There is a remarkably biased and deceptive piece in Foreign Policy with a critique of Cliodynamics, among other things.
The author writes, “Peter Turchin and his collaborators have championed a new approach in which history as a discipline will be replaced by cliodynamics”. This is an outrageous falsehood. The relationship between cliodynamics and history is a mutualistic symbiosis. I stress it every time I have an opportunity, for example, here:
https://peterturchin.com/cliodynamica/historians-and-historical-databases/
and here, again:
https://peterturchin.com/cliodynamica/the-mad-prophet-of-connecticut/
Next, commenting on the Seshat project investigating the relationship between moralizing gods and complex societies, the author states, “But under scrutiny, those patterns show themselves to often be just results of omissions and lacunas in the underlying databases.” This is a lie of omission. The author provides a link to the critique of our results, but conveniently omits mentioning our very robust rebuttal. In fact, our response involved an enormous amount of additional work. We have consulted with dozens more historians and scholars of religion and summarized their collective knowledge in an analytical narrative running at over 100k words. The resulting publication has been published as a preprint months ago:
https://osf.io/preprints/socarxiv/2v59j/
and is currently undergoing review in an academic journal. And what did it tell us? Far from weakening our main result, extensive buttressing of our data and improved analytical techniques have only strengthened it. We invited our critics to respond — so far they’ve declined to do so. But the author of this hit-piece has decided to ignore our response.
Finally, the author says, “History and historical data can still teach us so much if we take a guide with us on the way: a historian.” It may come as a surprise, but I am in complete agreement with this statement. But what the author (again, conveniently) fails to mention is that the Seshat project does precisely this. It would be impossible to build the Seshat Databank without historians and other experts on past societies. The author’s dismissal trivializes the enormous contributions to the Databank made by more than a hundred historians. You can see their contributions acknowledged here:
http://seshatdatabank.info/seshat-about-us/contributor-database/
Or you can simply look at the author list of Seshat publications, for example, this one:
Your work brings to the forefront the cyclical nature of societies; their waxing and waning. This challenges the current wisdom of the mob. Like Laocoon and Semmelweiss you will be attacked.
Please keep up your outstanding work. Perhaps, just ignore.
Not that you have a lot of time on your schedule to devote to such distractions still, it seems like “Foreign Policy’s” editors would be glad for a rebuttal piece, one that explains cliodynamics and Seshat (even if it is yet again), and moreover why “Foreign Policy’s” readers in particular should be interested in both.
Seems to me his issue is that he wants historians to stay the sole sages of the past.
Like wise mystical men who are the only ones capable of drawing any conclusions from their arcane sources
So history has always been subjective & has a bias based on whoever are the author(s) & for whom its target audience are intended for? Yet without it & those stories whether comfortable or not – what happened throughout man’s time on earth as a species , both domains of knowledge are symbiotic that is so obvious! What I would say there is history is underestimated with the impact & consequence on 21st century at our own peril that includes viewing cliodynamics & history are symbiotic but it is symptomatic & systemic over not having an integrated & holistic viewpoint needed in today’s world for everything we have to deal with, plan for & move forward?
The article, regardless of the conclusions was nicely written and is showing that like many studies, there are a myriad of opinions and a variety of possible conclusions. Ultimately I believe things stem to if one is trying to prove something is true/false or alternatively ego is trying to determine what kind of conclusions one can draw from the source material. Both have merit, although I find the former to be somewhat banal, however to each his or her own. Throwing stones can and do have consequences, so do forming conclusions based on subjective information. I suppose most of the folks who read this kind of material don’t need me to point out what is probably obvious, but I thought it worth taking a stab at it anyway.
You -PETER- should be honored to be part of the conversation. Someone casting aspersions on a project may draw others, who were unaware of the situation further into the discussion that may improve the overall results. That is the goal after all isn’t it ?
Personally I was a bit concerned with the suggestion that conservatives are “using” the data to “astroturf” historical context. In the end the specific systemic approach will either hold merit or it won’t. Anything else is just opinion.
Thank you for sharing the link to the article. I’m just a humble BA living on the Pacific coast of Japan trying to find his way in the world. Information like this is not only informative, but is quite exciting as well as concerning. Its the grist of life and I welcome the continued debate.
Your response is remarkably restrained considering the ideological agenda which permeated the article. Quite apart from the ‘normal’ academic territoriality – which sees a massive threat lurking in the data-driven mathematics – and the rhetorical tricks the author deploys throughout, there is the constant implicit suggestion that both you and Henrich have an ideological agenda. He doesn’t spell it outright but there is an unmistakable suggestion that your work sets out to ‘exonerate’ the sins (unspecified but we know what they are) of the West. If challenged on that Fafinski will claim that even if not intended it is the inevitable consequence of your methodology and naivete. Of course there are specific issues he raises which do merit a careful reply but they are drowned out by the bad faith throughout the article..
@Mike Berger
Yes it seems like an attack on the scientific method itself which is concerning to say the least
@johne
It really should interest foreign policy readers because the theory of cultural evolution and long term historical analysis could add so much to foreign policiy decisions, it should be at the forefront to make better decisions regarding other countries and cultures.
It’s worth so much more than the countless ideological opinion pieces which just act like echo chambers
Very briefly, the attack on the scientific method which you refer to in your reply, is latent in many prestigious supposed scientific platforms at the present time. It reflects an inherent tension between current ideological narratives sweeping the West and the traditional praxis of science. The big challenge, as you acknowledge in some more of your more recent articles, is to widen the ambit of causative factors beyond the ‘macro’ factors of the demographic-structural model.to include narratives/scripts and psycho-social dynamics which also impact on social tensions and divisions. A huge challenge!
That guy is living in the past. He reminds me of those naysayers for the electric vehicles.
History is a form of big data. Clyodynamics is the action of searching for either a pattern or a latent equation in the big data. It used to be a difficult task, but right now anybody can attempt its own set of equations and put it on github. I had it as an hobby for a while. I have been modelling with julia language. By the time that guy gets it , modelling historical patterns will be a training exercise for machine learning students.
The author is a recently minted 2018 PhD still trapped in fellowship and postdoctoral limbo according to his cv. Odds are he’s just trying to get his name out there in the historian establishment. Don’t get distracted.
The data quality issues raised by Fafinski are real ones which need to be understood. However without data and analysis, history is just a series of anecdotes. We always need to keep in mind that correlations can be the result of biased data availability, biased data interpretation or be spurious.
A few years ago I was working with the Maddison database (of historical GDP’s and populations). I was using it determine the economic capacities of countries during WW2. Even though this period is very well documented, I found significant errors in the data, from obvious typos, to abrupt changes in GDP and unexplained differences between similar countries. The lesson is to always be careful with data even when vetted by experts. And when the data confirms a hypothesis, always ask yourself it there another explanation possible.
I recently read your book, War and Peace and War. An interesting book and I generally agree with its conclusions. And yet there are historical errors (for example, the Ostrogoths were the German tribe farthest from the Romans, not a frontier people). There are also problems with ignoring negative examples, such the Kingdom of Sicily, a “meta-ethnic frontier” state by any definition, did not create empire. But a few outliers does not disprove a hypothesis, any more than a warm day in January disproves Winter.
@johne Most of Fafinski criticisms are aimed at the WEIRD book — I’ll let Joe Henrich respond to that. And Foreign Policy is not a good venue for a substantive discussion of issues, anyway. But I am thinking about organizing a discussion on a platform that would promote a good discussion that would move the field forward.
@Jacob: yes, this article has all the earmarks of defending the turf against interlopers
@Mike Berger: yes, as Walter Scheidel said on Twitter, “Tired standard tropes: 1) history is “irreducibly complex,” literally true but wrongly taken to mean that it can’t actually be explained, and 2) in certain cases, explanation is wrongly taken (as by Fafinski) to “exonerate modernity,” insinuating that explanation is immoral.” In other words, (2) is mixing up “what is” with “what should be”
@Craig Brackbill: Elite overproduction 🙂
@Roger Cooper: There are errors. But what Fafinski doesn’t understand (or doesn’t care about) is the cumulative nature of science, in which constructive critique leads to better data and methods.
Maddison database is flawed in many ways, but its dats have been refined and improved by the ClioInfra project. It’s a work in progress, and data are getting better.
As to your specific comments, Ostrogths — as Goths — may have started in Gotland far away from the Romans, but then they traveled to the Roman frontier. The Ostrogoths coalesced around the Amal family in Pannonia, right next to the Roman Empire.
It seems to me it will be important to include ideological elements into any future iterations of state fragility/resilience theory. I hope that some of the issues initiated by this hostile article, including the criticisms directed at Henrich et al’s work, will appear somewhere.i draw attention to the book Blind Trust by Vamik Volkan. It’s Freudian in orientation but very interesting in this context.
Foreign Policy has been practicing all the little deceptions and cheats Peter documents in how his work was treated. The difference being that Foreign Policy’s deceptions and cheats were about things like Bay of Pigs, Tonkin Gulf, Weapons of Mass Destruction and murder resulted [the civilians in Vietnam and Iraq were murdered, not ‘killed in war’ as the war was neither just nor lawful.Foreign Policy is, in its natural state, a full-on bull-horned cheerleader for illegal wars.
I have just reread the book and enjoyed it more than the first time. I find history an interesting subject but the historians tend to present the stories as they seem to prefer reality or perhaps to distance from competition. In any event, I view most histories with considerable skepticism and enjoyment (sometimes) is confined to finding obvious twisting.
Cliodynamics is clearly a better approach because it imposes constraints on its conclusions. A little analogous to operational vs. descriptive explanations. I would like a much deeper examination of the economics/industry of societies and individuals maybe in the appendices. There must be a great deal of commercial writing which can be mined.
A small nit about the book which I did not notice the first time is a reference to Democrats as ‘us’ or words to that effect. Gratuitous lapse in objectivity.
it was 118f in Siberia this week.
it was 116f in Las Vegas this week.
lol to eternity…..
In the Age of Discord you must be ready for such. People are less willing to consider things that go against their beliefs. As you know, less willing to compromise. Of course, such arguments of emotion tend to be weak as seen here, but no less persistent.
Ah, but there’s ‘history’ which shows the progressive march towards a grand Utopia, and ordinary history which explains what happened. Perhaps some people see Cliodynamics as contrary to their ‘progressive history’ narrative?
It’s a sad line of attack, also typical of the kind of groupthink you now seen in academia, the kind of behavioural pattern also evident in the circulation of petitions, statements etc. Of course one can see the advantage for the author of the FP piece in this regard.
The other side of the coin is that the humanities are under constant attack and are justifiably concerned with keeping the specialised knowledge required for studying ancient sources intact and vibrant. In that sense I do have some sympathy for someone who works with primary sources, including digital humanities techniques, he is far from a luddite.
It seems to me that the author of this blog has done more than most to move Cliodynamics into more rigorous territory. In particular I am thinking of the article ‘Fitting dynamic regression models to Seshat data’. That really represents a step forward from earlier work like that of Peter Pregrine, which is also fine but more limited.
That is, once I looked to the Seshat database entries in more detail I saw some strange things and inconsistencies. For Byzantine Crete “Rulers are gods” is inferred to be present because of someone’s (frankly convoluted) reasoning, but for medieval Paris basin this is inferred absent because of monotheism. You are in need of an editor!
Your books WPW and Ultrasociety are in my opinion masterpieces. Very much looking forward to another book for the “layman”.
I feel that in time these books will become much more widely acknowledged and appreciated.
Much respect.
@Marcus:
First, thank you for pointing this problematic data code. Indeed, it doesn’t make sense and we are going to change it.
Second, on a more general methodological note, we have plenty of editors and are constantly revising data values. In many cases data are changed as a result of others alerting us to errors (as in this case). Keep in mind that there are nearly 50,000 coded variable values in the Equinox 2020 data release. Naturally, there are errors that affect some small proportion of codes. An additional problem is that this variable (“rulers are gods”) haven’t yet been subjected to a Data Review Board treatment, so I’d expect that the proportion of errors would probably be higher than average.
Once again, thanks for pointing out this error!
Prof Turchin, what do you think of the Changing World Order by Ray Dalio?
https://www.principles.com/the-changing-world-order/
As you are, it seems like he’s trying to apply past historical data to determine what’ll happen in the near future, and chart out cycles in rise and fall of societies. Compared with your data, his data is only for the 1000 years or so rather than for the full span of human history. However, his data includes a lot more financial data, which your data doesn’t seem to cover as much. Curious as to your take.
War and Peace and War makes an interesting claim that empires tend to appear on metaethnic frontiers. The British empire came up in discussion, and it seemed unclear that it fit that mold. Perhaps it is an exception?
@Phil: I don’t want to write a negative review, but his work does not reach the level of scientific rigor I require for cliodynamics.
@J. Daniel: The British Empire is post-1500. It (and the Spanish Empire) fit the mold, but the mold changes around 1500, from steppe frontiers to gunpowder and sailing ship frontiers. I have plans to write Ultrasociety II that would talk about it, but don’t know when I’ll get around to doing it.
@Peter Turchin
So you are not writing Ultrasociety II right now but I read in your other blog post you are writing a new book .
What’s the new book about ?
so youre surprised (and indignant!) that the same oligo-neoliberal kakistocracy that regularly uses your work and your (farsical) arguments (i.e. elite overproduction) to attack and put a death knell into the heart of the humanities, in order to stifle creative thinking, in order to produce technocrats and engineers for their ultra-society, would then turn around and put your cliometrics or cliodynamics (a distinction without a difference) in its sights???
you like it when they pat you on the back and seem to want you as part of their club, but not so much when ypu realize they just want you to park their car.
these are the same socio-psychopaths who know we are past the climate event horizon but are hell bent on not giving 2 cents (literally) to try to mitigate the worst effects.
bloomberg and FP are cut of the same oligo-neoliberal kakistocracy, you werent complaining a couple of months ago about bloomberg when they were used your work to bash the humanities but now youre crying a river; perhaps the lady doth protest too much.
not that itll make 2 bits of difference when 5 gigatons of methane are released in the next couple of years.
then we can all sit around a wetbulb camp fire and argue all day.
you cant talk about population dynamics and in fact, human history without including sudden atmospheric weather effects (on the order of decades not hundreds or thousands of years even during the goldylocks holocene).
there is now plenty of literature on sudden climate change.
for example ENSO (el nino la nina) effects on the americas.
nowhere in your models or books do you try to capture this specific type of effect.
this seems to be a fatal gap in your work.
dont have to look far to find new research that directly refutes ‘elite overproduction’ idea.
here in a brand new preprint rossier & ellersgaard et al looked at 105 years of swiss elite and elite networks data.
one of their main conclusions is that during times of crisis i.e. ww2 2008 elite circles expand to include professionals and highly educated elites and consolidate precrisis.
this is further evidence which refutes the notion that elite overproduction as measured by the number of elite educated professionals i.e. attorneys phds postdocs fellows mbas etc leads to social instabillity.
From Integrated to Fragmented Elites. The Core of Swiss Elite Networks 1910-2015
https://osf.io/preprints/socarxiv/4y8dh/
“[We] project that future ecological disruption due to climate change during the 21st century will be abrupt [and] forecasts suggest abrupt ecological disruption beginning before 2030.”
how does the above fit into the cliodynamical elite overprodxn hypothesis?
The Projected Timing of Abrupt Ecological Disruption from Climate Change (2020)
https://www.nature.com/articles/s41586-020-2189-9
honest question: is china in ‘elite overproduction’?
what are the quantifiable and qualitative thresholds of being in or out of ‘elite overproduction’? as in, before and after.
i am trying to fit china into your narrative and models with my clearly limited understanding of your work.
china has over 500k phds. the chinese communist party has close to 1 million party members.
thats 1 million party members overseeing the lives of 1.4 billion peoples in multiple latitudes and times zones.
how does china fit into your model?
SURELY CHINA must be in ‘elite overproduction’???
or rather if we ditch your model and add elite dominance ala gilens & page with a dose of technic-capitalism (google dragonfly, nike is for china and the chinese, tesla massively expanding in china only to be forced to recall every car sold in china),
we get a better reality matching curve.