Wednesday, July 20, 2011

The Horse, the Wheel and Language

Archaeologist David W. Anthony’s The Horse, the Wheel and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World is a book about how archaeology and linguistics, working together, can reveal much about our past as well as a major contribution to the long-running scholarly (and not-so-scholarly) debate about the Indo-Europeans.

That Anthony—who clearly loves his discipline; his book conveys in loving (but accessible and engaging) detail how archaeology seeks to unveil and understand the past—holds that linguistics can tell us useful details about the past apparently puts him in something of a minority among archaeologists.

A name used to describe the original Indo-Europeans is Aryan, a term taken from the Rig Veda and used in the conflation of language group with racial group:
In the 1780s J. G. Herder proposed a theory later developed by von Humboldt and elaborated in the twentieth century by Wittgenstein, that language creates the categories and distinctions which humans give meaning to the world. Each particular language, therefore, generates and is enmeshed in a closed social community, or “folk” that, at its core, is meaningless to the outsider. Language was seen by Herder and von Humboldt as a vessel that molded community and national identities (p.8).
This leads naturally to a notion of cultural “authenticity”. Apply Darwinian notions of selection to the notion that language was central to national identity and a dangerous, indeed clearly potential lethal, reified conflation of language with race was set off, which included a search for the original homeland of the “Aryans” (Pp8ff).

As with many noxious ideas, the problem was that there is something to this. The evidence is that language does create “cognitive maps” which encourage enduring cultural traits (Pp19ff). The trouble comes when one conflates language with race or ethnicity (as linguist Max Muller put it in 1872, languages do not have skin colour or skull types) and ignoring, downplaying or denying the powerful commonalities of human existence: particularly if one starts assigning superiority to one race, ethnicity, culture or language over another. As Anthony writes:
The mistakes that led an obscure linguistic mystery to erupt into racial genocide were distressingly simple and therefore can be avoided by anyone who cares to avoid them. They were the equation of race with language, and the assignment of superiority to some language-and-race groups (p.10).
Muller was not the only prominent linguist who protested against this noxious nonsense:
While Martin Heidegger argued that some languages—German and Greek—were unique vessels for a superior kind of thought, the linguistic anthropologist Franz Boas protested that no language could be superior to any other on the basis of objective criteria (p.10).
The Rig Veda itself treated being Aryan as a linguistic-religious category, not a ethnic descent one. Anthony rejects any notion of an analytically useful connection between language and race (p.11).

So, if the focus is language, what can linguistics—the study of language—tell us about the past?

Anthony sets out with admirable clarity the ways linguists classify and group languages, including how languages change over time. Which permits “looking back” at the history of language according to standard patterns of how languages change over time (Pp11ff). The trick is connecting that to archaeological (i.e. material culture) evidence. However:
Where we see a very clear material-culture frontier … that persists for centuries or millennia, it tends to be also a linguistic frontier … such ethno-linguistic frontiers seem to occur rarely. But where a robust material-culture frontier does persist for hundreds, even thousands of years, language tends to be correlated with it (p.17).
So, some linguistic patterns can be connected to the archaeological evidence.

Anthony further identifies migration as something weakly understood in contemporary archaeology; an over-reaction to previous simple equating of (for example) changes in pot manufacture with changes in people. As migration is hugely important in understanding the path of human cultures, Anthony applies modern migration theory to the evidence. He also attempts to incorporate Soviet and post-Soviet archaeological literature in his analysis. (Pp17-8).

Then there is the difficult issue of horses. Where were they first domesticated? Chariots were the favoured weapons of rulers from Greece to China from about 1700BC to 700BC, before being replaced by “disciplined troops of mounted archers, the first cavalry” from around 800BC. Anthony argues that he and his research partner and wife Dorcas Brown have established where horses were first domesticated, and that horse-riding preceded chariots even though chariots preceded cavalry (Pp18-9).

The chapter How to reconstruct a dead language looks at “working backwards” from current languages to their predecessors (Pp21ff). Which raises the question of how long do languages last. Including notions of core vocabulary, how long resistant they are to change? The evidence of linguistic construction suggests that the original Indo-European language probably existed around 3000BC and lasted no longer than 2,000 years (Pp39ff). From the evidence of daughter languages, Indo-European probably stopped being a living language no later than 2500BC. Part of the evidence for this being the Avesta and the Gathas of Zoroaster who, on the evidence of where he named, seems to have lived in eastern Iran around 1200-1000BC, developing his moral and religious perspectives partly in reaction against the warrior-and-blood-sacrifice vision of the Rig Veda (p.51).

Working through the logic of the daughter languages and evidence on when various languages “split off” from their Indo-European “mother tongue” (Pp52ff), Anthony develops a helpful diagrammatic “language tree” that suggests that proto-Indo-European probably existed around 4500BC while Indo-Iranian and Balto-Slavic were the last “split offs” (p.57).

The language of wool and wheel dates proto-Indo-European to being spoken after 4000-3500BC, that is after sheep were domesticated, and wool textiles first used, and after wheeled vehicles developed (Pp59ff). The evidence on sheep and wool is limited, that on wheeled vehicles much clearer:
We can say with great confidence that wheeled vehicles were not developed until after 4000BC; the surviving evidence suggests a date closer to 3500BC. Before 4000BC there were no wheels or wagons to talk about (p.63).
By about 3400BC, there is widespread evidence for wheeled vehicles, the oldest well-dated image of a wheeled vehicle being from Bronocice in southern Poland (Pp66-7).

The wheel changed basic dynamics of human existence. Previously, barges, rafts or hauling things along with large groups were required to haul heavy or bulky items such as:
harvested grain crops, hay crops, manure for fertilizer, firewood, building lumber, clay for pottery making, hides and leather and people (p.72).
Living in large group villages was sensible just for being able to move things. Wheeled vehicles changed possibilities:
Although the earliest wagons were slow and clumsy, and probably required teams of specially trained oxen, they permitted single families to carry manure out to the fields and to bring firewood, supplies, crops, and people back home. This reduced the need for cooperative manual labor and made single-family farms viable. Perhaps wagons contributed to the disappearance of large nucleated villages and the dispersal of many farming populations across the European landscape after about 3500 BC (p.72).
The effect on the steppes was possibly even more profound:
Here wagons made portable things that had never been portable in bulk—shelter, water and food (p.73).
Herders could now move out into the great plains, permitting dispersal of communities:
across interior steppers that earlier had been almost useless economically. Significant wealth and power could be extracted from larger herds spread over larger pastures (p.73)
Bundling together the wheel, plough, wool sheep, dairying and the beginning of horse transport becomes the “secondary product revolution” which swept across European societies around 3500-300OBC:
“Secondary products” are items like wool, milk and muscular power than can be harvested continuously from an animal without killing it, in contract to “primary products” such as meat, blood, bone and hides (p.73).
We have no idea where wheel-and-axle technology originally developed, only that it spread rapidly. Mesolithic and Neolithic sleds were a possible prototype—and remained the better form of transport in snow and ice of Winter in Eastern Europe up to the C20th (p.74). Wagons and wheeled axes do, however, help date later Proto-Indo-European as probably being spoken after 3500BC (p.75).
Since pioneer farmers have been established as a major vector for language spread, the “first-farming then language dispersal” theory suggested an Anatolian origin for Indo-European, since that is from whence farming spread across Europe (p.75). There are, however, numerous problems with such an origin for Indo-European. Having, however, fixed a reasonable time-frame for when Indo-European was spoken, that provides a basis for working out where (Pp76ff).

But the notion of an Indo-European “homeland” has various problems, which Anthony outlines and considers, arguing for a Indo-European homeland in the Caspian steppes from the Dnieper to Urals and down to the Caucasus in 3500-3000BC, taking the reader carefully through the evidence and reasoning involved (Pp83ff). Including questions of frontiers:
Language is strongly associated with persistent material-culture frontiers that are defined by bundles of opposed customs, what I will call robust frontiers (p.105).
Anthony illustrates the concept with British examples—such as the Welsh frontier, which has persisted since the C6th. (That is, 1500 years and counting.) So:
Persistent, robust premodern ethnolinguistic frontiers seem to have survived for long periods under one or both of two conditions: at large-scale ecotones (forest/steppe, desert/savannah, mountain/river bottom, mountain/coast) and at places where long-distance migrants stopped migrating and formed a cultural frontier (England/Wales, Britanny/France, German Swiss/German French) (p.108).
Part of what is going on here is that geography constrains and enables possibilities at a given level of technology. The patterns of life for living in forests or desert are not the same as for steppes or savannah. This is an obvious basis for persistent cultural (and thus linguistic) difference.

While historians accept migration as a cause of persistent cultural frontiers, based on ample evidence, contemporary archaeology is much less keen, reacting against previous glib association of people movement with shifts in technology and aesthetics. Anthony examines ethno-linguistic frontiers within the US as evidence for the existence and persistence of such things. This leads into a discussion of the patterns and implications of migrations (Pp108ff).

Anthony points out that “ecotones coincide with ethno-linguistic frontiers at many places” (p.114). He cites the work of linguists Daniel Nettle (Oxford) and Jane Hill (Arizona) that the geography of language reflects an underlying ecology of social relationships:
Social ties require a lot of effort to establish and maintain … People who are self-sufficient and fairly sure of their economic future tend to maintain strong social ties with a small number of people, usually very much like themselves. …
But people who are moderately uncertain of their economic future, who live in less productive territories and have to rely on multiple sources of income … maintain numerous weak ties with a wider variety of people. They often learn two or more languages or dialects, because they need a wider network to feel secure. They pick up new linguistic habits rapidly; they are innovators (Pp115-6).
The former is a localist strategy, and fits with the patterns of many university educated people in the modern West. (There are some fascinating resonances with work on the differences between liberals and conservatives in the US.)

The upshot is that:
Where an ecological frontier separated a predictable and productive environment from one that was unpredictable and unproductive, societies could not be organised the same way on both sides. Localized language and small language territories were found among settled farmers in ecologically productive territories. More variable languages, fuzzy dialect boundaries, and larger language territories appeared among mobile hunter-gatherers and pastoralists occupying territory where farming was difficult or impossible. In the Eurasian steppes the ecological frontier between the steppe (unproductive, unpredictable, occupied principally by hunters or herders) and the neighbouring agricultural lands (extremely productive and reliable, occupied by rich farmers) was a linguistic frontier through recorded history. Its persistence was one of the guiding factors in the history of China at one end of the steppes and of eastern Europe at the other (p.116).
To see the power of Anthony’s point, consider how both China and Eastern Europe have strong patterns of centralised, autocratic rule in contrast to the more dispersed political structures of Western Europe and Japan.

As for the spread of language, that need not follow folk migrations, it can be a matter of small-scale, elite migrations. So
Out-migrating Indo-European chiefs probably carried with them an ideology of political clientage … becoming patrons of their new clients among the local population; and they introduced a new ritual system in which they, in imitation of the gods, provided the animals for public sacrifices and feasts, and were in turn rewards with the recitation of praise culture—all solidly reconstructed for Proto-Indo-European, and all effective public recruiting strategies. Later Proto-Indo-European migrations also introduced a new, mobile kind of pastoral economy made possible by the combination of ox-drawn wagons and horseback riding. Expansion beyond a few islands of authority might have waited until the new chiefdoms successfully responded to external stresses, climatic or political. Then the original chiefly core became the foundation for the development of a new regional ethnic identity (p.118).
This is a pattern which seems to have occurred in Mitanni, for example, a case Anthony had discussed earlier (Pp49-50).

There is a further factor:
Chronic tribal warfare might generally favour pastoral over sedentary economies as herds can be identified by moving them, whereas agricultural fields are an immobile target (p.119).
Which, of course, encourages centralised rule to defend and expand farming lands, using numbers assembled by concentrated organisational power to counterbalance mobility, as dicussed by Christopher I. Beckwith in his Empires of the Silk Road.

But before any such pattern can emerge, a herding economy has to be established, probably around 5200-5000BC across the river valleys from the Dnieper River and then across the Pontic-Caspian steppes:
a revolutionary event that transformed not just the economy but also the rituals and politics of steppe societies (p.119).
Having set the analytical scene in Part One Language and Archaeology, it is one to Part Two The Opening of the Eurasian Steppe.

Geography also constrains and enables technological possibilities. An isolated group has to invent or make everything itself. A group in contact with others can trade resources and acquire ideas and techniques. A larger group has more potential discoverers and internal trade possibilities than a small group. A socially complex group has more problems to solve, and possibilities to explore, the more socially differentiated it is. Literacy makes it easier for knowledge to be retained and diffused across time and space. Political stability encourages trade (internal and external), social differentiation and literacy. Given all this, the dominance of the Fertile Crescent in invention prior to c.500BC, and the dominance of China in invention from then until about 1500AD, is not surprising. Except, of course, for anything to do with horses, all of which seems to have between invented in Central Eurasia. Working out the basic history of humans and horses is a major focus of Part Two (reviewed in my next two posts).


  1. I read somewhere, perhaps even on this blog, that the medieval notion of "race" had nothing at all to do with the modern concept, but rather was based on a person's native language. So, in the middle ages one was considered a member of the English race or French race, etc.

    By that definition, Black and White Americans (along with Canadians...) are members of the same "race" - native speakers of North American English. Or, maybe all native English speakers worldwide would be considered members of this "race" or at least a closely related "race".

    I can't help but find the medieval definition immensely attractive and preferable to the modern one.

  2. Anon: it had some cultural element, but yes, that is basically correct it was based on language or "tongue". And the medieval definition was certainly far more practical than the modern. (Not that it could not also be a basis for status-claims and conflict.)

  3. Nice article. I think it is useful and unique article. I love this kind of article and this kind of blog. I have enjoyed it very much. Thanks for your website.
    carbon wheel rental