Thursday, April 12, 2012

The Art of Not Being Governed, III: Agricultural Technology makes state-resistant geographies?

Edit: Added art attribution. 

The argument here is that it is not particularly plausible that the Roman Empire in the west fell through either largescale invasion or endogenous economic or demographic decline. I expect military success to attend the armies of rich and populous nations, and such evidence as there is for Roman economic decline locates it in the second and third centuries, well before the state itself fell. As with so many proposed caused of the fall of the Roman Empire, cause comes far too early for effect.

On the contrary, trends cited as evidence of both feature prominently in James Scott's account of the "art of not being governed." Supposed barbarian invasions are better understood as ethnogenesis, an overall trend that nicely accounts for the loss of Roman family names within family descents, the rise of heresy, and the end of literacy in the secular Latin West. Deurbanisation and apparent site abandonment are similarly common symptoms of archaic (and more recent) state collapse. I'd even point to tax evasion, if I were not coming to that. According to James Scott, these indicate resistance to extractive state institutions. 

The problem is that Scott puts his model forward with some ideological naivete. For Scott, the story is of an underpopulated region, where there are always empty, or near-empty ecological niches into which to flee. It is not obvious that any such thing existed in Europe. In a Boserupian narrative, we salvage these state-free geographies with new technologies consequent to the very success of the Roman Empire in fostering economic growth. That is, we assume that the growth was unequally distributed, and that increasingly many western Romans saw an advantage in escaping the Roman social order. Better feudalism, they said, than Romanitas. 

It's not exactly a new argument. If I want to tell the story of the fall of Roman Empire in terms of disruptive technologies, I could just point to the stirrup and we'd be comfortably back in the 1950s, or the iron plougshare and take us back to the 1930s or earlier. These won't hold, but that doesn't mean that it is impossible to tell the story of pre-medieval Europe as a chapter in the history of technology.

So, yeah, I think that I have a skeleton of an account that I'm going to lay down, right after I expose some B.C. homeboys to my semi-worldwide audience of tens and tens of people. It's even vaguely relevant, although I mostly picked it for sentimental reasons that are pointless to explain.

 First, some housekeeping. Interest in the subject of the First Nations and Black DNA share of White Americans is great, the research effort is not so impressive. Per Wikipedia, this 2007 paper is still state of the art, but nowadays even dumb people like me can gin up a citation search and find later work in the same vein.

Next, we used the Native Americans, East Asians, Eurasians, and Sub-Sahara Africans from HGDP-CEPH as parental groups of the U.S. Americans (the genotype data of the 24 autosomal SNPs can be found in the Supp. Table S5) in a STRUCTURE analysis. Self-declared U.S. Europeans showed on average 93.2% of European ancestry (95% CI from 73.23% to 98.09%), self-declared U.S. Asians carried on average 89.5% of East Asian ancestry (95% CI from 37.43% to 97.46%), and self-declared U.S. Africans revealed on average 86.2 % Sub-Sahara African ancestry (95% CI from 47.82% to 98.5%)

It's interesting that by a naive reading of the data, which is not the only one possible, self-declared US East Asians would be more "Indian" than self-declared European Americans. I speculated about this possibility, obliquely, here. The virtual disappearance of the West Coast's Hawaiian minority is an interesting parallel.

So Europe's been populated by hominids for a long time, most of it labelled the Old Stone Age, or Paleolithic, by John Lubbock. There followed the Mesolithic and then Neolithic and then, according to Christian Jorgensen Thomsen, by the Bronze Age and Iron Age. The outline of the scheme is even older than the early nineteenth century Danish archaeologist, however, and it is today traditional to begin any account of prehistory with some handwringing over the appropriateness of the old, old scheme before settling in to use it. 

That is, we start with a history of technology, worry that it doesn't really apply, and give up; the evidence we have to hand is entirely technological. Turn the question around: why shouldn't history be a history of technology? There are two reasons. The first is the ensmartening thesis, the second that history is so long.

I'm being a little facetious in my first case. I think that I've earned the right to be so, because I have read a lot of arguments about how this or that crucial social reform or discovery that happened very recently first made it possible for human beings to innovate. I think that the claim is preposterous on its face, and I am only too aware of the way in which our great-grandparents' impulse to defend (and appropriate) the Reformation presses this argument forward. If we can't get past anti-Catholicism, the future bodes ill for dealing with more recent species of prejudice! haunts the English-speaking dream of the world. That said, the historian who asserts that we shouldn't talk about "technical progress" is right. There's way too much egg on too many faces at this point. Did the water mill replace the Antique slave state with the estate society of feudalism, and did the longbow instaurate bourgeois modernity? Maybe in the first case, no in the second; but the point is that making this argument when he did left Friedrich Engels' credibility the hostage of every generation of archaeologists to come after.

 The second is that history is so long. Why did the cavefolk of 73,000BC (perhaps the first 'modern' humans) invent pressure flaking rather than steam engines? The first only allowed them to carry toolblanks around that they could use to create precisely-formed, notched microtools in the very process of debitage; the second would have increased their per-head energy consumption! But we are on the wrong track when we suggest that the length of history invalidates history of technology approaches. On the contrary, they demand better history of technology!

Again, this is a long exercise in getting into Boserupp's tank. Growth promotes new technologies, which in turn allows for new growth, and new technologies. So why is history long? Because the process is incremental. What, in particular, retards the application of new technologies? The huge training burden that makes the transmission of skills so difficult. What solves this problem, and acts as a technological accelerator? Armies! The agonistic character of war frees up resources, places young men under compulsion, ensuring that the training actor gets a return on the investment of time, and, above all, gives young men (in particular) the thing they want and need most: erotic capital. Uniforms, competence, exercise under an adequate dietary regimen. Put it this way:

the artist intends to communicate the agonistic here.There are states and sides and ideologies and interests. The participants in this war are tools in the hands of state actors, implementing war. Certainly they are not  individuals. No human agendas to our policy making, please! But, but, butt ... (and pecs, too). You know, I think that the human agenda is going to win this one. War over; baby making time started! So the key issue of the Roman Empire in the west isn't this, but rather learning how to make aqueducts.

Okay then: the long Paleolithic was not without technological innovation. Pressure point flaking is pretty much essential if you're going to have bows and arrows, but, earlier, facilitated microlithic technologies that made mobile lifestyles more practical. An environmental pressure model is invoked to explain the innovations found at Blombos Cave, according to which the coast of Natal is a refuge from a Middle Paleolithic "Big Dry," but the refuge thesis doesn't rule out a local population intensification. Rapid movement up the Nile valley towards the rest of the world might make Wadi Kubbaniya the typesite for all out-of-Africa moderns. This would have been a no-brainer if Gordon Hillman's discovery of wheat and barley in a 25,000BC context had held up, but it appears that Kubbaniyites "only" ate large-seeded water margin plants, and not dry grass. This might suggest that they had boats, but, more generally, points to the possibilities of secondary resource utilisation at the water margin. For much of our Paleolithic past, humanity really does appear to have been mainly a big game hunter. Big game pays for itself across longer distances of transport, and is thus the most efficient resource for mobile hunting bands to focus on. The more complex technologies of Kubbaniya (harpoons, nets, grindstones) point to location-specific alternate modes of life in which a wider range of resources are routinely used.  Although Neanderthals probably did this too, so we are talking about a very early technological innovation here. 

David Lewis-Williams thinks that when this technology got to Europe, it was all very hard to learn and stressful and required ideological mobilisation by H. sapiens against their Neanderthal cousins, and he gives us an account of it here that is as cheerfully mad as it is rewarding to read. Be that as it may, the next technological stage in the history of Europe probably happened outside it, with the invention of agriculture in the Near East and its spread across the Balkans to the Atlantic littoral by any one of a number of contested mechanisms, culminating in a peak population in 5000BC, per this archaeologist who clearly has the booms and busts of the last generation on his mind, although he might not be wrong about a tremendous population crash to levels from which recovery was not made until after 3500BC, presumably because investors discovered, too late, that  agriculture had been oversold, and it took 1500 years for cave prices to recover.

Be that as it may, the 5000--3500BC period is the one in which the Neolithic gradually fades into the "Chalcolithic," so called because we start to find copper tools and artefacts, but aren't ready to declare the Bronze Age started until our cave mortgage is finally above water. Interesting fact: the earliest uses of copper tools are for stone toolmaking, including as punches for more accurate pressure flaking. We get a full-blown Bronze Age with the appearance of bronze in Greece.

 Again, we have a technological departure that cannot be isolated in its own silo. I've referenced before the possibility (likliehood) that bronze toolkits made chariot making practical. There's details to hash out with people who want to talk about an early "invention of the wheel" on the basis mainly of Soviet-era archaeology; I trust that stuff as far as I can throw it, but no doubt I'm an outsider bringing an insufficiently nuanced view of things to the table again. What's more important about who invented the wheel and when is posthole archaeology pointing to the rise of silled longhouses, the so-called three-aisle house. Bronze Age (and Neolithic) long houses might also have been "three aisled," and the site I'm going to direct you to discusses the Iron Age house, but that's not what's important right now. What is important is...

Oops, better back up here. Agriculture is not a unitary technology. It is the fitting of crops to climate and soil. The Atlantic littoral  is very wet, and, broadly speaking and leaving out Scandinavia and the Maghreb, divided into sandy lowland soils around the North and Baltic Sea, a zone of clay soils upriver; and lighter soils in the upland region extending from the northern foothills of the Alps across the Jura to the Vosges to form the Annales School's notorious "forest third" of France. The wetness makes for large areas of marshland in all three regions, especially the first two, but also promotes biomass accumulation.

We talk a great deal about soil fertility and its nutrient content, but we shouldn't lose sight of the fact that plants live on air and sunlight. Admittedly, they also need water and phosphorus for metabolising, and most plants need nitrogen, and these are drawn up through roots that, in turn, take them from the soil. One of the key tenets of the old "ensmartening" thesis was that it was discovered very, very recently that some plants, notably the legumes, fix atmospheric nitrogen and deposit it in the soil. Once this was finally learned, it became possible to rotate crops in the field, getting rid of the old fallows that depressed agricultural productivity and leading to an Agricultural Revolution!

In fact, legumes are part of the earliest agricultural practices, and crop rotation was discovered equally early. The problem has always been to find consumers. Grain is nice, while finding a consumer for the the volume of legumes produced in a close rotation is a dietary chore for the farmer and a drug on the market. Yes, legumes are protein rich; but so are dry grass grains. It has long been the practice of the poorest of the poor to eat mixtures of legumes and low grade grains as porridge, peasant bread, and even roasted mueslis, and these mixtures tended to be  intensively planted in garden plots as subsistence foods, while the fields were reserved for growing  grains in a more extensive way. This ancient tradition is still reflected in the breakfast and bread aisles of the supermarket, where these days old time peasant foods like muesli and multigrain sprouted breads have become the food of the rich. Bad years could, and did, lead to poverty. "Mixture stands" were hardy and weather-resistant, but scarcely invincible. But the more usual reason for famines, even in earliest times, tended to be misallocation of resources resulting from the failure of the market crop.  

Phosphate deficiencies are a more serious matter in farming. Ask any Australian --but don't ask the citizen of a littoral, because here, at least, phosphorus fertiliers have long been available in the form of fish byproducts. The real reason that some soils are "poor" is much more basic than the mysteries of chemical farming. It is that roots have a simple, mechanical problem when they try to penetrate down and hold the plant to the ground. Sandy soils provide no traction, and wick away moisture and trace nutrients; clay soils are too hard and impenetrable. Facing these problems, early farmers discovered complementary solutions. Sandy soils produce the greatest biomass when left undisturbed under a permanent layer of the sod that builds up under perennial grass meadows. People cannot, of course, eat perennial grasses. Fortunately, cows and sheep can, and cows and sheep are yummy! Well, I don't know about mutton, but....

As for clay soils, the roots will do fine if a little bulky matter is incorporated into them at ploughing season. The easiest such bulky matter is dropped by animals, and animals can pull ploughs! This means that you need pasture for your animals, but this can be found either under marshland, up in the downs (those crazy English with their reverse talk) --or just north of the claylands on the sandy soils. The "manure" dropped by livestock and incorporated into the soil provides the necessary spaces through which roots can penetrate, but also adds nitrates and ammonia to the soil, hopelessly complicating our discussion of soil improvement discussions by giving manure a double meaning as soil-bulking treatment and fertiliser. 

I hope that I've cleared that up, and you're welcome. How do I bring this back to the Bronze Age? Simple: the three-aisled longhouse is the first byrie, providing shelter for livestock in the winter in one wing, and for people in the other wing. Particularly well-made three-aisled houses (and here we are talking about further improvements in everyday carpentry practice, bringing us to the Iron Age) can even have lofts above the animals, letting them help heat the home in their own pungent way.

Before that could happen, though, we have to usher Andrew Sherratt's Secondary Products Revolution onto the state, and introduce an era in which livestock are used not just for their meat, but also for their dairy production, labour, and fibres. Clearly, this is going to be a huge change in the nature of European life; so will the rise of the horse as a security asset in the wake of the chariot. Here is an animal that is, at least compared to the sheep and the oxen, just not very good at eating pasture produce. On the other hand, it is faster on the draw than oxen. It's not an exaggeration to say that it is going to take thousands of years to work out the implications of this, because so much of that working out depends on the building up of a variety of quotidian skills of carriage and harnessmaking. 

Again, we want to bring on the Iron Age gradually, and with caveats, because ironmaking, and blacksmithing, are complicated technologies to learn. And, again, if Robert Drews is right, there is a specific and critical security technology that comes in a package with iron: cavalry. And the timing is right, too. As the great powers of the western Mediterranean ramp up their rivalries in the post-Alexandrian times, northern European mercenary cavalry begins to be available. At the same time, there appears across the "forest belt" and a little further north into the claylands, the first, incipient urbanisation of the Atlantic littoral, the oppidum movement, often associated with the rise of a unitary, ,continent-wide "Celtic" culture in a particularly gross form of archaeological essentialism. (My source for much of this.)

One of the most surprising things about the oppida and the Celts is that we have an account of them from Julius Caesar, and it seems to be wrong. Caesar repeatedly stresses that the Celts build oppidums and live in relatively sophisticated societies. On their border with the Germans, which runs along the Rhine, this changes. East of the Rhine, the Germans are swidden farmers who live a migratory life and live in small communities. In fact, however, the oppida extend east of the Rhine along as far as the Czech lands. The real boundary between oppidum and not isn`t cultural, but ecological. It is, roughly, the divide between clay and sand. But only roughly, because, I`m going to suggest, something else is going on.

Archaeologists of the Neolithic sometimes assert that virtually the whole of the old European forests were cleared at this time. The transition from Mesolithic to Neolithic is the crucial divide between two ways of life, and even today we are closer in lifestyle to our earliest Neolithic ancestors than to the Mesolithic hunter-gatherers who came just before. If you really bear down on this position, you even end up arguing that the peak population of `Gaul`must have been in the order of the peak population of Quatorzan France. (I think it`s in Braudel, Identity of France, but I`m not going to go check in case I`m wrong.) You can certainly end up arguing that the population of Roman Britain could have been as high as 7 million, which is the point where I started checking the footnotes in Heather`s latest, before giving up on the book a few pages later for other reasons. The problem for archaeologists is that it is hard to figure out what was growing on a particular site 5000 years ago, and the field systems they detect really do seem to operate in most of Europe -something that`s particularly easy to believe if there was an unsustainable agriculture boom that went bust in 5000BC.

Fortunately, the history of technology  has a (possibly) half-baked solution for you! It is the anthropogenic forest, and it also explains (as you knew it would) the geography of the oppidum. The claim is simple: the forest came back because it had been transformed by human intervention into a labour-efficient agricultural resource! Part of it was "outfield," used as pasture on which livestock could fatten and fill their colons. Part of it was coppiced to provide firewood and lumber, in particular to people without good woodworking tools; and part of it, and the most delicious part, was "mast" forest. Mast forest provides nuts for human consumption and food for wildlife, but it is most important as a habitat for  pigs. As livestock go, pigs don't seem particularly efficient, in that they don't have a secondary product, and compete with humans for food. (Cue crazy functionalist theorising!) However, all you need is plenty of salt, and the humble pig becomes a way of shifting high-quality* calorie availability from summer to winter. So guess when the salt mines of Europe first start operating, which culture they're associated with, and when the associated infrastructure of long-range salt transport, such as packhorses, seem to have come into use?

With nascent cities and long-range trade, both within Europe and with the Mediterranean, the oppida saw a rise in minting activity, with the obvious implication that bullion surplusses were building up in --oh, hey, why do I hear people singing Latin marching songs, and why do I see glints on the southern horizon?

So, anyway, the story I get from Edward Luttwak is that the Romans were super-smart strategical thinkers who came up with this brilliant idea to have an empire encompassing all the good bits of the world, with military frontiers with armies stationed along them to keep out the slavering barbarians who would otherwise invade in massive force. (Wait? What? How is this making "strategic" sense?) 

See, that's what you get for listening to a neo-con. The reality is much more simple. The last century of the Roman Republic was one giant banarama of civil wars. The only available conceptual model for understanding what was going on was that of the benevolent patron taking care of his clients and his extended household, which incidentally included lots of warriors who liked to hang out with him. Only the empire was too big, there were too many warriors, and eventually a general would put together an army, and off they went again. So when Octavian (you know; this guy) won the latest civil war in 31BC, he suddenly had 60 legions in his personal family. The ideal solution would have been to pay them all off; remember that among the concepts not available to the Romans was "professional army," after all. But even he didn't have that much money, and he ended up with 28 legions left over at the end of all the money. 

There's only one solution to imminent financial crisis: kick the can down the road and hope something happens. So the Senate gave Octavian "imperium" over the provinces that were "in revolt" against Rome, which made them liable to tribute exactions, and also places where Octavian could have authority over armies raised by the Roman state. Ta-da! A warehouse for his personal army. The troops were promised pensions at the end of 20 or 25 years of service, and meanwhile the tribute could keep them fed. Now he just needed to ensure dynastic continuity by giving his sons/adopted sons/whatever a chance to win military glory. And that is why, Augustus and his boys ended up in France, Germany, and Austria in 16/15BC at the head of 90,000 of the total of 150,000 men that comprised those 28 legions, looking for enemies worth punting. 

Ninety-thousand men; 90,000. And they were being bought off. That meant that they couldn't be shorted crap. And they were going to be used to buy dynastic legitimacy, so they would fight. And by fight, I mean blacksmithing and carpentry and masonry and all the other skillsets needed to keep a legion in action. That's not a big deal as long as you're talking about veterans of the civil war, here. But that's 15 years in the past now, and the legions still have to be up to strength. You are starting to need new legionaries, and more besides. Where are you going to get your new troops? I'm going to go out on a limb and say, from amongst the sons of the old ones.

But, of course, the legionaries aren't allow to marry. So obviously there aren't any sons, right?

What can I say? I come from an ironic generation. I'll pursue the point next time.

*I mean in terms of ease of  metabolic assimilation, of course. I'm certainly not just riding a played-out Internet meme! 


  1. The main thought this post induces is: clearly I need to learn more agronomy (if that's even the right term.) But then again, every time I do fieldwork I find myself thinking that. The "mast" bit reminds me of an old Kabyle saying along the lines of: Eat acorns in the mountains and be free, eat wheat on the plains and be oppressed.

    As far as the genes go though, I'm not seeing any support for your hypothesis in those papers, though neither of them is necessarily focused on the right subgroups to look at.

  2. "Following the algorithm depicted in Figure 1, we found 31 Amerindian/Asian (2.2%)
    and 13 African mtDNA lineages (0.9%) among the 1387 American Caucasian individuals catalogued in the FBI mtDNA population database."

    This is an mtDNA figure that only reproduces "admixture" in continuous maternal lineages. It is thus interesting to see that the more recent study using mtDNA and Y-chromosome studies (continuous male descent) shows a non-European residual of twice the level of the Goncalves study. That the Y and mtDNA signatures of "admixture" are equal reinforces Goncalves' point, that interracial relationships in North America were not sex-biased in the way that they were in Latin America.

    So what do these figures suggest about the extent of the "admixture" event? I assume that most of it took place before the spread of the birth certificate made dissimulation more difficult, that is, before the Civil War.

    After the Civil War, we get urbanisation and steam-powered immigration. The United States Census for 1910 reported a population of 92 million, of which 13 million were (almost exclusively) European immigrants. Between sex line attrition and this new population, the genetic signature of the American ethnogenesis is going to be a great deal more attenuated than the Latin American one.