For a modern European, how many centuries back before a decent %age of his....?

[Mr. Morden]

[Not sure if this belongs in Political Geography / Demographics or in Off-topic, but there’s a lot of ancestry talk here, so I’ll post it here, and if the mods want to move it, so be it.]

I’ve been involved in a lot of ancestry talk recently, and I had this one question that I haven’t seen an answer to anywhere.  I was curious as to whether or not anyone here had any thoughts on it, or knew where I might find the answer….

I’ve read how people say that every modern day European is a descendant of Charlemagne, because if you actually go back 1000 years plus, then you’re going through so many generations that the number of distinct direct ancestors you have from a single generation back then is at least in the tens of millions, such that almost every European who survived until adulthood and had children back in ~1000 will be a direct ancestor of every single European alive today (or similarly, a direct ancestor of everyone in the New World who has largely European ancestry).

But I can’t believe that the flow of ancestry would have any kind of sharp boundaries at Europe’s borders, and so I’ve got to assume that if you’re back to a time period when I (as someone with northwestern European ancestry) had tens of millions of ancestors alive at once, then at least a few % of those people would be outside the conventional borders of Europe, presumably mostly close to Europe, like North Africa, the Middle East, and Central Asia.  I would probably even have some in far-flung place like China.  It’s just that the number of lines of ancestry I’d have to, say, Charlemagne would be many times greater than the number of lines of ancestry I’d have to an ancestor in China.  And so my DNA would look overwhelmingly European, even though there would have been periods of time in the past when more than a few % of my distinct direct ancestors lived in either Asia or Africa.

But just how many centuries back do you have to go before a typical northwest European of today has more than a few % of his/her ancestors living outside of Europe?  And/or how many centuries back before he/she can find at least one common ancestor with a non-negligible number of modern day non-Europeans?  Anyone see anything like this question answered by a knowledgeable person on the web somewhere?

[kcguy]

I came across an unpublished academic paper years ago, in which the researchers used computer modelling to simulate ancestry.  On the Common Ancestors of All Living Humans

Quoting from Section 5.3 of the paper, "Figure 14 shows the corresponding ancestry for a randomly selected Norwegian. In this case, 92.3% of the ancestry in the year 5000BC is attributable to the 'country' in which the sim lives, in central Norway, and 96% to Scandinavia as a whole. . ."

[Mr. Morden]

I came across an unpublished academic paper years ago, in which the researchers used computer modelling to simulate ancestry.  On the Common Ancestors of All Living Humans

Quoting from Section 5.3 of the paper, "Figure 14 shows the corresponding ancestry for a randomly selected Norwegian. In this case, 92.3% of the ancestry in the year 5000BC is attributable to the 'country' in which the sim lives, in central Norway, and 96% to Scandinavia as a whole. . ."

That's interesting.  Thanks for that.  It sounds like that's talking about what %age of your DNA is traceable to those populations though.  I mean, most of the paper is about most recent common ancestor, but the particular stat that you're quoting above seems to be more "90%+ of a Norwegian's DNA comes from people living in Norway centuries ago".

In terms of the percentage of distinct ancestors you have though, that wouldn't be true.  1000 years ago, a modern day Norwegian would have no more than ~200,000 living ancestors in Norway (since that's about what the population of Norway was back then), but millions of ancestors living outside of Norway.  However, the Norwegians in 1000 AD would show up thousands of different times in the modern Norwegian guy's family tree because of pedigree collapse, whereas some guy living in Byzantium might just show up once.  Thus, the ancient Norwegians would contribute a highly disproportionate share of the modern Norwegian's DNA.

But what I'm wondering about is, if you forget about the %age of DNA contributed, and just count up the number of ancestors you have living at a given time, how far back do you have to go before the modern day Norwegian has a non-negligible number of ancestors living outside of Europe?

The paper you quoted flirts with that question in talking about the timing of "most recent common ancestor" between people in far flung regions of the Earth, but finding a time when 100% of today's population had a MRCA takes you farther back in time than what I was interested in.  OK, so to find a MRCA between every single European and every single Asian living today requires you to go back more than 2000 years at least.  But how long ago would you have to go for a typical modern European to have a MRCA with just, say, ~20% of people currently living in Asia?  Less than 1000 years?  That actually seems *possible*, but I have no idea if it's right.  I guess that info exists in the simulations discussed in that paper, but they didn't frame the question that way.

[kcguy]

I misunderstood your question the first time.  Instead of "ancestry", you meant "ancestors", as distinct people.  I didn't catch that the first time.

This time, I'm intentionally going off on my own tangent.

Years ago, before the birth of Prince William, I bought a book on the prince's ancestry.  As might be expected, there was a lot of pedigree collapse--it mentions that there are 27 lines of descent from Mary, Queen of Scots--but there were also some ancestors from much further afield.  (Of course, this doesn't really answer your question, since royalty are a bit less likely to marry someone from merely the next village over.)

Anyhow, here are some of the proposed lineages:
* The queen's grandmother, Queen Mary, whose grandmother was a Hungarian noblewoman, was descended from the half-brother of Vlad the Impaler, and through him she was descended from Genghis Khan.
* The queen is descended from a 14th-century count of Savoy, whose wife was from the Byzantine imperial family.
* The queen is descended from a tenth-century Duke of Hungary, whose wife was a Khazar princess and whose father had led the Magyars to Europe from the Asian steppes.
* The queen is descended, through medieval Crusader kings and kings of Armenia, from a second century king of Parthia.
* The author theorizes that one of the queen's ancestors, a tenth-century Spanish nobleman, was descended from the Caliphs of Cordoba, and through them from Abd Shams, the great-granduncle of Mohammed.
* The author also theorizes that another of her ancestors, an eighth-century French nobleman, was actually a Jewish "king" from Baghdad and a descendant of King David.
* The queen's mother was descended from a colonial Virginian, whose great-grandson was George Washington and whose sister was an ancestor of Robert E. Lee.  (OK, this one isn't as exotic, but I don't normally think of Europeans as having American ancestors.)

[Mr. Morden]

* The author theorizes that one of the queen's ancestors, a tenth-century Spanish nobleman, was descended from the Caliphs of Cordoba, and through them from Abd Shams, the great-granduncle of Mohammed.

I've read the claim that the Queen is a direct descendant of Mohammed himself, since some of Mohammed's descendants lived in Spain when it was the Caliphate of Cordoba, and you can purportedly connect from there to other European royal families.

But yeah, if it's really true that everyone with European ancestry alive today is a descendant of nearly every European who lived ~1000 years ago, then I guess that doesn't make her unique.  Almost every modern European would be a direct descendant of Mohammed.

[Yeahsayyeah]

The problem with "every living autchthon Central European is a descendant of Charlemagne is that it doesn't take social status in account. It is much more probable to have children with someone from the same or a compareable social status group. High noblemen have had some productive advantages in this regard. So even as the number of theoretically possible ancestors doubles with every generation, and at some point exceeds the number of the world population at this time, that doesn't mean that at every possible space there's a different ancestor.
The extreme would be a child of brother and sister which would cut the number of possible ancestors in half. But even if you have cousins as parents, which was quite common both in the nobility and for normal people, the child has only six grandparents instead of eight. If you do this over two generations it's possible it only has 8 great-grandparents instead of 16.

And if social and local mobility is not high, than there you have many, many generations in the same village or at least parish, that will lead to a huge ancestor implex, even if the canonical law of how close one may be related to marry, was adhered to. Of course, if one is related to Charlemagne in some generations all in this village or parish will be, e.g. if the local nobleman is a descendant of some high nobility bastards has an illegitimate child with his dairy maid (cliché, I know). Or as the last female descendants of the counts of Orlamünde married some wealthy confecitoner in a Thuringian town. So, as my wife has ancestors that are craftsmen in Thuringian towns, there is a chance...

In my known ancestry I have
Polish, German and Lithuanian peasants from East Prussia
evangelic exiles from Salzburg
Sorabian and German small peasants and home weavers from Upper Lusatia and the old markgraviate of Meissen
peasants from West Thuringia
German settlers and peasants from the Posen region

and my wife has
shepherds from West Thuringia, which given a name, that could be french (but there could also be other explanations) could be of French Huguenot origin.
peasants and workers all over the place from West Thuringia, Anhalt, the Halle region and middle Saxony
craftsmen from towns in the Vogtland, Thuringia and the Erzgebirge.

It just does not look like Charlemagne territory to me

[kcguy]

The problem with "every living autchthon Central European is a descendant of Charlemagne is that it doesn't take social status in account.

That's probably true.  Americans are used to thinking of English social mobility, but my impression is that Germany worked differently.

Most of England operated on the principle of primogeniture, so that the bulk of the family property went to the eldest son, and younger sons had to fend for themselves.  Similarly, titles of nobility followed the same principle, so that the eldest son of a baron would eventually become a baron himself, but the younger sons would remain commoners.  Because of this, it would become fairly easy for the youngest son of the youngest son of a English nobleman to fall into the middle class or lower.  (For example, Matthew Crawley on Downton Abbey.)

In contrast, my understanding of Germany was that property was much more likely to be divided among all sons.  I also believe that all sons of a German baron would be barons.  So I think the social barriers of marrying outside your class would be a bit more impenetrable than in England, where the social classes were a bit more informal.

At least that's my general impression, that it's much easier to be descended from the English upper classes than the German.

[Yeahsayyeah]

You must be logged in to read this quote.

This varies over time and space. There were primogenitur (oldest son inherits) and ultimogenitur (the youngest son inherits the property), equal division beween the sons and several modes unequal division in different spaces. Of course, in Germany east of the Elbe, there were also several forms of servitude. But even as primogenitur became common in the German nobility from the 15th century onwards, they still were a very distinct social group, where intermarriage with bourgeois was not very common  until the 19th century. Younger sons were often used for the military, (high catholic) clergy or high administration post and where economically secured this way, but stayed nobles and married nobles.

So for connections to the high nobility you would have to look for: children out of wedlock and morganatic marriages, intermarriages between nobles and (more or less) high bourgeois (but the nobilitation of high bourgeois is more common than the other way around), those who lost there status of nobility (which exists, but is rare).
. Or you would have to look at earlier times as the early and high middle ages seem to have been times of greater social mobility, but of course you don't find genealogical sources for that age (which is the basic problem for research, you have to find a connection to the high nobility at least in the fifteenth century to find a connection to Charlemagne.

Of course, my daughter has over 500 billion possible ways, where in one (or fiftythousand) you could find Charlemagne but it is not sure and the high number alone does not say too much about probability. There are, of course, other opinions, that say, that the high number alone makes it probable. But for instance, for my ancestral lines, it is probably cut in half alone because of my Slavic and Baltic ancestors, that most likely are not ancestors of Charlemagne.

[Mr. Morden]

The problem with "every living autchthon Central European is a descendant of Charlemagne is that it doesn't take social status in account. It is much more probable to have children with someone from the same or a compareable social status group. High noblemen have had some productive advantages in this regard. So even as the number of theoretically possible ancestors doubles with every generation, and at some point exceeds the number of the world population at this time, that doesn't mean that at every possible space there's a different ancestor.

The extreme would be a child of brother and sister which would cut the number of possible ancestors in half. But even if you have cousins as parents, which was quite common both in the nobility and for normal people, the child has only six grandparents instead of eight. If you do this over two generations it's possible it only has 8 great-grandparents instead of 16.

And if social and local mobility is not high, than there you have many, many generations in the same village or at least parish, that will lead to a huge ancestor implex, even if the canonical law of how close one may be related to marry, was adhered to.

There are two different, albeit related, issues here:

1) As you run the clock back in time, how fast does your ancestry spread geographically?

2) As you run the clock back in time, how often does your ancestry jump from one social class to another?

Both are interesting questions, but I personally find #1 more interesting, since I’m more curious about the geographic reach of ancestry than connection to nobility.

In my own family tree, I have a great grandfather who was born in Albringhausen, Westphalia, and his father (my great great grandfather) was born there as well.  But my great great grandfather’s wife (my gg grandmother) was born ~20 km away in Gerlingen.  I also have great great grandparents from Norway who moved quite a bit more than 20 km before marrying.  And then I have a great great grandfather born in Manchester, England whose parents were both born in Wales, and there’s also a great great grandmother who immigrated (with her husband) to the US from Vienna, Austria, but I’ve seen some hints that her family may have consisted of German-speakers living in Bohemia who migrated to Vienna in her childhood.

So from my own highly anecdotal sample, it looks like there was some geographic mobility in that era, though that era was the mid-1800s.  It seems plausible that geographic mobility would have been far less than that a few hundred years earlier, but I’m no expert on this.

Now, if it was the case that going back 1000 years, the typical child had one parent born in the same village as him/her, and another parent who had moved from, on average, ~20 km away, or heck, even ~10 km away, and if you’re assuming that your neighbors are also families where one parent has also moved from ~10 km away, then over the course of several centuries, it’s not hard to imagine that you’ve got ancestry spread across all of Europe.  Even if 10 km is the average, over many generations you’ll have a some big outliers in that average, so you end up with a big geographic spread.

But maybe 10 km is way too generous for the Middle Ages.  If it’s much less than that, then you get killed by pedigree collapse, as everyone ends up marrying their 5th cousins, and your ancestry’s geographic spread is too slow.

Btw, while most of my ancestors immigrated from Europe to North America in the latter half of the 1800s, a small number of them were living in the American colonies at the time of the Revolutionary War.  That’s as far back as any of the family stories go.  But when I plug the names and dates I have into the family trees I find online compiled by other people, I find that some of those Revolutionary War-era ancestors of mine connect back to the early Dutch settlers of New York and early English settlers of Massachusetts and Connecticut in the 1600s.  And from there, you can find people making alleged connections that do indeed trace back to Charlemagne via two independent lines (one via the English route and one via the Dutch route).  I have no idea if those online family trees are at all legit or not though.  It’s quite possible that one or both of those lines connecting me to Charlemagne are BS.