Only the Dead at Dawn (Part 2/2)
The second part of my Malthusian overview of human history. Covering the earliest states to the modern world.
[Part 1 argued that:
1.) tribal warfare emerged as a response to limits on food production and pre-dates the evolutionary split between humans and chimpanzees.
2.) agriculture emerged as a response to those conflicts over resources.
We left off after farming had replaced hunter-gathering in a few select regions, and had liberated early farmers from the Malthusian pressures that had constrained hunter-gatherer populations.]
V. States
Initially farmers would be free from the Malthusian pressures that had fuelled war since the dawn of time, they’d have plenty of effectively un-inhabited hunter-gatherer territory to expand into.
War between early farming societies would have been too destructive to be worthwhile, when so much free land was available to absorb the excess population. A truce would reign between groups of farmers, and individual farmers would at least have plenty of soil to cultivate.
Wars would occasionally be waged to clear more land of hunter-gatherers for farming colonies to spread into.
But a small area of hunting land can support many farmers. population growth would also have be slow as food surpluses would be very marginal, the early domesticated crops were nearly-wild plants with only small edible parts (see the ancestor of corn, teosinte, or the ancestors of any modern crop plants). Most farmers struggled to feed their children.
The bottle neck in food production, and therefore population growth, was labour (which scales with population, obviously) not land, so populations grew slowly.
In battle farmers would hugely outnumber hunters and have other advantages like specialist war equipment, such as shields and armour, which nomads can’t carry around all the time. They wouldn’t have more time to dedicate to war though.
Historians often say farming “provided a surplus that enabled specialist and non-food producers to make crafts and produce knowledge etc.”
That’s wrong, as we’ve seen hunters generally sustain themselves with comparatively little effort (less even than modern people), but have to fight for the privilege. The average hunter has much more free time to do things not directly necessary for survival, more “surplus”, than the average farmer who originally needed to work as long as possible just to stay alive and, as we’ll see, never escapes that poverty.
This stage of history played out recently enough in America and Australia to be well documented. Many Native American cultures were already farming in 1492, or were non-nomadic hunter-gatherers who lived in dense settlements around especially rich natural resources, like fisheries or herds of bison.
Many anthropologists have studied those cultures and made inferences about ancient hunter gatherers cough based on them, like in Dawn of Everything (a less than excellent book), but they’re essentially like farmers in their material conditions, you should expect their social structures to be more farmer-like.
When European farmers did encounter real hunter gatherers they annihilated them easily, due to the advantages of farming (admittedly they did also have guns).
Crucially though, these genocides were usually carried out by voluntary militias, not armies. Farmers exterminating hunters do so as individuals and without needing any encouragement, typically to get the land (and frankly probably because they enjoyed it), which they claim for themselves. There’re no states organising war at this stage of civilisation.
States arise when famers run out of new land to colonise and Moloch turns them against each other.
Archaeologists often remark on the period between 10,000BC (when farming starts) and 5,000BC in Mesopotamia (where farming starts), they note there aren’t any palace complexes, temples, human sacrifices, or any of the other signs of hierarchy that become common after 5,00BC. There also aren’t many city walls.
Often societies in this period are described as “anarchist”. I think that’s probably true. Maybe there were weak civil institutions (for the common good, naturally), but farmer’s still had open frontiers in the early stages of the agricultural revolution, it’s difficult to subjugate people if they can just leave to settle somewhere away from you. You won’t see anything like the Ancien Regime or contemporary America.
Also, I’ve emphasised that farming had low productivity to start with, selecting crops to be more edible, and better farming techniques, would improve yields over time. But early farming societies have low surpluses, you need almost everyone to till the earth or there’ll be famine, they’re simply not rich enough to support parasitic elites.
Once the land runs out though, peasants have nowhere to go, but more importantly there’s nowhere to absorb the excess population. Malthusian conditions always return in the long run, and, probably for the first time ever, agricultural societies need to fight each other regularly, and to the death.
War between farmers is completely different to war between hunters. The number of people involved is exponentially greater and it involves fortifications. The old raiding-and-ambushing-lone-individuals-and-small-groups tactics, that have been war’s only form for millions of years, no longer work.
You can’t ambush thousands of farmers one by one, and you can’t raid them if they have walls, lookout towers and a lot of people everywhere. The only forms of fighting that can curb the excess population fast enough are pitched battles and sieges, the kind you probably think about when you think about ancient wars. Hundreds of men lining up in the open and clashing lines, or storming city walls with ladders.
My advice is that it’s a bad idea to take part in a pitched battle (hopefully that’s not too controversial). This isn’t the same as the chimps and hunters dispatching defenceless enemies from a place of safety.
There’s no way for 1000 guys to kill another 1000, or even 100 guys, without taking casualties. The scale of armies prevent the instant-debilitation tactics that you need to dispatch an opponent safely, there’re too many enemies to surprise or neutralise before their comrades can retaliate.
Armies always pose at least some threat to each other, and since battle usually take place by mutual consent (both armies “agree” to fight there) they’re usually fairly evenly matched.
That makes battles absolutely terrifying, way scarier than the almost sport-like wars that chimps and hunters fight. Since the winners often massacres the losers, your chance of surviving an average battle aren’t much better than 50%. No one will do that voluntarily.
For a sense of how unpleasant war between agrarian states could get: when the Goths were displaced from their former territories by Hunnic invaders, in 376AD, they requested asylum from the Roman Empire, who denied the request but, as the Goths began to starve, they did offer to sell them dog meat in exchange for their children as slaves.
Moloch is driving the farmers to war, but if they want to fight effectively they need a way to coerce their soldiers to risk their lives, and a way for non-combatants to subsidies the soldiers.
The answer is obviously states, and eventually money. Creating the earliest conscription and professional soldiers (probably you’d call them warrior elites).
Massed armies don’t appear in history before this period.
A real version of the Wildling horde that attacks The Wall in a Song of Ice and Fire a) wouldn’t be able to supply itself (except maybe if they had reindeer herds, but we never see any), and b) would have no central authority to organise the army and coerce the warriors into risking their lives for the common good.
This theory of state formation predicts nascent farming societies on smaller landmasses should evolve into states more quickly than ones on larger landmasses (because there’s less space to expand into). Mesoamerica is a small geographic area (compared to the arable parts of Eurasia), the only neighbouring lands are to the north and south and, as Jared Diamond famously pointed out, it’s difficult for farming to spread north or south (because the climate changes more than it does east to west), so early Mesoamerican farming was effectively confined to a small area, roughly modern-day Mexico.
Did civilisation progress faster in the Americas than in Eurasia? The earliest urban civilisation, the Olmec’s, begins about 4500 years ago, only 2500 years after farming fist originated in the Americas, around 5000BC.
The fist Mesopotamian civilisation, Sumer, doesn’t emerge until around 5500BC, maybe as late as 4000BC, about 5000 years after the development of agriculture around 10,000BC.
Writing also emerges in about half the time, about 4000 years after farming in Mesoamerica, versus around 8500 years after farming in Mesopotamia.
Which is consistent with the theory that limited space shortened the time it took to reach carrying capacity for Mesoamerican farming populations, which hastened the onset of Malthusian pressures which then pushed them through the agricultural-but-anarchic phase of civilizational development about twice as fast.
It’s around this period, about halfway through agricultural history, that we see the first widespread human sacrifices in Mesopotamia and Egypt (there are older archaeological finds that could be human sacrifices but much less common, much smaller scale, and less clearly sacrificial), similar to what conquistadors would later document in Mesoamerica, clear displays of society’s ownership of human lives, and totally ad odds with the personal freedom hunter-gatherers enjoy.
In War in Human Civilisation, Azar Gat argues (compellingly imo) that the first cities were fortified settlements for the surrounding farmers to shelter in from raids by other, neighbouring farmers. Probably the first states were city-states like Uruk, which conquered much of Mesopotamia from the mid-4th millennium BC onwards. Perhaps later on they developed economic functions (although whether they ever were useful economically is still disputed) but cities, states and civilisation in general where all products of humanities perennial arms race.
In their project to strip back government services, even Libertarians acknowledge the military as one of the most necessary functions of a state, it stand to reason that if other functions are optional, armies are likely one of the most ancient and basic reasons sates came into existence.
VI. Malthus Returns
Until this point things had been improving for farmers, as their crops became more nourishing and produced higher yields, livestock-animals were domesticated, and land was still bountiful. Likely the average farmer in 5000BC was substantially richer than the earliest farmers had been, although life would still have been austere.
From here onwards though things only get worse. Farmers had bought a brief respite from the Malthusian world, by working long hours and eating a monotonous diet (plus a bit of genocide), but inevitably Malthusian conditions reasserted themselves, only this time at a lower base level of wellbeing. The grains gains farmers had bought themselves eroded away as people filled the available land, and, the now much larger, competing groups of people faced a new ultimatum: limit population growth, either by war, hunger, or pestilence.
As mentioned, people don’t enjoy this new kind of war, and decision making is centralised in states, so wars in this period don’t happen by default, a leader needs to initiate them; which they’re reluctant to because wars are risky and always incur costs. The overall level of violence per capita is lower than it had been for hunter-gatherers, which means famine and pestilence take up the slack in necessary deaths.
In the hunter-gatherer world population density was too low to sustain many contagious diseases, contact between people was too intermittent and there weren’t sufficiently large natural reservoirs for pathogens to jump between hosts before their current host’s immune system destroyed them or their current host died. More importantly though, most contagious diseases hadn’t evolved yet, or haven’t jumped over to humans. Most modern contagious diseases come from animals, as covid19 reminded us, mainly animal humans have a lot of contact with i.e. domesticated animals.
Contagions killed astronomical numbers of people after the advent of agriculture, and in the most horrible ways (DO NOT google small pox patients). New pathogens regularly jumped from livestock to humans and spread through the dense populations in waves. If you have an autoimmune disorder, it’s likely because your ancestors were subject to this selection pressure.
If you survived the epidemics and wars, the next Malthusian pressure trying to kill you would have been an old-fashioned shortage of land. Because rates of violence are low within states and between them (compared to between bands of hunter-gatherers) and because epidemics were sporadic, the main limit on population growth in this period is a shortage of food. Populations will grow until they reach a limit, if that limit is food consumption then average food consumption will tend to the minimum quantity people can survive on, a subsistence diet.
Farmers would inexorably be forced onto smaller and smaller plots of land, which they’d need to cultivate more and more intensively, to the point where they’re barely surviving and any children they have, above two per family on average, die off from malnutrition.
Here’s a graph of population and real wages (roughly indicative of food availability) in England, from How the World Works (highly recommended). Before the industrial revolution, wages and population are the the inverse of each other, wages rise substantially after the population collapse after the Black death in 1348, then fall again afterwards as the population grows.
Food consumption for the bulk of the population can’t rise much beyond a subsistence level, but economic output per capita as a whole can still rise under certain conditions, a subtle point many economic historians overlook when they discuss average wages or average consumption.
(Chart from Farewell to Alms)
For early farmers food production is limited by labour productivity. There’s no shortage of land, they can sow seeds 24 hours a day, the only limits to returns is the yield per seed (which is low), so even if other farmers die off, and competition for land weakens, it won’t raise living standards, because it won’t free up any labour (the bottleneck).
For farmers under states food production is limited by land productivity. They have a limited area of land, often much less than they could potentially cultivate with the labour time available (theoretically 24 hours per day). They have to invest the excess labour into improving yields per unit area of land, instead of just cultivating more earth, to get more output from the same plot. Things like weeding, ploughing and fertilising, which have diminish returns to labour inputs compared to just casting lots of seeds over naturally fertile soil (obviously not still true with modern tractors and chemical fertilisers). Farmers will be forced to use labour intensive production methods even if the extra returns are marginal.
If a technology improves land productivity, like fertilisers do for example, or farming itself did, by raising yields (the number of seeds you get back per seed sown), the increased food production will only raise food consumption and living standards temporarily. In the long term any improvement in yields loosens the constraints on population growth and the new people will compete for land until average plot sizes, and therefore per capita food production, are pushed back down to subsistence level. Typically this means a reduction in average economic consumption (for all the things that aren’t food), because now people need to invest more labour into exploiting the new technology, manufacturing fertilisers, digging up weeds, building fences etc., only a smaller plot of land, so they have less labour to devote to non-food production. It’s the same phenomenon as the initial drop in living standards farming caused in miniature. Any increase in food production will only feed population growth.
However, if a technology improves labour productivity, like tractors or ploughs for example, without raising land productivity, then food production doesn’t increase, it’s just easier to produce. If food production doesn’t rise then neither will the population. The extra labour that gets freed up by technology can then be devoted to any type of production that isn’t bottlenecked by land, like mining, manufacturing, or construction. Labour saving devices on the farm increase per capita economic consumption.
The Roman Empire was an example of a high population density, high yields per unit area of land, grain fed, lots of labour saving capital, labour-rich, land-poor economy. The average Roman was only 5’5’’, because of a poor diet, but Rome was rich in the material goods the excess labour could produce, infrastructure, weapons and armour, ships etc. Eventually Rome ends up with a large population of landless, sometimes jobless plebs, sustained on a dole of imported Egyptian grain.
By contrast many Germanic tribes held enough land that they could afford to graze cattle, and so had better nutrition and were much closer to modern people in height, Saxons averaged 5’8’’, but lacked a land-less surplus population, and so were materially poorer.
From here we can see why Europe, and especially Britain, has historically been wealthier per capita than East Asia.
It’s become a bit of a platitude among anthropologists that “we didn’t domesticate plants, they enslaved us.” Which is a bit too abstract to really mean anything, but if plants are our slave masters then the cruellest master has to be rice.
A rice crop yields about 11 million calories per acre compared to 6 million for wheat (unfortunately this is all modern data, ancient yields would have been much lower).
This meant rice growing societies had higher population densities than wheat growing ones, raising competition for land and all other finite (ordinal) resources. Asian peasants would have had small but highly productive growing plots and lacked access to most things that occurred naturally, like coal or building materials, because of rice’s high yields per acre.
The thing that’s most terrible about rice however, is that it takes a lot of labour to cultivate and harvest.
Which mean there’s less labour available to produce all the things that don’t occur naturally as well, like cloth or metals.
Another dreadful thing about rice is that you can grow it almost anywhere (with a warm, wet climate). You can convert hillsides and arid land into paddies by constructing flooded terraces. This means rice farming displaces almost all other uses for land that don’t produce food. In a European village you might see some woodland or a flock of sheep on hillsides, providing firewood or wool. You might even see cattle pastures on non-arable lowlands, providing meat and dairy with almost no labour inputs.
(I’m not completely sure if it would actually be possible to grow ancient wheat on terrain like this, if it is, it implies the higher population density caused marginal land to be brought under cultivation instead of the other way round, but I think the yields would be too low since they’re already fairly low in ideal conditions.)
(Warning: pointless digression)
[Incidentally, and this is especially conjectural, historians have long puzzled over why Britain has been an unusually wealthy part of Europe for so long, probably since at least the middle ages (and until Churchill threw that away); Adam Smith estimated average incomes in England were twice as high as in France. The East and Midlands of Britain is flat, arable land, the land here is extremely suitable for farming, naturally it developed a dense population with large urban centres (London, Birmingham, all the port towns etc.). A lot of the rest of the island is highlands that’s only suitable for raising sheep, and there also used to be a lot of bogs, marshes and ground too wet to till, that can be suitable for raising cattle. Perhaps that combination of terrain created the conditions for a developed urban centre, that would have been impoverished on its own, but was enriched by an inflow of low labour cost goods (beef, and especially wool and wood) from the regions that couldn’t grow wheat. By comparison, France for example, is mostly arable low lands.]
China had virtually no draught animals, oxen or horses, and very few food animals, the Chinese peasant diet was almost entirely rice, and without manure they had to resort to collecting sewage from the cities to fertilise crops. Human waste was big business, land lords often collected rents in it and mandated that tenants and workers were obliged to use designated toilets they could collect from.
R.H Tawney, the British economic historian, wrote after a visit to China in the late 1920s that the precariousness of Chinese agriculture was such that: “There are districts in which the position of the rural population is that of a man standing up to the neck in water. So that even a ripple is sufficient to drown him”
The fact nutrients needed to be collected from toilets to be fed back into the system, illustrates the extreme ecological limits population pressures would push agrarian economies to.
Pre-industrial China is an ocean of small rice paddies worked by immense numbers of emmiserated peasants. Malthusian farming reduced almost the whole world to a wretched existence, but in Asia things were about as bad as they can possibly get.
In A Farewell to Alms Gregory Clark demonstrate that livening standards were stagnant in Europe throughout the Malthusian period (from the early medieval period to the Industrial Revolution, except after the Black Death), and argues that long-term per capita income growth is impossible under Malthusian conditions.
My view isn’t quite so fatalistic, clearly incomes differed between societies, especially between the West and the East. I think the reason per capita consumption appears to stay constant at a local level, despite technological advances, must be because technologies improving land productivity were cancelling out the gains from labour improving ones. The main factor that determined the base rate of living standards was the labour requirements for different kinds of crops, relative to their yield per unit area of land. Paradoxically more productive crops (in terms of land) produced lower living standards because of the higher bound on population size they enabled. Wheat growers weren’t as poor as rice growers, but, almost everywhere that people had farmed for long enough, life was nasty and short. (The exceptions are places where disease or endemic war acted as the constraint on growth; things were better in places like Polynesia or, surprisingly, West Africa.)
VII. Why we’re here and where we’re going
You can probably pick holes in this narrative, I’m sure people are shouting “Neanderthals had weapons and small groups!”, “What about Bonobos?” and “We already have a theory of state formation!” Malthusianism wasn’t the rule everywhere and in all times, but the basic insight that the pre-industrial world was often caught in a Malthusian trap, and that it’s the fundamental cause of war, or at least of the instinct for war, and that it condemned pre-industrial people to subsistence poverty regardless of technological progress, as a general rule seems very true.
If human history has an underlying structure it’s one driven by violent conflict over the means of subsistence.
[This section gets a bit philosophical, maybe read it in the evening after a couple of drinks. Also the only philosophy I’ve read is Schopenhauer.]
War looks like a very strong candidate for the primary driver of human evolution and the development of civilisation.
In a way, this is also the foundation of my world view. It’s not the most optimistic outlook, but I’m convinced the natural state of replicating organisms is a bleak struggle for resources, and that the task of civilisation is to overcome the coordination problems and zero-sum games that nature pushes us towards.
Really bad books, like Sapiens, want to claim humanity is special because we “tell stories” or [insert vague bs about “what separates us from the animals” here], in reality it was war that drove our evolution and the development of civilisation, and war is just another version of zero-sum competition, the same thing that drives all evolution.
And the key development in that evolution, the thing that ignited the whole process, was coalitions, which evolved before our split with the chimpanzees; like them we’re fundamentally war-apes. It’s amusing to think that the most momentous event in human history, or at least the first, was probably just learning to hit other people with sticks, all the other developments, up to nuclear weapons, followed inexorably from there.
A struggle for survival between groups was our primary selection pressure since before the split with chimpanzees. It shaped many aspects of our psychology, social structures and history. Every questions in the social sciences should follow from that principle, if you read anything else on this blog you’ll find it permeates most of my thinking.
I want to emphasise that this view of human history is not Hobbesian. Life before civilisation may have been brutish, but it wasn’t nasty or short; and the core feature of civilisation, farming, was probably the worst thing to ever happen, even if it did rein in the violence inherent to the state of nature, perhaps even because of that.
In fact this view is at odds with a lot of common conceptions about civilisation and progress. Farming wasn’t a beneficial innovation in humanity’s long march of progress for example. Probably rudimentary forms of growing crops long predated the agricultural revolution, but were only relied on as a last resort. Farming was something people were driven to with antipathy, and facilitated by unconscious selection for more edible plants rather than inventing new ways of producing food.
Malthusianism created a world so unpleasant it was actually improved by war. People as modern (temporarily, not ethically) as Hitler and Churchill both described WW1, the quintessential hellish war, as the highpoint of their lives. It must have taken an extreme form of nihilism to think that, but reading about the standard of living in early 20th century Germany I find it completely understandable (as an aristocrat though, I’m not sure what excuse Churchill had).
The world recently passed 8 billion people in population. Certainly industrial society is not as Malthusian as agrarian society was, but it’s not clear that it’s entirely un-Malthusian. And if we are free from the population trap, we only escaped very recently. Only since about 100 years ago for most of the world, about roughly 1% of agricultural history, India still had famines as recently as the 40s and China as recently as the 60s.
Apart from the CCP, the modern world has taken a relaxed attitude to population growth (I’m looking at you Selfish Reasons to Have more Kids). But we’re talking about a law of nature that kept humanity enslaved for most of our existence, my instinct is to be fearful and that we’re playing with fire.
The natural state for mankind is a sad struggle to wrestle a worthwhile existence out of that merciless bitch Mother Nature, and usually at the expense of other men.
In the long run Malthusian condition are almost guaranteed to return, maybe in some future overcrowded by the religious with their high fertility rates, or some horrifying digital dystopia, like Robin Hanson’s Ems.
The main insight I’d like to emphasise is that farming was a massive shift in ecological niches for former hunter-gatherers, and how poorly adapted people are to the new environment it created. Hunter-gathering is our natural lifestyle, it provided freedom, and didn’t it ask much from people that nature hadn’t already equipped them to provide. Requiring people to live regimented lives, in alienated settlements, and to work long tedious hours doing unfulfilling work, in my view, is the fundamental cause of most of our modern social problems and discontent.
We paid a heavy price for that primordial Eden, the seeds of affluence were watered in blood, and it was a stagnant world that could never have led to something better. But compared to the deprivations of agricultural society it was a cost well worth paying. I often wonder if industrial society has yet managed to compensate us for the loss.
[Dramatic readings of Only the Dead at Dawn are available for children’s birthday parties and christenings.]