
Sign up to save your podcasts
Or


Malcolm Gaskill, Emeritus Professor of Early Modern History at the University of East Anglia, is the author of seven books, including Witchfinders, Between Two Worlds, and The Ruin of All Witches.
In this week’s conversation, Yascha Mounk and Malcolm Gaskell discuss how premodern people understood the boundary between the natural and supernatural worlds, why historians are right—and wrong—to complicate the Enlightenment’s story of progress, and what the persistence of superstition tells us about human nature.
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: I’m really looking forward to this conversation, in part because I always look forward to conversations where I feel like I learn new things, and this is a topic I don’t know that much about. Obviously generalizing over different places, different times, different individuals—what is the nature of the supernatural beliefs people had for a lot of history before “the Enlightenment”, the rational age? And what do we tend to get wrong about it?
Malcolm Gaskill: In the modern age, we tend to draw sharp distinctions between what we think is rational and irrational, natural and unnatural. If we went back 500 years, we would find varieties of opinion, skepticism, disagreement, and degrees of credulity—perhaps more than we might expect from our ancestors. But we would essentially find a kind of consensus that seems alien to us today.
The consensus is that their world—their visible, tangible world—was much more porous to the unseen world. People didn’t always understand mysterious forces; in fact, a state of puzzlement or bemusement was probably most common. But there was an assumption that there was an invisible world beyond this one, and that there was a relationship between daily life, existence, and some kind of world that existed beyond. That was a basic assumption people had, and it shaped all sorts of engagement that we would broadly call religious, or even superstitious. Essentially it was a relationship between the seen and the unseen, the tangible and the intangible, and the sublime.
Mounk: Part of what characterizes our beliefs today is that we’re quite tempted to believe in all kinds of supernatural, superstitious things, but we check ourselves. When we see a dark shadow at night, we might think for a moment, “gosh, that feels like a ghost,” and then say, “ha, silly me—that’s obviously not what it is.” If we look into the sky and didn’t realize there was going to be an eclipse, we think, “wow, that’s a pretty spooky natural phenomenon,” and then we think, “let me check whether there’s some kind of explanation for this.”
If you didn’t have that kind of mental filter, presumably every time you see something scary out of the corner of your eye, or something happens in nature you don’t have an explanation for because you can’t Google it, you’d naturally come up with stories about what might have caused it.
Gaskill: It’s an evolutionary advantage, in many ways, to try to make sense of chaos. But that often means getting it wrong—looking for patterns in things that aren’t there. There’s a condition called pareidolia, which is really where you join up the dots to make a picture where no picture exists. The mind always imposes patterns on strange, totally natural phenomena that aren’t necessarily understood, and out of them makes stories that are meaningful.
Going back to our late medieval ancestors, they didn’t make those stories out of everything. There was an acceptance that there were natural phenomena—extensions of God’s creation that weren’t direct providences, just things about nature that weren’t completely understood—and there wasn’t a constant need to impose some weird supernatural interpretation on them. This comes into witchcraft accusations too, where people often assume our ancestors didn’t understand so many things about illness and disaster, so they blamed everything on witches. That turns out not to be the case. There is a more rational sense to our ancestors than we might imagine, given the way our benighted forebears were produced by Enlightenment philosophes and rationalists into the 19th century, who wanted to force this difference between them and us so as to make us feel better.
Mounk: Let me push you a little on that. I think there are different levels, or different kinds of description, of what rationality might entail. I certainly think our ancestors were about as smart as we are—they had very similar mental faculties, just less knowledge. There were many parts of the natural world they didn’t understand, and they were raised in a much more religious world, where, as you said, the boundaries between the visible world and the invisible world were much more porous. So I agree that they had reasoning faculties broadly similar to ours, and exercised them. It’s not clear to me that we can get from that claim to the idea that they were somehow less superstitious, or believed in the supernatural any less.
There’s a story I learned in graduate school—I’ll keep it brief—about an anthropologist doing fieldwork, I think in the mountains of Ghana. He wants to be taken to see the rainmaker, but has to spend many months there before they trust him. Finally they take him—he crosses a desert, climbs a dangerous mountain—and he asks, “how do you make it rain?” The rainmaker says, “I have a secret stone. If I turn it a particular way and say a particular incantation, most of the time the next day it rains.” He asks to see the stone, and the rainmaker shows him a child’s marble, saying there’s only one of these stones in the world, it’s unique, and so on.
We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
Subscribe now
Set Up Podcast
If you have any questions or issues setting up the full podcast feed on a third-party app, please email [email protected]
Later, back in the city, the anthropologist acquires another marble, asks to be taken back to the rainmaker, and says, “look what I brought you—here’s another one of these stones.” The rainmaker is astonished—he thought there was only one. The anthropologist asks, “well, do you think it’s going to work?” The rainmaker thinks for a moment and says, “I don’t know—shall we give it a try?”
This was presented to me by Richard Tuck, who was leading the seminar, as a response to a classmate of mine who was arguing that means-ends reasoning is just a Western construct. It isn’t—this rainmaker has means-ends reasoning. He’s not an idiot; he’s capable of intelligent thought, he has a concept of one thing having a causal impact on another, he even has an empirical mindset: let’s see whether, if I do this, it rains tomorrow or not. What he doesn’t have is a scientific method, and a separation of the visible world from a non-visible one.
So what I easily buy is that our ancestors reasoned critically about phenomena in the world, but were still so steeped in a totally different thought-world that there’s a stark distinction between mainstream beliefs today—setting aside the other kinds of superstitious and religious belief that still exist today, which we can talk about—and the mainstream world then.
Gaskill: It is. There are a couple of things in that. One is the critical thinking required to distinguish between the likelihood of something being totally natural versus the imposition of a demon or an angel.
The other is that it’s very situational. This is where, in a lot of the so-called post-secular scholarship about these phenomena, emotions come in. People’s beliefs—a bit like your story, which is embedded in a very specific situation, a dialogue, an encounter between cultures—don’t float in the air. It’s easy to talk about belief systems and sets of ideas, but a lot of the time these are lived practices. One of the things anthropologists would often do—I think they’re perhaps more sensitive now—but think about someone like Evans-Pritchard going out to the Azande in Africa in the 1930s: really, it’s about trying to reduce everything they believed to functions. He was right, up to a point—these beliefs were hard to describe in the abstract, but you could see them in what people did. They weren’t inferior forms of reasoning; they were extremely embedded in culture and practice.
Sometimes what these anthropologists found—going back to your story—is that the peoples they studied were so baffled by the idea that someone was trying to fit these things into some kind of intellectual template that they had a bit of fun with them. One of the things anthropologists were slow to understand is that these people had a sense of humor, and realized they could play along.
So if we think of these only as sets of ideas and beliefs—belief systems—we slightly miss the point. We need to think of them as embedded in real-life situations and in emotions. Some of our own ideas and feelings—I’ll speak for myself, I won’t speak for you—that we’d espouse confidently and publicly might waver a bit under emotional pressure. It goes back to the old “no atheists in a foxhole” idea. We see this throughout the 20th century: men in battle become more superstitious. So people who, in daily life, wouldn’t hold these ideas, had feelings connected to them that were latent but surfaced because of the particular emotional stresses of the situation. I do agree with you—it’s just another way of thinking about these ideas and beliefs, as practices, as feelings and emotions, rather than codifying them the way anthropologists often wanted to.
Mounk: That I buy. Let me say something more general for a moment. I think there’s a broader historiographical tendency to complicate the easy distinctions we draw, and sometimes I worry it’s a bit of a historian’s idée fixe—partly because you want to rebel against the generation before you, and partly because historians love when things are complicated. The Dark Ages, it turns out, weren’t so dark; things were actually quite good. Secularization—well, in fact, today isn’t really secular, the continuity is much bigger than the discontinuity.
I’m always torn about those, because I think there’s a caricatured view of the Dark Ages as a miserable time of benighted people having terrible lives, which misses a lot of what’s interesting, fascinating, and valuable about those centuries. At the same time, if you compare how long people got to live, the condition of servitude a lot of people endured, and the extent to which the knowledge of the ancient world had been lost—with the rediscovery of these forms of reasoning, the Renaissance, and later the Enlightenment, not yet having happened—something about the short formula “Dark Ages” is not entirely wrong.
It seems to me that when it comes to these questions of how similar or how different the world of belief is, in the era before the Enlightenment and perhaps before the Renaissance, there’s a similar dynamic. I agree that “these people were stupid, superstitious morons, and today we have these great enlightened people who would never fall for anything like that” is far too simplistic—first, because those were people with our natures and our intelligence, not stupid morons, human beings trying their best to make sense of their condition the way we do today, and second, because there’s a lot of weird belief and superstition, modern-day witch hunts and so on, around today too.
Subscribe now
So I buy all that. And yet something in me rebels against the idea that there are no distinctions at all. The faculties of the human mind have, by and large, not changed—the mixture of incredible intelligence and venality that characterizes human beings today also characterized us as a species 500 or a thousand years ago.
The other thing you’ve alluded to—I may be getting my history of religion wrong, but I think this was originally Feuerbach’s idea—is that we tend to ascribe to God the faculties we wish for ourselves, and perhaps, more broadly, that God fills a psychological need for us. When we’re really scared, we cling to superstitious practices and beliefs to reassure us. That was true 500 years ago, and it’s true if you’re in a foxhole today, and it’s true whenever the world is scary and you don’t quite understand what’s going on in politics or society. That seems true today too.
So I buy the similarities. And yet it does seem to me very different to have a world in which all the authorities, secular and ecclesiastical, interpret everything through religion and belief in the supernatural, where everything is imbued with symbolic importance—versus a world in which, much as our human nature might tempt us toward that, the authorities pride themselves on not doing it, and we have a scientific method.
Gaskill: There’s simplification in both directions. A grand narrative that’s politically inflected, polemical, even propagandist—think of fascist narratives of history in the 20th century, or communist simplifications of history—these are highly tendentious versions, simplifications with a particular end.
We can go in the other direction too—importing backwards the idea that Newton was “a scientist” unlike anybody else, or importing back the idea that most people were kind of benighted peasants who were all stupid. This comes back to context. Historians do like to complicate things, but that’s not a bad thing—good historians, anyway, try to get between those two extremes: that our ancestors are just like us, or that our ancestors are totally different. It’s largely situational.
When one looks at witchcraft trials, and at the behavior of juries, they’re very discerning. There are different layers and degrees of rationality, and a bedrock of belief they don’t really have the intellectual wherewithal to challenge. There used to be a debate—I know it’s still going on—about whether there were actual atheists in the Renaissance, because it was supposedly impossible to hold that thought, to really believe there was nothing beyond. I never bought that. I always thought it was possible.
Mounk: Tell us about some of those witch trials. This is obviously meant to be a caricature, but I tend to think of the Monty Python sketch—“she’s a witch, she’s a witch,” and immediately the crowd rushes to drown the person, and then she drowns, and I guess she wasn’t a witch. There’s this image of an obviously stupid, idiotic process that everybody gets on board with. You describe the reality as much more complicated and subtle than that. So what did a witchcraft trial actually look like?
Gaskill: Most of the time there were no witchcraft trials. There may well have been suspicions within a village, but the assumption in that Monty Python version is that as soon as you couldn’t explain something, or there was someone you didn’t like, you found out the nearest witch, and there was no legal process—you chased them to the nearest tree and they were hanged, and that was it. Everybody felt an emotional release, and then you got back to normal until the next cow or chicken died, and then you found somebody else, or someone else you didn’t like.
It’s so far removed from the truth. I’ve spent the last 30 years telling people their myths about witchcraft are not right—not because I’m a natural killjoy, but because it misses the opportunity to understand other ways our forebears’ mentalities worked. Typically, there is a bedrock of belief in witchcraft, but most people had no direct experience of it whatsoever. There’s a point where this falls off the written record and we can’t be absolutely sure, but the written records we do have—which are actually copious for the 16th and 17th centuries—give us a pretty good view of the number of witch trials: probably about 110,000 across the 16th and 17th centuries. That seems like a lot for a crime we’d think imaginary, but it’s far fewer than you’d often read. From the Enlightenment onward, the figures got bigger and bigger—they went up into the millions.
Mounk: Also, by the way, against a population far, far smaller than today’s.
Gaskill: Infinitely smaller. The number of people who actually lived—not just the population at one time, but across, say, 1400 to 1700—is absolutely enormous. The number of witch trials is really very small, and very variegated over time and place.
Mounk: 110,000 doesn’t strike me as very small. I don’t want to nitpick, but 110,000 is a lot.
Gaskill: It’s a lot for a crime you and I would agree was imaginary, that people couldn’t possibly have committed. But it’s much smaller than people often imagine. There’s a kind of illusion, I think, between the idea of the Holocaust in the 20th century and what was happening before that—the numbers creep up into the millions and it feels like a very organized thing, a kind of technocratic genocide, which the Holocaust, in the 20th century, really was. But the witch trials are much more patchy. There isn’t a big central authority driving it. Within nations, typically, the central authorities are the restraining forces. The really bad witch hunts happen not because of the strength of law, but because of its weakness in that area.
So it does go against the prevailing narrative we’ve inherited about what the witch trials meant. What those stories have typically done is made our ancestors look stupid, venal, and weak. Whereas actually—taking England, which I know best, though not exclusively—you really have maybe a thousand witch trials across two centuries. That’s far fewer than people assume. It always surprises people that a thousand is a lot, but it also surprises them that you’re going to get 500 convictions out of that—500 executions for a crime that didn’t exist seems like a lot. But in half those cases, jurors stood there and said, “we don’t know,” and didn’t want to hang somebody for a crime they weren’t absolutely sure of, because the evidence wasn’t strong enough for the conviction needed to send someone to the gallows.
Mounk: I don’t know how extensive the record is on the trials that led to conviction versus the ones that didn’t. If you were defending, what kind of argument would you make? Is the prosecution’s case something like: she cursed this family, somebody heard her curse this family—here’s the witness who says, “yeah, I heard her curse her family,” and then three days later the youngest child fell ill and died? And the defense would say—what would they say? “If she didn’t curse them, there’s no evidence of that,” or, “she was seen in the woods collecting dark materials,” and the defense says, “she would always go there to get firewood”? Am I imagining something too muddled?
Gaskill: You are. There is no defense—not before the 18th century. In fact, typically the accused didn’t even get to say anything, because they weren’t expected to speak on oath about themselves—they weren’t on oath. Sometimes they were allowed to say something, and they’d just say, “I never did any such thing.” But it’s not the modern adversarial system. Quite apart from anything else, it’s all over incredibly quickly. The speed of it, I think, sometimes works in favor of the accused, and sometimes it doesn’t.
Mounk: I’m trying to move away from my memory of the Monty Python scene—“she’s a witch, let’s burn her”—and you’re telling me, well, actually there’s a trial, and half the time—but the more I learn about it, the more I feel like I’m getting back to the Monty Python scene. Perhaps there’s—I don’t actually know when British judges started wearing wigs, perhaps that’s later too—but there’s a theatrical nature to the court proceedings. It sounds like there’s some random accusation read out, then a very brief process that doesn’t seem to involve presenting evidence in any structured way, and then the jury somehow comes to a decision. It sounds like a coin flip.
Gaskill: No, it’s not a coin flip. The English legal system is built on an amateur principle—it is, really, even into the 19th century. There’s a lot of resistance in English culture to professionalism of any sort. It’s one of the reasons it was hard to introduce the police force in the 19th century—people thought you shouldn’t pay good citizens to enforce the law.
Mounk: I thought you were going to say it’s one of the reasons Boris Johnson became prime minister.
Gaskill: That’s another area we could go into another time. Corruption is inherent, and they sought to limit it by not paying people to do things. The judges are the only professionals in the system—they’re based in London, they ride out regularly on circuits, and “deliver the jail”: all the accused criminals who’ve gathered there over the last six months are delivered from the jail. That’s why they rattle through the cases—not just witch trials, but whatever’s coming their way. A clerk reads what’s called a jail calendar, a list of names, and people are constantly brought through the dock. So there’s a kind of farcical, almost tragicomic amateurism to all this, but within it there’s this strange discernment, this unexpected skepticism and hesitation.
Subscribe now
I think what the jurors feel is an abhorrence at the idea that people might engage with the devil to harm their neighbors, and they may well hold that bedrock assumption that it’s possible. But when they see this before their eyes, I think they, and magistrates, often think, God, this is just neighbors at war, and really you should shut up and go home. We don’t know how many cases were brought before a magistrate—just some bloke interrupting his dinner, a bang on the door, two neighbors standing there, “what is it now?” We don’t know how often he said, “you two really do need to live with each other, and she’s not a witch, look at her, she’s just a pathetic old woman.” It’s a shame we don’t have that record, because I firmly believe that kind of critical thinking, in that particular situation, knowing the personalities involved, was prevalent. It may even have been typical.
Mounk: Right—“this person is just a troublemaker, they continually accuse people of all kinds of things, they’re just querulous, shut up and go back home.”
Gaskill: I really think so. It just doesn’t leave any records.
Mounk: They do have very common characteristics. There’s nearly always a huge tension between religious visions of the future. In southern Germany, it’s often conflict within very small city-states for religious and political dominance. It isn’t theological as much as what they would call themselves “confessional states”—who’s going to be in charge, and what is their confession? Are they Lutheran, Catholic, or what? What religious politics are they going to impose on the whole of their territory, and how is that going to define them? Within southern Germany, in the era of the Reformation, there’s this very intense conflict between these states, and within them, and that produces a kind of toxic energy, a toxic set of emotions, which results in an atmosphere of the possibility that these areas are under a kind of demonic attack, and that it’s existential to do something about it. That characterizes Salem, Massachusetts, in the 1690s, as well.
But it also tends to be that these witch hunts burn themselves out quite quickly. Then there’s a moment of reflection, a restoration of order, because witch crazes are seen as emblems of disorder and chaos. In some of those cases in southern Germany, there are so many people being sucked into this vortex of accusation that there’s a moment of clarity afterward—what on earth are we doing? There are even cases in early 17th-century southern Spain where local people think they’re doing the work of the Inquisition, because they feel the devil is abroad, and they start rounding up accused witches, penning them up, waiting to be congratulated by the officials of the Inquisition, who come round and say, “what on earth are you doing? Let them go.” And they’re rather shame-faced, and they do.
So the historians, I think, looking back on the crazes—and they are exceptional, the exceptions that prove the rule, the rule being restraint—look back on them as symbols of weakness, religious disorder, and weaknesses of governance. That’s what happened at Salem, and it happened in England in the 1640s. England only has one really intense witch hunt, and during it the characteristics were civil war, disruption of the normal law courts, a heightened sense of religiosity, the feeling that England had been backsliding in religion and that Puritans needed to come forward. Young men took matters into their own hands. All those things made that witch hunt possible, but it burns itself out very quickly. In the very first witch trial, in the summer of 1645, you already find dissenting voices in the courtroom—not saying witchcraft doesn’t exist, but saying, hang on, does that testimony really stand up? Can we really execute someone on that basis?
Witch finders—professional witch hunters—always aim to be charismatic. They want to go amongst people and say, look, you probably have your doubts, let me galvanize you, let me lead you through this. Witch finders work precisely because people are naturally hesitant—they’re not sure whether this is witchcraft or not, or they don’t have the confidence to go to law because they feel they won’t be believed.
Mounk: Very interesting. We’re going to skip over a couple of centuries very rapidly here—Enlightenment, age of science—finally we’re all good rationalists who would never engage in witch hunts, who have banished superstition to the weird fringes of our society, where people dabble in tarot cards or horoscopes. What’s wrong with that picture?
Gaskill: It’s such an attractive picture, and it’s a kind of fable we tell ourselves, because it makes us feel like we’re the children of the Enlightenment, able to push that stuff into our background. Of course, that arc of progress into the 19th century was hugely broken by the First World War. Symbolically, one of the things it said was that maybe the age of reason and progress, and industrialization, had led to this—and this is often characterized as a moment not just of slaughter, but of a kind of collective madness.
The First World War also created—well, the myth of a lost generation, but also millions of grieving families who couldn’t deal with their grief the way they might have in an age of enlightenment, an age when spirits and things had been laid to rest. This is where you get the huge rise in the popularity of spiritualism, and indeed in all sorts of religions and practices we’d consider New Age—Buddhism, for example, really takes off in the 1920s, because somehow the old patriarchal forms of organized religion and the churches that had so stood behind the armies of the First World War seemed to have gotten it wrong.
So there’s this disruption in the confidence in that arc of progress, and I think we’ve been living with it ever since—and of course it goes into the Second World War. I think we still want to tell ourselves—and, going back to your earlier point, historians aren’t always these loose relativists about everything who say, well, medieval medicine was just different. No, it wasn’t just different—it was worse. You can make those kinds of value judgments. Nobody wants to be operated on by a medieval surgeon consulting the phases of the moon before making an incision on the operating table. That’s not fun.
But at the same time, I think the things we’ve wanted to suppress in ourselves can surface at moments of extraordinary societal and emotional pressure, and the First World War was certainly one of those. There’s also something about class and education in this—not just the forced suppression of these things within ourselves. Large parts of Western Europe remained relatively unchanged in terms of people’s reliance on harvests, the rhythms of life, the reliance on the seasons. I used to live in a village north of Cambridge—I lived there in the 1990s—and in the 1920s, the beliefs in witchcraft and magic in that village, only ten miles north of Cambridge, were almost completely unchanged from the 17th century. Why? Because people were uneducated, and the rhythm of their lives hadn’t really changed at all. They had the belief in maleficent witchcraft, in counter-magic, the idea that livestock were under attack. It’s extraordinary, but there it was.
Mounk: Whereas today one might argue that the villages around Cambridge may well be rational, whereas the fellows and faculty of Cambridge appear to be less and less so—but that’s a different point.
Gaskill: Well, I do have things to say about that too. I’m not part of Cambridge University anymore, so I couldn’t possibly comment. I think the world has changed. But there is the capacity to act in a mad, collective, emotional way. To take an example like Brexit—Brexit wasn’t a rational thing. Going back to Boris Johnson, and David Cameron—there were lies told about what this meant and what was happening, which people believed because they were very emotionally attached to a version of Britain that had existed during and after the Second World War. That weakness was exploited. In a sense, it conformed to many of the characteristics of superstition.
Mounk: That’s where I get off board a little. It’s interesting—perhaps you’re seeing the glass half full and I’m seeing it half empty, or the other way around, I’m not sure which way to put it. I recognize that in our society today we have a real attraction to superstitious beliefs—people talking about horoscopes get enormous audiences on TikTok and Instagram, and there’s a kind of new spirituality among a lot of younger people. But there’s a question about how seriously people really, deep down, take it. Do they engage with horoscopes as a kind of fun indulgence, or do they actually let it guide their lives in the same way?
By and large, when we’re talking about Europe, at least, and probably North America, people don’t let horoscopes choose their partners, and they certainly don’t consult them before a doctor performs an operation. Perhaps if you meet somebody you like, you look at the horoscope and get a little titillation, thinking about whether you’re well aligned or not. Perhaps if you really believe in horoscopes quite a lot, and you’re already in a fight with your boyfriend anyway, and the horoscope says he’s no good for you, you use that as an internal reason to say, you see, I really should break up with him. But I think the number of people who fall deeply in love with somebody they’re excited about, but don’t go for it because the horoscope says they’re the wrong person, is probably quite limited.
I agree there are these various forms of spiritual and superstitious belief people have—perhaps they don’t want to live on the thirteenth floor of an apartment, or sit in the thirteenth row on an airplane, or they wake up on Friday the thirteenth and think, ooh, got to be a little careful today. But even for the people who engage in that, it’s a little at the edges of their lives—which I think is different from, as you were saying, the doctor who consults the phase of the moon before doing an operation.
And then there’s this kind of metaphorical application of all this to our political adversaries or enemies, which I think is quite different. First, I was against Brexit—partly because I think, for all its flaws, the European Union is on the whole something sensible, and because, having grown up on the Continent, being among other things a German citizen, and having gone to university in England, having studied history at Cambridge, the idea that I could no longer move to Britain whenever I wanted, and feel welcome there the same way, was personally upsetting to me. But I just think that saying the people who voted for Brexit were superstitious, in the same way that people who believed in witchcraft are superstitious, that they weren’t being rational, flattens a distinction between people’s political values—the ways they might care about sovereignty, or have genuine concerns about a democratic deficit in the European Union, or the sense that they don’t really have control over migration, and that there’s too much migration too fast, and so on. All of which you can disagree with—you can say a lot of people regret the vote for Brexit given that a lot of what was promised didn’t come true, you can say there were falsehoods cited in the campaign—but to simply dismiss it as “people voted for Brexit because they’re superstitious”?
Gaskill: No, I agree with everything you’ve just said. I don’t think voting for Brexit was irrational. I think some of the evidence put before people was misleading, and they bought into it uncritically because it tended toward the way they already wanted to go. The idea of getting out of Europe is just an old-fashioned, jingoistic, little-Englander, xenophobic attitude that goes back generations, and I think it just touched that nerve. My father voted for Brexit, my father-in-law voted for it—it’s not like it’s divided our families. I don’t think they’re stupid; these are clever people. And I think, with the change of government, we might even be steering back toward the possibility—we’re told a majority of the population now want to rejoin the European Union. That doesn’t make me think, “look, see, you idiots, you got it wrong.” I think it just means the emotional complexion of the Brexit and post-Brexit era has steered back in another direction.
Maybe the thing is we’ve had a lot of evidence in the last ten years about economic growth, or the lack of it, which we can actually make decisions on. I just don’t think there was the evidence base to leave the European Union, but there was a very strong emotional, nationalist feeling, and nationalism doesn’t really care a lot about facts—what it cares about is inflaming, empowering stories. That’s what I think we had. So I wouldn’t characterize it quite that uncharitably. Any Brexiteers listening in the United Kingdom—I hear you, I understand you, I don’t agree with you, and I hope you think that maybe it’s time to try it the way we tried it before, and that really wasn’t quite so bad.
Going back to something you said before we got into Brexit—the difference between the symbolic, the metaphorical, and the literal: going back to anthropologists, I think sometimes they wanted to find myth and metaphor in sub-Saharan Africa, whereas actually these things weren’t as metaphorical, or as literal, as the anthropologists wanted them to be. They could be either. People could perfectly well tell a story, some origin myth, that was just a form of entertainment. There was an expectation that things needed to be literal or metaphorical that got in the way, whereas really it doesn’t necessarily need to be either.
This idea of—call it paradox, or cognitive dissonance, or aporia, whatever you want to call it—is true in 17th-century England, and you could find it true among the Azande of Africa: that you can think two things at once, that they can coexist. There’s another anthropologist, Nils Bubandt, a Danish anthropologist who wrote a book called The Empty Seashell—an absolutely fantastic book, about an island called Buli in the Pacific, where witchcraft is very sinister there precisely because it’s absent. That’s the paradox—it’s not because there are witches everywhere, it’s because people sense them peripherally but don’t actually have any direct experience of them.
So, again, we think about the ancients, we think about the myths, and we think, well, that was their religion. It could be—but actually, the possibility is that these are just symbolic stories that communicate moral truths, not actually about how the gods created the universe, and they could hold those two things together simultaneously, or depending on the context. I think that’s one of the things that’s hard to grasp—that they were working in a different kind of framework. It wasn’t just the ideas themselves that were different or similar.
In the rest of this conversation, Yascha and Malcolm discuss superstition in the present day and the complexity of human behavior in times of collective madness. This part of the conversation is reserved for paying subscribers…
Advanced AI systems are rapidly transforming the ongoing cat-and-mouse game that cybersecurity specialists have been playing with shadowy opponents for years. The latest skirmish is a dramatic opening scene for a high-stakes thriller coming to your screens, whether you like it or not.
This story starts, as it should, with the scary parts.
Two days ago, elite computer security experts announced the discovery of a critical vulnerability hiding in plain sight—one embedded in PHP, the software that powers an estimated 70% of the world’s websites. This bug allows bad actors to execute programs on the target machines, which is to say, take them over and do with your data as they please. The discovery instantly set off a mad scramble to patch the problem (just getting underway), as well as a nefarious effort by bad actors to exploit the bug for fun and profit before it gets fixed.
Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal values wherever they are threatened, please become a paying subscriber today!
Key to the plot are firms with suitably obscure names—Wiz and Calif. Both firms are straight out of central casting. Wiz is a giant Google-owned company founded by Israeli military intelligence veterans to help major companies secure their computer systems; Calif is an elite boutique cybersecurity firm that employs an international cadre of hackers and is deeply integrated with the top AI labs to help them test their products for vulnerabilities.
Wiz and Calif were using different advanced AI tools to run tests on PHP when they each independently discovered a flaw that had been lurking in the software for more than a decade. While the bug itself had been spotted as far back as 2020, researchers had mistakenly thought that it could only cause a software crash. But the plot twist was that the AI tools were able to show it could be weaponized to take over the entire machine, making the threat widespread and urgent—something that they estimated would have taken a team of human experts months to discover on their own, if ever.
The fallout of this discovery is now unfolding. Companies that monitor systems for cyber-intrusions, such as Cloudflare, soon detected millions of attempts to weaponize the flaw. In the UK alone, 32 government services were immediately taken offline out of caution until they could fix the problem. Many more organizations and governments around the world have since followed suit.
So what’s really going on, and what does this say about the risks of the new wave of large-scale generative AI?
Our first Vancouver meetup will be on Friday, October 23 at 6:30pm at the Wicklow Pub! Please email [email protected] if you want to join.
It has now been nearly a decade since Donald Trump first became president of the United States. We have become so used to his uncanny hold over the American political system that it is easy to forget how shocking and unexpected it was at the time. As I argued ten years ago, still fresh out of graduate school, this failure was not limited to the journalistic class; social scientists were also caught badly off-guard.
Donald Trump now looks as though he might finally be loosening his grip over America. He is less popular than at any point over the last decade. He is likely to experience a severe shellacking at the upcoming midterms. Soon, attention will turn to the 2028 presidential primaries.
But our collective failure to anticipate Trump’s rise has recently been on my mind in a perhaps unexpected context: As social scientists grapple with the ways in which artificial intelligence will transform our economy and our democracy, they are falling back into some of the same errors that led them to miss the imminent rise of populism: At a moment of disjunctive change, they are uncritically looking at recent data to predict what might happen in the future.
—Yascha
In the wake of the Great Recession, economists faced tough questions: How could they have missed the imminent collapse of the world economy? Of what use were all their intricate models if they couldn’t predict such a huge systemic crisis?
Just as economists failed to predict the Great Recession, so political scientists failed to predict a series of momentous political upsets. The most obvious is the dizzying rise of Donald Trump. But the story is bigger than that: In recent years, political scientists have overestimated the forces of stability time and again, failing to foresee Brexit, the chaos wrought in the Philippines by Rodrigo Duterte, and the serious threat posed to Polish democracy by its populist government, among other developments.
To be sure, it would be too much to ask political scientists to have predicted each of those twists and turns. For one, academics aren’t in the forecasting business. For another, there is such a thing as contingency: When the unlikely comes to pass, even correct models of the social world fail to see around the next historical corner.
And yet it would be a mistake to quibble away the failure of scholars to recognize that, say, one of the most seismic American political events in a half-century might be on the horizon. A year ago, most political scientists would simply have dismissed the suggestion that somebody like Trump could win the Republican nomination and prove competitive in the general election. Worse still, the dominant theories in the field actually seemed to predict just about the opposite of what eventually happened.
In one of the most respected recent works on presidential nominations, for example, Marty Cohen and his co-authors argue that, as their title has it, The Party Decides. In their view, party elites use endorsements and campaign contributions to help favored politicians in primaries, effectively preselecting approved candidates. And yet an outsider with virtually no institutional backing put up stiff resistance against Hillary Clinton in the Democratic primaries, and an even greater outsider with even less institutional backing swept the competition aside to become the Republican nominee for the presidency of the United States. In short, reality upended what we thought we knew even in areas where dominant theories did seem to make clear predictions.
How could this have happened?
Over the last three decades, political scientists have made impressive strides in understanding how the political world functions, especially in periods of relative stability. American politics is a good case in point. Scholars of the judiciary have shown to what degree the past ideological alignment of Supreme Court justices predicts their decisions in future cases. Scholars of Congress, meanwhile, can anticipate with great accuracy how representatives are going to vote on a pending bill by looking to such factors as their party ID, their past votes, or their biggest donors.
Three interlocking methodological shifts account for much of that progress.
Reaction to the riots in France has followed a familiar arc: shock and disbelief, followed by dismissal and even denial.
The wave of demonstrations began last month at a school in a Paris suburb, where teachers and students went on strike over staff shortages. What began as a local protest quickly morphed into a nationwide movement fueled by seemingly everything that ails France, turning particularly violent last weekend as rioters stormed city halls from Marseille to Limoges and Lille.
In the aftermath, social media was awash with testimonials from Parisians insisting that the unrest was much ado about nothing. Yes, the scenes from Limoges, where an inferno engulfed the historic town hall, culminating in the collapse of its ornate belltower to cheers from onlookers, were dramatic. But they were only a snapshot, we were told.
The reaction to the storming of the Bastille in 1789 was no doubt much the same, at least until the revolutionaries started chopping off heads.
While the guillotine may be unlikely to make a comeback any time soon, the long-term consequences of what’s unfolding across France shouldn’t be underestimated, either for the country itself or for Europe.
The immediate trigger of the protests was student discontent with the state of schools and broader frustration over the cost of living. But what’s really at stake is France’s economic future.
This article is brought to you by American Purpose, the magazine and community founded by Francis Fukuyama in 2020, which is proudly part of the Persuasion family.
This is the latest edition of “Diamond on Democracy,” a series by American Purpose at Persuasion. To receive future installments into your inbox—plus more great pieces by American Purpose and Francis Fukuyama’s blog—simply click on “Email preferences” below and make sure you toggle on the relevant buttons. And while you’re at it, why not subscribe to Persuasion, the Bookstack podcast, and Yascha Mounk’s column!
Subscribe now
It is increasingly apparent that a Blue Wave is coming on November 3. President Donald Trump’s approval rating has fallen to an average of 37 percent and may now be as low as 33 percent. His supposedly solid wall of support within his own party is eroding. Even in most deep red states, his approval ratings are underwater. Democrats have an 8-point average in the national preference polls for Congress. Fox News is predicting that the Democrats will win a solid majority in the House, picking up as many as 22 seats. There is at least an even chance Democrats could win the Senate, too.
Are the Democrats prepared for this new opportunity—and the potentially much greater opportunity of reclaiming the White House in 2029? Three questions loom large: First, what can and must be done to repair the damage to democracy and minimize the danger that an aspiring autocrat can once again severely degrade if not destroy American democracy? Second, what will the Democrats do to respond to voters’ concerns about high consumer prices, diminishing job opportunities, decaying infrastructure, and unaffordability of housing and other key components of a middle-class life? And third, what will be the political approach of a new Democratic majority? Will they try to build new coalitions from the center out that win over some Republican members of Congress, or will they meet polarization with polarization by trying to force through legislation with narrow majorities, even if it requires eliminating the Senate filibuster?
Barring a 1932-style electoral landslide that is unimaginable in the current political environment, the winning hand for Democrats will be to widen their base and, wherever possible, seek broader coalitions for policy innovation and reform. Come January of next year, it will be very difficult to pass significant domestic policy legislation that Trump will not veto, but the passage into law in June of a broad new bill to encourage housing affordability (which Trump passively accepted without signing) shows that occasionally partisan interests can converge. That chance will come again soon when the Senate considers the Bipartisan American Affordability and Jobs Act (BAAJA). It offers the best opportunity in decades to streamline the paralyzing thicket of permitting and review requirements that prevent us building the energy infrastructure (including renewable energy) we need to grow the economy and make electricity more affordable.
Come 2029 Democrats will also have the opportunity—and the historic imperative—to use their new Congressional power to promote significant reforms to protect and strengthen American democracy. They should use these next two years to stake their case for reforms that can garner bipartisan support.
The key to near-term progress is not to overreach but to push for reforms that can win broad backing from the Congress and the American people. The democracy-supporting philanthropist Rachel Pritzker has recently drawn a telling lesson from the fate of a recent reform bill. In 2021, the “For the People Act”—a sweeping piece of political reform legislation extending and modernizing voting rights and accountability measures—narrowly passed the House on a party-line vote but failed to overcome a Senate filibuster, with not a single Senate Republican supporting it. It had many good provisions, but it failed for lack of bipartisan buy-in. If the Democrats win a trifecta (White House, House, and Senate) in 2028, they will face some pressure to lift the filibuster for voting rights legislation in 2029 to pass such a bill with a narrow partisan majority if necessary. But what can be done by one partisan Congressional majority can be undone by the next.
American Purpose at Persuasion is a registered nonprofit that relies on reader support to pay our staff and keep our content free for everyone. If you value our work and want to fight for liberal values wherever they are threatened, please become a paying subscriber today!
In repairing American democracy, the priority should be reforms that can be sustained across swings of the electoral cycle. By contrast, the more narrowly tailored Electoral Count Reform Act of 2022 was crafted by a coalition of senators from both parties, and it then won substantial bipartisan support in the Senate to become law. As a result, it will be much more difficult in the future for a hyper-partisan minority to challenge and overturn the results of a presidential election when the Congress meets on January 6 every four years to count the electoral votes.
So, what kinds of reforms might garner at least some support across the partisan divide in both the Congress and the public arena? Below is a possible list. As with the Electoral Count Reform Act, these reforms might pass if Democrats work from the center out to build support from both parties. Each of the reforms addresses the high level of disgust that the American people now feel with political corruption and abuse of power and privilege in Washington. Donald Trump is the worst example of these sins in American history, but Democrats should approach the task with some humility. Presidents like Clinton and Biden have also abused the pardon power, for example, and the two parties are now engaged in an ugly race to the bottom for partisan advantage in gerrymandering.
A Viable Agenda for Democratic ReformEnd partisan gerrymandering of congressional districts. Pass a clean bill with only two provisions. The first should ban the use of partisan criteria in drawing Congressional district boundaries (with the hope that the same principle and procedure would be used for state legislative redistricting). The second should ban the redrawing of Congressional districts after the decennial redistricting (following a census), unless ordered by a court. If this legislation passed in 2027, Trump might veto it, but it would lay down a marker for a future administration. It could pass in 2029 and spare the country another disgraceful race to the bottom in the post-2030 redistricting. And it would brand opponents of a clean reform measure as practitioners of the worst and most cynical form of politics.
Ban the president and his immediate family members from profiting from deals with foreign governments and government-linked enterprises while the president is in office. The proliferating grotesque conflicts of interest of President Trump and his family, including cryptocurrency deals with government-linked billionaires from the Gulf kingdoms, constitute a “pattern of a presidency for sale” that compromises our values and our national security. Many Congressional Republicans are privately appalled by this sleaze, and as Trump’s popularity continues to weaken in his final two years, some may become emboldened to support a forward-looking law that imposes a ban on future actions. Were Trump to veto such a bill, it would further damage his hemorrhaging popularity and the Republican brand.
Ban Congressional stock-trading. A bill to ban purchase of stocks by members of Congress, their spouses, and their dependent children passed the House but failed in the Senate this summer because of a cynical effort by Republicans to insert in the bill a national voter ID provision (which could make it harder for low-income, disproportionately Democratic voters to vote). Once again, Democrats should press for a clean bill, with nothing else attached.
Protect the Federal Reserve. Under current law, members (governors) of the Federal Reserve Board serve 14-year, non-renewable terms, and may only be removed by the president “for cause,” which the Supreme Court has interpreted to mean “inefficiency, neglect of duty, or malfeasance in office.” But this is now under challenge by President Trump, who is seeking to remove Fed Governor Lisa Cook for politically opportunistic reasons. As America’s central bank, the Fed is arguably the most important independent agency in the Federal Government. It thus needs robust protection from political interference. Its autonomy could be strengthened if Congress passed a law giving Fed governors the same security of tenure enjoyed by the U.S. Comptroller General (the government’s supreme auditor), who can only be removed from a 15-year-term through impeachment or by a joint resolution of Congress.
Secure the tenure of Inspectors General. Soon after taking office in 2025, President Trump summarily fired at least 17 inspectors general of Cabinet departments and federal agencies, whose job it is under federal law “to prevent and detect waste, fraud, and abuse.” Controlling corruption and abuse of power requires greater protection for their tenure in office, and ideally more neutral means of appointment. The U.S. Comptroller General is appointed by the president from a short list of nominees submitted by an imaginatively constructed bipartisan commission (composed mainly of members of Congress) and can only be removed from office for cause by Congress, through the means described above. It may be wise to adopt similar appointment procedures and protections for the 37 inspectors general that are currently appointed by the president.
Protect the career civil service. The creation in the late 19th century of a professional and politically neutral federal civil service was one of the great steps forward for good governance and the rule of law in the United States. Now it is under unprecedented assault in modern times by the Trump Administration, which has prepared a new regulation (“Schedule Policy/Career”) to enable it to politicize large swaths of the career civil service. According to Democracy Forward, “Schedule P/C” would convert thousands, potentially tens of thousands, of “career federal jobs into positions that lack any protections and whose incumbents can effectively be fired for any reason at all.” The way to avert this descent into a new political spoils system is to pass, as soon as possible in the new Congress, the Saving the Civil Service Act, a bipartisan bill that would prohibit this kind of wholesale reclassification of merit-based, career civil service jobs.
These are modest but meaningful steps that Democrats should push, in partnership with principled Republicans, to secure our democracy and really drain the swamp in Washington. But we can go further. For quite some time, I have believed that we should forget about constitutional amendments and just focus on legislation to restore and protect democracy. But after the low point we have reached under President Trump, we may now be at a moment that could permit limited constitutional amendments to constrain future presidential overreach.
In thinking about this high-stakes terrain, it is crucial to bear in mind political realities. A constitutional amendment can only be sent to the state legislatures for consideration if it is adopted by a two-thirds vote in each house of Congress. This imposes a high bar for bipartisan consensus (but does not require presidential approval). So, there is no chance of advancing a constitutional amendment that one party sees as depriving it of a political advantage. Hence, eliminating the Electoral College is simply off the table: a waste of time. But there is at least one constitutional amendment that might garner two-thirds support in the Congress in 2029 for a simple reason: Sobered by the abuses of Donald Trump, Democrats want to protect democracy for the longer run. And, if there is a Democrat in the White House, Republicans will have a new incentive to constrain a president of the opposing party.
Which is why my final proposal is:
7. A constitutional amendment to constrain presidential pardons. At a minimum, such an amendment should forbid the president from pardoning himself or his immediate family members. I would favor the creation of a bipartisan independent commission (perhaps to be appointed in equal numbers by the majority and minority leaders of the House and Senate) to approve all proposed presidential pardons.
Finally, the political climate in 2029 may be favorable for more ambitious institutional and even constitutional innovations to control political corruption and possible future presidential abuse of power in the United States. Today, the United States ranks only 29th in the world in Transparency International’s Corruption Perceptions Index. Since 2016, its rank has fallen by 11 places. The experience of two Trump presidencies has taught us that strong independent institutions, with clear authorities and mandates and insulation from political control, are needed to control political corruption.
We’re expanding our events offerings! Please check out our events page to join Book Club, Ask the Author, and Intellectual Bootcamp—and to watch recordings of recent events.
Persuasion events
There is no one formula, but the least corrupt democracies, such as Sweden, Denmark, and New Zealand, have two types of provisions that may be instructive. One is a standing professional and independent prosecutorial authority, such as the Serious Fraud Office in New Zealand, to investigate serious instances of corruption and bribery. Recent experience suggests the value and indeed imperative of a body that cannot be ordered to back off by a president or a group of congressional cronies. The White House and the Congress cannot be relied on to investigate themselves, and so long as the Attorney General is a political appointee, there will always be questions if not constraints about the independence of the Justice Department when it pertains to the president and his appointees. The second (as in Sweden) is an office of Ombudsman, appointed by the parliament, to receive and investigate citizen complaints and ensure that high government officials comply with the law.
The corruption of the Gilded Age produced a wave of institutional innovations and reforms to improve the quality of government, including (as noted above) a meritocratic civil service in 1883, the Interstate Commerce Commission in 1887, the Sherman Antitrust Act of 1890, and the Federal Reserve System in 1913. Now we are living through a second Gilded Age, with unprecedented presidential corruption and abuse of power. If the Democrats win back the presidency in 2028, they should privilege reforms to ensure that a future president can never again so easily corrupt and threaten our constitutional system. What’s more, they may find surprising allies among the Republican ranks.
Larry Diamond is a senior fellow at the Hoover Institution and the Center on Democracy, Development, and the Rule of Law, both at Stanford University.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
Long-time readers will know that since shortly after the Russian invasion in February 2022, Persuasion has been following the war through the voice of one Ukrainian, Kateryna Kibarova. Kateryna’s accounts of the ongoing horrors—from the first days, to her return home to Bucha, to surviving the most recent summer amid constant bombardment—capture the seemingly never-ending nightmare facing civilians far from the frontlines.
Today, she writes about the prospect of yet another winter where daily essentials are in short supply, and what life is like in the deadly age of drone warfare.
Thank you for being part of our community, and for continuing to follow Kateryna’s story.
—David, Senior Advisor.
My dad in Zaporizhzhia says the birds that used to fly in the sky have been replaced by drones, chasing people and picking them off one by one. They’re very difficult to shoot down. They attack minibuses, cars. Drones and missiles are launched towards Kyiv from Bryansk, from Voronezh, from Belarus. Kharkiv is also under attack. Jet drones, the large ones, now carry explosives, and they also have cameras that allow them to strike precisely.
When the authorities find the wreckage of these drones, they can see from the parts that they were manufactured yesterday or the day before. As soon as Russia makes them, they launch them. It is the same story with the missiles. We have missile attacks every day. They hit Nova Poshta very hard, and they are destroying our very large hardware stores. They are targeting the huge hypermarkets covering hundreds of square meters. There is no single epicenter for the strikes anymore
Only one more-or-less-functioning shopping mall remains, and it’s small. Zaporizhzhia used to be a region of a million people; the city has one of the longest avenues in Europe. Now, when you drive down there, windows in all the buildings are blown out. Some districts of Zaporizhzhia are already considered uninhabitable. The city is gradually dying.
I recently had surgery at a private clinic. I was referred to a very good doctor for my knees. He didn’t even need the MRI report to understand what he saw. But I’ll tell you this: the surgery cost me around $1,500. We don’t have insurance-based healthcare. We don’t have a system where, say, you can get insurance if you’re retired. You’ll go to a regular hospital. What grandmother or retiree with a broken leg will be able to afford surgery in private clinics if their pensions are $100 a month?
When you’re in a state of high alert, your nervous system is constantly wound up. Once you’ve relaxed, you realize that you’re helpless. When you’re in stress every day, in survival mode, these quiet days are perceived by your body as something incomprehensible. There are attacks all the time, up to 15 alerts a day, lasting an hour or two. You sit down at your desk, start working, and then the air-raid siren goes off, you see that something is flying toward your district.
I haven’t even mentally prepared myself for this winter yet. Donald Trump keeps saying, “I’ll end the war in three days,” but nobody reacts anymore. Right now, people understand that their primary task is to survive the upcoming winter, to save themselves, their children, their homes. Not everyone will be able to heat their spaces. None of us are electricians or knowledgeable about generators or batteries. One moment they’re running, but the next it all burns out. People install solar panels, but if a drone hits them, that’s it, the whole system is ruined.
You need to think through every situation three steps ahead. For example: how to get to work? They close gas stations during air raids. One morning I woke up, went out, and saw that I needed to fill up the car. But I couldn’t because the gas stations were closed. If there is no hot water in the summer you can wash yourself outside, set up some kind of improvised shower. But what about in winter? You don’t plan your next vacation; you plan how to survive the next day.
There is already a noticeable food shortage in the stores. We’ve had this situation for about a month now. At first there were no vegetables or fruits at all. Thank God there are open-air markets, but grain products and everything else are severely affected. I stocked up on toilet paper, laundry detergent, grains, and everything else, because for the first time in five years I fear that famine may strike. I see that the stores, especially in the Kyiv region, are empty. My mom called me recently from Zaporizhzhia saying they too have nothing.
Businesses are in a dire state because, first of all, there’s a labor shortage—a large number of people have left Ukraine. And with constant air raid alerts, you have no customers. Then there are electricity issues and that means money must be spent on generators, on fuel. Transport has already become significantly more expensive in Kyiv. The metro used to cost eight hryvnias; now it’s 35.
I’ll put it this way: my state of mind at the beginning of the war was completely calm. There were enough air defense missiles and everything was shot down. Now, practically nothing is being shot down because there’s nothing to shoot them down with.
I happened to be at home during a recent attack. There were crazy explosions. I thought the house would be blown apart because everything was shaking. It was eight in the evening, people were coming home, and a large number of people died. In a nearby village, people were simply burned alive in their houses. Nearby, a car tire manufacturing warehouse was attacked. Everything was burning, everything was black.
Subscribe now
I see mothers with children leaving Kyiv and going either abroad or to Western Ukraine. I don’t know whether it’s temporary, just for the winter, or forever. Children can’t go to school normally when they are in bomb shelters.
Some friends who went abroad on summer vacation came back with the determination to leave Ukraine. When you live constantly in this horror, and you travel into normal life where people can walk around at night—well, these friends decided that’s it, they’re leaving. It doesn’t matter what they’ll be doing. Here they are big shots, their jobs are good. There, they’ll just be doing cleaning work or something like that.
You only have one life. I won’t be 38 again. People leave, and now I kind of understand why. During the first and second year of the war, we were all hoping it would end soon. Now, I think the people who are leaving are making the right choice, because of the children. In the Obolon neighborhood a drone killed a two-year-old girl. I mean, it was just a regular Saturday. A mother and daughter were sitting in a café, they didn’t even hear the air raid siren go off. Many drones are equipped with shrapnel, they send out bursts of iron fragments.
So I don’t judge anyone for leaving. I think everyone’s choice right now is their own. Everyone must figure out for themselves how they want to move forward.
Kateryna Kibarova is a Ukrainian economist and resident of Bucha.
Translated from the Russian by Julia Sushytska and Alisa Slaughter. This transcript has been lightly edited for concision and clarity.
About the Translators: Julia Sushytska was born in L’viv and is a Resident Associate Professor in Comparative Studies in Literature and Culture at Occidental College. Alisa Slaughter is Professor of Creative Writing at the University of Redlands. They co-edited and translated A Spy for an Unknown Country and Both Sides Face East: Durable Words.
Follow Persuasion on X, Instagram, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
Subscribe now
James Marriott is a London Times columnist and the author of the Substack Cultural Capital, covering society, ideas, and culture. He is the author of The New Dark Ages: The End of Reading and the Dawn of a Post-Literate Society.
In this week’s conversation, Yascha Mounk and James Marriott discuss whether the decline of reading is as severe as the data suggests, whether television or social media is more to blame, and why this is bad news for democracy.
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: You say in the introduction to your book that there’s a new genre of people wringing their hands about the end of reading, but you are a little earlier to that genre than others, and your book is really quite a lot more ambitious than just pointing that out. But let’s start with a straightforward factual finding: how reliable is the evidence that we have really seen a genuine decline of reading, and how worried should we be about that?
James Marriott: I do think the evidence is at this point really reliable. As you say, the non-viral essay on the decline of reading is pretty much its own genre of magazine journalism now. There have been lots of these sorts of articles. I always like to point out now that everybody has done this. I think I wrote one of the earliest ones back in 2023. I wrote an article about the future of the humanities in UK universities and spoke to lecturers and professors teaching English literature and other humanities subjects and began to find the sorts of comments and observations that have been reported in The New Yorker, in The Atlantic, everywhere now.
Professors told me that they’d spent their careers assigning books like Middlemarch or Ulysses and had assumed that this was pretty much what you did as an English literature professor. You assigned your undergrads these great monuments of English literature, and they were finding that they could no longer rely on their students to read the books. It seemed to them that their students were basically unable to read the books. They just couldn’t get through them.
I and probably everyone else who’s written this sort of article, was conscious that “university students aren’t as clever as they once were” and “the young are disappointing” is a sort of ancient theme of cultural commentary, as people like to say, going as far back as Socrates. I do think this is increasingly backed up by evidence. I guess the largest-scale survey for the decline of reading is this American Time Use Survey of almost a quarter of a million Americans, which found that in the last 20 years, daily time spent reading, defined quite broadly including audiobooks and newspapers as well as physical books, has declined by 40%. There’s a similar figure in the UK: a third of adults say that they’ve given up reading in the last 20 years.
On top of this we have these, I guess, probably more reliable studies about not only the decline of time spent reading but the decline of reading skill. So, declining literacy, which means—I always feel obliged to point out—not that people are literally becoming illiterate, but that our skills of interpreting, analyzing, and comprehending text are declining. There was an OECD report at the end of 2024 which went around the world with this really alarming headline that literacy was declining or stagnating in most developed countries. A really remarkable finding, for developed countries with good education systems to be sliding back in literacy. That’s some of the headline data on it.
Mounk: Let me stick to the data for a moment, and then let’s go to the wider significance of this. I’m a little bit torn on that data. I think there’s some data that is quite suggestive, and there’s certainly some anecdotes that seem very telling. At the same time, I worry that we’re always tempted to fall into these panics about things that are getting worse, and then ten or twenty years later you realize that actually we didn’t really change that much. I have to say that I’m struck that my students don’t seem to have gotten worse. Now, I’m lucky to teach at a pretty elite university and have pretty motivated students, but a lot of the anecdotes that you hear are also from people teaching at elite universities who should also have pretty motivated students. I also have shifted towards assigning some exams as pen-and-paper, in-class exams, and I really thought, my God, we have an hour in a classroom setting, in an exam, to write something coherent—we’re not really used to doing that that much, it’s going to be pretty rough. I found that actually the quality of the essays we produced in that time was pretty good. I find they’re able to read things and engage with them at a pretty high level.
There’s some piece of evidence that cuts in the other direction. For example, it was assumed that book sales were depressed for all kinds of reasons over the last 20 years, including competition from ebooks and audiobooks, but also obviously other forms of media. But my understanding is that by and large, the book market has sort of held constant, surprisingly well. It hasn’t grown very much—it’s probably still not a great business to be in, don’t become a writer, I say to my Substack audience—but it really hasn’t declined as much as you might expect if reading had really declined to that kind of extent.
The other point here is that a lot of the things that are replacing traditional forms of literacy still involve literacy. People are now spending more and more time interacting with AI agents, and even though we can talk to them, and perhaps eventually AI will be some sort of hot avatar that is following you around the kitchen as you’re preparing your lunch, or as you watch a robot preparing your lunch, for now most of the way that people are interacting with AI agents is reading and writing. So I’m trying to steelman the case of a skeptic here. What would you say to those responses?
Marriott: Really good points. I think there are a few things to say in response. The first thing is the health of the publishing industry. I think it’s important to disentangle that from the number of books that are being sold. My understanding is that publishing has either, in terms of profits, stagnated or slightly declined over the last 20 years. It’s important to separate total profits in publishing from volume of books sold. What we find when we look into the disparity between those figures is that the volume of books sold has declined, but publishers have managed to maintain profits by bumping up cover prices. Books are more expensive relative to inflation than they were 20 years ago.
There was actually a report at the Frankfurt Book Fair, I think last week, that was one of the first concrete pieces of evidence of the publishing industry admitting that it has a problem. The headline was that in the EU, publishing turnover had declined in real terms 30% in the last 20 years, which is a pretty severe decline. One of the points that article was making was that some of the difficulties in publishing have been slightly masked by inflation and increased cover prices, and that there was an increase, not only in this sort of literacy stuff, but that it was beginning to affect the publishing industry.
We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
Subscribe now
Set Up Podcast
If you have any questions or issues setting up the full podcast feed on a third-party app, please email [email protected]
There are other factors that have kept the publishing industry a little more buoyant than it might otherwise be. One of them is that in most Western countries, certainly here in Britain, a lot of public libraries have closed, and public library usage has really—certainly in Britain—fallen off a cliff. I think a lot of people who would once have borrowed books from public libraries are now being driven to bookshops because that’s not an option. So I think publishing has been soaking up some refugees from public libraries, and that has kept the headline figures healthier than they might otherwise have been for a while.
“Big Library” is finally being defeated—it’s what we all wanted as book lovers. I guess the other thing would be, you could also look at the quality of books that are being sold, whether we believe that publishing is shifting high-quality text of the kind that idealistic people like me might hope they’re shifting, or whether books themselves are getting a bit simpler and a bit less of the kind of complicated, rich text that we might hope for. I think there’s evidence that that is happening. So that’s my brief overview on why things in publishing are more melancholy than the publishing PR machine might want us to believe.
Mounk: I really felt like I was playing devil’s advocate when I was saying that things are fine in publishing. That’s a tough case to make.
Marriott: No, it’s true. I think it certainly has been the case that, until recently, it seemed that publishing profits have not been tracking the supposed decline in reading. I think they are now, and I think that previous phenomenon is explicable.
The second point I would make, as to the utility of the data overall, is a general point that I actually wish I’d made clear in the book: you don’t have to believe in every single data point about the decline of literacy to believe in the overall story that’s going on here, which is a story about the shift from a culture whose emphasis is on the written and printed word to a culture which has an emphasis on the image and the moving image. The great book on this subject, by Neil Postman, Amusing Ourselves to Death, was the first polemic to sound the alarm, and I’m really just following in that book’s wake.
Neil Postman’s anti-television polemic, written in 1984—it’s striking that Neil Postman doesn’t cite a single piece of data for the decline of reading, the decline of literacy. We know now that he could have done; reading did decline when television arrived. But he was making, I guess, a wider and more vibey, but perhaps also, I think, more compelling and more watertight point, which is that it’s not necessarily that you can measure these things in data, but the emphasis of the culture is shifting. People are moving from a world in which they understood and comprehended the world through text to one in which they understood and comprehended the world through image. He would have said that images are more visceral, more powerful, easier to watch, and therefore they will tend to overwhelm text just because they have more immediate power. So that would be my attempt to steelman the case.
Mounk: Where do you locate the beginning of the end of reading? I grew up in the 1990s, I suppose—I remember, which tells you a lot about the kind of teenager I was, that I had some t-shirt that said something like “movies are art, theater is life, TV is furniture,” or something like that. Neil Postman identifies television as the enemy of literate society. But if you think television is the beginning of the end of reading, then it goes really all the way back to the 1950s and 1960s in the United States—the sixties, seventies, and eighties perhaps in Europe—which, by the way, is also coincidentally a period when universal literacy in many countries was just emerging. So the time period in which most people are literate and television isn’t yet the main form of entertainment is like seven years.
Now, it seems to me that really your concern is driven more by the internet and social media and short video, TikTok and Instagram, and perhaps podcasts and all of those kinds of things. But there’s this kind of tension between relying on Postman and the emergence of television for the theoretical and conceptual frame, and pointing to social media and so on as the thing that really is going to put an end to literate society. So are we going to be nostalgic for the world of television, where at least we had a unified culture and broadcast news as a kind of thing that gives us a unifying sense of the world, as we see in Britain? Was that actually part of a golden age, or was that already part of a fall from the golden age? And if so, when was the golden age?
Marriott: When you talk about the way that technology has disrupted literacy, you can find yourself becoming ludicrously nostalgic, and if you’re not careful you end up railing against radio. Neil Postman began his story with the electric telegraph, which he thought fragmented news and put us in a world dominated by useless information because you could just communicate too much. I think you can overdo it.
I think this story does begin with television, basically. The first really significant drops in literacy and reading are dated to the 60s and 70s. I think those declines—I don’t have the statistics at my fingertips, I’m afraid—but I think those declines, some of the data seems to suggest, may actually be more severe in terms of the loss of reading time than the declines we’ve seen with the advent of the smartphone in the last 20 years.
I think if there was a golden age of literacy, it probably occurred in the early 20 or 30 years of the 20th century. I think it was a short window before books and literature seriously had to begin to compete with radio and cinema. As to whether TV was a golden age, I think it’s mixed. I think one part of the story is that literacy and reading undoubtedly declined with the arrival of television—the drops are huge. It’s not hard to imagine why: when you have an incredibly entertaining screen in your front room, suddenly the most entertaining and diverting object in your house is no longer a book, it’s the TV in your front room, and understandably everybody was watching television.
Mounk: Some of the most striking studies are on the effect of having a television on the number of children women have—I think it’s the famous study in rural India—and it’s just like getting a television is the most amazing contraceptive. The average number of children goes down from something like eight to five the year the television is introduced. Part of it may be the spread of cultural values and things like that, but a lot of it is literally just that you now have an alternative form of entertainment to the thing that produces children.
Marriott: Yes, I think that’s absolutely true. The data for the way that TV changes social life, the way it changes friendships—Derek Thompson writes about all this very well—I think is fascinating and overwhelming, and should lead us to believe that it also had this huge impact on reading. It’s no surprise that it also has a huge impact on reading; it’s a kind of direct competitor to reading in a way that you wouldn’t necessarily expect it to be a direct competitor to sex.
Mounk: Derek points out, just for those who don’t know that line of writing, which is really interesting—for example, I think the number of barbecues in America has declined by over fifty percent, the number of dinner parties has declined by sixty percent, I’m making up these numbers sort of from what I remember as best—huge drops over the last twenty, twenty-five years. Perhaps that’s because everybody is sitting at home alone rather than socializing, and we’re doing that, and perhaps we’re also reading books, but it’s quite plausible to think that really the effect here is that it’s being eaten up by this entertainment time, which nowadays is no longer the television, but mostly online streaming, TikTok, Instagram, and other forms of social media.
Marriott: Yeah, absolutely. In my book I quote some of the observations made by social anthropologists in the UK going to people’s houses and looking at how working-class life was changing in Britain. There’s a really terrifying quote from a famous anthropological study done, I think, in the early 1960s: they go to a working-class house and everyone’s sitting around watching the TV, and the father makes some comment like, “It’s brilliant, we’ve got the television, we never have to go outside now,” or something sort of unintentionally dystopian and revealing.
As to your point about whether TV constitutes a golden age, I would say there are two things to say. I think TV sets us off on a route towards a visual culture—I think it’s the beginning of the decline of literacy. But I think it has two advantages over the world we’re currently in. The first, I think, is that television, especially in its early decades, was run according to values that were still very much shaped by literacy. It’s hard to say that every technological revolution instantly changes a culture in every single way—that’s true in some ways, and we can talk about these striking statistics about birth rates and socializing, but in some ways it takes a while for the values of the technology to be absorbed by the culture.
Subscribe now
If you look at a lot of especially factual TV programming broadcast by the BBC in the UK, it seems obvious, watching these programs, that this is a kind of television shaped very deeply by the values, the culture, and the assumptions of literacy—people talking in coherent sentences, debates that take a long time. There’s a famous debate about Britain’s membership of the EEC between Roy Jenkins and Tony Benn that goes on, I think, for more than an hour, and they get into economic statistics, and it’s sort of print-inspired in its mode of communication. A lot of early factual programming is less like the sort of stunt-driven, very wacky kind of TV that you’d get nowadays if you were trying to popularize an academic subject to a mass audience—it really is just a guy almost reading a lecture to an audience on television with some moving images in the background. I think the fact that that age of television was shaped by the values of print made it a better engine for democracy than modern 24-hour cable news is.
The other point to say is that a really fundamental thing about television is that it’s static. A TV sits in the front room of your house—you can’t take it on the bus with you, you can’t take it on the train with you—and if, in the 1980s and 1990s, you were bored commuting to work or bored in an airport, the most entertaining thing available to you in that situation was a book or a magazine. I think literate culture had a crucial place to cling onto in our culture. I think what the smartphone does that’s important and perhaps undernoticed is it just makes screens ubiquitous, and it wipes out all the other cultural space that literacy used to occupy—literacy has nowhere to thrive now.
Mounk: I remember getting on the subway, I think in New York, a couple of months ago, and seeing somebody with a newspaper—a physical newspaper. I remember thinking, wow, how eccentric, having a newspaper on the subway. But of course, it would have been the case that you look at the subway and half the people, more than half the people, are reading either a newspaper or a book or something like that. It’s funny how quickly that has happened—what used to be ubiquitous has come to be exotic in a certain kind of way.
I think you set out very nicely both the fact that this is happening and some of the broader effects it’s having. But again, there’s a bunch of think pieces that are going on about the decline of reading. You really make a more ambitious argument that is trying to explain what the historical role of literacy has been beyond the world of literature, and why, therefore, we should expect much more far-reaching consequences from these transformations. Tell us how it is that the spread of mass literacy and the role of the book has really shaped the things we take for granted about the modern world.
Marriott: Really, following people like Marshall McLuhan and Walter Ong, who are the pioneers of these arguments—our culture is shaped by text in ways that we’re so used to, being surrounded by books, that reading and writing seem so obvious to us, almost so banal to us, that we forget to understand how profoundly they’ve shaped our civilization. I’m here in the Times Book Cupboard, and I am literally surrounded by books.
My book is trying to assemble an argument that I personally don’t think is particularly controversial, which is—you can go into history and look at, I guess, the two pillars of my argument: the phenomenon of the Enlightenment and the phenomenon of modern mass democracy. I think those two stories don’t happen at all without literacy. One of the fundamental points that media theorists like Marshall McLuhan and Walter Ong would make is that there are certain forms of knowledge which basically don’t exist without literacy. It’s easy to assume that things like philosophy, complex narrative history, logic, and mathematics are innate forms of human knowledge and intelligence that spring out of our brains fully formed, and that any human community will be able to have these kinds of intellectual activity. They argue, I think very persuasively, and I really agree with them, that’s not the case at all.
Mounk: Why is that? Explain the reasoning.
Marriott: One of the great ways to prove the advantage of literacy over talking and oral culture is to have to explain your ideas on a podcast, and you find yourself doing a much worse job of it than you can on the printed page.
Mounk: It’s funny, I actually find the relationship between oral culture and the written word to be the other way around. Some people tell me, “I’m bad at writing,” and I’m always too polite to say it, but I always think, well, that just means you haven’t thought through this very carefully. The point actually is that when you’re speaking, you can jump over holes in your argument quite easily—you have a little rhetorical flourish, you distract the person you’re talking to, even yourself, with some kind of little joke, and then you go on to the next point. Whereas in writing, it becomes really obvious when one thing doesn’t follow from the other, and that’s where people start getting stuck—that’s where they’re like, “I’m a bad writer because I can’t put my idea on paper.” Well, that’s because you hadn’t actually thought through the idea.
So presumably part of the reason why we need writing, for example, for a really complicated mathematical proof, is that it’s just impossible to keep everything in our mind unless you’re really a savant, to a sufficient degree, to go back and double-check each step and make sure that you haven’t jumped in some way where you’re introducing an error into the proof or the equation.
Marriott: Yeah, I think that’s really well put. The phrase that Walter Ong uses is that writing allows us to externalize part of our minds. You can take ideas that, in an oral culture, before the arrival of literacy, only exist in your head, and you can put them on paper and look at your thoughts from the outside. You can submit them to criticism, you can rearrange them, you can analyze them, you can stare at them and reread them again and again and discover the flaws in them. You can have another thought and then join that to the next thought.
This can easily seem like a slightly banal point to make, but that simple action of being able to criticize your own thought, to externalize your own thoughts, to join together your thoughts, is impossible without literacy, and that simple act is the beginning of history and philosophy and mathematics and logic. The other part of that equation of literacy is not only the writer but the reader. Readers are able to submit written text to much greater scrutiny than a listener is to a spoken text. When you’re talking to me, I’m rolling on, going at your pace, having to follow every word you say, and you have all these persuasive techniques that have nothing to do with logic—you can be charming, you can be funny, you can skip around, and I just have to follow you, going forward and forward and forward.
One of the crucial things that we do as readers—and we know this now because scientists have tracked the way our eyes move while reading—is that we’re constantly skipping backwards and forwards over text. We’re not necessarily moving in this completely linear way; we’re looking backwards and forwards, and we’re always checking those connections, bringing greater powers of criticism and analysis to the written text. It’s just impossible to bring that to a spoken text. I think those two crucial things mean that literacy is just this huge driver of richness, complexity, and detail of thought.
Mounk: I’m convinced by that—though it’s two literature heads convincing each other, so it doesn’t take much. There are two different senses in which you’re talking about literacy, and I want to make sure that we don’t jump between them. The first one is just that you need the ability to write things down, the ability to fix our thoughts in this way, to make them subject to auditing in a way that they wouldn’t be in spoken language. But that could remain the preserve of specialists. When you’re talking about needing written mathematical notation in order to do complicated mathematical proofs, I’m sure that’s right, but it doesn’t matter if the average American or the average Brit reads fewer books—mathematicians are still going to be able to use mathematical notation, and they don’t become somehow less capable of communicating with each other about high-level math because the average person spends less time reading novels or magazines.
So, unless we really think we’re about to live in Idiocracy, the threat from the modern world presumably isn’t to the existence of the written word, or to literacy in sense one. The threat is to the broader culture in which communicating through writing and reading plays a very important role. So tell us a little bit about, beyond the importance of writing to mathematicians being able to develop those kinds of sophisticated proofs, how this culture of mass literacy—or, when we’re talking about the 18th and 19th centuries, semi-mass literacy—helped to inspire the modern world, and why, therefore, we should worry if fewer and fewer people are reading, or the number of people who buy books is going down slightly, and students at colleges read summaries of Ulysses rather than reading the whole of Ulysses.
Marriott: Of course. Telling the story of mass literacy and finding a starting point is always a somewhat arbitrary pursuit. You could begin anywhere in the last 500 years. The obvious place to begin is the invention of the printing press in the 15th century, which massively increases the volume of books produced in Europe—almost exponentially. It’s one of the great transformative technologies in history. It doesn’t necessarily lead to something that looks much like mass literacy immediately. Even in the aftermath of the printing press, it takes a while for books to be widely adopted; literacy remains a surprisingly elite pursuit, even though it’s growing in the first 100 to 150 years after the printing press.
My story starts in the middle to late 18th century, in this phenomenon known as the reading revolution—this point in history at which a number of factors seem to be driving the adoption of literacy in the wider population outside the elite. You have not only the printing press and the production of more books, but the emergence of increasingly complex modern states and economies, which demand literate bureaucrats and merchants who can read and write to conduct trade. You have the influence of Protestantism, a version of Christianity which places a huge emphasis on the written word. All these things culminate in an inflection point in the late 18th century, when more people are reading, and they also read more books, so more books are circulating and being read by more people.
It seems pretty obvious to me that it’s no coincidence that that time in history is also the time in which you get the Enlightenment and the beginnings of democracy. More different ideas are diffused; ideas can be compared and criticized in a way that’s just difficult when you have access to fewer books. You get the emergence of newspapers, the beginnings of the public sphere. You get information diffused down through society—political information which, for most of history up until that point, had been the preserve of an aristocratic elite, becomes capable of being understood, interpreted, and criticized by ordinary people.
This is just a transformative moment in our history. There are kinds of arguments about politics that you can’t have without mass literacy—not only in terms of the diffusion of political information among a wide population who can then read it, absorb it, and draw conclusions about it, but also in terms of the participation of ordinary people in political debates. I always think it’s interesting to consider some of the democratic thinkers of the 18th century: people like Thomas Paine and Jean-Jacques Rousseau came from backgrounds which, had they been born into those families 100 or 200 years previously, would never have been literate, and would never have made their arguments about democracy and rights to a large audience beyond the local pub, because they wouldn’t have had the means to. I just think there’s this inflection point when political debate rapidly expands, and information, criticism, and knowledge spread to the population. That is basically the point at which Europe and North America begin to roll out of an age of monarchical autocracy toward the democracies we know today.
Mounk: Not to go all social science on you, but obviously there’s a question about whether the spread of literacy is the independent or the dependent variable here. Is it that the spread of literacy leads to more democracy, perhaps more economic growth, more inventions, and so on? Or is it that the spread of literacy among an elite group in 16th-, 17th-century Europe allows inventions like the steam engine and other kinds of things, which create huge economic transformations, and at the end of those economic transformations is the rise of universal elementary school and so on, which actually tends to happen in the second half of the 19th century?
So, is it that literacy rises and that leads to the creation of democracy, or is it that the spread of the Industrial Revolution leads to a much more affluent society, which leads to more democratic institutions, and part of what those democratic institutions institute is much more universal education? I think it’s probably kind of impossible to fully disentangle that, but why should we think of literacy as the motive force of those transformations rather than the result of those transformations? Or is it a little bit of both?
Marriott: I think that’s really well put. I agree it’s impossible to disentangle these incredibly complex historical phenomena, and I guess one of the problems of trying to make a polemical argument out of history is that history is just not simplifiable to provide useful strands of argument.
I do think that there is a very compelling case to be made that the Industrial Revolution and the Enlightenment are basically the products of the growth of literacy. The guy who I think makes this argument most compellingly is the economist Joel Mokyr, who’s got a book called The Enlightened Economy, another book called The Culture of Growth, and he won the Nobel Prize for his work on this stuff a couple of years ago. The case of these books is, if you are looking to explain why Europe had the Industrial Revolution, the predominant factor is its culture of literacy and its culture of the Republic of Letters—this community of thinkers, scientists, and inventors across Europe who are exchanging ideas, building on each other’s work—and without that culture of scientific exchange and intellectual interaction and criticism, you don’t really get the Industrial Revolution. He makes the case more compellingly than I could in five minutes.
There are a few interesting instances where particular innovations that were foundational to the Industrial Revolution do seem to have depended on this literate culture. For instance, I think the example he gives is that the first steam engine would have been impossible had its inventors not had access to a Europe-wide tradition of debate and criticism about issues like air pressure and valves. When they came to their invention, they weren’t just tinkering in a workshop—they were able to draw on technical information argued over by natural philosophers for the last five decades. That is a situation that can’t exist without literacy, because you just have no access to that information.
The other really interesting data point that I like—Joseph Henrich makes this point, or draws attention to the study, in one of his books—is that there does seem to be a causal connection between the number of copies of Diderot’s Encyclopédie, his great Enlightenment encyclopedia, that were diffused in various towns at the end of the 18th century, and how many inventions those towns and cities went on to send to the Great Exhibition of 1850. Which suggests—not that there were loads of kids who read the encyclopedia and then became inventors—but that it’s a kind of sign of a place that has access to knowledge and an intellectual life, that is plugged into the Republic of Letters. And being plugged into the Republic of Letters does seem possible to draw, if not a straight line, then a kind of causal connection to better inventions, better science, better technology.
I would say, given all that, I agree that literacy is profoundly important. I would not begin to say it’s necessarily the most important thing. I think you could argue that if you’re telling the story of modern democracy and you were to extract the single most important factor, perhaps it’s economic growth, and perhaps it’s the rise of the bourgeoisie and their increasing economic competition with the aristocracy that is the beginning of democracy. I can still buy that. I still think literacy is important. The argument of the book is not that it’s the only factor, but that it is a very important factor. That would be my answer.
Mounk: That makes sense. I guess I’m trying to think about what kind of literacy is necessary, and in what way, for democracy to be likely to sustain itself. Clearly there are some extreme examples in which you can get a reasonably functional democracy with very low levels of literacy. India is the obvious case. That’s atypical for a whole set of reasons—partially its democratic institutions are a strange kind of colonial inheritance from the United Kingdom. Partly, I think, the degree of what some political scientists call ethnolinguistic fractionalization—the degree of regional, linguistic, ethnic, religious, and to some extent caste difference—makes it such a fragmented society that nobody can really hope to dominate, and that helps democratic institutions. It tends to be the case that democratic institutions persist when you have very homogeneous societies, or really, really diverse societies, and the trouble often starts when you’re somewhere in the middle.
There are all kinds of reasons why India set out on its democratic experiment, and all kinds of reasons why it was able to sustain it so far, by and large. But it’s still striking that this is a case in which, going back to 1948, you had very low levels of literacy, and yet the country was able to sustain this democratic experiment for a very long time, with a giant population. Of course, you’re right that by and large we see that the more educated and more literate a population is, the higher the chances of democracy surviving have been. But what exact kind of form of literacy do you need in order to sustain that?
Subscribe now
That, I think, does depend on the kind of causal account we give about the beginning of the 19th century. If we’re really saying literacy rose and that led to all of these other changes, and then you had democracy at the end, then you might say, well, if we have a lot fewer people reading, and they have fewer critical faculties, that is immediately going to lead to a potential decline of democracy. But if you’re saying actually what happened is a much more complicated, intertwined process, where of course you need literacy among some small elite group that was responsible for those technological innovations, and the result of that process was to spread literacy from 10 to 15 percent of the population—which I guess it would have been in Britain in the 18th century, I don’t have those figures to hand—to 90-plus percent by 1900 or 1950, then literacy here is as much the thing that is being caused as the thing that’s causing it.
You might say, okay, we’re not going to go into a world where the elites in society are not able to read and write. We might read a little bit less for pleasure, and there might be a larger portion of the population that really has trouble reading in any significant way, but that just gets us back to what the origins of democratic societies looked like. At the same time, we’re going to get other forms of engagement which can also serve to spread information. I’m listening to a lot more audiobooks than I did in the past. Is that exactly equivalent? Perhaps not, but I certainly get something out of listening to a French novel I’m listening to at the moment, or listening to the standard history of China that I listened to a few months ago. It’s not like none of that information enters my mind—it’s still a way of engaging with serious information. And of course, perhaps the median YouTube channel, and even the median Substack, might be full of empty inanities, but there are also conversations like these, full of great insights. So I guess the question is, we’re clearly not going to become a completely illiterate society.
We’re always going to use reading and writing in certain specialized contexts, in order to do things like mathematical proofs, or write the instructions for how to administer the nuclear codes, or whatever other kind of context. There are going to be other kinds of forms of intellectual exchange—a lot of the technologies now exist. So why should we be as concerned as you are that this slight change in the mix of different forms of communication, and the exact role that literacy plays in our society, is going to be that damaging?
Marriott: A really good question. I guess in my head I split the beneficial democratic effects of literacy up into two camps. The first is the democratization of information, which took place in the 19th century, and the democratization of the skills of analysis and abstraction that are associated with literacy. I know that sometimes sounds like a slightly woolly thing to say, but I think, being trapped inside our literate society, we can fail to see how the ways we think about politics are shaped by literacy.
One of the things—there’s interesting writing by an Edwardian journalist called George Bourne, who was going to these rural communities in the UK where literacy had just arrived, and he was talking about the transformative political effect of the spread of even quite cheap newspapers. People were accessing information about their governments, and they were accessing what to us seems an obvious, more abstract way of thinking about politics. Something is going wrong in your local area, and it’s not just an item of gossip, not just something that’s arrived out of nowhere and exists in your life that you complain about and it makes things worse. The skill of literacy allows you to join it up, because you have more information, because of more abstract thinking, to a wider social story, and you can think about it as part of a systemic problem. I think literacy diffuses information and increases critical thinking through the population, and that’s its main role in the rise of democracy.
There’s a slightly separate point to be made about the way that it sustains democracy, which is becoming more apparent to us as we move into this post-literate age—which is that one of the most striking things that Marshall McLuhan says about literacy is the way it emotionally desaturates a culture and a society. He makes a really good point, I think it’s at the beginning of The Gutenberg Galaxy, that if you come from a traditional pre-literate culture, an oral culture as we call them, the thing that would strike you about a highly literate culture—like the one that existed in the beginning and middle of the 20th century—would be how eerily emotionless it is.
When you’re communicating in an oral culture, all the information you ever receive comes to you face to face from a person, and you receive every item of political gossip or knowledge about the world in a highly emotional context. You’re talking to each other, you’re excited, people have different personalities, some of them are charming. When people are trying to persuade you around to their point of view, they’re shouting at you, or charming you with their charisma, or maybe they’re just really tall, and we’re more inclined, irrationally, as human beings, to be persuaded by tall people because they seem more authoritative and we assume they have higher status.
What literacy really effectively does to our civilization is strip a lot of these irrational forms of argument out. When you’re trying to persuade someone of a point through writing on the printed page, you can’t charm them, you can’t shout at them, you can’t cry at them, you can’t impress them with your physical presence. Writing is by no means perfectly rational, but it places a much greater emphasis on rational argument than face-to-face communication does. I think one of the ways literacy has really benefited democracy is by biasing our discourse—not in every instance, but overall—more towards less emotional, more reasoned forms of communication and argument.
McLuhan was making this point in the ’50s and ’60s, and I think it’s becoming obvious to us now quite how right he was. I think a lot of what’s going on with our politics now is—I think not only that reliable information is becoming less available to people for various reasons, but that we’re just seeing the emotional temperature of our whole culture whoosh back up again, as this whole human world, which literacy had tended to exclude—of people’s voices and faces and personalities—is just rushing back in. If you want to make a point in person or over a video, you have a range of extra-rational or irrational techniques of persuasion you can draw on, compared to if you’re making a point in a newspaper column or a book, where you’re much more restricted in the number of irrational techniques you can draw on. I just think this much more personal, much more emotional culture is much more susceptible to the kinds of irrational and charismatic forms of persuasion that I think are very dangerous for democracy. It seems like an almost banal point to make, because to me the evidence is everywhere.
Mounk: I’m a little bit torn about this, because you meet certain polemicists who sound incredibly macho on the page, for lack of a better word—who dress down their enemies with the confidence of a field marshal—and then you meet them in real life and they’re just sad little homunculi. You think that clearly, all through their middle and high school lives, they dreamt of taking to the page to eviscerate the bullies, and now they’re hiding behind the featurelessness of writing, exactly as you describe it, in order to do that.
Perhaps it’s good that we’re able to look beyond the homunculus in the writing, or perhaps it’s actually hiding something deep about the character. But it’s helpful to know, because you’re like, you’re not the kind of self-confident, boasting writer you pretend to be. You’re just a sad little hurt man who was bullied too much in middle and high school.
Marriott: T.S. Eliot has a great line about the bombastic self-confidence of the mild-mannered man safely entrenched behind his typewriter, which has sometimes occurred to me as I’ve written some of my more bombastic opinion columns safely behind my desk.
In the rest of this conversation, Yascha and James explore whether artificial intelligence will impact literacy, how we can preserve the art of reading, and why the decline of literacy is bad for democracy. This part of the conversation is reserved for paying subscribers…
We are the writers behind “Balance ton Claude,” an anonymous X account that, on September 21, published evidence that C’était ça ou mourir, the breakout novel of France’s literary season, had been largely generated by AI.
The book, by the Haitian-Canadian author Thélyson Orélien, had just won several prizes, was the favorite to win the Goncourt (France’s most prestigious literary prize), and was being translated into twenty languages. We ran it through Pangram v.4 (by far the most reliable AI detector, with false positive rates in the order of 0.0041%). Every page came back at or near 100% AI. In the following days, the author was found to have plagiarized much of his literary production before the advent of LLMs (and admitted as much). Furthermore, readers who are used to working with AI quickly identified, throughout the book, the obvious LLM writing habits. In light of the overwhelming evidence, the Académie Goncourt removed the book from its longlist.
Subscribe now
Here is why we made the revelation. We have been writing for several months about our fear that LLMs could end the civilizational era that was based on reading and writing. A reader of our work came to us with his doubts about Orélien’s novel. We chose to reveal the truth. Actually, we have no problem with people using AI to create content—we are generally tech-positive and have written at length about our fascination with AI—but because readers care about whether words in a book were chosen by a human or by a machine, it is dishonest to lie to them.
This, we believe, is the necessary conversation. What matters here isn’t Orélien’s race or nationality, or our political views (or the political views that are sometimes attributed to us). The issue is far more universal than that. It’s about whether we are willing to have our literary and public discourse overwhelmed by AI writing—and whether we really will deprive ourselves of the discernment to tell the difference. Since we are not opposed to AI by itself, our argument here may appear a little subtle, but we do believe that there is a line to draw, and that the line has intellectual coherence to it—and, certainly, the “Balance ton Claude” collective have been willing to enforce that line. What matters here is that LLMs pose an existential threat to literature.
As the scandal has shown, AI can already produce writing not only good enough to pass the Turing test, but to win national literary prizes. Models will keep improving, their writing ticks will fade, and they will increasingly one-shot content indistinguishable in form and substance from the work of the most talented human authors. Pangram could unfortunately become fallible (it already generates some false negatives, and many are fine-tuning open-source models so as to make its writing undetectable to Pangram). Furthermore, people who pride themselves on always spotting LLM prose probably exhibit a misplaced confidence: they don’t account for the progress to come, and they never notice the times they are fooled. As Benjamin Todd from “80,000 Hours” put it on X: “Plastic surgery is most noticeable when it’s bad, so it’s more widespread and successful than it looks. Same with AI writing.”
So, what will happen? LLM-written text will make up a growing share of what appears on social media, in academic journals, and in bookshops. If even 0.001% of AI output is good enough to pass as good human content, this will represent a massive volume because AI-writing can be produced in a few seconds, at almost no cost, without limit. The logic of that is it will eventually crowd out human content. An ever-smaller share of human attention will go towards human-written text. Human writers will be discouraged, and many will stop writing. Why spend weeks on an essay that will be suspected of being AI-generated anyway, that may be no better than what a model could produce in a few seconds, and that will struggle to find readers?
There will always be some demand for human content, but if readers cannot tell human and AI content apart, that demand will not be able to be expressed. The supply (texts whose words were chosen by a human) therefore won’t be produced.
The Market for LemonsThere is a strong parallel here with the situation described in George Akerlof’s famous 1970 paper, “The Market for Lemons.” Imagine a market for used cars. Most cars are good, worth say $10,000. They trade at that price; both buyers and sellers are happy. One day, the owners of bad cars (“lemons,” worth close to nothing), find a way to make them indistinguishable, to the buyer, from good cars. They start selling. Word spreads that some lemons are masquerading as good cars. Buyers now ask to be compensated for the risk of getting a lemon and price every car at the expected value of a random car on the market, say $8,000. But at that price, some owners of good cars (who know what they hold) refuse to sell and withdraw from the market. The average quality of cars on sale falls. Buyers observe this and revise their offers down again, say to $6,000, to reflect the new expected value of a car. More good cars withdraw. The spiral continues until the market unravels: only lemons are left, and nobody buys.
Now consider the marketplace for texts on social media. The currency in which readers pay is attention (which, like money, is a finite resource). Before LLMs, the probability that a given block of text was a good car (a text whose words a human had chosen) was 100%. The reader applied no discount in the attention allocated to the block of text. One day, LLMs appear. The probability drops slightly, say to 80%. Facing the same block of text as before, the reader becomes slightly less likely to read. Some writers, losing audience, stop expending effort to write. The share of human text falls to 70%. The reader applies a further discount in the attention allocated, disincentivizing more potential writers. The spiral continues—the same for writing as for used cars. (In this example, AI texts are lemons not because they are objectively of lower quality, but simply because some readers have a subjective preference against them.)
Note that the reader’s discount falls on every text, human or machine, but a reduced reward eliminates only human writers, because they alone have costs to cover. Each drop in attention pushes some human writers below their break-even point, while no drop in attention can push an AI copy-paster below theirs. So, the composition of the pool keeps worsening, with human writers exiting from the content producer pool first, followed by effortful prompters, and finally leaving only the cheapest producible text. (Those who defend AI writing would do well to recognize that the same dynamics affect the effortful prompters.) At some point the reader understands that the probability of facing human text is close to zero and stops allocating attention to text altogether: the chance that what is being read is ChatGPT prose is simply too high. Without readers, the remaining writers quit. The market unravels.
We’ve tried to put this dynamic in visual form with the following graph:
One source of stability in the market is writers with pre-AI reputations, who have proved they can write without LLMs. They will benefit from a presumption of authenticity, and their texts will not be discounted as much as the average text by readers (perhaps in the short-term, these writers will even benefit from an attention premium!). But unknown writers will face a massive suspicion of copy-pasting AI. Soon every aspiring author will think: “why spend weeks to do marginally better (perhaps worse) than an LLM, only to be suspected of having used one anyway, and to fail to reach an audience anyway?” When legacy authors have retired, few will write.
In places where that sort of earned reputation is absent by design, the unraveling is already visible. Geoffrey Litt, who did graduate work in human computer interaction at MIT, reports that he takes no more pleasure in reviewing anonymous academic submissions in the peer-review context: he always wonders whether he is reading a colleague’s thinking or Claude’s, and fears he is spending more time reading than the author spent writing. He predicts anonymous peer review in its current form will not survive AI. What has already happened to anonymous submissions will happen to every text by an author with no pre-existing reputation, meaning, eventually, to almost every text. One possible exception is that we may continue to read people whose standing comes from something other than writing: central bankers, politicians, CEOs, famous actors, etc, either because their public record makes them trustworthy or because their position makes their opinion worth knowing even if an LLM or ghostwriter chose the actual words. Note that this scenario implies that standing becomes a precondition for being read, and standing cannot be acquired by writing—and that (very-plausible) scenario serves as an answer to the oft-heard argument that “AI will democratize writing.” Here, in the long run, only the famous get read. The anonymous but talented young thinker has no chance.
Bookstack at Persuasion is delivered to you every other Sunday at 6am EST. Grab a coffee, settle onto the couch, and see your weekend out with a deep dive into the best new book releases…
This week, Richard talks to Peter Bergen, author of All the Presidents’ Wars: Twenty-Five Years of America's War on Terror.
Host: Richard Aldous
Producer: Bhaasita Athani
Follow Persuasion on X, LinkedIn, Instagram, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below. Bookstack listeners can take advantage of a 20% offer for paid Persuasion membership, with access to special columns and bonus Good Fight with Yascha Mounk podcast content.
20% off for Bookstack listeners!
From the publisher's feed