PDA

View Full Version : Scottish and Manx DNA



sktibo
09-05-2019, 04:29 AM
The recent study which included Scottish and Manx DNA references brought up some questions I am hoping to discuss.

Here's what the supplementary info had to say about the island's history:

"The earliest posited language was Manx Brythonic, from the Iron Age, but later Irish influence formed the modern Manx Gaelic language. Subsequently the island experienced Viking rule (indeed many place-names are of Norse derivation, interestingly these vary across the isle39), until sporadic English rule in the modern period. The genetic ancestry within Isle of Man is about half English, with just over a quarter Irish ancestry, reflecting the Gaelic and English histories. Indeed, four individuals with ancestry from the Isle of Man are instead grouped in the England cluster, supporting recent English ancestry within the Isle of Man. The Norse genetic influence on the isle is much smaller than that observed within the other regions formerly under the rule of the Kingdom of the Isles (the Hebrides) or the Jarldom of Orkney (Orkney, Shetland, Caithness, and Sutherland). This suggests that the Norse influence either was diluted by subsequent migrations and conquests, or the genetic impact was minimal as for example with an elite dominance model of rule, whereby numerically inferior newcomers exert control over the people, their customs and language, but have little demographic effect. A larger Manx sample in the future may help to better understand the impact of this complex history."

If we look at the Isle of Man cluster's admixture percentages (the model using Norway, England, and Wales) they have 42% English, which is way more than I would have thought this would be:

32926

The pattern we've seen so far is that the "Celtic" or the peoples who were native to the Isles held out in a genetic sense much more strongly than previously thought, but in the case of Man it looks like there was a large migration from England. Did the English immigrate to the Isle of Man in such numbers to alter the genetic situation this greatly? Could this be representative of something else? I think the population of Man that spoke Manx Gaelic was still fairly significant up until the end of the 19th century which I thought would have indicated a stronger hold out on the genetic situation as well, but I do not know much about the island's history.

09-05-2019, 09:02 AM
Hi Sktibo, I’m not familiar with this new study, but thinking back to the POBI study and Fine Scale where the populations of the Isles started to split to clusters, Is this not the best way to look at the Manx also? It could be they most resemble Cumbrians? Or Lancastrians?, maybe they were always like that from the original founders? Then with a topping of Gaels, and then Norse?

spruithean
09-05-2019, 01:30 PM
The recent study which included Scottish and Manx DNA references brought up some questions I am hoping to discuss.

Here's what the supplementary info had to say about the island's history:

"The earliest posited language was Manx Brythonic, from the Iron Age, but later Irish influence formed the modern Manx Gaelic language. Subsequently the island experienced Viking rule (indeed many place-names are of Norse derivation, interestingly these vary across the isle39), until sporadic English rule in the modern period. The genetic ancestry within Isle of Man is about half English, with just over a quarter Irish ancestry, reflecting the Gaelic and English histories. Indeed, four individuals with ancestry from the Isle of Man are instead grouped in the England cluster, supporting recent English ancestry within the Isle of Man. The Norse genetic influence on the isle is much smaller than that observed within the other regions formerly under the rule of the Kingdom of the Isles (the Hebrides) or the Jarldom of Orkney (Orkney, Shetland, Caithness, and Sutherland). This suggests that the Norse influence either was diluted by subsequent migrations and conquests, or the genetic impact was minimal as for example with an elite dominance model of rule, whereby numerically inferior newcomers exert control over the people, their customs and language, but have little demographic effect. A larger Manx sample in the future may help to better understand the impact of this complex history."

If we look at the Isle of Man cluster's admixture percentages (the model using Norway, England, and Wales) they have 42% English, which is way more than I would have thought this would be:

32926

The pattern we've seen so far is that the "Celtic" or the peoples who were native to the Isles held out in a genetic sense much more strongly than previously thought, but in the case of Man it looks like there was a large migration from England. Did the English immigrate to the Isle of Man in such numbers to alter the genetic situation this greatly? Could this be representative of something else? I think the population of Man that spoke Manx Gaelic was still fairly significant up until the end of the 19th century which I thought would have indicated a stronger hold out on the genetic situation as well, but I do not know much about the island's history.

I think it could be a combination of a few factors? I mean we know the Isle of Man had Brythonic settlement, followed by Gaelic, Norse and eventual English. Could it be that the common Brythonic and Norse (Danelaw + Norse settlement in NW England) ancestries of England are skewing (or over inflating) the "English-like" ancestry component for Man?

sktibo
09-05-2019, 05:23 PM
Hi Sktibo, I’m not familiar with this new study, but thinking back to the POBI study and Fine Scale where the populations of the Isles started to split to clusters, Is this not the best way to look at the Manx also? It could be they most resemble Cumbrians? Or Lancastrians?, maybe they were always like that from the original founders? Then with a topping of Gaels, and then Norse?

Perhaps, here is the link to the study: https://www.pnas.org/content/early/2019/08/27/1904761116?fbclid=IwAR2XINtXlv1K5DnrwdQNWXGeuQgeZU sg-m1QJKmRaqdarwj8dI3N1YpQRDM

I find it a bit odd because the "Celtic" source population is Welsh in the three population admixture analysis.


I think it could be a combination of a few factors? I mean we know the Isle of Man had Brythonic settlement, followed by Gaelic, Norse and eventual English. Could it be that the common Brythonic and Norse (Danelaw + Norse settlement in NW England) ancestries of England are skewing (or over inflating) the "English-like" ancestry component for Man?

Good point, perhaps the mixture of the original layer (So the Isle of Man was originally Brythonic speaking?) plus the Norse could resemble an English admixture is what you are saying?

spruithean
09-05-2019, 05:35 PM
Perhaps, here is the link to the study: https://www.pnas.org/content/early/2019/08/27/1904761116?fbclid=IwAR2XINtXlv1K5DnrwdQNWXGeuQgeZU sg-m1QJKmRaqdarwj8dI3N1YpQRDM

I find it a bit odd because the "Celtic" source population is Welsh in the three population admixture analysis.



Good point, perhaps the mixture of the original layer (So the Isle of Man was originally Brythonic speaking?) plus the Norse could resemble an English admixture is what you are saying?

Yeah, that's sort of what I was thinking. Especially given the geographic positioning and settlement activity in the regions.

sktibo
09-05-2019, 06:11 PM
I'm most interested in the data we now have from Perthshire, which of all the regions analyzed in the English-Welsh-Norse analysis received the highest percentage of English admixture. This is particularly surprising because much of Perthshire, including areas which were sampled, were Gaelic speaking relatively recently. IIRC the study collected data from people with four grandparents born in the area instead of eight, which I believe should take us back into the genetic situation of the 1800s? There is a paper which has a lot of data on Gaelic in Perthshire called "A Geography of Language: Gaelic-Speaking in Perthshire, 1698-1879."

32949

32946 The parish map of Perthshire

32947 The Gaelic and English speaking boundaries of Perthshire in 1873 and in 1879

32948 A chart with the percentages of English speakers in the parishes of Perthshire from 1755-56. Note at this time the parishes with the highest percentages of English speakers are still recorded to be below 50%.

On top of this, Gaelic in Perthshire may not have died out as early as previously thought as a native Gaelic speaker of the Perthshire dialect was recently discovered in Strath Tummel. I believe this area would be covered by the northern samples collected in the region called "Tayside" in the study.

So the study did collect samples from people who fall within the formerly Gaelic speaking areas of Perthshire, yet they do not differ genetically from those in lowland Perthshire or even those in Fife. I'm completely surprised that an area with relatively recent Gaelic speaking roots (For mainland Scotland) received the highest percentage of English-like ancestry in the three population analysis, and I'm hoping others will suggest some ideas as to why this may be. It seems to me that the fact that Gaelic survived for so long in these areas indicates that the English did not mass migrate into these parts of Perthshire.
I suppose that as we saw with the original POBI study which suggested that the English are not primarily Anglo Saxon genetically, again the linguistic situation does not reflect the genetic situation: the gaels of Perthshire must have been of different stock than the gaels of the west.

spruithean
09-05-2019, 06:32 PM
Hmm interesting question, again I think it's probably something to do with the general similarities between the population of England prior to the Anglo-Saxons, being perhaps somewhat similar to the Pictish population, but also that this region of Scotland (including Tayside-Fife and IIRC Aberdeen) experienced some Anglo-Saxon and Anglo-Norman (including Flemish) settlement in the time of king David. The People of Medieval Scotland project website has a fair number of important individuals from various transactions and deeds in the Perthshire, Tayside-Fife, Aberdeen areas who appear to have origins either in Southern Scotland, England or directly from Normandy (they have Anglo-Saxon or Norman names, or sometimes have mixed Gaelic-Germanic names).

Jessie
09-05-2019, 06:43 PM
I'm most interested in the data we now have from Perthshire, which of all the regions analyzed in the English-Welsh-Norse analysis received the highest percentage of English admixture. This is particularly surprising because much of Perthshire, including areas which were sampled, were Gaelic speaking relatively recently. IIRC the study collected data from people with four grandparents born in the area instead of eight, which I believe should take us back into the genetic situation of the 1800s? There is a paper which has a lot of data on Gaelic in Perthshire called "A Geography of Language: Gaelic-Speaking in Perthshire, 1698-1879."

32949

32946 The parish map of Perthshire

32947 The Gaelic and English speaking boundaries of Perthshire in 1873 and in 1879

32948 A chart with the percentages of English speakers in the parishes of Perthshire from 1755-56. Note at this time the parishes with the highest percentages of English speakers are still recorded to be below 50%.

On top of this, Gaelic in Perthshire may not have died out as early as previously thought as a native Gaelic speaker of the Perthshire dialect was recently discovered in Strath Tummel. I believe this area would be covered by the northern samples collected in the region called "Tayside" in the study.

So the study did collect samples from people who fall within the formerly Gaelic speaking areas of Perthshire, yet they do not differ genetically from those in lowland Perthshire or even those in Fife. I'm completely surprised that an area with relatively recent Gaelic speaking roots (For mainland Scotland) received the highest percentage of English-like ancestry in the three population analysis, and I'm hoping others will suggest some ideas as to why this may be. It seems to me that the fact that Gaelic survived for so long in these areas indicates that the English did not mass migrate into these parts of Perthshire.
I suppose that as we saw with the original POBI study which suggested that the English are not primarily Anglo Saxon genetically, again the linguistic situation does not reflect the genetic situation: the gaels of Perthshire must have been of different stock than the gaels of the west.

The more I think of this study the more I'm baffled why they used English, Welsh and Norse to study the admixture. I don't think it's accurate that so many of these populations have such high English admixture. The problem is that while Welsh are Celts like the Irish and Scots it doesn't mean that a Welsh population will stand in for ancient Gaels. If you look at the Traveller study Scots and Irish are closer to the English than they are to the Welsh and the Welsh are closer to the English than they are to Irish and Scots. So it might be more likely that in admixture Scots and Irish would end up looking more English than Welsh. I don't think it is good to use other populations of Britain and Ireland to look at the admixture question. The Irish DNA Atlas also used French as a proxy for Celts which is not correct either. The problem is that Bretons have a lot of British/Irish admixture so that similarity is due to migrations from Britain to Brittany. Anyway I really don't think you can assess ancient admixture by using modern populations.

English, Welsh, Scots and Irish are all similar populations so I don't understand why they use them working out admixture. They are all going to swamp each other's results. Too much over fitting.

sktibo
09-05-2019, 06:46 PM
Hmm interesting question, again I think it's probably something to do with the general similarities between the population of England prior to the Anglo-Saxons, being perhaps somewhat similar to the Pictish population, but also that this region of Scotland (including Tayside-Fife and IIRC Aberdeen) experienced some Anglo-Saxon and Anglo-Norman (including Flemish) settlement in the time of king David. The People of Medieval Scotland project website has a fair number of important individuals from various transactions and deeds in the Perthshire, Tayside-Fife, Aberdeen areas who appear to have origins either in Southern Scotland, England or directly from Normandy (they have Anglo-Saxon or Norman names, or sometimes have mixed Gaelic-Germanic names).

So the early Burgh system might have something to do with this.. good suggestion. My ancestors from Stirling / Perthshire (Who can mostly be traced to Perthshire but some have brick walls in Stirling) do have a lot of surnames which do not appear to be of Gaelic origin, such as Brown and Christie, even in the highland areas of Perthshire (in my case Killin.) I noticed when I was there in Perthshire, searching through graveyards, that names like these were quite common.

Returning to the conversation about the Isle of Man, I checked out the Manx Y-DNA webpage which estimates that 15% of the Manx population's Y-DNA belongs to I1, which I think suggests a higher Scandinavian genetic influence than the 5% estimate given in the recent study. I think this can be considered evidence for your suggestion as to the large English percentage in the Manx population.

sktibo
09-05-2019, 06:54 PM
The more I think of this study the more I'm baffled why they used English, Welsh and Norse to study the admixture. I don't think it's accurate that so many of these populations have such high English admixture. The problem is that while Welsh are Celts like the Irish and Scots it doesn't mean that a Welsh population will stand in for ancient Gaels. If you look at the Traveller study Scots and Irish are closer to the English than they are to the Welsh and the Welsh are closer to the English than they are to Irish and Scots. So it might be more likely that in admixture Scots and Irish would end up looking more English than Welsh. I don't think it is good to use other populations of Britain and Ireland to look at the admixture question. The Irish DNA Atlas also used French as a proxy for Celts which is not correct either. The problem is that Bretons have a lot of British/Irish admixture so that similarity is due to migrations from Britain to Brittany. Anyway I really don't think you can assess ancient admixture by using modern populations.

English, Welsh, Scots and Irish are all similar populations so I don't understand why they use them working out admixture. They are all going to swamp each other's results. Too much over fitting.

I think that the admixture charts using England Wales Scotland Denmark Sweden and Norway (B ) and the one which uses some Scottish, English, Welsh, and Norwegian samples as admixture sources (C) were poor choices and I agree with you on that. You make a lot of good points, and I don't think the Welsh will stand in perfectly for the Gaels - but - I think it's an improvement over trying to model these populations using the French as a source for the "Celts." Ultimately I think you are absolutely correct when you say that we cannot assess ancient admixture with modern populations. While I get carried away by these admixture charts, I suppose at the end of the day the spanning trees are a better way to look at these things. I really hope we will get more Iron Age or Early Medieval samples, but I thought it was very cool that it seems they do have some ancient Gaels from Iceland which were included. What were your thoughts on the assessment they did with that?

Jessie
09-06-2019, 01:57 AM
I think that the admixture charts using England Wales Scotland Denmark Sweden and Norway (B ) and the one which uses some Scottish, English, Welsh, and Norwegian samples as admixture sources (C) were poor choices and I agree with you on that. You make a lot of good points, and I don't think the Welsh will stand in perfectly for the Gaels - but - I think it's an improvement over trying to model these populations using the French as a source for the "Celts." Ultimately I think you are absolutely correct when you say that we cannot assess ancient admixture with modern populations. While I get carried away by these admixture charts, I suppose at the end of the day the spanning trees are a better way to look at these things. I really hope we will get more Iron Age or Early Medieval samples, but I thought it was very cool that it seems they do have some ancient Gaels from Iceland which were included. What were your thoughts on the assessment they did with that?

I think that is the way to go and using the ancient Gaels was a good start. It is also interesting that they found the most isolated areas more similar to these ancient Gaels i.e. Donegal, Hebrides and Argyll. Could this also be because there is more admixture in the other areas compared today than there was in the past? Donegal, Hebrides and Argyll would be more similar to the ancient Gaels.

sktibo
09-06-2019, 03:29 PM
Another point of interest is the data from the Fair Isle. Looking at the comparison to the ancient Icelanders (unless I'm reading these charts incorrectly) it seems that Shetland actually has less shared drift with the Norse samples than Orkney does. That caught my eye and I went back to looking at the data for Shetland of which the region that stands out most in the admixture charts is the Fair Isle - which in the chart in the OP has more "Welsh" admixture than the Outer Hebrides and less "Norse." Further, only one of the three Fair Isle samples actually show significant Norse admixture, the other two show very little. These Fair Isle samples still cluster with the Shetlandic samples despite there being a pretty significant difference. This got me wondering if the reason that Shetland and Orkney are so unique compared to Scotland and Ireland isn't actually because of their Norse admixture, perhaps even prior to the colonization of these islands by the Norse they would have been genetic outliers still. I often find that I want to group these things based on cultural or material evidence, Shetland and Orkney having evidence of Pictish culture prior to the coming of the Norse made me think that without this Norse influence they would cluster with the other peoples who shared that culture on the Scottish mainland. However, I now think that I must re-think ideas such as these, as the same line of thought led me to think that the Gaelic speaking areas of Perthshire would have had much more in common with western Scotland, which did not turn out to be the case.

Dubhthach
09-06-2019, 04:14 PM
The Isle of Man had significant migration from Northern Britain during the 19th century, particulary due to the growth of Tourism business at time. This is somewhat reflected in the accent in 'Manx English' which sounds like North of England accents. It's quite interesting listening to speaker of Manx language (which has been revived) speaking a Goidelic language with quite a 'north of England' type accent.

Older recordings of Manx speakers in comparison have a more standard accent (between Irish and Scottish).

Dubhthach
09-06-2019, 04:20 PM
It's worth remembering that their 'England' component as shown in that chart is obviously based on modern English population. In sense you could argue that's a bad approximation for Anglo-Saxon input, given that the English are clearly an admixed population where a large chunk (if not a majority) of their ancestry would have come from an Insular speaking population.

As a result it's possible that some of percentages showing up in other population might be reflective of shared Iron age ancestry. Recall after all the Welsh cluster are probably the most isolated of all clusters showing drift. The question is does this reflect the overall pre-AS population of Lowland Britain (eg. modern Britain) or was some of that drift already present at time of Roman conquest. A Better proxy would have been to used the ancient DNA from York and Hinxton to provide a 'pre-AS' baseline.

MacUalraig
09-06-2019, 07:54 PM
POMS isn't a particularly good guide to the peopling of Perthshire in general as their database is compiled from charters. But yes there were Norman families including our landlord Menzies of Weem.

spruithean
09-06-2019, 08:10 PM
POMS isn't a particularly good guide to the peopling of Perthshire in general as their database is compiled from charters. But yes there were Norman families including our landlord Menzies of Weem.

I wasn't saying it was, but that there is evidence of a presence, however as mentioned earlier I think this large "English-like" result in this paper for several regions is something older that populations in Britain and Ireland share.

sktibo
09-07-2019, 08:54 PM
I wasn't saying it was, but that there is evidence of a presence, however as mentioned earlier I think this large "English-like" result in this paper for several regions is something older that populations in Britain and Ireland share.

It is quite interesting that all the areas which are clearly representative of the Pictish parts of Scotland all have a significant "English" admixture on the three population chart, even "N Scotland" ...
[of which it isn't clear if this is referring to the samples around Inverness or Caithness, I suspect Caithness, because Inverness is used as a sample in the third chart. The main figure calls this "Inverness" while "N Scotland" represents Caithness, but the more detailed spanning tree in the appendix refers to the Inverness-area samples as "N Scotland" and the Caithness samples as "Highland." It would be nice if they didn't mix up their labeling like this.]
This could mean that the Picts had a significant enough difference in their genetic composition to the Gaels (perhaps they share a common lineage with the ancient peoples of the Anglo-Scottish borders?) but I also wonder about the possibility of the Gaelic leadership of the early Kingdom of Scotland bringing in many people from outside of Scotland to these Pictish areas (as discussed before, Flanders, England, France, ect.) in order to help deal with or suppress problems due to cultural differences.
What would have really helped answer this question would have been if they included admixture charts for the Sco-Ire region, and of course if they had used ancient DNA to model this instead of Wales. The Iron Age English samples were mentioned but it seems to me that from this study that the Gaelic Icelandic samples are genuine medieval Gaelic samples, which would have been a really good choice. Heck, they could have done a three way admixture chart using both the ancient Gaelic and Norse samples along with the Anglo-Saxon samples we also have. Why they didn't decide to do that is beyond me... An analysis like that probably would provide some real answers to many of the long and ongoing discussions on this forum.