On the Hydraulic Thesis
Jul 1, 2011

On the Hydraulic Thesis

The Hydraulic Thesis is related to the interplay between the rise of states in certain societies and the production of large scale irrigation and other 'hydraulic' systems that increase per area food production. Historically we see there is a loose temporal correlation between the rise of both elements of civilization. Depending on the various ways it is stated, there is an implied causation arrow whereby the increased food production afforded by hydraulic systems causes the increasing size and stratification of hierarchal social structures. Ultimately, the validity of the thesis rests upon the strength to which one asserts the causation.

In Jared Diamond's Guns, Germs and Steel, and in David Landes' The Wealth and Poverty of Nations, both men take opposing views on the Hydraulic Thesis (Diamond against, Landes for). Unfortunately, I find that both essentially straw-manned the opposing side and while perhaps appropriately criticizing the strawmans don't address more substantive criticisms. Without claiming any special knowledge of the subject beyond analyzing the statements in these two books, the logical conclusion to me appears to be somewhere in the middle where I believe a version of the thesis is true provided it is stated in relatively weak terms.

Landes was the more egregious of the two, despite me siding with his view more. He says that opponents of the thesis are "zealous in their political correctness" and are only "presumably" disagreeing with the thesis out of a sense that attributing non cultural factors to the rise of states like China's is in some way racist. While this may be true of the person he quotes, it is undeniably possible to object to the thesis on intellectual grounds - as Diamond does who is certainly not racist and clearly strives in his book to assert environmental not racist factors that a priori would be entirely in keeping with the hydraulic thesis were it not for Diamond simply thinking it is false. To ignore the tempered, intellectual arguments in favor of burning the arguments of an extremist strikes of intellectual dishonesty.

As for Diamond, he makes two major criticisms of the thesis which are both invalid in my view. Firstly, he notes that the theory doesn't explain the larger progression from bands to tribes to chiefdoms and finally to states. True, it doesn't; it provides an explanation for the last step of coalescing into states. However, unless one claims the thesis as the sole determinant of civilizations' complexity growth from the Neolithic onwards this criticism is irrelevant. I find it merely to be one important factor, in certain specific civilizations like China and Mesoamerica, relating to only the last step of state formation.

Secondly, Diamond essentially asserts that irrigation development could not have precipitated state formation because people had no previous knowledge of irrigation upon which to know that if they formed a state they would have the capacities for large scale hydraulic systems. Hence, it was states (formed due to other factors Diamond identifies such as population density) that formed first and hydraulic systems are a lagging consequence of that. Before addressing my criticism of this, let me introduce my view of the correct answer.

It is worth noting what both authors say in common; namely, existence of states without larger scale hydraulic systems for a long time in regions where it would have been beneficial, continued small scale and local hydraulic systems, and despite a loose temporal coincidence the occasional lag time between the two. It seems they both agree that a state - with its excess workers and centralized bureaucracy - needs to be in place for a major hydraulic system to occur. One causation arrow is established: the existence of states causes major hydraulic systems. It is the reverse arrow that is in question.

On an entirely different issue - technological development - Diamond introduces the concept of a random variable. He notes that given how Technological Historians have identified some fourteen factors that result in societies being innovative and that there is lots of variance among all these factors within any large region and long breadth of time that we can essentially treat technological development as a random variable. While perhaps not predictable as we may wish, we can say that over enough time and space there will be local extrema in the variability of technological development; i.e., it is random. This concept applies more generally. Namely, in any stage of social development in history while there may be some broad overarching ultimate causes (as Diamond attempts to discover with things like crops and animals available to be domesticated) there are always a large range in the proximate factors to development that can be treated as a random variable.

Let us consider this idea in terms of the coalescing into city-states in ancient China. We can expect that there will be local tribes and chiefdoms which, for a host of proximate factors, such as higher than normal fertility that we can simply attribute to being random, some will make marginal increases in social stratification and size of their society that others will not. Of those that do, perhaps most will move back down when the transient proximate factors that led to their rise remove themselves. But some will exist to the point of attempting or stumbling upon a marginal increase in hydraulic complexity that would not previously have been possible.

Like many other factors, hydraulic management is an autocatalytic process with regards to civilization development. That is, both factors are mutually reinforcing and provide causal feedback. As Diamond firmly establishes, increased food production is one of the major drivers of civilization. It is what allows primitive agrarian societies, for instance, to begin having specialized hierarchal roles like "chief", "priest", "bronze worker", or "warrior" for the surplus food to support these roles now exists and without it only largely egalitarian societies with all participants devoted to food production are possible. Increasing large scale hydraulic management in irrigation and the like increases the food productivity of the societies thus allowing more complex, larger and more stratified societies. Conversely, increasing size and centralized structure makes it possible to build these kinds of larger scale human works.

Together the two factors reinforce each other allowing increasingly complex societies with increasingly large scale hydraulic management. The autocatalytic process is seeded by the previously mentioned underlying random variability that will naturally create pockets whereby the process can begin and take off. Finally, societies exist in competition with each other and social Darwinism at the level of societies determines the final makeup of societies in a region. The societies that have extensive food production from hydraulic management, and the increased size, complexity and stratification that goes with it, will be the most prepared to out-fight, out-populate and out-compete their competitors leading to a preponderance of such societies in the future.

We have now developed the concepts to be able to state my position succinctly: the hydraulic thesis is an autocatalytic feedback process precipitated by transient variability and perpetuated by social Darwinism. Moreover, it is but one important factor among many others that manifest itself in varying relevance in different regions and times.

Let us return now to Diamond's second criticism about how the lack of model societies prevent hydraulic management preceding states. Firstly, as soon as such irrigation/states codeveloped in some places then there absolutely would be precisely such a model for neighbors to assimilate or be forced into irrelevance, just as the spread of any other technologies. There still remains some question of the initial cases but, this only remains a valid criticism if one takes a discontinuous view where either a state exists or it doesn't and irrigation either exists or it doesn't.  Instead, if one takes a more continuous view that smaller changes are going to occur due to variability and those changes reinforce each other then it is absolutely possible for both phenomena to grow in tandem with all kinds of proto-states and proto-irrigation intermediaries. It isn't an issue of asking which phenomena causes the other but taking the view that they both reinforce each other.

Diamond's goal in his book is to identify ultimate causes - as it turns out largely in terms of environmental factors relating to food production - that explain the significant differences among major continents and regions in their historical development and in that regard he did an excellent job. Just as the available cereal/pulse domesticatable crops turned out to be relevant distinctions, so too I think is the distinction discussed by Landes. Namely, the fact that crops like rice are particularly susceptible to increased productivity via increased labour in relatively high tech ways such as irrigation compared to crops like wheat which are less so is an important distinction that explains in part the differences between Asian and European historical development. The hydraulic thesis, in the form I have described, explains loosely the mechanism by which that works.

Thoughts on this post? Comment below!

Share this post:

Tweet It! Facebook Add Feed Reddit! Digg It! Stumble Delicious Follow

Post a Comment

Frequent Topics: