Disclaimer: I'm not a philosopher and I'm not a native speaker, so apologies if my question is somehow flawed as a result. I'm happy to clarify anything that is unclear. So here it is:

Reason's own standards of justification require that any claim meet objective criteria independent of the believing subject. But reason cannot justify itself according to these standards. It cannot provide an objective, non-circular proof of its own legitimacy and authority. Using reason to show that reason is indeed reliable is like asking a man whether or not he is honest. In either way he'll say that he is.

So it relies on an initial "act of faith". From a purely rational perspective, this seems problematic and suggests reason falls short of its own criteria for knowledge and justification. It fails to meet the very standards it sets for other claims. Doesn't this reveal a kind of "irrationalism" at the core of rationalism - an inability of reason to be fully self-critical and self-grounding according to the objectivity it demands of other beliefs? I got the intuition if you have to begin with an act of faith, everything that follows is spoiled by it.

I know one might object that one cannot conclude from this argument that reason is indeed unreliable or at least dubious as it uses reason itself to make this point. It would be self-defeating to do so. On the other hand, it might be thought of as a reductio. We take reason's premises and bring them to their logical conclusion. Is this really self-defeating? And even if it were, is this worse than the circularity or the ad-hoc assumptions that the "defender of reason" would have to rely on?

It seems to me that the worst result the reason skeptic has to expect here is an impasse. And I don't think that this is problematic for the skeptic as she doesn't need to "win" the argument, it's sufficient for her that the anti-skeptic doesn't win either. If the question remains undecided, doubt remains. The draw is a win for the skeptic, so to speak.

I would really like to read your thoughts on this. So thanks in advance for your replies.


6 Answers 6


Your entire post starts with a suspect claim.

Reason's own standards of justification require that any claim meet objective criteria independent of the believing subject.

Says who? Is this a quote from the Handbook of the International Body of Official Reasoning Reasoners?* :D The term reason is not defined and does not demarcate practice in quite the crisp manner you believe it does. From WP:

Reasoning is associated with the acts of thinking and cognition, and involves the use of one's intellect. The field of logic studies the ways in which humans can use formal reasoning to produce logically valid arguments.2 Reasoning may be subdivided into forms of logical reasoning, such as deductive reasoning, inductive reasoning, and abductive reasoning. Aristotle drew a distinction between logical discursive reasoning (reason proper), and intuitive reasoning,3 in which the reasoning process through intuition—however valid—may tend toward the personal and the subjectively opaque.

What this means is that your belief about what reasoning is is itself a subjective and normative description, and it highlights how linguistic artifacts are often proxies for political and conventional ideas by way of nuance of definition. In fact, it goes to the heart of what the term objective really means, which is not to say that it really means this or that, but that what the term reason means itself is open to interpretation.

So, first, the claim "reason's own standards of justification require..." is itself questionable. What standard? Who is responsible for codifying standard? When was this standard codified? What are your sources about the standards? At best, we can argue that logicians have a general practice and corpus to which they refer. But let's for the sake of comity accept that there is a mythical fictional Average Logician that represents the best practices of modern logicians who study reason. Does that mean that the contemporary Average Logician believes fully in the objectivity of logic? Are there really Laws of Thought?

The laws of thought are fundamental axiomatic rules upon which rational discourse itself is often considered to be based. The formulation and clarification of such rules have a long tradition in the history of philosophy and logic. Generally they are taken as laws that guide and underlie everyone's thinking, thoughts, expressions, discussions, etc. However, such classical ideas are often questioned or rejected in more recent developments, such as intuitionistic logic, dialetheism and fuzzy logic.

The contemporary Average Logician is more of the mind, perhaps, that logic and reasoning are not universal and objective, but are particular and somewhat subjective. In fact, now is the good time to introduce the notion of intersubjectivity:

In the debate between cognitive individualism and cognitive universalism, some aspects of thinking are neither solely personal nor fully universal. Cognitive sociology proponents argue for intersubjectivity—an intermediate perspective of social cognition that provides a balanced view between personal and universal views of our social cognition. This approach suggests that, instead of being individual or universal thinkers, human beings subscribe to "thought communities"—communities of differing beliefs. Thought community examples include churches, professions, scientific beliefs, generations, nations, and political movements.

Thus, reasoning is not some magical, universal practice that floats about the universe that requires rigid and formal epistemological proof, but is rather a process that exists independent of its justificatory foundations. People have been reasoning long before the Pre-Socratics of Asia Minor and philosophy was born. What you believe to be true about reasoning, that it is objective and disembodied is a particular fiction that is popular, but not necessarily true. In fact, reasoning is an empirical fact in that it regardless of what you claim reasoning is, it occurs independent of you and is an aspect of the Wittgensteinian Sprachspiele.

We don't believe that there is a formal justification for claiming singing songs makes people feel emotions, nor is there one for claiming that feeding people keeps them from feeling hungry. Reasoning exists and is used in its most rudimentary forms without a justification for the claim that it helps produce more reliable truth about the world. Stone-age cultures reason, and little children reason. Non-logicians reason and monkeys reason. And none of that requires a thorough justification. The project of justification is a philosophical program for logicians and philosophers, and while it may add value to reasoning (I believe it does), isn't strictly speaking relevant to most people who reason.

So, what I am saying is that there is no monolithic practice called Reason. That is a trope of a narrative of a particular Weltanschauung that comes from the Enlightenment (the Age of Reason). It's a popular "cinematic trope". Laws of Thought. The Objectivity of Reason. Universal Engines of Rational Thinking. These are stories, not properly speaking facts.

* No such body exists. ;)

  • 1
    I think what I had in mind was some idealized form of rationalism that aims at acquiring knowledge in the form of justified true beliefs and in accordance with the laws of classical logic. I didn't meant monkey reason, that I can tell you. ;) Maybe I should have written rationalism instead of reason?
    – Numa
    Jul 28, 2023 at 6:50
  • +1 for Weltanschauung - "What is necessary is never unwise." - Spock's father
    – Scott Rowe
    Jul 28, 2023 at 22:35

It is rather too strong to require that any claim meets objective criteria. But you are correct to say that if we ask what rational grounds there are for being rational, it is difficult to supply a non-circular answer. Any proposed response just invites the retort, "And what reason do I have to accept that?"

There is a closely related issue within logic, which is: how can you justify deduction? Anyone attempting such a justification will find themselves using deduction and so inviting an objection of question-begging. There are several answers on this site relating to that question, e.g. this one.

This does not mean that a justification of reason is impossible or that skepticism is inevitable. Various types of answer have been offered:

  1. Skepticism about reason is self-defeating. If you try to offer me a reason to be skeptical about reason, then why not be skeptical about that reason?

  2. Reason is justified pragmatically. It just works. Do you have anything that works better?

  3. Reason is justified naturalistically. If reason didn't work, at least approximately, we would not have survived and prospered as we have. Reasoning well is positively selected for.

  4. Reason is justified because it is based on fundamental intuitive principles that are self-evident and indubitable. We just cannot imagine it being any other way.

  5. Reason and its standards are the product of a social convention. If you reason badly you will be sanctioned by the tribe you belong to.

  6. Reason is part of the scientific enterprise. Our knowledge of logic and of how to reason well advances in the same way our scientific knowledge advances. There are competing accounts of how scientific knowledge advances, but there is little doubt that it does.

I daresay there are other responses. If you google "why be rational?" you will find a number of resources on the subject. Some authors will disagree with your initial assertion that any claim should be justified by objective criteria. Maybe the best we can hope for is consistency and coherence.

  • I already granted that a skeptic conclusion might be self-defeating. But I don't know if this really does away with the skepticism. Any other response to the problem is question begging in some form. Also if we define skepticism not in it's negative-dogmatic form ("I know that knowledge is impossible.") but in it's pyrrhonian form (uncertainty about any knowledge claims), I think the Pyrrhonian has the only response to the problem that doesn't beg the question.
    – Numa
    Jul 28, 2023 at 7:04
  • 1
    I agree that any rational argument for using reason is open to the challenge of being question-begging. Being open to uncertainty about any knowledge claim is not really what I would call skepticism. That is just fallibilism: the acknowledgement that whatever I hold true, I might be mistaken. If knowledge required certainty, we wouldn't know much, if anything.
    – Bumble
    Jul 28, 2023 at 10:27
  • The fallibilist would claim that some of his beliefs are justified, if only in a tentative, imperfect manner. The Pyrrhonian wouldn't claim that any beliefs are (or aren't) justified. He suspends judgement about that. The difference is subtle and maybe fallibilists are justd skeptics being shy. :)
    – Numa
    Jul 28, 2023 at 11:27
  • Nice comprehensive answer. However, I would question the statement "if you reason badly you will be sanctioned by the tribe you belong to": When I look at politics today, this fails abysmally. If that were true, people like Donald Trump should never manage to get to the top. And he's only the highest profile offender. There are populists in just about every country, and they succeed precisely because they don't give a beep for reason. Jul 28, 2023 at 11:44
  • 1
    @cmaster-reinstatemonica Yes, I agree. The answers I listed were just possible options that have been advanced and defended by some people or other. For myself, I wouldn't put any weight on 1, 4 or 5. I also place only a little weight on 3, since we know from experience that humans are not naturally good at reasoning and our evolutionary success may be mostly down to other factors.
    – Bumble
    Jul 28, 2023 at 13:08

I'm not a philosopher, either. I think you're basically right. Philosophers and mathematicians have been trying for centuries to formalize reasoning, and while they've added significant structure to it, and discovered many useful properties, they've also discovered that there are always some holes. Gödel's Incompleteness Theorem proved this about mathematical logic.

All systems of logic and reasoning are ultimately based on some premises, AKA axioms. These have to be accepted as given, and they also include the processes that are used to manipulate the symbols in the system. But then you can reasonably ask, "Why are these the right axioms?". The answer is generally something along the lines of "Because they're the most useful for answering questions about the real world." But how do we arrive at that conclusion? Essentially, it's just intuition and concensus. Yet we know that human intuition is often not very good (although this often applies to conclusions made quickly, not conclusions that are reasoned out carefully), so why should we accept that as a basis for anything? No matter what you do, eventually you keep explaining a previous conclusion on some other arbitrary reason that intuitively seems right.

It's kind of like when a child keeps asking a series of "why?" questions, and eventually the parent resorts to "because I said so" or "because that's the way it is". Or see Turtles all the way down.

Ultimately we just have to accept that our reasoning processes are "good enough" -- they work for what we need them to do. While we sometimes make mistakes in reasoning, that doesn't seem to be a fault of the systems, but just our ability to use the systems -- "we're only human" -- and it doesn't happen enough to cause overall failure.

  • 1
    So your response would be that the "search for truth" is ultimately subordinate to practical considerations?
    – Numa
    Jul 28, 2023 at 7:08
  • 1
    Nice Answer. However, I do not think that "it's just intuition and concensus[sic]". Reason itself is also "proven" in the same way that any physical theory is: It has made countless correct predictions, and has not once been proven to produce a wrong result that had not to be blamed on something else. Like physical theories, that does not prove reason is correct (a failure to falsify is not a verification!), but it's a damn useful tool. Jul 28, 2023 at 11:55
  • @cmaster-reinstatemonica That's my point. Eventually it all comes down to "it works", but defining that rigidly is ultimately circular. "How do we know it works?" "Because our machines do what we want them to." "How do you know that?" "Because I can see it with my eyes." "How do you know your eyes are telling the truth?" ....
    – Barmar
    Jul 28, 2023 at 14:31
  • 1
    @Numa I think so. It's kind of like cosmology. We're naturally curious about how the universe was created, and increasing knowledge is a good thing, but you don't need to know it to put food on the table.
    – Barmar
    Jul 28, 2023 at 14:34
  • I think you misunderstood me slightly: Theories cannot ever be proven in physics. Physicists can only fail to find places where those theories make wrong predictions. However, once those physicists have been unsuccessful often enough, we gain confidence in a theories' predictions. This is not a circular process. It starts with something weird like "moving clocks tick slower", which prompts experimenters to check all kinds of different moving clocks, and since all these experiments fail to show deviations from the theory, we use Einstein's theories with confidence. Jul 28, 2023 at 15:40

Put another apple next to an apple, and you've got two apples.

You can therefore reason that 1+1=2, and that if you do the same with oranges you will get the same result.

Reason (facts, universal laws, evidence, truth) are all objectively, intuitively, and independently observable, no faith required.

  • How do you know that there are two apples? How do you know that the way your mind carves up reality is "objectively right"? Maybe it would be more appropriate to treat reality as an undifferentiated whole. If so there are no two apples or two anything really. The very concept of numbers would depend on there being a multiplicity of ordinary objects. It would be an abstraction derived from an evolutionary contingency about how our mind perceives reality, not how reality IS. So no, I don't think it is obvious that 1+1=2.
    – Numa
    Jul 28, 2023 at 7:18
  • That's just our normal definition of +. There are systems where 1+1 = 1, 1+1 = 0 and 1+1 = -1. And who says that applying + to apple counts should result in a total apple count? Jul 28, 2023 at 12:00
  • What is your basis for believing that apples and oranges behave the same way when you put one of them next to another one? And even if you've tried it with both apples and oranges, maybe pears are different.
    – Barmar
    Jul 28, 2023 at 14:40
  • Don't go grocery shopping with these guys, or at least avoid the Produce section. Philosophy is a pax on people who are too good at Math.
    – Scott Rowe
    Jul 28, 2023 at 22:05
  • Numa - the topic is about "reason" (defined as: a cause, explanation, or justification for an action or event) not "knowing" (whatever strange way you define that...) So the collective shared whole lives experience of billions of humans and apples certainly qualifies as "reason" that apples and maths work without faith of any kind. Jul 29, 2023 at 11:06

Over many years I have come to the conclusion that most questions in philosophy are best answered- or at least considered- through a Darwinian lens. Sure, our thoughts are always questionable- but there are some patterns of thought that seem more reliable than others, in the sense that they lead the thinker to conclusions that are more likely to ensure the thinker's survival. You can, in that sense, consider reason not to be some method of thought that is correct in an absolute sense, but instead to be a way of thinking that is more likely to guide a human out of trouble. We have evolved to be able to reason and to consider that reason is better than irrationality. In that sense, reason can be viewed as a useful recipe for making sensible decisions, and the position of the skeptic becomes entirely academic.

  • 2
    And for the other folks, there's the Darwin Awards.
    – Scott Rowe
    Jul 28, 2023 at 22:01

[some] bootstrapping amounts to a vicious form of epistemic circularity [emphasis added].

I do not know if that is the specific form of epistemic circularity you had in mind, but

this problem is not specific to reliabilism (Cohen 2002). Indeed, the problem arises for all theories that allow for “basic justification”—that is, justification that is obtained via some process or method X without antecedent justification for believing that X is reliable.

So some justifications don't have autonomous reasons to believe them, and

As van Cleve (2003) forcefully argues, theories that do not allow for [these] basic justification seem to lead to wide-ranging skepticism.

The problem seems to be

can we still give a principled explanation of why some forms of bootstrapping seem illegitimate [emphasis added]

Which is an open question.

Whether or not a solution to that, an account for when we are just ignorant of something, will defeat other arguments for radical skepticism, such as "dialectical ineffectiveness and the failure to defeat defeaters", probably depends on the nature of philosophy in general.

I personally/naively don't find them very convincing, as I don't see how different it is to doubting a premise of absolutely any argument: so what? Consistency - absolutely nothing works - so I don't have to respond - is not a virtue if it's feigned. And antiskeptics can make the alternative appealing, by e.g. showing some circular reasoning is not vicious but virtuous

A virtuous circle is a circular definition or argument whose circularity does not matter for the purpose at hand.

  • I hope that helps clarify what I was saying in the comment, and sorry if redundant and you already did.
    – user66760
    Jul 28, 2023 at 20:24
  • 2
    "The Unreasonable Ineffectiveness of Reasoning in Metaphysics"
    – Scott Rowe
    Jul 28, 2023 at 22:03
  • yup that's a good one @ScottRowe
    – user66760
    Jul 28, 2023 at 22:11
  • 1
    You're on a roll @ScottRowe. This comment but even more the 'Darwin awards' 🤣
    – Rushi
    Jul 28, 2023 at 23:09

You must log in to answer this question.

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