From discs in the sky to faces in toast, learn to weigh evidence sceptically without becoming a closed-minded naysayer

by Stephen Law

Stephen Law is a philosopher and author. He is director of philosophy at the Department of Continuing Education at the University of Oxford, and editor of Think, the Royal Institute of Philosophy journal. He researches primarily in the fields of philosophy of religion, philosophy of mind, Ludwig Wittgenstein, and essentialism. His books for a popular audience include The Philosophy Gym (2003), The Complete Philosophy Files (2000) and Believing Bullshit (2011). He lives in Oxford.

Edited by Nigel Warburton

10 NOVEMBER 2021

Many people believe in extraordinary hidden beings, including demons, angels, spirits and gods. Plenty also believe in supernatural powers, including psychic abilities, faith healing and communication with the dead. Conspiracy theories are also popular, including that the Holocaust never happened and that the terrorist attacks on the United States of 11 September 2001 were an inside job. And, of course, many trust in alternative medicines such as homeopathy, the effectiveness of which seems to run contrary to our scientific understanding of how the world actually works.

Such beliefs are widely considered to be at the ‘weird’ end of the spectrum. But, of course, just because a belief involves something weird doesn’t mean it’s not true. As science keeps reminding us, reality often is weird. Quantum mechanics and black holes are very weird indeed. So, while ghosts might be weird, that’s no reason to dismiss belief in them out of hand.

I focus here on a particular kind of ‘weird’ belief: not only are these beliefs that concern the enticingly odd, they’re also beliefs that the general public finds particularly difficult to assess.

Almost everyone agrees that, when it comes to black holes, scientists are the relevant experts, and scientific investigation is the right way to go about establishing whether or not they exist. However, when it comes to ghosts, psychic powers or conspiracy theories, we often hold wildly divergent views not only about how reasonable such beliefs are, but also about what might count as strong evidence for or against them, and who the relevant authorities are.

Take homeopathy, for example. Is it reasonable to focus only on what scientists have to say? Shouldn’t we give at least as much weight to the testimony of the many people who claim to have benefitted from homeopathic treatment? While most scientists are sceptical about psychic abilities, what of the thousands of reports from people who claim to have received insights from psychics who could only have known what they did if they really do have some sort of psychic gift? To what extent can we even trust the supposed scientific ‘experts’? Might not the scientific community itself be part of a conspiracy to hide the truth about Area 51 in Nevada, Earth’s flatness or the 9/11 terrorist attacks being an inside job?

Most of us really struggle when it comes to assessing such ‘weird’ beliefs – myself included. Of course, we have our hunches about what’s most likely to be true. But when it comes to pinning down precisely why such beliefs are or aren’t reasonable, even the most intelligent and well educated of us can quickly find ourselves out of our depth. For example, while most would pooh-pooh belief in fairies, Arthur Conan Doyle, the creator of the quintessentially rational detective Sherlock Holmes, actually believed in them and wrote a book presenting what he thought was compelling evidence for their existence.

When it comes to weird beliefs, it’s important we avoid being closed-minded naysayers with our fingers in our ears, but it’s also crucial that we avoid being credulous fools. We want, as far as possible, to be reasonable.

I’m a philosopher who has spent a great deal of time thinking about the reasonableness of such ‘weird’ beliefs. Here I present five key pieces of advice that I hope will help you figure out for yourself what is and isn’t reasonable.

Let’s begin with an illustration of the kind of case that can so spectacularly divide opinion. In 1976, six workers reported a UFO over the site of a nuclear plant being constructed near the town of Apex, North Carolina. A security guard then reported a ‘strange object’. The police officer Ross Denson drove over to investigate and saw what he described as something ‘half the size of the Moon’ hanging over the plant. The police also took a call from local air traffic control about an unidentified blip on their radar.

The next night, the UFO appeared again. The deputy sheriff described ‘a large lighted object’. An auxiliary officer reported five lighted objects that appeared to be burning and about 20 times the size of a passing plane. The county magistrate described a rectangular football-field-sized object that looked like it was on fire.

Finally, the press got interested. Reporters from the Star newspaper drove over to investigate. They too saw the UFO. But when they tried to drive nearer, they discovered that, weirdly, no matter how fast they drove, they couldn’t get any closer.

This report, drawn from Philip J Klass’s book UFOs: The Public Deceived (1983), is impressive: it involves multiple eyewitnesses, including police officers, journalists and even a magistrate. Their testimony is even backed up by hard evidence – that radar blip.

Surely, many would say, given all this evidence, it’s reasonable to believe there was at least something extraordinary floating over the site. Anyone who failed to believe at least that much would be excessively sceptical – one of those perpetual naysayers whose kneejerk reaction, no matter how strong the evidence, is always to pooh-pooh.

What’s most likely to be true: that there really was something extraordinary hanging over the power plant, or that the various eyewitnesses had somehow been deceived? Before we answer, here’s my first piece of advice.NEED TO KNOWTHINK IT THROUGHKEY POINTSWHY IT MATTERSLINKS & BOOKS

Think it through

1. Expect unexplained false sightings and huge coincidences

Our UFO story isn’t over yet. When the Star’s two-man investigative team couldn’t get any closer to the mysterious object, they eventually pulled over. The photographer took out his long lens to take a look: ‘Yep … that’s the planet Venus all right.’ It was later confirmed beyond any reasonable doubt that what all the witnesses had seen was just a planet. But what about that radar blip? It was a coincidence, perhaps caused by a flock of birds or unusual weather.

What moral should we draw from this case? Not, of course, that because this UFO report turned out to have a mundane explanation, all such reports can be similarly dismissed. But notice that, had the reporters not discovered the truth, this story would likely have gone down in the annals of ufology as one of the great unexplained cases. The moral I draw is that UFO cases that have multiple eyewitnesses and even independent hard evidence (the radar blip) may well crop up occasionally anyway, even if there are no alien craft in our skies.

We tend significantly to underestimate how prone to illusion and deception we are when it comes to the wacky and weird. In particular, we have a strong tendency to overdetect agency – to think we are witnessing a person, an alien or some other sort of creature or being – where in truth there’s none.

Psychologists have developed theories to account for this tendency to overdetect agency, including that we have evolved what’s called a hyperactive agency detecting device. Had our ancestors missed an agent – a sabre-toothed tiger or a rival, say – that might well have reduced their chances of surviving and reproducing. Believing an agent is present when it’s not, on the other hand, is likely to be far less costly. Consequently, we’ve evolved to err on the side of overdetection – often seeing agency where there is none. For example, when we observe a movement or pattern we can’t understand, such as the retrograde motion of a planet in the night sky, we’re likely to think the movement is explained by some hidden agent working behind the scenes (that Mars is actually a god, say).

One example of our tendency to overdetect agency is pareidolia: our tendency to find patterns – and, in particular, faces – in random noise. Stare at passing clouds or into the embers of a fire, and it’s easy to interpret the randomly generated shapes we see as faces, often spooky ones, staring back.

And, of course, nature is occasionally going to throw up the face-like patterns just by chance. One famous illustration was produced in 1976 by the Mars probe Viking Orbiter 1. As the probe passed over the Cydonia region, it photographed what appeared to be an enormous, reptilian-looking face 800 feet high and nearly 2 miles long. Some believe this ‘face on Mars’ was a relic of an ancient Martian civilisation, a bit like the Great Sphinx of Giza in Egypt. A book called The Monuments of Mars: A City on the Edge of Forever (1987) even speculated about this lost civilisation. However, later photos revealed the ‘face’ to be just a hill that looks face-like when lit a certain way. Take enough photos of Mars, and some will reveal face-like features just by chance.

The fact is, we should expect huge coincidences. Millions of pieces of bread are toasted each morning. One or two will exhibit face-like patterns just by chance, even without divine intervention. One such piece of toast that was said to show the face of the Virgin Mary (how do we know what she looked like?) was sold for $28,000. We think about so many people each day that eventually we’ll think about someone, the phone will ring, and it will be them. That’s to be expected, even if we’re not psychic. Yet many put down such coincidences to supernatural powers.

2. Understand what strong evidence actually is

When is a claim strongly confirmed by a piece of evidence? The following principle appears correct (it captures part of what confirmation theorists call the Bayes factor; for more on Bayesian approaches to assessing evidence, see the link at the end):

Evidence confirms a claim to the extent that the evidence is more likely if the claim is true than if it’s false.

Here’s a simple illustration. Suppose I’m in the basement and can’t see outside. Jane walks in with a wet coat and umbrella and tells me it’s raining. That’s pretty strong evidence it’s raining. Why? Well, it is of course possible that Jane is playing a prank on me with her wet coat and brolly. But it’s far more likely she would appear with a wet coat and umbrella and tell me it’s raining if that’s true than if it’s false. In fact, given just this new evidence, it may well be reasonable for me to believe it’s raining.

Here’s another example. Sometimes whales and dolphins are found with atavistic limbs – leg-like structures – where legs would be found on land mammals. These discoveries strongly confirm the theory that whales and dolphins evolved from earlier limbed, land-dwelling species. Why? Because, while atavistic limbs aren’t probable given the truth of that theory, they’re still far more probable than they would be if whales and dolphins weren’t the descendants of such limbed creatures.

The Mars face, on the other hand, provides an example of weak or non-existent evidence. Yes, if there was an ancient Martian civilisation, then we might discover what appeared to be a huge face built on the surface of the planet. However, given pareidolia and the likelihood of face-like features being thrown up by chance, it’s about as likely that we would find such face-like features anyway, even if there were no alien civilisation. That’s why such features fail to provide strong evidence for such a civilisation.

So now consider our report of the UFO hanging over the nuclear power construction site. Are several such cases involving multiple witnesses and backed up by some hard evidence (eg, a radar blip) good evidence that there are alien craft in our skies? No. We should expect such hard-to-explain reports anyway, whether or not we’re visited by aliens. In which case, such reports are not strong evidence of alien visitors.

Being sceptical about such reports of alien craft, ghosts or fairies is not knee-jerk, fingers-in-our-ears naysaying. It’s just recognising that, though we might not be able to explain the reports, they’re likely to crop up occasionally anyway, whether or not alien visitors, ghosts or fairies actually exist. Consequently, they fail to provide strong evidence for such beings.

3. Extraordinary claims require extraordinary evidence

It was the scientist Carl Sagan who in 1980 said: ‘Extraordinary claims require extraordinary evidence.’ By an ‘extraordinary’ claim, Sagan appears to have meant an extraordinarily improbable claim, such as that Alice can fly by flapping her arms, or that she can move objects with her mind. On Sagan’s view, such claims require extraordinarily strong evidence before we should accept them – much stronger than the evidence required to support a far less improbable claim.

Suppose for example that Fred claims Alice visited him last night, sat on his sofa and drank a cup of tea. Ordinarily, we would just take Fred’s word for that. But suppose Fred adds that, during her visit, Alice flew around the room by flapping her arms. Of course, we’re not going to just take Fred’s word for that. It’s an extraordinary claim requiring extraordinary evidence.

If we’re starting from a very low base, probability-wise, then much more heavy lifting needs to be done by the evidence to raise the probability of the claim to a point where it might be reasonable to believe it. Clearly, Fred’s testimony about Alice flying around the room is not nearly strong enough.

Similarly, given the low prior probability of the claims that someone communicated with a dead relative, or has fairies living in their local wood, or has miraculously raised someone from the dead, or can move physical objects with their mind, we should similarly set the evidential bar much higher than we would for more mundane claims.

4. Beware accumulated anecdotes

Once we’ve formed an opinion, it can be tempting to notice only evidence that supports it and to ignore the rest. Psychologists call this tendency confirmation bias.

For example, suppose Simon claims a psychic ability to know the future. He can provide 100 examples of his predictions coming true, including one or two dramatic examples. In fact, Simon once predicted that a certain celebrity would die within 12 months, and they did!

Do these 100 examples provide us with strong evidence that Simon really does have some sort of psychic ability? Not if Simon actually made many thousands of predictions and most didn’t come true. Still, if we count only Simon’s ‘hits’ and ignore his ‘misses’, it’s easy to create the impression that he has some sort of ‘gift’.

Confirmation bias can also create the false impression that a therapy is effective. A long list of anecdotes about patients whose condition improved after a faith healing session can seem impressive. People may say: ‘Look at all this evidence! Clearly this therapy has some benefits!’ But the truth is that such accumulated anecdotes are usually largely worthless as evidence.

It’s also worth remembering that such stories are in any case often dubious. For example, they can be generated by the power of suggestion: tell people that a treatment will improve their condition, and many will report that it has, even if the treatment actually offers no genuine medical benefit.

Impressive anecdotes can also be generated by means of a little creative interpretation. Many believe that the 16th-century seer Nostradamus predicted many important historical events, from the Great Fire of London to the assassination of John F Kennedy. However, because Nostradamus’s prophecies are so vague, nobody was able to use his writings to predict any of these events before they occurred. Rather, his texts were later creatively interpreted to fit what subsequently happened. But that sort of ‘fit’ can be achieved whether Nostradamus had extraordinary abilities or not. In which case, as we saw under point 2 above, the ‘fit’ is not strong evidence of such abilities.

5. Beware ‘But it fits!’

Often, when we’re presented with strong evidence that our belief is false, we can easily change our mind. Show me I’m mistaken in believing that the Matterhorn is near Chamonix, and I’ll just drop that belief.

However, abandoning a belief isn’t always so easy. That’s particularly the case for beliefs in which we have invested a great deal emotionally, socially and/or financially. When it comes to religious and political beliefs, for example, or beliefs about the character of our close relatives, we can find it extraordinarily difficult to change our minds. Psychologists refer to the discomfort we feel in such situations – when our beliefs or attitudes are in conflict – as cognitive dissonance.

Perhaps the most obvious strategy we can employ when a belief in which we have invested a great deal is threatened is to start explaining away the evidence.

Here’s an example. Dave believes dogs are spies from the planet Venus – that dogs are Venusian imposters on Earth sending secret reports back to Venus in preparation for their imminent invasion of our planet. Dave’s friends present him with a great deal of evidence that he’s mistaken. But, given a little ingenuity, Dave finds he can always explain away that evidence:

‘Dave, dogs can’t even speak – how can they communicate with Venus?’

‘They can speak, they just hide their linguistic ability from us.’

‘But Dave, dogs don’t have transmitters by which they could relay their messages to Venus – we’ve searched their baskets: nothing there!’

‘Their transmitters are hidden in their brain!’

‘But we’ve X-rayed this dog’s brain – no transmitter!’

‘The transmitters are made from organic material indistinguishable from ordinary brain stuff.’

‘But we can’t detect any signals coming from dogs’ heads.’

‘This is advanced alien technology – beyond our ability to detect it!’

‘Look Dave, Venus can’t support dog life – it’s incredibly hot and swathed in clouds of acid.’

‘The dogs live in deep underground bunkers to protect them. Why do you think they want to leave Venus?!’

You can see how this conversation might continue ad infinitum. No matter how much evidence is presented to Dave, it’s always possible for him to cook up another explanation. And so he can continue to insist his belief is logically consistent with the evidence.

But, of course, despite the possibility of his endlessly explaining away any and all counterevidence, Dave’s belief is absurd. It’s certainly not confirmed by the available evidence about dogs. In fact, it’s powerfully disconfirmed.

The moral is: showing that your theory can be made to ‘fit’ – be consistent with – the evidence is not the same thing as showing your theory is confirmed by the evidence. However, those who hold weird beliefs often muddle consistency and confirmation.

Take young-Earth creationists, for example. They believe in the literal truth of the Biblical account of creation: that the entire Universe is under 10,000 years old, with all species being created as described in the Book of Genesis.

Polls indicate that a third or more of US citizens believe that the Universe is less than 10,000 years old. Of course, there’s a mountain of evidence against the belief. However, its proponents are adept at explaining away that evidence.

Take the fossil record embedded in sedimentary layers revealing that today’s species evolved from earlier species over many millions of years. Many young-Earth creationists explain away this record as a result of the Biblical flood, which they suppose drowned and then buried living things in huge mud deposits. The particular ordering of the fossils is supposedly accounted for by different ecological zones being submerged one after the other, starting with simple marine life. Take a look at the Answers in Genesis website developed by the Bible literalist Ken Ham, and you’ll discover how a great deal of other evidence for evolution and a billions-of-years-old Universe is similarly explained away. Ham believes that, by explaining away the evidence against young-Earth creationism in this way, he can show that his theory ‘fits’ – and so is scientifically confirmed by – that evidence:

Increasing numbers of scientists are realising that when you take the Bible as your basis and build your models of science and history upon it, all the evidence from the living animals and plants, the fossils, and the cultures fits. This confirms that the Bible really is the Word of God and can be trusted totally.
[my italics]

According to Ham, young-Earth creationists and evolutionists do the same thing: they look for ways to make the evidence fit the theory to which they have already committed themselves:

Evolutionists have their own framework … into which they try to fit the data.
[my italics]

But, of course, scientists haven’t just found ways of showing how the theory of evolution can be made consistent with the evidence. As we saw above, that theory really is strongly confirmed by the evidence.

Any theory, no matter how absurd, can, with sufficient ingenuity be made to ‘fit’ the evidence: even Dave’s theory that dogs are Venusian spies. That’s not to say it’s reasonable or well confirmed.

Of course, it’s not always unreasonable to explain away evidence. Given overwhelming evidence that water boils at 100 degrees Celsius at 1 atmosphere, a single experiment that appeared to contradict that claim might reasonably be explained away as a result of some unidentified experimental error. But as we increasingly come to rely on explaining away evidence in order to try to convince ourselves of the reasonableness of our belief, we begin to drift into delusion.

Key points – How to think about weird things

1. Expect unexplained false sightings and huge coincidences. Reports of mysterious and extraordinary hidden agents – such as angels, demons, spirits and gods – are to be expected, whether or not such beings exist. Huge coincidences – such as a piece of toast looking very face-like – are also more or less inevitable.
2. Understand what strong evidence is. If the alleged evidence for a belief is scarcely more likely if the belief is true than if it’s false, then it’s not strong evidence.
3. Extraordinary claims require extraordinary evidence. If a claim is extraordinarily improbable – eg, the claim that Alice flew round the room by flapping her arms – much stronger evidence is required for reasonable belief than is required for belief in a more mundane claim, such as that Alice drank a cup of tea.
4. Beware accumulated anecdotes. A large number of reports of, say, people recovering after taking an alternative medicine or visiting a faith healer is not strong evidence that such treatments actually work.
5. Beware ‘But it fits!’ Any theory, no matter how ludicrous (even the theory that dogs are spies from Venus), can, with sufficient ingenuity, always be made logically consistent with the evidence. That’s not to say it’s confirmed by the evidence.

Why it matters

Sometimes, belief in weird things is pretty harmless. What does it matter if Mary believes there are fairies at the bottom of her garden, or Joe thinks his dead aunty visits him occasionally? What does it matter if Sally is a closed-minded naysayer when it comes to belief in psychic powers? However, many of these beliefs have serious consequences.

Clearly, people can be exploited. Grieving parents contact spiritualists who offer to put them in contact with their dead children. Peddlers of alternative medicine and faith healing charge exorbitant fees for their ‘cures’ for terminal illnesses. If some alternative medicines really work, casually dismissing them out of hand and refusing to properly consider the evidence could also cost lives.

Lives have certainly been lost. Many have died who might have been saved because they believed they should reject conventional medicine and opted for ineffective alternatives.

Huge amounts of money are often also at stake when it comes to weird beliefs. Psychic reading and astrology are huge businesses with turnovers of billions of dollars per year. Often, it’s the most desperate who will turn to such businesses for advice. Are they, in reality, throwing their money away?

Many ‘weird’ beliefs also have huge social and political implications. The former US president Ronald Reagan and his wife Nancy were reported to have consulted an astrologer before making any major political decision. Conspiracy theories such as QAnon and the Sandy Hook hoax shape our current political landscape and feed extremist political thinking. Mainstream religions are often committed to miracles and gods.

In short, when it comes to belief in weird things, the stakes can be very high indeed. It matters that we don’t delude ourselves into thinking we’re being reasonable when we’re not.

Links & books

The Atlantic article ‘The Cognitive Biases Tricking Your Brain’ (2018) by Ben Yagoda provides a great introduction to thinking that can lead us astray, including confirmation bias.

The UK-based magazine The Skeptic provides some high-quality free articles on belief in weird things. Well worth a subscription.

The Skeptical Inquirer magazine in the US is also excellent, and provides some free content.

The RationalWiki portal provides many excellent articles on pseudoscience.

The British mathematician Norman Fenton, professor of risk information management at Queen Mary University of London, provides a brief online introduction to Bayesian approaches to assessing evidence.

My book Believing Bullshit: How Not to Get Sucked into an Intellectual Black Hole (2011) identifies eight tricks of the trade that can turn flaky ideas into psychological flytraps – and how to avoid them.

The textbook How to Think About Weird Things: Critical Thinking for a New Age (2019, 8th ed) by the philosophers Theodore Schick and Lewis Vaughn, offers step-by-step advice on sorting through reasons, evaluating evidence and judging the veracity of a claim.

The book Critical Thinking (2017) by Tom Chatfield offers a toolkit for what he calls ‘being reasonable in an unreasonable world’.