Falsifiability and the Burden of Proof
“Not believe in Santa Claus! You might as well not believe in fairies. You might get your papa to hire men to watch in all the chimneys on Christmas eve to catch Santa Claus, but even if you did not see Santa Claus coming down, what would that prove? Nobody sees Santa Claus, but that is no sign that there is no Santa Claus. The most real things in the world are those that neither children nor men can see. Did you ever see fairies dancing on the lawn? Of course not, but that’s no proof that they are not there.”
—from the famous editorial “Yes, Virginia, There Is a Santa Claus”, published in 1897 by the New York Sun
Today’s post on critical thinking concerns a fundamental principle of rationality, yet one that many believers get wrong, to their detriment. That is the principle of burden of proof: the person who makes a positive claim has an obligation to support it if they wish others to believe them. It is not the responsibility of others to prove that claim false. Closely related to this is the principle of falsifiability – that is, what sorts of claims it is possible to prove or disprove. This post will discuss both of these ideas.
Even today, there are many who misunderstand the burden of proof. A Christian made the following comment to me last month in an e-mail:
And certainly, unless you have visited an appreciable portion of the cosmos and whatever else, there is at least a reasonable possibility that a creator exists.
My correspondent felt this was a compelling reason not to be an atheist. But this claim has the burden of proof backwards.
Evidence is the sole link to truth. If a person makes an extraordinary claim, we don’t have to search the whole universe to disprove it – we just have to ask that person how they know that, what facts they have that led them to believe in that way. If they can’t produce such facts, then we’re justified in concluding that they don’t know what they’re talking about, and there’s no reason to believe them.
This is why the person who makes the positive claim has the burden of proof to support it. Since we’re not justified in believing something for which we don’t have evidence, a person who claims to be justified in their belief is under the obligation to present their evidence. By contrast, a lack of belief in a given proposition requires no positive justifying evidence, only the absence of evidence to the contrary. Therefore, it is the default position, and does not incur any special burden of proof.
This basic fact is widely recognized in human inquiry. In law, it takes the form of the bedrock legal principle “innocent until proven guilty”. At trial, the prosecution – the side that makes the accusation of guilt, which is the positive assertion – has the burden of proof to support their claim. Similarly, in science, the “null hypothesis” is the default assumption. A researcher who defends the positive claim of a causal connection between two phenomena is obliged to present evidence supporting that hypothesis. Even in medicine, any diagnosis should be supported by observations of symptoms – not the silly claim that you can’t disprove a diagnosis of what the patient might be suffering from.
A proposition is falsifiable if there is at least one definitive test that could refute it. If there are no such tests – if, like Last Thursdayism, the proposition is compatible with any state of evidence we could possibly find – then it is useless to consider or believe. There are an infinite number of unfalsifiable propositions that can be imagined, many of them mutually contradictory. If we’re willing to give assent to ideas that cannot be verified by testing or evidence, then we’ll either be lost in a hopeless morass of confusion, or else choose arbitrarily and hope that somehow we’ve hit on a true idea out of the infinite number of wrong ones.
As above, evidence is the sole link to truth, and the only path out of this impossible epistemology. The existence of unfalsifiable claims further underscores why the person making the positive claim has the responsibility to support it. If it were otherwise, we’d be obliged to believe an infinity of mutually contradictory, unfalsifiable ideas – fairies on the lawn, dragons in the garage, teapots orbiting the sun, the universe created last Thursday, and many more – just because we can’t disprove them. Clearly, this is absurd. It’s futile to believe unfalsifiable propositions, and merely asserting that an idea can’t be disproven is pointless. The only propositions to which we should give our assent are those that are falsifiable and for which the burden of proof has been met.
Other posts in this series: