Martin has responded at length to my posts where I argue for an inclusion of genetic data in a synthetic model of human history and development. There are multiple issues here where we disagree, or differ in our interpretations. First, as Martin admits at one point, “Those are all my words put into Razib’s mouth.” Much of the post is a misunderstanding. Because I would rather devote my time to discussing positively my ideas, as opposed to misunderstandings or misconceptions of what I believe, I’m not going to deconstruct every single point which Martin attributes to my own sense of the past. Don’t misconstrue this for an avowal of the positions which I leave on the table. Time is finite and I simply don’t feel obligated to clear up confusions deriving from one post. I have commented on these topics long enough and with enough detail that I suspect regular readers can naturally derive the implicit background axioms with little difficultly. Second, Martin states “I myself haven’t studied natural sciences in any organised way since high school.” This is a problem, because his idea of what I believe in regards to history and social dynamics aren’t the only points where his interpretation is off; I think he signals some misunderstandings of biological dynamics as well. Finally, in regards to the Etruscan & genetics, Martin states, “I just don’t find it very interesting that this confirms statements in a poor historical source.” Much of the disagreement derives from a different emphasis of the ends of historical research, from what I can tell. That is, we’re interested in different questions. I do happen to think it is very interesting that genetic science has falsified the consensus of archaeologists and validated a myth recorded by Herodotus.
On the issue of ethnic identity, I’m an instrumentalist. I’m clearly not an essentialist as Martin assumes I am, I am not an essentialist on species concepts, so it makes no sense to say that I’m an essentialist on cultural categories which are by their nature more fluid. Use of the term “Slav” is shorthand, and it does not imply certainty that a group or individual fits all the criteria which one adduces to a Slav. Rather, it is a label for a correlation of traits which span geographic space and imply a cultural complex. The analogy to racial/population groups is relevant; these are simply labels for statistical constructs which trace a distribution of gene frequencies across populations. And, unlike genetic variation cultural variation can manifest sharp differences and crisp boundaries because of the nature of inheritance. While one by definition receives half of one’s genes from two parents, there may be a bias in inheritance of cultural traits from one line of descent. For example, a region where patrilocality is dominant may show little genetic variation because of movement of females from one village to another, but because the offspring are raised in the cultural unit of the father there still remains strong differences in dialect and religious custom. As an example, the cultural difference between Anatolian Turks and Greeks is far greater than the genetic distance. This does not mean that cultures are “turnkey” modules, where characteristics perfectly integrated into a functional whole. The Pomaks are Muslims in Bulgaria who speak a Slavic language. They share a religious identity with their Turkish neighbors, and a linguistic one with their Orthodox Christian neighbors.1 Here are the current ethno-linguistic and religious demographics for Bulgaria:
According to the 2001 census…Bulgaria’s population consists mainly of ethnic Bulgarian (83.9%), with two sizable minorities, Turks (9.4%) and Roma (4.7%). Of the remaining 2.0%, 0.9% comprises some 40 smaller minorities, most prominently in numbers the Russians, Armenians, Vlachs, Jews, Crimean Tatars and Karakachans. 1.1% of the population did not declare their ethnicity in the latest census in 2001.
96.3% of the population speak Bulgarian as their mother-tongue.
…
Most Bulgarians (82.6%) belong, at least nominally, to the Bulgarian Orthodox Church, the national Eastern Orthodox church. Other religious denominations include Islam (12.2%), various Protestant denominations (0.8%) and Roman Catholicism (0.5%); with other denominations, atheists and undeclared totalling approximately 4.1%.
It is generally appropriate to assume that Turks, if they have a religious identity, will be Muslim.2 But this leaves around 3% of the Muslim population which by the numbers must be non-Turkish. Some may be Roma, and some are likely Pomak. The Roma are religiously diverse, so many of them may be Orthodox. Where does this leave us? Obviously religion, language and likely region play a major role in various identities in Bulgaria today, there are complicated contingencies here, but that does not mean that we can not engage in generalization to capture the patterns we see before us! If you know someone speaks Turkish you can be assured that they are likely Muslim (or Muslim identified). The certitude is lower that one is Orthodox Christian if one speaks Bulgarian, but, if one adds a geographic qualification then your probability calculation shifts because Pomaks tend to be concentrated in the Rhodope mountains.
The point is that there are patterns extant within non-essential statistical categories. Characteristics are not distributed randomly, but exhibits known distributions. Even if we can not speak with total certainty because of one characteristic, a sequence of conditional probabilities generated by the intersection of characteristics can allow us to make a strong inference. The probabilistic nature of inferences will result in the inevitable error, but the key is to attain an understanding beyond our null hypothesis and random expectation. If we take an extreme rigorous/experimentalist interpretation of science then we must discard much of geology, paleontology and astronomy. I do not believe most would be satisfied with this, so we make due to with the tools we have to generate inferences when experiment is not available. Similarly, in the historical sciences the data is scattered, noisy and subject to equivocal interpretation. But that does not mean that general truths do not exist and can not be discerned with cautious analysis.
When Martin contrasts the nature of archaeology, and social science in general, with an understanding of species and biological dynamics, he manifests ignorance of the latter topics. Consider:
Archaeological cultures and linguistic areas are not like biological species. One end of a culture blob often hates the guts of the other end, insisting that although the two have similar languages and pottery styles, they are in fact not related at all. One end will happily join forces with part of an adjoining blob to annihilate the other end of the home blob. As a result, a third blob may be born, consisting of ten sub-blobs that fight constantly among themselves while exchanging wives and porphyry axe-preforms with the two parent blobs.
In reality, as I have implied elsewhere, I am interested in the dimension of within species competition. To be succinct about it a great deal of evolutionary process occurs because of competition of individuals & groups within species. Assuming that cultural dynamics are a sharp contrast illustrates a lack of knowledge of the complexity of multiple-levels of biological dynamics. The similarities between the two processes have even resulted in the production of cultural models which borrow heavily from population genetics.
The final issue is one of emphasis. Martin has brought up the example of Croats and Serbs, the Cantonese and Mandarin dialects (I am not totally comfortable with the 1:1 comparison between Cantonese and “Mandarin,” since the latter is geographically fragmented and many different regional identities are subsumed into this one dialect label). I really don’t know where that’s coming from, but I suppose he presumes that I have a naive conception of ethnic identity. As someone who has visited villages in Bangladesh where locals have asked if I was a “foreigner,” but which they meant someone from across the river (in primary river in their district), I would reassure him that I have no such illusions. Rather, it seems to me, as Martin implies, many identities are newly constructed and are quickly “fit” into national narratives. That being said, I also think that time and space are parameters we need to take into account; the post-French revolutionary nation-state is one extreme model, which most “modernizing” nations have taken up when possible, but it obviously isn’t appropriate for the ancient, let alone pre-literate, societies which are at issue in this post. Ancient Turks and Slavs surely spoke mutually intelligible dialects (unless linguistic homogenization has occurred), but they were not nation-states. Pan-Turkism and Pan-Slavism are the children of the Enlightenment; not echoes of ancient days.
With that being said, I still think a term like “Slav” is appropriate and useful. Even if a tribe or clan were primarily identified with their local identity, that does not mean that their macrosocial group has no relevance. Antelope do not think of themselves as antelope, but it is appropriate to make generalizations about antelope.3 When I speak of religion I don’t talk about something that I think has a grounding in reality; I don’t believe in the unique truth claims of any religion. On the other hand, believers do accept their own truth claims, and believe that the differences are so great that they are important enough to kill over on occasion (and certainly important enough to make or break a marriage possibility). The preconceptions of a conscious agent is one thing, but the mass action of thousands of conscious agents is another thing altogether. I am not particularly interested in how Slavs viewed themselves because as a methodological matter I think across much of human history conscious reflection is overrated.
A clear example of the dynamics I’m getting at already exists in the literature. In the 1970s L. L. Cavalli-Sforza proposed the model of demic diffusion for the expansion of agriculture into Europe from Anatolia and the Levant. The extreme cases would be that agriculture expanded through cultural diffusion or that it spread through the migration of peoples. Cavalli-Sforza suggested that population expansion would have resulted in the spread of agriculture as demes, populations, which practiced the new cultural technique grew at a faster rate than hunter-gatherer groups. This is not a purely theoretical model; there is a body of anthropological literature which contrasts the pro-natalism of farming peoples (more young hands to work) with the anti-natalism of hunter-gatherers (more young mouths to feed). This is where genetics comes in. The short of it is that the most probable scenario is that 25% of the ancestry of Europeans, as a whole, descends from a population expansion out of the Middle East in the last 10,000 years. Critics of Cavalli-Sforza have used this number to suggest that it was cultural diffusion which spread agriculture, but he has pointed that the 25% number is totally acceptable in a wave of advance model where the initial genetic signal will become weaker with admixture every generation. Another key point is that the frequency of “Neolithic” genes seem to decrease not as a pure linear function of distance from the point of origin (Anatolia); rather, coastal regions and river valleys in southern Europe tend to exhibit higher frequencies. What the genes are telling us here is that the balance between migration/demographic expansion and cultural diffusion was likely influenced by geographic parameters. Ideas were more apt to climb up mountains than people!
Is this interesting? I think it is. I don’t know if Martin would think so. Obviously the peoples who expanded into Europe from Anatolia did not think of themselves as “Anatolians.” It seems plausible that they spoke related languages, but the overall point is that they spread a cultural toolkit which eventually conquered Europe. Part of this conquest was that they reproduced at high rates. And part of it was the spread of the new lifestyle to the indigenous population. The balance between the two can be answered, or inferred, in part from analysis of patterns of genetic variation.
At this point it should be clear that these were the questions I was probing in my initial posts on the Slavs; why is it that over the past 2,000 years Slavic languages have spread at the expense of Finno-Ugric languages? The historical linguistics and the geography of Finno-Ugric strongly implies this (the expanse of Russian has embedded within it many Finno-Ugric dialects). One could posit that this was a random event, though I would point that that a parallel process occurred in the Balkans and central Europe. Slavic languages are roughly intelligible to this day, which strongly implies a recent ancestry. Some of the genes, ancestral lineages, which are exist across groups we label “Slavic” are not shared across Slavic groups, and some of them are. It seems that a parsimonious explanation of this pattern would be the expansion of Slavic groups and their genetic assimilation of indigenous peoples into their cultural complex. I would leave the Croat-Serb differences to political scientists.
1 – These religious identities themselves may be notional and not strictly adhered to.
2 – Prior to the ethnic exchange of the 1920s large numbers of Turkish speaking Christians and Greek speaking Muslims existed. But the literature I have seen suggests that even if an Orthodox Christian Anatolian spoke Turkish as their first language, they did not identify as Turkish in the manner that their Muslim neighbors did. Note that in much of Europe “Turk” is synonymous with Muslim identity despite the fact that those Muslims may not be ethnic Turks!
3 – Please don’t object that antelope can only breed with other antelope, it’s actually more complicated than that. Biological species concepts are quite often fuzzy as well.
Comments are closed.