My take on burden of proof

While I’m criticizing Austin Cline, I also want to say something about his article on burden of proof in the context of the atheism vs theism debate.  Again, I have nothing against Cline, and in fact he brings up several points that I agree with:

A more accurate label would be a “burden of support” — the key is that a person must support what they are saying. This can involve empirical evidence, logical arguments, and even positive proof.

The “burden of proof” is not something static which one party must always carry; rather, it is something which legitimately shifts during the course of a debate as arguments and counter-arguments are made.

The part I disagree with Cline’s assertion that the (initial) burden of proof “always lies with the person who is making a claim, not the person who is hearing the claim and who may not initially believe it.”

This leaves open the question of who has the initial burden of proof when both people are making claims.  For example, what if the theist claims there is a god, and I claim there is no god?  According to Cline, atheism refers to people who make no claims about gods, and thus atheists don’t have the initial burden of proof.  However, I am part of the subset of atheists who positively claims there are no gods, so where does that leave me?

In my analysis, burden of proof is the answer to three different questions:

  1. Who wins if no further arguments are made?
  2. Who should win if no further arguments are made?
  3. Whose turn is it to advance the argument?

1. Who wins if no further arguments are made?

If no arguments are made, it is easy to see who will win: absolutely nobody.  Everybody will just continue to believe the things that they believed before.  An absence of argument leads to the status quo.  Thus, you could say that the person with the burden of proof is the person who wants to change the status quo.  If I want to persuade a theist, and the theist doesn’t really care, guess what, I have the initial burden of proof.

2. Who should win if no further arguments are made?

If no arguments are made, it is harder to say who should win.  Imagine that you are an external observer to an argument, and you have to decide who wins purely based on the arguments presented.  And then nobody decides to present any arguments.  How should you decide?

Your decision can only be guided by heuristics.  For instance, Occam’s razor suggests going with the simpler theory.  A world without wizards is simpler than a world with wizards, because all that magic stuff is rather complicated.  (Which are more complicated: gods or wizards?  In my humble opinion, gods.)

From a Bayesian perspective, you’re judging prior probabilities.  For instance, the prior probability of wizards is fairly low, and the prior probability that there are no wizards is fairly high.  Thus, the side that believes in wizards has the burden of proof.  Note that (contra Cline) it doesn’t make a difference whether I believe there are no wizards, or if I simply lack any belief in wizards.  The burden of proof is the same.

3. Whose turn is it to advance the argument?

The trouble with burden proof as discussed so far is that nobody wants it.  It becomes a way of avoiding responsibility.  I say, “you have the burden of proof!” and you say “no, you have the burden of proof!” and we get nowhere.  Therefore, I want to present a perspective of burden of proof where it’s not considered a burden.  Basically, if you have the burden of proof, that simply means it’s your turn to speak.

The trouble with atheist vs theist arguments is that different people believe in a lot of different gods, and for many different reasons.  Usually when I discuss with a theist, I don’t know beforehand what particular god they believe in, or what particular arguments they prefer.  Thus, it makes sense for the theist to go first, unless they really want to hear my refutation of the Ontological Argument Nobody Ever Makes (which is a real thing that I do).

That theists have the burden of proof is not a disadvantage, unfortunately quite the opposite.  I don’t know where to start when arguing whether gods exist.  Gods: less plausible than wizards.  And there’s the problem of evil.  Your turn.


  1. Pierce R. Butler says

    Very nicely put (though I gotta quibble about the difference between “who’s” and “whose”, ‘cuz copyeditors are like that…).

  2. Emily (luvtheheaven) says

    Thanks for all your posts about atheism lately. They make me happy and give me a fresh new perspective on things that are personal and relevant. 🙂

  3. brucegee1962 says

    Interesting post. I’ve got some further thoughts about the heuristics used for your question #2. My sense would be that the heuristic that most people instinctively use is the argument ad populum. That is, the burden of proof lies on whichever position is held by a minority within the society.

    After all, that’s pretty much the position in science: there’s a much higher level of proof expected from Einstein to argue AGAINST Newtonian physics than there is for Newtonian physics to defend itself, because Newtonian physics is what everyone believes.

    So if a theist were to make the argument “Regardless of whether religion is true, churches do a good job of promoting morality. Most people in our society believe this, even some atheists who drop off their kids at Sunday School at churches they themselves don’t attend,” then the burden of proof would be on the atheist to disprove this, because that really is the belief of most people in our culture.

    Of course, from the position of both logic and history, we know that the argument ad populum isn’t a particularly good one, as we can see from looking back at times when the majority believed in things like vampires and witches. But it’s an extremely convenient heuristic, because if you DON’T use it, and start off by a priori rejecting and wanting to test every single assumption that your society believes, then you’re likely to have a fairly miserable and lonely life.

  4. EnlightenmentLiberal says

    IMO, “the burden of proof” is an ambiguous phrase.

    One version of “the burden of proof” is a cultural rule that governs participants in the context of a conversation in the interest of achieving fairness. It’s the rule that says: If you make a claim in a conversation which is controversial, the onus is on you to cite evidence and sources and your justification, and specifically it is a disallowed move to tell the opponent to “go google it” because that would be an unfair imposition on the other person’s time.

    This version of “the burden of proof” is not a rule about epistemology. In particular, all participants in a conversation are morally obliged to use all of their available background knowledge; ex: it would be extremely silly to say “yes, I know that this premise is true, but I won’t accept it for the purpose of the conversation because you didn’t provide a citaiton”.

    Another version of “the burden of proof” is a rule of epistemology. It’s the rule that says: Ration your beliefs and levels of convinction according to your known evidence. Do not have beliefs and levels of convinction that are unsupported by your known evidence, and do not withhold belief and levels of convinction when your known evidence is sufficient. In other words, proper Bayesian epistemology.

    In particular, consider a discussion between a theist and someone who specifically adopts the position “I don’t know”. The theist and atheist should adopt proper Bayesian epistemology, which means that each should need evidenec and reason that justifies a belief before adopting that belief. Thus, the atheist’s first move in a “purely logical discussion” should be “I don’t currently have good evidence and reason to believe that there is a god. Do you have any?” If the theist tries to say “go google it”, that is not a violation of rules of epistemology. That’s a violation of morality and ethics, specifically a violation of cultural norms, specifically the cultural norm known as “the burden of proof”. This cultural norm “burden of proof” is basically entirely separate from epistemological “the burden of proof” of the atheist who needs evidence of a god before adopting a belief of the existence of a god. Two different concepts operating under the same name.

    In other words, the atheist could choose to “go google it”, and if they found sufficient evidence for a god, they would be obliged by the epistemological “burden of proof” to adopt a belief that there exists a god, and yet the theist would still be in violation of the cultural norm “burden of proof” by demanding that the atheist “go google it”.

    As for who “wins” an argument: I quit my formal debate club in high school for a reason. I’m not interesting in “winning” in that abstract sense. I’m interesting in having reliable and true beliefs, and I’m sometimes interested in the art of effective persuasion in order to make the world into a better place.

  5. quackepistemologist says

    Cline is wrong. The burden of proof is only ever on the claimant, not the believer. People can debate about anything they want, but no one incurs a burden of proof for simply believing God/gods exist

Leave a Reply

Your email address will not be published. Required fields are marked *