When two parties are in a discussion and one makes a claim that the other disputes, the one who makes the claim typically has a burden of proof to justify or substantiate that claim especially when it challenges a perceived status quo.[1] This is also stated in Hitchens's razor, which declares that "what may be asserted without evidence, may be dismissed without evidence." Carl Sagan proposed a related criterion – "extraordinary claims require extraordinary evidence" – which is known as the Sagan standard.[2]

Shouldn't both sides have to justify their claims instead of one party having to do it? I don't understand, because from a skeptic's point of view there's no absolute truth, so both parties should have the burden to prove their position and not only one. I don't understand this logic and it seems like a fallacy to me. Am I correct?

  • 8
    Just from a logistical standpoint, it makes more sense for the person who believes the claim to demonstrate the proof for it simply because they're likely more familiar with the evidence needed to do so. May 30, 2021 at 4:08
  • It can't be a fallacy unless there is a logical error.
    – user207421
    Jun 1, 2021 at 3:20
  • 7
    "It seems like a fallacy to me". Well, you have to prove it's a fallacy. We don't have to prove anything. ;) Jun 1, 2021 at 8:55
  • 4
    "because from a skeptic's point of view there's no absolute truth." <- A statement of absolute truth. :) Jun 1, 2021 at 13:48
  • 2
    "Shouldn't both sides have to justify their claims instead of one party having to do it?" See Russel's teapot, good luck with proving it doesn't exist.
    – njzk2
    Jun 1, 2021 at 20:21

11 Answers 11


You intuition makes some sense, and to clarify burdens of proof you can get some hint from the next paragraph of your same reference:

In a debate it is possible that there is a single claim (one party claims there is a chair, while the other party has the position there might or might not be a chair), or that there are multiple claims (one party claims there is a chair, while the other party claims there is none). In the latter case, both parties have the burden of proof - as the burden lies with the person who makes their respective claim. It is an argument from ignorance to argue your claim should be considered true because the opposite claim is easier to prove and has not been proven.

So it depends on your debate or conversational context, if you and your counterparty have two or multiple exclusively different claims both of which are not status quo commonly accepted knowledge then both parties have the burden to prove. But if one party's claim is status quo, or is non-exclusively different from the other's and also containing the other's claim as above first example, then obviously the burden of proof lies in the other more specific claiming party.

Finally as described in the same reference, if you try to shift the burden via arguing the other party's proposition is false because it has not yet been proven true and insist the other party to prove first (and thus feel or claim vindication by embarrassing the other party) then you'll commit the argument from ignorance informal fallacy.

  • 1
    Wait a minute this opens up the "you proof that god does not exist" type of frames: if enough people believe god exists suddenly you have to proof he doesn't exist - which is impossible to proof. Burden of proof cannot depend on the majority.
    – paul23
    Jun 1, 2021 at 12:59
  • 1
    @paul23 I think it has to - without some majority "default", there can be no "claim" that needs to be supported with evidence. A claim of X that needs to be proven only makes sense if not-X is the prevailing belief. If not-X is the prevailing belief, the burden is on X. The burden of proof can't be universal regardless of prevailing belief. An assertion can only be considered "extraordinary" in the context of some belief system. Jun 1, 2021 at 15:28
  • 2
    No the burden of proof needs to lay with the claim that adds most "magic" to a system. So that adds most changes to already proven systems.
    – paul23
    Jun 1, 2021 at 15:33
  • 1
    @paul23 You have to have some set of axioms, though. Occam's Razor is a good start, but isn't sufficient.
    – wizzwizz4
    Jun 1, 2021 at 16:31
  • 1
    @JohnBollinger if that would be the real world the scientific method would never exist. A statement needs to be falsible, and when you make a statement falsible you actually put on proof (or make it trivial to provide proof.
    – paul23
    Jun 1, 2021 at 20:13

The rule is more of a practical guide for how to live than a philosophical statement about how truth works.

Eccentric: "You should use healing crystals to treat that disease."

Doctor: "I doubt that would help. Can you prove that works?"

Eccentric: "Can you prove it doesn't?"

Doctor: "It would be impossible to prove that. Even if I ran an experiment where I surrounded my patient with healing crystals and nothing happened, that wouldn't prove anything; maybe I was just using the crystals wrong."

Eccentric: "Then, since we both have equal amounts of evidence for our theory, you should buy these crystals, just in case."

Doctor: "No, you're the one making the extraordinary claim. Get some evidence or stop wasting my time. I have work to do."

Even if the claim is not so extraordinary - for example, if a researcher claims that a vaccine is effective against a virus, or if I accuse you of stealing my wallet - we should probably default to doubting the claim unless some evidence is presented. Doubting a claim is not the same as disproving it, but it allows us to get through life without spending all our time trying to disprove wild unsubstantiated theories.

  • 2
    That's a perfect demonstration of a fallacy though. Neither side can prove a negative, because (with the exception of mathematical theorems) that is always impossible. The question then is whether either side can demonstrate evidence for some probability of a positive. The reason that "extraordinary claims require extraordinary evidence" is that the status quo, by definition, already has substantial evidence. This doesn't mean the status quo is correct, but it does mean that the evidence for the claim needs to be at least as good as the evidence for the status quo.
    – Graham
    May 31, 2021 at 10:49
  • 11
    @Graham I think there's a subtlety here that's going unrecognized. While you're correct that "You didn't prove your assertion, therefore your theory is false" is fallacious, that's a substantially different sort of argument than "You didn't prove your assertion, so I'm going to carry on as if I never heard it (until you do)." I'm not sure the later assertion can rightfully be claimed to be a fallacy.
    – R.M.
    May 31, 2021 at 16:42
  • Well, stealing a wallet is a bit different. I cannot prove you stole my wallet unless you cooperate and allow someone to search you. Unless there was a camera recording everything, that is.
    – o0'.
    May 31, 2021 at 19:35
  • 1
    If someone claims they can cure a disease using healing crystals, they are expected to provide the details: what kind of crystals, how to use them (orally? topically?), etc. Then it can be proven in research (like clinical trials) that crystals don't work any better than placebo. So the actual argument should be "Research could determine if it works, but I see no reason to spend my time, money, etc. on such research. If you believe it works, provide your own research." TL;DR the burden of proof is just a heuristic that prevents you from spending resources unwisely.
    – user31389
    Jun 1, 2021 at 16:24

Shouldn't both sides have to justify their claims instead of one party having to do it?

Yes, that's exactly what happens here.

For example, Joe claims there's an invisible hippopotamus on Jane's head. Jane claims Joe's assertion shouldn't be accepted and shouldn't have been made because Joe hadn't met his burden of proof here.

Jane is doing exactly what you said she should do. She is making a claim (that Joe shouldn't have made the claim he did) and meeting her burden of proof for making that claim (by pointing out that Joe has not met his burden of proof).

By pointing out that Joe has not met his burden of proof, Jane has met her burden of proof. As you say, both sides have a burden of proof and Joe has not met his and Jane has met hers.

Pointing out that the other side has not met its burden of proof meets your side's burden of proof if your claim is that the other side was not justified in making its claim. This is precisely because, as you said, every claim has some burden of proof.


Shouldn't both sides have to justify their claims instead of one party having to do it?

Any claim - any claim, no matter how you phrase it or what its subject - is subect to a burden of proof. If I claim that the sky is blue, I have a burden to prove that claim. If you claim that the sky is blue and I claim that it isn't, we both have a burden.

I don't understand, because from a skeptic's point of view there's no absolute truth, so both parties should have the burden to prove their position and not only one.

'Absolute truth' is a red herring here. Regardless of whether or not such a thing as absolute truth exists, any claim still has a burden of proof. Unless one is a solipsist, absolute proof is not a requirement for acceptance of a claim.

But not all positions are claims about the subject.

Let's say that you claim that the sky is blue and I say that I don't believe you. The only claim I am making is that I don't believe, not that the sky is not blue, or that the sky is a different color than blue. I can support my claim to disbelief simply by referring to the only definitive authority that can possibly exist: myself. You can then conduct the trivial proof of your claim by pointing to the blue sky and if I am benig honest then my disbelief would be dispelled.

I don't understand this logic and it seems like a fallacy to me. Am I correct?

It depends on what you're specifically referring to.

If someone makes a claim and then says that they have no burden to prove that claim, then they are simply wrong. Their claim automatically attracts a burden of proof, by virtue of being a claim. This isn't necessarily an onerous thing, and they may be able to meet their burden with a trivial demonstration - such as pointing to a blue sky after claiming that the sky is blue.

If however the person does not make a claim about the subject, then that person does not have a burden of proof.

The confusion often arises because a person who holds a position such as "I don't know" or "I don't believe that claim" is not making a relevant claim, they are stating that they do not accept that the original claim is true. While this could be viewed as a claim about their internal state, that is hardly relevant.

For example: A states "I flipped a coin and it landed on heads." B says "I don't believe your claim." A counters with "Prove that it landed on tails!" Regardless of whether or not B does in fact believe that "it landed on tails," at this point B has not made such a claim.

Note that this is not about the ontological truth of the statement, but about the belief states of the disputants. While it is true that the claim (the coin landed heads) is either true or not true, without some evidence to support the claim the respondant is not required to accept any particular truth value.

It is important to note that "I do not (believe/accept/understand/etc) your claim" is not the same as "your claim is false" or "a counter-claim is true." A statement of disbelief or non-acceptance of a claim does not automatically constitute either assigning a truth value to the original claim or assigning some truth value to a competing counter-claim, and it is an error to assume that it does so.

A classic example of this error is the many creationists who respond to "I don't believe you" with some variant of "prove that God does not exist" or "prove that the universe created itself" or some other such statement. This is a fallacious shifting of the burden of proof away from the claimant. The claimant has a burden to prove the claim, nobody else is required to disprove it, or to prove an alternate claim they did not make.

If on the other hand the response was "God doesn't exist" then the respondant has made a claim of their own, and thus acrues a burden to prove that claim.


The burden of proof is not part of logic per se. In the course of any logical work what is proved, is proved; and what is not, is not.

This is instead a rhetorical convention in dialectics meant to forestall a common fallacy getting out of hand. By deciding that there is a thing to be proven, and you either succeed or fail at proving it, you clarify problems surrounding the fallacy of ignoratio elenchi: that the nonexistence of evidence is evidence of nonexistence. If you are not careful, this seemingly silly idea gets the better of a lot of people, who walk away from merely refuting someone, believing they have proven something, instead.

If you have kept proper track of the burden of proof, the person who disagrees is less likely to come away with the impression that the opposite of what was asserted has been proved when he has only prevented the assertion from being proved.

Only if he has claimed the burden of proof, is he going to positively prove something. And he can do that. The best way to prove someone wrong is to prove they are entirely wrong, and some alternative thing that contradicts their proposal is correct, instead. But usurping the burden of proof is a clear reversal, and it generally gets mentioned and agreed to.

It still works best if you have a single assertion in your sites, and you know who has asserted it, because taking the floor also means that you are not on the defensive, and you are not stuck with the phrasing and framing chosen by the other party. When you take up the burden of proof, you get to say what you intend to prove. So a lot of times it is less of a burden, and more of a tool.

  • I disagree with the sentence "burden of proof is not part of logic per se". I think logic is the rule of the game named "debating"; and that the rules force the game to be asymmetrical: there's a "prover" and a "skeptic", and the "skeptic" player just points out to parts of the argument that he or she doesn't believe/is convinced of.
    – Plop
    May 31, 2021 at 14:44
  • 1
    @Plop. I am writing from a position that disagrees with you. "Debating" contains logic, but it also involves dialectic, rhetoric, style, and criticism. It seems obvious that those are separate things from the pure logic that makes up things like mathematics and propositional reasoning. In mathematics, there is no burden of proof. There are just proofs, nobody is counter-arguing. To claim that this is part of logic per se says that kind of math is not logic. The burden of proof is a part of dialectic -- a way arguers keep mutual interactions productive. And it is not always necessary. Jun 1, 2021 at 15:13

Some of the respondents here are seemingly mis-reading the original question: the given alternative to "the other person has the burden of proof" is not "I must accept what they say without proof", but rather, "each of us has a burden of proof for our respective positions". That of course leaves open another option: that "I have no position" - typically being undecided or indifferent.

Since there's an implied "unexcluded middle", assertion of burden should not be construed as a logical proposition.

Rather, it a pragmatic cost/benefit evaluation: proofs of statements that change our world-view are far more valuable than proofs of those that do not. Accordingly, it is worth investing effort in proving such world-view-changing statements, provided that one estimates that its truthfulness is sufficiently likely.

Value is of course personal to each participant, but where there's a "commonly held view" then it seems reasonable to assume that its value summed across all people is quite large, even if it's not large for any individual. (It's non-zero for any individual who has committed some tiny fraction of their life to memorizing and believing it.)

The act of assertion implies at least some value, and assertion contrary one's peers implies significant value, since it's risking reputation.

Indifference is also a value assessment: that neither the affirmative nor the negative would, if proven, have significant value.

Assertion of burden can also be taken as a directive, for which we have the pithy English phrase "put up or shut up".


If something is important to you, and I don’t care, why would I have any obligation to spend any effort on disproving it?

  • 2
    The question says "two parties are in a discussion", so this answer is irrelevant.
    – user76284
    May 30, 2021 at 7:13
  • 2
    This is a bad answer because it falls precisely into the fallacy: it is, for example, the perspective of flat-earthers. If people say that the earth is round, and I don't care, why should I offer any proof?
    – RodolfoAP
    May 30, 2021 at 15:23
  • 4
    @RodolfoAP The flat-earthers do care, though, or they wouldn't make so much noise about it. May 30, 2021 at 16:52
  • The answer, albeit simple, is spot on: if two people are in a discussion and one has a point to make, the other party does have to change their mind until an actual argument is produced. And since by hypothesis the former party has the desire to sway the latter party's mind, it's their responsibility to provide the argument. Flat earthers who really don't care about what not miss think don't speak out. Those who go out of their way to publicly say they don't care are hypocrites who need to cope with failing to convince anyone.
    – armand
    May 31, 2021 at 3:45
  • @JosephSible-ReinstateMonica So do the round-earthers.
    – user253751
    May 31, 2021 at 9:47

Where the burden of proof lies depends on the context and reason for the discussion.

If discussion is informal and conducted at the leisure of all parties then the where the burden falls depends on how much interest each party has in the the topic and in convincing the other party and any listeners to share their view. It's reasonable for either party to walk away when their interest in the discussion is satisfied or exceeded.

In more formal discussions, such as legal proceedings there is more likely to be an explicitly defined burden of proof, and a party who has been given that burden and refuses to attempt to prove their claim will be seen to be acting unreasonably and in some cases their claim will be assumed false.


The most simple way to approach this is, if we were to accept any claim not supported by evidence then there is no limit to what we would have to believe, as the number of fanciful unsupported claims to be made is infinitely larger than the number of truthful claims. (Compare "Kant was born in 1724" to the number of wrong claims that could be made on the same topic).

That's why "what is asserted without evidence can be dismissed without evidence". It's simply a common sense heuristic, and already the way we instinctively handle any claim that does not conform to our confirmation bias (Think about how believers in any religion handle the claims of other religions, although both are often equally unsubstantiated).

From a theory of communication viewpoint, consider also how any person making a claim is in an active role: they are actively trying to convince other people. (As making a claim about X is not the same thing as silently believing X is true). Other people are in a passive role: they wait to be convinced, or not, by the new claim. If theclaimant wants to convince people it is only normal that they also produce the means of convincing, i.e. the evidence or an argument. And if they are not trying to convince, then why make the claim in the first place?

By the way, consensus has nothing to do with it. Consensual but wrong unsubstantiated claims happen all the time. Think about superstitions. It might be the consensus in the kindergarten that Santa is real, yet if one kid were to challenge this notion (i.e. not believe in Santa, which is different from saying there's no Santa), it would be on the believing kids to provide evidence. Otherwise all the kids might as well believe in every other claim, like baba yaga, the tooth fairy or the moon is made of cheese. Consensual claims should be supported by evidence too, it's just that we don't feel the urge to challenge them because of our confirmation bias.


This would literally be absurd

It is fairly easy to construct a thought experiment where this would lead to an oscillation between all states of belief, for all uncertain states. I say that you are a Stone. You say you are not. You fail to disprove it. You are then defined to indeed be a Stone. Subsequently, you can claim that you are not a stone. Since I cannot disprove your statement, it is then so. Then my cousin Gary argues that you are, in fact, a Stone. And we end up in a round robin of failed proofs.

The only way to construct a heuristic of argument so that we avoid absurdity is to make a conservative barrier to progress. In Science, a "theory" must not only be novel, it must have more utility or be of less complexity. Three, rather conservative, tests to bulwark against pseudoscience, quackery and nonsense.

I mention science because the statements of Carl Sagan and Cristopher Hitchens must surely be read as statements of and from the philosophy of science. I would hazard to guess that it would even be statements of scientific realism, but leave that as my guess.


Simple answer -- yes. Asserting that one's opponent has the burden of proof, IS a fallacy. That of declaring that one's views are correct until proven wrong.

In evaluating a claim there are four logic states one may be in: A) currently unable to evaluate as to its truth/falsehood, b) reasonably confident of its truth, c) reasonably confident of its falsehood, and c) reasonably confident that the claim is incoherent, or otherwise intrinsically unevaluable.

All four states relative to a claim need to be supported. For instance, if someone says they have confidence in the truth of a claim, the rejection of that, and assertion of current unevaluability, itself needs to be supported.

As a pragmatic aside: When in a discussion, if a disputant resorts to a burden of proof fallacy, and demand that you do the logic equivalent of running laps to demonstrate each step of a discussion, they are likely just trolling you and are not honest interlocutors.

  • 1
    "That of declaring that one's views are correct until proven wrong." It's simply not the case. Being unconvinced about a claim until it is proven does not imply being convinced about another until proven wrong. My uncle Joe currently has either an odd or even number of $ in the bank. That you can't convince me the number is even does not mean I believe the number to be odd.
    – armand
    May 30, 2021 at 1:33
  • Your "being unconvinced" does not constitute evidence that some other reasonable person cannot draw a conclusion about any thesis about the evenness of your uncle's bank account. Your reasonableness has not been established, hence your asserting your POV should be accepted until someone else provides evidence to counter it is-- indeed -- to commit a fallacy.
    – Dcleve
    May 30, 2021 at 4:38
  • 1
    Absolutely not. By being unconvinced i am not forcing my POV on any one, just stating that I have none. This is what you miss. Every yes/no question has 3 possible answers: yes, no, and "I don't know".
    – armand
    May 30, 2021 at 4:45
  • Your "being unconvinced" has no relationship to the truth or falsity of either of the four possible logic states for a thesis. That you did not realize that I SPECIFIED those four states, and did not pay close enough attention to the reply you downvoted, AND responded to twice, to realize either this, or that that there are more than three, is noteworthy.
    – Dcleve
    May 30, 2021 at 4:51

Not the answer you're looking for? Browse other questions tagged or ask your own question.