The Skeptics Society & Skeptic magazine

When I taught my Harvard course Rationality,¹ wrote my book Rationality: What It Is, Why It Seems Scarce, Why It Matters,² and hosted my BBC Radio and podcast series Think With Pinker,³ I got a good sense of people’s curiosity about human rationality. They don’t care so much about logical fallacies like Denying the Antecedent or statistical biases like the Gambler’s Fallacy or Base Rate Neglect. They want to know why, as Michael Shermer’s book title puts it, “people believe weird things.”⁴ So, I began one episode of the BBC series with this flourish:

Our species is smart enough to have plumbed the nature of the universe, life, and mind. But then why do so many of us believe in so much quackery and flapdoodle?

There are conspiracy theories, like the idea that COVID-19 was a subterfuge by Bill Gates to implant trackable microchips in people’s bodies, that Donald Trump fought a cabal of Satan-worshiping pedophiles in the “deep state,” or that jet contrails are mind-altering drugs dispersed in a secret government program.

There’s fake news, such as Joe Biden Calls Trump Supporters ‘Dregs of Society’ and Yoko Ono Had an Affair With Hillary Clinton.

And there’s paranormal woo-woo, such as beliefs in possession by the devil, extrasensory perception, astrology, and spiritual energy in mountains, trees, and crystals.

How can we explain this pandemic of poppycock?

After the episode aired, this admittedly purple prose drew objections from the biologist Rupert Sheldrake and the Nobel prize-winning physicist Brian Josephson.⁵ Belief in extrasensory perception (ESP) does not belong in that list, they said. There is evidence from everyday experience and controlled experiments that dogs can sense when their owners will return, people can sense when they’re being stared at from behind, and, in famous studies done or inspired by the social psychologist Daryl Bem, people can “feel the future,” by predicting where a random algorithm will display an erotic photo on a computer screen or improving their performance in a memory test by rehearsing the words after the test has been given.⁶ Josephson directed me to a 2022 talk delivered at Hereticon (“A Conference for Thoughtcrime”) by the writer Mitch Horowitz called “Case Closed: ESP Is Real.”⁷ To dismiss ESP in the face of this evidence, they asserted, was itself irrational. Doing so flouted my own advisory to examine the evidence with an open mind and admit when one is wrong—which I should do, they said, in a public retraction on the BBC.

Josephson and Sheldrake are certainly correct that belief in unpopular, controversial, or even mistaken ideas is not in itself irrational. For a belief to be irrational, it must flout a normative canon of reasoning, such as a law of logic or principle of probability. But by these standards, I maintain, belief in ESP really is irrational, given everything we know about the world and how to reason about it.

* * *

The normative canon in question has been stated in many ways, most precisely in the eponymous theorem of Thomas Bayes: one’s degree of credence in a hypothesis should be estimated by starting with the prior probability of the hypothesis—how well supported, credible, or plausible it is before looking at the evidence—multiplied by how likely the evidence would be if the hypothesis is true, scaled by how common that evidence is across the board (i.e., how easy it is for that evidence to turn up whether the hypothesis is true or false). It was expressed in different words by Bayes’s contemporary David Hume: “No testimony is sufficient to establish a miracle, unless the testimony be of such a kind that its falsehood would be more miraculous than the fact which it endeavors to establish.” It was stated more pithily by Pierre-Simon Laplace: “The weight of evidence for an extraordinary claim must be proportioned to its strangeness.” And it was boiled down to five words by Carl Sagan: “Extraordinary claims require extraordinary evidence.”

In the case of ESP, the empirical case would have to be stupendous to outweigh the overwhelming prior odds against ESP existing. All of our experience, and all of our understanding of the physical universe, speak against both the possibility of the future affecting the past, and against an ability to sense the state of the world without the transmission of information by physical energy. If ESP really existed, not only would the laws of physics have to be overturned, but life would be unrecognizable. The knowledge afforded by telepathy or precognition could easily be exploited to bankrupt casinos, undermine poker tournaments, and make fortunes in financial markets, making them obsolete. Experts who developed their ESP would be snapped up by the intelligence community, weather forecasting, and the criminal justice system. This is not the world we live in.

Of course, we can’t give precise numerical estimates to our prior credence in the existence of precognition or telepathy, but as the psychologist Eric-Jan Wagenmakers and his colleagues point out, even rough napkin estimates show what the hypothesis of ESP is up against.⁸ Consider a flawless, well-designed experiment that provides strong evidence for ESP—data, say, that are 19 times more likely if ESP exists than if it does not. If our prior belief is very low, say, 00000000000000000001, then the experiment requires us to update it to a posterior belief of 00000000000000000019—far higher, yes, but still leaving us pretty confident that it does not exist. More charitable priors, up to any reasonable estimate, still require extraordinarily strong evidence—far stronger than shown in any experiment or meta-analysis.

* * *

To be fair, one can challenge the priors. As Josephson pointed out to me, the fact that Wall Street hasn’t been destroyed by clairvoyant stock pickers may show only that ESP is rare, not that it doesn’t exist at all.⁹ As he put it, “Crossing a valley on a tightrope is possible, but most people would end up in the valley if they tried. Some people may do well in the stock exchange through precognitive abilities that most people don’t possess for example.”

But in fact it would take only a few percentage points of accuracy above the base rates to make a killing in Las Vegas or in the stock, commodity, or currency markets. (Imagine, for example, if instead of anticipating whether some porn will appear on the left or right of a computer screen, a psychic trader predicted whether a stock price will go up or down, and sold long or short accordingly.) Is there even a single person who has gotten rich—or for that matter made a living— from the markets through precognitive abilities?

It’s also true that that many phenomena in physics challenge common sense, so our intuitions about what is physically possible can’t be taken at face value to set a Bayesian prior on ESP. But that doesn’t mean we can point to some counterintuitive physical phenomenon from quantum mechanics or relativity and conclude that any weird thing is possible. This is what Horowitz did at the end of his talk when he invoked time dilation near the speed of light in special relativity to explain how a student might effectively study for an exam after it’s over. To call this “physics for poets” would be a disservice to poets.

But should we go meta, and adjust the priors to acknowledge the possibility that our understanding of physics is incomplete and yet-to-be-discovered laws might explain how psychic powers are possible? Though many phenomena at extreme scales of space and energy—near the Big Bang or a black hole, at the size of a photon or of a galaxy—are incompletely understood, this cannot be said about the physics of everyday life. As Sean Carroll shows in The Big Picture, on these scales, from nanotech to moon rockets, the laws of physics are completely understood. We aren’t in need of strange new forces or fields to explain how a bicycle works, or why eclipses happen. Carroll takes the argument a step further: our understanding is so complete that if there were as-yet unidentified fields in addition to those underlying gravity, electromagnetism and so on, we would be able to detect them, and we don’t.

Of course, one can always go meta meta, and concede that maybe our entire understanding of physics, including its implications for what we do and don’t understand and may or may not discover, is fallible, and that we can never rule out yet another scientific revolution. But does anyone really think that the most radical revolution in human understanding in four centuries will be fomented by the claim by dog owners that their pets can sense when they will come home? Or is it more likely that the claim is fishy?

* * *

Which brings us from the priors to the evidence. There are many reasons to suspect that the evidence adduced for ESP can be explained by more mundane causes, inflating the denominator of Bayes’s fraction and yanking the posterior down still further.

To begin with, we’re talking about a phenomenon that, if it existed, would be tiny in magnitude, on the order of one tenth of a standard deviation.¹⁰ Moreover, the instances of “paranormal” phenomena are, by definition, identified post hoc, as anything and everything we can’t explain by what’s “normal.” They are a miscellaneous collection of oddities and anomalies rather than a systematic phenomenon whose conditions and outcomes are identified a priori. Small effects identified after the fact look suspiciously like random happenings that no one has the patience or resources to get to the bottom of.

Also, the classic claims for ESP in controlled experiments cited by Horowitz, such as those of J. B. Rhine and his intellectual descendants, have been exposed as artifacts of investigator bias, leakage of information, selective reporting, overinterpretation of coincidence, questionable research practices (such as post hoc data exclusion), and outright fraud.^{11, 12, 13, 14}

The temptations to engage in such chicanery, conscious or unconscious, are obvious. From time immemorial people have devoutly wanted paranormal phenomena to exist, as we see in the miracles recounted by the world’s religions. These desires may be motivated by a longing for immortality, for contact with the dead, for a realm in which divine justice might be meted out, and for an expansion of the horizons of consciousness.

Finally, there are the notorious Bem experiments. The 2010 originals were shown to be tainted by questionable research practices, and the first published replication attempts failed.¹⁵ Horowitz’s “closed case” thus rests on Bem’s own meta-analysis of subsequent studies, most of them unpublished, and itself announced in a self-publishing platform rather than a conventional peer-reviewed journal.¹⁶ This meta-analysis was then examined by the psychologist Daniel Lakens in his article “Why a meta-analysis of 90 precognition studies does not provide convincing evidence of a true effect.”¹⁷ Contradicting Horowitz’s summary of “90 replications in 33 labs,” Lakens notes that “only 18 statistically significant precognition effects have been observed in the last 14 years, by just 7 different labs…72 studies reveal no effect.” He concludes: “The results are clear: the estimated effect size when correcting for publication bias is 0.008 [in standard deviations], and the confidence intervals around this effect size estimate do not exclude 0. In other words, there is no good reason to assume that anything more than publication bias is going on in this meta-analysis.”

I conclude: when we follow the normative standard of Bayesian reasoning, use the facts and physics of everyday life to estimate the prior probability of paranormal powers existing, and consider the myriad ways in which the evidence claimed in their support could have other explanations, we must conclude that there is no good reason to believe ESP exists, and that insisting that it does is irrational. That is why I have not asked the BBC for airtime for a retraction.

———

Pinker and the Paranormal
Some Critical Comments

Brian D. Josephson Responds to Steven Pinker

In a talk in his BBC Radio 4 series Think with Pinker, Steven Pinker asked “why do so many of us believe in so much quackery and flapdoodle?”, characterizing extrasensory perception as “paranormal woowoo”.¹ I can imagine such language slipping out in the course of casual conversation, but on the BBC, in a programme where the text must have been carefully thought out in advance?

Something must have led to this being said in such an uncritical manner. So, I thought I’d email Pinker to find out what had led him to speak in this way in regard to the paranormal. In response he came up with two arguments.

The first has, at first sight, a degree of plausibility, and is the following: if there really are people with the claimed paranormal abilities, they could use these to win consistently at betting, and we would learn about that. However, as described in a recent Guardian article,² it seems this does not happen, because when such people start to win significant sums of money the bookies take note, responding to the threat that they pose by imposing limits on how much they are allowed to bet. As a result, we cannot safely infer that there are no people who can use their paranormal abilities to win large amounts at betting.

What about the second? In an email Pinker wrote:

When in my book Rationality I cite Sean Carroll’s arguments in The Big Picture³ in support of the claim that ESP is incompatible with the laws of physics, this is not an argument from authority.

That may be so, but the fact is that Pinker’s position presumes the validity of Carroll’s analysis. Was that analysis valid in fact? In response to my asking for more detail, this was his response:

It starts from the commonplace observation that in everyday phenomena at humanly relevant scales, from nanotech to moon rockets and everything in between, the laws of physics are completely adequate. We aren’t in need of strange new forces or fields to explain how my bicycle works, or why eclipses happen. Carroll’s argument, explained in the book, is that the understanding is so complete that we can go one step further, that current laws predict if that if there were as-yet unidentified fields in addition to those underlying gravity, electromagnetism and so on, we could be able to detect that, and we don’t.

This article appeared in Skeptic magazine 27.2
Buy print edition
Buy digital edition
Subscribe to print edition
Subscribe to digital edition
Download our app

Is that a good argument? No. It happens all the time in science that people consider, on the basis of the evidence available at the time, that they have a good understanding of some particular state of affairs. But then something new comes up that doesn’t fit the existing scheme, and as a result the models have to be adjusted to take them into account. Future physics just cannot be predicted on the basis of the past in this way. Furthermore, physicists derive their laws by studying situations where some model of concern is easy to test, and this tells us little about the general situation. Thus the idea that we can have “complete understanding” of nature in a particular domain is a misconceived one.

Pinker is far from being the only one to dismiss the paranormal on the basis of inadequate reasoning in this way. In part this is the outcome of this kind of reasoning being promulgated by organizations such as the Skeptics Society and this magazine, whose Bayesian approach presumes with inadequate support that there are genuine reasons for regarding the paranormal as essentially impossible. Where such reasons cannot be produced, the application of the Bayesian method amounts to little more than an assertion “I don’t believe it, therefore it is not true.”

The QAnon organisation, critiqued by Gabriel Gatehouse in his BBC series The Coming Storm,⁴ propagates its “truths” in a rather similar way, a situation highlighted by aphorisms such as Paul Simon’s “a man hears what he wants to hear, and disregards the rest” (The Boxer), or The Coming Storm’s “He or she who drives the narrative drives the outcome,” and Marshall McLuhan’s “the Medium is the Message.” We deserve better.

———

Steven Pinker Responds to Brian D. Josephson

by Steven Pinker

I thank Professor Josephson for his reply. But the fact that the gambling industry clamps down on successful sports betters cannot explain why we don’t see anyone using psychic powers to make a killing from gambling or investing. Outcomes in sports are not random. Successful betting on cricket or the horses depends not on “feeling the future” but on deploying information that is unavailable or imperfectly factored in by the oddsmakers. The Guardian article made it clear that the winners penalized by the bookies enjoy their edge through clever but decidedly ordinary means: seeing outcomes on TV or at the venue seconds before it gets to the bookies; bribing trainers or stable hands for inside information; arbitraging imperfectly estimated odds. If the casinos needed to handicap uncanny winners in truly random gambles like slot machines or lotteries, that would be different. And of course, financial markets, where ESP would reap billions, don’t handicap winners at all.

Josephson is welcome to dispute his fellow physicist Sean Carroll on whether our current understanding of physics rules out undiscovered fields that apply at the scales of everyday life. But it’s a cherry on top of the Bayesian argument, which requires only that psychic powers are incompatible with the laws of physics as we now understand them. Carroll, mindful of the dangers of premature triumphalism, tops it off with a carefully reasoned argument why those laws are almost certainly not going to be replaced by something completely different. But even if his meta-argument is wrong, at present there is no reason to doubt those laws. Causes precede their effects; information is transmitted by patterns in matter and energy.