Endogamy and DNA
As we SPRING FORWARD today, let’s look at endogamy and you.
Yes, this is about DNA. Really. No, actually, the word endogamy isn’t Latin and it has nothing to do with the law. Well, then again, it may have been the law in some cases that caused the endogamy, so…
Endogamy affects DNA results
Here’s how it works. We all have two parents, four grandparents, eight great grandparents, 16 great great grandparents and so on — each generation doubling the number of ancestors. But pretty soon — as recently as the Middle Ages for us folks alive today — the number of ancestors in your family tree would be larger by far than the number of human beings alive on the earth at that moment.
The reason we don’t actually see this is that some of our ancestors married their cousins, so lots of individuals show up in our family trees more than once. Martin Davenport Sr. of Virginia, for example, is both my sixth and my seventh great grandfather. His grandson David Baker married a first cousin once removed, Martin’s great granddaughter Dorothy Wiseman, and that’s an example of endogamy. It was simply the custom to marry in the local community and that often meant cousins.
In some populations, however, endogamy was more than just custom: sometimes people were restricted in where they could live and who they could marry. And in those populations, because of those restrictions, the number of shared ancestors soared. It had to. You either married within the group, small as it might be, or you didn’t marry at all.
Now I knew all of this, intellectually, but it didn’t really hit me until I started managing a Family Finder project that includes kits for my uncle, my brother, and a large number of cousins ranging from first cousin to third-cousin-once-removed, in addition to my own kit. And the results in that project really show the DNA effects of endogamy.
For my uncle and many of my cousins, all 16 of their 2nd great grandparents were born in the United States and many, if not most, and sometimes all 32 of their 3rd great grandparents were born in America as well. And the further you go back towards colonial times, the smaller the available pool of marriage partners there was.
So for these members of the Family Finder project, there are on average 250-300 people in the autosomal DNA database at Family Tree DNA whose DNA matches theirs closely enough to be reported as a likely relative. On FTDNA, where there are 10 matches per results page, these folks are likely to get 25-30 pages of matches (a typical example is the first one in the chart on the left: 28 pages, 280 matches), and on average perhaps 50 will show up with a predicted relationship of fourth cousin or closer.
And a lot of those people show up as matches only because of endogamy. The chances of fifth cousins showing up as an autosomal DNA match is only about 10 percent, meaning only about one in 10 fifth cousins will have enough DNA in common with you to show up in the test.4 But if another person who’s also descended from Martin Davenport Sr. in multiple lines does the Family Finder test, we might be no more than seventh cousins but we’d likely have a better chance of more DNA in common than even fifth cousins would have… and perhaps enough that we’d be reported as a match.
And if we had lots of cousins marrying cousins, or simply neighbors marrying neighbors, we might show up as predicted third or fourth cousins when we were actually sixth or seventh cousins or even more distant than that. One test-taker in 2010 discovered that she and a match had four ancestral couples in common, were no more than seventh cousins once removed in any one line, and yet were predicted to be fourth cousins because of the amount of DNA they had in common.5
The situation is the reverse for another member of my project. All of his 2nd great grandparents (and, it appears, even all eight of his great grandparents, though there’s one we’re still chasing) were born in Europe. He has fewer than 40 FTDNA matches, and only three closer than distant cousin. (His match page results are second in the chart.) Part of that is because there are far fewer Europeans than Americans in the FTDNA database. And part of it is also because far fewer of his ancestors were geographically in a position to marry each other. So his results are the least affected by endogamy of us all.
My own results (third in the chart) are in between those two. My mother’s family has been in America as far back as I’ve been able to trace, but my father and his parents were 20th century immigrants from Germany. I have 163 matches as of March 2012. I have many fewer predicted third or fourth cousins than my all-American maternal uncle and cousins.
And then there’s my cousin Dick. Dick’s paternal ancestry is Ashkenazi Jewish. Throughout history, there have been fewer potential marriage partners for the Jewish population of Eastern Europe than for many other European population groups. Well beyond the usual tendency of people to marry within their own group, their own culture and their own religion, this group was often confined to ghettos and often restricted by law in who they could marry.6 Endogamy is a big issue in DNA testing for this population.
Dick’s match page count is on the bottom of the graphic to the left. Look at that. Eighty-four pages of results. 840 matches in all. Dozens upon dozens reported as predicted third or fourth cousins. And that’s after an adjustment by FTDNA to down-weight his matches to “prevent over predictions due to intermarriage and reflect more accurately (his) relationship to other Ashkenazi Jews” who’ve done Family Finder testing.
I know in the long run Dick will appreciate having a lot of fairly close genetic cousins. From a genealogical perspective it will be both a challenge and a joy.
But the simple fact is, the common ancestors his mother and mine shared had vastly more freedom in who they chose to marry than his father’s people could have dreamed of. And it’s absolutely chilling when you think of some of the reasons why that is so.
SOURCES
- Online Etymology Dictionary (www.etymonline.com : accessed 10 Mar 2012), entry for “endo”. ↩
- Ibid., entry for “gamete.” ↩
- Merriam-Webster Online Dictionary (http://www.merriam-webster.com/dictionary/ : accessed 10 Mar 2012), entry for “endogamy”.)
To some degree or another, we all have endogamy in our family trees. As genealogists, we usually call this pedigree collapse.[4. Wikipedia (http://www.wikipedia.com), “Pedigree Collapse,” rev. 13 Feb 2012. ↩
- “What is the probability that my relative and I share enough DNA for Family Finder to detect?”, Family Tree DNA (http://www.familytreedna.com : accessed 10 Mar 2012). ↩
- Roberta Estes, “Cousinship in Endogamous Populations,” Rootsweb, Genealogy-DNA mail list, posted 17 May 2010. ↩
- See e.g. the discussion of the Hapsburg “familial” laws in Heiko Haumann, A History of Eastern European Jews (Budapest : Central European University, 2002), 93-96. ↩
Thank you so much for this post – it explains a lot to me! The Family Finder test predicts a bunch of possible 4th cousin matches for me that I have been unable to verify through traditional genealogy. But since my father’s side of the family were all in SE Massachusetts and Rhode Island for several hundred years, it makes sense that I have distant cousins marrying distant cousins in my tree that I do not know about. I had wondered why I couldn’t verify any of the predicted relationships – endogamy may explain it some of the reason.
It’s the most likely reason, Wendy… and it sure plagues me in some of my North Carolina lines!
Very interesting. I am often frustrated by the supposed “matches” generated for my husband’s BALL line. His family lived in New Jersey and Pennsylvania, and there is no family tradition of anyone coming up from the south or moving south. Yet nearly all of the close DNA matches for him come from Virginia. Since Ball is such a common name, I still think it likely that this was due to closely related English Balls settling in different parts of the USA. But perhaps endogamy still explains it.
And remember: if there’s just one sister you didn’t know about who married and moved south, it could be the Ball line after all!
I came across this posting years later but it is just what I needed to read.
Endogamous groups that could show matches “closer than actual” are on both sides of my tree. Previously I had hesitated to test DNA due to costs beyond my hobby budget.
My mother’s peoples come from 2 tightly knit ethnic groupings. Her maternal side is French-Canadian & her paternal side is Breton from France…
So I’m currently going through hundreds of pages of images of parish records for possible family surname groupings… to get some paper trails so I can find some possible cousins to persuade to have DNA testing done…
My father’s side is Germans from Russia — who apparently pioneered in new countries every few generations & named nearly every male Johann & every female Elisabetha or Katharina!!! Here I hit brick walls in just 3-4 generations back from me.
Long time post but I’ll add a comment anyway. I also have some GFR ancestors. I had to laugh at the lack of creativity with names. Whatever records exist are in the possession of the AGHS. Genetic genology is a challenge because they were quite endogamous even after arriving in America. I match alsmot as strongly to third cousins on that side as I do my actual first cousin.
Much of the GFR
One grandmother’s parents: Germans from Russia; other grandmother: 273 years’ worth of Colonial Americans. No shortage of matches!
I’m thinking this may explain why my son has many DNA matches on both my side and his dad’s where he has more DNA with the match than we do – and his dad and I have zero matches. We both have a lot of Britain ancestry and very early US colonial roots.
No. Your son could not have inherited any DNA he did not get from you (in which case you should also match the person) or from his father (in which case he should also match the person). If your son has more DNA “in common with” the other person than either of his parents, some of it is being counted wrong — either too much to him or too little to the parent.
I do comprehend that “you can’t get more DNA than your parents give you”. But this seems to be a lot of errors then (because so far my son’s dad and I don’t match). Is this a common occurence, this many errors in testing? We tested on Ancestry and uploaded to Gedmatch. Not seen this with my mother, sister, cousins or others – not with this many! There are over 100 cases on my side where my son has more DNA than I do and so far over 50 on his father’s side (we just got his test results back) where he has more than his father. So far only a couple of cases where we share a DNA match in these cases, but those appear to be because we match on different sides of the match’s tree.
Different platforms, different analysis, there are too many possible issues to diagnose this. You might want to hire a DNA professional to do an analysis for you.
Back to the drawing board. Just when I thought I’d found my son’s Jewish grandfather, turns out it’s probably endogamy in play. A man matched my son at 124.8 cM but his son only matches my son at 17 cM. I thought that I had found my son’s first cousin twice removed which might have meant that he was a first cousin of my son’s grandfather. His son doesn’t match my son at 2nd cousins once removed so I’ve come to the conclusion that it is endogamy. The higher match came from an endogamous group of people in Poland but married outside of this group in America. Would that account for the apparent watering down of cM? Still trying to get my head around it all.
Your best bet for help understanding the endogamy issues in Jewish genealogy and DNA is the Facebook group, Tracing the Tribe – Jewish Genealogy on Facebook (https://www.facebook.com/groups/tracingthetribe/).
Interesting. I checked my Family Tree DNA account and I have 192 pages worth of matches. On Ancestry I had 120,000 matches until the recent culling update. Endogamy. Both my parents have families who go back to early Colonial Virginia with more than a few ancient planters including a bulk arriving up into the mid 17th century. Now I know why there are so many cousin matches.
Regarding my parents, it turns out they are distant cousins themselves sharing several common set of grandparents going back on average six or seven generations.
Anyways, having done my family tree, the level of cousin marriages are large and seem to be the result of some type of kinship group clustering. After getting to the point of doing research on great so and so aunts and uncles and seeing how these families were intermarrying across the board with little change in outlier individuals surname wise, it became clear that these marriages were very deliberate and local.
The kinship groups themselves being comprised of roughly five to ten families on average with a few outliers and largely being local to one another, not only geographically but also in regards to wealth and what little evidence of status through public office, militia rank or that of occupation.
I can trace these sort of kinship group structured relationships being somewhat intact for roughly 400 years (Surry, Isle of Wight and Southampton Counties) and in some cases through genealogical research, discovering that some of these same families were intermarrying in England before coming over to Virginia.
Now this makes better sense.
Thanks for the good article. I learned something new.
My mother was adopted in Columbus, Ohio in May 1907. She was given a Star of David pendant when she turned 12 and was told that her birth mother wanted her to know that she was Jewish. I have he adoption paper but have been unable to find a birth record for her. Why does my Ancestry DNA ethnicity result not show that I am Jewish?
It is possible for someone to be Jewish by religion and not Jewish by ethnicity. It’s not common, but it is possible.