I’m making my way through the Oct. 9, 2015 TLS, and have just read Patrick Sims-Williams’ review (available here to subscribers) of a British Museum exhibition on the Celts; I thought the last couple of paragraphs worth reproducing:
The Director of the British Museum introduces Celts: Art and identity as “not so much a show about a people as a show about a label”. An exhibition may not be the best way to explore the “Celtic” label because the clearest evidence for long-term Celtic identity is linguistic, not visual. Language is the basis for modern pan-Celticism. Buchanan (1582), Leibniz (1699) and Lhuyd (1707), who first applied the “Celtic” label to Gaelic, Welsh and Breton, did so not because Celtic was a prestigious term to appropriate but because they saw that these languages are related to the language of the Gauls who called themselves “Celtae” according to Caesar. Already in the ninth century the element dunum in Augustidunum (modern Autun) had been correctly identified as “hill” in Celtica lingua. The obvious next step was to link it with Gaelic dùn (“fort”) and start to map Celtic place names across Europe. Thus Celts as Celtic-speaking peoples were in place long before any notion of Celtic art. That only arrived from 1851 onwards with Daniel Wilson, J. O. Westwood and J. M. Kemble. Tipped off by Kemble, Ferdinand Keller finally attached the Celtic label to the La Tène art of Lake Neuchâtel in 1866, the year in which Matthew Arnold could still take the ancient Celts’ “inaptitude for the plastic arts” as a given.
In view of the primacy of language, it is a shame that the exhibition is so reticent about it. It is said that the Celts “left almost no written records”, and “Celtic-speaking communities” are hardly mentioned before the medieval sections. Just one inscription is included, the sixth-century Irish “Maccutreni Saliciduni” inscription from Powys (note dunum again). It would have been good to show a Celtiberian bronze inscription and a pre-Herodotean Celtic inscription from the Italian lakes. Both would come from areas devoid of “Celtic art” as understood here and would enhance the exploration of the label Celtic. The fact is that while most ancient Celts spoke what we call Celtic languages (so far as we know), only some Celtic-speakers went in for “Celtic art”. Conversely, “Celtic art” probably appealed to many who were not in any sense Celts, as it still does today.
An earlier paragraph doesn’t have to do with language, but puzzles me:
Reaching the Roman period, the focus narrows to Britain and Ireland. When the Roman army left Britain, “new pagan leaders gradually established Anglo-Saxon kingdoms”, reads the caption. Perhaps it would have been helpful to let on that these pagans had crossed the North Sea? But “migration” and “invasion” have been taboo words in British archaeology for decades, and are avoided throughout this exhibition. The curators are even partial to the groundless speculation that the Celtic languages originated c.2000 BC on Atlantic shores and were never carried there from anywhere else.
Anybody know why “migration” and “invasion” are taboo words in British archaeology? Surely nobody is claiming the Angles, Saxons, et al. were native to the isles!
I’m not an expert in the field by any means, but in the context of the Celtic language in Britain, I believe the prevailing viewpoint used to be that Celtic languages and culture were forced onto preexisting cultural groups by waves of invaders. Then in the 70s it switched to more of a sort of cultural borrowing idea, where there was contact between Continental groups and Isle groups. Invasion as a word tends to have a complex relationship with archaeology, because it brings to mind groups killing all the men and taking over, and that’s not usually the whole story. Migration I’m less certain of why it’d be taboo, unless it’s because somebody decided it brought to mind images of far more people than the archaeologists believed actually moved in to the isles.
But yes, I’m fairly certain that almost everybody admits that the Saxons and so on were not native (I won’t say everybody, because archaeology has a tendency to always attract that one person who wants to argue that space aliens, Egyptians, Vikings, the tribes of Israel, etc, were really the people who were there, so it wouldn’t surprise me if somebody is arguing that the Saxons were really native and then fought their way back).
I’ve always been curious about why the Lindisfarne Gospels look so Celtic, even though they were produced in Northumberland by a person named Eadfrith.
Not entirely, no. But the majority view is that most Britons are descended from pre-Celtic settlers, and that both the Celtic- and the English-language shifts were a matter of elite dominance rather than displacement, much less genocide. Things went similarly with the Normans, after all, except that there was no language shift.
Just you wait.
I expect soon this viewpoint will be adopted across the Atlantic and it will become commonplace to write how Americans abandoned tribal wars, adopted a Constitution and established the United States, without ever mentioning that the Americans of the scalping and tomohawking kind and those who wrote the Constitution were somewhat different Americans…
Surely nobody is claiming the Angles, Saxons, et al. were native to the isles!
While Saxons being from Saxony (in what is now Germany) and Jutes being from Jutland (in what is now Denmark) is obvious from the (fairly well-known) names, the original homeland of the Angles is, while almost as obvious, somewhat more obscure (it’s Angeln, a small peninsula off Jutland).
Patrick Sims-Williams obviously alludes to the influence of Sir Colin Renfrew and his consistently anti-migrationist stance. As you perhaps know, Renfrew has also written (and pontificated) on linguistic issues such as the origin of the Indo-European family (the Anatolian hypothesis). According to him, once the initial farming expansion took place in the seventh millennium BC, most of the known branches developed more or less in situ, and e.g. the Celts consolidated into a linguistic unit on the Atlantic seaboard (from Iberia to the British Isles) and throughout much of Western and Central Europe by developing “cumulative mutual Celticity” from ca. 4000 BC onwards. Likewise, the ancestor of Indic was spoken in and around the Indus Valley already in the sixth millennium BC.
@SFReader, I’m not saying you are wrong about willful ignorance past, present and future, but the mid first millennium cultural / linguistic transfer into Britain didn’t bring with it technology that allowed a five-fold increase in population, nor diseases to almost exterminate the previous inhabitants. So without further data, it is more likely that the present UK population has a strong pre-Germanic genetic component than that the US population is dominated by pre-Columbian genes.
The stereotype of a short-statured, pasty-faced, black-haired working class and a tall, lighter-haired aristocracy does show a consciousness of non-assimilation between aborigines and invaders. The latter would currently be dominated by Normans, but the former could conceivably have resisted change for several millennia.
Other hatters probably have real genetic data at hand.
Genetic data do indeed exist, and the Britons are about the same mixture (very roughly 1 : 1 : 1) of Western Hunter-Gatherers, Early European Farmers and Yamnaya people as everyone around them far & wide (Basques included). Wave after wave of invasion, yes; the occasional river of blood, yes; wholesale population replacement – not since Britain’s ice cover retreated last time.
Concerning Renfrew, I agree with him that we should expect the introduction of agriculture – which after all involved considerable population movement into Europe: the EEF – also introduced a language family. It just wasn’t IE. More likely, its extant representatives this side of the Caucasus are Basque and those Basque-like words in Sardinian. Likewise, agriculture may have reached India together with the Dravidian languages.
The people of the Andronovo culture, thought to have spoken something in the Indo-Iranian branch, had some EEF ancestry, implying some degree of back-migration from central Europe to the steppe. I think that’s in the original paper on the genetics of the Yamnaya and Corded Ware/Battle Ax people.
Y: There were strong links in both directions between 7th century Northumberland and 7th century Iona – exiled Northumbrian royals in Iona and Hebridean Irish monks at Lindisfarne.
SFReader is obviously joking, but some people actually have tried to make a case that intermarriage had a larger role to play in forming colonial America than we normally assume. If you find this sort of thing interesting, try Gathering the Bones 4 or To Our Mother of the Lakes. But genetic data would be a lot more useful here…
@David, my understanding is that genetic-‘Yamnaya’ has to a large extent been taken as a proxy for ‘IE peoples.’ But if it’s evenly present all over Europe, including the Basques, either that’s wrong or it demonstrates that linguistics and genetics do not correlate at that time depth. (Or I’m missing something).
But while those old genomes are fascinating, are there any markers for more recent splits that could be used to separate ‘Celtic’ and ‘Germanic’ and ‘other’ and tell us to what extent more recent cultural takeovers in Britain brought in new genes as well? Or is anything gained by looking for differences between the southeast and the north and west, even Ireland?
the Americans of the scalping and tomohawking kind and those who wrote the Constitution
To be fair, a scalping-and-tomohawking criterion probably wouldn’t be of much use in determining who in colonial North America had European ancestry or not.
Just what I was going to say!
@Y:
Christianity was spread among the northern Anglo-Saxons primarily by Irish missionaries, so it’s not that surprising that northern A-S religious art has a strong resemblance to Irish religious art.
When I was at school in England we learned that our people were the outcome of repeated invasions and incursions — Romans, Angles, Saxons, Danes, Jutes,* Vikings, Normans etc etc. I don’t know if professional archaeologists have come up with a different story sometime during the past half-century.
In any case, invasions came to an abrupt end after 1066, so that by the time the Tudors appeared on the scene the English had miraculously morphed into the genetically pure and superior race that later built the empire (occasional German kings and queens notwithstanding).
*aka Burlaps
There was a really great Stewart Lee standup routine about British attitudes towards the successive waves of immigration to the islands (Bulgarians, Huguenots, Anglo-Saxons, Beaker people, fish): https://www.youtube.com/watch?v=6KVO378tjsw
That is great. Bloody Anglo-Saxons with their epic poetry!
Or that they correlate at that time depth – but not so much later!
This isn’t done with specific markers in the first place, but with statistical similarity (composed of thousands of unspectacular markers that nobody looks at in detail).
That should help; I don’t know if it’s already been done.
Well, if Britons have the same mixture as other populations around, then couldn’t have been large-scale population replacements by those other, similar groups, without leaving much of an observable signal in this analysis?
P.S. I thought the Yamnaya-like population includes Western Hunter Gatherer ancestry and that’s the only significant source of WHG in modern populations, so I wouldn’t juxtapose WHG with Yamnaya.
statistical similarity — yes, I didn’t think there were actual single markers, some sort of statistical procedure must be necessary to get signal from the noise. I imagine that it’s akin to principal component analysis, aka rebasing your Hilbert space to maximize energy in the lower dimensions — with the better known ‘types’ identified as the major components in world- or continent-wide variation, and then if you zero out those and look at variation over a smaller area there might be new signals there.
…Oh. I remembered two results from two different papers, years apart, and managed to keep them in separate compartments of my mind. I think you’re right.
No, Yamnaya is a mixture (about 1 : 1) of Eastern Hunter-Gatherers and Caucasian Hunter-Gatherers. EHG are intermediate between WHG and Ancient North Eurasians (Mal’ta).
Yes, I’ve seen a website that argues precisely this at length based on hydronymy inter alia. I’d google up the link, but, on the other hand, what’s the point? Your imagination of what it would be like is probably not a lot less informative than the actual content.
I’ve read people suggesting that the henge-building people may have spoken a language akin to Berber. However the argument is based on a small number of placenames that appear to be pre-Celtic.That’s not much to go on.
It used to be common practice to attribute cultural change in Britain and/or Ireland to migrations and/or invasions from the mainland, neglecting trade, local development, etc. So to restore balance, it is common practice to refer everything to trade and/or local development. Grahame Clark’s “The Invasion Hypothesis in British Archaeology” discusses both the older and newer tendencies as of 1966.
The Celts are the salt of the earth. (Or the silt perhaps.)
I read not long ago about migration of OE speakers back to the northern mainland, leaving some linguistic traces in what would become later continental Germanic languages. I can’t find the reference, and I may have gotten the story wrong.
The Saxon invaders didn’t depart from the modern German state of Saxony. The Saxony of the day was on the coast, just below Angeln and Jutland (edited to say I’m a goofball – the ancestral area of the Saxons is of course still considered lower Saxony). The Saxons later pushed up the Elbe – forcing the Slavs out, or perhaps imposing themselves by replacing the Slavic elite.
David, you may be more up on this than I, but I had thought some of the WHG ancestry was picked up en route, post steppe but before the full expansion. And my understanding is that it is not a homogenous mixture WHG across the continent. There was some local admixture, in some cases continuing for a number of generations, particular in marginal areas, resulting in things like a higher WHG ancestry in some Scandinavian populations.
Also the Indo-European invasions were “y-chromosome mediated,” meaning, essentially that while there is WHG ancestry, it’s overwhelmingly female. The genteel 20th century archaeological consensus that cultural assimilation was far more powerful than migration and invasion has been pretty much destroyed by barbarian geneticists. The WHG males were for the most part, ahem, “not assimilated.” And it’s doubtful that WHG women were simply exercising their free will to choose higher status mates.
News to me, but I’ll read up on this.
Indeed not; there’s a northwest-southeast gradient, inverse to that of EEF ancestry. It’s fine-grained enough that there’s a tiny bit more EEF ancestry in Andalusia than in the Basque Country. On top of that, there’s a northeast-southwest gradient of Yamnaya ancestry, nowadays peaking in Estonia.
Do you have a reference for this? It’s not in the much-discussed papers from a few years ago, but I may well have missed more recent findings.
I can’t quite figure out what Bench Grass is arguing.
http://proto-english.org/ argues that Germanic-speakers had arrived in the British Isles before the Roman conquest, and generally that language replacement isn’t as common as in mainstream interpretations. But there are an awful lot of initial K-s in the personnal and tribal names. An early Germanic language might have a borrowed K like in Kent or an eroded Sk, but why would they have many borrowed Ks in Caratacus, Camulodunum, Cantiaci, Catuvellauni, Corieltauvi, etc.?
I think certain other people insist that the Anglo-Saxons are THE indigenous people of the British Isles, because otherwise they’d be immigrants. Like the Welsh. And as noted above, the Fish!
The very earliest version of Tolkien’s mythology had the English as the original humans of Britain, leaving and then returning. Shortly thereafter, thank goodness, he abandoned the whole ethnic aspect of his mythos. It’d be interesting to know if he was drawing inspiration from some previous notion.
David,
Here’s a Science magazine version:
http://www.sciencemag.org/news/2017/02/thousands-horsemen-may-have-swept-bronze-age-europe-transforming-local-population
“Even more lop-sided than the mostly male wave of Spanish conquistadores …”
And here’s Razib Khan on the contemporary South Asian migration:
>In sum, the balance of evidence suggests male mediated migration into South Asia from Central Asia on the order of ~4-5,000 years ago. There are lots of details to be worked out, and this is not an assured model in terms of data, but it is the most likely. In the near future ancient DNA will clear up confusions.
I’m not a fan of most of his politics, but he’s certainly the best, most accurate web-distiller of the latest genetic data that I know of. This is the consensus that is building, as far as I can tell. Male dominated mass migration and conquest is responsible for the initial trans-Eurasian expansion of the Indo-European languages, full stop. Other, later examples of language spread are likely more complicated.
I think this is the full subtext of the passage in the Celtic art review that Hat asks about. It’s a catty quote, underlining the degree to which genetics and much of archaeology remain out of synch, and the degree to which the 20th century archaeological point of view has already lost, though its defenders haven’t yet entirely succumbed.
The one thing I’m mystified by is the quote in my previously linked article fro David Anthony, an archaeologist whose work is otherwise in complete concordance, and in some ways established the foundations for, the genetic orthodoxy, in the the Horse, the Wheel and Language. Anthony suggests that with male migration continuing for generations, something must have been wrong on the steppe to keep driving them out. I wonder what Anthony would posit was wrong with Iberia at the height of Spanish Habsburg dominance of Europe, that kept driving Spanish men to find their fortune and their mates in the Americas.
Morally wrong, perhaps, but I think Anthony is suggesting something economically wrong. Alas, that doesn’t seem in keeping with human nature.
I wonder what Anthony would posit was wrong with Iberia at the height of Spanish Habsburg dominance of Europe, that kept driving Spanish men to find their fortune and their mates in the Americas.
Primogeniture? Anyhow they weren’t just “Spanish men,” they were disproportionately Extremeños. Not exactly the most economically prosperous part of the peninsula.
Marja Erwin: In a maddeningly allusive and roundabout fashion, Bench Grass is arguing that most 18th century “white” Americans were actually mestizos, that marriages between European and Native American big shots (think Pocahontas) were a key part of how certain families managed to claim large tracts of land, and that both these facts were massively downplayed or outright suppressed for various historical political reasons. I find the idea thought-provoking and not inherently absurd, but boy does it need some quantification.
This paper, based on 23andme data, says:
That’s hardly consistent with the term mestizo as generally understood from Mexico and points south. In any case, there was no cover-up: in the 1920s the Virginia aristocrats with actual or claimed Native ancestry lobbied the legislature so hard that the 100% blood quantum for whiteness had to be reduced to 15/16 with the remaining 1/16 Native. Rich Virginians did not want themselves and their families treated legally as “Negroes, Mongolians, American Indians, Malayans, or any mixtures thereof, or any other non-Caucasian strains”.
I can’t see what’s supposed to be suspicious about /k/ in Germanic; that’s beyond my ken. 🙂 Proto-Germanic *k comes from PIE *g and *gʲ, which weren’t particularly rare.
Initial *p would be suspicious, because PIE *b more or less didn’t exist.
With link to the PNAS paper – thanks!
Interesting. Do you have a direct link?
JC: I actually sent him that paper a while back, as it happens, and he posted on it a few days ago. Unfortunately it’s inconclusive for one crucial reason: we (or at least I) don’t know what proportion of European Americans are descended from people who got here before 1800, or in what ratio. That could probably be estimated by cross-checking with genealogies, but really we need to dig up a few Colonial graveyards and do DNA tests on the results. After all, if you’re gonna DNA test Native American remains, it’s only fair to do the same to white folks’… ; )
There is an overview of the history of stories about Migration Age Britain in Guy Halsall’s “Worlds of Arthur.” He lets himself get peevish and bluster rather than coolly laying out the evidence for some specific points, but he lays out how the stories which scholars tell have changed since the 19th century.
Two key concepts are that scholars between say the 1970s and 2000s realized that earlier generations were relying on =folk theories= (the ‘Sid Meier’s Civilization’ view of migration, where ethnic groups bounce around like billiard balls driving other ethnic groups before them and occasionally knock one off the table) and that =their archaeological criteria for identifying ‘Germanic invaders’ failed= when you applied the same criteria to modern societies or to earlier periods of ancient history when nobody claimed there were massive invasions. That movement won, and now there is a counter-movement pointing out that just because ethnic groups appear and disappear and merge and transform doesn’t mean that we can dismiss all the evidence for migrations of large populations in antiquity.
I would recommend that anyone interested in prehistory read it before they look at any of those generic studies, because he shows what you need to know to think effectively about ethnicity and migration.
“Initial *p would be suspicious, because PIE *b more or less didn’t exist.”
Proto-English’s reconstruction of a Germanic-speaking zone in Britain includes the land of the Parisi on the coast.
Doubtless from the otherwise wholly unknown and probably phonotactically impossible PIE root *bh₂er- (not to be confused with *bʰer-), no doubt.
Indubitably!
I’m for it.
Well, actually, you could craft *|ph₃er|- = */bh₃or/-, which would totally become *par- in Germanic. I won’t bother looking up if such a PIE root has ever been postulated, though.