Say it with me: Völkerwanderung

By Razib Khan | May 10, 2010 4:47 am

toneePeter Heather’s new book, Empires and Barbarians: The Fall of Rome and the Birth of Europe, exhibits none of the minor faults which I noted in Diarmaid MacCulloch’s Christianity: The First Three Thousand Years. Heather manages to robustly balance the need for both breadth and depth, and I would even offer that this semi-sequel to his previous book, The Fall of the Roman Empire, is a superior piece of scholarship in relation to its predecessor (if a bit less compelling as narrative because of the weighting toward archaeology as opposed to literary sources). The author reports that he’s been working for 15 years on Empires and Barbarians, and it shows in the wide spread of sources and multidisciplinary nature of the argument. And that argument is in short an overturning the post-World War II orthodoxy among archaeologists, and a lesser extent historians, that cultural evolution occurs overwhelmingly through a process of the diffusion of memes, and is rarely accompanied by the flow or replacement of genes. This model is a counterpoint to the pre-World War II conception of the shift of language being a consequence of the shift of nations; ergo, it was once presumed that the rise of the English and the fall of the Celtic British occurred via the driving out of the latter by the former toward the maritime fringes of Wales and Cornwall. After World War II the sources were reinterpreted so that the Anglo-Saxon tribes were refashioned into very small compact bands of warriors who toppled the old Roman-British elite, and imposed their own language and cultural forms on the local populace (Norman Davies’ takes this model as the default in The Isles). If this was the outlook when it came to Britain, which became England and witnessed the extinction of the Celtic and Latin languages as well as the Christian religion with the arrival of the Germans, then naturally an even more skeptical take on mass migration would hold for the post-Roman German states of the Franks, Visigoths and Lombards who had a far more marginal cultural affect on the local Roman population (in late antiquity and the early Dark Ages the sources distinguish between the indigenous Romans and the various Germanic tribes decades after the fall of the Western Empire). In Empires and Barbarians Peter Heather reiterates that the view that the German tribes replacing the Roman era populations is false. But, he also objects strongly to the post-World War II consensus which would tend to minimize the extent of migration, population movement, and demographic displacement. In short, Heather wishes to rehabilitate the Völkerwanderung.

800px-Invasions_of_the_RomaIn previous posts I have outlined a theoretical framework which implies non-trivial migration of peoples, so I am amenable to Heather’s revisionism. In fact, I was a bit surprised that Heather outlines a process very similar to what I envisage in its broad sociohistorical parameters, though his knowledge of the particular instantiations of the general processes in the context of post-Roman Central Europe naturally surpasses what little I knew. But before I get to that, I would like to enter into the record a major objection I have to the argument in Empires and Barbarians. Many times within the text Peter Heather contends that the centuries long linguistic continuity of particular Germanic tribes, for instance the Burgundians, necessarily entails that the barbarians had to have brought women on their migrations. He marshals plenty of other literary and archaeological data to support this contention. For example, literary sources and analysis of burial grounds of the Goths from this period in the Balkans attest to the existence of a wagon train of women (and children) who followed the barbarian warbands along the Roman roads. But the argument from linguistics seems very weak. We have copious cases where native-speaking women are not necessary, at least in preponderance, to perpetuate a language. Heather gives one example within the text itself, he notes that the current data seem to imply that the majority of the women whom the Norse brought to Iceland were not of Norse origin. Rather, they were likely to be Irish to British. And yet no one doubts Icelandic’s Scandinavian affinity as a language. Similarly, across much of Latin America the vast majority of the population derives from the unions of Spanish men with indigenous women. The offspring, and the societies they created, are Spanish-speaking (excluding the Guarani bilingualism in Paraguay). Someone with a better grasp of the details of sociolinguistics can enlighten us on the exact details of how language is transmitted, but I’m rather sure that women are not a necessary precondition for linguistic continuity. In fact, parts of Latin America, such as Argentina, offer up an example where a continuous flow of men could have resulted in a post-Roman Germanic society where most of the ancestry was German, even if all the female ancestors during the founding generation were Romans (Heather observes that in some cases such as England and northern Gaul it looks as if there was a continuous migration of Germans for decades, if not centuries).

But that is a minor quibble in a book which is dense with data and rich with analysis. Heather’s argument is eminently reasonable and moderate. Many of the more extreme advocates of a post-modern understanding of ancient tribal identity presume that groups such as the “Goths” could emerge almost spontaneously from a welter of infinitely diverse populations (you know someone has lots of method but little knowledge when they constantly use quotations around the most banal and unproblematic terms). Examples of Roman senators raising their sons wearing trousers and speaking in Gothic can be offered up as the norm, so that a Gothic elite could emerge from the local population almost immediately. All that was required was a tiny elite of warriors to trigger the emulation from below. In this way cultural forms of the Vandals, Goths, and Anglo-Saxons swept across Europe with only trivial flows of people. It also dovetails with the revision that the Roman Empire did not fall, that it was not invaded, rather, it evolved and transformed. Heather is not convinced that the persistence of tribal identities such as that of the Goths for centuries in an environment where they were heavily outnumbered by Latin speaking natives could have persisted if their original identity was so tenuous, fluid and open. Rather, he seems to argue that there was an ethic core, to which some could assimilate, but which had at its heart an original demographic pulse from Central Europe. I find this persuasive because I have become convinced that cultural ideas move across societies far less quickly, at least in the pre-modern past, than we had long assumed. The Goths and other barbarians brought a suite of particular, distinctive, linguistic, religious and sartorial characteristics as an integrated unit, and persisted in their distinctiveness for generations, and sometimes centuries. In the post-Roman world of continental Western Europe they were eventually assimilated by the Roman substrate, and I see little evidence of their genetic impact. They were truly dwarfed by the local populations, but that can remain true even if a whole Central European tribe moves en masse into the Roman Empire. A tribe of 50,000 is a drop in the bucket demographically in many Roman provinces, but if 10,000 of those were men under arms, in the late Roman period this entailed significant capability of projecting and enforcing force. In Britain the case is somewhat different, there are genetic data which imply that some substantial replacement did occur, in particular on the “Saxon Shore.” These data are perfectly understandable when one considers the near total abolition of Romano-British norms and forms from the lowlands of what became England by the 7th century, at sharp variance with the dominance of Romanitas among the Franks, Lombards and Visigoths.

To me the falsity of the post-World War II archaeological consensus in the case of the fall of Roman Empire is so probable that Heather’s debunking is not of particular interest. Rather, I was more curious as to the underlying rationale or causes he provides of the migration. His argument is complex and multi-layered, but one aspect which I found congenial was his contention that the relatively low intensity form of agriculture practiced in German Central Europe did not produce sufficient surplus to satisfy the demand for luxury goods by the free class of German males. The taste for luxury goods emerged due to proximity to the Roman Empire, which exported them in return for region-specific luxuries (amber) or commodities. The migration of adventurers and soldiers from beyond the limes into the Roman Empire, often in military service, predates the barbarian migrations. Rather, Heather argues that the push of the Huns in the late 4rd and early 5th century, combined with the economic pull of the wealth of the Roman Empire, caused mass simultaneous movement of warrior elites from the German heartland during this period. A movement of peoples, not a band of brothers. While the Roman state could have handled one or two tribes, as it had in the 2nd and 3rd centuries, the simultaneous push from the mouth of the Rhine down to the mouth of the Danube was too much. Though some of the German groups were defeated, others were not, and once the initial breakthrough occurred there was a positive feedback loop as other tribes rushed in to take advantage of the weakness of the Empire.

The selective migration of warrior elites and their families during this period had major long-term effects. Heather suggests that it was during this period that one saw the transition of much of East-Central Europe, from Poland down to Bohemia and the north Balkans, from being one of Germanic speech to Slavic speech. German peasants no doubt remained after the emigration of their elites after the collapse of the Roman limes. And Heather does not believe they were exterminated, rather, he points to literary and archaeological evidence which suggest that there was a set of norms among Slavic migratory bands to absorb and assimilate other marginalized groups (interestingly, this was definitely the case with the Russian expansion into Siberia, where many Muslim Turkic groups were absorbed into a Russian Orthodox identity and became Cossacks). Heather implies that part of the Slavic expansion was fueled by a change in mode of production, a switch to more intensive farming techniques which produced population growth and demic diffusion in all directions, in particular toward Poland and the Baltic more generally. Because of the relative lack of literary evidence this section of the book is not totally persuasive, but the fact remains that much of what was German in 500 was Slavic in 1000.

Empires and Barbarians concludes at the year 1000. By this time intensive farming and urban civilization, at least in fragments, had reached most of Europe. Local elites were no longer transitory in their expectations, so a mass migration of a whole ethnic group was no longer in anyone’s interest (much to gain, but much to lose!). The non-Mediterranean farming system of three-fields, as well as improved plowing technologies, had shifted the demographic center of gravity north. Extreme gradients of elite wealth and social complexity which had characterized the Europe of the Pax Romana were no longer operative. Without gradients there would naturally be less flow. The great chaotic demographic transient between the rise of Rome and the emergence of medieval Europe was over.

Note: Empires of the Silk Road and The Horse, the Wheel, and Language are excellent complements to Empires and Barbarians .

Image Credit: Wikipedia:


Comments (9)

  1. bioIgnoramus

    I recently read an account of a Romano-British will from the Dark Ages, which (presumably) gave a good idea of what a farming family would have owned. The author of the book pointed out that the total of archaeological finds from that era for that group added up to less than the items listed on that one will. Something pretty conclusively destructive had happened in post-Roman Britain. (He also pointed out that we wouldn’t have very much evidence of what the German invaders owned if it hadn’t been that they were pagans and therefore buried grave goods.)

  2. Bolek

    It’s pity that historians don’t read papers on genetics and therefore produce outdated books from the very beginning. In recent paper Underhill at al. 2009 it was shown, that typical Slavic clades of hg R1a1 were present in Poland much earlier than migration period:

    It is therefore very unlikely that Central Europe was Germanic before 500 A.D. and became Slavic after 500 A.D. can. There is nothing in genetics, anthropology, language or archeology that would support such theory.

    Linguists like F Kortland say:
    ”Gothic is closer to Upper German than to Middle German, closer to High German than to Low German, closer to German than to Scandinavian, closer to Danish than to Swedish, and that the original homeland of the Goths must therefore be located in the southernmost part of the Germanic territories, not in Scandinavia (1982, 1984, 1987a, 1987b, 1992). I think that his argument is correct and that it is time to abandon Jordanes’ classic view that the Goths came from Scandinavia”.

    So it looks like Goths didn’t come from Scandinavia. No surprise as Jordanes’ stories are quite often fantasies. He wrote for example about some struggle of Goths with Egyptian pharaoh after living their homeland.
    Most likely Germanic tribes which later became Goths originated somewhere in southern Germany and went along Danube before reaching Ukraine. They had nothing to do with Sweden or Poland.
    There is no much evidence for very old Germanic presence in Poland. No trace of Germanic language in hydronyms, no influence on Polish language which is considered to be the most archaic among Slavic languages and closest to PIE.
    There is also no evidence of any substantial and not recent Germanic genetic input. There is very little I1 in Poland, and not all of it can be linked to Germanic I1 as MRCA for typical Polish clade I1-P is calculated to be older than typical Scandinavian clades I1-Norse or Germanic I1-AS. See:

    R1b-U106 is also very scarce among native Poles. No Goths related haplotypes have been identified so far.

    Historians base their speculations mostly on Tacitus’, Jordanes’ or Ptolemy’s writings. But Tacitus or Ptolemy were not researching ethnicity of various tribes living in geographical region called Germania Magna. They didn’t study their languages or genes. It seems now that most of those tribes if not all living east of Elbe river and north of Carpathian mountains were actually R1a1 and speaking Slavic before 500 A.D. and after 500 A.D.

    We know genetic makeup of Germanic tribes of that time. In Capelli at al. 2003 migrations of Germanic tribes into central England were studied:

    It is evident that we have there negative correlation between Germanic and R1a1. By studying PCP of Figure 3. and hg. frequencies in Figure 1. we can see that the most Germanic areas of England (ie. the regions where Germanic input is over 60% like York, Norfolk, Western Isles etc.) have R1a1 level of barely 2%, much below the country’s average and thirty times lower than Slavic countries like Poland.

    The conclusion is that Germanic people cannot be associated with R1a1 and indigenous Germanic tribes(i.e. coming from territories where German language originated) invading England– Jutes, Saxons, Angles, Frisians – at that time had very little of R1a1, if anything at all.
    It is good that you mentioned books like “Empires of the Silk Road” and “The Horse, the Wheel, and Language”. Modern genetic studies showed that only R1a1 had been identified among aDNA of IE steppe people. Not a single haplotype of hg. R1b-M269 or I1 – which are Celtic and Germanic markers – has so far been found among Indian higher caste Brahmins (descendants of IE Aryans, which are linked with Andronovo and Yamna cultures).

    This proves that Celtic and Germanic ethnos cannot be related to Tocharian and Yamna cultures as many fantasized. There is no evidence for this. Genetics shows that it is rather impossible. Those books are therefore wrong in their speculations concerning ethnicity of Tocharian, Yamna, Maykop or other steppe cultures and linking them with Celts or Germans. Anthony’s book is also completely wrong in its linguistic part. It is a total nonsense to ignore most important IE languages, most archaic and closest to PIE with all the necessary vocabulary for wheel, horse etc. from which all other IE words can easily be derived. In my opinion such mistakes and disregard for genetics make it a crap worse than Wikipedia or some other internet “science” produced by Celtic or Germanic nationalists. It is sad, that some people treat seriously all those fantasies about Celtic or Germanic ‘history’.

    I haven’t read Peter Heather’s new book yet, but from your description I can see that it is repeating old mistakes and has disregard for genetics and scientific discoveries of recent years. So I don’t expect much from it.

  3. Eric Johnson

    I’m sure the size of a foreign pinnacle-elite matters in itself. But what about endogamy. Perhaps Italy etc would have been very slowly germanized (genetically and culturally) if the Germanics there had remained endogamous? Is there any evidence that the Germanics did/didn’t remain endogamous in England, in Italy/etc?

  4. Is there any evidence that the Germanics did/didn’t remain endogamous in England, in Italy/etc?

    the literary evidence from italy exists because italy was a literate society, and the royal family of theodoric managed to intermarry with other visigoths. some of this obviously due to religion, the goths were arian, the romans catholic. OTOH, the royal family of the vandals accrued some prestige by intermarrying with the imperial families of the west. additionally, i believe that the christian bias toward monogamy took a few generations to “take hold” in many of these barbarian european tribes, who were loosely polygynous quite often at the elite levels. so the german males could have it both ways (look at charlamagne’s numerous concubines, 4 centuries after the christianization of the franks).

    in england, the genetic data, which is subject to great deal of interpretation, slightly indicates that the male ancestry is more likely to be german than the female. so i don’t see support for endogamy there. in fact, some of the male forebears of the dynasty of wessex had celtic first names, suggesting that elite british males were integrated into the west saxon order.

  5. Eric Johnson

    > i believe that the christian bias toward monogamy took a few generations to “take hold” in many of these barbarian european tribes, who were loosely polygynous quite often at the elite levels.

    I’ve heard similar about the vikings, when they were first christianized, later on in history.

  6. Eric Johnson

    > so i don’t see support for endogamy there. in fact, some of the male forebears of the dynasty of wessex had celtic first names, suggesting that elite british males were integrated into the west saxon order.

    Fools! If only had read Greg Clark, they would know they could probably have replaced the whole country if they stayed endogamous, thus maxing out their fitness.

    Maybe such norms aren’t easy to enforce even if someone wants to, though. It must often go against the individual interest.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar