My Ancestry

Inspired by Karin Bojs's and Peter Sjölund's recent book Svenskarna och deras fäder, I've looked into my ancestry by means both genetic and genealogical. Here's a few highlights.

  • Like most Stockholmers, I'm of mixed rural Swedish stock. My great grandpa's generation contains 16 people born mainly in the 1880s. Only one of them was born in Stockholm. His parents were born in Värmland and Södermanland provinces. The other 15 were born all over rural southern Sweden: Bohuslän (two people), Småland (two people), Södermanland, Skåne and Närke. They went to Stockholm to find work, met and got married.
  • My Y chromosome is type R1b-M269, which is the second-most common one in Sweden and the most common one in Western Europe. My closest modern matches form dense clusters in England and New England. There's clearly an Englishman in my recent pedigree, most likely in the 15th or 16th centuries judging from a combination of genetic statistics and genealogy. In the mid-1600s my paternal line was already in Värmland with Swedish names.
  • My mitochondrial DNA is the very common type H with my closest modern matches clustering in Finland. This means that my maternal line points east to a very great grandma in West Asia about 25,000 years ago. Of Europe's three original major population components, this would represent the Ancient North Eurasians.
  • I found the first Rundkvist! In the 1800s a lot of rural Swedes quit using the patronymic and took family names instead. My grandpa's grandpa Johan Jansson (1853-1925) took the name Rundkvist and moved to Stockholm from Fryksdalen in Värmland. His brother Magnus Jansson instead chose Söderqvist for some reason.
  • Update 14 March: Aard regular Lassi pointed out something enlightening. Parts of modern Sweden saw state-sponsored immigration from Finland in the decades around 1600. This is the simplest explanation for why I have a Finnish maternal line. Its earliest member known to me, Helena Helgesdotter, was born near Gothenburg in 1775.

More like this

Couldn't quite catch a word in an old Blur song. Turned out to be "jumbojet" with the stress placed on the wrong syllable. JumBOjet. Grötrimslyriker, as we say in Swedish. The Swedish Anti-Theft Association offers tags for keyrings. You put the tag on your keyring and pay an annual fee, and then if…
I don't read much in Swedish. On a whim I decided to check what recent Swedish books I've read and liked outside work. Turns out they're all popular history. Alla rekommenderas varmt för den som delar mina intressen! Kring Hammarby sjö. 1. Tiden före Hammarbyleden. Hans Björkman 2016. Local history…
This past weekend saw my seventh annual boardgaming retreat: 43 hours in good company at a small hotel (in Nynäshamn for the first time), all meals included. My buddy Oscar organises everything. This year we broke the attendance record, with 28 participants, mainly guys in our 30s and 40s. Before…
My personal genealogy has never interested me much, knowing as I do that the number of ancestors multiplies by a factor of two with each generation. Thus in AD 1800 someone born in 1975 had about 2^8=256 ancestors of child-bearing age (or slightly fewer if someone has been productive in more than…

One of these days I should get my mother to write down a bunch of family history that she has told me and/or my sister. My mother is the last surviving member of her generation, mostly because she was the youngest (several of her cousins were a couple of decades or so older than she; she was the youngest of seven children, born to a 38 year old mother).

Some of my father's cousins are still alive, including one who lives about an hour drive from me (coincidentally; my father's side of the family lived mostly in northeastern Colorado, about 3000 km west of where I live now).

By Eric Lund (not verified) on 12 Mar 2017 #permalink

Record an interview with her about the family! I've been wondering idly about the first Rundkvist for decades, and now it turns out that I used to meet his daughter once or twice a year when I was a kid. Much family lore was lost because we never thought to record what she knew.

Useful resource, but it carries the risk of being drowned in a flood of information:

On uniparental DNA, beware sample bias in the sample of genomes tested by whichever service you used. It can be misleading. Autosomal DNA, which makes up by far the most part of your total genome, is a better bet to get a read on your total ancestry composition. If indeed you had a British ancestor, it should show up in your total ancestry composition as distinct from Scandinavian, although if it was that far back and only one individual, it will be a small % of the total. Full genome sampling and testing costs more, of course, but just uniparental is not too informative, except for general interest.

By Aspidistra (not verified) on 13 Mar 2017 #permalink

I've done Y chromosome and mitochondrial typing. Now I'm waiting for the general SNP scan that Family Tree DNA calls "Family Finder".

I have just read the book "Germans" which clears up a lot of misconceptions about the peoples with that particular language group.
BTW are there any particular genetic markers that are more common for the group that would later be known as "celts" ?

-- -- -- -- -- -- -- -- --
John, regarding the difficulty of interesting people in reading up on facts.

In my experience, it helps to start with a simplified summary in the first one or two paragraphs, because 99% of readers will just skim through the text.

So if you -for instance- are publishing a cure for cancer, readers will ignore it unless the first paragraph reads WE HAVE A FUCKING CURE FOR CANCER. It is no point being subtle when confronted with stressed-out readers.

And, yes, I know that some discoveries are so inherently complex that it is hard to boil them down to a one-paragraph summary up front. that is why it is so rare to find good science jounalists, it is a bit of a black art.

By BirgerJohansson (not verified) on 14 Mar 2017 #permalink

We could be related :-)

In the 17th century the King of Sweden promised tax breaks to anyone who moved to various uninhabited parts of the kingdom. From Savonia in eastern Finland some people went to Karelia (both were parts of Sweden at the time) and become my ancestors, others moved to Värmland. Maybe your maternal Finnish ancestor was one of them.

BTW, Finnish was spoken in Värmland up to 1950s. They even sung poems that were parts of the Kalevala cycle.

Maybe some day I'll have my DNA analysed...

By Lassi Hippeläinen (not verified) on 14 Mar 2017 #permalink

Excellent point, Lassi. The slash-and-burn Finns are of course the easiest way to explain why my Finnish mitochondria are near Gothenburg in 1775.

We could be related

Go back 40 generations--about a thousand years. In theory you should have 2^40 ancestors in that generation, which is a bit over a trillion. That's 3-4 orders of magnitude more people than were alive in the world at the time, so obviously there is considerable duplication in your family tree. And given that some of my ancestors were Swedish, it's likely that I share some of those ancestors.

By Eric Lund (not verified) on 14 Mar 2017 #permalink

I remember when it was big news that Ronald Reagan and Francois Mitterand had a common ancestor in the 9th century...

By Lassi Hippeläinen (not verified) on 14 Mar 2017 #permalink

Martin@4 - That's it - the 'general SNP scan' will give you the autosomal DNA, which they should interpret for you in terms of ancestry in relation to various reference populations.

When you get that, if you can, you should download your own raw data file, if indeed they don't give that to you automatically. After all, they are your data, and there are other interesting things you can do with them, using kosher software that is freely available (like medical information, if indeed you want that - my own experience with that is that the medical information is mostly not too helpful because they only sample about 2,400 SNPS. Typing a full genome is still way too expensive for most customers of consumer genomics).

Birger@5 - re. Celts, no, not unique individual uniparental markers (Y DNA and mtDNA). Remember me babbling a while back about the 'great homogenisation'? They have to use autosomal DNA and compare with reference populations, using statistics. The Celts are not recognised as a distinct identifiable culture until the Iron Age; the great homogenisation was over some time during the Bronze Age.

I'm assume I'm not babbling to a bunch of average punters here - at least you, Martin and Eric have fully functional brains (plus I am keen to see the intersection and synthesis of genetic and archaeological information - most linguists seem just not interested), and I presume (perhaps wrongly) that you have at least some prior relevant knowledge. I have no interest in trying to enlighten the general readership here, let along anywhere else. When I try to talk to intelligent friends who have not taken any interest in developments in modern genetics, their eyes just glaze over.

By Aspidistra (not verified) on 14 Mar 2017 #permalink

Birger@4 - Further to my verbose babbling above, to try to clarify - please appreciate that it is difficult to be succinct about a branch of science that has absolutely exploded since the 1990s, but about which most people seem to lack even a fairly fundamental understanding, so let me start by using this as a hook: CELTS WERE NOT FUCKING UNIQUE. In fact, I have read that they were a culture, rather than a 'race', which is probably as good a way to put it as any.

The most dominant Y DNA (i.e. male line) haplogroup in Western Europe is R1b (but only second most dominant in Sweden). Martin belongs to R1b - M269, which is the most common haplogroup subgroup in Western Europe. I belong to R1b - L21. Among ancient remains identified archaeologically as 'Celtic', R1b is highly represented, as it is in all European groups dating to the Iron Age.

That includes the Basques, who are notable for speaking a non-Indo-European language. People were misled by that for a long time and thought that the Basques must be a remnant of some very old population in Europe, but they are not - somehow, the Basque language remained intact, but the population changed with time to look like pretty much most other populations in Western Europe. The stand-out group among Western European populations are the Sardinians, who lack the Ancient North Eurasian component which is present in all other European populations to some extent, thought due to geographic isolation during the period around the end of the Neolithic/beginning of the Bronze Age when pastoralists from the Pontic-Caspian steppe entered Europe in a wave of migrations. But, just to be contrary and confuse everyone, although the Sardinians are genetically distinctive and lacking ANE ancestry, they speak an Indo-European language, the language group that is associated with the steppe migrations, which are also associated with a substantial component of ANE ancestry.

So, Genetics 101 - languages do not map particularly well to populations in terms of genetics, as the Basques and Sardinians demonstrate. The Celts had a distinctive material culture and languages, but they are not notably distinctive genetically - certainly not like the Sardinians.

Ötzi the Iceman matches most closely to Sardinians, i.e. he had no Ancient North Eurasian ancestry, which places him among the Neolithic wave of migration into Europe from the Near East. He had among the possessions he carried a copper-headed axe, which is yet further confirmation that at least some Neolithic groups in Europe had copper, before the migrations of the steppe peoples.

Ancient remains identified archaeologically as Celtic cannot be identified as such by a unique set of genetic markers, just as the Basques cannot be. They can only be identified genetically by the relative composition of the group as a whole. Of course, the remains of an ancient Celtic individual can be identified wrt to Y DNA, and that male line can be tracked down to modern times. Apparently 12% of modern Irishmen can trace their male lineage to someone called Niall of the Nine Hostages, whoever the fuck he was (I have no idea, and no wish to find out), to the extent that it has become something of a joke on the Internet discussion boards - if you're Irish, the genomics companies will tell you that you are descended from this Geezer, along with seemingly everyone else in Ireland.

In the British Isles and Ireland, on a population basis, modern Irish, Scots and English cannot be differentiated (much to the chagrin of some people). However, if a single person is genotyped, then by reference to ancient population groups, that individual's ancestry can be broken down so that he can be identified as so many % Anglo-Saxon, so many % Celtic, so many % Danish, etc. because reference population sets have been established for those ancient groups, now that enough ancient remains have been genotyped who are identifiable by archaeology, e.g. associated with grave goods or whatever. So, people in East Anglia have a relatively high proportion of Anglo-Saxon ancestry; people in the north of England might have a higher proportion of Danish ancestry, people in the far south of England who might have a higher proportion of Celtic, etc. But none of them is 'pure', they are all mixtures.

So, the typical English geezer who proudly pronounces that he is 'descended from a Viking' is just being a delusional dick telling self-aggrandising stories in the pub to anyone not too bored to listen - on a whole genome basis, he is no doubt just plain white modern English bread like everyone else. My favourite story is from a Singaporean Chinese woman I know who lived in England for a while - when people used to start that sort of one-upmanship in the pub, she used to say "I'm descended from a shipping clerk and a bus conductor." In fact, her personal ancestry is actually more interesting than most of them - her ancestors come from Hainan Island, which is a whole other long-winded subject.

It's complicated. If you sample from the Eurogenes Blog, run by the Polish blogger who calls himself Davidski, you will quickly see just how complex the whole field has become.

Lassi@7 - Go for it. Endless hours of fun.

When Martin gets his ancestry composition it will predictably set him off on another journey of discovery, depending what his results show; especially, of course, if he comes out as something less than, say, 95% Scandinavian. I have some revealed ancestry for which I have absolutely no explanation at all - it is totally absent from the family genealogical myths that I was told by all of the older relatives who are now dead. They would have been as mystified as I am.

My wife finally opted to spit into a sample tube a month ago, so we are waiting to get her results. She is getting impatient to know. My anticipation is tempered by the knowledge that the American company we used does not have a really good reference population for East Asians, aside from Asian Americans, who are dominated by Southern Han (they have sampled over one million individuals, but their total sample is overwhelmed by Irish and North Americans of Irish ancestry, who all seem to be mad keen genealogists). But of course they have access to the global GWAS database, like all of the genomics companies. So their analysis of ancestry is improving all the time. Research on the origins of East Asians has lagged badly, as has genomic research in Africa, but it is now in the process of catching up.

The basic message I have been trying to push on this Blog for some time is that all of this delving into ancient genomes from the remains of anatomically modern humans in Europe NEEDS THE FUCKING ARCHAEOLOGY TO ANCHOR IT. The two disciplines are complementary and potentially mutually enlightening and confirmatory. Actually three, if the linguists were willing to engage, but most of them appear not to be - they seem to fear that their discipline will be 'contaminated' if they allow it to be informed by these other disciplines. I suspect they are just pissed off that it is the geneticists and archaeologists who together have settled the question of where the Proto-Indo-European Urheimat was, while they have argued endlessly and fruitlessly about it among themselves. Those linguists who ignore the mass of evidence now available are the ones who seem to continue to get it wrong.

If you think this is not a succinct attempt at a summary, try reading Davidski some time. There is no short cut for stressed readers into the field of modern genomics, any more than there is any short cut into the field of, say, Scandinavian archaeology - you need to read the stuff, even at the popular science level.

By Aspidistra (not verified) on 15 Mar 2017 #permalink

Celts "were a culture, rather than a ‘race’".

An archaeological culture is a list of types of object and structure that occur reliably together: all material stuff. The Hallstatt and La Tène archaeological cultures were probably largely associated with Celtic languages.

But we're pretty sure that there were groups who spoke other languages yet used variants of the material culture. And conversely, not every single Celtic-speaking group is likely to have used the material culture. Further, we know from social anthropology that tribal societies like these encourage multilingualism. Are you a Celt if Etruscan is spoken regularly in your household?

So the only really undisputed way to use the word "Celtic" is in reference to language. Which does not survive much from the period in question.

I was thinking Anglo-centrically, where people totally conflate Celtic material culture, languages and genetic populations (and notably art among the New Age crowd), which I think they can just about get away with (please correct me if I'm wrong).

I don't even want to think about 'Celtic' material culture and languages in Europe - that just gets far too confusing for me, having being taught in an Anglo-centric tradition from a very young age that Celts were an identifiable population grouping who spoke an identifiable group of languages.

By Aspidistra (not verified) on 15 Mar 2017 #permalink

Yeah, in the British Isles there is to my knowledge no sign of any pre-Celtic language stratum, so that's what people spoke there until the Roman invasion.

There's a recent school of thought that argues that the language group originated along the Atlantic coasts of Europe and then expanded east into Central Europe, rather than the reverse which is the established idea.

But if there were a pre-Celtic language in the British Isles, how would we know? Neither group had writing at the time, so if the invaders thoroughly defeated the previous indigenous group, the latter group may not have left any trace.

Of course one clue might be geographical names. In the US and Canada, many rivers, lakes, and even cities and states/provinces have indigenous names. Likewise in Japan, as you go northeast you are increasingly likely to encounter geographical names which are of non-Japanese origin (presumably derived from the ancestral language of the Ainu). It helps to have some kind of gradient or abrupt transition so that you can identify this situation.

In the case of the Celts, this signal is further attenuated by the subsequent Anglo-Saxon and Viking invasions of the British Isles. Certainly the Celtic influence in place names is visible: multiple rivers named Avon, which was an indigenous word for river, and the possibly apocryphal Torpenhow Hill ("Hill-hill-hill Hill").

For many of the indigenous peoples of North America, particularly in the eastern part of the continent, place names are the only trace of the indigenous language. Contemporary accounts by European settlers note cases where as much as 90% of the indigenous population died in disease outbreaks. In some cases enough native speakers remained to keep the language alive despite the efforts to force children to attend "Indian schools" where they were forced to use only English. At least two, Cherokee and Inuit, acquired their own writing systems (the former is obsolete, but you may see the latter in parts of Arctic Canada), though most of the remaining indigenous American languages have adapted the Latin alphabet to their languages, similar to how Portuguese missionaries created a Latin-derived alphabet for Vietnamese.

By Eric Lund (not verified) on 15 Mar 2017 #permalink

Birger@5: Still trying to find a rational response to your question, you can refer to Eupedia here:

To quote: "Genetic studies determined that most of the ancient Celtic men belonged to the Y-DNA haplogroup R1b-S116 and its subclades. Two Early Bronze Age migrations brought the L21 subclade to north-west France and the British Isles, and the DF27 subclade to south-west France and Iberia. The third major Celtic subclade is S28 (aka U152), which is associated with the expansion of the Hallstatt and La Tène Celts, as well as with Italic tribes."

But the writer(s) don't explain how they identified those 'ancient Celtic men' as being Celtic, or what they mean by Celtic, and they don't identify where the 'two early Bronze Age migrations' came from - maybe we can infer from the steppe, given that Celtic languages were Indo-European languages; they appear to conflate material culture with language, and it all pretty much dissolves into hyperbole after that.

By Aspidistra (not verified) on 16 Mar 2017 #permalink

No, that was a ridiculous thing to say - they can't have been steppe migrations.

If the Atlantic coast theory is right, they would be coming from the other direction, anyway.

I assume 'they' (whoever they are) are assigning 'Celtic' to these subclades based on an association with material culture, which is interpreted to be 'same' as Hallstata and La Tène. But as 'Celtic' refers to a language group, at least in modern parlance, that's dodgy.

It hasn't escaped my eagle eye that they include the R1b-L21 subclade in this, which is my own. Well, north-west France could work.

By Aspidistra (not verified) on 16 Mar 2017 #permalink

We also know that the Celts didn't call themselves that. As I understand it, the term is derived from the Greek κελτο&iota (keltoi), which means "barbarian". Which implies that at the time there were Celtic people living in what is now the Balkans.

There is a portion of northwest France called Brittany. It was one of the last places in continental Europe where Celtic languages were spoken, along with Galicia in northwest Spain.

By Eric Lund (not verified) on 16 Mar 2017 #permalink