Another interesting and helpful link via a post on The Monkey Cage (which has recently moved to the Washington Post, and unfortunately seems to have become much slower and less reliable as a result), this time to a new paper in Proceedings of the National Academy of Sciences of the USA that seeks to understand global historical development via a cultural evolutionary model. It’s a nice short paper (http://www.pnas.org/content/early/2013/09/20/1308825110.full.pdf+html), so you can and really should read it for yourselves, even if you skip the equations, but in brief the authors constructed a model of the Afroeurasian landmass as a grid of 100x100km squares:
Each grid cell is characterized by existence of agriculture, biome (e.g., desert), and elevation. At the beginning of the simulation, each agricultural square is inhabited by an independent polity, and the cells adjacent to the steppe are “seeded” with military technology (MilTech) traits, which gradually diffuse out to the rest of the landmass.
Once the model was set in motion (from a starting-point of 1500 BCE), each polity had a certain chance of conquering every other polity to build a larger polity, and each conquered polity had a certain chance of taking over the ‘cultural genome’ of its conqueror – i.e. adopting its traits of military technology. The result is a pattern of the diffusion of different forms of military technology and the expansion of complex multi-cell polities (empires) from three cores in the 1500-500 BCE period (Egypt, the Fertile Crescent and eastern China) to encompass most of Europe, the Middle East and China by the end of 1500 CE. This pattern was then compared with the ‘actual’ course of imperial expansion, put together from historical atlases, and the authors claim a 65% rate of prediction of which cells would have which level of ‘imperial density’ at a given date; a re-run of the model, omitting the military technology factor, had only a 16% success rate. It is these claims that have led the study to be reported in the non-academic press as ‘explaining’ history and ‘predicting’ the rise and fall of empires.It is these claims that Erik Voeten in his blog piece (http://www.washingtonpost.com/blogs/monkey-cage/wp/2013/09/25/no-math-cannot-predict-the-rise-and-fall-of-empires/) seeks to question, without a great deal of difficulty. In brief, he suggests that the argument verges on circularity:
[T]hey used information that we know historically contributed to the rise and fall of complex societies to model the evolution of complex societies. The inclusion of distance from the Steppe is especially problematic given that this is a geographic variable that is used to “predict” the geographic spread of something that we know spread around areas close to the Steppe.
In other words, it’s actually more interesting to wonder about the bits of the model that didn’t match the historical development – either because the model failed to predict the rise of a complex polity in one area, or it predicted one that never actually developed – than to imagine that presenting the importance of the geographical advantages of the Fertile Crescent in mathematical terms amounts to anything more concrete than a rhetorical technique. To be fair, the authors seem to be aware of this issue, and actually do have a more sophisticated idea of what they’re trying to do:
Social scientists have proposed a number of theories to explain the emergence of large-scale societies, emphasizing such factors as population growth, warfare, information management, economic specialization, and long-distance trade. However, because existing theories are usually formulated as verbal models, the causal mechanisms underlying these theories are not always made explicit.
The aim, then, is to differentiate between the different factors that have been suggested, and to reach a more precise idea of their interrelation and relative significance, by running different scenarios. None of this, as Voeten notes, puts them in position to predict anything, let alone future developments – and one suspects there’s a very good reason why the whole thing stops at 1500 rather than attempting to tackle modern developments. Actually one immediately thinks of ways in which taking things any later would instantly raise doubts about the model as a whole: how does a model apparently predicated on land-based movement and communication (on my reading, the sea seems to be treated as a barrier in the same way as mountains and deserts are) deal with the European expansion overseas, and how does a model predicated on the spread of military technology cope with the gunpowder question – the Chinese invent it but the Europeans employ it?
It’s difficult to escape the sense that the model is based on a very old-fashioned model of societal expansion – from a specialist perspective, the bibliography is equally striking for its omissions (Tainter on the collapse of complex societies might have offered some useful insights, for example, and a whole raftload of works on comparisons between Rome, China and the Mughal empire) and its inclusions (Toynbee’s Study of History, and Fukuyama). It is undoubtedly symptomatic that the authors think of themselves as engaging with a tradition of social-scientific research, not with historians (see the quote above). As regular readers know, I have a lot of sympathy with the social-scientific project to develop broader theories, rather than the historicist obsession with detail and specificity, but this does often come across as an example of excessive generalisation and simplification. The most obvious example is their fundamental assumption that “the force that favors the spread of ultrasocial traits is warfare” – no mention at all of the possibility that ‘culture’, in the broadest sense, might be spread by other means, which fits all too well with their apparent neglect of the sea as a source of connectivity (they do note that littoral cells can attack other littoral cells, but precisely because mounting such a sea-borne attack is a major enterprise and so limited in range, this must surely under-estimate the possibility of diffusion).
In brief, the model explores the possible dynamics of a taken-for-granted process of aggressive expansion, apparently without questioning whether non-aggressive cultural diffusion might also play a role. From a historical perspective, it’s a fascinating exercise, and food for thought, but the results are liable to seem to most pre-modern historians either rather banal or hopelessly flawed.