This was a superb read, thank you. Are you by chance familiar with Scheidel's Escape From Rome? He deals with much overlapping themes of European comparative prosperity and China's hardships with centralisation and steppe hordes.
The book argues that the fall of the Roman Empire and the resulting political fragmentation in Europe was the best thing that ever happened to us. It fostered competition and innovation, leading to the region's eventual economic and technological sucaess. Unlike centralised states like China, Europe's divided states encouraged military and technological advancements, preventing monopolisation of power. This fragmentation laid the groundwork for capitalism, scientific progress, and modern political institutions, making the collapse of Rome a crucial turning point for Europe's prosperity.
I found many, many same things mentioned in this post as was in the book.
Big fan of Escape from Rome. Scheidel's work gets referenced in a few of the papers I cited here, so it's also definitely been a big indirect influence on this post.
I couldn't help but be reminded of Taleb's theory of anti-fragility. It seems The West was more anti-fragile because of its diverse geography and the multi-variate threats against it, while China was bigger and better organized but ultimately more susceptible to frequent( every hundred or occasionally more years) Black Swans because of its geographical unity and the uni-dimensional nature of the threat from nomads.
An unanswered question is if there is an ongoing cultural difference between the East and West that will shape the future development of China vs. the US?
The footnote on historical climactic variability predicting current trust levels implies geography impacts culture over centuries.
I think there are definitely some cultural differences that will impact the future paths of different regions, peoples, etc., especially if they try to stay distinct, like China seems to be doing.
Thank you so much for keeping these posts free! Your writing has definitely been played an important role in my thinking around the kinds of questions I would like to study in my future career.
The whole Asian despotism thing goes back to the Greeks vs Persians. As a lot of even contemporary Greeks noted, despotism had a specific political implication. Namely, absolute (arbitrary) rule with no enforceable constitution (which can't be changed on the rulers whim).
Even amongst the Greeks, lots could see that the Persians were relatively benign rulers and that their state functioned very well for their people. The actual lived experience of Persians vis a vis their state may have been less onerous than that of a Greek citizen to their city state. But they were the ruled, not citizens.
And so you get a situation where extremely despotic societies like the Spartans are on team freedom because you've defined freedom in a certain way.
-Population density explains patterns of state formation better than anything else. Where there are dense populations, we see states form, even in totally isolated places like Rwanda.
-Population distribution is sufficient to explain patterns of Chinese unification. The North China Plain will naturally tend to be dominated by a single state; any state which dominates the North China plain will be in a good position to dominate the smaller population centers around it. The Yangzi is enough of a natural barrier that a strong state based to its south could hold out against a united north for a long time; but a state controlling the northern plain still had an advantage.
-The Manchus were not steppe nomads, and during periods of turmoil China was vulnerable to decapitation strikes by sedentary states. In a plausible alternate timeline, the Sengoku period in Japan lasts another ~40 years, alt-Hideyoshi's invasion of Korea coincides with Li Zicheng's rebellion, and the Qing Dynasty is Japanese.
-Chinese unification made each cycle of collapse worse in raw human terms, as it put the whole subcontinental region into turmoil at the same time; but China never suffered a full regression of urban civilization like what happened in Western Europe after the fall of Rome. Are there any examples of major technologies or productive skills that China lost over time? I don't think the Song lacked any technologies or skills the Han had.
-The Industrial Revolution was a weird thing that required very specific technologies that did not exist anywhere on Earth before the Scientific Revolution; it wasn't just a later stage of the agrarian civilization tech tree. We wouldn't expect an Industrial Revolution anywhere without a culture of scientific inquiry for its own sake.
1. This is a level of confusion I don't rightly know if I can correct. Please think about how you get to dense populations in the first place and about how you need mechanisms to explain state formation.
2. The population distribution is insufficient to explain patterns of Chinese unification, as the article should have made apparent. You cannot just say 'X will naturally happen' without some mechanism and expect such an empty statement to float. The population distribution also follows from state actions. Please consider the history here.
3. No one said the Manchu were steppe nomads and what substance is there to this "decapitation strikes" notion?
4. Unification did not make each cycle of collapse worse in raw human terms. You're thinking about population growth doing that. As regards putting the region into turmoil, sure, every dynasty's fall did that, but if you read the article and follow the graphs, you'll perhaps notice that population size didn't augment re-unification timing. And yes, there are plenty of examples of technologies being lost, rediscovered. As regards the Song, this is true plenty of times for weaponry alone.
5. The technology required for the Industrial Revolution existed in Rome, it was just a children's toy and apparently no one with the know-how and gumption thought to scale it up. Perhaps I was unclear in the article though, because I use the Industrial Revolution synonymously with "modern economic growth." In one of the sections I've pushed to the appendix, I discuss where economic growth began in China in reference to the advent of the Gongsuo and Chinese resistance to railroad construction.
"the Spanish collected 275% more revenue per head in 1700 than China did in the same year"
It would be interesting to know if this includes its American colonies. I'm guessing it doesn't. The Spanish Empire struggled with tax collection in its colonies, and encountered fierce opposition when it sought to increase taxes:
> In 1700, England, France, the Dutch Republic, and Spain combined had a population that wasn’t as large as China’s, but together, they collected 14% higher total tax revenue. By 1750, they had 3% higher total revenues [...] These ratios became more extreme with time
3% is less than 14%. Was that a typo and the former was supposed to have another digit?
Regarding the advent of agriculture, pretty much all at once in many places, prior to the Holocene, CO2 levels were very low, perhaps too low to make agriculture productive enough to be worth doing. With warming, CO2 levels rose to levels that permit productive agriculture. Prior to the Holocene, agriculture was a waste of time. A simpler explanation.
There's an even simpler (and more robust) explanation for why Europe industrialized while China didn't: The Chinese language is logographic and could not be printed (with Gutenberg-level tech)
If you look at a graph of European vs. Chinese innovations, the point of divergence aligns with 1450 almost to a T.
The printing press hypothesis also explains why Middle Eastern civs stagnated alongside China. Arabic contains many special cursives and notations that could not be accurately printed. And instead of revamping the language to fit the printer, the Ottoman empire's scribe guild outright banned the tech.
>“These locations did not adopt agriculture from one another, were not engaged in long-distance trade with each other, and weren’t even related by conquest; there was no cultural exchange that explains how all of these places took to the plough and the field simultaneously, and yet they did.”
They did have contact, and we know that because the first domesticate, the dog. There's a lot of debate, but the dog was domesticated 40-15kya in Eurasia. So you can tell a story about a single culture going from 0 to 1 (a la Thiel) by domesticating dogs 15 kya in Siberia, then that domestication-capable culture spreading over the whole world.
Additionally, there is a lot of support for religion being a prerequisite for agriculture as argued in anthologies like Religion in the Emergence of Civilization.
"For about 140,000 years before the start of the Holocene, anatomically modern humans lived in small groups of relatively mobile hunter-gatherers. Then in a relatively short time after 12,000 BC, human groups began to settle down, adopt agriculture and take many of the steps that we associate with ‘civilization’. The reasons given for this shift have predominantly been climatic change, population increase and economic and ecological factors, although social and cognitive factors have increasingly been included (Bender 1978; Hayden 1990; Renfrew 1998). The aim of the proposed study is to explore the extent to which spiritual life and religious ritual played a role in this momentous shift."
Of course the religion could have been an independent development, but it could also have spread at the end of the Ice Age, as a number of anthropologists have argued. The best evidence for that is the bullroarer, which is an instrument used in men's mystery cults the world over in very similar ways, and with very similar surrounding mythology. Specimens have been found at Gobekli Tepe and Kortik Tepe, which just precede agriculture in the Near East.
A couple caveats. It would be interesting (maybe demanding a post) to look at the earliest evidence of the domesticated dog near each center of agricultural development. Is the dog always first? I think that holds but I'm not 100%.
Second, some of the disagreement is definitional about what counts as "independent." If, say, the technology of calendars spread and then that allowed domestication, is that independent? What about dogs and religion?
> In the premodern era, populations would technologically progress as they grew, but when they shrank, living conditions frequently worsened enough that people would be forced to give up using, working on, and transmitting newly-learned techniques and newly-minted technologies in favor of simple farming, and the knowledge related to those things would be lost to subsequent generations. Likewise, the demand for new technologies and techniques would fall, and those who knew them would fail to transmit them to the next generations because there’s no time or need. When those subsequent generations reversed the declines that caused people to drop new technologies, they wouldn’t be able to just pick them up again, so their productivity growth rate over the years would almost-certainly have been negatively impacted relative to the counterfactual where the division of labor hadn’t shrunken.
This has disturbing implications for the future considering the state of global fertility.
Richard Pipes theorized that societal despotism is primarily a function of the strength or weakness of their property rights. With China being such a large realm, land was plentiful, and thus had less value than a realm where land was scarce, such as Japan.
I wonder if their weakly developed property rights also disinventivized them from innovation. Admittedly, Japan was rather backwards until the Perry Expedition shook them out of isolation. But this was primarily due to depriving themselves of the global community of knowledge. They still grew in GDP even under the isolation period, and had been growing steadily well prior to such, too:
Had they not cut themselves off like they did, there's a strong chance they would've been the pre-eminent power of the world: strong property rights, a culture of hard work and innovation, the highest average IQ of any country on Earth, a fairly large population, damn near endless port cities. Who could've stopped them?
Very cool post! I was going to mention Scheidel’s work, but I see it’s already been mentioned in the comments.
I do wonder about the Koyama et al model, though. It takes the existence of and conflict between political communities as a given at one scale, while treating the formation of political communities as endogenous at another scale (ie, within a continent). That seems odd, and it makes me wonder why the different treatment within vs outside of a continent is justified. This doesn’t necessarily mean the model isn’t useful, it’s just a puzzling feature for me.
This was a superb read, thank you. Are you by chance familiar with Scheidel's Escape From Rome? He deals with much overlapping themes of European comparative prosperity and China's hardships with centralisation and steppe hordes.
The book argues that the fall of the Roman Empire and the resulting political fragmentation in Europe was the best thing that ever happened to us. It fostered competition and innovation, leading to the region's eventual economic and technological sucaess. Unlike centralised states like China, Europe's divided states encouraged military and technological advancements, preventing monopolisation of power. This fragmentation laid the groundwork for capitalism, scientific progress, and modern political institutions, making the collapse of Rome a crucial turning point for Europe's prosperity.
I found many, many same things mentioned in this post as was in the book.
Big fan of Escape from Rome. Scheidel's work gets referenced in a few of the papers I cited here, so it's also definitely been a big indirect influence on this post.
Yes, that is a great book.
If you enjoyed Scheidel’s book, you might also enjoy my article on the topic:
https://frompovertytoprogress.substack.com/p/was-the-fall-of-the-roman-empire
I couldn't help but be reminded of Taleb's theory of anti-fragility. It seems The West was more anti-fragile because of its diverse geography and the multi-variate threats against it, while China was bigger and better organized but ultimately more susceptible to frequent( every hundred or occasionally more years) Black Swans because of its geographical unity and the uni-dimensional nature of the threat from nomads.
Excellent write up. A good complement to the book Why the West Rules, for Now (https://www.amazon.com/Why-West-Rules-Now-Patterns-ebook/dp/B003VTZSFY/) well.
An unanswered question is if there is an ongoing cultural difference between the East and West that will shape the future development of China vs. the US?
The footnote on historical climactic variability predicting current trust levels implies geography impacts culture over centuries.
Also, yes, raiding barbarians suck.
I think there are definitely some cultural differences that will impact the future paths of different regions, peoples, etc., especially if they try to stay distinct, like China seems to be doing.
Thank you so much for keeping these posts free! Your writing has definitely been played an important role in my thinking around the kinds of questions I would like to study in my future career.
Excellent Post!
The whole Asian despotism thing goes back to the Greeks vs Persians. As a lot of even contemporary Greeks noted, despotism had a specific political implication. Namely, absolute (arbitrary) rule with no enforceable constitution (which can't be changed on the rulers whim).
Even amongst the Greeks, lots could see that the Persians were relatively benign rulers and that their state functioned very well for their people. The actual lived experience of Persians vis a vis their state may have been less onerous than that of a Greek citizen to their city state. But they were the ruled, not citizens.
And so you get a situation where extremely despotic societies like the Spartans are on team freedom because you've defined freedom in a certain way.
-Population density explains patterns of state formation better than anything else. Where there are dense populations, we see states form, even in totally isolated places like Rwanda.
-Population distribution is sufficient to explain patterns of Chinese unification. The North China Plain will naturally tend to be dominated by a single state; any state which dominates the North China plain will be in a good position to dominate the smaller population centers around it. The Yangzi is enough of a natural barrier that a strong state based to its south could hold out against a united north for a long time; but a state controlling the northern plain still had an advantage.
-The Manchus were not steppe nomads, and during periods of turmoil China was vulnerable to decapitation strikes by sedentary states. In a plausible alternate timeline, the Sengoku period in Japan lasts another ~40 years, alt-Hideyoshi's invasion of Korea coincides with Li Zicheng's rebellion, and the Qing Dynasty is Japanese.
-Chinese unification made each cycle of collapse worse in raw human terms, as it put the whole subcontinental region into turmoil at the same time; but China never suffered a full regression of urban civilization like what happened in Western Europe after the fall of Rome. Are there any examples of major technologies or productive skills that China lost over time? I don't think the Song lacked any technologies or skills the Han had.
-The Industrial Revolution was a weird thing that required very specific technologies that did not exist anywhere on Earth before the Scientific Revolution; it wasn't just a later stage of the agrarian civilization tech tree. We wouldn't expect an Industrial Revolution anywhere without a culture of scientific inquiry for its own sake.
1. This is a level of confusion I don't rightly know if I can correct. Please think about how you get to dense populations in the first place and about how you need mechanisms to explain state formation.
2. The population distribution is insufficient to explain patterns of Chinese unification, as the article should have made apparent. You cannot just say 'X will naturally happen' without some mechanism and expect such an empty statement to float. The population distribution also follows from state actions. Please consider the history here.
3. No one said the Manchu were steppe nomads and what substance is there to this "decapitation strikes" notion?
4. Unification did not make each cycle of collapse worse in raw human terms. You're thinking about population growth doing that. As regards putting the region into turmoil, sure, every dynasty's fall did that, but if you read the article and follow the graphs, you'll perhaps notice that population size didn't augment re-unification timing. And yes, there are plenty of examples of technologies being lost, rediscovered. As regards the Song, this is true plenty of times for weaponry alone.
5. The technology required for the Industrial Revolution existed in Rome, it was just a children's toy and apparently no one with the know-how and gumption thought to scale it up. Perhaps I was unclear in the article though, because I use the Industrial Revolution synonymously with "modern economic growth." In one of the sections I've pushed to the appendix, I discuss where economic growth began in China in reference to the advent of the Gongsuo and Chinese resistance to railroad construction.
"the Spanish collected 275% more revenue per head in 1700 than China did in the same year"
It would be interesting to know if this includes its American colonies. I'm guessing it doesn't. The Spanish Empire struggled with tax collection in its colonies, and encountered fierce opposition when it sought to increase taxes:
https://en.wikipedia.org/wiki/Revolt_of_the_Comuneros_(New_Granada)
> In 1700, England, France, the Dutch Republic, and Spain combined had a population that wasn’t as large as China’s, but together, they collected 14% higher total tax revenue. By 1750, they had 3% higher total revenues [...] These ratios became more extreme with time
3% is less than 14%. Was that a typo and the former was supposed to have another digit?
No typo. The delta momentarily shrank.
Interesting. Thank you.
Regarding the advent of agriculture, pretty much all at once in many places, prior to the Holocene, CO2 levels were very low, perhaps too low to make agriculture productive enough to be worth doing. With warming, CO2 levels rose to levels that permit productive agriculture. Prior to the Holocene, agriculture was a waste of time. A simpler explanation.
Plant productivity 50% lower under low-CO2:
https://nph.onlinelibrary.wiley.com/doi/full/10.1111/j.1469-8137.2010.03441.x
There's an even simpler (and more robust) explanation for why Europe industrialized while China didn't: The Chinese language is logographic and could not be printed (with Gutenberg-level tech)
If you look at a graph of European vs. Chinese innovations, the point of divergence aligns with 1450 almost to a T.
https://x.com/x_spinoza/status/1830511573754114245
The printing press hypothesis also explains why Middle Eastern civs stagnated alongside China. Arabic contains many special cursives and notations that could not be accurately printed. And instead of revamping the language to fit the printer, the Ottoman empire's scribe guild outright banned the tech.
>“These locations did not adopt agriculture from one another, were not engaged in long-distance trade with each other, and weren’t even related by conquest; there was no cultural exchange that explains how all of these places took to the plough and the field simultaneously, and yet they did.”
They did have contact, and we know that because the first domesticate, the dog. There's a lot of debate, but the dog was domesticated 40-15kya in Eurasia. So you can tell a story about a single culture going from 0 to 1 (a la Thiel) by domesticating dogs 15 kya in Siberia, then that domestication-capable culture spreading over the whole world.
Additionally, there is a lot of support for religion being a prerequisite for agriculture as argued in anthologies like Religion in the Emergence of Civilization.
"For about 140,000 years before the start of the Holocene, anatomically modern humans lived in small groups of relatively mobile hunter-gatherers. Then in a relatively short time after 12,000 BC, human groups began to settle down, adopt agriculture and take many of the steps that we associate with ‘civilization’. The reasons given for this shift have predominantly been climatic change, population increase and economic and ecological factors, although social and cognitive factors have increasingly been included (Bender 1978; Hayden 1990; Renfrew 1998). The aim of the proposed study is to explore the extent to which spiritual life and religious ritual played a role in this momentous shift."
Of course the religion could have been an independent development, but it could also have spread at the end of the Ice Age, as a number of anthropologists have argued. The best evidence for that is the bullroarer, which is an instrument used in men's mystery cults the world over in very similar ways, and with very similar surrounding mythology. Specimens have been found at Gobekli Tepe and Kortik Tepe, which just precede agriculture in the Near East.
A couple caveats. It would be interesting (maybe demanding a post) to look at the earliest evidence of the domesticated dog near each center of agricultural development. Is the dog always first? I think that holds but I'm not 100%.
Second, some of the disagreement is definitional about what counts as "independent." If, say, the technology of calendars spread and then that allowed domestication, is that independent? What about dogs and religion?
Surely there are many causes for the agricultural revolution, and perhaps Jupiter is among them. But cultural diffusion is criminally underrated. There is surprisingly good evidence for it! https://www.vectorsofmind.com/p/evidence-for-global-cultural-diffusion
> In the premodern era, populations would technologically progress as they grew, but when they shrank, living conditions frequently worsened enough that people would be forced to give up using, working on, and transmitting newly-learned techniques and newly-minted technologies in favor of simple farming, and the knowledge related to those things would be lost to subsequent generations. Likewise, the demand for new technologies and techniques would fall, and those who knew them would fail to transmit them to the next generations because there’s no time or need. When those subsequent generations reversed the declines that caused people to drop new technologies, they wouldn’t be able to just pick them up again, so their productivity growth rate over the years would almost-certainly have been negatively impacted relative to the counterfactual where the division of labor hadn’t shrunken.
This has disturbing implications for the future considering the state of global fertility.
Richard Pipes theorized that societal despotism is primarily a function of the strength or weakness of their property rights. With China being such a large realm, land was plentiful, and thus had less value than a realm where land was scarce, such as Japan.
I wonder if their weakly developed property rights also disinventivized them from innovation. Admittedly, Japan was rather backwards until the Perry Expedition shook them out of isolation. But this was primarily due to depriving themselves of the global community of knowledge. They still grew in GDP even under the isolation period, and had been growing steadily well prior to such, too:
https://academic.oup.com/ssjj/article/23/2/147/5912248?login=false
Had they not cut themselves off like they did, there's a strong chance they would've been the pre-eminent power of the world: strong property rights, a culture of hard work and innovation, the highest average IQ of any country on Earth, a fairly large population, damn near endless port cities. Who could've stopped them?
Very cool post! I was going to mention Scheidel’s work, but I see it’s already been mentioned in the comments.
I do wonder about the Koyama et al model, though. It takes the existence of and conflict between political communities as a given at one scale, while treating the formation of political communities as endogenous at another scale (ie, within a continent). That seems odd, and it makes me wonder why the different treatment within vs outside of a continent is justified. This doesn’t necessarily mean the model isn’t useful, it’s just a puzzling feature for me.
Alternative title: One Small Step for Jupiter; One Giant Leap for Mankind.
Typos:
>Interstate competition is robust because is there are threats
>Third, check out this cool paper that argues that argues historical...