A Collection of Unmitigated Pedantry · 392 posts · started 3 May 2019, one post every Friday, almost without a gap · latest post 11 September 2026, still running · posts run 5,000 to 12,000 words; a series often runs eight to twelve weeks · one clear turn: in the summer of 2020 it shifted from debunking history in pop culture to explaining how ancient societies actually worked, and from 2022 it added modern war and geopolitics · this page is built on 13 posts read in full, spanning May 2019 to July 2026, six of them from the earliest 20% of the blog's lifespan, with samples on both sides of the turn · everything on the site is free; there is no paywall
In May 2019, a man with a fairly fresh PhD in ancient history sat down to watch The Return of the King and started counting orcs.
The effects team had said in interviews that they built the host of Minas Morgul at around 200,000. On screen, the road lets the orcs march five abreast. Give each orc four feet — which is already generous. 200,000 divided by five is 40,000 rows; 40,000 rows times four feet means the column is thirty miles long.
Thirty miles means the tail of the army cannot break camp before the head has finished marching for the day.
But the good part isn't "the movie got it wrong." It's the next calculation. Minas Morgul is about sixty miles from Minas Tirith, and an infantryman can carry roughly ten days of rations, enough for about 120 miles. So as a matter of supply, this works. And that same arithmetic quietly explains the Witch-King's behavior — why he storms the city instead of settling in for a proper siege:
Because the pack animals which pull wagons full of food eat food themselves, there is literally no amount of wagons which would enable an army of this size to sustain itself indefinitely in a long siege.Collections: The Siege of Gondor, Part I, 10 May 2019
The paragraph that opens that section reads: "They say 'amateurs talk tactics, but professionals study logistics.' Well, pull up a chair at the Grown-Ups Table, and let's study some logistics."
It reads like elaborate nit-picking. Seven years and 392 posts later, the blog has become something else entirely.
It will now tell you things like this: before about 1750, in every society — rich or poor, urban or rural, "civilized" or not — roughly half of all children died before adulthood. Which is why the familiar figure "life expectancy in antiquity was about 25" misleads. Survive to age five and your life expectancy jumps to about forty. Death was not spread evenly across the ages; it was concentrated, brutally, in the first few years.
What Devereaux has kept coming back to for seven years is a single yardstick: to judge a society, don't look at the most glamorous one percent. Look at how the median member lived. And by the numbers, the median member was a farmer.
The most famous use of that yardstick is Sparta. He did not argue that the sources are unreliable. He did the opposite: he granted the admirers their sources entirely — and then started counting people. The helots, Sparta's collectively-owned slaves, made up 65 to 85 percent of that society. For comparison: Roman Italy under Augustus, maybe 25 percent. The American South in 1860, 32.3 percent. He knows of no other state enslaved to that degree.
What about the military? He splits war into three layers — tactics (how you fight once you're there), operations (how you get the army to the fight and keep it fed), strategy (why you're fighting at all) — and then points out that Thucydides records, year after year, the Spartans invading Attica and going home because they ran out of food. 431, 430, 428, 427, 426, 425. Six years. Their ally Corinth's great port was two days' march away.
A mediocre soldier on the field beats a super-soldier sitting at home, every time.This. Isn't. Sparta. Part VII: Spartan Ends, 27 September 2019
His verdict leaves no room: Sparta was "little more than an extraordinarily effective prison, metastasized to the level of a state." Not a model. A cautionary tale.
Then he went after the bigger thing: the meme that circulates online as "hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times." He notes its source — a novelist, G. Michael Hopf, "perhaps conspicuously, not a historian." He named the whole complex the Fremen Mirage, after the desert warriors in Dune, and then went to count who was actually winning. The answer is deflating: agriculture began in the richest zones, and so did the state. The "hard men" pushed out into deserts and mountains were the losers of that process, protected by the poverty of their land rather than by their prowess. Hence the line that works almost as an axiom:
no one chooses to live in a resource poor zone if other options are available.The Fremen Mirage, Part I, 17 January 2020
In 2021 he gave the ideology a name — the Cult of the Badass, whose central tenet is that the capacity to produce violence is the highest human virtue — and then walked it, point by point, through Umberto Eco's 1995 essay "Ur-Fascism." He was careful: he states explicitly that he does not think the novelist he is criticizing is a fascist. His claim is that the ideology shares most of fascism's DNA.
It is a modern lie, masquerading as timeless wisdom.The Universal Warrior, Part III: The Cult of the Badass, 19 February 2021
But his best work isn't in the polemics. It's in making you do the arithmetic on a peasant household's grain. The harvest comes in once a year; the family eats every day. Stored grain starts spoiling faster after a year. Money can be stolen or taxed — and worse, the price of food rises exactly when you need it most, so cash goes useless at the moment of crisis. So farmers banked their good years in relationships: they banqueted neighbors, funded village festivals, married their daughters off with the best dowry they could manage, and paid back favors. Which diffuses the surplus, which is why only about ten percent of it ever reached the cities.
Which is why pre-industrial armies had to take food as they went — everything that can carry food also eats food. He calls a village a "food piñata" for an army: full of food, some violence required for extraction. And he refuses to let the bloodless phrase "ravaged the district" cover it, writing a full account of what foraging actually involved — the looting, the burning, the torture, the rape — and then closing here:
Every so often someone asks "If you were in [era in the past], what would you be?" and folks respond with kings and knights and clergy and artisans. But you wouldn't, would you? By the numbers, we wouldn't be the soldiers in gleaming armor, we'd all be the poor peasants desperately grabbing what few valuables they had and their families and sprinting in the woods as everything we had built over decades burned behind us.Logistics, How Did They Do It, Part II: Foraging, 29 July 2022
The same post contains a shorter line: "If this needs saying: war is terrible, we should do less of it."
A blog that began by counting orcs in a movie ends up here. That isn't drift. It is the same project throughout: take a past that has been told as a thrilling story and put it back on its real distribution of people.
There is a city-building game about ancient Egypt called Children of the Nile. Devereaux says that when he played it, the moment his ruler turned about twenty, the game informed him that he was now old and must think constantly about death.
Whoever wrote that line, he realized, had been fooled by a number.
The number is "life expectancy in antiquity was about 25." It is true, and it misleads. Using the model life table ancient historians normally reach for: at birth, 25.0 years for women and 22.8 for men; survive the first year and it jumps to 34.9 and 34.1; reach age five and it is 40.1 and 39.0 (on top of the five years already lived). The longer you have lived, the more years you have left — the exact opposite of the intuition.
There is only one reason for that, and it is the single fact on this page most worth keeping: of all live births, fully half were dead by age ten; only about 45 percent reached adulthood. This is not one unlucky society. He states it flatly: it holds for every society before about 1750 — hunter-gatherer, pastoralist, farming; urban and rural; "Western" and not. The richest aristocracies could buy no exemption, because until the 1700s no society's medicine could change the pattern at all.
So the twenty-year-old ruler did not think death was around the corner, because it usually wasn't. The real fact is harder: half the people born alongside him were already gone before he could remember them.
…it is one of the most striking ways in which "the past is a foreign country," although in this case the past is even more alien than any foreign country on earth.Life, Work, Death and the Peasant, Part II, 18 July 2025
Bret C. Devereaux holds a PhD in ancient history from UNC Chapel Hill and an MA in classical civilizations from Florida State, and is a Teaching Assistant Professor at North Carolina State University — a teaching post off the tenure track. Here is how he puts that himself:
While I do teach as the academic equivalent of a tenant farmer, tilling the Big Man's classes, this project is my little plot of freeheld land…Life, Work, Death and the Peasant, Part II, 18 July 2025
His research sits at the intersection of the Roman economy and the Roman military. His official bio solemnly records that he has beaten Dark Souls (and now also Elden Ring), and that he "enjoys good music, bad jokes and writing about himself in the third person."
The blog opened on 3 May 2019 with three posts in a single day. Since then it has run one post every Friday, reaching 392 posts by September 2026, and it is still going — the most recent went up nine days ago. He gets his entire audience by word of mouth, funds the work through Patreon, and when a topic ought to be chosen by readers he actually holds the vote.
He coined a phrase for his enemy: "a boy's view of military or social history," with a matching "boy's sociology." His judgment on it is blunt — it is "unfit for use by serious adults."
There is a very specific moment where he ran into it. Teaching Sparta to undergraduates, he found the standard approach was to hand students Plutarch's Life of Lycurgus — or describe it to them — with no apparatus for questioning what Plutarch said, and, more importantly, none for following that testimony to its necessary conclusions. Students left with a hazy, utopian feeling about Sparta.
And that feeling has legs. As he writes, the Sparta of the undergraduate survey becomes the Sparta those undergraduates teach when they become high-school teachers, which becomes the Sparta of Frank Miller, Steven Pressfield and Zack Snyder. For years he joked with colleagues that he wanted to build a syllabus for a course titled "Sparta Is Terrible and You Are Terrible for Liking Sparta." In August 2019, two duelling essays in The New Republic and a subsequent Twitter argument with one of the authors turned the joke into seven posts.
But Sparta is only the door. The root question he has not put down is: whose experience gets to stand in for "the past"?
In agrarian societies, farmers were 80 to 90 percent of the population. The people who wrote, and whose writing survived, were the thin layer on top. So "history" defaults to the history of elites, of wars, of the parts you can film — and then that version gets used as an instruction manual for modern life. "What we need is a Caesar." "Hard times create strong men." "The Spartans were real men."
His counter-thesis has two halves, and it only has force when both hold. The first is a yardstick: judge a society by its effect on the typical member — not by how well its aristocrats lived, not by which battles it won. The second is a diagnosis: the popular histories of hard men and golden ages were never descriptions of the past. They are modern feelings wearing period costume.
Miss the root question and the blog reads like an expert nit-picking movies. The nit-picking is the entrance. It is not the work.
Over seven weeks in August and September 2019, he wrote Sparta. The series became the second most-read thing on the site — at the three-year mark he looked it up and found just over 415,000 page views.
What makes it work is that he gave up the easy route at the start. One camp in the scholarship (he calls it the Hodkinson camp) argues that our sources exaggerate Sparta's peculiarity — that much of what looks exceptional wasn't. Taking that road, all he had to say was "Plutarch is embellishing." But he judged that road a dead end rhetorically, because it demands that your reader throw out every ancient source he has ever read before you've said anything else.
You are asking him (it is usually a him) to begin the meal with the most difficult to swallow part of the argument.This. Isn't. Sparta. Retrospective, 19 August 2022
So he inverted it: I accept your sources in full. Now let's finish the arithmetic together.
Step one is counting people. Recent scholarship has been trimming helot numbers — from Talbert's c. 200,000 to Hodkinson's c. 162,000 to Figueira's low estimate of 75,000–118,000. Devereaux notes that a lower bound is a lower bound: even at the floor, the enslaved share of Spartan society is 65 to 85 percent. He supplies the controls too: Roman Italy in 225 BC perhaps 18 percent, rising to around 25 percent by Augustus — the fruits of two centuries of successful war; the American South on the eve of the Civil War, 32.3 percent.
Which turns the scholarly argument about whether the citizen class was exceptional into a question of posture:
it is quibbling over pebbles while tripping over a boulder. That boulder is the helots. And it is a big boulder!This. Isn't. Sparta. Retrospective, 19 August 2022
Step two is the military — and this step is the most portable thing on the site. He refuses to measure military effectiveness by battles won, and insists on splitting the question into three layers: tactics is how an army fights; operations is how you move it, supply it and keep it in the field; strategy is what wars are worth fighting and for what ends.
Split that way, Sparta's report card reads: tactics, modestly better than the average polis, and only modestly. Operations, a disaster. Thucydides records the invasions of Attica year by year: in 431 they besiege a minor town, accomplish nothing and leave; again in 430; in 428; in 427; in 426 they turn back early after earthquakes; in 425 they are back — and every time they leave because supplies run out. They are two days' march, on foot, from a major friendly port at Corinth. Six years after the first invasion, they still had not solved the logistics of sustained operations in what was effectively their own backyard.
At the strategic layer he offers one of the better images I've read for a state with one tool: Sparta is the fellow in the aphorism about having only a hammer, dropped into a world of screws. Any problem that could not be solved by driving a phalanx at it — maritime enemies, fortified positions, holding an alliance together, raising money — either defeated them or provoked them to try to compel friendship with hoplite armies, which worked exactly as well as you'd expect.
Then the ending, which he writes almost like fiction. After Spartan hegemony shattered, Athens remained a center of commerce and culture and Corinth a crucial port, while Sparta became a theme park — a destination where wealthy Greeks and Romans came to gawk at quaint rituals. Plutarch and Cicero both went. The markets catered to Roman tourists.
Thus ended Sparta: not with a brave last stand. Not with mighty deeds of valor. Or any great cultural contribution at all. A tourist trap for rich and bored Romans.This. Isn't. Sparta. Part VII, 27 September 2019
The thing he will not let go of is the bargain itself. Many people are willing to forgive Sparta everything on condition that it really did produce super-soldiers. He dismantles the condition: it didn't. Spartan soldiers were average.
The horror of the Spartan system, the nastiness of the agoge, the oppression of the helots, the regimentation of daily life, it was all for nothing.This. Isn't. Sparta. Part VII, 27 September 2019
Four years later he finally wrote the principle behind the yardstick as a standalone piece. To the objection that historians shouldn't make value judgments, or should judge only "by the standards of the time," he answers on two levels. First: historians of the past made judgments too, they just didn't announce them — it is no accident that Droysen and Tarn, upper-class men writing inside hegemonic imperial monarchies, thought Alexander's hegemonic imperial monarchy was wonderful. Second, and sharper:
what people mean when they complain that now historians are 'imposing' their values on history is really, 'in the past I agreed with the values historians imposed on history, but now I do not.' This isn't a plea for objectivity, but for special consideration.Fireside Friday, June 23, 2023 (On Historical Judgement), 22 June 2023
So what should the standard be? He offers two, both practical. One is "did this person succeed at what they were trying to do?" — Octavian largely did (re-establish the Roman state around his sole rule), Sulla largely did not (restore the Republic to durable working order). The other is the subject of this section: the impact on the well-being of the typical member of a society. Because by the values of many Greek elites Sparta was notionally splendid — and those elites never wrote down the condition of most people who lived there.
In January 2020 he picked a larger target: the meme about hard times and strong men. He supplies the attribution — the novelist G. Michael Hopf, "perhaps conspicuously, not a historian" — and adds an aside I'm fond of: "one also wonders what the women are doing during all of this, but I have to admit, were I they, I would be glad to be left out too."
His first move is his signature one: translate a mood into propositions that evidence can hit. He breaks the "theory" into four claims — less settled peoples are made tougher by hard conditions; therefore they fight better than their richer neighbors; therefore they inevitably overrun them; therefore history is a cycle. Only once it is in that shape is there anything to shoot at.
His second move is a concession. He notes openly that the framework assumes a society's chief value is how efficiently it externalizes violence — a premise he rejects entirely — and then grants it anyway, for the duration.
Let us fight the Fremen on ground of their own choosing. I am, for reasons that will soon become quite obvious, still fairly confident that our sophistication will prevail.The Fremen Mirage, Part I, 17 January 2020
Third, he counts winners — starting further back than anyone expects, with the spread of farming itself. If the Mirage held, we should see farmers repeatedly overrun by tougher hunter-gatherers. What we see is the reverse. Farming begins in the most resource-rich zones, which were already the densest and most socially complex. And it isn't farming that spreads, it's farmers — using population density and the military advantage it brings to displace lower-density neighbors. The hunter-gatherers who held out weren't those in the hardest places; they were the ones on rich coastlines dense enough to resist.
The rise of the state is a sequel to the same story: pristine states appear in exactly the richest, most intensively farmed zones. The pastoralists pushed into deserts, mountains and arid belts are the losers of that process, protected not by their prowess but by land that agrarian states cannot follow them into. Which yields the line that works as an axiom:
no one chooses to live in a resource poor zone if other options are available.The Fremen Mirage, Part I, 17 January 2020
He plants one lovely counter-example along the way: Neanderthals were more Fremen than the Fremen — stronger and more robust than anatomically modern humans — and were pushed onto progressively more marginal land anyway, by a species better at forming large groups. Hardness was not the advantage. Numbers and organization were.
Fourth, and this is where the series is most honest: he brings out his opponent's best case himself. Steppe nomads. He concedes without hedging that the hype in the sources "was real" — a mounted steppe archer was probably the most dangerous man-for-man combatant on any battlefield of his day.
Then he takes it apart. Mongol power came from a very specific intersection of skills — riding, horse archery, and using the logistics of daily pastoralism as military logistics — not from hardship. The evidence is that harder people existed: the "forest people," pushed out of steppe society into more isolated and difficult lives, could not keep herds, and so had no horses, no horse archery, no steppe logistics, and were weaker. As for scale, he is equally clear: Chinggis succeeded largely because he rewrote steppe culture, abolishing the traditional hierarchy and imposing a decimal organization that obliterated clan and tribal lines, solving the scaling problem that limits most non-state peoples. That makes the Mongols an outlier among steppe polities, not their exemplar.
Then he counts China. He lays out a timeline from the Spring and Autumn period to the Qing and totals it up: across 2,683 years, complete steppe control amounts to about 98 years, plus roughly 540 years of partial control, often over far less than half of what we now call China. Give the partial years half credit and nomad control is about 13 percent of the timeline; generously add the Qing and it reaches 24. For the other three-quarters or more, the most serious threat came from where? From something exactly like itself — another rich, complex, highly stratified agrarian state. Rome is the same story: its peer competitor, Sassanian Persia, was the real danger, not the "barbarians" on the frontier.
In the modern era he cites Max Boot's tally: of 443 insurgencies, almost two-thirds unambiguously failed and only about a quarter unambiguously succeeded. So much for the assumption that the insurgent is more dedicated, more willing to endure, more willing to take casualties — most insurgencies are outlasted by the states they fight. And he adds a methodological barb: you functionally never see anyone try to assess these questions carefully. They are assumed.
Finally the origin, which is the turn of the whole series. Where does the Mirage come from? From "decadent" societies' self-criticism — the Romans complaining about their own softness while standing at the dawn of a long period of power and effectiveness. Their own future defies the theories they advanced. Those ancient texts were then picked up in the nineteenth century as raw material for nationalist and frequently racist identity construction, mixed with orientalism and "scientific" racism, and baked into public education. The pure form died in the Second World War; the residue survived as a thing "everyone knows" about history.
The Mirage is broken at the keel, at its foundation – the ship's super-structure may not be made more seaworthy by sawing it off the hull.The Fremen Mirage, Part IV, 28 February 2020
A year later he pushed the thread to its end. In the Universal Warrior series he first spends three weeks establishing that there is no universal warrior experience or set of warrior values (one culture's definition of courage isn't merely different from another's, it is sometimes diametrically opposed), and only then turns to the ideology underneath. He names it the Cult of the Badass and states its creed in two lines: Heinlein's Premise, that "violence is the supreme authority from which all other authority is derived," and what he calls Pressfield's Corollary — that the capacity to produce violence is therefore the highest human virtue.
Refuting the first, he offers an example I have not been able to forget. In March 1783, officers of the Continental Army were entertaining a coup. Washington did not hold the balance of violence — had the officers chosen to act, he could not have stopped them. What he did was produce a pair of reading glasses and say: "Gentlemen, you will permit me to put on my spectacles, for I have not only grown gray but almost blind in the service of my country." That ended it. Any Roman, Devereaux notes, would have recognized what was mobilized: auctoritas, an authority earned through self-sacrifice that cannot be enforced by violence at all. The Romans had a whole vocabulary for non-violent authority — virtus, honor, dignitas, gravitas, auctoritas — and that vocabulary was written down by a society run by military aristocrats. If anyone were going to describe their world as pure violence, it would have been them. They didn't.
Refuting the second, he does the thing the series is remembered for: he matches the Cult of the Badass point by point against Eco's fourteen features of Ur-Fascism — the cult of tradition, rejection of modernism, action for action's sake, life as permanent warfare, contempt for the weak, the cult of heroic death, machismo, the appeal to a frustrated class. He is careful about it: he says outright that he does not think Steven Pressfield is a fascist, notes that Pressfield seems committed to democracy, and observes that his whitewashing of Sparta's uglier features suggests genuine discomfort with them. His claim is that the ideology shares most of Ur-Fascism's DNA, and that the missing elements follow near-necessarily from the opening premises. He also quotes the Eco line that echoes through everything else he has written: "Fascist governments are condemned to lose wars because they are constitutionally incapable of objectively evaluating the force of the enemy" — which is precisely what he had found in Sparta, whose ideology made it a poor strategist.
The section lands on two sentences. One: "a free society has no need for warriors. Not among its soldiers, not among its police, not among its civilians. At times, a free people may need to become soldiers, or police officers, but always to return to being civilians again, either at the end of the day or at the end of the war." The other is quoted above: a modern lie, masquerading as timeless wisdom.
In July 2026 the thread reached its current form. He wrote an anthropology of the online "statue pfp" classics crowd — he calls them chuds — and noticed something odd: they aren't deep about anything. The ordinary self-taught enthusiast is a shallow puddle with very deep potholes (can give you the hour-by-hour position of every Maine regiment at Gettysburg, but is hazy on Reconstruction), while a trained historian's knowledge is a swimming pool sloping smoothly from the shallow end to the deep.
But the chuds…it's just a puddle. Not especially wide or deep.Fireside Friday, July 10, 2026, 10 July 2026
His explanation: they are working backwards — the modern ideology comes first and antiquity is recruited afterwards as an anchor. Which is exactly why they never go deep: up close, antiquity doesn't fit. And the misfit he names is a testable positive claim: the clear tendency from classical antiquity is that diversity was a winning strategy — societies that more successfully and more fully incorporated culturally and ethnically different groups won, and societies that stayed small and homogeneous lost. Livy, Polybius, even Philip V of Macedon noticed it at the time.
In July 2020 he began a series called "Bread, How Did They Make It?" The first installment contains almost nothing about how to farm. It is about farmers.
He starts with structure: in an agrarian society, farmers are 80 to 90 percent of the population. The typical unit is a large family on a small farm — pre-Han Chinese documents assume a household of eight (two elderly parents, a son, his wife, four children); Ptolemaic and Roman census data show a bewildering variety of compositions but the same rough size. The land: the standard estimate for a Roman smallholding is 5 to 8 iugera, roughly 3 to 5 acres; the pre-Han figure of "one hundred mu" comes to 4.764 acres — essentially identical. A 1460 survey of Saint-Thibery in Languedoc found 118 of 189 households on under about 12 acres.
Then the counterintuitive point: these households are badly configured as units of labor. A family with two adult men could work close to 20 acres; it is likely to be sitting on a quarter of that. Too many mouths, not enough land. But the farmers aren't optimizing for efficiency. They want to marry, have children, stay near the people they love. "Farmers, after all, are people, not mere tools of agricultural production."
And then the axis of the whole series: the grain is harvested once a year, but the family eats every day.
After that sentence, every apparently foolish thing these farmers do becomes sensible. They do not maximize yield; they minimize risk — because improving yields by 5 percent means nothing if a tail risk starves everyone in year three. So they plant wheat and barley and beans together (wheat is more valuable, barley more drought-resistant: if the rains fail the barley survives; if it rains too much the barley is up on the high ground anyway). So they deliberately scatter their plots across different micro-climates — a bit of the valley, a bit of the flat, a bit of the hillside. Every one of those choices costs efficiency. Every one buys survival probability.
These are not irrational, unthinking people; they are poor, not stupid – those are not the same things.Bread, How Did They Make It? Part I: Farmers!, 24 July 2020
So how do you bank a good year? He closes the exits one by one. Store grain: losses are tolerable for the first year, but after that pests and spoilage accelerate; and doubling storage is a real capital improvement (he has calculated a Roman family of five at about 200 modii a year — 1,746 liters, enough to fill a small cellar once divided into containers). Store money: it can be stolen, seized by passing armies, and taxed, often extortionately. Worse, the disasters you most want to insure against — bad harvests, war — are precisely the ones that contract the food supply and drive the price up, so money goes useless at the moment it is needed most.
So they banked it somewhere else:
Grain rots and money can be stolen, but your neighbor is far likelier to still be your neighbor in a year, especially because these relationships are (if maintained) almost always heritable and apply to entire households rather than individuals, making them able to endure deaths and the cycles of generations.Bread, How Did They Make It? Part I: Farmers!, 24 July 2020
Which gives village life its shape. In a good year they banquet the neighbors, pay into the village festivals, marry off sons and daughters with the best dowry they can manage, and settle the favors they have recently called in. This is not folklore; it is insurance, and better insurance than a granary or a purse, because it survives death and the turning of generations.
But it has a consequence, and the consequence is the real point of the essay: the surplus gets diffused. There is a ton of food in this countryside and hardly any surplus — because production is organized for minimum risk, because families grow to the size the land can support, because anything spare goes straight back into the social safety net. His back-of-the-envelope figure for what finally reaches the non-farming cities: something like ten percent.
That constraint grows upward into his two most forceful conclusions.
First: the tyranny of the wagon equation. Everything that can move food — people, mules, oxen, draft horses — also eats food. So a pre-industrial army cannot live off supply lines running back to a base; it must gather locally, continually. Which is why he wrote twelve thousand words on foraging, taking apart a thing military history usually disposes of with the phrase "ravaged the district."
Taken apart, it looks like this. Villages are not scattered at random; they cluster and connect by paths. An army with cavalry can forage a front roughly twenty miles wide, moving ten miles a day — about 200 square miles, holding perhaps 5,000 people at pre-industrial densities (around 25 per square mile), who just after the harvest are sitting on roughly 1.35 million person-days of food. His sample army of 19,200 combat effectives, 4,000 non-combatants and 9,800 animals needs about 61,850 kg per day — roughly 5 percent of what's locally available. He calls the village a food piñata: full of food, some violence required for extraction.
Those numbers explain a stubborn fact of the record: the ceiling on field armies in the broader Mediterranean. 20,000 is normal, 40,000 is big, 80,000 is unusually huge, and more than 80,000 is unsustainable in almost all circumstances.
Then the violence itself, which he will not skip. Foraging parties could be up to a third of a medieval army's strength, split between fast outriders who hit the village before news of them could arrive and slower foragers who came behind to carry things off. Captured peasants were tortured to reveal hidden seed grain; captured women were very frequently raped. He quotes the thirteenth-century Chanson des Lorrains on the scene, and quotes Fulcher of Chartres praising crusaders who, having captured Muslim women, "did nothing evil to them except pierce their bellies with their lances" — a churchman recording this as restraint.
If this needs saying: war is terrible, we should do less of it.Logistics, How Did They Do It, Part II: Foraging, 29 July 2022
And the final paragraph of that post is the "you wouldn't be the knight" passage quoted at the top of this page. It isn't a flourish. It is what all the preceding arithmetic — 80 to 90 percent of people farming, ten percent of surplus extracted, five percent consumed per day — becomes once you write it in the second person.
Second: why there was no Roman industrial revolution. In August 2022 his Patreon supporters voted for the question. His answer breaks "the industrial revolution" out of its vagueness into a chain of specific links, each of which requires three things at once: the principle, a workable design, and an economically viable use case.
The chain runs: Britain's forests were largely cleared in antiquity (by 1000, England was about 15 percent forested, later dropping to around 5), so from the sixteenth century people heated and cooked with coal; surface seams ran out and mining went deep; deep mines below the water table must be pumped; and pumping by steam only pays where the fuel is nearly free — Newcomen's 1712 engine was so appallingly inefficient that it made sense only sitting on top of a coal mine. That use case bought the decades of iteration that produced Watt's engine in 1776. But you also need the other end: an industry where rotational motion is the bottleneck. That is spinning, which consumed 80 percent or more of the labor in textile production — and you need Hargreaves's 1765 spinning jenny to turn rotation into thread at scale, and you need Britain to be the world's textile center (Scottish and Welsh wool plus the cotton of empire in India). You even need Europe's kings to have spent three centuries obsessed with casting pressure-containing cylinders, because a good cannon and a good steam cylinder are the same craft problem.
What Rome lacked was not cleverness. It was every link. They used some coal, but marginally. They used watermills for grain, not for thread. They had no spinning wheel at all (still drop spindles), and their warp-weighted loom was roughly an order of magnitude less efficient than a flying-shuttle loom, so the economy could not have absorbed the thread anyway. And they had put essentially no effort into pressure cylinders, because they had no reason to.
So what was unusual about the Roman economy? He doesn't deny it: the empire created a linguistic, customs and monetary union across the Mediterranean and kept it relatively free of pirates and bandits, which made markets reliable enough for real regional specialization, spread innovations, and allowed capital to accumulate. But he draws the boundary with an image I find very good:
much like how realizing design efficiencies in an (unmotorized) bicycle does not bring it any closer to being a motorcycle; you are still stuck with the limits of the energy that can be applied by two legs.Why No Roman Industrial Revolution?, 26 August 2022
And then the conclusion with the most political weight in it, which ties all three ideas together in a sentence:
Roman economic growth, because it was a product of institutions instead of technologies, was not durable in the face of those institutions collapsing.Why No Roman Industrial Revolution?, 26 August 2022
What transfers from a long-running writer is usually not the conclusions but the moves. His are unusually legible.
Build your opponent's strongest version first, then fight on ground of his choosing. In the Sparta series he accepts the admirers' sources wholesale. In the Fremen series he grants a premise he rejects — that a society's value is its efficiency at externalizing violence. In the Universal Warrior series he spends three weeks establishing whether the historical claim is true before saying a word about the ideology, because "truth is an absolute defense in such matters." It is an expensive way to write. It also leaves nowhere to retreat to.
Split into layers. Tactics / operations / strategy is the crowbar he reaches for most often. It works because popular narrative lives almost entirely on the first layer, and outcomes are decided on the other two.
Turn a mood into propositions. "Hard times create strong men" is a feeling; he renders it as four declarative claims, and only then is there something evidence can hit. He does the same to the universal warrior (seven claims) and to the Cult of the Badass (Heinlein's Premise plus Pressfield's Corollary).
Count it. Four feet per orc, two thousand wagons, 1,746 liters of grain, 1.35 million person-days, 443 insurgencies, 98 years out of 2,683. Not a display of precision — every one of those numbers carries wide error bars — but because once you start counting, the vague adjectives have nowhere to hide.
Distinguish disputes settled by evidence from disputes where several expert positions remain live. He borrows a term from law: colorable, an argument that is legally plausible even if the person describing it isn't sure it would win. His rule for himself in public writing is to stay inside the colorable range and, when that range is wide, to say so. And when scholars accuse each other of being out of date, his response is clear-eyed: do not mistake the weight of publication for the weight of scholarly opinion, because revisionist positions out-publish orthodox ones by nature — "it is very hard to get a job arguing, 'that old fellow over there actually solved my topic 40 years ago, go read him (or her).'"
Put the bibliography at the front. His long posts open with a bibliography note before the argument starts, rather than burying it at the end. Sources first is a posture.
Say you don't know. How much did the Roman economy actually grow moving from a "low equilibrium" to a "high equilibrium"? "some estimates put it around 25% but these are very speculative; in practice we really don't know." That sentence sits in the middle of a post otherwise made entirely of causal reasoning.
Counterfactuals are for showing contingency, not for telling stories. If Spain or Portugal rather than Britain had ended up controlling India, would the cotton have flowed to a place where coal wasn't cheap and abundant? He says plainly that the question is unanswerable. He raises it to make you see how much contingency is hiding underneath a chart of global GDP.
The most honest criticism is the one he wrote himself. In the 2022 Sparta retrospective he admits that his choice between the two scholarly camps was partly rhetorical: the Cartledge position (Sparta as exceptional and exceptionally bad) is a far better persuasive tool than the Hodkinson position, because it shares an evidentiary base with the admirers. He defends the choice well, but its implication can't be dodged: this is history writing organized to win an argument. In the same piece he concedes that his first post on the agoge was unclear (he does not think the agoge taught combat skills, but the text didn't say so; he added a clarifying footnote years later), and that the legendary Spartan infanticide probably did not happen — we ought to have evidence for it and we don't.
A second self-audit: in February 2022 he wrote a long explainer on the war in Ukraine; reviewing it in May he wrote that he had "clearly underrated the probability" of the outcome that actually occurred. He had expected Russian forces to win the opening battles and to encircle Kharkiv, stalling only in the big cities on the Dnieper. Instead the offensive culminated before it got there. He is specific about which assumption failed.
Two external criticisms I read in the original. The first is Elizabeth Van Nostrand's line-by-line fact check from November 2020. She verified a great many specific claims and found most of them sound — boys taken into herds at seven, deliberate underfeeding with theft encouraged, the whipping at the altar of Artemis Orthia, the krypteia's killings, the spartiate collapse from 8,000 to 1,500, helots as state-owned slaves. But she flagged two kinds of problem: details like the "five villages" (her sources described four), and, more importantly, a methodological one.
Along the way he sometimes presents facts as more certain/less controversial among historians than they are. He's not hiding that he does this, but on first read through I did walk away with the impression that certain things were more settled than they are, and unprepared to argue for their truth.Elizabeth Van Nostrand, Epistemic Spot Check: This Isn't Sparta, 25 November 2020
The second is sharper, from a guest post by Topher Brennan: when Devereaux defends medieval Europe against the charge that A Song of Ice and Fire exaggerates the brutality of medieval war, he reaches for arguments he would never accept elsewhere — while his own logistics series documents in detail how medieval armies looted, burned and raped.
I am quite certain Devereaux knows better. His arguments in defense of the honor of medieval Europe are the sort of thing he'd have no patience for if made in defense of Sparta, or ancient Rome for that matter.Topher Brennan, Guest Post: Medieval Warfare Was Bad, Actually
Two more are structural, and readers should watch for them. First, this blog is fundamentally a polemical form. Its best pieces are all "this myth is false," and demolishing a myth is easier than building a positive picture. His genuine positive picture — peasant life, logistics, demography — is very largely a synthesis of other people's research. He says so, every time (that is what the front-loaded bibliography notes are for), but readers easily quote a synthesis as though it were a primary finding. Second, his subject matter has been steadily widening: nuclear deterrence, Ukraine, the US Navy and global trade all sit well outside the economy and military of the ancient Mediterranean. He marks his uncertainty. Readers often don't.
One last observation is mine rather than anyone's accusation. In the 2026 piece on the chuds he announces that he is no longer trying to persuade his opponents — "someone cannot be reasoned out of a position they did not reason themselves into" — and that his goal is only to put up a flare for bystanders, signaling that a more rigorous past exists elsewhere. As strategy that may well be right. But it means the blog is increasingly not arguing so much as drawing a boundary; and a boundary is a comfortable posture in which to stop being refuted.
On the Hodkinson camp's scholarly case about Sparta, I have read only Devereaux's own account of it plus the bibliographic facts (Hodkinson, Property & Wealth in Classical Sparta, 2000; his essay in Spartan Society, 2004; and so on). I have not read the originals, so I report the bibliography and do not repeat arguments I have not read.