Friday, August 01, 2008

Cliodynamics, the rise & fall of empires and asabiya   posted by Razib @ 8/01/2008 04:37:00 PM

Yesterday assman recommended Peter Turchin's oeuvre as a nice theoretical overview of world history, in particular Historical Dynamics: Why States Rise and Fall. Unfortunately it was checked out at the library, so I've ordered it, but his more popularly oriented War and Peace and War: The Rise and Fall of Empires was available and I read it last night. With rather large font, copious endnotes, and a great deal of quotation of ancient and early modern historians and cultural observers it really doesn't measure up to 416 pages. It's a quick read. For those who are interested in world history and have some background knowledge I would recommend it, but its value is more in clarification than in originality (if you don't have much background knowledge, skip it, you won't know if Turchin is full of it or not, so why bother?). As a mathematical biologist by training Turchin's program seems to be how to translate verbal models into formal ones (he has Santa Fe Institute connections in case you're curious, so be warned or heartened, depending on where you stand). The reasoning behind this project is so obvious that I won't repeat or elucidate it. Overall his goal is to flesh out a discipline of cliodynamics which can complement cliometrics. There's a strong undercurrent within War and Peace and War that the reliance of cliometricians upon economic theory as a framework to make sense of empirical macrohistorical data needs to be complemented copiously by methodologies from a host of other disciplines.

If you read this blog regularly you'll know that I'm generally skeptical of the power of theory in history. Not only am I skeptical of theory in history, but I'm skeptical of the empirical data which has been collected in the historical disciplines, at least in terms of its operational utility in generating a model of the past as it was, and aiding us in projecting the future as it is likely to be. The rise of economic and social history within the past few decades points to the fact that traditional textual scholarship left something to be desired; in short, it looked through the glass with elite eyes. If you conceive of historical and social processes as purely a function of elite dynamics then so it goes, but if you reject that then you are missing out on much of the picture. You could attempt to understand Christianity as it was in the 4th century when the Roman Empire began sponsoring it as the imperial cult simply by reading the New Testament and the commentary of the early Church Fathers, but I don't think this would truly make things that comprehensible. Rather, Christianity was also social phenomenon of Jews and gentiles who lived and died from the 1st to the 3rd century, and these people had an important influence upon the practices and norms which were associated with believers in the Christian religion.

The muddiness of the glasses through which we view the past, and the power of distortion of one's own and societal biases, are critical to keep in mind whenever engaging in analyses of humane topics. Historical scholarship requires a recursive level of self-criticism and skepticism which is likely not at issue when it comes to statistical mechanics. The reality that skepticism is warranted has led some to abandon the idea of empirical truth altogether, contending that all past is "fiction." I won't engage with this viewpoint except to say that it's lazy. There are more things between heaven and earth then dreamed of in such a philosophy. Granted, the extreme skeptics may be right, true knowledge as we understand it through common sense may elude us or be a fantasy, but whether this is so is an open empirical question, not established truth. Nothing ventured, nothing gained.

Peter Turchin ventures a great deal in War and Peace and War. There are grand declarations in the prose about all the insight gained through the theoretical models which he's brought to bear on historical dynamics. To some extent it smells of Lakoffian hubris, but the difference is that Turchin is attempting to synthesize real formal tools, dynamical systems, with real data, economic, diplomatic and social history. In contrast, George Lakoff's conceptual metaphor theory isn't widely accepted, and his empirical data seems made up (his characterizations of the psychology of conservatives is based on a Berkeley liberal's preconceptions; not that there's anything wrong with that, but that's not ethnography). I bring this up because the style of exposition in War and Peace and War might seem alternatively bombastic and turgid, but there's some substance underneath it all. Keeping scratching.

As I said, the value-add here is clarity and precision. Turchin for example draws heavily on a conceptual framework created by the 14th century Arab intellectual Ibn Khaldun. The text is littered with references to Polybius and Alexis de Tocqueville (both of whom I think said a lot that could be said, especially Polybius). War and Peace and War doesn't present new hypotheses as much it injects them into quantitative analytic frameworks amenable to testing. I recall a conversation I once had with a phylogeneticist who explained that before the cladist revolution taxonomy was mostly a game of credential and age. Differences in opinion were resolved by the weight one would attribute to the "Cuz I said so!" contention of a Great Thinker (from Haeckel to Mayr). Whatever shortcomings there are to the hypothetico-deductive system promoted by the cladists, and despite their occasional veering into fanaticism, the reality remains that operationally their emphasis on a systematic framework as opposed to gestalt intuition revolutionized the field and allowed for a transparency which facilitated communication.

War and Peace and War is focused on the eternally recurrent cyclical social-historical dynamics. It isn't, for example, going to put the spotlight on the great divergence because at this point that is sui generis. Turchin outlines three cyclical dynamics which tie together the narrative of War and Peace and War:

1) Asabiya cycles
2) Secular cycles
3) Fathers-and-sons cycles

The general concepts aren't that obscure; asabiya is a term formulated by Ibn Khaldun to describe the rise and fall of polities in the Maghreb and more generally the Islamic world. Asabiya as defined by Turchin is roughly social cohesion. Finland would be a society with a great deal of asabiya, and Somalia would be one with very little. Northern Italy has it more than southern Italy. Japan has it more than China. Khaldun contended that nomads with asabiya invariably conquered farmers and city-dwellers who lacked it, but within a few generations the new conquerors would lose their asabiya and so fall prey to a new set of conquerors. Turchin expands this to a macrohistorical and imperial scale, while Khaldun's conception spans decades Turchin contends that the can dynamic span centuries (up to 1,000 years in fact in a cycle of decline).

The secular cycle is temporally shorter than the asabiya cycle, and is driven by social and demographic parameters. While asabiya may characterize the rise and fall of civilizations, secular cycles can easily be mapped onto the rise and fall of a particular polity or dynasty. Its time scale is on the order of a few hundred years. The endogenous parameters which drive the secular cycles are in Turchin's model demographic and economic; the surplus progeny of the elite, combined with the inevitability of inequality of material wealth across society. It is on this scale that the title of the book is most appropriate, as the shocks of societal collapse and anomie may allow the cycle to begin anew as the old order is shattered and the leaner and meaner new order picks up the pieces.

Finally, there is the shortest cycle, the fathers-and-sons. This is rather easy to comprehend because it is something we can viscerally comprehend. A generation which was preceded by peace may wish for some war simply because it does not comprehend the downside of aggression. In contrast, a generation which has experienced conflict may be very cautious in the future because of past experiences. This caution of course fosters peace in the present and may result in complacency in the youth who are not cognizant of the costs of disorders.

The asabiya theory is the most interesting from the point of view of the origin of empires and states: Turchin's claim is the genesis of new imperial elites is on the metaethnic frontier. These groups who face off against outgroups develop a great deal of internal cohesion, and they leverage this group level social capital into collective action. This is not a new idea, you will find an allusion to this in Guns, Germs and Steel. But War and Peace and War explores the hypothesis in detail through a systematic analysis of the cross-cultural data. I found it broadly persuasive. Here are some supports for the "frontier" theory of empire mentioned by Turchin, or that I can think of

1) Rome. The city of Rome is the borders of Latium, with the Etruscan territory to the north. Additionally, Turchin emphasizes the chasm which separated Mediterranean civilization from Gallic barbarism which dominated the Po river valley during the middle and late Republic. In the north central portion of the peninsula, the Roman Republican elite fixated on these Gauls as the Other, and their own identity emerged in part as a response to the Gaulish Sack of Rome.

2) Byzantium. Turchin argues that the origins of Byzantium can be found among the Illyrian emperors who flourished around 300. The early Byzantine elite were battle-hardened soldiers who faced the barbarians across the Danube.

3) China. The early Chinese dynasties had a tendency of being drawn from the semi-barbarian northwest fringe. In fact, the rivals to these powers were often semi-barbarian polities which bordered the Yangtze river valley! The states of the Chinese heartland on the other hand were generally conquered by these marginals. A traditional explanation is one of geographic determinism, but asabiya theory seems to suggest by facing the barbarians group cohesion is generated which is then easily used to conquered the civilized heartland.

4) The post-Roman states. I'm going to gloss over this quickly, but Turchin spends a great deal of time focusing on the fact the barbarian successor states which filled the vacuum after the fall of the West Roman Empire. By and large they were from the semi-Roman borderlands. The German tribes such as the Goths were from the frontiers of the Empire and had to develop their own identity and reorganize their polity to withstand the sallies of the Roman legions (the Goths were often federates of the Romans, as were the Franks). German groups who remained far behind the lines remained stateless, later to be conquered by the larger and more cohesive tribal confederations which arose along the Roman frontier.

5) The Islamic Empire. The Arabs arose along the borders of three major states, the Byzantines, the Persians and the Ethiopians. Some revisionist scholars such as Patricia Crone actually make arguments which dovetail even more perfectly with Turchin's thesis as to the origins of asabiya on the metaethnic frontier. For nearly 1,000 years the Arabs existed on the margins of civilized empires. Their identity as distinct from the sedentary Aramaic speakers of the Fertile Crescent no doubt emerged in large part because of their exclusion from participation in civilization.

6) The Persian Empires. The Achaemenids, Parthians and Sassanids all arose along the metaethnic frontier. The Parthians faced off against barbarian peoples to the north on the margins of northeast Iran. The Achaemenids and Sassanids in contrast came from the Fars, bordering the ancient civilized Semitic speaking lowlands.

7) The Mauryas. The first major Indian Empire was from the Indo-Aryan marchland of the eastern Gangetic plain, not from the heartland of Aryavarta (the Doab). The Mauryas in fact were likely Aryanized, as opposed to Aryan.

8) The Ottomans, who were the furthest northwest of the Turkic politices which rose up after the collapse of the Seljuks. The Ottomans were long known as ghazis who were both the frontline soldiers of Islam against the Christians, while at the same time being less authentically Turkic because of the assimilation of Christian elites into their power structure over time.

9) Russia. Turchen spends a great deal of time emphasizing the importance of the Cossack frontier in generating the Russian Empire. Today Moscow is in the middle of Russian Slavdom, but during the medieval period it was somewhat on the margins to the north and east, abutting up against the lands of the Finns to the north, and within reach of Turkic nomads just to the southeast.

10) The Carloginian Empire. The Franks came to power along the border of Roman and German cultures. One of the issues one confronts with the early Frankish kings was that they were often fluent in both German and Vulgar Latin. This emphasizes that their power base was neither in the Germanic forest to the north or the safely Roman south of Gaul.

Turchen's idea of the metaethnic frontier needs some elaboration. There are copious references to Clash of Civilizations and Bowling Alone: The Collapse and Revival of American Community. Asabiya doesn't emerge from a vacuum, social cohesion is a response to the Other. Let's focus on Byzantium and a few emperors to illustrate the sort of individuals who might be indicative of asabiya in the elite.

Justinian the Great was a man of modest origins from the Thracian hinterlands. He was termed the Last Roman because with him the dream of a unified Roman Empire died. He was the last East Roman/Byzantine Emperor for whom Latin was his native language, though he ruled a state which was dominated by Greek speakers. Heraclius, the early 7th century Emperor who defeated the Persians was from an Anatolian Armenian family and was exarch of Carthage. Leo the III who beat back the Arabs during the Second Siege of Constantinople was of Syrian Christian background. Basil II was from an ethnically Armenian family. And so on.

The Byzantine Empire was Greek speaking and Chalcedonian Christian. But many of its emperors were not from Greek backgrounds ethnically, and the Armenians were generally not originally from Chalcedonian Christian backgrounds (the Armenian national church is still not Chalcedonian today). But they identified far more strongly with Greek Christian civilization of Byzantium than they did with Arab Islam or Persian Zoroastrianism! One of Turchin's major points is that those on the metaethnic frontier might not be exemplars of the ethnic identity to which they are affiliated; rather, they are often quite marginal in the distribution of characteristics. But, they stand astride the region of greatest change, the fault-line across civilizations. Because of their position along high tension fault-lines asabiya emerges naturally because of functional necessity.

Turchin spends a chapter elaborating where he thinks asabiya comes from, and much of it is actually a dismissal of standard explanations of human prosocial tendencies. He does not believe that rational choice theory, kin selection or reciprocal altruism, scale appropriately to explain the dynamics operant along the metaethnic frontier. He appeals rather vaguely to multi-level selection driving the emergence of mixed behavioral strategies which utilize human intelligence to generate group cohesion and punish cheaters. I approach multi-level selection with a great deal of skepticism. That being said, I'm more skeptical of the ability of simple rational choice, kin selection or reciprocal altruism to scale well enough to explain the baroque social complexity which is characteristic of mass cultures.

The devil is in the details as they say, but it seems that in War and Peace and War the case is being made that excessively simplistic, universal and reductionistic explanations for human behavioral complexity can't be made to work. Evolutionary psychologists and economists often operate under the assumption of cognitive uniformity (or, at least did until recently). Turchin illustrates the likelihood of the importance of cognitive variation and personality differences via the ultimatum game. He argues that asabiya can emerge when a critical mass of moralists willing to sacrifice to punish free riders foster virtuous circles of altruism. He doesn't push this very far, but leaves it up to the reader to search for the literature. I personally think that there's something to this, though I think the argument at this point has more of a feel that x, y and z can't explain the complexity, so it must be a, b and c.

A background paradigm here is Robert Boyd and Peter Richerson's body of work on cultural evolution. Turchin suggests that the arguments over obscure religious theologies in early Christendom had less to do with details of theology as opposed to the importance of these as markers which divided groups. I agree with this. More trivially, but just as significantly, in
War and Peace and War there is reference to the particular hairstyle which one German confederation promoted to separate them from another group of tribes. There's a more recent example of this sort of thing resulting in ethnogenesis; the Zulu ethnicity arose in large part in the wake of the warlord Shaka's conquest of a collection of Bantu tribes (he was of the small Zulu tribe), and one of the ritual changes that he fostered at this point early in the 19th century was the abandonment of circumcision. To this day the Zulu in South Africa are uncircumcised, while their Xhosa neighbors are circumcised.

The emphasis that the author places upon homogeneity and group identifying markers is obviously uncomfortable for the dominant consensus in American academe. War and Peace and War seems laced with muted apologia for the rather negative implications of the importance of metaethnic frontiers in terms of a multicultural society which has no core central identity. In some ways this is an exact replica of Robert Putnam's discomfort with the finding that diversity tends to lead to reduced social trust. Turchin has a solution for this: a diverse group can coalesce around a common ideology against the Other.

This common ideology is generally religious. The Byzantine Empire crystallized around Orthodox Christianity, the Caliphate around Islam, the Gupta's revitalized Hinduism while the Maurya's patronized Buddhism, the Tang alternatively favored Taosim and Buddhism, etc. One could go on and on. This is, I think, a case where the solution is just as unpalatable as the problem from most academics. An ethnically diverse frontier, such as that of the Byzantine with the Arab Muslim world, can adhere together, but only with a well demarcated religious identity. Or, one might have a relatively religiously unenthusiastic society such as Japan or Scandinavia which nonetheless has a great deal of asabiya, but here one sees that there is another sort of homogeneity. Ultimately, it seems there's No Free Lunch. Secular ideologies such as Communism or Zionism likely can fit the bill, but again, for those who wish to promote value-neutral pluralism these systems of social organization have their downsides.

The other two dynamics are much simpler to understand. For secular cycles Turchin presents some simple phenomenon which occur in many societies

1) Elites overproduce and become top-heavy during times of plenty
2) Inequality in material wealth lead to resentment and social discord

The metastable situation will generally shift at some point and switch to rebellion and civil war. This tends to kill off much of the elite, and naturally redistribute wealth with the collapse of law and order. Obviously there's much more, but that's the gist of it. The empirical examples aren't too hard to dig up (this is where having a background of knowledge helps, if not, take the author's word for it).

As for the generational cycles; let's just say that I suspect my generation of Americans will be a little less enthusiastic about foreign adventurism after what happened in the last decade. I'm sure that 20 or 30 years from now we'll be cautioning the young ones who want to Do Something in the face of uncertainty, an we'll be dismissed idiotarians or some other neologism.

...I'll probably post on
Historical Dynamics in 2 weeks....