Paternity most assured

By Razib Khan | April 23, 2012 12:32 am

The myth that 10 percent of children the product of ‘non-paternity events’ is rather persistent. I have no idea why, but I do know that even biologists accept it. But how we can we continue to accept this when surnames can provide population genetic information 400 years after the fact? The population of Belgium is famously divided between Latinate Walloons and Germanic Flemings. But is notable that a substantial number of Flemings carry surnames of clear Romance origin. This is in large part due to acculturation. Nevertheless, even 400 years after the largest of the migration and assimilation events males with Romance-origin surnames reflect their genetic background:


If non-paternity events occurred at a rate of 1 out of 10 the correlation between surnames and genetic lineage would have been decoupled long ago. These results have been confirmed in other societies. I predict that low non-paternity rates will also be confirmed in China; as that nation has a long history of surnames. Of course, one might posit a scenario where males who are the products of non-paternity events tend to be less fit than those who are not, so over the long term these estimates based on present day Y chromosomal lineages may not be appropriate reflective of the frequency of events at some point in the past.

MORE ABOUT: Non-paternity

Comments (22)

  1. But what if the cuckolders were disproportionately likely to be from populations with the same Y chromosome as the supposed fathers (i.e. other immigrants from the same place, or men from ‘the old country’ of origin.)?

  2. Sandgroper

    My daughter was absolutely delighted to discover recently that the originator of her mother’s Chinese surname about 2,500 years ago was a bastard, and that his father was a murderous adulterer. The bastard adopted a new name to disassociate himself from his father. People of the same name have lived in the same part of China ever since.

    Her mother was somewhat less delighted. All this time she’s been telling me she’s descended from a great philosopher, and it turns out she’s really descended from a murderous adulterer and a bastard LOL!

    But then that means so was the great philosopher.

  3. April Brown

    @ #1 – yeah – I wonder if it wouldn’t be useful, for discussions of genetic trends, to have a unique term for non-paternity events where the biological father is an uncle or a cousin of the legal father. No idea what the frequency would be on events like that, but they might not show up in genetic studies because the kid would have the expected set of grandparents, etc.

  4. Dm

    For a link between a _set_ of surnames attributed to a common subculture, it isn’t even necessary to postulate a higher rate of kin-nonpaternity in the mold of #1, #3. Even if nonpaternity events happen outside of the family group, but inside a broader language / culture community, or simply locally withing the groups sharing the same neighborhoods, churches, or festivals, then we may still observe a remarkable conservation of Y-chromosome composition after many generations no matter the imperfect true paternity.

  5. The genetic data is clouded by the fact that adoption would have the same effect as cuckolding, but considered very different culturally.

    Apparently adoption of completely unrelated children is/was rare in Belgium though, or this dataset would be different.

  6. Karl Zimmerman

    Adoption was very rare in the West until the modern period. Most foundlings were given to the Church, and could either remain a Brother or Sister (and hence have their official line end), or be apprenticed if male to a craftsmen. However, the latter was an economic, not an emotive, affair, and I see no reason to assume most would adopt their master’s surname.

  7. Miley Cyrax


    Murderer and adulterer from her mom’s side, sandgroper from her dad’s side ; )


    That was my thought as well.

  8. in many societies adoption was generally of relatives (e.g., caesar’s adoption of his grand-nephew). in any case, please note that a lot of the paternity stuff on extant populations involves STRs, not Y chromosomes.

  9. Ed

    Or, what if you have a racially ambiguous name like ‘Lee’ or ‘Gee’ which can be an English or Chinese last name? For instance a lot of people assume that my white/chinese friend with the last name Lee has a white father because wm/af couples are more common, when in fact his father is Chinese (among other Asian ethnicities).

    I always thought it was amusing that he was a white/asian guy with a name that could be either white or Asian.

  10. Euler

    #1 and 4
    This still wouldn’t explain how very rare surnames have even higher correlations with genetics, as mentioned in the abstract in the last link.

  11. Winston Churchill’s younger brother wasn’t very much like him at all, leading biographer William Manchester to assume he was only a half-brother. But this kind of behavior seemed more common amidst the upper reaches of society than among the middle classes.

  12. “If non-paternity events occurred at a rate of 1 out of 10 the correlation between surnames and genetic lineage would have been decoupled long ago.”

    To elaborate on Neuroskeptic’s point at #1, #3 and #4, only if the non-paternity events involve people with distinguishable Y-DNA haplotypes. If cryptic non-paternity events (i.e. unknown ones) involve people who are ethnically identical, one can discern non-paternity rates from these data. And, fine scale social structure within which extra-marital affairs happen, and the fact that a woman who has an affair with a genetically similar “third man” is much less likely to be caught when a child is born than a woman who has an affair with a genetically dissimilar one, both favor just this kind of non-randomness in “non-paternity events.”

    For example, if you have an affair with your husband’s full brother, nobody is likely to be able to tell after the baby is born that he isn’t your husband’s. But, if all of you and your husband’s blood relatives are blond, and you have a black haired child, people might start suspecting that you have engaged in infidelity. And, of course, the fact that you married your husband sheds light on what “your type” of man is.

    Also, I suspect that a model in which most non-paternity events happen at a constant rate over four hundred years is probably flawed. I’d expect that a punctuated, community wide surge model might make more sense. For example, maybe a village is subject to three rape and pillage Viking raids (or Hundred Year’s War battles) over a century that lead to a lot of non-paternity events and a lot of dead children. One might imagine tolerance for those events being greater than their frequency being greater than at other times, allowing introgression into the pool of haplotypes found in people with a given last name, making the introgressed genetic markers look like part of the norm for that last name. But, in times of peace, stability and relative prosperity, non-paternity events might be much more rare. This would look like a low rate for the entire period statistically.

    Also, given the frequency of premature death of parents in a society where the life expectency is about forty years, I would suspect that a great many children would end up living with their stepparents and it isn’t clear what last name would be used for a person like that in the era from ca. 1600 to ca. 1870 when surnaming practices weren’t so rigidly implemented bureaucratically. The social history of the working and lower classes is less comprehensively and accurately recorded than political history and the lives of the elites. This would cut the other way to some extent (tending to exaggerate the number of non-paternity events), but the point is that there are a lot of moving parts that make a determination of non-paternity rates from this kind of data conjectural.

  13. Sandgroper

    @7 – Alas, I am at pains to conceal the shameful fact that my paternal grandfather and greatgrandfather were both Banana Benders.

  14. Roger Bigod

    The Sykes study is pretty good. The sample was of males with that surname from a rural part of Yorkshire. The estimated non-paternity rate was 1.3% per generation over 700 years. This group may not be typical, but there’s no obvious difference from the conditions of most of humanity for the last few millenia.

    Another data point is blood groups (antigens on human red cells). These were mostly worked out in the mid-20th Cent by British blood bankers. They would collect a pedigree, test for presence of the alleles of interest and demonstrate Mendelian inheritance. As a screen, they would run all the previously known blood groups and look for violations of the Mendelian pattern. The lore among blood bankers was that non-paternity showed up at a rate of 2-3%.

    Adoption was much more frequent in previous centuries because of parental mortality. People didn’t record it or get recognition by a court.

  15. Social collapse.

    Now non paternity events are frequent.

    In JD Unwin’s survey, societies with high rates of non paternity events tend to disappear from history. He has a freudian theory to explain this, that easy sex drains energy.

    My explanation of his observation is that paternal uncertainty leads to low investment in posterity and children, resulting in conquest by societies with higher levels of chastity and monogamy. Look around. What we see is unprecedentedly low investment in posterity among the ruling class.

  16. #16, rate and citation please.

  17. Miguel Madeira

    #12, Also, given the frequency of premature death of parents in a society where the life expectency is about forty years

    I think that this life expectancy of 40 years was largely a life exectancy of 5 years for most people and of 60 years for the few survivors. If the “age of procreation” was relativly low in these times (let’s say, 16 y.o.), probably there was not much cases of orphans because that.

  18. #16 : Hello, is that correlation? It’s causation here. Yeah, hi. Just wanted to check, are we the same? No, didn’t think so. OK bye.

  19. “I think that this life expectancy of 40 years was largely a life exectancy of 5 years for most people and of 60 years for the few survivors. If the “age of procreation” was relativly low in these times (let’s say, 16 y.o.), probably there was not much cases of orphans because that.”

    Death in childbirth leaving men as widowers with children was a pretty ordinary phenomena in early modern Europe and colonial America. There were certainly people who were living to 60 and 70 and older in pre-modern times, but while child mortality was great there was also a lot more death among people in their prime particularly women.

    The biggest sources of increased life expectency have been a decline in a wide variety of infectious diseases that are deadly to people who aren’t particularly frail, and a decline in deaths from trauma.

  20. Anonymous age 70

    I am going to admit there might be somewhere a low rate of paternity error. But, it is not true overall. A few years ago, I had my y-markers tested. Since then, I have had hundreds of close matches reported, and almost all of them have a different surname. In fact, a very large number of surnames.

    Others have reported the exact same thing.

    I do accept the possibility that current rates are very low. With things like the pill, it makes sense. It is the historical claims I simply do not find consistent with their claims.

    Yet, here we have a highly mathematic claim there are few paternity errors. Hogwash.

    I have concluded that y-marker testing, the basic test, is useless in genealogy due to the large number of paternity errors. The exception is a new test which claims to identify kinfolk to the degree of fifth cousin.

    I do not begrudge the several hundred dollars I spent on y-marker, because I learned a lot. But, it was no help at all for my family history.

    In fact, the company experts essentially admit this. They advise if there is no paper trail to support the DNA results to ignore them. An outrageous statement.

    In 1993, the MSM reported that a retiring doctor had said in 1953 he monitored blood types in an East Coast hospital, and 10% of the children born had blood types not possible with the blood type of the husband.

    I do not know what these folks are trying to do with their mathematic studies, but I for one do not buy their claims.

    Let me add to the list of paternity errors,of course war rapes. Rapes. Adoptions. Secret adoptions; if a girl got pregnant, the mother sometimes took the baby as her own to make it possible for the daughter to someday marry. Foundlings. Also, recording of erroneous names by immigration officials who did not understand the accent of an immigrant. And, in Ireland, around 1100 when the priests insisted on surnames, sometimes every tribe member took the surname of the chief even when they were not blood kin.

    Other studies have reported the same CURRENT paternity error rates, for the more developed societies, but as high as 30% for less developed societies.

  21. Sandgroper

    “I am going to admit there might be somewhere a low rate of paternity error.”

    Somewhere where – in the world?

    “as 30% for less developed societies.” – such as?


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar