Are the Irish Celts?

Expansion of Hallstatt and subsequent La Téne horizons across Europe.

Probably the most controversial question asked of mainstream Irish academia over the years is as to whether or not that Irish people can be categorised as a definitively 'Celtic' meta-ethnic group.

Until the past few decades, the general consensus amongst Irish scholars was that a migration from somewhere on continental Europe occurred c.500BC and thus with it bringing the precursor of the Irish language. However this theory has since waned in support due to the subsequent backlash of PC revisionism amongst academics during the 1990's. The word 'Celt' was added to the ever increasing list of racial slurs that you couldn't say anymore because it triggered people.

So are we or aren't we? And if not then what?

Well after spending a considerable amount of time playing around with different models on Global 25, I would actually in earnest say yes......but also no.

You see due to the dearth in available samples from prehistoric Ireland at this current moment in time, at least until may of next year when the embargo is lifted on the ancient Irish DNA compendium courtesy of Trinity college's science department, I have to improvise and make do with samples from across Bronze age Britain as substitute proxies. But beggars cannot be choosers I suppose.

In a previous post I mentioned that between now and the early bronze age period, at some point there was an admixture event which led to a 'southern shift' noticeable in contemporary Irish genomes relative to the more 'northern' Rathlin samples who share most drift with Dutch beakers of which Irish score around 80% as on Global 25.

As you can see being isolated on an island for
four millenia has led to a fair bit of drift since then.
Dstats courtesy of Eurogenes blog.

So what caused this southern shift? I think the answer at this stage is fairly obvious. Last year foremost population geneticist David Reich of Harvard commented that during the Iron age in south-east England samples show evidence of a resurgence in EEF like admixture. One of the scenarios postulated was introgression from a remnant neolithic survivor population subsisting on the island, but I think this is unlikely. Given it's closer proximity to the continent, I believe a population of Hallstatt or perhaps even later La Téne Celts (or both) arrived, and mixed with the LBA population, whose mixed descendants then took over the island with their superior Iron technology and chariots.  Gaius Julius Caesar comments in his 'De Bello Gallico' how certain Gallic tribes such as the 'Belgae' made frequent sorties across the channel to Britain.

During the period of the EIA after the collapse of the Bronze age trading networks which connected all of Europe and plunging temperatures resulted in a collapse of the Irish population from a BA high of around 100,000 to perhaps as low as 20,000 people by 300BC. It would not have been particularly difficult for a mannerbund of Iron sword wielding elites from abroad to take over and spread their language, culture and of course genes too.

Something like this is what I believe probably occurred in Ireland which led to a more fractured and tribal lifestyle were petty kingdoms waged war via launching surprise cattle raids on each other. And when modelling Irish using Irish_EBA and various Hallstatt and another attested and surprisingly northern Celtic sample from the Greek colony of Empuriés in Catalonia, this geneflow is very apparent.

In a blog post back in 2018 Davidski (see here) using qpAdmix modelled Iron age samples from Hinxton obtaining similar scores to the one shown above, though on Global 25 the models are not immune to being skewed by genetic drift which these Bronze Age 'Irish' naturally share with these Iron age samples from the continent. So for the most part we can be chalked up as 'Celticised Beakers' (Celts came from Bell beakers anyway) with some later introgression from Norwegian Vikings (I'll be doing a post on this later) and subsequent Anglo-Normans whom resemble the Irish autosomally anyway. 

Thankfully Davidski (PBUH) went through the trouble of revealing this genetic drift between Bronze, Iron and Modern era populations on a PCA chart.

PCA courtesy of Eurogenes blog.

What is apparent to me first and foremost from this is that Irish, Scottish and Welsh cluster closer to the BA cline than both IA Britons and modern English, whom are noticeably shifted towards the Belgian like Hallstatt samples from Bylany. This seems to verify what David Reich stated last year that an admixture event occurred somewhere in the vicinity of SE England and from there, Celtic culture spread out across the Isles and by 300BC La Téne artefacts start popping up in Ireland, probably from across the Irish sea via somewhere in North England. Interestingly a tribe called the 'Brigantes' make a sudden appearance in SE Ireland around this time. Identical in name to the one from northern England fleeing Roman expansion a few centuries later. These Brythonic tribes may have come as refugees and adopted the more archaic Gaelic language, which appears to represent a more basal layer of Celtic language that probably spread a few centuries earlier.

They may also have brought the famed R1b-M222 snp found predominatly in Ulster of which basal subclades are found in Britain, perhaps spreading with La Téne migrants?

But of course aren't we all immigrants afterall? (lol jk)

Tracing the neolithic migration from Anatolia to Ireland

By 10,000 BCE the neolithic revolution had already gone well underway amongst a population of mesolithic levantines known as the natufians. These previously forager people began to sow crops such as einkorn, emmerwheat, barley, rye, lentils and as a result this led to an increase in the effective population size which in turn led to the formulation of the Pre Pottery Neolithic culture (PPNA).

By 8,500 BCE this new way of living spread to Anatolia. But was it spread via migration?

The latest studies seem to say no surprisingly. Mesolithic Anatolian samples from Pinarbasi are typically little to no different from later neolithic Anatolian samples from sites such as Boncucklu and Barcin, give or take some very minor natufian and CHG/Iran_N introgression.

Barcin Hoyuk, Bursa c.6500BC
ANF that initially enter SE Europe resemble this population.

You can read more about this particular study here, it appears that for whatever reason farming developed amongst these Anatolians independent of any foreign influence, or via some sort of 'cultural diffusion' as the "pots not people" crowd would say. It seems this particular population of mesolithic/neolithic Anatolians evolved from a clade separate to that of natufians but rather of upper paleolithic hunter gatherers from the Caucasus mountains (Dzudzuana) according to the recent Lazaridis preprint. (See here)

Then roughly around 6,500 BCE these northwest Anatolian neolithic highlanders push west into the Aegean, introducing agriculture to Europe for the first time in prehistory, bringing their cattle, pots and thicc venus figurines with them.

From there, this migration (or invasion if you're an Anprim) diverges into two differing branches stemming from the Aegean launching point, with one arriving into central Europe around 5,500 BCE forming the LBK (Linearbandkeramik) horizon, and another directly across the Mediterranean forming the Cardial ware culture, also known as 'Impressed Ware' by archaeologists due to the impressions on the ceramic type utilised. 

By around 4,000 BCE these farmers finally reach Ireland, characterised by their series of megalithic henges dotted across the island the most famous of which being that of Newgrange in county Meath.

But from which migration path did this new EEF population arise from? 

Thankfully we have a high resolution genome available from Ballynahatty in county Antrim available to play around with in order to find out.

Reconstruction of Ballynahatty woman, her phenotype appears
to be that of Gracile Med type, common amongst EEF.

Reconstruction of a cardial EEF woman from Gibraltar

nicknamed Calpeia who lived around 5,500BC.

Her relatives would eventually repopulate Ireland

over a thousand years later.

To do this I will be modelling her using samples from cardial Iberia for her 'southern' admixture and LBK Germany sample for her central european EEF admixture.
And the results are indeed quite interesting.

Cardial agriculturalists absorbed WHG groups more or
less on contact during their expansion across the Mediterranean
as opposed to more insular LBK farmers who harboured only ~5% WHG
including older magdalenian foragers (El Miron) meaning
Irish neolithic farmers had some Aurignacian ancestry
from Europe’s first AMH population 40kya.

She does come out as mostly of cardial stock, which would confirm the findings of the 2015 Trinity college study that analysed her genome showing her to have shared more drift with EEF from Iberia and not central Europe. The closest contemporary population to her being Sardinians, she seemed to have had a relatively swarthy southern european complexion. Although it is plausible that a separate migration from central Europe made it into Ireland, as it is confirmed to have in Britain by the recent Brace et al study from earlier this year. (Yeah you can read about that one here too.)

Until we get more samples from the period we won't know. Although some from Carrowmore and Primrose in county Sligo have been published in this recent study, see here.

R1b-L21 Y-snp frequency by province

As most of you should probably know, the vast majority of Irish men (nearly all of them in some counties particularly in Connaught) will be downstream somewhere of R1b-M269, most likely belonging to the L21/M529 subclade. Once upon a time during the dark ages of population genomics it was believed that this particular haplogroup spread with post LGM hunter gatherers who migrated from the Franco-Cantabrian refuge during the pleistocine/holocene transition, known as the Bolling-Allerod warm period, which saw temperatures soaring before they plummeted again to LGM levels during a period known as the younger dryas. This theory was popularised for a few years by geneticists such as Stephen Oppenheimer and Brian Sykes whom came to the conclusion that the inhabitants of Britain and Ireland where predominantly the descendants of these people, an assertion now proven false.
It appears now thanks to improvements in technology within the field, that the source of R1b-L21 derived haplotypes and autosomal dna in Ireland can be attributed to the beaker people that arrived 4500ybp.

The subsequent founder effect resulted in the replacement of nearly all previous neolithic paternal lineages on the island (predominantly I2a2a1 see here) with the phylogenetic landscape remaining relatively unchanged over the next four millenia, no thanks to the clan system which ensured the dominance of this Indo-European line. Only the Basque have a higher proportion of haplogroup R1b.

As you can see later migrations of vikings,normans,gallowglasses etc left their mark, but these groups integrated relatively quickly and even amongst modern surnames with norman and viking origins R1b is still interestingly quite prevalent. 

During the Bronze age it seems that the most common terminal snp was probably R1b-DF21 found in Rathlin 1 (tall blonde fella they found round the back of a pub) which is most commonly found in men with surnames such as O'Carroll, Dooley, Bowes and Flanagan to name a few. Overtime this lineage was overtaken by later founder effects during the Iron age (most famously R1b-M222 associated with the O'Neill clan and found in over a million Irish men) due to various bottlenecks which saw the population shrink probably due to a cold spell at the time.

In no particular order, L21/M529 peaks in both Connaught and Munster at around 73% each, followed by 70% of men in Leinster and about 61% of men in Ulster.

So it seems like men in those two provinces were cucked harder over the years, for obvious reasons.

These numbers are based off of Family Tree DNA databases and the 2011 Busby et al study, which might be somewhat dated, however I believe these numbers to be fairly accurate.

More or less. 

Feasible Global 25 models

As promised, I will now be modelling contemporary Irish using a host of ancient samples which I believe reflect best the population genomic history of Ireland using the best admixture tools at my disposal.

For the first plausible model I will start off with the most basic threefold Steppe_EMBA/EEF/WHG using various proxies available currently on the Global 25 database. Nothing fancy.

For this first run I will be using yamnaya samples from samara kurgans, neolithic samples from Barcin Hoyuk, Turkey and a mesolithic sample from Loschbour, Belgium as a standin for WHG input.

 "distance%=5.5949"

 Irish:

Yamnaya_Samara,49.8

Barcin_N,37

WHG,13.2

Not bad as far as fits go, but the target distance is a tad too high. Around 2-3% is the optimum range we want to be reaching, but I am fairly happy with this result. However there are samples available on Global 25 to me where I can get a closer fit than this.

"distance%=4.5636"

Irish:

RUS_Poltavka,48.6

UKR_N_o,37

WHG,14.4

The results are near identical for this model, however the fit has drastically improved somewhat. Probably due to sharing more recent drift with these samples from poltavka (essentially descendants of the yamnaya who continued to live on the pontic steppe instead of migrating west), and the neolithic outlier sample from Tripolye, Ukraine.

For this next run I will be using more recent populations which should yield better results, this model will consist of Corded Ware samples from Germany, and EEF samples from Globular Amphora and Grooved Ware horizons respectively. The former from Poland and the latter from Ballynahatty, Ireland.

"distance%=3.8649"

     Irish:

Corded_Ware_DEU,65

POL_Globular_Amphora,35

IRL_MN,0

Unsurprisingly, modern Irish do not score anything from Ballynahatty, as her population was near totally replaced by incoming Rhenish beakers, who were themselves most likely of single grave origin, although somewhat more farmer shifted, perhaps harbouring some funnelbeaker input from Holland. Even when using these beakers as a proxy instead of CWC the same happens.

"distance%=2.934"

      Irish:

Bell_Beaker_NLD,79.8

POL_Globular_Amphora,20.2

IRL_MN,0

As you can see Irish have shown a great deal of continuity since the Bronze age, although there has been a pretty obvious bit of turnover since then from a southern shifted source.

Hmmmmmm I wonder what could've caused this? 

I think I may know the answer, which I will discuss in a later post.

Fáilte

This blog will be dedicated to observing the population genomics of Irish people, as well as focusing on other ancient dna samples from the island that are published in the latest peer reviewed studies as they arrive. I will be tracing the various demographic movements into the island that occurred during the various Mesolithic (WHG), Neolithic (EEF) and most importantly in the Irish case, the Bronze Age (Bell beakers) periods. I will be using the most up to date admixture tools to model both ancient and modern Irish samples. To do this I will be relying predominantly on the nMonte software created by huijbregts which relies on the Monte Carlo statistical method. Mainly using Global 25 coordinates courtesy of Davidski of the Eurogenes blog. (That man does some phenomenal work btw).

I will also be doing my utmost to dispel and debunk some of the romantic old wife’s tales that have been circulated regarding Irish origins in recent years. And of those there are many.


My blog will also be paying very close attention to ancient samples from the continent that are relevant in the ethnogenesis of whom we now know as the Irish.

Ancient Irish DNA - Rathlin 1