Fireside Friday: October 15, 2021

Fireside this week! We’re in that mid-semester crunch time with students turning in papers and exams which need grading, but fortunately Ollie is getting into the fall season:

The pumpkin spice cat-toy was a gift from a friend.

For this week’s musing, I want to discuss in a fairly brief way, my views of ‘megahistory’ or ‘cliodynamics’ – questions about which tend to come up a fair bit in the comments – and also Isaac Asimov, after a fashion. Fundamentally, the promise of these sorts of approaches is to apply the same kind of mathematical modeling in use in many of the STEM fields to history with the promise of uncovering clear rules or ‘laws’ in the noise of history. It is no accident that the fellow who coined the term ‘cliodynamics,’ Peter Turchin, has his training not in history or political science but in zoology; he is trying to apply the sort of population modeling methods he pioneered on Mexican Bean Beetles to human populations. One could also put Steven Pinker, trained as a psychologist, and his Better Angels in this category as well and long time readers will know how frequently I recommend that folks read Azar Gat, War in Human Civilization instead of The Better Angels of our Nature.1

Attentive readers will have already sensed that I have issues with these kinds of arguments; indeed, for all of my occasional frustrations with political science literature (much of which is perfectly fine, but it seems a frequent and honestly overall positive dynamic that historians tend to be highly critical of political scientists) I consider ‘cliodynamics’ to generally take the worst parts of data-driven political science methodologies to apotheosis while discarding most of the virtues of data-driven poli-sci work.

As would any good historian, I have a host of nitpicks, but my objection to the idea of ‘cliodynamics’ has to do with the way that it proposes to tear away the context of the historical data. I think it is worth noting at the outset the claim actually being made here because there is often a bit of motte-and-bailey that goes on, where these sorts of megahistories make extremely confident and very expansive claims and then when challenged is to retreat back to much more restricted claims but Turchin in particular is explicit in Secular Cycles (2009) that “a basic premise of our study is that historical societies can be studied with the same methods physicists and biologists used to study natural systems” in the pursuit of discovering “general laws” of history.

Fundamentally, the approach is set on the premise that the solution to the fact that the details of society are both so complex (imagine charting out the daily schedules of every person one earth for even a single day) and typically so poorly attested is to aggregate all of that data to generate general rules which could cover any population over a long enough period. To my mind, there are two major problems here: predictability and evidence. Let’s start with predictability.

And that’s where we get to Isaac Asimov, because this is essentially also how the ‘psychohistory’ of the Foundation series functions (or, for the Star Trek fans, how the predictions in the DS9 episode “Statistical Probabilities,” itself an homage to the Foundation series, function). The explicit analogy offered is that of the laws that govern gasses: while no particular molecule of a gas can modeled with precision, the entire body of gas can be modeled accurately. Statistical probability over a sufficiently large sample means that the individual behaviors of the individual gas molecules combine in the aggregate to forma predictable whole; the randomness of each molecule ‘comes out in the wash’ when combined with the randomness of the rest.2

I should note that Turchin rejects comparisons to Asimov’s psychohistory (but also embraced the comparison back in 2013), but they are broadly embraced by his boosters. Moreover, Turchin’s claim at the end of that blog post that “prediction is overrated” is honestly a bit bizarre given how quick he is when talking with journalists to use his models to make predictions; Turchin has expressed some frustration with the tone of Graeme Wood’s piece on him, but “We are almost guaranteed” is a direct quote that hasn’t yet been removed and I can speak from experience: The Atlantic‘s fact-checking on such things is very vigorous. So I am going to assume those words escaped the barrier of his teeth and also I am going to suggest here that “We are almost guaranteed” is, in fact, a prediction and a fairly confident one at that.

The problem with applying something like the ideal gas law – or something like the population dynamics of beetles – to human societies is fundamentally interchangeability. Statistical models like these have to treat individual components (beetles, molecules) the way economists treat commodities: part of a larger group where the group has qualities, but the individuals merely function to increase the group size by 1. Raw metals are a classic example of a commodity used this way: add one ton of copper to five hundred tons of copper and you have 501 tons of copper; all of the copper is functionally interchangeable. But of course any economist worth their pencil-lead will be quick to remind you that not all goods are commodities. One unit of ‘car’ is not the same as the next. We can go further, one unit of ‘Honda Civic’ is not the same as the next. Heck, one unit of 2012 Silver Honda Civic LX with 83,513 miles driven on it is not the same as the next even if they are located in the same town and owned by the same person; they may well have wildly different maintenance and accident histories, for instance, which will impact performance and reliability.

Humans have this Honda Civic problem (that is, they are not commodities) but massively more so. Now of course these theories do not formally posit that all, say, human elites are the same, merely that the differences between humans of a given grouping (social status, ethnic group, what have you) ‘come out in the wash’ at large scales with long time horizons. Except of course they don’t and it isn’t even terribly hard to think of good examples.

Take, for instance, Chinggis Khan (born Temujin; I am going to use Temujin here to mean the man himself and Chinggis Khan to mean his impact as a ruler once the Mongols were fully united). The conditions for Chinggis Khan were not new in 1158; the basic technological factors with made the Steppe way of war possible had existed in the Eurasian Steppe for at least two thousand years by the time Temujin was born. Political fragmentation was also an important factor, but this was hardly the first time that nearby China had been politically fragmented (at the very least the periods 771-221BC, 220-280AD, 304-589 and 907 through to Temujin’s birth in 1158 all qualify) and the steppe had effectively always been politically fragmented. Our evidence for life on the steppe is limited (we’ll come back to this in a second) but by all appearances the key social institutions Temujin either relied on or dismantled were all centuries old at least at his birth.

What had been missing for all that time was Temujin. To buy into the strongest form of ‘cliodynamics’ is to assume that the Steppe always would have produced a Temujin (in part because his impact is so massive that a ‘general law’ of history which cannot predict an event of such titanic import is not actually a functional ‘general law’). And to be fair, it had produced nearly Temujins before: Attila, Seljuk, etc. But ‘nearly’ here isn’t good enough because so many of the impacts of Chinggis Khan depend on the completeness of his conquests, on a single state interested in trade controlling the entire Euasian Steppe without meaningful exception. The difference between Temujin and almost-Temujin (which is just basically ‘Jamukha’) is history-shatteringly tremendous, given that both gunpowder and the black death seem to have moved west on the roads that Chinggis opened and the subsequent closure of those routes after his empire fragmented seem to have been a major impetus towards European seaborne expansion.

Moreover, it is not at all clear that, absent Temujin in that particular moment – keeping in mind that Temujin hadn’t appeared in any other moment – that there would have inevitably risen a different Temujin sometime later. After all, for two millennia the steppe had not produced a Temujin and by 1158, the technological window for it to do so was already beginning to close as humans in the agrarian parts of the world (read: China) had already begun harnessing chemical energy in ways that would eventually come to rob the nomad of much of his strength. If Temujin dies as a boy – as he very well might have! – it is not at all clear he’d be replaced before that window closed; his most obvious near peer was Jamukha, but here personalities matter: Jamukha was committed to the old Mongol social hierarchy (this was part of why he and Temujin fell out) and was so unwilling to do the very things that made Chinggis Khan’s great success possible (obliterating clan distinctions and promoting based on merit rather than family pedigree). Jamukha could have been another Seljuk, but he could not have been another Chinggis Khan and in this case that would make all of the difference.

To get briefly into a bit of historical theory, Chinggis is an individual whose actions in life fundamentally altered many of what the ‘Annales School’ of history would call the structures and mentalités of his (and subsequent) times. The Annales school likes to view history through a long duration lens (longue durée) and focus on big shaping structures like climate, geography, culture and so on. The difference between this and cliodynamics is that Annales thinkers propose to describe rather than predict, so it is not fatal to their method if there are occasional, sudden, unpredictable alterations to those underlying structures – indeed those are the moments which are most interesting. But it is fatal to a cliodynamics perspective, which does aim for prediction since “our prediction is absolutely right unless it is completely wrong” hardly inspires confidence and a “general law” of anything is only a “general law” in that it is generally applicable not merely to the past but also to the future.

In short, Chinggis Khan wasn’t a commodity; he couldn’t be replaced by any other Mongol warrior. And figures like that abound through history (for Roman history, it matters greatly for instance that Marius, Sulla, Pompey, Caesar and Octavian had very different personalities when they found themselves in a position to dominate the Republic with military force). Moreover, the figures like that who we think of, generally capital-g ‘Great Men,’ are hardly the only such individuals like that. They’re the only ones we can see. What of, for instance, the old Argive mother – her name lost to history – who killed Pyrrhus of Epirus, considered the greatest general of his generation, with a lucky throw of a roofing tile, both ending his career but also setting in motion a chain of events where the power vacuum left by Epirus would be filled by Carthage and Rome in a way that would bring those former allies (allied against Pyrrhus, in fact) into a shattering conflict which would then pave the way for Roman dominance in the Mediterranean? History must be full of innumerable such figures whose actions created and closed off courses of events in ways we can never know; how do we know that there wasn’t some would-have-been Temujin on the steppe in 100AD but who was killed in some minor dispute so very minor it leaves literally no evidence behind?

(The fancy way of putting the influence of all of those factors, both the big structural ones and the little, subject-to-chance ones, is to say ‘history is contingent’ – that is, the outcomes are not inevitable but are subject to many forces large and small, many of which the lack of evidence render historically invisible.)

Both cliodynamics and psychohistory assume these differences and problems ‘come out in the wash’ over a long enough period and a big enough sample. It doesn’t take much of a counterfactual thought experiment about how small changes by individuals could lead to enormous historical differences to see that they don’t. The defense that cliodynamics only deals in probabilities is little comfort here: in fact the apparent randomness (which one may argue is merely complexity on a scale that is beyond simulation) swallows the patterns. One could easily argue, for instance, that the extremely unlikely career of our fellow Timujin is a necessary cause (albeit merely one of many, several of which might be considered deeply improbable) for the fact that all commercial pilots and air traffic controllers worldwide have to learn a form of English (which one may well assume has its own structural knock-on effects in terms of the language used for business and from there the outsized cultural impact of English-speaking countries).3 No one in 1158 was likely to have supposed that English – a language at that time not even spoken by the English nobility (they spoke French)! – would become the first truly global lingua franca (and arguably the only one, though here caveats may overwhelm the claim) and thus the language of aviation. But that is precisely the kind of big structural change that is going to be really impactful on all sorts of other questions, like patterns of commerce, wealth, culture and influence.

Such complex causation defies general laws (even before we get into the fact that humans also observe history, which creates even more complexity) with such tremendous import from such unlikely events in an experiment which can only be run once without a control.

The other problem is evidence. Attempting to actually diagnose and model societies like this demands a lot of the data and not merely that you need a lot of it. You need consistent data which projects very far back in history which is accurate and fairly complete, so that it can be effectively compared. Trying, for instance, to compare ancient population estimates, which often have error bars of 100% or more, with modern, far more precise population estimates is bound to cause some real problems in teasing out clear correlations in data. The assumptions you make in tuning those ancient population figures can and will swallow any conclusions you might draw from the comparison with modern figures beyond the fairly obvious (there are more people now). But even the strongest administrative states now have tremendous difficulty getting good data on their own lower classes! Much of the ‘data’ we think we have are themselves statistical estimates. The situation even in the very recent past is much worse and only degrades from there as one goes further back!

By way of example, I was stunned that Turchin figures he can identify ‘elite overproduction’ and quantify wealth concentration into the deep past, including into the ancient world (Romans, late Bronze Age, etc). I study the Romans; their empire is only 2,000 years ago and moreover probably the single best-attested ancient society apart from perhaps Egypt or China (and even then I think Rome comes out quite solidly ahead). And even in that context, our estimates for the population of Roman Italy range from c. 5m to three to four times that much. Estimates for the size of the Roman budget under Augustus or Tiberius (again, by far the best attested period we have) range wildly (though within an order of magnitude, perhaps around 800 million sestertii). Even establishing a baseline for this society with the kind of precision that might let you measure important but modest increases in the size of the elite is functionally impossible with such limited data.

When it comes to elites, we have at best only one historical datapoint for the size of the top Roman census class (the ordo equester) and it’s in 225 BC, but as reported by Polybius in the 140s and also he may have done the math wrong and it also isn’t clear if he’s actually captured the size of the census class! We know in the imperial period what the minimum wealth requirement to be in the Senate was, but we don’t know what the average wealth of a senator was (we tend to hopefully assume that Pliny the Younger is broadly typically, but he might not have been!), nor do we know the size of the senatorial class itself (formed as a distinct class only in the empire), nor do we know how many households there were of senatorial wealth but which didn’t serve in the Senate because their members opted not to run for public office. One can, of course, make educated guesses for these things (it is often useful and important to do so), but they are estimates founded on guesses supported by suppositions; a castle of sand balanced atop other castles of sand. We can say with some confidence that the Late Roman Republic and the Early Roman Empire saw tremendous concentrations of wealth; can we quantity that with much accuracy? No, not really; we can make very rough estimates, but these are at the mercy of their simplifying assumptions (we’ve actually discussed this problem before).

And this is, to be clear, the very best attested ancient society and only about 2,000 years old at that. The data situation for other ancient societies can only be worse – unless, unless one begins by assuming elite overproduction is a general feature of complex, wealthy societies and then reads that conclusion backwards into what little data exists. But that isn’t historical research; it is merely elevating confirmation bias to a research methodology.

As noted, I have other nitpicks – particularly the tendency to present very old ideas as new discoveries, like secular cycles (Polybius, late 2nd century BC) or war as the foundation of complex societies (Heraclitus, d. c. 475 BC) without always seeming to appreciate just how old and how frequently recurring the idea is (such that it might, for instance, be the sort of intuitive idea many people might independently come up with, even if it was untrue or that it might be the kind of idea that historians had considered long ago and largely rejected for well established reasons) – but this will, I hope, suffice for a basic explanation of why I find the idea of this approach unsatisfying. This is, to be clear, not a rejection of the role of data or statistics in history, both of which can be tremendously important. Nor is it a rejection of the possible contributions of non-historians (who have important contributions to make), though I would ask that someone wading into the field familiarize themselves with it (perhaps by doing some traditional historical research), before declaring they had revolutionized the field. Rather it is an argument both that these things cannot replace more traditional historical methods and also that their employment, like the employment of any historical method, must come with a very strong dose of epistemic humility.

Psychohistory only works in science fiction where the author, as the god of his universe, can make it work. Today’s psychohistorians have no such power.

From XKCD. In a hilarious irony, I have had this very webcomic cited to me as an example to support the notion that STEM folks in general and physicists in particular were just smarter and more capable than all other kinds of specialists; it does not appear to have occurred to the person in question that their own inability to manage the basic humanistic skill of reading subtext undermined their argument.
To be clear, I have tremendous respect for physicists (and zoologists, for that matter) when it comes to their mastery of physics (or zoology); I just wish they extend to historians the same respect (many do, of course).

On to recommendations:

Over on YouTube, Drachinifel has an interesting discussion of how the Royal Navy might have responded to Germany’s ‘Plan Z‘ (a plan to radically build up the Kriegsmarine) over a ten year period. The counter-factual is interesting for two reasons. First, I think the analysis that essentially no version of Plan Z, either the initial proposal or the scaled down version put into production, was likely to actually produce a German navy which would have an even playing field (much less a decisive upper hand) with the royal navy is fundamentally correct. The problem here is an asymmetry of interests combined with the fearsome technical ‘catch up’ the project demanded. The reason for this is the broader interesting point here which is that military plans cannot be considered in isolation because chances are any major build-up is going to be noticed by potential adversaries who will then seek to do something to prepare for it (a dynamic now quite visible in the Indo-Pacific).

Speaking of which, over at Foreign Policy (paywall, alas), I found “China Is a Declining Power – and That’s the Problem” by Hal Brands and Michael Beckley to be a useful read. It is both a good discussion of ‘power transition theory’ which happily moves beyond just using oversimplified and inaccurate descriptions of a supposed ‘Thucydides trap’ which also notes that not only can rising powers disrupt international systems, but so can once rising powers who now see themselves stalling out.

Also in Foreign Policy lately is me, writing on the United States’ pattern of failure in training foreign auxiliary forces and suggesting a range of historical models built either around incorporating (rather than bypassing) local social institutions or completely deracinating the soldiers in question (essentially obliterating, rather than bypassing those institutions). The halfway approach that the United States has often used of trying to impose military institutions developed for radically different cultural environments leads to situations in which local social institutions and ties reassert themselves in the form of debilitating ‘corruption’ which has simply become our name for the mismatch between our expectations and local customs.

Also by me is an essay in The National Interest (not pay-walled, but multiple pages long), comparing the American All-Volunteer Force (AVF) with the Roman experience of transitioning from a conscript citizen militia to a professional, all-volunteer army. I suggest that many of the drawbacks which boosters of the AVF insisted could be overcome are in fact inherant in the structure of any standing volunteer army and so cannot be avoided but must be accounted for and that we should take seriously the notion that the AVF might not be the correct force structure going forward into a period of intensified partisan polarization and great power competition.

Meanwhile, over at the Ancient World Magazine, editor-in-chief Josho Brouwers writes that as of next month he plans to begin focusing on AWM full-time in an effort to get the platform to really take off. AWM is a public-facing ancient world publication (so broadly, but not exclusively, what we’d call ‘Classics’) with its own distinctive editorial tone and a particular emphasis on debunking ideas about the ancient world. Given the planned intensification effort over there, now might be a good time to give it a look and decide if that’s something you’d be interested in reading; AWM is free but supported via Patreon. I certainly know how challenging it can be to build a public education web platform that is truly self-supporting, so I wish them the best of luck.

Also, because we have been talking quite a lot about the First World War, I thought I should mention The Great War YouTube channel. The channel is still active, bu its great fame came from the run from 2014 to 2018 where the channel ran week-by-week summaries (on the 100-year anniversary) of the events of WWI, along with in-depth looks at the weapons, tactics, people and countries involved. It is a truly incomparable resource, though I do find that following the events of the weekly summaries can be difficult if you are coming at it with no knowledge of the war, but see below on that score.

Finally, for the book recommendation, again as we have been talking a lot about the First World War, so I wanted to make two recommendations. The first recommendation is simply my favorite readable summary of WWI, which is the somewhat aged but still relevant J.L. Stokesbury, A Short History of World War I (1981). Truly following the whole of WWI, with its bewildering array of fronts and battles, can be a real challenge; Stokesbury’s history is well written (with a fair bit of understated sardonic wit that can come off as very dry but that is absolutely my jam) and very clear in ways that I have found other summaries of the war not to be. It is by no means a ‘last word’ on the war (indeed, some of its points are a touch out of date) and it can be quite dry, both in its humor but also in the sometimes flat presentation. But I found it importantly easy to follow as an introduction to the war back when I was assigned it years ago and I go back to it fairly regularly. It also has one of my favorite extended examples of Clausewitzian friction.

The more substantive (and more recent) recommendation here is Robert A. Doughty, Pyrrhic Victory: French Strategy and Operations in the Great War (2005), which you will note I leaned on quite a bit in my discussion of the trench stalemate. Pyrrhic Victory is clearly written, easy to follow though not exactly riveting reading, but then for a complicated issue I would rather have solidly functional writing than something more fun but also more confusing. It isn’t a complete history of the war (and one would be advised to have at least a basic grasp of that first), but rather a fairly in depth focus on the way that the French strategy and operational doctrine changed over the course of the war. It is this focus that I think makes it a particularly interesting work because it gets away from merely going through ‘they fought X battle; it didn’t work, then they fought Y battle; it didn’t work’ and instead gets into how each phase of the war forced changes and adjustments within French military thinking as they tried to come to grips with the novel tactical and operational problem that was trench warfare on the western front. Doughty’s approach is very detailed, often narrowing down individual orders and guidance offered by senior French commanders and decisions made on a day-by-day basis (which explains the book’s 578 page length), but that kind of detail is welcome in a topic that is, in the end, all about the details. Most doctrine-studies at this level of detail are written for a very narrow audience and consequently hideously expensive but Pyrrhic Victory is mercifully fairly cheap in both physical and electronic form and so makes a good entry-point (assuming you have a baseline familiarity with WWI) for both the concept of doctrine and also the practical ways that doctrine is actually implemented in an army.

  1. Yes, I am aware that Gat was consulted for Better Angels and blurbed the book. This doesn’t change my opinion of the two books. my issue is fundamentally evidentiary: War is built on concrete, while Better Angels is built on sand when it comes to the data they propose to use. As we’ll see, that’s a frequent issue.
  2. Of course the predictions in the Foundation series are not quite flawlessly perfect. They fail in two cases I can think of: the emergence of a singular exceptional individual with psychic powers (the Mule) and situations in which the subjects of the predictions become aware of them. That said Seldon is able to predict things with preposterous accuracy, such that he is able to set up a series of obstacles for a society he knows they will overcome. The main problem is that these challenges frequently involve conflict or competition with other humans; Seldon is at leisure to assume such conflicts are predictable, which is to say they lack Clausewitzian (drink!) friction. But all conflicts have friction; competition between peers is always unpredictable.
  3. For those confused by the causation, the Mongols are considered the most likely vector for the transmission of gunpowder from China, where it was invented, to Europe. Needless to say, having a single polity spanning the entire Eurasian Steppe at the precise historical moment for this to occur sure seems like a low probability event! In any event, European mastery of gunpowder led directly into European naval dominance in the world’s oceans (its impact on land warfare dominance is much more complex) which in turn led to European dominance at sea. At the same time, the English emphasis on gunnery over boarding actions early in this period (gunpowder again) provided a key advantage which contributed to subsequent British naval dominance among European powers (and also the British navy’s ‘cult of gunnery’ in evidence to the World Wars at least), which in turn allowed for the wide diffusion of English as a business and trade language. In turn, American and British prominence in the post-WWII global order made English the natural language for NATO and thus the ICAO convention that English be used universally for all aircraft communication.

304 thoughts on “Fireside Friday: October 15, 2021

  1. Thank you for this article, I’ve greatly enjoyed it.

    May I ask a follow-up question? What do you think in general about Annales school of thought?

  2. About Psychohistory though. As it turns out it doesn’t even work in fiction.

    Spoilers for a super old book series.

    The Foundation members realize that there is something wrong with the story they have been told about Sheldon and his Psychohistory after the Mule incident is resolved. There was no Sheldon prediction about the Mule and how to defeat him, which should mean that every prerecorded prediction made by Sheldon forever after would be wrong. Instead the next prediction is suddenly right again. From this the characters conclude that there is a Second Foundation guiding them and making it look like Sheldon was able to predict everything in advance. This leads into a whole book about the two Foundations conflict between each other.

    1. Asimov does the same to psychohistory as he does to the Three Laws of Robotics. He takes an idea that seems like it would work, and shows how it struggles when confronted with reality.

      It is interesting to note that the first couple of crisises that the Foundation society is confronted with are actually solved by “Great Men”. While the “recording of Seldon” says that the solution was inevitable about one of them, and the stories presents it in one layer as the inevitable march of history, the solutions are devised by individuals and implemented despite large-scvale opposition in their society.

      The Foundation stories are from the very beginning about the limits of psychohistory and shows that it is not good enough. What it gets to achieve uncontested is near future modelling of existing trends in society, i.e. the breakdown of the Empire that had been going on for decades when Seldon started building his modelling theories.

    1. One thing about Marx, as I understand it, is that the cornerstone of his view of history is competition between different class interests.

      Which at least has the virtue, as a model, that some people in any given era are consciously trying to uphold the interests of their classes. And that nearly every person tries to control and optimize their own lives. So in much the same sense that modeling human societies as being subdivided into competing states has value because humans themselves believe in this and act accordingly… Sometimes you can get at least a moderate degree of predictive power out of expecting class interests or class competition to play a role.

      The problem with the “cliodynamics” approach is that it essentially tries to look at a wiggly line on a graph, then infer a causal mechanism from the shape of the wiggly line. When the wiggly line itself is based on unreliable data, this is utterly pointless.

      To the extent that even just staring at the society itself for ten minutes and trying to figure out who’s the economically dominant class and then analyzing everything else in terms of that? May well be more cost-effective than trying to read tea leaves or wiggly lines.

      1. Personally I think a desire to move up in class is more common than a desire to promote class interests. Social mobility tends to undermine the conditions necessary for Marxist class conflict

        1. I think this is an important point. I think Marx felt he covered it by his references to the petit bourgeoise as he saw few distinctions in the needs of the vast urban working class (which is weird to me) and thought the peasants useless in revolution (even as they were far more likely to not see a chance for change but revolution). It is a failing of Marx I believe. Because of course in urban areas where people will be concentrated on moving up and in rural areas in their social connections (which are in a different way a way to move up).

        2. I’m not super well read on Marx, but my understanding is it goes about like this:

          Very few Individual people consciously desire to promote their class interests, it’s rather that individuals’ understanding of their personal interests and place in the world gets filtered through their class position. It’s essentially the “It is very difficult to get a man to understand something, when his salary depends on his not understanding it” effect.

          The class interest of the upper class is to reduce social mobility (because social mobility dilutes their power and wealth, and introduces the risk that they could become downwardly mobile). Individuals in the upper class further this interest by hoarding their wealth and passing it on to their children, engaging in snobbery and elitism, defending their privileges, and so on.

          Individuals in the lower classes want to move up in class, but they don’t have as much power as the upper class individuals who want to stop them, so they usually fail.

          When the upper class is acting in its class interest (not consciously, but as a product of individual upper class members’ perceptions of their own self-interest) and the lower class is not, the upper class pulls even further ahead. The class conflict at this stage is essentially one-way (in Warren Buffett’s quip, “There is a class war in this country, and my class is winning”). This immiserates the lower classes, who (in Marx’s formulation) are eventually made so miserable that they realize that they have a collective class interest and can only improve their position by acting on it. Since the lower classes vastly outnumber the upper classes, once this realization is widespread and acted upon, you get a revolution.

          Obviously as a proscription this did not hold up in practice.

          tl;dr I interpret Marx’s position as being that social mobility does indeed undermine the conditions necessary for (lower-to-upper) class conflict, but (upper-to-lower) class conflict undermines social mobility, and upper classes generally prevail. That lower class individuals should realize and act in their class interest is, in Marx, not a description of how the world is, but a normative proscription of what should happen in order to bring about a better world. (Though of course, he did also think that this ought-to-is transition would inevitably come about due to the immiseration of the lower classes and so forth.)

          1. “Obviously as a proscription this did not hold up in practice.”

            Or it did happen, but not the way he wanted. Compared to the mid-1800s, European societies have been revolutionized. Aristocratic classes are gone or irrelevant. The dictatorship of the proletariat — universal suffrage — has been instituted everywhere, with even women voting. Several points of the Communist Manifesto have been implemented, like progressive taxation and government control or regulation of much of the economy. I wouldn’t say such countries are socialist in a strong sense — worker control of the means of production is not in play, and social control only in a loose sense — but they could be; even the USA has established that it can have wage and price controls, or break up large capitalists, when it feels the need.

        3. It’s important to bear in mind that Marx’s work was focused on Europe, which had (and still has!) a hereditary aristocracy. No amount of individual wealth guarantees the ability to “move up” into a class whose membership is largely defined by birth.

          Marx’s analysis of class conflict in his own time was that the bourgeoisie weren’t trying to move *into* the aristocracy but to displace it as a center of economic and political power, an outcome that was already finished or nearly so in some places (Britain) rapidly approaching in others (France) and looming on the horizon in still others (Germany, Russia).

          One weakness of this analysis is that it is quite local to Europe. I haven’t read Marx’s work on America, so I don’t know how he proposed to adapt his theory to historical conditions here. It seems to me (and I’m hardly the first to make this observation) that in America race comes closest to filling the role of class in Marx’s Europe: a status group that’s largely fixed at birth regardless of individual wealth or accomplishment.

          1. Matirialist analysis seems to miss social caste wholesale. (Which is what the US has, a racial caste system.) The shoehorn solution is the idea of ‘false conciousnous’ but elides the very real economic value in the caste system to the donminate caste.

            I’m a rather unremarkable attorney, in a common discipline. But if I’d lived 70 years earlier, when the caste system was more strictly enforced, I would not have had to compete economically with some rather spactacular individuals in my field, say Barack and Michelle Obama, because the caste system. Let alone all the other more run of the mill attorneys (black or female), that currently populate the industry.

            For firms and institutions interacting with the top caste(s), this creates a signifigant artificial labor scarcity. Such that when the caste system is threatened or begins to break down, many people in the top caste(s) will see more economic risk from the collapse of the caste system much more profound than from any strictly Marxist class conflit. This destroys the concept of ‘false conciousness’, as this is real economic livelyhood threat.

            And, those in the lowest caste(s) will often gain more from the under mining and destroying the caste system then from assaulting the class system.

          2. Gee, as an attorney, unremarkable or not, I’m not sure competition is really a problem. You know the joke about the lawyer who moved to a small town, and was starving to death, until a second attorney moved to town.

        4. The problem with Marx isn’t that he ignored social mobility, his theory of the future depends on it. The issue is he’s rigged the social mobility element to get the result he wants, either consciously or unconsciously. He has the petit bourgouise decline entirely to be come proletariat, while the grand bourgouise also squeeze each other out and become a socially declining class of increased individual wealth until the economy becomes a tiny few capitalists controlling a small number of gigantic hyper efficient firms, which can be easily taken over by a revolution of the proletariat.

          Needless to say, the next 150 did not play out that way. Reading him now, the results of his theory seem rigged to produce the communist outcome. Marx is a guy playing chess against himself to produce the endgame he likes then calling it science. The tragedy being he convinced a lot of other folks that it was science too.

          1. If you look at the division of wealth tween middle and upper classes these days, his predictions may yet not prove wrong

          2. The premise was it all goes in one direction, so his results are inevitable. That its a lot more fluid and can go many different ways is a major flaw.

            Likewise was the Marxist and then Marxist-Leninist prediction that factory workers were the revolutionary class whose interests would produce the socialist classless society. Turns out they’re a normal socio-political class with normal socio-political class interests, who are also not particularly interested in being a revolutionary vanguard.

    2. I think Popper was wrong when he argued that Marxist theory was unfalsifiable – this can be seen from the fact that it’s been thoroughly falsified.

      (To be fair to Popper, he was talking more about the special pleading that takes place once the predictions fail to turn out. rather than the original Marxist predictions, which are strong but false.)

      1. Since we’re on the topic of special pleading, I think it’s very telling that you see so many arguments of the form “Marxist theory (chapters 1-10) must be wrong because in the section of chapter 6 where he makes some predictions, some of them are accurate, some kinda right and some of them are just wrong. So we can skip reading any of the rest; because anyone who made a wrong prediction can’t possibly say anything worth thinking about.”

        Marxist theory is not, of course, primarily concerned with the generation of prophecies. Rather, it’s a description of how certain economies function and the effects on human beings. Now, are there branches of Marxism that are mostly about special pleading, as above? Sure, but they’re silly, and if “Theory X has some silly loudmouths” is disqualifying then I have some bad news about… well… everything.

        So if you meet someone who wants to tell you about Marx’s prophecies, yes, you should absolutely follow Popper in ignoring them. But if you meet someone who wants to tell you about the labor theory of value, say, or Gramscian hegemony, and your response is “but the guy who described that also made a guess that was wrong so why are you still talking,” you’re the silly loudmouth in that conversation.

        (For the record, you actually can’t tell if Marxism was falsified or not. You see, only economists would really know, but economics is thoroughly falsified by Adam Smith’s failed predictions, so…)

        1. No.

          The labor theory of value and Marx’s predictions are a common structure that stands or falls together. The silly loudmouth is the one who claims that his analysis is valuable even when all its conclusions are wrong.

          1. One of my favorite “what ifs” is if Marx had accepted the regular supply and demand model that was new but not *super* new at the time, instead of the labor theory of value. It would have put everything on a much sounder footing, rather than building the entire theory on a fundamentally flawed theory of value.

          2. JohanL, you act as if supply & demand hadn’t occurred to Marx. All of the classical economists were aware that supply & demand caused prices to fluctuate. What the classical economists, including Adam Smith and David Ricardo, did not think supply & demand could explain was the long-term center of gravity of prices, or what Adam Smith called the “natural price,” or what Marx calls the “exchange-value” of commodities—distinct from the turbulent market price that a commodity might have at any moment.

            Identifying the source of the “natural price” of commodities is important because you can’t explain the accumulation of monetary value, society-wide, through supply & demand. Sometimes market scarcity favors suppliers, and prices go up. Sometimes a market glut favors buyers, and prices go down. But for every buyer there is a seller. It is a zero-sum game, society-wide, unless new monetary value is added into the economy somehow. Keep in mind that we are also not talking about how new “utility” originates (because even Marx understood that utility was not zero-sum in exchanges), but rather how monetary value can accumulate in society.

            Look up Anwar Shaikh’s lectures on youtube about this. Marx was light-years ahead of all of this.

          3. The labour theory of value originates as far as I can tell with Adam Smith, Wealth of Nations, Volume 1, Chapter V, paragraph 1.
            As I understand it, Marx uses the theory to explain how business can even make a profit in a free market – something that supply and demand theory actually finds hard to explain.
            (The problem with Marx is that people tend to see him as either an infallible prophet or as a total charlatan, rather than as a thinker who was as brilliant and occasionally wrong-headed as any other.)


            I appreciate the point you’re making, and I’d have broadly agreed with it before, but that AskHistorians thread I’ve linked changed my opinion about this a while back.

            But also, I feel there are issues the simple approach to falsifying Marx run into. Again, I’m sometimes confused by why the humanities don’t “deprecate” older works, and I can understand some of the sentiments behind trying to deprecate Marx, but it doesn’t seem like Marxist theory is being used in practice in the way we’d think it is.

            Also, there’s the issue of whether we’re dealing with the normative or descriptive content of Marxism; sure, the predictions can be falsified, but falsifying the predictions doesn’t need to make the approach to moral and political philosophy etc obsolete.

            Also consider a case from philosophy; Nietzsche and his work is paid a lot of attention (even in the analytic tradition as far as I’ve heard), despite lots of it going strongly against the current of, say, most discussions in contemporary philosophy. The humanities just seem more adept at drawing out the valuable bits of older work, and less inclined to just deprecate and reinvent them.

          5. I stand corrected. I presume, then, that we should have thrown out Newtonian mechanics as soon as we got good data on the orbit of Mercury, and gone back to just guessing how gravity worked? Obviously the Principia Mathematica is even more of a unitary structure than the collected works of Marx, and so if any of its predictions are falsified then the whole thing must be useless.

            Or did we do right to keep using the parts that worked even though there was, evidently, some further element yet to be worked out?

            Also falsifiable: your claim that “all its conclusions are wrong.” Many things which Marx predicted actually did come to pass, so… I guess by your measure, that means you had a bad conclusion and all your thoughts are valueless?

            (A further wrinkle: the labor theory of course predates Marx and is usually credited to Ricardo, and hegemony comes much later with Gramsci. Now, I can kinda see how if Marx is aLl WrOnG and Gramsci follows or amends Marx, maybe that disqualifies him in your mind. But are you really saying that we can dismiss Ricardo’s work because *after he wrote it,* some other guy who agreed with him came along and made some bad guesses?)

          6. So Marx had a *slew* of correct predictions and got caught out in an edge case?

            Name one non-trivial one.

        2. Predictions are a costly signal that you’re correct. A theory that delivers wrong predictions is a wrong theory. A theory that doesn’t concern itself with predictions is not even a theory. It’s just a set of preferences.

          1. I mean, this sounds like this would throw out literary analysis, a lot of philosophy, formalisms etc. And if we take mathematics, there’s still “theories”, especially theories in the sense of conceptual frameworks.

            Fundamentally, making predictions is just one goal among many; making graphs, drawing pictures, choosing simplifying assumptions and categories of analysis etc. are all non-predictive but are still absolutely essential to understanding things in a lot of domains.

          2. “making graphs, drawing pictures, choosing simplifying assumptions and categories of analysis etc. are all non-predictive but are still absolutely essential to understanding things in a lot of domains”

            Doing such things help you make predictions; predictions are a great way of showing that you *do* understand things.

          3. @mindstalko

            Sure, predictions can help confirm theories. But not all theories are predictive. Heck, evolutionary theory struggles to make predictions


            And there’s still the issue of stuff like literary analysis, philosophy etc.

            Finally, in a sense, historical lenses probably ARE predictive, at least to the extent that they incorporate information about historical correlations. The point seems to be that it’s simply intractable to turn this into a quantitative model of history

            For analogy, there’s no big Theory of Statistics (there’s things like Benford’s Law sure, but those work to limited degrees in a very very small set of cases) that explains in general what sorts of things show up in our statistics, and why; instead there are approaches (Bayesian, Machine Learning etc) that can shed light on particular cases, and methods of analysis (summary statistics, box plots, principal components analysis) that MIGHT help for particular questions about a specific dataset.

            But all this requires lots of human judgement, guided by background theories (mean square errors, plots of residuals, eyeballing histograms), grounded on even more abstract conceptual frameworks (integrals, laws of probability, notions of bias and efficiency of estimators. Note that a whole bunch of these DO rely on assumptions or principles that don’t have anything to do with prediction, at least without a whole host of empirical auxiliary facts). In this very restrictive sense, history might well function predictively; for example Dr Devereaux does seem to make historically-informed predictions about the future of Westeros, or about potential future consequences of a volunteer military force.

    3. “Marx’ view of history” might mean his particular story about the past and future development of humanity, or it might mean his theoretical views about the causal priority of the economic base over the superstructure.
      A lot of our host’s theoretical tools seem to me to have roots in Marx: putting a series of posts on wheat farming on a military history blog is a decidedly post-Marxist move as I understand it, as is I believe the assertion that certain kinds of military organisation or technology require certain economic structures to support them. But saying that an approach is post-Marxist isn’t the same as saying that it’s Marxist. I would presume that Marx’ views on the economic structures of pre-capitalist societies was based on even less data than we have now.

      1. I can’t speak for Prof. Devereux, but a doctrinaire Marxist would say that relations of (economic) production create the fundamental structure of society. Prof. Devereux has made the much less reductionist claim that broader social structures determine military structures. The fact that he talks about grain production doesn’t mean that he thinks it is more fundamental to social structure than, say, religious beliefs, which he has also written about.

      2. There’s Marx the social thinker and Marx the revolutionary – and many of his writings mingle the two. As a thinker many of his ideas are seminal – a strictly materialist view of history (this does not mean no religion but no supernatural forces, whether divine plans or Hegelian ‘world-spirit’); base and superstructure viewed dialectically; the changes in consciousness that comes with forms of production (eg reification) and more. Much of Marx is now intrinsic to sociology – and unacknowledged but there in economics. He also set a trend with his dedication to empirical research – his works are stuffed with an enormous range of data.



            The links above go into the topic of Marx in economics and sociology a bit, but in general I see no reason to think that the ideas that survive enough scrutiny by the academic community to be incorporated into the cores of well-established fields of study are doing those fields damage. The burden of proof to establish something like that seems rather high.

          2. You seem to be quite unfamiliar with the history of ideas that have been incorporated into the cores of well-established fields of study

          3. If you’re alluding to the fact that terrible ideas have been integrated into the cores of established academic disciplines in the past, sure, that’s true.

            But that’s not enough to overturn the general principle that in a sense, the fact that a given idea is central to a healthy and well established field (in this case economics or sociology) is a strong point in favor of that idea. I suppose this justification is defeasible; given a substantive intrinsic case against the idea, it would fail. But in general the burden of proof is on the critic to provide that case.

            Sure, terrible ideas with little justification have sometimes made it into the cores of fields of study.

            But one hand, there’s no reason to believe this has happened for a majority of those ideas (Most ideas that are central to current healthy disciplines are probably good ones; I take it there’s good epistemic reasons to go with the disciplinary consensus, but also intrinsic reasons justify various such ideas).

            And on the other hand, I don’t see that there some approach that outperforms critical vetting (especially in the context of ideas making it to the core of relevant disciplines) by the scientific community when it comes to discriminating between good and bad ideas (at least in this particular context).

          4. When I claim they’re healthy, what I mean is that they make progress on solving their problems and seem to be able to convince prospective entrants of their validity (as well as other scholars in other fields).

            Further, even if Marx in sociology or economics was detrimental to the field, that’s still not going to determine the overall health of the field. Marx’s influence on sociology could be to it’s detriment without that detriment undermining the field as a whole. I see no prior reason to think modern sociology or economics is unhealthy as a field. (Of course, since I think the fields are healthy, that in turns leads me to believe they utilize what they do use from Marx for good reason. This sort of a matter of our priors about the state of sociology or economics).

          5. ” What problems have they objectively solved? ”

            I don’t think I have to elaborate much on which problems economics has solved but here’s an example of a intuitive question it can explain;


            For sociology there is, for example, explanation of how inequality emerges (or why social mobility is lost or maintained)



            “Convincing other scholars is circular. ”

            It’s a valid heuristic that we rely on all the time.

            To offer a much clearer example, we justifiably believe Fermat’s Last Theorem is true because it’s proof has been examined by mathematician and the relevant mathematical community and found to be valid.

            We rely on academic consensus when highly specialized public policy is made in consultation with economists.

            We know Newtonian physics is not completely correct essentially relying on physicists as a community doing the work of disconfirming it.

            I see no good reason to think relying on the community opinions here is invalid. The alternative is engaging in specialized argument and research of the exact kind those academic communities carry out. This not feasible for a vast amount of our knowledge, and would require throwing away a substantial portion of scientific knowledge that we have but are not capable of individually verifying.

          6. We?

            I rely on it in fields of thought where the scholars have proven themselves reliable.

            The replication crisis is severe exactly in the fields that you are touting.

          7. One of the comments of the very link you provide picks apart the solution you are offering.

          8. The interpretation of the replication crisis and its implications aren’t exactly that clear cut.


            Further, how exactly are we deciding which fields have reliable scholars?

            I’d also note that my argument about community opinion isn’t that sensitive to the reliabilities of individual scholars (we only need look at things like the Condorcet-Jury Theorem to know that a group as a whole can be more reliable than its constituent members).This idea is very well explained and further elaborated here;


            Also, could you clarify which comment you mean in your second reply?

          9. They are not entitled to the benefit of the doubt, and granting it would be the end of their fields.

          10. @Mary

            Unfortunately, at this point I don’t think we have much to productively discuss on this topic. We just seem to have reached an impasse. I value having discussed this; thank you for taking the time to reply to these comments.

    1. Thanks, I gives some insight into US policy thinking. I dan’t buy most of their arguments for reasons I am too lazy to go into but it reaods a lot like the “Russia is doomed” theme we also see. Patrick Armstrong in his blog post RUSSIA, RUSSIA, EVER FAILING has a probably incomplete list

      A compendium of doom from the “experts”: Russia will fail in 1992, finished in 2001, failed in 2006, failed in 2008, failing in 2010, “rapid deteriorating economy” in 2014, failed or declining in 2015, failing in 2017, negligible economy and “rusted out” military in 2017 (“Russia’s coming attack on Canada” is an exceptional fount of worthless analysis: hardly a correct statement anywhere, starting with the sub-head), falling behind in 2018; headed for trouble in 2019. Russia’s isolation, ancient weapons, instability. A gas station masquerading as a country. Doomed to fail in Syria and losing influence even in its neighbourhood in 2020. One “expert” repeats himself as if the intervening decade had not passed.

      1. That blogger seems to have responded to American ignorance with Russia to with full on Putin-Russia fanboyism.

        1. Well yes. He is a retired Canadian Government “Russian analyst” who collects Western idiotic statements. I believe it’s a hobby with him. Of course when working, one of his jobs probably would have been to help ensure that the Canadian cabinet and the Prime Minister had a bit more realistic understanding of the USSR or Russia strengths and weaknesses.

      2. I think my favorite formulation of this idea is how western experts have predict 10 of the past 3 chinese crisis.

  3. Is there somewhere that I can get an overview of the efforts to predict what Roman Italy’s population was? I’m vaguely aware of ML modeling of China’s population in the present and I’m wondering if that has been applied elsewhere to see if ML can deduce anything.

    1. Probably the best place to start in terms of getting a sense of where the issue stands today is L. de Ligt, Peasants, Citizens and Soldiers (2012). Brace yourself, the issues the estimates get based on are fairly technical.

  4. There’s a story that Bismarck was asked what was likely to have the greatest impact on the future of Europe. He replied, after thought ‘that the United States speaks English’. On can easily envisage a world where a large north American state speaks French, or Spanish, or Iroquois (or pidgin Arabic – see the hilarious Journey to Fusang by William Sanders).

    A Muslim historian thought that the greatest impact of the Mongols was the sack of Baghdad. It not only extinguished a major centre of thought and learning, it called into question the fundamental premises of the Islamic project, with consequences that echo down to the present day.

  5. Never go full Platonist.

    “Raw metals are a classic example of a commodity” — except even they aren’t, if one looks hard enough. E.g.
    “Commodity” is not a trait of a thing, but a decision (modeling assumption) to treat things as identical.

    “Both cliodynamics and psychohistory assume these differences and problems ‘come out in the wash’ over a long enough period and a big enough sample. […] apparent randomness (which one may argue is merely complexity on a scale that is beyond simulation)”
    Indeed, what other interpretation is there, that doesn’t make absurd metaphysical claims? People tend to use “random” in a few related meanings: “representative/prototypical example” (e.g. “Bibulus was some random senator”), and “I won’t bother to model this in more detail because I expect that doing so would not be worth the effort”.
    (Computing/cryptography would have done everyone a favor if its term would be “sacred/profane numbers”, because it makes a claim about the provenance of the numbers and the requisite handling procedures.)

    “The geography atlas is not the terrain.” To be useful, maps necessarily omit details.

  6. “That said Seldon is able to predict things with preposterous accuracy, such that he is able to set up a series of obstacles for a society he knows they will overcome.”

    Really depends on how you view the First/Second Foundation relationship, because in cynicism;

    A bunch of psychics herding the intellectual and technical populace, who in turn herd the manufacturers and producers, only need to have a steady guideline that they all follow in order to avoid infighting. Whether or not it’s wholly accurate doesn’t track – the philosopher-psychics will ensure fidelity.

  7. I know physicist Sean Carroll was asked about psychohistory and he had another problem with it as well – in a gas, the interactions between particles are very simple and very weak. Humans can interact very strongly, and they can have a wide variety of different interactions with a wide variety of outcomes. This means that the equations of state would be so complicated as to make the project impossible. Presumably, if you knew that original state of the universe and all the laws of physics you could make predictions (statistical predictions, assuming quantum mechanics is truly random), but that is not a practical research program.

      1. For what it’s worth, Asimov’s psychohistory doesn’t work well with planet-sized populations; it needs to go bigger.

      2. Indeed, a part of Asimov’s background given for psychohistory was that it was supposed to model the society of an entire galaxy, with all the orders of magnitude involved. Make of that what you will, but at the very least, it’s some lip service paid to the problems with the concept that our host describes.

        1. And the Galactic Empire might have had a lot more reliable data for the thousands of years of history between now and then.

      3. The best defense of psychohistory is Doylist, but Watsonianly I think it is worth pointing out that psycohistory was designed for dealing with galactic events. IIRC the “it will come out in the wash” assumption explicitly relies on a scale where you can treat inhabited planets as the smallest political unit to be explicitly modeled.

        1. I think the better watsonian issue is that Seldon secretly sets up a group to push things along to make sure that galactic macrohistory falls in accord with what he ‘predicted’.

      4. The human population is large enough that, if we were gas particles, we could be modeled using statistical mechanics.

        The clearest way to see this is in the use of PIC (particle in cell) codes in plasma physics (plasmas are electrically charged gases). The code approximates the motion of the ~10^20 particles in the plasma using only millions of particles. For a lot of circumstances, there isn’t much difference between the statistics of millions of particles and the statistics of 10^20 particles.

    1. i can’t believe any technical person with even a passing knowledge of non-linear phenomena be it turbulence or what have you can entertain the ability to predict all of history with fancy math; we can barely predict where an airfoil will stall or how much noise it will make with the substantial computing we have at our disposal.

  8. Typos:

    “career of our fellow Timujin is a necessary cause” – Timujin or Temujin?

    “The channel is still active, bu its great” – but

  9. Temujin’s success appears to the casual reader to be equivalent to Asimov’s ‘The Mule.’ A person who breaks the predictions, a confounding variable. Or arguably, an outside context problem, something that happens that no one could see coming, because it is beyond the imagination or ability to predict.

    A Black Swan event.

    However, while I would agree that cliodynamics makes claims it cannot prove, these are made by human beings, not the actual discipline that is evolving. If it is truly a science, and not wish fulfillment, over-time the rigorousness will strengthen.

    Whereas ‘Classics’ will always be subject to new interpretations, and just because the current model seems to be more grounded by evidence, there remains the hole that there’s no theory/model of classics that can be used to test the validity of the model in the way that physics can test the validity of a theory/model.

    Arguably, this is what differentiates a STEM subject from an Arts subject, and like religion versus science, one should not be used to question the other, because they’re not speaking the same language.

    I don’t know, but at least I know I don’t know.

    1. “If it is truly a science, and not wish fulfillment, over-time the rigorousness will strengthen.”

      The problem isn’t just lack of rigor, but lack of material. The necessary data on Rome (or the Ottomans, or the Aztecs, or premodern China. . .) is *gone*, and can’t be gathered now. Maybe 5,000 years from now the US Census will still be in operation, and we’ll be able to draw inferences like “eight out of ten civil wars were preceded by X, Y, and Z.” But that’s a sci-fi scenario only slightly less fanciful than Asimov, not a useful guide to the present reality.

      1. Absolutely. I don’t imagine that cliodynamics will have much to say about data-sparse historical events that classics cover, but if civilization doesn’t collapse, and if we maintain our ability to collect data, then I could imagine a future where it would be a useful tool. YMMV of course.

      2. “eight out of ten civil wars were preceded by X, Y, and Z.”

        There is something similar in a previous Fireside Friday discussing stasis and the fragility of self government. He doesn’t quantify the fraction of democratic backslidings that are well described by stasis, but this seems doable even with the limited classical data. Civil wars are things that people are likely to write about.

        1. I will avoid all political (in the sense of partisan) commentary, and note that I didn’t find that post very impressive. Modern republics, populated by millions of WEIRD people, are so different from ancient democratic city-states that the latter don’t really furnish useful models.

    2. If your predictive model isn’t developed enough to predict anything with any accuracy, what value is it? I’m not saying that resources shouldn’t be allocated to “pure science” research, but the advocates of cliodynamics are pushing for that discipline to replace the current model of historical analysis. And cliodynamics simply doesn’t work for “black swan” events like Temujin; except Temujin are all throughout history! So it’s a discipline that doesn’t work for its stated purpose of prediction.

      At best, cliodynamics is an area of study that deserves further development but it’s hardly the sole tool either for historical analysis or predictions of the future.

      1. “[T]he advocates of cliodynamics are pushing for that discipline to replace the current model of historical analysis.”

        Really? Name one. I don’t recall Turchin saying anything to that effect in anything I’ve read about cliodynamics.

      2. Temujin, Alexander and Napoleon are part of a very small group of extraordinary individuals which happened to be in an extraordinary context. Napoleon would not have been chief of state under the Bourbons, Alexander would have probably reduced the size of his conquest if Phillip lived for another 5 for 10 years (he could have disinherited Alexander or killed him). Temujin might have been stopped cold on his tracks by the Tanguts if he lived 50 years earlier.

  10. I think you are strawmanning the cliodynamics a bit. A model doesn’t need to be universal to be useful. To be more concrete, a model that describes a certain behaviour of an agricultural society (as Turchin does in his Secular Cycles) says nothing about such shocks as a horde of steppe warriors appearing out of the blue and conquer everything. As an analogy, classical physics is inapplicable at near-light-speed velocities but there are plenty of useful things you can build using its concepts.

    1. Regarding the incompleteness of quantitative data in the past, it’s less important to establish a baseline than to understand the trend. Would you say that it’s impossible to say anything definite about trends for the ancient Rome or medieval societies? Things like silver content of coins, provenance of elites, coin hoards, wages and land ownership?

      1. Well isn’t that Brett’s point… the majority of that ‘data’ is conjecture. An error bar greater than 100% is the difference between ‘is there’ and ‘isn’t actually there’.

        There is knowledge in the past, but the past is both so vast and so poorly recorded that the knowledge will always be at best a guess. Or, at least, a guess until some point 1,000 years from now when every human being on the planet has left behind 400,000 pages of records for 1,000 years.

        1. Probably I wasn’t clear enough. Turchin’s arguments don’t require exact knowledge of any of the parameters but rather of how they changed. To take the census as an example, the historians interpret the results differently and hence there is a 3-5 times difference that Brett mentions. However it’s not one census, we have a pretty long data series, and what is important for Turchin’s model is how the population grew, peaked and declined.

          It seems like the trends should be more reliable than point estimates, as whatever problems affect the accuracy would presumably affect all the data points similarly. Turchin says that the consensus among historians is that the republican census numbers “can serve as a reasonably solid basis for reconstructing the *dynamics* of the Roman population,” citing Scheidel 2001, 2004. I’m not an expert so I’m wondering if Brett would agree with this.

          1. IIRC (I studied the debate on Roman census figures about ten years ago), there is a debate whether the population of Italy was increasing or not in the century before Augustus’ putsch. One problem is that on the order of a third of the population had probably avoided registration just before the Gracchan laws which gave registered citizens access to free land. We would expect non-registration to increase at times when Roman wars were not very rewarding to Roman citizens, and decrease when there were rewards which you had to be on the census rolls to claim. Then the census was politicized in the lead-up to and aftermath of the Social War. So the accuracy of the census probably varied due to specific historical circumstances, and not just at random based on how competent this year’s censor was.

    2. Similarly, let’s talk about the ideal gas law since Brett uses it as an example of a scientific idea that can’t be ported over to humans. Perhaps Brett hasn’t studied the sciences thoroughly enough to learn that the ideal gas law is never true! And its deviations from physical reality are often not small! It’s reasonable to assume that Turchin understands that his efforts will no lead to perfect predictive power.

      Overall, this is a notably weak takedown from Brett. I haven’t seen anyone, even Turchin, claim that cliodynamics has achieved its aims. The idea that it’s useless simply because it hasn’t or can’t predict Genghis Khan is laughable.

      1. Even if Turchin really was that modest, unfortunately the popular view of the field is that similar grand pseudo-mathematical claims ARE reasonable – even for more minor things, like predicting who will win the next presidential election because all of the last 14 elections have favored the candidate with (insert subjective, overfitted model here). And it’s just… not true.

        There are two big issues with such theories: do they work “normally”? And just how frequent are the Black Swan events like Genghis Khan that maybe we’ll give them a mulligan for? I’d argue that Black Swans are exceedingly common – in just the past 30 years, we’ve had Gorbachev + a peaceful fall of the USSR, the Internet, cheap cell phones, Brexit, President Trump, and many more things that were hard to predict yet amazingly influential. But I don’t think these models even have much useful to say ignoring these. They’ll predict a recession or a war or a technology, and then just wait until it actually happens, never mind if it takes 2 years or 20. It’s like an old trick psychics use, just make a bunch of predictions then declare victory when one of them comes true.

        1. There’s a “popular view” of Turchin’s work? Come on.

          Also, it’s hilarious that you point at psychics throwing s**t at a wall to see what sticks, while simultaneously identifying events from the last couple years as “amazingly influential” on a historical scale. Who is going to remember Donald Trump in 1,000 years? Even Gorbachev is likely to be an obscure footnote.

          1. No, I said “popular view of the *field*”, not Turchin. He’s just one example of many authors who go on TV to trot out Foundation-esque predictions for presidential elections or the economy or whatever based on historical cycles or some such, and are unfortunately taken seriously. (Not that all future prognostication is bad, just not this flavor of it.)

            You’re claiming cliodynamics is potentially useful but incomplete. I’m saying it will never be useful, ever, and using the utter inability to predict even smaller timescale stuff as an example.

          2. @SnowFire

            “Many authors” whom you choose not to name. Someone like Nate Silver is doing something entirely different from Turchin, for the record, but he is the only pundit I can think of who cloaks his predictions in substantive data. Other pundits who say things like “the incumbent party will lose the coming election because incumbent parties typically lose during these economic conditions” aren’t actually engaging in serious data-driven political science (they just want you to think they are).

            If you think that short-term events can prove or disprove the viability of Turchin’s work you are missing the point entirely. A question like “will Donald Trump win the 2024 presidential election?” is way, way outside the scope of what he is trying to do.

          3. Then you get into the problem identified by John Maynard Keynes, “In the long run, we’re all dead.”

            On some level these subjects are interesting because the influence and inform real life. The point where we are ignoring a Gorbachev because he will be a foot note in a 1,000 years, is also the point where we ignoring someone who profoundly effected the lives of probably billions of people, not to mention was arguebly THE leading figure in the peacfule end of a conflict that threatened the continued existance of hummanity as a species.

          4. “Even Gorbachev is likely to be an obscure footnote.”

            Who was one of the people who ensured that the fall of the USSR was not WWIII.

          5. Depending on what happens in the next 1000 years, the most well-known person from our time might well be not someone who is famous today, but some random dude whose diaries happened to survive the Great Purge/Fire/Cataclysm.

        2. Anyone who is pretending that some kind of model can predict the outcome of a presidential election is either a fool or a crook. But Brett does not talk about these people in this post. He has mentioned Turchin several times who has never claimed any such thing.

          Regarding the black swans, it’s obviously true that they happen and that we have had a lot of technological black swans in the last 100-200 years. But they impact any prediction based on history, whether quantitative or qualitative. I’ve enjoyed the article on the All-Volunteer army, in which Brett compares the US army with the Roman one, making the case that the US army will have the same problems that the Romans had (an implicit prediction). If in 20 years all the fighting will be done by autonomous nano-bots then the Roman experience might be less helpful.

          In other words, this critique can be employed against anyone who is trying to apply history’s lessons to the present and future.

        3. Covid-19 for another black swan event. Which interacts with the result of unlikely electoral outcomes (Brexit, Trump’s victory) — President Clinton’s covid response would have been drastically different.

          And global warming looming over us, for yet more electoral interactions — imagine Al Gore had become President in 2000. Very different executive actions, at a crucial time.

          1. I see the glimmerings of a theory here: that the level of religious dissent was constant, as was the Church’s ability to co-opt it (as in the Franciscans) or suppress it (as in the Cathars), but the printing press upset the old equilibrium. Unfortunately, it’s difficult, probably impossible, to set up a scientific theory with predictive power that includes unpredictable technological innovations.

        4. And, in another era, we had Luther, Calvin et al. Did anyone in 1500 predict the religious fracturing of Europe as the central event of the next 150 years? Does anyone now have a theory (that isn’t pure special pleading) that generates such an event?

          1. This is similar to asking Newton mechanics or classical thermodynamics to predict tomorrow’s weather. They are unsuitable for this but useful for other things.

          2. I think that is a case where you can see the fractures starting before, but the exact nature of the faultlines is clearly shaped by individual people: The 1400’s and 1500’s saw an intensification of various kinds of religious activity among laypeople, so I think *some* kind of change from the medieval status-quo (if such a thing can be siad to exist) was coming, but it’s not certain it would have been anything like the Reformation(s) we got.

          3. Arilou is quite right. There was a lot of religious experimentation going on in the fifteenth century. Sooner or later the Church was going to lose control. Luther gave voice to a general dissatisfaction with the Church hierarchy. The Reformation might have taken different form without him, but something of the kind would have occurred.

          4. Was there measurably more religious experimentation in the 15th century than in the 13th or 14th? Cathars, Franciscans, etc.? (Some of those groups ended up inside the church, others outside, but none until the 16th ended up creating a durable separate church.) Where is the theory that predicts that outcome, or the wars, revolutions, etc. that followed it?

          5. The religious experimentation was going mainstream as dissatisfaction with the corrupt Church hierarchy grew.
            As I said the church managed to deal with earlier dissent by persecution or cooption but the momentum of resistance was growing and beginning to infect conservative thinkers. Erasmus for example, or Marguerite of Navarre. When members of the establishment start getting radicalized the tilting point has been reached.
            The printing press with it’s rapid disemination of ideas was a major contributor.

  11. The comments about trying to build complicated castles of analysis on foundations of statistical sand reminds me of a lot of discussion in (especially) the early stages of the COVID-19 pandemic. I remember LONG discussions in which people were combing through information put out by different countries and building up lots of ideas about why the death rate was so different in different countries…when by far the biggest difference was how good different countries were at detecting positive cases. Same with things like hospitalization rates, hard to use that for analysis when you have things like South Korea hospitalizing pretty much everyone who tested positive until very recently.

  12. In defense of Asimov, I would point out that Seldon cheats. He has also set up an inner circle of psychics who exercise considerable power in the dominant political unit when he starts out his little project who have the job of gathering immense amount of data, formulating new predictions, refining old predictions and the like. They also, when necessary, push things along so that the political state of the galaxy conforms to the predictive set. If those historians with their over-expansive models had time machines to go back and ‘fix’ all the messiness in history away, who knows what they could do?

  13. If Peter Turchin actually said “a basic premise of our study is that historical societies can be studied with the same methods physicists and biologists used to study natural systems” that would suggest that he’s either unaware that physicists and zoologists don’t use the same methods, or that “natural systems” means different things in physics and zoology, then he’s either under-read in physics, zoology or both, or basically dishonest. Either way he can probably be written off with no further consideration.

    My opinion only, your mileage may vary.

  14. I haven’t read the Foundations books in over a decade but I seem to recall that people with mind control abilities from the Second Foundation are helping keep the whole thing on track which provides at least a partial explanation for why Seldon keeps getting things right. Although even then their intervention doesn’t look like it would really be enough to keep events so closely aligned with Seldon’s predictions.

    1. Sort of. The Seldon plan works fine without the Second Foundation’s intervention until the Mule emerges as a 1-in-a-million outlier to ruin it. After that, the 2F intervenes more actively, but is a desperate action on their part to salvage a broken plan. That’s one and a half books (and 300 years) of Seldon succeeding on his own terms.

      1. n.b., this is based on the original Foundation trilogy, not the later books which I haven’t read but which I understand to retcon a bunch of what was going on in the originals.

    2. AFAIK Asimov wrote the Foundation stories on the fly. So the original idea of psychohistory was to stand on its own. The Mule, and then Second Foundation, being psychic, came in only later.

  15. I read through your article on the peacetime draft in the National Interest, comparing it to the abolishment of Roman conscription as part of the Marian reforms and I find myself unconvinced.

    The presence of peacetime conscription into federally-contrlled armed forces in the United States is an anomaly rather than the rule, with the vast majority of American growth in the 19th century (including virtually all of its foreign wars) happening under the all-volunteer army. This is very much unlike Roman conscription, where, as you have pointed out in the past, was very much a constant thing for the majority of the Republic’s existence, due to the Republic basically always being at war. In a way, the transition from American Volunteer forces to Conscript forces in the aftermath of the Korean War is the opposite of what happened in the Roman Empire – with the US, having obtained its hegemonic position mostly without a conscripted army, now turned to conscription in order to maintain theri hegemonic position.

    I also disagree with your charactherisation of the failure of US Army expansion during the Iraq/Afghan Wars as being related to the all-volunteer force.. All-Volunteer forces have, in the past been able to transition to large, mass conscripted armies in the way described by the Gates Comission in times of crisis, producing effective fighting forces in a (relatively) short space of time (a few years) – the vast expansion of the previously all-volunteer British and US Armies on their nations’ entry into WW1 and WW2 come to mind (you could even cite the example of the full, all-volunteer expansions of Canadian/Australian/New Zealander forces in both wars). This was despite the British defence spending pre-war, at about 3% of GDP (same as the US now) I’d argue the far bigger factor in the inability of the Roman Empire to expand its forces was due to the far bigger burden placed on the Roman state by the army – the vast majority of Imperial revenues (80%) were already being poured into the army, and the low productivity and inefficiencies of pre-modern agricultural economies would make substantial increases in Imperial revenues impossible. In contrast, the military today is only about 10% of US Government spending, and the ability of modern governments to borrow money far more easily than pre-modern states would allow this epxansion to take place. You could argue that the political will doesn’t exist to expand defence spending to a significant degree for a long unpopular war like in Iraq/Afghanistan, but a clear and present danger tends to produce this political will (see examples from the world wars above)

    As for political control over the military, a fully volunteer force doesn’t necessarily lead to issues with breakdowns in the trust between the military and civilian leadership. Both the United States and United Kingdom had all volunteer militaries for the majority of the 19th and early 20th century, and in neither situation did civil-military relations break-down in the way we’re seeing now. I’d argue that what we’re seeing now is a different case in a different set of circumstances related to the rise of a political movement that has worked to erode trust in non-miliary governmental institutions, rather than an inherent problem in professional all-volunteer armies themselves.

    1. I’ll tackle just one point you raised: the capability of US and British military to expand with conscript man power. That was not achieved easily. First of all, it was based on using the existing Anglo-Saxon custom of local magnates recruiting their own forces. Such units were used in Boer War and the Spanish War, which also demonstrated the limits of the practice. Subsequently, the yeomanry and state militias were regularised into Territorial Army and National Guard, respectively. Then, prior to WWI, it took over ten years to create the structures that allowed conscription to be introduced. In fact, some of those features were created only between the world wars: neither Anglo-Saxon country had a proper reserve officer training system during the WWI, unlike the Continental belligerents.

      Now, the structures upon which the previous expansion of the conscription was based are only a faint memory. There are no structures or cultural traditions whatsoever that would allow the massive numbers of people to be trained, and no ethos that would could be used to give legitimacy for the draft: “conscript” is used generally as a synonym for “badly-trained” and military service is seen as a specific profession requiring clearly unusual suitability. There is no upper class that would have a tradition of military service, nor the idea that officer background would be a general advantage for one’s career: thus, no ready pool of leaders upon which to draw for the cadres of an expanded military.

      1. Those structures still exist, however- the US still has its National Guard, many NG units were deployed to Iraq/Afghanistan, and the ROTC still provides a number of reserve officers.

        I mean, sure, an American mass mobilisation for a possibly unlikely near-peer conflict (that doesn’t end in a nuclear exchange) might not produce an army ready to fight a war in a few months – but does it have to, given the US’s unique geographical advantages? (i.e. being oceans away from any potential land war).

        And in the meantime, why is it justifiable to deprive hundreds of thousands of young men and now women of the first years of their adulthood to sit around in some military camp when they could be discovering the next cure for cancer?

        1. In its current shape, the National Guard is sort of military reserve, but it is so actually very close to being a regular force, exactly because it is deployed with some regularity. The same goes for ROTC: the ROTC doesn’t really train reservists: most graduates serve at least four years in active service. It does give some officer reserve, but nowhere near the number of people that a drastic expansion would require. (My own country, Finland, has a rather extreme example: the Finnish Defence Forces has, in full mobilised strength, about 95 % conscripted reservists. Even officer corps would consist around 80-90 % of conscripted reserve officers.)

          The whole structure of an army based on conscription is very different from a professional army. The peace-time establishment of a conscription-based force is centered around training function. The conscripts and the units consisting of them are a product of the system: they are inducted, trained and (perhaps after a relatively short time of providing an active force), discharged to serve in the reserves, then, called up every few years to refresh, until they are old enough not to be capable any more. Similarly, most units exist only in deactive mode, their personnel being reservists working in civilian occupations.

          When you do it well, the conscripts are not “sitting around” but their service is, very much, a single long boot camp, with every day of service programmed to produce, at the end, a fighting unit for the reserves. Most importantly, even if you wish only to have a capability to enlarge a military to, say, ten-fold from its normal strength in a year or two, you must have the necessary training infrastructure and personnel readily available. (A well-functioning conscript force can do this in the order of days, by mobilising the reserves.) We both know this infrastructure and personnel don’t exist in the US. The US Army took about three years to reach an actual mobilisation man power levels in WWII, beginnig this work in 1939 (with the CCC building large parts of the infrastructure from mid-30’s onwards). In the current situation, the infrastructure for mobilisation is much worse. It is not the question of inducting young people to serve. It is the question of not having places to have them sleep, weapons for them to use nor people to train them. And because a professional army looks entirely different from a conscript-centered force, it definitely doesn’t want to have this infrastructure.

          Of course, the actual answer is in the end of your text: the USA doesn’t need a large army for strategic reasons, and quite as well, military service is not really a worthwhile pursuit. This, if anything, is a very Roman mindset: essentially just the thing that Dr. Devereaux was referring to.

  16. I’ve been looking for a good history of WWI that includes things like what the heck was happening in Eastern Europe, not just what was going on in Flanders.

  17. Regarding your recommendation of The Great War on Youtube, it’s worth noting that the same team is currently in the middle of WW2, on the same one-video-a-week basis. is their channel for that. They’re doing things 79 years later, which puts the current timeline into the middle of Stalingrad.

    (They also have an inter-war series on their “home” channel, TimeGhost History. But that compressed 21 years into a few dozen videos, so it clearly wasn’t real-time.)

  18. In short claiming that impersonal and somewhat magical forces shape history and individuals have no impact is just wrong. Obviously there is interaction between social conditions and individuals; Napoleon himself claimed that he could not have achieved what he did without the conditions created by the French revolution, but contra-wise the French Revolution didn’t gaurentee a Napoleon figure. In all likelihood without Napoleon there would have been no French Empire. Without Alexander there would have been no Hellenistic kingdoms. If he had lived longer and left a mature heir there might have been an Alexandrian Empire instead and what effect would that have had on the development of a Roman Empire?

    Personally I just love playing around with alternate histories.

    1. Well there will always be individuals to exploit a moment or rise to prominence. But without Napoleon there would have been Gen. Alexander Dumas and it is highly unlikely, he would have chosen to conquer Europe or having done so, march into Russia despite his supply lines being broken and his troops being routed by dysentery. He would no doubt have made piece with some constitutional monarchy, had a very different relation with French slave holdings in the Caribbean(having been born a slave himself) and the history of Haiti in particular might be very different. The history in Africa might be very different. Yes it might have still had many wars with Germany and two world wars or perhaps not. Perhaps the revolution would have succeeded in Germany had he risen to power not Napolean and so there were no Hapsburg left to reign or marry. That might leave a different bloodline with Victoria (if she were even Queen) and so no bleeding prince in Russia, a more rational Tsar and Tsarina and perhaps even a willingness to compromise with the democrats–or a revolution before Marx so no Lenin, no Stalin. There are darker turns of course the lack of Napoleon could go as well but I’d rather not argue for the benefit of dictators. The point is that as you say, individuals matter. And really, not just one individual. Robespierre gets tagged by non-historians with the terror but he in fact was not looking for bloodshed. That was Marat, he who was killed in the bath. Robespierre disapproved of the French Revolutionary Wars too and without them, no Napolean. So had we simply had Robespierre and a different mix of people, history might be quite different.

      1. Yes, exactly! Personalities matter. Unfortunately being enslaved oneself doesn’t necessarily make one anti-slavery, as history shows with great clarity.
        There wasn’t much if any Hapsburg blood in the Hanoverians and anyway Victoria’s haemophilia seems to have been the result of a mutation in her or her mother. Victoria herself pointed out that there was no trace of the disease in her family or Albert’s. Now if Princess Charlotte had lived to become Queen. Or if King William IV’s little daughters had lived history might be very different.

        Isn’t this fun?

        1. Now if Princess Charlotte had lived to become Queen.

          Steve Jackson Games did an ebook game supplement about that world, which they called Britannica-6.

          tl;dr: “On the timeline known as Britannica-6, the 19th century is a time of great inventions and huge engineering projects – but this isn’t the Victorian Age! Queen Victoria was never born, and the Georgian period never ended. Royal “Bloods” and outrageous dandies gamble fortunes on aerostat races and sponsor gigantic war machines, while Luddites grumble and the secret agents scheme.”

        2. Either Catherine of Aragon dying before Henry VIII decided to divorce her, or Henry VIII dying after Mary was born but again, before he decided to divorce Catherine. . . .

          1. Or that son born in 1511 surviving. The existence of a healthy male heir would have changed everything as far as Henry was concerned.
            If Jane Seymour had survived to give Hentpry a second son Mary Tudor might have been allowed to marry Philip of Bavaria and been a much happier woman with the children she longed for.
            A really fascinating possibility is if Henry VIII had succeeded in marrying his elder daughter Mary to Charles V and his younger daughter Elizabeth to Philip II in 1545/46. The possibility of Elizabeth at the head of the Spanish Empire is a fascinating one.

    2. I had intended to bring up Napoleon myself, because he really did change history single-handedly. he did not start the French Revolution of course. I think most historians conclude that some kind of radical shake-up of the Ancien’ Regime was inevitable just for the sheer economics of it.

      But Napoleon completely changed the board in such a way that the Revolution survived (and then he killed it later and that’s an entirely different crazy story). There was a very realistic chance that the First Coalition would have defeated the Revolutionaries if not for The Boney Man’s utter and repeated humiliation of the coalition in Italy. Napoleon would, for the rest of his life, rack of up one of the longest and most brilliant military records in history (and getting hundreds of thousands or millions of soldiers killed). However, while his subordinate commanders were often quite good, not one of them displayed the kind of military genius of Napoleon. He seemed to understand war of that era on an almost-instinctual level. And even at the end, both at Paris and, later, Waterloo he kept somehow conjuring victory and staving off defeat out of almost nothing until overwhelmed by numbers and sheer bloody-minded tenacity.

      And going along with this, he completely upended the political situation in Europe and much of the world. Just for one specific instance, the United States had a relatively quiet border with Spain at the Mississippi. When Napoleon claimed back that territory, he offloaded it at a bargain price o the U.S. of A because he needed the money to carry out his wars. Napoleon’s personal ambition took the land from the Spanish Empire, and also created the situation where America had a legal title to the territory (and nobody asked the current occupants their opinion, whether they were French, Spanish or Native American or American Indian) while ensuring that neither Spain or France were in a position to hold the territory AND that Britain had enough to deal with and wouldn’t even think about intervening. This also caused a cascade of events later by placing American immigrants in the prime position when newly-independent Mexico wanted to strengthen its hold in the Tejas region and invited in settlement. And then all that eventually led to Texan independence from Mexico, the rise of President Polk and the American conquest of the south and west.

      Would some of this have happened anyway? Probably, but how can you even define which part? Would all of it have happened absent Napoleon? Well, we can’t reroll the dice on history but you’d* need a huge argument to explain it. And this is only one small way that Napoleon had an impact on events, which might not have happened at all had he, say, been shot by stray musketry in one of his battles.

      *The abstract “You”, not a particular individual here. I’m not putting these words into anyone’s mouth but askign a rhetorical question.

      1. H. Beam Piper’s Paratime series includes a short story where Napoleonic war diplomat Benjamin Bathurst walks into an alternate timeline where there is no war, his diplomatic pouch causing consternation among the German, French and English ministries. Sir Arthur Wellesley of the Foreign office writes that there is an Artillery general, Napoleon Bonaparte, in French service who is attracting attention for his brilliance but is unquestionably loyal to Louis XVI.

  19. I am quite reminded of Karl Popper’s “The Poverty of Historicism” about how these grand theories, which he terms “historicism”, never pan out. I feel like Turchin is a remarkably anachronistic example in that mold, and at the same time I wonder if that means he will be as pernicious as hegelism or historicist marxism.

  20. On the XKCD comic, it’s not that physicists are smarter than anybody else. Math talent is more easily detected and ranked at a very young age than most other forms of talent. (Athletic talent and beauty also fall into this category.) Young people tend to arrogance, and XKCD follows. On the other hand, I know more people with serious math talent who are quite literate than people with serious humanities talent who are quite numerate. This may not be a matter of innate characteristics. I suspect that this might be due to a combination of sexism and aristocratic posing. And maybe I’m not a very good observer.

    Math talent, btw, has a heavy overlap with physics talent but is not quite the same thing. A number of very good physicists are relatively weak on math (Faraday had almost none), and a number of very good mathematicians have little physical intuition.

    1. The point of the comic is how annoying and ignorant the physicist is. This is most obvious in the alt text.

    2. I would take strenuous objection to the claim that math talent is easily detected and ranked at a very young age.

      Rather, I think the better frame is that we have a particularly arcane and unintuitive manner of teaching mathematics, which appeals to certain people very early and misses most others entirely. Some young people are fortunate enough to come across teachers early on who manage to connect with them and transcend the indecipherable abstractions; most others are not so lucky. But it’s entirely possible to develop facility much later in life–what the field calls “mathematical maturity”–particularly when one finally encounters means of explaining the underlying concepts in a way that can generate the proper intuition.

      That “early talent” thing is not some unique property of the individual; it’s a fortunate match between teaching methods and the individual’s comfortable modes of understanding.

      (I say this as both a sometime math educator and a current machine learning practitioner.)

      1. Oh . . . thank you, thank you, thank you! I am one of those whose early mode of understanding had virtually no connection to those traditional teaching methods. Sometime in my late 50’s I was exposed to the idea of using manipulatives to provide a conceptual tool for learning math, and it was as if the proverbial light bulb went on. A bit too late to influence those low grades throughout my school years, but at least I finally saw what *might* have worked.

  21. A good example of the virtues and limitations of using statistics to predict the future is playing out right now: the 2021 Major League Baseball season is drawing to a close as the League Championship Series begin.

    There are basic statistics for every MLB game going back to at least the 1890s (the National League was founded in 1876, resulting in six or seven Civil War veterans playing in the MLB, and the rules and statistics have been largely laid down since 1895) with even more detailed and elaborate records being produced and kept as time goes on. In addition, minor league, college and some high school baseball games are also fastidiously recorded. Innovations continue to happen — pitch counts and velocity became common in the 90s, the StatCast tool began measuring the velocity and launch angles of batted balls in the 2010s and very recently the RPM of pitches began being tracked. There are few things on the planet, much less sports, that have been as thoroughly recorded and analyzed as baseball. It’s not surprising that many of the top executives in the game have backgrounds in finance; the next generation will probably have PhDs in statistics.

    But, as the old cliche says, there’s a reason we play the games. Players who put up amazing numbers in the minor leagues often struggle in the bigs and players who put up amazing numbers during the regular season sometimes struggle in the post-season. Within a season, players can have ups and downs. It’s entirely possible to go 4-4 (four for four or four hits in four at-bats) with infield singles or lucky bounces a statistic like BABIP (batting average on balls in play) would say are almost certainly outs and 0-4 with four screaming line drives that should be hits. In 1978, the Boston Red Sox and New York Yankees finished tied at the end of the season, forcing a one game playoff to determine the division champion (who would automatically go to league championship series as the playoff format was different). With the Red Sox leading 2-0, Yankees shortstop Bucky Dent hit a three-run home run to give the Yankees the lead, which they never relinquished. Dent hit just five home runs all year and 40 in his entire career. Statistically, that home run was one of the most unlikely things to happen.

    At a more recent Red Sox-Yankees matchup, in the regular season, Boston third basemen Rafael Devers, who bats left-handed, hit a home run against left-handed Yankees pitcher Aroldis Chapman, who threw a 102 mph fastball. At that speed a batter has just 0.38 seconds to determine if and where they should swing. For a left-handed batter against a left-handed pitcher, it’s somewhat less because the ball is obscured by the pitcher’s body for much of their motion. According to Fangraphs, it was pretty much impossible:

    Statistics might tell you how likely a thing is, they don’t provide certainty.

    The other thing is that those stats don’t just go to feed just front offices or the industry of old players complaining about how much the game has changed. Players use them. Managers use them. They learn from them. Hitters have turned themselves into legitimate power threats by studying launch angles, pitchers transform themselves by improving their pitch RPM. Shifting used to be something you did when a big left handed slugger was at-bat, now there are defensive shifts for pretty much every batter.

    Can the course of a season be predicted? To an extent — many sports websites have pre-season predictions. Most predicted that the Yankees, Rays, Astros, Braves and Dodgers would make the post-season. Almost none predicted that the Red Sox and Giants would do so well, or that the Twins, Mets and Padres would do so poorly. Nobody thought the Yankees would almost fail to make the postseason or that the Dodgers would only get in by wild card. And baseball is much, much simpler than history.

    1. And who could’ve predicted that the final nail in the coffin of the Giants’ season would be driven by one of the worst check-swing missed calls of all time? (Not salty or anything, obviously.)

  22. It’s been observed before that Asimov’s original conception of psychohistory and the Seldon Plan in the first book can be interpreted an early precursor to the original Pentagon-birthed vision of the Internet — a set of algorithms that incorporate granular population-scale data from a broad cross-section of databases to predict and preempt social unrest, intended as a tool for top-down population control if not straight-up counterinsurgency — but the vision as laid out in the subsequent books (i.e. psychohistory doesn’t necessarily work as such, but the idea of the Seldon Plan is itself a tool in the direct, ongoing, and extensive manipulations of a powerful psychic cabal actively running society from behind the scenes) is possibly more telling about the modern Internet of commercial social media platforms run by like Google and Facebook, where we’re all supposed to be increasingly terrified about the specter of the Internet as a powerful toolkit warping our brains and society through disinformation or whatever, and tech companies don’t really seek to assuage this fear but instead actively encourage it, as part of their sales pitch to advertisers whose belief in the power of online manipulation is necessary to keep the ad revenue flowing.

  23. I wonder if cliodynamics is one of the possible sciences of the future. Let’s assume that for the next few hundred-thousand years, human civilization survives in a state somewhat like today, where information is easily accessible and pretty accurate. You could get accurate census and economic data for quite a long period and maybe you’d be able to make worthwhile predictions.

    Although, I guess, these predictions will be as valid as weather predictions are today (and probably worse). Fairly accurate in the very short-term, accurate general high-level predictions, but still quite unpredictable in every other case.

    1. The other direction is that psychohistory would be more like climate prediction or insurance companies, you predict big sweeping trends like “Eurasian people would dominate the world” or “adding energy will warm up the planet” but not minor details like whether they would come from Europe or China or exactly how climate changes in a small region.

  24. A robust historical refutation of cliodynamics. I’m also interested in how cliodynamics falls apart when encountering the interface between history and the natural sciences. Some immediate examples:

    Since 2000, we’ve had a handful of novel coronaviruses with pandemic potential. (SARS CoV-1 and SARS CoV-2, MERS, etc.) It happened to be the case that there’s an inverse correlation between mortality and concealed spread, and it furthermore happened that the breakout pandemic occurred in 2019 rather than 2003 or 2012. But surely, which coronavirus we get is almost purely random—and it’s not hard to imagine a breakout pandemic in a different year, or with a different mortality profile, would effect massive structural changes. And this is true of every plague throughout history! Cliodynamics can’t anticipate the relative mortality of novel pathogens, or their effect on the large structures of history. (c.f. THE YEARS OF RICE AND SALT for a fictional treatment.)

    Nor are the chaotic effects of plague restricted to the macro scale! Trump, Johnson, and Bolsonaro all caught COVID and recovered; there was no sense in which cliometrics could constrain this outcome. Even if you figure that all such leaders would “come out in the wash”, well—Hitler, Lenin, Stalin, Churchill, Chamberlain, FDR, Hirohito, Yamamoto, De Gaulle, Tito, etc. etc. all lived through the flu pandemic of 1919, many at a particularly susceptible age. It’s hard to imagine that ALL of them would come out in the wash. For that matter, who knows which of the millions of dead would otherwise have wound up with a name I’d casually list among that company.

    Speaking of the flu pandemic, Einstein, Curie, Bohr, Oppenheimer, etc. all lived through it as well. The precise state of world affairs at the moment of the development of atomic weapons likely had a crucial impact on history. Imagine a 19th century cliodynamicist trying to pinpoint the arrangement of states at the moment of the Trinity Test!

    Having unlocked physics, we can go nuts in sciences beyond biology. Imagine trying to predict the past few centuries of history of the Arabian Peninsula and Persian Gulf before the commercial exploitation of oil. Imagine trying to predict the next few centuries of the history of the globe without knowing how alternative energy and carbon capture will shake out!

    Even purely natural disasters get into the game. There’s some recent discussion of the possibility that thousands of years ago, an asteroid airburst destroyed a city in the Jordan Valley. Imagine the impact of a Tunguska-scale impact on Classical Rome, Victorian London, Cold War era DC or Moscow. When the Yellowstone Supervolcano goes off it will neatly tear the book of history in two, dividing it into “before” and “after”—would cliodynamics care to give us the year this will happen? And so on, and so on.

    Probably their are meaningful things that can be said about large-scale social dynamics and their consequences. These statements are no-doubt heavily conditional, come with big error bars, cannot be stacked, and rely, always rely, on the caveat of “all else equal”. But all else is never equal.

    1. oh, and: I am skeptical that this works even that well for beetles.

      one of the most interesting experiments in biology, in my opinion, is the study of the Malagas Island/Marcus Island seabeds, as summarized in “When Snails Attack”.

      In brief: two island seabeds, similarly situated and physically connected, were found to have completely different ecosystems: one barren and lobster-dominated; the other diverse. Researchers were puzzled, but when they moved some lobsters to the other island they found that the whelks there would attack the lobsters and eat them, by virtue of their superior numbers. In short, the natural system had at least two different stable states; for all we know, they were separated by nothing more than a contingency of history.

      Since then there’s been some attempt to attribute the distinction to some less-contingent force (e.g., the presence of a certain kind of algae), but to my mind for the scholar of biology, the lesson is almost identical. Otherwise indistinguishable environments can vary radically in their ecosystem: whether it’s a strict question of contingency or the hidden result of obscure deterministic factors matters little to the observers. Prediction is still overthrown.

    2. I’ve actually participated in excavation at that site (Tall el-Hammam) in Jordan. One of the very first finds there, even before excavation officially started, was a piece of pottery that had been melted on one side, just enough to start to drip over one edge. To do that would require temperatures of a few thousand degrees (Celsius or Fahrenheit, doesn’t matter at this scale), but only for a few seconds. (The shard was actually casually mistaken for a piece of Trinitite while being analyzed in a lab in the US; Trinitite is glass formed from the sand at the site of the first nuclear bomb test in Trinity, New Mexico.) Now there are two ways we know of that the observed melt pattern could happen: nuclear weapons (unlikely in the Middle Bronze Age), or an airburst impactor from outer space. Now that they’ve released the full report, including things like the sheer scale of the (nigh-instantaneous) destruction and finding micro-diamonds, another tell-tale feature of terrestrial impacts, I think we can safely rule out the former scenario.

  25. Re: Afghan auxiliaries, what do you think of the argument that US military (either deliberately or through force of habit) trained the Afghan National Army to be more or less totally reliant on the constant air support availability afforded by total US aerial supremacy, thereby leaving them essentially screwed not just on a broad “war and society” level but even on a narrowly “pure” tactical level the instant that air support started to be withdrawn?

    One could extend that question into a broader problematization of the US military’s own reliance on assumed air supremacy, thus rendering its effectiveness potentially more questionable than we might like to believe against any enemy with even a narrow range of remotely effective air defense options (from the obvious Great Power peer-conflict scenarios to even a regional-power conflict like Iran or Korea) but that seems like a much bigger bite to chew.

    1. The ANA routinely got defeated by the Taliban before the Americans left. I have also never read anything complementary about the fighting power of “our allies” written by low-ranking soldiers or journalists who spent much time in Afghanistan. Instead, there were tales of rampant corruption, desertion and cowardice, gross incompetence and a fondness for terrorizing the local population as well as raping little boys.

      Lack of airpower seems like an excuse. The real mystery is how the American leadership convinced itself the ANA was worth much.

      1. I think as historians dig into the records over the next 20 years or so you’ll find that they knew exactly how bad the ANA was. What they balanced it against was the financial and political (time) investment necessary to build something better.

        As I noted on another comment section for this blog, the US Army was given the manpower resources comparable to the NYPD in order to pacify a nation of 30 million people in a country the size of Texas. And, to do this without making any poticially embarissing diplomatic concessions.

        It was not a solveable problem on the resources allocated. Something that Joe Biden recognized as Vice President back in 2009 when the Obama Admin. was doing their strategic review.

      2. The analysis I find most convincing is a dysfunctional officer corps culture. A combination of strong up-or-out policies, huge rewards for making it to flag rank, and a culture that tends to punish “defeatism” or “not being a team player” ensures that junior officers don’t challenge the official line. So if the brass says the ANA is good, no one dares to point out that the emperor has no clothes.

        If someone did point that out, that might in turn force the US to figure out what Bret seeks, a way of building militaries that works in non-Western, non-industrialized countries.

    2. A sound point. But, no institution has thought more about how to achieve and maintain air supremacy then the US Military. From the development of Wild Weasel tactics, to the HARM anti-radiation missiles, and stealth tech. The US Gov. is always imagining and re-imagining how to deny and gain air suppirority.

      We probably saw the future of this at the openeing of the NATO intervention in Lybia when the USS Florida launched a barrage of cruise missles to functionally destroy all known Lybian air defense sites on the first night of the intervention. And the USS Florida is a kluge, a traditional Cold War Ohio Class ballistic missile submarine converted into a mass curise missle platform. I’m sure there is someone in the Navy right now working on purpose built subs that can launch near simulataneous 100 missle salvos for the purpose of suppressing coastal air defense and inland air defense.

      1. I do agree, I don’t think the US is likely to lose an air supremacy war anytime soon, and I suspect most of their potential enemies are thinking about ways to reduce the effects of said air supremacy rather than contesting it per se. (except perhaps, very much locally)

  26. A lot of this argument can be summarized in more mathematical terminology as:

    History does not satisfy the ergodic hypothesis.

    The statement “the differences between humans of a given grouping (social status, ethnic group, what have you) ‘come out in the wash’ at large scales with long time horizons” is almost the same as “all accessible microstates are equiprobable over a long period of time”. Large scale differences cannot be the result of small scale variation.

    The ergodic hypothesis is broken by fruit flies, permanent magnets, and many other things. It would not surprise anyone familiar with ergodic theory that history is not ergodic.

    More information:
    See especially the part starting with the ammonia molecule.

    1. The corollary to this is that I expect that the worst parts of data-driven political science to come from people about equal distant between math and history, rather than from the people with the strongest mathematical backgrounds.

      1. Yeah, I think this is the right lens. There IS a field of mathematics that fits well with predicting the course of human civilization. Unfortunately, it’s Chaos Theory! Some fun results of Chaos Theory:
        – No matter how detailed and accurate your simulation is, it will only match reality for a short time before errors exponentiate out of control. See “Butterfly Effect”
        – Causal attribution is dicey in the short term and nonsensical in the long term. Narratives are Just-So stories and can’t predict the future.
        – Something can look stable until some unobserved effect causes it to spontaneously dissolve. See “Bifurcation Diagram”

        1. Or as they put in Behavioral Studies: ‘Subjects under carefully controlled conditions behave as they darn well please!’

          1. Researcher:

            No, no, NO! Subject 24277 why are you interacting with those other humans?


            Because I can.


            This is going to take SO much more compute to model now that you’ve made it exponential in N.


            How dreadful. Oh the horror.


  27. There is in fact a robust debate in evolutionary biology about the importance of historical contingency. Famously, Stephen Jay Gould wondered what would happen if you „replayed the tape of life“. I haven’t read Turchin myself, but from what you describe I‘m rather surprised that he doesn’t acknowledge this question – he certainly ought to be aware of it.

    1. I mean, when Elon Musk comes out with a “new realization” about AI futures that were already not only thought of but also being explored and parodied in Aasimov’s time*, you lose hope that anyone pays attention to the past even in their own field.

      *And by Aasimov, for article pertinence

  28. Part of the joke for that XKCD is that the have been some examples of physicists radically remaking an ancient field using essentially this technique. The clearest example of this is physical metallurgy, although physical chemistry and semiconductor physics deserve mention too. These examples are all in the middle/late 1900s.

    There is a straightforward explanation for this. There are a lot of problems that require quantum mechanics to solve, but quantum mechanics is very difficult to figure out, for both mathematical and philosophical reasons. Once quantum mechanics was developed by physicists, physicists could use it to revolutionize many other fields. “It was very easy in those days for any second-rate physicist to do first-rate work.” – Dirac.

    Since then, the boundaries between fields have largely stabilized. There certainly is still good work to be done by bringing ideas from one field to another. But it is unlikely to have the same sort of impact that the first wave of quantum mechanics had.

    1. Most of the joke, though, is that physicists tend to do this without actually understanding the field itself first.

      Especially in light of the alt text: “If you need some help with the math, let me know, but that should be enough to get you started! Huh? No, I don’t need to read your thesis, I can imagine roughly what it says.”

      1. Yes. Most of the times a physicist tries this, it fails. It is a low probability, high reward approach.

    2. The impact of physicists in genetics has been huge, but not because of new discoveries in physics. The quantitative skills of physicists just happen to be highly useful there, unlike most old-fashioned biology.

      Since ambitious research efforts usually fail, all those physicists who embarrassed themselves might give a misleading impression when base rates are ignored. Had Alvarez’s asteroid hypothesis turned out to be wrong, paleontologists could still tell the story about an insufferably condescending physicist who was knew nothing about animals but imagined he could just tell them what killed the dinosaurs. But indeed he could. Nowadays, we don’t associate his discovery with an invading physicist.

    3. A physicist can get all the way to a Ph.D. without ever reading a book written by someone who was lying. Reading through lies, I’ve concluded, is the missing skill that made all my attempts to be the guy in the XKCD cartoon end in disaster.

    4. Most legendary figure of this character is, of course, John von Neumann. Slightly exaggerating, whenever he sneezed while contributing to Manhattan project, new revolutionary fields of study (computing! game theory!) sprang into being. Of course, many physicist may embarrass themselves thinking themself as the next von Neumann.

      John Maynard Smith also comes to mind; while famous fro contributions in biology and genetics, he was originally an aircraft engineer.

  29. The most interesting part of Asimov’s psychohistory is that it never actually abandons the key assumption that individual people are effectively gas molecules. The Second Foundation is a few hundred people, maybe generously a few thousand, and they are able to retain total unity and also control the entire galaxy to such an extent that the prearranged Seldon Plan recordings are never disrupted because some breakaway state on the other side of the Milky Way has aligned with a state within reach of the First Foundation and so has disrupted the balance of power.

    Or to put it another way, human agency only really matters in the Foundation books if you’re some kind of superhuman psychic (of which there are eventually multiple types). Otherwise your behavior is predictable enough for the Second Foundation to be able to have one person per planet doing some mind-control work and ensure that you follow the plan of the Seldon recordings.

    Of course, Asimov was writing it in dialogue with Gibbon, not even Carlyle, let alone Butterfield or the Beards.

    1. I don’t know if you ever read Foundation’s Edge, or consider it a “proper” foundation book, since the original trilogy was written in the 40s and early 50s, and Edge wasn’t published until 1981 I think. But he actually does move away from that in the much later sequel. Part of the plan was always centered around the Second Foundation, with its (assumed) monopoly on psychic powers, being able to push the much more physically powerful First Foundation in the direction they wanted indefinitely. Some 500 years into the plan, the First Foundation has been able to develop a shield that protects them from mentalics, operated even without really understanding the principles of psychic phenomena involved.

      That is treated as a Very Big Deal that can derail the whole plan, and it’s only because the plan gets derailed in a completely different fashion by other actors before the psychic shield does so that you don’t see the full use of it. And Seldon never conceived of something like that; having lived in a time of near technological stagnancy, he doesn’t fully appreciate how a technologically revolutionary society like the First Foundation can operate, and that misapprehension is baked into the calculation structure.

        1. I don’t really consider it that hot of a take. You do have significant stylistic differences and his approach to continuity over that time period is really rather slapdash.

  30. Coming in to criticize “psychohistory”-esque takes from the STEM side:

    Small variations and uncertainties “coming out in the wash” is not actually a feature of all physical systems, only of the relatively simple ones that humans could model with minimal computing power in the sci-fi Silver Age. Reduced uncertainty/complexity in the aggregate is destroyed by chaos (high sensitivity to some variables in very specific an regions) and non-locality (everything affects everything, therefore a larger instance of the same problem requires you to model *more* interactions rather than fewer).

    A great example is the Nice Model of the solar system’s evolution, and it’s half-dozen variations; simple compared to social systems, yet very much not consistent with the “everything-gets-simpler-at-scale” ideal of Asimov. Even in an arena of very well-understood basic mechanics, it can only give very rough impressions of the dynamics of the whole system.

  31. This is a great post, but you overstate your case. The kind of… instability of historical predication, at least as you are describing it, fits within the branch of mathematics known as chaos theory. That is; systems can exhibit a strong sensitivity to initial conditions. However, chaotic systems are, actually, in a sense, predictable. So your claim that “Such complex causation defies general laws…” simply does not follow.

    Just as background, I suggest you watch the youtube video on Newton’s Fractal by 3blue1brown (

    My case for how this would work in the realm of historical analysis goes back to your previous post on what the Romans thought of the republic’s collapse: the Romans tended to view the collapse through a lens of personalities behaving badly and ignoring republican ideals, which blinded them from the real structural factors that made the collapse likely.

    If I wanted to develop a mathematical theory of history, I would start in the same place. Like Newton’s Fractal, there are regions of calm predictability wherein the system behaves as expected. However, drift too closely to the “edge” – the point where the old structure forces are breaking down – and literally anything can happen. Then, at that critical point, personalities and/or freak events can be decisive and prediction is impossible, even for disconnected parts of the system.

    In other words, perhaps history can be predictable, as long as we make our predictions contingent. Or, perhaps not, but if not, the reasons would need to be deeper than simple complexity or instability, etc.

    1. “That is; systems can exhibit a strong sensitivity to initial conditions. However, chaotic systems are, actually, in a sense, predictable. So your claim that “Such complex causation defies general laws…” simply does not follow.”

      From my understanding, that’s what makes prediction hard.

  32. Thank you for linking us to your two articles Bret. It is very enjoyable to see your thoughts as a military historian appear in more prescriptive works.

    1. Also it was interesting to see references to Jugurtha appear in both works! I had never heard of him before and it was interesting to learn how he fought for and against Rome at different times.

  33. I think some of these criticisms of Turchin are fair, but some are off the mark, and as a whole the discussion mostly misses what Turchin is doing that is interesting and useful.

    Turchin is doing something more specific than just trying to turn history into a predictive science using data to generate laws. Turchin is trying to develop an evolutionary science of history. This means representing historical change using evolutionary concepts and methods. I personally think this is a great idea, and Turchin is providing a useful intellectual service by doing work in this area. But it is VERY hard to explain to people what an evolutionary science of history is in the sense that Turchin means it, not so much because the ideas are so very technical, but because there is a natural tendency to jump very quickly to false assumptions about what it must be (progress! stages of civilization! laws of development!). This is worsened by the fact that there is not intellectual consensus about a lot of the details of what evolutionary explanatory frameworks look like, so it is not like Turchin’s version of it is a widely accepted and easily generalizable approach.

    With regard to the specific criticisms Brett levels here, my broad take is that the first set about contingency and the representation of historical facts can be fairly rebutted by Turchin’s defenders (including me, I guess), but the second set about the quality of historical evidence represent real problems that Turchin has not always dealt with well, although the upshot of those problems is less heinous than Brett suggests.

    To rebut the first set of points: the exact same arguments can be used to establish that we could never have a data based, predictive science of biology. And yet we do. Yes, populations of living organisms are complex, yes founder effects and weird accidents and Genghis Khans cause various unpredictable wiggles and crashes and bizarre trajectories. But the entire apparatus of evolutionary biology is constructed to describe and predict exactly such systems. It turns out that there are a lot of interesting things we can learn by building mathematical models of these complex and to-some-degree-unpredictable systems, including predicting with surprising accuracy various patterns to be found in their development over time. Turchin sometimes says incautious things, but he is genuinely a trained and competent evolutionary biologist and his work does a skilful job of applying the techniques of the field to human history. It doesn’t and won’t get you psychohistory, but it could develop into a science (or, if you prefer, school of history based on methods adapted from evolutionary science), which would be very exciting and cool.

    That said, the devil is in the details, and while I am sympathetic to Turchin’s approach, a lot of his actual conclusions are very questionable, partly for the reasons Brett raises. The predictive quality of a model tends to be very dependent on the quality of the data that it is based on, and for a lot of historical periods various interesting data is lacking.

    But, evolutionary modelling is extremely flexible, and it does not always need to make use of data of the sort Brett discusses. Censuses and GDP figures are great if you can get them, but there is a lot that can be done with estimates or with qualitatively derived categories. Furthermore, as Turchin argues in a lot of places, a lot of the point of modelling isn’t so much to run the model and predict the future (despite what journalists writing about him always glom on to): a lot of the point of model building is conceptual clarification. Evolutionary thinking offers a particular perspective on history that can sharpen certain causal issues, and evolutionary ideas can be used as a basis for historical theses even if no quantitative model is ever developed.

    A good example of this last claim is Azar Gat, who Brett so highly (and rightly) praises: Gat’s approach to the history of warfare is explicitly evolutionary and his whole explanatory framework is based on evolutionary principles. In my view Turchin and Gat are quite close in terms of their theoretical orientation: the big difference is that Turchin is interested in mathematizing his evolutionary explanations via explicit models, whereas Gat uses evolutionary concepts only as part of a causal narrative.

    1. Hmm, this strikes me as what Brett meant by motte and Bailey arguments because if it is just this small, tiny claim, then what is the point of the exercise. What does it give us that historians and philosophers, indeed just armchair readers haven’t already? It’s rather like bitcoin enthusiasts (sorry just read the article in NYT) who are so certain it is just an amazing technology that will do everything including expand artist ownership blah-blah but when you really get them down to brass tacks what it means is that you can pay a lot of money for a photo of a tweet anyone who took a screenshot could and launder money more effectively. The point of putting math into history seems really to be because math is considered real, ‘hard’ thinking and history is considered soft and just reciting facts. Also if a computer does it, it must be objective.

      But we all know, or should all know, that this isn’t true. What we choose to input matters. How do we know that those broad data points really are the important stuff and not for example a common co-morbidity/ride-along of an entirely different data point nobody thought to enter but nobody did. We think they are correct because of assumptions we make about what must matter.

      And evolutionary biologists are not immune to this problem. They seem to be especially afflicted though when they transfer what they try to practices within their field and glom them unto others. The madness that is Evo-psych stands out strongly here.

      Math and computer modeling are very useful tools but they are not the only ones in our intellectual toolbox.

      1. I don’t think it is a motte-and-bailey switch to say ‘the approach is promising, even if the results are thus far of dubious value’. As for Turchin’s claims, the fact that he doesn’t really think he can predict that, say, Britain would dominate the seas in the 19th century, hardly leads to the conclusion that his real ideas are just itty bitty piffles. Turchin is plenty ambitious: I have doubts about most of his substantive claims. But the problem is about whether his specific claims are argued for sufficiently well, not about whether his general style of argumentation (relying on quantitative analysis and evolutionary models) could possibly work in the context of history. I personally, am very much in the ‘yes it absolutely can’ camp.

        Evolutionary approaches to history, in the mode that Turchin and Gat use, are about taking seriously the idea of cultural evolution: that elements of culture are passed from generation to generation and from community to community, spawn variants, and then succeed or fail partly through random chance but partly because they have advantages compared to alternatives in the environment. This is an approach that has pretty substantive impacts on how one understands change and continuity in history: at the same time elements of it are already present in a lot of very traditional historical writing. Suffice to say, I think its important and interesting enough that even when exponents of the approach have their flaws (as Turchin does), the approach itself has value.

        Also, Evolutionary psychology is playing a different game, since they are not at all interested in cultural evolution. They are telling just-so adaptationist stories, which is a pitfall that naive evolutionary approaches are prone to.

  34. For analyses of two papers which came out of the SESHAT research program, see Some Comments on Turner on Old World Iron and Be Careful with Rein Taagepera’s List of Largest Empires Ancient historians have embraced big data since Mommsen’s Corpus Inscriptionum Latinarum in the 19th century, as long as those data can survive the merciless painstaking critique which historians apply to anything brought forward as evidence. (Its a good sign that this project has a journal which publishes criticisms of the data underlying projects of this sort).

  35. I find it comical that you offer Gat’s work as an antidote to Turchin and Pinker. Most of your polemic against Turchin could be repurposed against Gat with only a quick find-and-replace. The misuse of terminology from hard science mixed with epistemic arrogance and projection onto the past from the future that is so annoying in “cliodynamics” is absolutely a feature of evolutionary psychology. Just like Turchin, Gat offers tidy just-so stories for every feature of society and grand predictions, he’s just a better liar. You accept it from Gat because his cynicism validates yours, and because as a former historian he knows how to talk to historians.

    Your observations about the sociological implications of conscription vs a volunteer military are well-supported (though lacking comparisons to empires not built on conscription, and empires that retained it), but mostly pointless, in my opinion. The reason states transition to a volunteer military is that conscription is morally and politically untenable unless you can point to a clear and present danger. When a state becomes an empire, it ceases to be able to do that. Calling for the reintroduction of conscription is about as realistic as calling for the abolition of the military, and the latter option is a lot cheaper.

    1. Yes, Brett’s total lack of epistemic humility leaps from the page. But, fairly typical of someone who has devoted a lot of time to a narrow field and hasn’t had the opportunity to expand their horizons yet.

    2. Why do you characterise Gat as an evolutionary psychologist? He doesn’t seem like one from my (admittedly not very deep) reading. There are a wide range of thinkers who make use of evolutionary ideas, and I can’t see much overlap between Gat and Evo-Psych. In particular, while he does discuss the evolutionary basis of human psychology a bit, his discussion is not reminiscent of evo-psych (e.g. no discussion of specialized mental modules). But the big difference is that evo-psych is not interested in and does not generally discuss cultural evolution, whereas cultural evolution is very salient in Gat’s views.

      Also, it seems weird to characterise his ideas as ‘lies’. Why dishonest, and not just mistaken?

  36. I read OP and found its critique highly agreeable, though I have no familiarity with the thing being critiqued.

    All persons who wish to propound a grand unifying story about history ought to be required to read Toynbee, and reflect on what a waste of time that was.

    With respect to science and predictions, “prediction” does not always mean what you might think it does. Sometimes, like with weather prediction, it really is about predicting the occurrence of future events. More often maybe it means being able to predict answers to questions you might not have thought to ask without the predictive theory. The theory that all biological diversity is derived from a common ancestor by descent with modification and natural selection is an example of this. Its predictions are not about what kinds of life forms will evolve in the future, but about what kinds of evidence we will find if we compare living things.

  37. It seems to me there may be some value in looking at (any) system from 40,000 ft so as to see general trends without distraction by the many details that dominate on the micro-scale, but do in fact largely wash out on the macro scale. Example: stellar dynamics have been well-understood for many decades by astrophysicists. We can understand the general life cycle of very distant stars. But we cannot understand or predict the day to day behavior of the sun. Scientists observe and document the day to day dynamics of the sun, but predicting that behavior: coronal mass ejections, solar flares, etc is beyond their capacity, and likely always will be.

    There is great value to observing and understanding the small scale phenomenon for their own sake. But that lens does not necessarily provide a general understanding of the large scale behavior.

  38. Sorry it took me so long . . . here’s my list of typos for you to consider:

    every person one earth -> on earth
    to forma predictable whole -> to form a
    factors with made the Steppe -> factors which made
    controlling the entire Euasian Steppe -> entire Eurasian
    after his empire fragmented seem to -> empire seems to
    still active, bu its great fame -> active, but
    a fairly in depth focus -> in-depth
    and consequently hideously expensive -> consequently are
    consulted for Better Angels -> Better Angels
    two books. my issue -> books. My

    You wrote: “to see that they don’t”
    What? Is that what you meant? I thought you were saying that they *do* in fact (sometimes)
    lead to enormous historical differences?

  39. The movie “Men In Black”, of all things, has a rather deep quote on this. “A person is smart. People are dumb, panicky dangerous animals.” And it’s pretty much true – one person, you can maybe research their desires and habits and predict them. Put that person in a group, and who knows where the hell things will go.

    For a rather sad example of how general societal conditions can be a poor guide, see relations between majorities and disliked minorities (Jews in the Russian Empire, freed slaves in the Redemption-era American South, Muslims in India, Ahmaddiya in Pakistan, etc.). It’s not uncommon for there to be a grudging live-and-let-live situation that may well be a boring peace if not upset, but can go very, very wrong if it is. Which provocations are ignored/settled and which ones lead to awful things? Dumb luck, pretty much. Are the local leaders responsible, apathetic, or actively seeking an excuse to start a pogrom? How about the government / military? Does anyone stop for a second and point out that the latest outrage is crazy and false? I read a book awhile back, “Blood At The Root”, about a Georgia county that essentially exiled all its black residents on the basis of completely BS rumors, while an adjoining Georgia county with an actually responsible sheriff (probably not a great person, it was the old South, but at least not completely unhinged) didn’t, despite many similar rumor-powered incidents and the same societal dynamic. Good luck accounting for that in a mathematical model.

    1. The wisdom of crowds say the opposite. So does the record of history. Some people are dumb and panicky. A crowd without a purpose moves in random directions. A crowd with a purpose is united to a degree only seen in the social insects. That purpose may be vile – a lynching, a genocide – or noble, or directed to some goal (say, raising a megalith or a skyscraper) but the crowd will shape itself into an organised whole. Warfare is simultaneously the most destructive and the most organised thing we do.

  40. Cliodynamics brings to mind the discussion of statistical urban planning in Jane Jacobs’ The Death and Life of Great American Cities. She posits three main phases of scientific discovery: simple direct relationships, like Newtonian mechanics or Maxwell’s equations; statistical relationships, like quantum mechanics; and complex interrelated systems, like genetics. Her main argument is that urban planning in her time tried to use the tools of statistical analysis when it should be treating cities as complex interrelated systems, and it sounds like you have a similar argument around history.

  41. Oy. The Better Angels book. I checked it out to see what all the fuss was about. Expected to broadly agree with its conclusions. Found such a mess of fallacies and special pleading that I gave up on the book after a couple of hundred pages. People in the past entertained themselves with violent narratives like Grimm, which shows they were violent; when we play much more graphically violent video games it’s proof that we’re better at compartmentalizing. It doesn’t make sense to say nukes caused the end of wars between major industrialized powers, because poison gas didn’t and they’re both WMDs. The French Revolution got derailed into tyranny and the American didn’t because the French listened to the wrong philosophers (not because the two wars were totally different types of thing which happened to share the name “revolution”). The silliness just never stopped.

    1. Yes, the French Revolution was totally different from the American because the former was about restructuring an entire society while the latter was about removing a layer of foreign control. Obviously restructuring society is the more difficult goal with a greater opportunity for things to go totally pearshaped. The same is true of the Russian revolution, and just about all transitions to a Marxist system. The fact that Marxism is by and large being applied to agrarian societies not the highly industrialized systems of the theory certainly didn’t help. But we’ve seen that Marxism does not work in reality. That it is in fact hugely destructive and invariably leads to poverty and dictatorial government.

  42. Long-time (sporadic) reader, first-time commenter.

    Briefly, my view is that physics envy is a real problem, but some of the broadest aims of cliodynamics may have value as long-term avenues of research, given enough time, interaction with established fields, and a greater diversity of research perspectives. I’ll draw a few analogies.

    Consider psychology. As an academic discipline, its roots are mired in the sort of intuitively appealing but untested or untestable ideas that Brett rightly criticises. Even today, there is a level of unevenness in its methodology that would be unacceptable in more mature areas of research. But it would also be wrong to say that it has not progressed in the intervening time, or that the field as a whole is entirely meaningless.

    Now, evolutionary biology. As with human history, we have only one grand real-world experiment to observe; our data, though much less lacking than the latter field, are still woefully incomplete. But we can clearly see numerous important interactions between contingencies (chance mutations, abiotic events, etc.) and real trends or “pseudo-laws” – pseudo by physics standards, that is. And, as with history, counterfactual thinking and thought experiments are an increasingly popular way of exploring proposed trends within contexts where real-world data are not available.

    Next I’d like to consider computational sociology and economics. These, too, are young fields, much younger than the others I’ve just mentioned. And they are correspondingly immature – lacking, for example, in comprehensive accounts of the limitations of their own methodologies. But even with those real and important limitations, they are still a way to obtain new information about some of the ways that certain principles can interact, so long as we remember to take all of those limitations into account.

    This need for counterbalancing within a field is one common benefit of “network effects”: the larger a field grows, and the more it interacts with adjacent fields, the more opportunities there are for ideas to be roundly critiqued, and in response to grow more nuanced or ultimately retreat to the fringes. This can be seen in action in every sufficiently diverse part of academia. But to become diverse, a field first needs the opportunity to grow and evolve.

    How to encourage new and risky ideas, and give them the soil they need to flourish (or perish), without going overboard and pouring money down the drain, is a perennial challenge in the world of research, made no easier by the limited availability of funding in general and the understandable risk-aversion of funding bodies. But it’s a conversation worth having.

  43. I feel that anyone inspired by physics to develop a Grand Unified Theory of History must be a bit optimistic, given that we don’t currently have a Grand Unified Theory of Physics.

    On the other hand, any discipline that presumes to be of any use to anyone is prepared to make some predictions. After all, if you offer advice about what or how to do anything, you are making a prediction about what actions are most likely to achieve the best results.

    Doesn’t matter if you are studying literary theory, history, weather forecasting or seismology; if you are offering advice, you are making at least an implicit prediction based on your subject. A seismologist, for example, might not have much idea about when The Big One will happen on the San Andreas fault, but he might be prepared to stick an upper limit on how bad it might be, which is probably more important information, for the people who write California building regulations.

    The question is: Can you tell how good the advice is?

    Weather forecasters, of course, advice people about the probably near-future weather all the time. They generally attach a certain probability to the likelihood of events within a certain range; for example, 70% chance more than X rain on date Y. And if, on 70% of occasions they say that, there is more than X rain on date Y, their advice is perfectly reliable.

    The Good Judgement Open is an attempt to spread this mechanism from near-future weather to near-future history.

    Works of history generally offer a causal model, in which A happened because of B and C, but it would be better if we could easily test these models, and see how well each one worked.

    1. I actually disagree with the idea that advice (or useful perspectives more broadly) always implicitly corresponds to clear predictions. Sometimes it’s more like saying “this is a good policy to have, in general” without being specific about how or why. I think of the “advice” evolution has given to organisms in this sense, and I think normative ethics can be similar.

      1. If something is a good policy to have in general, it should, on average, have better results than alternative policies. And that is a prediction, although it might not be an easily tested one.

        1. But who decides what “on average” or “better” mean? One depends on a choice of context, the other on a choice of values. To treat something as a prediction in the quantitative sense you seem to be using, we must make those choices. But advice / perspectives often point toward concrete policies without committing to particular choices.

          1. That is just a matter of defining what you mean by better. For instance, increased welfare benefits were sold in the sixties as decreasing dependency and poverty. So we know they didn’t work on their own terms.

            This is a separate issue from whether we want them to work

          2. If an advisor cannot agree with the advised about what “better” means for purposes of giving advice, they are in no position to presume to give advice.

          3. ad: what about when the advice concerns which values to pursue?

            Mary: I would say there are and have always been a range of narratives about the welfare state, and that those which come out overall in support are not so monolithic.

          4. “what about when the advice concerns which values to pursue?”

            Robin, if the advice has some justification, that justification can be tested, and the advisor implicitly predicts that it will pass those tests.

            If the advice has no justification, it is just a demand that the advised obey some of the advisors desires for no reason. In which case, it seems generous to call it advice.

          5. Again, what you’re saying depends on the justification being explicit. I’m not saying advice shouldn’t be justified, but that some justification can’t be cleanly spelled out. That shouldn’t be controversial; implicit learning is a well-studied phenomenon, many aspects of which can themselves be explicitly described and tested.

    2. We do test them – against the data. Which is expanded and refined all the time – by adding to the corpus of inscriptions, or catalogues of grave goods, or dna analyses, or archaeological surveys and site reports, or radiography of manuscripts or…You get the picture. Like any discipline worth its weight, historical narratives are constantly tested and re-evaluated.

        1. None in high-level approach (all valid disciplines aim at cumulative correction of error). The key differences are that we only get to do history once (no experiments), and the data is scatty and heterogeneous.

          1. I’ll grant you that is a difference between History and Physics. But it is not a difference between History and Cosmology. Or Geology. Or Palaeontology. So are we saying that the problem with cliodynamics is that it should be likened to cosmology, rather than physics or zoology?

Leave a Reply