Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics

Modeling social dynamics (a long and detailed suggestion post)

A topic by shinshimon created 41 days ago Views: 154 Replies: 4
Viewing posts 1 to 2

Mechanical Turchin Social Engine:
This is based largely on Peter Turchin's work on social dynamics. If you're interested, recommended reading is at the bottom. Turchin's insights can be boiled down to two key points:
1. Agrarian societies follow regular, endogenous, overlapping cycles. The longest of these, based on Malthusian dynamics, Turchin calls secular cycles. The downturn in a secular cycle is accompanied by great social unrest, often including civil war.
2. The level of social cohesion, or *asabiya* (after Ibn Khaldun), is extremely important to the success of any human group. Asabiya builds in the face of threat from an outside enemy. Asabiya reduces in each downturn of a secular cycle, and has a general tendency to decline over the long term.

These phenomena are very pertinent to SotE. I will give a more extensive, but still brief, explanation of how each one works according to Turchin, and how I think they could be modelled by game mechanics in SotE.


Things are great for farmers as long as there is open land left to be cultivated. Every son can move out to found his own farm, and all can enjoy the same standard of living as their parents.

Things start to suck when you run out of new land to cultivate. If a family has 5 sons, there won't be enough land to go around, and each son will be living in poverty and precarity. If a bad year hits and he loses most of his crop, he will need to go into debt to feed his family. His farm will be the collateral. His creditor will likely be a nobleman, or another peasant who didn't have any siblings, and inherited a full-sized farm that can afford to loan out surplus.

If another bad year hits, the poor man loses his farm as collateral, and is now a landless laborer. The nobleman or rich peasant now has even more land. Iterate over a few generations, and you have a very unequal society, with a few very rich individuals, and a lot of landless poor. This situation is more or less inevitable once an agrarian society hits the Malthusian limit.

The poor keep getting poorer, because more landless laborers = more labor supply = lower wages. This usually leads to peasant revolts, but the upper classes, if united, can almost always put such a revolt down.

The upper classes, glutted with wealth (more land + lower labor costs), experience a brief golden age. This is when a lot of world-famous art is patronized. However, copious reproduction and ever-growing conspicuous consumption vastly increase the operating costs of the upper classes, while the agrarian economy that they skim off of isn't growing any more. Soon enough, the elites enter their own crisis. They need to spend like in the golden days in order to keep their status, but they are increasingly strapped for cash. They turn to the state for support, and turn on each other to monopolize state access among a sustainably small patronage network. In desperation, social cohesion falls apart. This is when you get a civil war, and possibly successful peasant revolts.

Conflict follows a generational "fathers and sons cycle", where one generation learns the horrors of war and eventually works towards peace, while their ignorant sons rekindle things one generation later.

This process continues until the underlying problem of elite overproduction solves itself, with fewer births (due to precarity) and more deaths (due to war) bringing the elite population back to healthy numbers. The elites have to die. If the literal Black Death hits and the peasant population falls well below Malthusian limits, this actually makes things worse in the short term, because the elites now have an even smaller economy to extract from.

Per Turchin, this process explains, in whole or in part, the late-republic Roman civil wars, the Crisis of the 3rd Century, the unrest all over Europe in the 14th century, the unrest all over Europe in the 17th century, and most other periods of intense internal instability in agrarian societies. The timeline of the rise-and-fall of Chinese dynasties follows the same frequency, and from what I know of it, the specifics seem to be in order as well, at least for the fall of the Western Han.

How does this translate into SotE? The following need to be modeled:

Rural carrying capacity/arable land still available for cultivation.

Economic inequality, either abstracted or represented as actual population wealth stats.

A natural process whereby poor pops become poorer and wealthy pops become wealthier when arable land runs out.

Pop fertility levels that match economic fortunes. This will naturally soft cap post-Malthusian commoner populations without reaching famine states, and it will naturally increase elite population even after the general population has stagnated.

Unrest as living standards decline. I'm sure you guys are planning on this already.

Conspicuous consumption (including art patronage!) among the elites as their wealth increases. Effectively, per-capita demand should grow with per-capita wealth, but NOT decrease with it under normal circumstances. As elite demand exceeds their ability to meet it, they should naturally shift into the same "angry and poor" state as the peasantry, despite not actually being poor. Then the fun starts!

When elites enter the "angry and poor" state, they begin petitioning the state for handouts and sinecures and other forms of support. They also divide into factions, centered around powerful men who promise gains to their dependents. Giving patronage to one faction causes the other faction to get angry, and undermines their support for the state.

When one faction's loyalty is low enough and an event sets off the powderkeg, they revolt. Ideally, the most powerful families within a state - as well as their local holdings/power centers - will be modeled, and the actual holdings/power centers of the rebel faction will provide the backbone of the revolt.

As the events of war cause general population decline, elite population should fall faster than the general population - they are more likely to be fighting and dying in battle, and are more likely to be targeted by rampaging armies, at least in per capita terms.

As war wages throughout a country, pops should grow war-weary, reducing unrest levels. The modifier should wear off completely within 20-30 years however, causing war to be rekindled until the underlying problem of elite unrest is taken care of.

At some point, elite demands need to come down back to normal levels. TBH, I don't know what specifically to peg this phenomenon to, but it needs to happen. Once elite pops are able to meet demands again, they should naturally leave the "angry and poor state", and all the plotting and feuding that undermined the stability of the state should vanish.

A side note: modeling elite unrest as a function of demand vs. ability to meet demand gives you lots of interesting ways to manage it, at least in the short term. A big war that brings in lots of booty might buy you the time you need to prepare for the shit really hitting the fan. If a disproportionate number of nobleman happen to die in that war, so much the better...

Now onto...

This is going to be a lot shorter, because I'm at 2000 words already and I'm kind of tired.

Asabiya is, in a sense, a measure of how well the members of a society are willing to put the good of society ahead of their own private interests, or that of their family and friends.

Asabiya grows in the face of a threat from an outside enemy - Turchin uses the term metaethnic threat. The greater the threat, and the more alien the enemy, the better. Alienness can mean several things. Are they barbarians, with a fundamentally different social structure (and economic base, if they come from the steppe)? Do they follow a foreign religion, perhaps one that you especially hate?Gaulish barbarians attacking Rome, Vikings attacking most of Northern Europe, Qiang and steppe peoples attacking various (North) Chinese states, Muslims attacking Christian kingdoms in Spain - all of these supercharged the asabiya of the societies on the defense.
But asabiya declines over time. It hits temporary low points during the periods of unrest in secular cycles, but it never fully recovers to its previous point without a recharge from foreign threat. By the year 400, Italy was an asabiya black hole, as it had been the center of the Roman Empire for nearly a millennium. Asabiya in the Balkans, however, was constantly being recharged by barbarian threats, which goes a long way to explaining why so many late Roman emperors came from that region. Being from a society with high asabiya has important effects on individuals, and it tends to produce strong leaders. I don't think it's a coincidence that the Barca family had their base in wild and unpacified Spain, rather than in the African heartland of Carthage.

How to model all this mechanically?

I don't think there's a way around an abstracted "asabiya" modifier attached to each pop. Gradual decline can be modelled by having the unrest period of each secular cycle chip away at the base asabiya value (in addition to much larger temporary reductions).
Base asabiya would only be recharged by metaethnic threats, which could operate through several mechanisms. Are you at war with an alien enemy? Asabiya ticks up slowly over time for all pops, faster if the enemy is more alien. Are alien armies rampaging throughout your territory? Pops gain an asabiya boost proportional to the alienness of the enemy, the damage inflicted, the size of the enemy army, whether or not they experienced the horrors firsthand, and if not, how close they were from the carnage. Are your pops in a border region next to a metaethnic Other that you are not on good terms with? They enjoy a passive asabiya boost.

Asabiya should grow and fall logarithmically, so that - for example - border regions don't always hit max asabiya, and so that you don't immediately lose all social cohesion in the imperial core just because you made it safe. Looking at the Roman example, it should take about 6-800 years to deplete your asabiya battery under ideal circumstances (the last sack of Rome before Alaric was in 390BC, and I believe no foreign army seriously threatened Italy after Hannibal withdrew in 203BC).

That's all for now... I'll leave this at 3100 words. Thanks for reading! If you want to read more, I highly recommend Peter Turchin's *War and Peace and War* or *Secular Cycles* (*Secular Cycles* is very math heavy, *War etc.* is a more readable summary that doesn't get into the mathematical models). Turchin's book *Ultrasociety* deals with the origins of large-scale societies among humans, which is also highly relevant to SotE's simulation, but beyond this post.

Very interesting concept ShinShimon, I hadn't really thought of imperial social cohesion from this cyclical kind of perspective.  What would be especially interesting is to observe this kind of "secular" social cycle when it intersects with resource depletion cycles, like soil depth and fertility.  The trick would be to represent the secular cycle you describe in the form of individual powerful families and their attitudes.  I'll have to read up more on Turchin, thanks for bringing this to our attention!

No problem! Turchin's work is really interesting - he's spearheading the effort to turn the study of human societies into a hard science. He comes to it from a background in population ecology, and actually has his PhD in zoology, so he thinks about human societies differently from the conventional historian. Of course, he also works very closely with conventional historians, especially in the construction of the Seshat historical databank (a huge and fascinating project, whose data is available publicly here

Wow, thanks for bringing this to my attention.  I know in the humanities there is often a lot of resentment at the notion that the study of human societies can be a hard science, but this is the approach I am interested in.  Thanks for the find.  Is there any other literature of Turchin's you can point me to?  Would be interested in reading up.  Also, are you a part of our other communities?  Would like you to stick around.  It's helpful to have people with your perspective.

He has a lot of published journal articles dealing with specific matters related to his broader research goals, and a few other books that I haven't read yet. Of these, Historical Dynamics seems to be the most relevant to SotE - it deals with a lot of questions related to large-scale human societies, with the same analytical and mathematical rigor as his other work. The bibliography on his Wikipedia page has a list of his most important articles, and you can read brief descriptions of all his published books on his website ( That website also includes a semi-regularly updated blog, which has a big backlog of interesting posts related to his work.

I've become a somewhat active user of the Discord, and I'm sure I'll stick around. I've been following SotE since you left M&T to start on it, but now that I can play around with the 2.0 world generator, it's grabbing my interest more and more.