Devereux on Big History

General
, ,
1

This recently was posted on WhatsApp. Adding a thread here for deeper treatment.

2

Just wrote a long reply to Devereux and accidentally deleted most of it. Probably a message from the universe in there somewhere …

In that deleted post, I quoted Turchin’ preface in End Times (pp. ix - xii) about what cliodynamics really is. It’s basically Big Data. It’s also exactly what LLMs, do, BTW.

Here is what I said about all that in my forthcoming open-source textbook on IT:

“In considering all the possible information technology careers, “historian” is not likely to come immediately to mind! Yet Peter Turchin, one of the historians whose work is cited in this chapter (Turchin, 2016), has made data science central to the practice of historical analysis. Turchin and many other collaborators have established a database called “Seshat” (Turchin et al, 2020). (Seshat was the ancient Egyptian goddess of writing and measurement). The idea behind the Seshat database is to comb through historical archives all over the world to create a repository of numerical data for historical trend analysis. The chart in the image above illustrates this approach. In this chart, Turchin and his research team are measuring social stress levels in the United States decade by decade through quantitative indicators like income distribution, strikes, and protests. Based on such numerical trend analysis derived from historical data, Turchin has identified cyclical patterns in the history of many nations featuring growing inequality, rising conflict, crisis moments, and more evenly distributed income in the aftermath of the crisis. If artificial intelligence makes any significant contribution to the practice of historiography, it will only be because data gathering teams such as those contributing to Seshat have provided all this numerical data to analyze in the first place!”

3

Also, contra Devereux, I think conflating Peter Turchin and Steven Pinker is rubbish, and exactly the sort of big, sloppy, over-generalization that Devereux is at such pains to warn us against.

1 Like
4

I had to take another look at this and see if he had a point. Essentially he was just gate-keeping; what Turchin and Pinker have in common is, they are not historians.

5

How such gate-keeping is helpful for anything practical in current circumstances escapes me. The general trend of things seems to be the dissolution of academic silos and the diminishing sway of the tenured class.

More substantially, here is my take on the usefulness of historical studies for current matters, and a recommended approach to history across disciplines.

  1. future-oriented decision making requires world models with generalized frameworks. (See Vervaeke on recursive relevance realization). No such framework can be expected to capture all the details. However, the absence of ANY such framework makes decision making computationally intractable and subject to sheer randomness. Radical Humean skepticism about cause and effect suffers from similar drawbacks. If A → B is not at least likely on some probabilistic grounds, why do we ever choose the brake over the gas pedal or steer right as opposed to left?

  2. there is no guide to the future other than modelling based on experience. However, experience itself suggests that sheer repetition or simple trend projection is unlikely to transpire as expected. Heuristic pattern matching is about the best we can do. But imperfect as such pattern matching is, it generally outperforms sheer randomness.

  3. the most likely historical patterns are those best attested by empirical evidence. I would certainly include Turchin’s work in that well-attested class. He would be joined there by Luke Kemp, Ian Morris, Niall Ferguson, Ferndand Braudel (and the entire Annales school), Jared Diamond, and quite a few others. None of the abstract patterns teased out by these historically grounded, but none the less theoretically inclined thinkers, is a perfect fit for all historical data at all times and in all places. No serious historical scholar would support such totalizing pretensions. Again, the name of the game is suggestive heuristic fit and improvement over typical naive “lessons of history” analogizing.

I’m fine if people want to improve on Turchin by finding better data or specifying potentially more refined explanatory hypotheses. If Turchin’s generalizations are off target, please suggest better ones. Noting that everything is unique and unrepeatable offers no guidance at all however about how to plan a career, what sort of house to build, or how one might conduct one’s life in general, not to mention how one might address the great existential questions of our moment.

6

Turchin predicted that the height of political instability in the US would be around the year 2020 didn’t he? I guess it’s open to interpretation, but it looks worse now from my pov, and is likely to continue getting worse. I’m anticipating ground troops in Iran, leading to a draft, leading to War Powers Resolution enactment and the suspension of the Constitution, and likely a civil war. Of course I could easily be wrong – but if even half of this comes to pass, it would seem that the apex was not 2020, but rather the present or near future.

Did Turchin have specific predictions for 2025/2026? I think you stated at some point last year that 2025 was proving to be a pivotal year for the collapse of the current world system. (Sorry to put words in your mouth, please correct me if I’m wrong). The forecasts I know of that do line up most accurately with the level of tumult we see specifically this year and last are astrological. A great war for the US was predicted by astrologers before WWII occurred sometimes between 1941 - 1948 (Uranus in Gemini) – and again, another pivotal war occuring between 2025 and 2032 has long been predicted.

gemini_return
gemini_return1920×805 104 KB

7

Here is a Google summary (see below). To me, Turchin’s model is more like earthquake forecasting. Pressures build in predictable ways. How and where the fractures happen is far less precise. Turchin notes the US might have had something like Civil War in the 1930s, but it didn’t. (The pressures got vented in other ways). In any case, Turchin’s model is per-nation, and that is surely incomplete. Pogany’s global model, for example, is a necessary supplement.

Querty: Did Peter Turchin make specific predictions?

Yes, Peter Turchin made specific, quantitative predictions based on his field of “cliodynamics,” most notably forecasting in 2010 that the United States would experience a significant spike in political instability, social unrest, and violence beginning around 2020. His models pointed to a 10–20 year period of turmoil caused by elite overproduction, declining living standards, and state financial distress.

Key Specific Predictions and Findings:

  • The 2020 Prediction: In a 2010 Nature article, Turchin predicted that the 2010-2020 decade would see increasing instability in the US and Western Europe, specifically citing that the peak of this turmoil would start around 2020.

  • Quantitative Thresholds: In 2017, he explicitly updated his forecast to predict more than 100 “instability events” (such as riots or political violence) per 5-year period in the 2020s.

  • Driver Identification: He identified that high levels of “elite overproduction”—too many elites competing for too few positions—would cause intense infighting and weaken political systems.

  • Duration: Turchin has anticipated a period of, at minimum, 5 hellish years, and likely a decade or more of high-intensity conflict.

Turchin has noted that these trends are not inevitable fates, but rather trends that follow historical cycles (roughly 50-year periods of social pressure build-up). He argues that these periods of unrest, which he calls an “age of discord,” are driven by structural factors that, if understood, could be managed to avoid the worst outcomes.

8

Yes, this is what I thought. But doesn’t peak imply… peak? I.e. escalating in the 2010’s (as it did), then declining in the 2020s (as it isn’t)? As I was saying, this seems off, a bit. Still impressive, but not as impressive as predicting the onset and duration of all of America’s most conclusive wars, or predicting that this year, specifically, would be the most pivotal year in our history. You can find about a million videos going back years on YT predicting just that due to the truly unprecedented alignments in effect this year.

I still find Turchin’s work impressive, but I think just about anyone paying attention could have predicted more than 100 instability events per 5-year period in the 2020s.

One aside: you mentioned before that you ascribe this to correlation rather than causation. Me too - and I don’t know of any mundane astrologers this century who believe the planets are causing these developments in a mechanistic sense. It’s more like symbolic resonance. Here’s one possible interpretation: Planets don’t cause events. They index the simulation’s global state. Collective human behavior is an output of that state. Mundane astrology is an attempt to read the index.

9

I let Gemini do the heavy lifting, but my gut feeling about Pluto (or any non-casual grand cycle correlation) is this sort of thing is ripe for cherry-picked data and convenient over-fitting on literally cosmic scales.

Look, Turchin cycles don’t always pan out according the calendar. But I won’t be reaching for epicycles! Population growth, income concentration, and greedy nobles eating the peasants are real forces, but they are not the only forces. Throw in Black Swans like the Black Plague or Pearl Harbor and the dial gets reset! Also, Turchin takes note of, but does not model international forces. One time he noted the English cycle got delayed 50 years or so by collapse of France (giving English nobility a new feeding ground).

Let’s put it this way - for my current CS students - what’s a better roapmap - elite overproduction or the transit of Pluto? Elite overproduction seems both correlated AND casual to what is happening with CS entry level hiring just now. Moreover, viewing matters that way allows for constructive strategizing - namely, CS students need a counter-elite strategy, they need to “cheat” the prior game, they need to map out a new model of what “elite” might become in some new dispensation. A lot of my educational radicalism is precisely for that reason. I’m not waiting for the old job market to come back. It won’t.

Prompt: Criticize the correlation of the mundane astrology of Pluto with historical events.

Gemini said

To criticize the correlation of Pluto’s mundane astrology with historical events, one must move beyond the “cosmic synchronicity” narrative and look at the structural, logical, and historical flaws in the methodology.

While Pluto correlations are popular for their “grand scale,” they face significant skepticism from both historians and traditional astrologers.

1. The Confirmation Bias and “Texas Sharpshooter” Problem

The primary criticism is selective data mining. Pluto transits a sign for 12 to 30 years. In any three-decade window of human history, one can find “transformative” events (wars, revolutions, or collapses).

  • The Fallacy: If Pluto is in Capricorn (Sign of Government), and a government falls, it’s cited as proof. If a government falls while Pluto is in Sagittarius (Sign of Ideology), it is also cited as proof of “ideological transformation.”

  • The Criticism: The definitions of “Plutonian” events (death, rebirth, power, transformation) are so broad that they can be retroactively applied to almost any major historical shift.

2. The Discovery and “Retrofitting” Issue

Pluto was only discovered in 1930. This creates two distinct problems:

  • The Lack of Tradition: Unlike Mars or Saturn, which have 2,000+ years of recorded observations in mundane astrology, Pluto’s “meanings” were assigned during the rise of Fascism and the development of the Atomic Bomb. Critics argue these meanings are projections of the 1930s zeitgeist onto a rock, rather than inherent qualities of the planet.

  • Historical Retrofitting: To validate Pluto, astrologers look back at the 1700s or 1400s and apply these 20th-century meanings to events that people at the time didn’t associate with those themes.

3. The “Pluto Return” Skepticism

The “Pluto Return” (a ~248-year cycle) is often used to predict the collapse of the United States or the fall of Rome.

  • The Counter-Argument: Empires and nations are constantly evolving. The Roman Empire “fell” and was “reborn” dozens of times across 1,000 years. Picking the specific 248-year mark as the “true” collapse is often a result of adjusting historical dates to fit the astrology, rather than the other way around.

  • The England Example: England has survived multiple Pluto returns. Astrologers point to the “Elizabethan Golden Age” as a positive Pluto return, but critics note this is a “pivot” that conveniently explains away why the nation didn’t collapse, making the theory unfalsifiable.

4. Comparison: Mundane Astrology vs. Historical Complexity

Historians argue that attributing major shifts to a single planetary body ignores the complex, multi-causal nature of history.

Feature Astrological View (Pluto) Historical/Critical View
Causality Planetary cycles trigger collective shifts. Economic, environmental, and social pressures drive events.
Timeframes Specific 20-year “transits” define eras. Historical eras (e.g., The Industrial Revolution) span multiple transits.
Orb of Influence Events “count” if they happen near the aspect. “Orbs” are often stretched (up to 10 years) to force a correlation.

5. The Dwarf Planet Dilemma

Since Pluto’s reclassification as a dwarf planet in 2006, a rift has grown between “Modern” and “Traditional” astrologers:

  • Traditionalists argue that if Pluto is the “Lord of Transformation” because of its distance and size, then Eris (which is more massive) or Ceres should have equal or greater mundane impact.

  • The Critique: If astrology is based on the visible sky (Thema Mundi), including an invisible dwarf planet while excluding others is logically inconsistent.

“Pluto is often the ‘scapegoat’ of mundane astrology; whenever a historian can’t explain a sudden shift, an astrologer blames Pluto. It is the ‘God of the Gaps’ for the celestial world.”

Would you like me to compare Pluto’s historical correlations specifically against the cycles of other outer planets like Neptune or Uranus to see which holds more “weight”?

10

Given the way LLMs work - necessarily producing an output that fits with the input - the following isn’t a great way to arrive at an impartial analysis.

You can give that same prompt to criticize anything with a well-established correlation and still get a critique.

11

I’m just letting it fill in details. Putting “Rome” on a 248 year cycle, for example, is going to hit something more or less every 248 years. To say those are some sort of special “crisis” moments ignores lots of other things at other times that may be worthy of the “crisis” label just as much.

Without spending a lot of time lining up the details, the frequency of most things historical speeds up after the Industrial Revolution for obvious reasons. Pluto still orbits at the same rate. Likewise, Pluto had a 248 year orbit for the dinosaurs, during the Paleolithic, in early civilizations, etc. Did something magic engage lately to synch up Pluto with human affairs? There are dozens of cyclical history theories, but to me all these 80-year, 50-year, 30-year and other cycles are mostly a Rorschach test. If some gross pattern appears on a Gestalt level (like animal shapes in the clouds), then the research challenge is to dig into data and find some plausible mechanism “why”? Generational models have a sort of surface plausibility because repopulation, memory, learning, etc. have generational correlates. Orbits of this planet or that don’t point in any such casual direction.