodology for reasoning with Euclid, and his pick of them turned out to be wrong because space is only approximately Euclidean away from large masses, making the Parallel Postulate wrong. You say 'is vs not', how about was? Which category is that in? Are mathematical truths counted in 'is', without a specific material instantiation? Infinity is considered by many philosophers not to be capable of full instantiation in our physical universe, but then that argument also applies to 'true perfect circles'. An axiom that many fans of logical inference like to champion is the Law of the Excluded Middle, but that fails to grapple with the world of our experience rather than the specific limited categories of logic. It's better thought as a choice for simplifying reasoning, than as recognising something about the world.

You say

The infinite regress, being of no beginning or end, cannot actually be situated as anything. It is formless, and yet in doing so, we are still talking about it somehow.

This is a very bad picture of infinity. The whole-number line is infinite, are we incapable of doing math? Think of the Fibonacci, we choose a starting point, begin an algorithm, and it can be taken to infinity. Pi is like that too.

You pick up on hinges. A more classic challenge to categories is thinking about what a chair is, because defining it is very contextual. Speaking of the old poker-wielder though:

"In this sort of predicament, always ask yourself: How did we learn the meaning of this word ("good", for instance)? From what sort of examples? In what language-games? Then it will be easier for you to see that the word must have a family of meanings."

-Wittgenstein, in Philosophical Investigations

Things like 'not' don't make sense in isolation. They are part of language games. See this answer discussing the language game 'opposites': Life and Death as one and the same?

So. Actual ways to deal with the challenge to Foundationalism in reasoning, which is what Munchausen's Trilemma presents.

You are doing Coherentism, turning one mode of reasoning on another. Hofstadter notes that when we do this we don't simply form linear relations between them, but what he calls Tangled Hierarchies, which allow Strange Loops, which he identifies as core to how minds work, and rather than circular reasoning more like a woven fabric of reasoning. First Order logic seeks to avoid such complications, but the world we actually live in requires Higher Order logics, that allow such self-referencing structures. Minds avoid Turing's Halting Problem, through their ability to 'step outside' of a current rule set. This is the basis for Godel Incompleteness too, where we find that we can recognise new truths within a system not deducible from it's axioms - because we can keep adding axioms, vstepping outside' of the system to reformulate our abstractions about it.

For further discussions see:

Does the first Godel's incompleteness theorem forbids the existence of a Theory of Everything?

Are there contemporary analytic defenders of the view that pattern/meaning is metaphysically fundamental and directly knowable?

Why is it impossible to hold both horns of a moral dilemma?

According to the major theories of concepts, where do meanings come from?


It has been solved. There is a beginning to the Universe. Existence, then reasoning start there. It was answered with the completion of Jewish prophecy for the Messiah, predicted by our Creator some 2000yrs ago by my estimation -- that at some point Wo/Man could complete their journey of knowledge.

To my knowledge the answer has not been published. The prophecy was pretty much abandoned by everyone in the West, both Left and the Right. We might be going a new Dark Age where the quest for knowledge was abandoned or created so much despair that no one will deserve the Answers.

  • 1
    Are you saying the Messiah was predicted during Jesus' lifetime? The prophecies of a messiah go back to the time of the end of the Davidian line being monarchs, when Judah fell to the Neo-Babylonian Empire in 586 BCE.
    – CriglCragl
    Jan 13 at 12:05
  • @CriglCragl: I would say that it was predicted AFTER the cruxifiction, Only after such a sacrifice could YHVH have then predicted that Man would complete the journey of knowledge. I know that it is completed now, because I completed it myself. I am holding the answers that have plagued Man for millenia, but no one seems to be looking for them.
    – Marxos
    Jan 15 at 4:58
  • "Well I came upon a man, at the top of a hill He called himself the savior, of the human race Said he'd come to save the world, from destruction and pain But I said How can you save the world from itself?"
    – CriglCragl
    Jan 15 at 8:08
  • @CriglCragl: Sabotage wasn't off the table. Waiting in limbo seems to be the actual case. "...'Peace on Earth' was all it said. One tin soldier rides away...".
    – Marxos
    Jan 19 at 3:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .