/Man as a Rationalist Animal

Man as a Rationalist Animal

seeing like a state
On Seeing Like a State

Subconsciously or not, most of us presuppose malice behind failure. This goes doubly for historical failures, and quadruply for political failures. The daily form of this hisses about “corrupt politicians” (past and present), perhaps about “businessmen and special interests”. The more extreme forms fall into conspiracy theory. Often this is diagnosed as a form of pessimism, especially “pessimism about politics”. That’s wrong; it’s optimism.

The pessimistic view is this: “Everyone is just trying their best.” If the horrors of history are the result of ill will then we should take comfort. It may not always be possible to avoid evil dictators, but at least we know that human agency has some power. An evil person realizing their evil machinations implies that perhaps a good person can successfully realize a good plan. Stalin may have been mean and bad, but if we just get the right people in there (read: me), then surely The Good will result. But if everyone is just “trying their best” then none of this is assured. Indeed – something is so broken that our best intentions still produce misery. So… what happened?

Seeing like a State sets out to answer this question. Namely: why do we see large state schemes cause so much misery even when guided by good intentions and (seemingly) careful design? And that also explains its subtitle: How Certain Schemes to Improve the Human Condition Have Failed.

If I had to give a one-sentence explanation of the book, it would be: “The effects of technocracy on a polity are almost always negative. Of course, that argument is detailed across four-hundred pages, and nothing but the book itself can really capture that analysis – I’ll do  my best, but just read the book.

Its popularity means that many other bloggers have attempted more detailed analyses. These two are particularly good: a Ribbonfarm piece by Venkatesh Rao, and the more critical Slate Star Codex review by Scott Alexander. The SSC review, in particular, goes into a lot more detail than I will. On the economic side, J. Bradford Delong writes a some-what skewed (but good) analysis, which is here corrected by Crooked Timber. Finally, if you read anything, here’s James C. Scott’s own overview of Seeing Like a State.

The plethora of reviews also means a plethora of criticism. This is helpful: I don’t want to describe the book but explain its import, and contrasting analyses are better for that than a cursory retelling. But since explaining the book is going to take a lot of time, I’m going to have a whole other post replying to criticisms I’ve seen levied.


A state has a plan – could be despotic or desirable, doesn’t matter. In order to put that into action, the state needs more control over society than it previously had. This is obvious – you can’t institute a plan unless you control enough elements to get the plan to work. That means it needs to understand society, at least in the cursory “this many people live here”, “that region is mountainous” kind of way. But most states inherit societies that are messy, mostly because social relationships are messy and nature is also messy. To combat this, the state tends to create broad, synoptic views that only ever approximate the (actually) irreducibly complex realities.

Scott (writing his own overview):

Intervention in society (or nature) for whatever purpose (e.g. delivering welfare benefits to those with particular disability or keeping watch on political enemies) requires creating the mapping or optics necessary to legibility. In Seeing Like a State, and as a student of politics, I concentrate on state-making and government.

Missing here is the fact that those purposes tend to bleed into one another. Part of complexity is overlap, which means that it’s not really possible to isolate The Problem and surgically remove it even if you have a razor-sharp scalpel and a steady hand. Scott’s examples are Soviet collectivization, Tanzania, early forestry, Colonial regimes, the city of Brasilia, but each of those has specificities I’d prefer to avoid.

Here’s an example I made up:

The urban center of an old city is a hopelessly chaotic slum. The denizens can navigate it without trouble, but without proper street names (or even streets) it’s much harder for the state to get a proper reading. Now, the government was elected on the promise of better healthcare for all, and they’ve partially delivered. There are more hospitals, and they’ve increased the number of ambulances. But people in the densely packed center still aren’t getting proper aid. Ambulances keep getting lost, or they’re losing critical minutes because of the labyrinthine old streets, or unmarked buildings and houses make it impossibly difficult to determine who is actually in need and which apartment they’re in.

The citizens themselves don’t understand this – they grew up in there, so it’s obvious who’s where. They try and explain this to the city, but it’s based in highly specific information and historical details that don’t make sense to outsiders (go left at Greg’s place, right at the spot of the Best Marble Game of All Time, straight through to what-used-to-be-the-old-pool-hall-but-isn’t-now, etc.). But the ambulance dispatch and the state use a different model: they view things from the perspective of a map, which is only reasonable given that they have a much wider area to concern themselves with. They need some ordering mechanism that will mesh with those of other districts.

At first, the state just puts random names on the streets. This helps some, but the residents still colloquially go by the old terms they know, which causes problems for dispatch. Moreover, most of those alleys are still too narrow for ambulances to get through. The state decides on a more radical project: it’s going to plow through what it can and build new, ordered streets based on a grid. While they’re at it, they decide to make one commercial district and one residential district – it’s just a better system.

Now you have a pretty resentful populace. “But why?” The local economy has been disrupted, for one, even if no one else sees it. Turns out that a lot of the interweaving between commercial and residential was actually necessary for those local businesses – not to mention the young people who hung around getting paid to help move a thing here or there when the shipment arrives. But that’s a State’s reason. More pressing to the residents is the loss of their local knowledge, what we might call “folk monuments” that create senses of identity. Most pressing is their loss of power.

This last one is important in two ways. The first is both obvious and not. A large part of our powers lie in our familiarity with surroundings. When those have changed, we lose a lot of the familiarity that brings us that. Here’s an example: an old man has occupied an apartment for his whole life, and he’s maintained his self-sufficiency due to familiarity. He’s always been able to hobble to the store (right down the way) without aid, and when he does need it the same neighbors are always there. But the new plan places a commercial zone some ways away from the residential, and the new neighbors aren’t familiar to him. Moreover, the old store would always stock [thing] because it knew the old man and expected his business. But the new one has to deal with everyone, so it diversifies in a way that can’t suit his needs. The old man has suddenly become dependent on someone else (in this case, probably the state itself). Weird, small effects like that happen all over. “Small” is the key word, because they do look meaningless in the grand scale.

The second only reveals itself when things get really bad, so let’s work up to it.

To fund the ambulances you have to tax. That’s always been a problem in the slums (who lives where? how do you tax?) but suddenly you find that it’s much easier to get quantification. It’s a small amount, but people unused to paying taxes are suddenly hit with it. Not to mention, many of them are poorer in total because the local economy was interrupted. At the same time the district looks wealthier on average, because the few who successfully transitioned to the commercial zone now have a larger (alien) clientele. They employ less from the neighborhood, and the money is shared less often (the old man no longer has anyone to tip a dollar for helping up the stairs), but from the outside everything looks better. Either way, the citizens get angry, and they still aren’t seeing the benefits, and all of their local knowledge has been messed with, so they riot.

We all know the cliched MLK quote: “A riot is the language of the unheard.” I think this is true, but a better way to phrase it in Scott’s terms would be: “A riot is the language of the unintelligible.” The citizens may be rioting over taxes, but it’s not really that, and if they could even explain “why” they were rioting, it wouldn’t make sense to outsiders. “They tore down the old-pool-hall!” is actually a better summation of it than “taxes”, but that just seems… well, weird and petty. After all, to understand why that’s important, you need to have an intimate knowledge of the history and economy of the society that no one outside it has. Even that old man’s complaints look odd. He lost autonomy, but he’ll probably live longer because of the ambulances. A statistician might be able to look at all the confounders and draw out these explicit economic effects to determine what side they fall on, but even that wouldn’t really get to the heart of it. The heart of it is a lot more psychological, to use a word that mostly fails to capture it.

Either way: the slums riot. Suddenly you find that your ambulance routes are highly efficient for combating these riots. In the old slums, citizens had all the power. They knew where to run, and how to dodge out of the light, and who would hide them, etc. Now it’s in the hands of the police. And, of course, the police force costs money, but luckily the new names and addresses make it much easier to tax the citizenry, so taxes go up to combat the riots from the taxes going up. But to ensure that citizens pay those taxes…

One wanted to establish efficient healthcare, but to do that you had to create a state apparatus that was able to control people enough to develop it around them. It’s about ambulances, technically, but somehow ambulances and cops and increased taxes and housing development all blended into one process, and every single part is required to make the others work. If you stopped just one, the whole project is lost, and all you did was cause damage with no benefit. At a certain point, it starts looking weirdly humane to accelerate it, if not just to finish and get to the benefits. But, of course, accelerating it can mean exacerbating the bad effects. And it definitely means removing more power from the hands of the populace.

Now, I intentionally chose (what looks to me) to be a highly ambiguous example. You have a statistically wealthier and healthier populace, but it’s also somehow more immiserated. And a lot of that immiseration was do to these weird cadastral surveys that failed to account for how the location actually worked, and even more had to do with the way that implementing that synoptic knowledge deprives people of their own powers. But still… I can easily see arguments for both sides.

Scott is not writing anything as simple as “Big Government is always awful“, and he’s not a libertarian (well, not in the way we commonly use that term). But let’s imagine that we were debating this: very quickly it would turn into a numbers game (number of lives saved by ambulances vs. economic displacement effects vs. cost of development vs. tax burden vs. [etc.]). Make no mistake: that’s the language of the state. It might be right in this instance (depending on your politics, of course), but it’s not something that a rioter would understand. There’s a much more ambiguous power disparity that isn’t getting mentioned, because how would you put that in state language? Moreover – that doesn’t seem to be very relevant here.

This is all, in a nutshell, a process of legibility. That word is going to come up a lot – it’s Scott’s answer for why things always seem to go awry. Everything from naming the streets, to assigning numbers for taxes, to replanning a city to be intelligible to an outside eye (how the state sees, of course), is part of it.

Legibility is a way of viewing things, but it’s also a tool to change things. That tool is essential in state craft, for reasons that should be obvious, and its necessity is the reason it keeps happening.


A common response to power deprivation (well, depending on your politics) is moral: one just shouldn’t do that to people. But that won’t work for everyone. Someone with slightly less trust in the masses might easily say: “So what? They’re near-sighted, and the state should take away their control.” I think that’s the person who might actually get the most from the book, because it’s not just about “how” authoritarianism develops, it’s about how authoritarian rules are often much less efficient.

The book is about failed technocracy, but that’s based on another proposition: The people are not stupid. Collectively, they’re probably smarter than you, and that’s 100% certain when it comes to their daily existence.

Scott distinguishes between metis and epistemic knowledge. Epistemic knowledge is from the state, and it’s often called “rational”. It’s based in scientific or “scientificish” knowledge. It’s so general as to apply everywhere, which means it kind of winds up applying nowhere. Sometimes this looks like farming techniques that are great in theory, but in practice are incapable of adapting to any single locale. Other times it looks like geometric land distribution that ignores local conditions.

Metis is much more ambigious, a strange mix of hyper-empiricism and tradition and encoded ritual that has adapted to produce best results within a specific context. The origins of it would take a whole other book (presumably filled with Darwinian metaphors), but the point is that it works. It’s a reason that (using repeated examples of Scott’s) small village farms tend to vastly out-produce large epistemic plots; it’s also (less discussed) a reason that people tend to resent deracination even if it’s into a “nicer” context.

In a rare example of both social and agricultural metis vs. epistemic knowledge, Scott talks about a Russian village that had parceled out land based on growing conditions. All the villagers get a sliver of swamp for one crop, some forest for another, pasture for yet a third, and a few logs for some mushrooms – they get equality of outcome, because they all know about how much yield there is. But that looks messy and, well, unequal on a map. So the state just cuts the land into geometrically equal plots makes sure everyone sticks to theirs – as a result, no one gets what they need and nearly everyone gets less (except for a few who have a surplus of things they didn’t even want a surplus of) even if it looks more equal on a map. On top of that, everyone has to work harder because the labor is no longer equally distributed between tasks.

(That’s also a kind of whimsical example – the rule is generally famine and oppression.)

Metis is also social – that’s going to become more important for my view of the book than anything. Scott points out that seemingly every village in pre-modern France had its own law or custom, which was often not written but was understood. At other times, it might look like a religious ritual: one plants crops at this specific time for a deity. That might be effective (and have developed to be at that exact time) because it’s the right time to plant crops, but you’d never understand that if you asked a villager. They’d simply say that it was for the goddess. Planting at a specific time for a deity makes sense for the village, and preserves its crops. But for the state, it looks totally backwards. Hence we get legibility, as explained above.

In general: people have a reason they do things the way they do. The issues that arise aren’t just “humanitarian” in the sense of “people don’t like it. They’re actually pretty pragmatic, and governments that fail to recognize this (or ignore concerns as merely “petty humanitarianism”) tend towards economic and agricultural chaos. Disrupting this from the outside normally means that you don’t actually understand them. Almost always, that hubris will make life less productive and less enjoyable.


It gets worse.

Here’s an easy way to make the earlier example about ambulances unambiguous: the next year, a fascist party wins elections and begins imposing absolutely terrifying laws (pick your least favorite and assume that’s what they’re implementing). All of a sudden, you really wish the chaotic old city was back. It would give resistance a fighting chance. As it is, the people are cooked.

That form is what Scott spends most of the book on. It doesn’t have to be fascist, but fascism is an example of it. It’s when a state decides to completely remodel itself based on abstract plans. This he calls “Authoritarian High Modernism”.

High Modernism is something like scientific state planning on steroids. It’s equally obsessed with “science” and a certain “sciencey aesthetic” (grids, strict geometric shapes, etc.), that took hold in the first half of the 20th century. It only sees through epistemic knowledge (or “rationalism”). In theory, it wanted to harness the powers of rational calculation to make human life better and more efficient. In practice, “good” and “efficient” tended to collapse into one another, and the state set about creating the most utilitarian system possible, where utilitarian really means utility rather than “greatest good for all”. If I had a word for this, it would be “ergonomomania”, but that word looks dumb so let’s never discuss it again. Also in practice: changes this radical require near complete control of the society.

High Modernism fails, and it fails catastrophically. Inherent to High Modernism is a desire to “better” people and states along rational lines that don’t take into account: a) local variation, b) human preferences and behavior, and c) its own ignorance. (c) probably includes both (a) and (b), but its meta form is something like “ignorance of the possibility of its own ignorance.”

I talked about (a) above. It’s metis. The complexity of nature makes it very resilient, and all parts of it tend to have a role. Everyone knows this. But this means that changing any of it changes the rest, and standardization practices have very far reaching consequences. Sometimes that just means everything stops growing after a point, because you failed to plant “useless” flora that inject nitrogen into the soil. Sometimes it means the Dust Bowl.

(b) is metis when it meets the authoritarian nature of High Modernism. High Modernism tends to view people as something like a summation of data. Each datum is quantitative component (needs this much food; require this much socializing), and things that can’t be quantified (either due to their qualitative nature or to the limits of the science used) are considered “irrational” and ignored. You see where this is going.

Whether those “qualitative” things are “irrational” or not is immaterial (also probably based in your personal philosophy). What’s really important is that any complaint will be in qualitative (or irrational) language, and hence will be ignored and/or suppressed. Even when, say, “social needs” are taken into account, they’re quantified in ways that the state can understand and use. The complexity of human networks is reduced and distilled into hopelessly useless forms. Again, this this is theoretically for the good of the citizens (“We’re freeing them from the oppression of false idols!”) but it has long-lasting psychological effects. That is, it would if people lasted, because many people prefer to resist rather than lose their “irrational” practices, and then [the history of 20th century utopianism] happens.

The resistance, it should go without saying, looks highly irrational.

(c): [everything below]


It’s natural to focus on High Modernism’s failures. They’re unambiguous examples of the state’s mentality, and they really draw out its shortcomings. But they’re historical, which can leave one wondering about how the book relates to our time. The examples Scott gives are things we do know scientifically now. We’re much better at understanding natural complexity and adapting to it rather than trying to make it adapt to us. This gives the (in my view) false impression that the “rational science” we have now is actually up to the task. coming clean: I worry a lot about our confidence in this regard.

At the same time, my concerns are not merely with the science we have (especially psychology, but w/e). I think a lot of the problem  comes from “legibility” itself. Namely – legibility is a kind of language, and so any resistance that isn’t in it will look crazy. At best, that means we ignore it (now). But at worst, and especially with the assumption that we “know everything now” (implicitly or explicitly), that enables us to just do the exact same thing again. 

Let’s consider High Modernism. Certain adherents (for instance, Le Corbusier) were obsessed with the aesthetic element, but others were more empirical. I think this is more of a spectrum, with High Modernism on one end and… I dunno, maybe Paul Feyerabend on the other? Either way, almost none of us are perfectly one way or another.

I think this is helpful to understand why certain aesthetic ideals emerged. Many people maybe started on the more-empirical side, but then noticed that all of the research started looking the same. I’ve called this “quantification”. It probably looked geometric, “simple” (think Occam’s razor), etc. Much like you’d imagine scientific papers to look today. When confronted with a situation where they didn’t have data, but still knowing that in the past all the good data looked like “quantification”, there’s a pretty natural instinct to assume it should have a certain look. This is, of course, separate from the obsession with grids – what I’m really interested in here is why empirical data overlays metis in activities, not the preference for what a “clean city” would look like.

Where does this get us? From the perspective of early 20th century rationalists: global commerce and dazzling new technologies vs. peasants who insisted on setting their own crops on fire for a bizarre religious ritual. That’s not hyperbole, by the way, that’s an example from the book. The rationalists were maybe aware that they had spotty or incomplete data on growing practices, but the data they did have was quantified and supported by labs, whereas the villagers couldn’t explain their practice using any kind of data. The peasants, it turns out, were right, but would you have guessed that? More importantly: would you guess that now?

The only way that I can see that one would is if there’s a genuine admission of ignorance on the part of the “scientists”. Even reconsidering one’s position requires some kind of counter-argument, and the problem here is that the counter-argument doesn’t make sense in the language of the scientists. Indeed – a solid counterargument requires some kind of power (you’re not going to listen to some random guy on the street  screaming about [whatever]), but legibility inherently reduces that power.

A common takeaway from Seeing Like a State is “reality is a lot more complex than the state simplifications”. And that is extremely true. But I think the far more interesting, under-discussed, and disturbing takeaway is about mutual incomprehensibility. Note that this part is still quite relevant. Even in a democracy where the citizens have a voice – and let’s say that politicians actually listen – the way they explain things will not make sense to outsiders. This limits their ability to counteract a given proposition.

We’re lucky enough to have a (relatively) robust system of rights at the moment. But part of legibility is that it actually results in at least the partial prostration of civil society itself. Namely, it highly privileges one part of the exchange (the “rational” one, however we’ve decided that), and hence disadvantages whoever is on the outs with the back-and-forth.

This isn’t not Scott’s argument – it’s my own – but I think it’s heavily implied in the work. Not to get ahead of myself, but there’s a reason that he isn’t a (right-)libertarian: much of what he thinks we’re doing at the moment is economic simplification. But that also requires a kind of psychological simplification.

Either way – the first task is then to show hidden “rationality” in things that we consider highly irrational. If those exist, then I think there’s a good chance that there are other understudied things going on with other [outgroups; who am I kidding].

Accordingly, much of the rest of this review will be about that.


In the [cool kids scene] of the 2000’s, opposition to Christianity was a given. No subject was more likely to unite than a good old bashing of fundamentalists. Marxists, Anarchists, Libertarians, Liberals, Randians – it crossed political borders. (I’m from California, if that helps.)

It was hard not to – it’s not like the fundies had any coherent reasons they could articulate. Every other day some [class indicator] pastor would announce opposition to “the gay agenda” or seek to return to “a Christian nation” as though that were a desirable thing that had ever existed (like, deism duh, etc.). If that wasn’t bad enough, when pressed for an explanation, they’d just read some passage from John or Corinthians. “And? Was that all?” Readers take note: if someone is busy mocking you for your holy book, justifying your actions based on your holy book is a terrible rhetorical strategy.

Standard interpretation: “The olds are lamenting the loss of an oppressive institution that has no objective value, right?” Right?

So this is what churches do in our language: they’re probably the single most important economic institution in rural America. Period.

Here are some obvious economic effects: Nearly every church functions as a community safety net, where tithes collected are distributed to poor members or members experiencing sudden economic shock (disemployment, medical issues, etc.). Depending on the church, this is actually a lot more immediate and a lot larger than government distributions that approximate the same thing. They also function as labor banks, wherein members help one another with projects that they could otherwise not afford (think of home improvement projects coordinated through the church, wherein one can afford to repaint their house or call on the labor expertise of a fellow congregation members [say a plumber] to perform a simple  but otherwise costly repair). Hell, one of the biggest things they do is something almost no one seems to think about: most churches provide free after school programs for poor congregation members, which is a humongous cost for parents. “Big deal.” Yeah, but the cost of childcare is actually fucking enormous.

Note that most of those are vastly more important for the old and the retired (“it’s just the olds complaining!”), both in terms of cost (local members helping for free) and autonomy (one is less likely to have to enter the anonymity of a retirement home, etc. if community members are there to help).

Churches have many more nebulous effects that are even larger: they improve social trust, which has a stupidly powerful economic effect. They provide local networking effects, allowing members to find new jobs and move up using church connections. Many studies relate churches to decreased violence and drug problem in communities (although I suspect this is confounded by social trust and the kinds of people who tend toward religiosity). A lot of these aren’t going to have great studies attached, because they’re under the radar and understudied. But want to see something that will knock your socks off?

Gruber’s results suggest a “very strong positive correlation” between religious market density, religious participation, and positive economic outcomes.” People living in an area with a higher density of co-religionists have higher incomes, they are less likely to be high school dropouts, and more likely to have a college degree.” Living in such an area also reduces the odds of receiving welfare, decreases the odds of being divorced, and increases the odds of being married. The effects can be substantial. Doubling the rate of religious attendance raises household income by 9.1 percent, decreases welfare participation by 16 percent from baseline rates, decreases the odds of being divorced by 4 percent, and increases the odds of being married by 4.4 percent.

Source, emphasis mine. Note the most important part of that: one has to live nearby other co-religionists for these effects. “The kids are turning to atheism/the Devil” is a sign that all the coreligionists are going away.

In other words, churches are a cornerstone of local economic activity but they have to be churches. You can replace a Widget Factory with a Zigdet Factory and it’ll be the same, but we have no idea how to replicate church attendance.

This has another effect: any attempt to “stop” these negative effects won’t work. They only work based on the logic of the community. Because metis is both the worldview and the action, undoing one messes with the other. You can literally see this happening right this very moment in the midwest: church attendance collapsing is deleting a lot of those benefits even while the government tries to fill the void. It doesn’t “work” in the same way, because a lot of those effects are based on shared culture and trust that a government agency just can’t replicate. They only come with a shared worldview.

In a weird way, maybe that welfare participation is the (attempted) replacement, but then it’s hard to square the simple economic replacement with this:

And, attending religious services weekly, rather than not at all, has the same effect on individuals’ reported happiness as moving from the bottom to the top quartile of the income distribution.

“Man, this sounds like something everyone should know!” I agree, but also HA! That will never happen. know how to argue for cultural conservatives to my left-wing, coastal audience. But how do you think the average actual conservative argues for that? “Faith”, “family values”, “God”, i.e., irrationally.

I know this is hard, but imagine actually being a conservative Christian in a dying town. Everything I just described is going away, nothing seems able to replace it, and things are just getting worse. The most noticeable difference by far is going to be “cultural” – what language would you use? “Loss of faith and family” is actually pretty apt. Let’s say that their arguments are identical to mine, just shrouded in local language. Fine – all that means is that In the final analysis, the conservative christian recognizes that they’re being deprived even of the power to complain, which is to say, even of the power to explain their powerlessness.


“Ew. Are you saying-”

No, I’m not saying that the government is oppressing American Christians. That’s stupid. They lost the culture war, but no one is tearing up their actual communities. There’s a broader social pressure now that has, well, social effects. I do think that two hundred years from now when we have a better handle on psychology and economics everyone is going to look back at this time with total confusion. Like – how did no one notice? Didn’t you see this economic and social collapse? They were even yelling at you about it! We will confidently aver: “Yes, but when they were yelling they had the impertinence to quote the Bible, and so we knew that they were wrong.” And the person from the future will, quite reasonably, call us complete fucking twats.

But that is still not oppression. Most churches are dying from… well, the opposite of oppression: the free choice of individuals to leave them (also economic collapse, but whatever).

You can sub in nearly any dissatisfied group for my church example, but I selected it for specific reasons: 1) It’s language that is probably considered the most irrational to anyone outside the group; 2) It’s still a powerful enough faction that someone has bothered to research it and give me (vague) data; 3) the jargon used is almost entirely that of a distinct social dynamic, rather than something related to, say, farming.

What I am implying is that whatever the modern form of legibility is, religious institutions are far outside of it. They may not be the only ones, but they’re definitely there. I can already feel people gearing up to scream at me for [something], so let me say something: I am not a fundamentalist Christian. I am also not particularly inclined to suddenly adopt their social dynamic (which, anyway, is impossible). I also do not think that they’re oppressed. But they certainly are on the outs culturally, inasmuch as it takes no special bravery to be an atheist these days and the idea of a Hollywood Film coming out with strong Christian values is basically unimaginable. That gives us a privileged view at what happens when a culture – in modernity, like, right this moment – starts to get replaced without the replacement understanding it. Turns out that doesn’t look great.

But it’s most helpful in showing what it looks like (vividly) when groups cannot communicate.


Of course, if the government did suddenly become a whole lot more authoritarian, and for some reason decided to persecute Christians, it would also be pretty clear that no one would understand that. By which I mean: the kinds of things that a group will complain about won’t make sense to outsiders, and if they do, it’s probably not the real reason.

Legibility is a process of power, but it’s also a channel for it. People with epistemic knowledge can argue within civil society using it, but those outside cannot. The kinds of argument they make might be translatable, but they won’t know how to translate them (if we even have the rational knowledge to do so). Given that when power crushes someone, it tends to crush the powerless, we have a pretty serious problem: no one with power will even understand the powerless when they are being crushed.

Ignore the example of Christianity now: that was only to show that problems of legibility still affect us. The point is that one might not even realize that they’re acting in an authoritarian manner, because from your view the complaints arising aren’t real complaints (if they even look related, i.e. church to local economy may not be obvious).

This has a flip-side, one which my readers really might not like. Namely: be suspicious about what problems angry groups focus on. Christian groups genuinely are angry about the loss of churches as regular institutions but, as we saw, the building “church” isn’t a large enough structure to contain everything meant by that. The complaint is economic and psychological and social and religious and [many other here]. This doesn’t mean that you, personally, understand them any better, but it does mean that we rarely understand ourselves. Most of these  are like… x-order effects, and even people trained to look for those generally have an extraordinarily hard time sorting out what it affecting what. Metis is holistic, which makes it very hard to pinpoint specific, translatable items. “Why do you think I won’t like that?” Think about it some more and get back to me.

That flip-side also has an important distinction. I’ll mention it here, but it’ll only be clearly developed in later posts. Namely: there’s a difference between asking a group why something works and asking them what to do. If you asked a villager “why” their crop-system works, the answer they give you will almost certainly be wrong (objectively, scientifically, whatever here). And yet if you ask them what to do, i.e. the government crops are failing, how do we make this work? they’d (in given examples) be correct. This is highly speculative, but I suspect that a lot of the problem with “rationalism” boils down to this.


Let’s recap.

  1. States use legibility to understand and control a society, for the purpose of [anything].
  2. That process inherently disadvantages groups who use local knowledge (metis) and removes some of their power. When done by authoritarian societies, this is one of the explicit aims of legibility, but it is not done solely by authoritarian societies. When done by [democratic?] societies, it looks more like “what is compelling evidence in a political conversation”.
  3. Power is important for two major reasons: a) groups are more likely to rebel when they lack it (resentment) and, b) metis is actually pretty efficient, and you lose a lot of productivity/have to waste your own resources when you disempower people (because they cannot use metis, duh).  I’ve only really discussed (b).
  4. Legibility problems persist in modernity, and they look a lot like social dynamics crashing into x.
  5. When translating those social dynamics into my own language (legibility), I’m using economic and “scientific” psychological language.

This takes us to a final question: that of incentives.

The High Modernist states (largely) wanted complete social renovation. But that isn’t the only thing that brings up issues of legibility. Where the state wants something done (say, increased yield from the land; ambulances), there’s a specific incentive to work with “the best knowledge available”. Moreover, it’s much less likely that the state’s behavior will be “conservative” in the sense of “moving slowly and really observing how these things work.” Where something must be done, and you only have two options: data vs. fire-rituals, data is going to win out. To put it plainly: legibility is going to be a problem in the regions where the state is incentivized to change something. Now that I wrote that, it seems pretty painfully obvious.

A good indicator of incentives is in what kind of arguments are considered useful, or legitimate (this probably aligns with the “why”/”how to do this” distinction I made earlier). The way I had to translate the Christianity example should give you an indication of own epistemic knowledge. Namely: when we make political arguments, nearly always the “best” looking form of it is going to be in economic terms.


So what is x?

There are a few answers that should be obvious. But given that this review is (theoretically) part of a longer series, we’ll get to that.

I’m not a relativist, and definitely not a cultural relativist. I don’t accept that Christian Fundamentalists have an equally valid interpretation of the world, (that would be the worst way to take my argument). I’m proposing that the epiphenomena that we like (economic prosperity, psychological wellbeing, etc.) are not so easily separable from that culture. Moreover, we may not even have words for them because they’ve been dismissed as “irrational”. And they definitely aren’t able to be replicated or explained when we don’t take the time to really reexamine what that culture is doing for its people.

Because I’m not a relativist, I think those “epiphenomena” are actually mostly what people miss when the culture disappears. I don’t think people are happier in churches because Evangelic Protestantism is correct – I think it’s a more complex interaction of community and coordination and [weasel words here] that is interpreted as “Evangelic Christianity” itself. But I do think that telling people that it’s entirely their own culture is a pretty terrible idea. This should be obvious but isn’t: if people don’t think it’s these broader epiphenomena (even if we can’t replicate them yet), and thus think it really is something about that specific culture, then politics will get weird.

The next book I’m going to write on is The Great Transformation, by Karl Polanyi. Quite a lot of that book deals with what he calls the “economic bias”, where Industrial Revolution states (and their later scholars) tended to neglect massive social upheaval in favor of desultory charts that showed “growth”. But even though the book is historical (as is Seeing Like a State), it’s not simply a work of economic history. What concerns Polanyi is not something that Scott really addresses: what is the powerless’s reaction to loss of power, especially when that is economic?

Let’s say that in a given state, there are two cultures: Culture A and Culture B. They’re quite different, but both have adapted to their circumstances. They don’t necessarily understand one another, and they conflict on a lot of values. Let’s say that Culture A has a society based on metis, whereas Culture B is more willing to adopt “scientific” methods.

Culture B somehow gains power over the state, and chooses to “enlighten” Culture A. This might be for good or bad reasons – it doesn’t matter. Either way, the oversimplification results in a lot of negative-but-not-quite-measurable effects, or at least not using the tools of Culture B. Hence, according to all the demographic rundowns that State B can provide, Culture A is doing fantastically. They’ve been bettered! Culture A is clamoring for more autonomy, but that looks like madness to Culture B. There’s no explanation for their resentment – they have ambulances and they’ll live longer now!

How do you think the State (Culture B) is going to explain it? I don’t just mean in propaganda – I mean newspapers, because most journalists are trained in universities and tend to think in the state’s logic. “It’s because these [reactionaries] are obsessed with their [reactionary] culture.”

Ok, now let’s look at it from Culture A’s side. From their perspective, all of that culture disappeared and that’s the most obvious difference. The epiphenomena that they’re really missing are not only unstudied – they’re explicitly denied by the entire intellectual apparatus. All of us a sudden, it starts to look a lot like the culture itself – and preferably the most chauvanistic type available – is the exact thing that will solve all their problems. Note that in Scott’s analysis (and my own), this should have been avoidable. And yet…

How is the interplay going look between Culture A’s complaints and the mockery of State B’s newspapers? The most obvious answer is the right one: it will strengthen Culture A’s belief that their current, fallen state can only be reinvigorated with… Well, what kind of politics develops out of that?


I heard a sociologist joke once about how “culture” was sort of the social science’s god-of-the-gaps argument. When we didn’t really understand why something was happening, we’d just assign it to culture and move on. That joke takes on a very different light in view of Scott’s work.

As I’m writing this, an article came out in the Atlantic. It’s the standard “This is why Trump voters were Trump voters” pablum, and was shared on my facebook by everyone I expected to share such a thing. It proposes “cultural anxiety” as the explanation.

Note that using Scott (and, well, really any economist), that’s not easily separable from “the economy”. But ignore that. Note also that there are a few pretty glaring problems with how they measure “cultural anxiety” (for instance: not buying into colleges is a sign, but that implies that “going to college” is a necessary part of American culture, which is a pretty strong blue-tribe bias). Ignore those, too.

This has another effect: “culture” is eminently useless, mockable, etc. in the state’s language. It’s a lacuna. “Something that should be overruled by sound, rational reasons.” I’m pretty sure I didn’t see a single person share the article out of “compassion”, whatever that is. It was entirely to point out that “those fucking [whatevers]” don’t have any good reasons.

Trump is a buffoon, etc. But note something interesting here: what is considered a good reason? What’s the “logic” behind a vote? And, far more disturbing, what is the predictable response from someone whose (frankly, legitimate and economic) anxieties are all mocked as “cultural”? Will it be a return to the worst parts of that culture, or will they “come to reason”?

top image from Akira Kurosawa’s Dreams

Part of the Uruk Series