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:
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.
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.
- 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.
- 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.
- 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.