Do extraordinary claims require extraordinary evidence?
This particular catchphrase has spread like wildfire in atheist and sceptical corners throughout the Internet. It is easy to see why – at first glance, it has a certain plausibility to it.
But in practice, it is often used as a rhetorical club rather than as a method to find truth.
For example, David Deming has called attention to the misuse of the ECREE principle in science. While he is not against any use of the principle, he argues that its misuse can make science too conservative. He writes:
ECREE is frequently invoked to discredit research dealing with scientific anomalies or any claim that falls outside the mainstream. The statement is usually made without justification or explanation, as if the mere invocation were enough to stifle debate and disqualify any legitimate opposition.
Many principles that seem plausible at first glance collapse on further investigation. So let’s examine this and see what we find.
The first hurdle: what does “extraordinary” mean here? Usually it remains undefined. Let’s go through some possible definitions:
- Rare. This seems to make sense: if something rarely or never happens in our experience, we are more likely to be sceptical of it. We are sceptical of Bigfoot claims, but this would not hold if we encountered bigfoots every week. However, while individual rare things are rare, as a set rare things happen frequently. Donald Bradman had a rare talent, for example. Additionally, all things are rare if described in enough detail. While I have written plenty of blog posts, none of them have had the same number of words as this one.
- Unprecedented. If something is not known to ever have happened, we might be sceptical of new claims for it. But of course, new things do happen regularly. No one had landed on the moon prior to 1969; nevertheless, it still happened.
- Outside of the normal capacity of the cause of the effect. For example: the claim that “Julius Caesar killed someone by firing lightning from his fingers” would be an extraordinary claim, because human beings do not have such capacities. However, this definition would not include acts of God, for if God exists, he is perfectly capable of performing miracles.
- Completely unexpected, given our background knowledge. This would cover miracles, and also such things as the reappearance of long-extinct animals. This does make “extraordinary” entirely subjective, though. The moon landing would be extraordinary to a lost Amazonian tribe. Additionally, one could argue that claiming background knowledge of places and times distant to ourselves is an unjustified extrapolation.
The next question: what counts as extraordinary evidence? Here are three possible definitions of that:
1. Quantity. The same kind of evidence, except more of it.
2. Thoroughness. In some circumstances, one might do only a cursory check to satisfy ones curiosity, in others one might use a fine tooth comb.
3. More spectacular. So if investigating whether Jesus rose from the dead, one might want evidence as miraculous as the claim. God might write ones name in the stars, for example. But this notion of extraordinary evidence is deeply problematic. Firstly, evidence for any large or complex claim usually takes the form of many small pieces of evidence, rather than one spectacular atom-bomb. Secondly, if you really do see God write your name in the stars, that’s only evidence that you should be in the nuthouse. And thirdly, there is a weird sense of entitlement involved. Who expects God to perform magic tricks for them?
How do we know things, anyway?
Think through your life, from your childhood, through school, talking with your parents, friends, reading books and newspapers. Throughout that time, how many claims have you heard? Many thousands. How many have you actually checked? A very small percentage, no doubt.
We are social animals. As Westerners, we think of evidence-gathering as a solitary, individual pursuit. But in reality, much of what we know we know via authority. As individuals we don’t see the evidence, and for anything specialised, we wouldn’t understand it even if we did.
In addition, many of our beliefs cannot be tested by evidence, yet we cannot do without them. These include the adequacy of the human mind, the reality of causation, human agency and so on.
In short: for most claims, the normal amount of evidence we get as individuals is zero evidence.
To what extent does extraordinariness guide our decisions to look for evidence?
Ordinary claims are not interesting enough to check. Nor do we usually check claims just because they’re bizarre. Usually we dismiss them as crackpot and move on. If bizarre claims are held by people we respect, we are more inclined to respect them, although more often we might only raise an eyebrow.
The decision to investigate a claim, and the amount of evidence one might seek, is mostly goal-driven. For example, a claim like “I ate eggs for breakfast” you would take or leave, without checking. But if a suspect in a murder trial used “I ate eggs for breakfast” as an alibi, the police might spend a lot of time testing that claim, despite its ordinariness.
Apart from that, we are driven by cultural biases. It never occurred to me to check claims I heard in English class, history, mathematics or physics etc. But I did check my religion. Why? My culture told me to. It doesn’t tell agnostics they should consider getting religion, so by and large they don’t.
How much evidence should we look for?
With the previous considerations in mind, how to decide how much evidence to look for?
1. The first and most decisive consideration is: how much evidence would there be if the claim was true? This will differ from claim to claim. Evidence for “are there any Tasmanian tigers left” will be far less than the evidence for the theory of evolution, for example, no matter how thorough the investigation. Claims for which repeated experiments can be done will have much more evidence than one-off historical events.
2. A more thorough search for evidence is entirely appropriate for claims with important ramifications. But this will have limits. To expect more evidence than a thorough investigation of the claim can give is self-defeating.
3. Finally, we need to be aware of cultural biases and act to mitigate these in the course of any investigation. Unfortunately, the “extraordinary claims” test is often used to entrench our Western cultural biases rather than come to an objective assessment of the truth, which must be the aim of any rational enquiry.
2. Extraordinary Claims And Cumulative Case Reasoning, by Tim McGrew (video, semi-technical) – A look at the “extraordinary evidence” test and Bayesian methodology