Saturday, October 12, 2013

Playing the cycles game

A flippant title for a serious subject. In the course of a thoughtful post about 'state collapse' viewed from a systemic perspective, Jay Ulfelder links to an article in Nature from last August about Peter Turchin's work on political instability, which Turchin argues goes in 50-year cycles. (The Nature piece footnotes a Journal of Peace Res. article by Turchin that looks at U.S. data from the late 18th century to the present. Presumably the idea is that this is a global phenomenon (extending how far back?), but he began with data on a single country.)

Though my inclination is to be skeptical, I have not even read the whole Nature piece, let alone the footnoted article. It struck me as interesting, however, that it was left to a commenter on the Nature article to mention Kondratieff cycles, which are posited 50-year swings in economic activity. Apparently the author of the piece didn't think the parallel was worth noting. And indeed, there are significant differences between (postulated) economic cycles and (postulated) political cycles. The former are somewhat less obviously and directly linked to or entangled with individual agency. (Not everyone will think that a valid point, of course.)

The notion of cycles of unrest/stability is not new; I believe that, w/r/t the U.S., Huntington's American Politics: The Promise of Disharmony made this kind of argument. Schlesinger's The Cycles of American History did too, though, I would guess, in a somewhat different way. 

But, to quote the Nature piece:
What is new about cliodynamics isn't the search for patterns, Turchin explains. Historians have done valuable work correlating phenomena such as political instability with political, economic and demographic variables. What is different is the scale — Turchin and his colleagues are systematically collecting historical data that span centuries or even millennia — and the mathematical analysis of how the variables interact.
Especially since I'm not going to understand the details of the mathematical analysis, who am I to say he (or they) shouldn't do this? Let a hundred flowers bloom, and it'll all come out in the wash. Or something like that. 

9 comments:

Kindred Winecoff said...

I'm a big fan of Turchin, although this isn't my favorite of his work. But you should definitely check out his book War and Peace and War and his book on American history (http://cliodynamics.info/PDF/SDAAS_Sep17.pdf). He had a very interesting article in PNAS recently: http://www.pnas.org/content/early/2013/09/20/1308825110.abstract. His blog -- Social Evolution Forum -- is consistently interesting.

LFC said...

Kindred,
Thanks for the tips. I think I was vaguely aware of him but haven't read his work.

LFC said...

P.s. I'll try to get over to Symposium to read your article there.

Peter T said...

I read War and Peace and War some time back, and Secular Cycles recently. I thought the latter the better book - the thesis more fully developed, and the history used more carefully. I like the approach, but I think the claim to quantitative rigour stretched - we lack the data to test the claims in this way, so what we end up with is good history if done well, and bad history if done poorly.

LFC said...

Peter T,
'War and Peace and War' is available at the library I have easiest access to, so maybe I will check it out ('Secular Cycles' I would have to expend a little bit more energy to get my hands on).

My (former) advisor's first book was called 'Long Cycles'; I've read some of it, a long time ago. My own grad-school work did not deal with cycles and was not quantitative. Kindred is someone you might have a discussion about Turchin with, but his visits to this blog are, I think, infrequent. (If you want his coordinates, send me an e-mail and I'll give them to you. A link to my e-mail is at the profile/"about me" page.)

Kindred Winecoff said...

I read the blog via RSS but only check comments here occasionally. I am easily reached on Twitter (@whinecough) or e-mail (kindredwinecoff@gmail.com).

LFC said...

Kindred,
Nice to hear you're reading the blog via RSS.

Speaking of cycles, I'm feeling in a bit of a down phase right now w/r/t blogging. I'll maybe put up one more post this month and then take a break.

Peter T said...

Secular Cycles is on Kindle.

More generally, my issue with this kind of thinking is that it overlooks that any reiterated multi-player interaction where the players have any considerable degree of freedom pretty quickly becomes an exercise in strategic creativity, evolving over time. Which is why, when they teach tactics, they give half a dozen general rules, and then lots of case studies showing how different trade-offs and opportunities worked out.

Another issue is boundaries - Turchin's father-son cycle works one way in practice when the sons reject the fathers' practices, but another when they imitate them. With the boundary set by changing social tolerances more than by any absolute standard.

The upshot is that, in practice, cycles are an occasionally useful rubric, not a predictor.

LFC said...

Peter T,

Appreciate the comments. Hard to engage here as I've not read Turchin (nor have a Kindle, as it happens). People can get attached to their schemes of explanation and claim too much for them, that's true as a general matter I think. (Cf. Toynbee as criticized by Pieter Geyl, for just one, rather old example.)

Btw what are you interested in reading about here? (Or what are your interests in general? I have some sense from yr comments at CT and here, but it's a little hazy.)

I ask b/c there's a small # of regular readers of this blog and they all seem to be interested in rather different things. Which is why I just generally post about whatever I want w/o trying to fit an audience. But if there are specific issues that you would like to see broached here, or want to guest-post about, let me know. (I'm not promising, but inquiring)