PDA

View Full Version : Probability of cuckoldry in geneaology



Hando
05-23-2015, 03:42 AM
Hi, I have been told that there is a 2% chance of cuckoldry in every generation.
1) Is this true or does this 2% include all other sources of NPE such as adoptions, children born to mistresses etc?
2) And if there is a 2% chance of cuckoldry in every generation, does this mean that after 50 generations there is a 100% guaranteed chance of cuckoldry in a lineage?
3) The Danish monarchy has been around for 1,000 years. That is approximately 50 generations. If the statistics are true about the 2% cuckoldry rate, then doesn't this mean that the current head of the Danish Royal family is not biologically descended from the original founder of the dynasty? Everyone's genealogical search seems meaningless if this is the case.

Megalophias
05-23-2015, 05:19 AM
If Danish royalty has the same chance of NPE as other people, which I doubt, and it is actually 50 generations, then it would be a 36% chance that the current guy is in fact a real patrilineal descendant.

Because as said in the other thread, probabilities do not add like that. If they did, a coin could never come up heads twice in a row.

Hando
05-23-2015, 06:05 AM
If Danish royalty has the same chance of NPE as other people, which I doubt, and it is actually 50 generations, then it would be a 36% chance that the current guy is in fact a real patrilineal descendant.

Because as said in the other thread, probabilities do not add like that. If they did, a coin could never come up heads twice in a row.
Sorry, but why would you doubt that the Danish monarchy has the same chance of NPE as other people?
And how did you calculate it to be 26% chance that the current Danish monarch is a patrilineal descendant?
1) Does this 2% chance of NPE include all other sources of NPE such as adoptions, children born to mistresses etc or just cuckoldry?

Megalophias
05-23-2015, 07:19 AM
Sorry, but why would you doubt that the Danish monarchy has the same chance of NPE as other people?
Because people care a lot more about the parentage of the heir to the throne than of some random schmoe. Because queens are likely to be surrounded by handmaids and guards and shit, giving them less chance to screw around. Because royalty does is a big deal and tends to get recorded, so if the King decides to adopt someone as the heir to throne chances are someone is going to write it down somewhere.


And how did you calculate it to be 26% chance that the current Danish monarch is a patrilineal descendant?
If there is a 2% chance of an NPE in a generation, there is a 98% chance that there isn't an NPE, right? So there is a 98% chance that the father's patrilineage is being passed down.

So in the first generation after the founder there is a 98% chance that the heir is carrying the original Y DNA. Then in the next generation there is again a 98% chance that the father's Y DNA is being passed on. So now it is a 98% of 98% chance (96%) that the grandfather's Y DNA is being passed on. And then in the next generation it will be 98% of 98% of 98%, and so on down the line. When you get to 98% of 98% of 98%... 50 times, that's 36%.

Hando
05-23-2015, 08:15 AM
Because people care a lot more about the parentage of the heir to the throne than of some random schmoe. Because queens are likely to be surrounded by handmaids and guards and shit, giving them less chance to screw around. Because royalty does is a big deal and tends to get recorded, so if the King decides to adopt someone as the heir to throne chances are someone is going to write it down somewhere.


If there is a 2% chance of an NPE in a generation, there is a 98% chance that there isn't an NPE, right? So there is a 98% chance that the father's patrilineage is being passed down.

So in the first generation after the founder there is a 98% chance that the heir is carrying the original Y DNA. Then in the next generation there is again a 98% chance that the father's Y DNA is being passed on. So now it is a 98% of 98% chance (96%) that the grandfather's Y DNA is being passed on. And then in the next generation it will be 98% of 98% of 98%, and so on down the line. When you get to 98% of 98% of 98%... 50 times, that's 36%.
Thanks, that makes a lot of sense.
But what about the last question. Does this 2% chance of NPE include all other sources of NPE such as adoptions and name changes etc or just cuckoldry? In other words, is it more than 2% if we were to include adoptions and name changes?

Baltimore1937
05-23-2015, 08:53 AM
Back when I thought my direct maternal line descended from a daughter (Joan, Lady of Wales) of a mistress of King John, I saw an amusing tidbit connected with that. A high class (knight or lord, etc) man was found in her bedroom, whereupon he was promptly hanged by her husband Prince Llewelyn ap Iowerth (spelling?) of Wales.

Dave-V
05-23-2015, 01:44 PM
Hi, I have been told that there is a 2% chance of cuckoldry in every generation.
1) Is this true or does this 2% include all other sources of NPE such as adoptions, children born to mistresses etc?
2) And if there is a 2% chance of cuckoldry in every generation, does this mean that after 50 generations there is a 100% guaranteed chance of cuckoldry in a lineage?
3) The Danish monarchy has been around for 1,000 years. That is approximately 50 generations. If the statistics are true about the 2% cuckoldry rate, then doesn't this mean that the current head of the Danish Royal family is not biologically descended from the original founder of the dynasty? Everyone's genealogical search seems meaningless if this is the case.

1) Most of the studies I've seen look at all breaks between Y-chromosome descent and presumed male descent, not just cuckoldry. Often it is broadened to breaks between Y-descent and surname descent which widens it further to include name changes etc. See for instance the referenced studies and discussions here: http://www.isogg.org/wiki/Non-paternity_event. Edit: Some of those studies looked just at cases where there was a high confidence of paternity (i.e. where fathers were pretty sure they were the father). That tends to narrow down the cases to mostly cuckoldry but not necessarily only that.

Whether the 2% is true depends on culture and historical period, and even if it's an average the standard deviation must be pretty wide. The percent has also probably gotten smaller in recent centuries too in Western culture, since in medieval periods there was less stigma about illegitimacy and people changed surnames more frequently of course.

2) The p*n formula (i.e. 2*50=100) would only work if you were counting every occurrence of an NPE (non-paternal event) as a separate event, not calculating the chances of it happening at least once. The probability that an NPE occurred at least once is 1- (1-p)^n. where p is the probability and n is the number of generations. So for example over 50 generations you'd get 1-(0.98^50) = 64% chance of an NPE (or 36% chance of NO NPE as has been quoted here already).

That math still assumes things like that the probability of an NPE occurring doesn’t change over the generations or that the occurrence of an NPE in one generation doesn’t affect the chances of it occurring in the next generation (i.e. they're always independent events) etc.

3) You're basically right (IMO). The point isn't that the probability is 36% versus 45% versus 55%, it's that everyone has a whopping big chance of their genetic family tree not being identical to their genealogical family tree. Some people know that already (e.g. because of recent adoptions). But (in my view) we ALL have NPEs (at least in the widest sense) among our millions of ancestors. Does that make genetic testing useless? Not to me. Just more complex.

Hando
05-23-2015, 03:55 PM
3) You're basically right (IMO). The point isn't that the probability is 36% versus 45% versus 55%, it's that everyone has a whopping big chance of their genetic family tree not being identical to their genealogical family tree. Some people know that already (e.g. because of recent adoptions). But (in my view) we ALL have NPEs (at least in the widest sense) among our millions of ancestors. Does that make genetic testing useless? Not to me. Just more complex.

Thanks for your considered reply. I don't know what the ^ symbol is for the formula 1- (1-p)^n

According to the link you provided adoption is one of the NPE scenarios in the context of genetic genealogy. Let us imagine as an example, that we have the genealogical records (paper trail) of a family lineage that can be traced back 25 generations. Let us further imagine that at the 12th generation and 13th generation of this lineage there were two adoptions. So the 12th man and the 13th man both adopted sons to carry on the line, but these adopted sons are younger brothers' sons. (Ie the 12th and 13th man's nephews become their respective adopted son). Would these two adoptions from within the circle of blood relatives count towards the 2% NPE rate we are talking about? I ask this question just to confirm whether I am correct in thinking that the adoption example I provided is a valid example of a break between Y-descent and surname descent.
Someone (name withheld to respect member's privacy) on this forum has also said that in their studies they have found a 2% NPE rate that only includes cuckoldry and that if one were to include all forms of NPE the rate would increase to 3.6%. I am wondering what the chances of cuckoldry at the rate of 2% would be in 25 generations.

Dave-V
05-23-2015, 06:51 PM
I don't know what the ^ symbol is for the formula 1- (1-p)^n

The "^" was my attempt to mean "to the power of" without a superscript. I see this editor does have superscript so you could also write the formula 1-(1-p)n.


Would these two adoptions from within the circle of blood relatives count towards the 2% NPE rate we are talking about?

The chances of an NPE (i.e. the 2%), depending on what you include in the definition of "NPE", is based on human tendencies from societal, religious, etc factors, and if we assume that percentage doesn't change over the generations (as the formula does) then it's fixed in the same way that a coin's 50% chance of coming up heads is based on the structure of the coin. And in the same way as the 50% chance of coming up heads is independent of the number of times a coin actually comes up heads, the 2% chance isn't affected by the number of adoptions (or cuckoldry, etc) in anyone's line.

If it helps, think of flipping a coin twice. It has a 50% chance of heads (H) or tails (T) so the four possible outcomes are HH, HT, TH, and TT. No matter which of those outcomes you get, the chances are still 50% of getting heads or tails in any one try.

Your example is interesting though...In your example if the men adopted their brothers' sons, then the adopting father and the biological father (as brothers) would have the same Y-DNA and presumably the same surnames, so the adopted sons would share both with their adopting father and there would never be a break between Y-chromosome descent and surname descent. But if the men adopted their sisters' sons then the adopted sons would have different Y-DNA from the adopting fathers but presumably take his surname and there would be a break.

The second case would be called an NPE (actually 2 NPE events, I guess), but I don't think the first case would. And actually if the first case wasn't recorded it would probably never come to light through genetic testing or otherwise.

I wouldn't get too hung up on the 2% and what kinds of NPEs (or not) it includes though. It's just an estimate for modeling purposes. The more causes you include obviously the more the percentage will go up, but that just mean it might vary between 1% and 3% in some cultures or time periods as you include more causes and in others it might vary from 1% to 10% or more if certain causes are more prevalent at that time and place. A good example is the changing of surnames in the Middle Ages... since it happened so frequently, including that cause in your definition of NPE probably increases the chances of an NPE significantly by several percentage points. In today's world where surname changes are much rarer, you might be able to say it didn't matter if that cause was in the definition or not because it wouldn't greatly affect the 2%.



I am wondering what the chances of cuckoldry at the rate of 2% would be in 25 generations.

At the rate of 2% in 25 generations it would be 1-(1-0.02)25 or 40% (39.65%).

apophis99942
05-23-2015, 10:59 PM
Hi, I have been told that there is a 2% chance of cuckoldry in every generation.
1) Is this true or does this 2% include all other sources of NPE such as adoptions, children born to mistresses etc?
2) And if there is a 2% chance of cuckoldry in every generation, does this mean that after 50 generations there is a 100% guaranteed chance of cuckoldry in a lineage?
3) The Danish monarchy has been around for 1,000 years. That is approximately 50 generations. If the statistics are true about the 2% cuckoldry rate, then doesn't this mean that the current head of the Danish Royal family is not biologically descended from the original founder of the dynasty? Everyone's genealogical search seems meaningless if this is the case.

Do you think that ideas like three-parent babies are novelty? I know that they recently passed a law in the UK about this. Because of this, I'm unsure of the relevance.

Hando
05-24-2015, 05:28 AM
The "^" was my attempt to mean "to the power of" without a superscript. I see this editor does have superscript so you could also write the formula 1-(1-p)n.



The chances of an NPE (i.e. the 2%), depending on what you include in the definition of "NPE", is based on human tendencies from societal, religious, etc factors, and if we assume that percentage doesn't change over the generations (as the formula does) then it's fixed in the same way that a coin's 50% chance of coming up heads is based on the structure of the coin. And in the same way as the 50% chance of coming up heads is independent of the number of times a coin actually comes up heads, the 2% chance isn't affected by the number of adoptions (or cuckoldry, etc) in anyone's line.

If it helps, think of flipping a coin twice. It has a 50% chance of heads (H) or tails (T) so the four possible outcomes are HH, HT, TH, and TT. No matter which of those outcomes you get, the chances are still 50% of getting heads or tails in any one try.

Your example is interesting though...In your example if the men adopted their brothers' sons, then the adopting father and the biological father (as brothers) would have the same Y-DNA and presumably the same surnames, so the adopted sons would share both with their adopting father and there would never be a break between Y-chromosome descent and surname descent. But if the men adopted their sisters' sons then the adopted sons would have different Y-DNA from the adopting fathers but presumably take his surname and there would be a break.

The second case would be called an NPE (actually 2 NPE events, I guess), but I don't think the first case would. And actually if the first case wasn't recorded it would probably never come to light through genetic testing or otherwise.

I wouldn't get too hung up on the 2% and what kinds of NPEs (or not) it includes though. It's just an estimate for modeling purposes. The more causes you include obviously the more the percentage will go up, but that just mean it might vary between 1% and 3% in some cultures or time periods as you include more causes and in others it might vary from 1% to 10% or more if certain causes are more prevalent at that time and place. A good example is the changing of surnames in the Middle Ages... since it happened so frequently, including that cause in your definition of NPE probably increases the chances of an NPE significantly by several percentage points. In today's world where surname changes are much rarer, you might be able to say it didn't matter if that cause was in the definition or not because it wouldn't greatly affect the 2%.




At the rate of 2% in 25 generations it would be 1-(1-0.02)25 or 40% (39.65%).

Razib Khan's article is interesting. It says that "societies such as England where surnames have a relatively deep history in some families (generally high socioeconomic status ones) the vast majority of men share the same male ancestor hundreds of years in the past."
And among Sephardic Kohanim Jewish priests the rate of NPE is only 04%. Much lower than the 2% we were discussing.

http://blogs.discovermagazine.com/gnxp/2010/06/the-paternity-myth-the-rarity-of-cuckoldry/#.VWJCMlLFq9Z

Does this mean that even if the average rate of NPE was 2%, certain socio-economic circles such as the nobility, would have a much lower rate of NPE than 2% per generation? And therefore that the chance of NPE in the genealogical lineage of 25 generations I referred to above could be much lower than 2% NPE generation?

Hando
05-24-2015, 05:39 AM
http://www.livescience.com/40627-cuckolds-rare-in-belgium.html
This article about the Flanders study states "In a second experiment, the researchers looked at a population of men with French last names whose ancestors (like Larmuseau) emigrated from Northern France during turmoil in the 1600s. Even after 400 years, DNA from the male sex (Y) chromosome looked very different between those with French versus Flemish last names, suggesting the rate of misattributed paternity was low.

Both methods showed a rate of mistaken paternity of about 1 to 2 percent.

The findings suggest the rate of cuckoldry has been consistently low in this population for the last 400 years. Still, it's hard to draw conclusions about past or current levels of promiscuity, because the records can't reveal why a man would raise a child who was not genetically his. Perhaps a man was duped into thinking he was the father, or he knowingly adopted the child. Either way, amateur and professional genealogists can breathe a sigh of relief. "Genealogical records are mostly correct and you can use these for sociological research," Larmuseau said."

Doesn't this suggest that if one has genealogical records dating back 25 generations (600 years) as I referred to above, then it is safe to state that one is genetically descended from one's founding ancestor from 25 generations ago as claimed in the official genealogical records?

George Chandler
05-24-2015, 05:53 AM
http://www.livescience.com/40627-cuckolds-rare-in-belgium.html
This article about the Flanders study states "In a second experiment, the researchers looked at a population of men with French last names whose ancestors (like Larmuseau) emigrated from Northern France during turmoil in the 1600s. Even after 400 years, DNA from the male sex (Y) chromosome looked very different between those with French versus Flemish last names, suggesting the rate of misattributed paternity was low.

Both methods showed a rate of mistaken paternity of about 1 to 2 percent.

The findings suggest the rate of cuckoldry has been consistently low in this population for the last 400 years. Still, it's hard to draw conclusions about past or current levels of promiscuity, because the records can't reveal why a man would raise a child who was not genetically his. Perhaps a man was duped into thinking he was the father, or he knowingly adopted the child. Either way, amateur and professional genealogists can breathe a sigh of relief. "Genealogical records are mostly correct and you can use these for sociological research," Larmuseau said."

Doesn't this suggest that if one has genealogical records dating back 25 generations (600 years) as I referred to above, then it is safe to state that one is genetically descended from one's founding ancestor from 25 generations ago as claimed in the official genealogical records?

It's not safe to state that one is genetically descended from one's founding ancestor unless you have multiple converging lines which line up genetically. At that point you can only say it's highly probably the line isn't broken from that point. There is always the possibility if a close relative jumping into the gene pool which is why you can't say with absolute certainty the line isn't broken you just increase the probability that it hasn't been broken. When it comes to descending from an ancient line the only way to be 100% certain of that specific ancient ancestor's DNA is to test the remains.

George

Hando
05-24-2015, 06:10 AM
It's not safe to state that one is genetically descended from one's founding ancestor unless you have multiple converging lines which line up genetically. At that point you can only say it's highly probably the line isn't broken from that point. There is always the possibility if a close relative jumping into the gene pool which is why you can't say with absolute certainty the line isn't broken you just increase the probability that it hasn't been broken. When it comes to descending from an ancient line the only way to be 100% certain of that specific ancient ancestor's DNA is to test the remains.

George

If that is the case, then doesn't it mean that most or all of those ancient lines such as the British Royal family, the 1,000 year old Danish Royal family and members of the nobility such as the Earl of Spencer and others who claim descent from such historical figures as William the Conqueror are in fact not descended from their lineage founder?
And what about the Belgians in this Lamurseau study that claims that Belgian men with French last names whose ancestors (like Larmuseau) emigrated from Northern France during turmoil in the 1600s are mostly descended from their purported ancestors from 400 years ago?

jbarry6899
05-24-2015, 01:18 PM
I'm reminded while reading this thread of a few ways in which uncertainty and evidence are expressed.

First, there are the legal standards, "preponderance of evidence," and "beyond a reasonable doubt." It would seem to me that a well-documented pedigree, combined with membership in a population with a historically low NPE rate, would meet the first standard. To reach the second, however, would require additional evidence, such as the convergence of multiple genetic lines, as suggested.

Second, back in the days when I was a professional analyst, we made an effort to standardize our expression of uncertainty in conclusions. If we felt that there was about a 3-1 chance that something was correct (75% probabilty), we used the qualifier "probably." Only if we believed that there was a 90% or greater probability would we say that something was "almost certainly" true.

Those with statistical backgrounds will see these standards as roughly equivalent to one- and two-sigma confidence intervals.

A couple of other thoughts: It's important, I think, to recognize that the creators of midaeval pedigrees did not employ the standard that we would consider today for historical research. In fact, pedigrees were political, not historical documents, intended to bolster a claim to a title or property. In my paternal ancestry, in fact, we have two certified pedigrees, one by French authorities and the other by English, that give contradictory information about the family's origin. There is another pedigree for this family in which a later reviewer found that information was omitted that he characterized as "embarassing to both parties." Thus, even an apparently well-established pedigree may contain accidental or deliberate errors.

As we are doing genetic genealogy, a useful yardstick is the Genealogical Proof Standard of the Board For Certification of Genalogists. It is as follows:

Reasonably exhaustive search

Assumes examination of a wide range of high quality sources
Minimizes the probability that undiscovered evidence will overturn a too-hasty conclusion

Complete and accurate citation of sources

Demonstrates the extent of the search and the quality of the sources
Allows others to replicate the steps taken to reach the conclusion. (Inability to replicate the research casts doubt on the conclusion.)

Analysis and correlation of the collected information

Facilitates sound interpretation of the data contributed by each source
Ensures that the conclusion reflects all the evidence

Resolution of conflicting evidence.

Substantiates the conclusion's credibility. (If conflicting evidence is not resolved, a credible conclusion is not possible.)

Soundly reasoned, coherently written conclusion.

Eliminates the possibility that the conclusion is based on bias, preconception, or inadequate appreciation of the evidence
Explains how the evidence led to the conclusion

Full disclosure: I am the person who posted the 3.6% NPE rate from all sources. It is a back of the envelope calculation of the mean for 17 Irish families over 35 generations and has a 95% confidence interval of roughly 1.5%. The context is in another thread: http://www.anthrogenica.com/showthread.php?2921-NPEs-in-an-Anglo-Irish-Family

Jim

Dave-V
05-24-2015, 03:06 PM
First, there are the legal standards, "preponderance of evidence," and "beyond a reasonable doubt." It would seem to me that a well-documented pedigree, combined with membership in a population with a historically low NPE rate, would meet the first standard. To reach the second, however, would require additional evidence, such as the convergence of multiple genetic lines, as suggested.


I agree with you in principle but we could debate if the standard of "beyond a reasonable doubt" is ever reachable. Even the Jewish priests example that Hando quoted at 0.4% (not 0.04) per event has a 10% chance of an NPE across 25 generations, and that's assuming all 25 generations are from that group. To me it's still not beyond a reasonable doubt if one in ten men are probably wrong.

Uncertainty is nothing new in genealogy, however. Church charters have been forged, government officials make mistakes and even tombstones have wrong dates. The comment "DNA doesn't lie" is sophistry. There are no absolutes in this business.

Hando
05-24-2015, 03:38 PM
Hi, what DNA test should I test for myself to determine if I am indeed genetically descended from my purported ancestor from 25 generations ago? Luckily, the Y DNA subclade for my surname has been tested for, so I don't need to test relatives.

jbarry6899
05-24-2015, 03:54 PM
You will need YDNA tests, both STR and SNP, but in addition you will need to identify other men descended from that purported individual along different lines of descent and obtain test results for them. Also, you should investigate both your pedigree and those of the other test-takers to assess the quality of documentation and the integrity of the conclusions. If several men have different, well-documented, ancestral lines originating with the same individual, and they also have a strong YDNA match at high resolution (ideally 111 STR markers) and are in the same haplogroup and subclade as identified by SNP testing, you can have good confidence in your common ancestry. There is no precise way to calculate the degree of certainty but by comparing it with the Genealogical Proof Standard I mentioned upthread you can determine whether your research meets those criteria.

Hando
05-25-2015, 01:22 AM
You will need YDNA tests, both STR and SNP, but in addition you will need to identify other men descended from that purported individual along different lines of descent and obtain test results for them. Also, you should investigate both your pedigree and those of the other test-takers to assess the quality of documentation and the integrity of the conclusions. If several men have different, well-documented, ancestral lines originating with the same individual, and they also have a strong YDNA match at high resolution (ideally 111 STR markers) and are in the same haplogroup and subclade as identified by SNP testing, you can have good confidence in your common ancestry. There is no precise way to calculate the degree of certainty but by comparing it with the Genealogical Proof Standard I mentioned upthread you can determine whether your research meets those criteria.
Thanks for the advice.
1)What are the test names I need in order to test for both STR and SNP YDNA?
2)My main problem will be in convincing other men descended from that purported ancestor from 25 generations ago. Since they would be very distantly related to me after 25 generations it would be difficult to allay their suspicions. Has anyone dealt with this sort of thing before?
On the other hand, I have been told that they have tested people with my surname and found the cluster.
That is why I wondered if it was enough just to test myself and compare it with the cluster that is identified with my surname. How accurate would that be?

jbarry6899
05-25-2015, 02:59 AM
1 You need to begin with a YDNA test of at least 37 markers. That would enable Family Tree DNA to predict your haplogroup, after which you can order specific SNPs and STR upgrades as necessary.

2. The first step is to identify those persons by traditional genealogical research. Start with your ancestor of interest an follow as many lines as possible forward to living descendants. Once you have done that you can contact them, describe your goals and ask them to test. You can offer to pay for tests if that would help them to participate.

3. If your surname is very unusual you could start just by looking at others who share it but if it is at all common there could be many origins and family lines so you would have no assurance that you and any matches share the distant ancestor in which you are interested.

Hando
05-25-2015, 08:05 AM
Your example is interesting though...In your example if the men adopted their brothers' sons, then the adopting father and the biological father (as brothers) would have the same Y-DNA and presumably the same surnames, so the adopted sons would share both with their adopting father and there would never be a break between Y-chromosome descent and surname descent. But if the men adopted their sisters' sons then the adopted sons would have different Y-DNA from the adopting fathers but presumably take his surname and there would be a break.

The second case would be called an NPE (actually 2 NPE events, I guess), but I don't think the first case would. And actually if the first case wasn't recorded it would probably never come to light through genetic testing or otherwise.
Just to play Devil's advocate here, if adopting one's brother's son (nephew) is not considered part of the 2% or 3.6% or whatever NPE rate you adhere to, then what about being cuckolded by one's brother? I know that cuckoldry is part of NPE and it should be because it indicates that you are not the father of your son, just as adopting your nephew also falls into the category of a "non son event." If such is the case, then shouldn't the NPE rate of 2% or 3.6% or whatever NPE rate you adhere to, also include adopting your nephew? I come from a society where in the past male and female contact was very strict and patrilochal. It wasn't as strict as Islamic culture, but adopting outside of the kin was extremely rare. That is why if one could not produce a son, one adopted from one's own family. Preferably a nephew. If no nephew was available, then from some other male blood relation. The 0.4% NPE rate for Sephardic Jewish Kohanim priests quoted by Razib Khan comes to mind...

As such I would think that
1) the NPE rate was less than the 2% average
2) adopting one's nephew or other kinsman should be also considered NPE at least in practice.
Any thoughts? The reason why I am asking this is because I can trace my lineage back 25 generations. We have the family official genealogical records that record most of the individuals that descend from an ancestor in the late 1300s AD.
At the 12th generation there was an adoption. The 12th ancestor did not have a son, so he adopted one of his younger brother's sons. Namely his nephew.
At the 13th generation there was another adoption. (By the adopted nephew.) This time a male cousin's son was adopted. So it was also within the family. They could all trace their descent from the lineage founder from the 1300s AD. These are two cases of a break in a father and son relationship in 25 generations. If we are to go by the 2% NPE per generation probability, then this would be double the number of NPE for 25 generations. I wonder if I can therefore count these two adoptions from within the kin/blood circle towards the 2% NPE rate. If we go by the 1% NPE rate it would be far more sufficient. Even if we go by the 3.7% NPE per generation rate, it would be enough.
Thanks

George Chandler
05-26-2015, 04:55 AM
If that is the case, then doesn't it mean that most or all of those ancient lines such as the British Royal family, the 1,000 year old Danish Royal family and members of the nobility such as the Earl of Spencer and others who claim descent from such historical figures as William the Conqueror are in fact not descended from their lineage founder?
And what about the Belgians in this Lamurseau study that claims that Belgian men with French last names whose ancestors (like Larmuseau) emigrated from Northern France during turmoil in the 1600s are mostly descended from their purported ancestors from 400 years ago?

My reply was in a more general terms and meant to deal with the degree of certainty needed to establish proof of a certain lineage. It doesn't provide information, clues or guesses as to what specific lineage may or may not be legitimate. Family lines become broken no matter of class or social status for various reasons and as mentioned in the other post sometimes the genealogy is wrong.

George

Hando
05-26-2015, 12:23 PM
My reply was in a more general terms and meant to deal with the degree of certainty needed to establish proof of a certain lineage. It doesn't provide information, clues or guesses as to what specific lineage may or may not be legitimate. Family lines become broken no matter of class or social status for various reasons and as mentioned in the other post sometimes the genealogy is wrong.

George

1)When you talk about family lines becoming broken, are you including adoptions as well, or just cuckoldry?
2)And I beg to differ. I think some social classes have a higher chance of becoming broken than others. Just take a look at the stats in that Razib Khan link. Michigan black populations in Table 1 have a 10.1% rate of NPE per generation, while Michigan whites have a 1.49% rate of NPE. I also think that nobility have a much lower NPE per generation rate than commoners. As for royalty Megalopolis has mentioned that they were under more scrutiny so the chances of cuckoldry are not as easy.

George Chandler
05-27-2015, 12:23 AM
No I wasn't including adoptions. My statement "Family lines become broken no matter of class or social status for various reasons" is a general statement which could include children such as William The Conqueror who was his fathers son through another mother (though not an NPE). It does stand to reason that if the male in a household (no matter the social class) is out sowing his seed elsewhere that what was good for the goose was also good for the gander..so to speak. I doubt any body guard who wanted to keep his head would tattle on the wife during past centuries (especially if they felt sorry for her). I haven't run any numbers on broken lines within the historic upper class and won't be doing so anytime soon.

You could be correct that they have a lower statistical NPE per generation? I really don't know.

George

Táltos
05-27-2015, 03:06 AM
As for royalty Megalopolis has mentioned that they were under more scrutiny so the chances of cuckoldry are not as easy.

Well I wouldn't be surprised if a *few* romances occurred between the royalty and the guards!

Hando
05-28-2015, 03:30 AM
I'm reminded while reading this thread of a few ways in which uncertainty and evidence are expressed.

First, there are the legal standards, "preponderance of evidence," and "beyond a reasonable doubt." It would seem to me that a well-documented pedigree, combined with membership in a population with a historically low NPE rate, would meet the first standard. To reach the second, however, would require additional evidence, such as the convergence of multiple genetic lines, as suggested.

Second, back in the days when I was a professional analyst, we made an effort to standardize our expression of uncertainty in conclusions. If we felt that there was about a 3-1 chance that something was correct (75% probabilty), we used the qualifier "probably." Only if we believed that there was a 90% or greater probability would we say that something was "almost certainly" true.

Those with statistical backgrounds will see these standards as roughly equivalent to one- and two-sigma confidence intervals.

A couple of other thoughts: It's important, I think, to recognize that the creators of midaeval pedigrees did not employ the standard that we would consider today for historical research. In fact, pedigrees were political, not historical documents, intended to bolster a claim to a title or property. In my paternal ancestry, in fact, we have two certified pedigrees, one by French authorities and the other by English, that give contradictory information about the family's origin. There is another pedigree for this family in which a later reviewer found that information was omitted that he characterized as "embarassing to both parties." Thus, even an apparently well-established pedigree may contain accidental or deliberate errors.

As we are doing genetic genealogy, a useful yardstick is the Genealogical Proof Standard of the Board For Certification of Genalogists. It is as follows:

Reasonably exhaustive search

Assumes examination of a wide range of high quality sources
Minimizes the probability that undiscovered evidence will overturn a too-hasty conclusion

Complete and accurate citation of sources

Demonstrates the extent of the search and the quality of the sources
Allows others to replicate the steps taken to reach the conclusion. (Inability to replicate the research casts doubt on the conclusion.)

Analysis and correlation of the collected information

Facilitates sound interpretation of the data contributed by each source
Ensures that the conclusion reflects all the evidence

Resolution of conflicting evidence.

Substantiates the conclusion's credibility. (If conflicting evidence is not resolved, a credible conclusion is not possible.)

Soundly reasoned, coherently written conclusion.

Eliminates the possibility that the conclusion is based on bias, preconception, or inadequate appreciation of the evidence
Explains how the evidence led to the conclusion

Full disclosure: I am the person who posted the 3.6% NPE rate from all sources. It is a back of the envelope calculation of the mean for 17 Irish families over 35 generations and has a 95% confidence interval of roughly 1.5%. The context is in another thread: http://www.anthrogenica.com/showthread.php?2921-NPEs-in-an-Anglo-Irish-Family

Jim
Just to play Devil's advocate here, if adopting one's brother's son (nephew) is not considered part of the 2% or 3.6% or whatever NPE rate you adhere to, then what about being cuckolded by one's brother? I know that cuckoldry is part of NPE and it should be because it indicates that you are not the father of your son, just as adopting your nephew also falls into the category of a "non son event." If such is the case, then shouldn't the NPE rate of 2% or 3.6% or whatever NPE rate you adhere to, also include adopting your nephew? I come from a society where in the past male and female contact was very strict and patrilocal. It wasn't as strict as Islamic culture, but adopting outside of the kin was extremely rare. That is why if one could not produce a son, one adopted from one's own family. Preferably a nephew. If no nephew was available, then from some other male blood relation. The 0.4% NPE rate for Sephardic Jewish Kohanim priests quoted by Razib Khan comes to mind...

As such I would think that
1) the NPE rate was less than the 2% average
2) adopting one's nephew or other kinsman should be also considered NPE at least in practice.
Any thoughts? The reason why I am asking this is because I can trace my lineage back 25 generations. We have the family official genealogical records that record most of the individuals that descend from an ancestor in the late 1300s AD.
At the 12th generation there was an adoption. The 12th ancestor did not have a son, so he adopted one of his younger brother's sons. Namely his nephew.
At the 13th generation there was another adoption. (By the adopted nephew.) This time a male cousin's son was adopted. So it was also within the family. They could all trace their descent from the lineage founder from the 1300s AD. These are two cases of a break in a father and son relationship in 25 generations. If we are to go by the 2% NPE per generation probability, then this would be double the number of NPE for 25 generations. I wonder if I can therefore count these two adoptions from within the kin/blood circle towards the 2% NPE rate. If we go by the 1% NPE rate it would be far more sufficient. Even if we go by the 3.7% NPE per generation rate, it would be enough.
Thanks

Dave-V
05-28-2015, 05:43 AM
...These are two cases of a break in a father and son relationship in 25 generations. If we are to go by the 2% NPE per generation probability, then this would be double the number of NPE for 25 generations. I wonder if I can therefore count these two adoptions from within the kin/blood circle towards the 2% NPE rate. If we go by the 1% NPE rate it would be far more sufficient. Even if we go by the 3.7% NPE per generation rate, it would be enough.
Thanks

I don't think I really understand your question. If you mean, can you count the two known adoptions as NPEs and therefore be reasonably sure that no OTHER NPEs exist in your line, then no, that's not how probability works. Just because you find an NPE in one generation doesn't mean there is less of a chance of NPEs occurring elsewhere. Each generation has an independent chance of an NPE.

If you flip a coin 50 times and get 40 heads and 10 tails, it doesn't mean you will start getting more tails than heads if you flip the coin 50 more times. Each flip has a 50% chance of heads or tails that is not affected by the outcome of the flips before it. You COULD in theory flip a coin 100 times and get 100 heads (although I'd start to doubt the coin too before then :-) ).

Now, if you believe that your past generations have a culture that would yield less than a 2% per generation NPE rate, who am I to judge. You may well be right.

Hando
05-28-2015, 11:15 AM
I don't think I really understand your question. If you mean, can you count the two known adoptions as NPEs and therefore be reasonably sure that no OTHER NPEs exist in your line, then no, that's not how probability works. Just because you find an NPE in one generation doesn't mean there is less of a chance of NPEs occurring elsewhere. Each generation has an independent chance of an NPE.

If you flip a coin 50 times and get 40 heads and 10 tails, it doesn't mean you will start getting more tails than heads if you flip the coin 50 more times. Each flip has a 50% chance of heads or tails that is not affected by the outcome of the flips before it. You COULD in theory flip a coin 100 times and get 100 heads (although I'd start to doubt the coin too before then :-) ).

Now, if you believe that your past generations have a culture that would yield less than a 2% per generation NPE rate, who am I to judge. You may well be right.

Thanks for your well reasoned reply. I was actually wondering if these two known adoptions in my lineage could be considered NPE or not? I know there was no break in the Y DNA because the younger brothers' son was adopted into the older brother's family on both occasions, but the adopted nephews were not biological sons of the adopter/father. I was wondering if these two scenarios were examples of the 2% NPE (or 04%, or 1% or 3.7% etc etc) per generation we are discussing.
As for the issue regarding cultures that have a lower than 2% NPE, what are your thoughts on the 0.4% NPE per generation rate for Sephardic, Kohanim Priest lineages? If they can be as low as 0.4% NPE, wouldn't it be possible for other cultures such as mine?
Thanks

Kopfjäger
06-03-2015, 03:32 AM
Now, if you believe that your past generations have a culture that would yield less than a 2% per generation NPE rate, who am I to judge. You may well be right.

I think a lot of this NPE stuff is a bit of hyperbole. Joseon Korea (and by extension, modern Korea) comes to mind when thinking about the potential for NPEs. This is a society with strict, cultural norms, and one could easily become a social pariah by engaging in "non-sanctioned" relationships. This held especially true for various female members of the Joseon royal family, who held with high regard in keeping the legitimate, kingly lineage intact. Even in modern times, the sense of community and adherence to social standards act as strong inhibitions to behavior resulting in NPEs.

In short, I think the NPE rate does depend on cultural values and norms, although these have changed so much in the West that sometimes we forget that it too shared similar social norms with places like Korea in the past.

Hando
06-03-2015, 12:08 PM
I think a lot of this NPE stuff is a bit of hyperbole. Joseon Korea (and by extension, modern Korea) comes to mind when thinking about the potential for NPEs. This is a society with strict, cultural norms, and one could easily become a social pariah by engaging in "non-sanctioned" relationships. This held especially true for various female members of the Joseon royal family, who held with high regard in keeping the legitimate, kingly lineage intact. Even in modern times, the sense of community and adherence to social standards act as strong inhibitions to behavior resulting in NPEs.

In short, I think the NPE rate does depend on cultural values and norms, although these have changed so much in the West that sometimes we forget that it too shared similar social norms with places like Korea in the past.
I agree with you Kopfjaeger.
In addition many North American, Australian etc forum members have surnames from European countries such as Germany and UK and they can trace their paternal genealogies back centuries to these European mother countries. These members have tested for their Y DNA which turns out to roughly match the country of origin back in Europe. So for example, if your paternal line came from Cornwall during the 1600s as suggested by your surname and your Y DNA haplogroup and subclade indicate an origin in Cornwall, then how would that be possible if it wasn't for the fact that there was an unbroken line of descent stretching back 12 generations. If the chance of an NPE is 3.7% per generation the likelihood of an NPE should be high, yet many members with such scenarios managed to share the same genetic Y DNA as their ancestor 12/13/14 etc generations ago. This should indicate that the NPE rate is much lower than has been suggested on this thread by some. So I think even a 1% NPE is too high for many families. The average is brought up to 1% because members of certain social groups who have a high NPE rate has brought the average up.

It should also be noted that we should not project modern say NPE rates unto traditional societies from the past. Much has changed in the past 100 years and let us not forget that free love, women's lib etc are practices from the 60s and these have raised the NPE rate relative to historical NPE rates.

Megalophias
06-03-2015, 04:23 PM
I agree with you Kopfjaeger.
Much has changed in the past 100 years and let us not forget that free love, women's lib etc are practices from the 60s and these have raised the NPE rate relative to historical NPE rates.

Do you have any evidence for that? Committing adultery and then passing off the other guy's kids as your husband's is still totally unacceptable in modern culture. Also, for those who *are* sleeping around, we now have effective contraception, which didn't exist before. More premarital sex, yes, but I don't see how that translate into more adultery.

jbarry6899
06-03-2015, 04:46 PM
Do you have any evidence for that? Committing adultery and then passing off the other guy's kids as your husband's is still totally unacceptable in modern culture. Also, for those who *are* sleeping around, we now have effective contraception, which didn't exist before. More premarital sex, yes, but I don't see how that translate into more adultery.

Not adultery, but the percentage of births to unmarried women in the United States is increasing. How many of those bear their father's surname is unknown, but this complicates analysis of NPE rates. From the CDC:

"In 2007, nearly 4 in 10 births were to unmarried women: The proportion of all births to unmarried women was 39.7%, up from 34.0% in 2002. The 2007 proportion was more than double that for 1980 (18.4%)." http://www.cdc.gov/nchs/data/databriefs/db18.pdf

Táltos
06-03-2015, 05:08 PM
Not adultery, but the percentage of births to unmarried women in the United States is increasing. How many of those bear their father's surname is unknown, but this complicates analysis of NPE rates. From the CDC:

"In 2007, nearly 4 in 10 births were to unmarried women: The proportion of all births to unmarried women was 39.7%, up from 34.0% in 2002. The 2007 proportion was more than double that for 1980 (18.4%)." http://www.cdc.gov/nchs/data/databriefs/db18.pdf

Page 5 of your link also points out that the trend is not just unique to the USA.

Dave-V
06-03-2015, 11:14 PM
It should also be noted that we should not project modern say NPE rates unto traditional societies from the past. Much has changed in the past 100 years and let us not forget that free love, women's lib etc are practices from the 60s and these have raised the NPE rate relative to historical NPE rates.

While this may or may not be true in recent generations I don't think you can conclude recent NPE rates have risen relative to historical NPE rates unless you're talking about very specific traditional groups that have been very isolated from outside influences until recently.

For the wider populations I think it is more likely that rates for all the various causes of NPEs (however you choose to define it) fluctuate semi-independently over the centuries and the overall rate may be greater or smaller at times as a result, and not necessarily lower (or higher) than present.

Just from research done in my own surname group we have examples of battle survivors adopting the children of killed friends as their own, people adopting new surnames or the surname of their lord, landlord, or clan chief, and other cases of non-cuckoldry NPEs that are certainly less frequent today.

On the cuckoldry side we have at least two or three documented cases of illegitimate sons recognized by our ancestors from medieval times through the 1700s, all times when such things were much more openly acknowledged. It is difficult for us to consider those times by our own moral codes since we have changed so much.

For instance, I don't honestly think that jus primae noctis (droit du seigneur, i.e. lords having their way first with new brides) was ever really a widespread practice but just the fact that it was mentioned in texts as real for some 3000 years (http://en.wikipedia.org/wiki/Droit_du_seigneur) says that past cultures were considered to have practices that would have been judged completely immoral even in our supposedly ultra-liberal 60s free love era. By contrast, no-one who knew our modern Western morality would even think of suggesting that that practice has existed there in the past 100 years.

So I don't think we're projecting modern NPE rates onto the past if we said NPE rates could have been higher then at times. But here I would also agree with Kopfjaeger that there is plenty of hyperbole on this topic since we'll never really have useful data on the subject going back in time.

Hando
06-04-2015, 04:17 PM
While this may or may not be true in recent generations I don't think you can conclude recent NPE rates have risen relative to historical NPE rates unless you're talking about very specific traditional groups that have been very isolated from outside influences until recently.

For the wider populations I think it is more likely that rates for all the various causes of NPEs (however you choose to define it) fluctuate semi-independently over the centuries and the overall rate may be greater or smaller at times as a result, and not necessarily lower (or higher) than present.

Just from research done in my own surname group we have examples of battle survivors adopting the children of killed friends as their own, people adopting new surnames or the surname of their lord, landlord, or clan chief, and other cases of non-cuckoldry NPEs that are certainly less frequent today.

On the cuckoldry side we have at least two or three documented cases of illegitimate sons recognized by our ancestors from medieval times through the 1700s, all times when such things were much more openly acknowledged. It is difficult for us to consider those times by our own moral codes since we have changed so much.

For instance, I don't honestly think that jus primae noctis (droit du seigneur, i.e. lords having their way first with new brides) was ever really a widespread practice but just the fact that it was mentioned in texts as real for some 3000 years (http://en.wikipedia.org/wiki/Droit_du_seigneur) says that past cultures were considered to have practices that would have been judged completely immoral even in our supposedly ultra-liberal 60s free love era. By contrast, no-one who knew our modern Western morality would even think of suggesting that that practice has existed there in the past 100 years.

So I don't think we're projecting modern NPE rates onto the past if we said NPE rates could have been higher then at times. But here I would also agree with Kopfjaeger that there is plenty of hyperbole on this topic since we'll never really have useful data on the subject going back in time.

1) I believe that members of the nobility in Europe and other groups such as the Korean royal family had a much lower NPE rate than the general population.
2)When you talk about illegitimate sons from cuckoldry are you talking about fathers adopting boys who aren't their sons or accepting bastard sons?
3)If you believe that there is plenty of hyperbole on this topic of cuckoldry/NPE rates, then do you feel that the 2% cuckoldry rate may have been lower among the historical European nobility? I am specifically talking about cuckoldry.
Thanks

Dave-V
06-04-2015, 10:08 PM
1) I believe that members of the nobility in Europe and other groups such as the Korean royal family had a much lower NPE rate than the general population.
2)When you talk about illegitimate sons from cuckoldry are you talking about fathers adopting boys who aren't their sons or accepting bastard sons?
3)If you believe that there is plenty of hyperbole on this topic of cuckoldry/NPE rates, then do you feel that the 2% cuckoldry rate may have been lower among the historical European nobility? I am specifically talking about cuckoldry.
Thanks

1) That's fine. I believe it was likely higher than today's rates among nobility at least in the British Isles and Western Europe through to around 1700. Nobles were watched more but they were also allowed more, and royals (who were watched most of all) certainly had many mistresses and fathered many illegitimate children. Whether their rates was higher or lower than the general population at those times, I don't know. But since we will probably never have data to quantify it we'll never know for sure.

2) By "illegitimate" I mean children fathered by married men with a woman who was not their wife. Among the noble lines we have documented in our surname project there are several cases where men in their wills or land transactions referred to other men specifically as their "illegitimate son". Only rarely was the "other woman" identified. In one instance the King of England issued an order to place a whole family under house arrest and threw in "and his illegitimate son" just for good measure.

We have other cases of adoptions too, but I was separating those out.

3) No, it would not surprise me to learn that more than one out of every 50 births among Western European nobility during certain centuries, especially before 1700,was fathered by someone other than who the mother was married to. That's not because I think the men or women were "looser", it's just because from what I have read the rise of monogamy as a moral code (as opposed to something encouraged by the church) in Western society was associated with the social and religious equality movements (some earlier than 1700 of course) and the growth of the nuclear family in importance above other family or social ties.

But I honestly have no data or special insight into the question, so my belief is just that - a belief.

Hando
06-05-2015, 04:43 AM
1) That's fine. I believe it was likely higher than today's rates among nobility at least in the British Isles and Western Europe through to around 1700. Nobles were watched more but they were also allowed more, and royals (who were watched most of all) certainly had many mistresses and fathered many illegitimate children. Whether their rates was higher or lower than the general population at those times, I don't know. But since we will probably never have data to quantify it we'll never know for sure.

2) By "illegitimate" I mean children fathered by married men with a woman who was not their wife. Among the noble lines we have documented in our surname project there are several cases where men in their wills or land transactions referred to other men specifically as their "illegitimate son". Only rarely was the "other woman" identified. In one instance the King of England issued an order to place a whole family under house arrest and threw in "and his illegitimate son" just for good measure.

We have other cases of adoptions too, but I was separating those out.

3) No, it would not surprise me to learn that more than one out of every 50 births among Western European nobility during certain centuries, especially before 1700,was fathered by someone other than who the mother was married to. That's not because I think the men or women were "looser", it's just because from what I have read the rise of monogamy as a moral code (as opposed to something encouraged by the church) in Western society was associated with the social and religious equality movements (some earlier than 1700 of course) and the growth of the nuclear family in importance above other family or social ties.

But I honestly have no data or special insight into the question, so my belief is just that - a belief.
1) I thought NPE was about the male line being broken, but in the first paragraph you are talking about the male being the genetic father regardless of legitimacy. So are you also including illegitimate sons from mistresses outside of the marriage, but still sharing the real father as part of your NPE definition?
2)By more than one out of every 50 births being a product of cuckoldry, are you suggesting a rate of about 2% NPE per generation or more? I know you said there is no data, but I wonder why you specifically brought up one out of 50 as a number?
Thaknks

Dave-V
06-05-2015, 02:47 PM
1) I thought NPE was about the male line being broken, but in the first paragraph you are talking about the male being the genetic father regardless of legitimacy. So are you also including illegitimate sons from mistresses outside of the marriage, but still sharing the real father as part of your NPE definition?
2)By more than one out of every 50 births being a product of cuckoldry, are you suggesting a rate of about 2% NPE per generation or more? I know you said there is no data, but I wonder why you specifically brought up one out of 50 as a number?
Thaknks

1) Barring immaculate conception, there's always a genetic father :). The "illegitimate son" cases I was quoting aren't NPEs in our own lines, but presumably they are NPEs in the line of the cuckholded husband.

Technically an NPE by the usual definition is an unexpected break in the Y-DNA descent, so one could certainly argue that a well-documented illegitimate son isn't counted as an NPE because you can still follow the Y-DNA ancestry back through the genetic father. But I would still think illegitimate births occurred more often than was openly recognized, and of course not all illegitimacies openly recognized have documentation surviving to this day, so the documented cases should just be the visible part of the iceberg.

2) Yes, sorry, I was translating the 2% to the equivalent number of births. If 2% of births are NPEs that equates to 2 out of every 100 births or in other words 1 out of every 50 births.

Hando
06-07-2015, 05:46 AM
1) That's fine. I believe it was likely higher than today's rates among nobility at least in the British Isles and Western Europe through to around 1700. Nobles were watched more but they were also allowed more, and royals (who were watched most of all) certainly had many mistresses and fathered many illegitimate children. Whether their rates was higher or lower than the general population at those times, I don't know. But since we will probably never have data to quantify it we'll never know for sure.

2) By "illegitimate" I mean children fathered by married men with a woman who was not their wife. Among the noble lines we have documented in our surname project there are several cases where men in their wills or land transactions referred to other men specifically as their "illegitimate son". Only rarely was the "other woman" identified. In one instance the King of England issued an order to place a whole family under house arrest and threw in "and his illegitimate son" just for good measure.

We have other cases of adoptions too, but I was separating those out.

3) No, it would not surprise me to learn that more than one out of every 50 births among Western European nobility during certain centuries, especially before 1700,was fathered by someone other than who the mother was married to. That's not because I think the men or women were "looser", it's just because from what I have read the rise of monogamy as a moral code (as opposed to something encouraged by the church) in Western society was associated with the social and religious equality movements (some earlier than 1700 of course) and the growth of the nuclear family in importance above other family or social ties.

But I honestly have no data or special insight into the question, so my belief is just that - a belief.

Here's a link that shows Switzerland's NPE rate at 0.83% and Sephardic Kohanim priests at 0.4%
http://faculty-staff.ou.edu/A/Kermyt.G.Anderson-1/papers/worldwidepatconf.pdf
These are members from a wealthy GDP (in the case of Switzerland) and a conservative social group (in the case of the Jewish priests). I believe that the European nobility were both wealthier and more conservative than modern Europeans. So I believe that their cuckoldry rate would have been lower than that of modern Europeans.

And the Kohanim priests have a 0.4% NPE rate. Although only around 44 were tested, I bleieve that the European nobility were as concerned about bloodline and lineage as much as the Kohanim priests, so I would expect the European nobility NPE rate to be as low as the Kohanim NPE rate of 0.4%. But then again I am only guessing just as you said it is only a belief.

crossover
08-14-2015, 03:07 PM
Hi, I have been told that there is a 2% chance of cuckoldry in every generation.
1) Is this true or does this 2% include all other sources of NPE such as adoptions, children born to mistresses etc?
2) And if there is a 2% chance of cuckoldry in every generation, does this mean that after 50 generations there is a 100% guaranteed chance of cuckoldry in a lineage?
3) The Danish monarchy has been around for 1,000 years. That is approximately 50 generations. If the statistics are true about the 2% cuckoldry rate, then doesn't this mean that the current head of the Danish Royal family is not biologically descended from the original founder of the dynasty? Everyone's genealogical search seems meaningless if this is the case.

considering royaly often married their cousins, even if a a mom of a patrilineal ancestor had an affair and gave birth to another guy's kid, the current head would still have biological ties to original founder

JohnHowellsTyrfro
08-14-2015, 05:41 PM
From what I've read the aristocracy and royalty probably had as much of an opportunity for infidelity as the average person, probably more. "it is a wise child that knows his own father". Shakespeare.

Tomenable
08-18-2015, 12:49 AM
3) The Danish monarchy has been around for 1,000 years. That is approximately 50 generations.

A more realistic estimate for royalty is that 1000 years would be around 33-35 generations.

Check for example this line of descent from Mieszko I (900s) to Elisabeth II (1900s), which is 33 generations long:


I'm not sure about Charlemagne, but I know that Elisabeth II is descended from Poland's first monarch - Mieszko I - and not via just one, but through at least two distinct genealogical lines. One line is through Canute the Great (Canute was Mieszko's grandson - son of his daughter), the other line is via the Piast dynasty and then the Hohenzollerns. Including Siemowit I of Mazovia (Piast). I will focus on the latter, since the tree of Canute's descendants is relatively more well-known.

Mieszko I was Elisabeth's great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great-great grandfather (that is, 30x great-grandfather); through at least 2 lines of descent.

Here is the line of descent from Mieszko I x his wife Doubravka to Elisabeth II through Siemowit I (male line plus female line wherever male line was broken):

30x great-grandparents Mieszko I ( - died 992) x Doubravka Premyslovna ( - died 977)
29x Boleslav I Piast, king of Poland (born 967 - died 1025)
28x Mieszko II Piast king of Poland (990 - 1034)
27x Casimir I Piast duke of Poland (1016 - 1058)
26x Vladislav I Piast duke of Poland (1043 - 1102)
25x Boleslav III Piast duke of Poland (1086 - 1138)
24x Casimir II Piast supreme duke of Poland (1138 - 1194)
23x Conrad I Piast supreme duke of Poland (1187 - 1247)
22x Siemowit I Piast duke of Mazovia & Sieradz (1215 - 1262)
21x Boleslav II Piast duke of Mazovia & Sandomir (1251 - 1313)
20x Troyden I Piast duke in Mazovia (1284 - 1341)
19x Euphemia princess of Cieszyn & in Mazovia (1310 - 1374) x Casimir I Piast duke in Silesia (1280 - 1358)
18x Premislav I Piast duke of Cieszyn & in Upper Silesia (1332 - 1410)
17x Anne princess in Silesia (1374 - 1420) x Henry IX Piast duke in Lower Silesia (1369 - 1420)
16x Louis III Piast duke in Lower Silesia (1405 - 1441)
15x John I Piast duke in Lower Silesia (1425 - 1453)
14x Frederick I Piast duke in Lower Silesia (1448 - 1488)
13x Frederick II Piast duke in Lower Silesia (1480 - 1547)
12x Sophia princess of Legnica (1525 - 1546) x John George Hohenzollern of Brandenburg (1525 - 1598)
11x Joachim III Frederick Hohenzollern elector of Brandenburg (1546 - 1608)
10x John Sigismund Hohenzollern elector of Brandenburg, prince of Prussia (1572 - 1619)
9x George William Hohenzollern elector of Brandenburg, prince of Prussia (1593 - 1640)
8x Frederick William I Hohenzollern elector of Brandenburg, prince of Prussia (1620 - 1688)
7x Frederick I Hohenzollern king of Prussia (1657 - 1713)
6x Frederick William I Hohenzollern king of Prussia (1688 - 1740)
5x Sophia Dorothy of Brandenburg-Schwedt (1719 - 1765) x Frederick William Hohenzollern (1700 - 1771)
4x Sophia Dorothy princess of Württemberg (1736 - 1798) x Frederick II duke of Württemberg (1732 - 1797)
3x Louis Frederick duke of Württemberg (1756 - 1817)
2x great-grandfather Alexander duke of Württemberg (1804 - 1885)
great-grandfather Francis Württemberg duke of Teck (1837 - 1900)
grandparents Mary of Teck (1867 - 1953) x George V Coburg Windsor king of the UK (1865 - 1936)
father George VI Windsor king of the United Kingdom (1895 - 1952)
Elisabeth II queen of the United Kingdom (born 1926 - )

AFAIK, Elisabeth II has ancestors from every single dynasty of Europe.

Baltimore1937
08-18-2015, 11:04 PM
It's not necessary to laboriously track every generation all the way back to Charlemagne. All you have to do is connect to someone whose pedigree is already established, such as Eleanor of Aquitaine and husband Henry II. In my case, I connected to their daughter Eleanor.

One case of cuckoldry along the way is well recorded. In this case, the father had a son out of wedlock. The son changed his surname to Green, which (downstream) was the surname of the second colonial governor of Maryland. The son became knighted himself (as was his father), so the undisclosed mother must also have been of high class.

ffoucart
10-17-2015, 06:25 PM
I remember some papers (perhaps 10 years ago) from medical studies about NPE. There is a social bias: the lower the social class is, the more higher is the probability of NPE, and reverse. At this point, statistics were saying that near 25% of the children in the lower classes are the results of NPE, versus less than 1% in the higher classes.

Number were revised because the study was made with a sample of divorced mothers (in England, in the early 2000's).

Instead of 25% in the lower classes, it's more 10% (in later studies).

But the conclusion is the same: social bias of NPE.

It's not surprising: a woman has far more to lose in the higher classes than in the lower with infidelity.

botoole60611
11-11-2015, 07:15 PM
I just caught this thread. Here is my real life example through two generations, although one probably doesn't count.

I am adopted and 4 years ago reunited with my birth mother who gave me my bio-dad's name, and I found all his side. Bio-dad was born two years before his parents married, but his sister always assumed that her parents had an affair, since her father's first wife was dying and eventually died two years after my dad was born. Get along with them just fine. Put together a tree on ancestry.

Two years ago a first/second cousin popped up on Ancestry DNA, and this match's tree did not line up with mine, but it turned out my paternal grandmother was from the same small town of Kingsley, MI, as her father. Piecing things and family lore together, it looks like her uncle is my grandfather. DNA tests with three separate family first cousins, once removed, each showed about 450 cM, which is in the right . So I had my dad's paternal line solved. Had to saw off a big branch from my tree and replace it. Ended up with a new family surname of Smith.

I had also DNA tested a first cousin once remopved on my dad's maternal side. She tested only about 133 cM, but I never pursued it. My thinking was that her mother and my grandmother were half sisters, but my grandmother was not widowed until after her youngest was 12.

The breakthrough in finding out that great-grandma had an affair came in the form of two additional Ancestry DNA matches. I had a 3rd cousin showing up with 90 cM, and a tree of only 13 people. However these people lived in Kingsley, MI, as well. My closest 4th cousin with 64 cM also had family in Kingsley. These two also showed up as matching each other. I threw what I could find about their Kingsley family into my tree and gound their common family - Gray, and the great great GF of one of them was the brother of the great GF of the other. Both of the Gray families lived down the road from great grandma's, and two of the daughters os my now presumed grat grandfather Gray were witnesses to my great grandmother's marriage.

What I know is that there were four Gray men, brothers and cousins, living in Kingsley at the time my grandmother was conceived. Based on the DNA relationships between my two matches I took a guess and named one as the culprit.

What I also know is that one of the three Smith brothers is my grandfather. Tested DNA with the daughters of two of the men. The third one is long dead and I know of know children.

geebee
12-03-2015, 05:43 PM
As I understand it, the purpose of "three-parent" babies is simply to allow women who may be carriers of a serious mitochondrial disorder to have children without the risk of passing on the disorder.

All of the child's chromosomes come from just two parents. As I understand it, the third parent is an ovum donor. But in this case, the ovum is only used for its mitochondria; its nucleus is replaced with that of the intended mother. Since the mitochondria contain their own DNA, it will be passed on by any daughters who have children of their own.

The 23 pairs of chromosomes, however, will come from only two parents.

In theory, it might be possible to "fix" damaged chromosomes, however, by replacing individual chromosomes with undamaged ones from another individual, or individuals. In other words, it would be possible to have *more* than three-parent babies.

One more point ... the present form of three-parent babies would not affect NPEs in any way, but they might be seen as NMEs or "nonmaternal events". This is because up to now mtDNA has been seen as tracking the mother's mother's mother's line, and it would no longer do so in the same way.

As a man, my X chromosome and my mtDNA came from the same woman. If I were a three-parent baby, each would be from a different woman.

EDIT: http://www.medicalnewstoday.com/articles/288943.php

It looks as if there are two basic techniques for producing "three-parent babies", but both involve a transfer of the mother's nuclear DNA into the ovum of another woman. The basic difference seems to be whether it's done before or after conception. In either case, only the mtDNA comes from a "third" parent.

geebee
12-03-2015, 06:46 PM
My paternal grandmother was born out of wedlock, though perhaps her birth really shouldn't be considered an NPE because (1) the identify of her father was never any secret; and (2) she would not have carried on his surname in any case.

It also appears that my maternal grandfather's birth may have involved an NPE. Unfortunately, while I've always been able to track up the tree from my paternal grandmother, the only thing that suggests an NPE with my maternal grandfather is DNA evidence.

There are a number of people on my grandfather's side who are related to each other and to me, and closely enough that I *ought* to be able to place them in my grandfather's tree *if* that tree is accurate. By inference, I think I may even have been able to find who may be my maternal grandfather's actual paternal grandfather, but that may be as close as I can get it.

Piquerobi
11-16-2019, 01:33 PM
This is from a recent study on cuckoldry:


Highlights • Combining genetic and genealogical data illuminates our ancestors’ sexual behavior •Gene-genealogy mismatches imply extra-pair paternity (EPP) • Historical EPP rates were low overall (∼1%) but varied depending on social context • EPP rates were highest (∼6%) among urban families with low socioeconomic status

Paternity testing using genetic markers has shown that extra-pair paternity (EPP) is common in many pair-bonded species. Evolutionary theory and empirical data show that extra-pair copulations can increase the fitness of males as well as females. This can carry a significant fitness cost for the social father, who then invests in rearing offspring that biologically are not his own. In human populations, the incidence and correlates of extra-pair paternity remain highly contentious. Here, we use a population-level genetic genealogy approach to reconstruct spatiotemporal patterns in human EPP rates. Using patrilineal genealogies from the Low Countries spanning a period of over 500 years and Y chromosome genotyping of living descendants, our analysis reveals that historical EPP rates, while low overall, were strongly impacted by socioeconomic and demographic factors. Specifically, we observe that estimated EPP rates among married couples varied by more than an order of magnitude, from 0.4% to 5.9%, and peaked among families with a low socioeconomic background living in densely populated cities of the late 19th century. Our results support theoretical predictions that social context can strongly affect the outcomes of sexual conflict in human populations by modulating the incentives and opportunities for engaging in extra-pair relationships. These findings show how contemporary genetic data combined with in-depth genealogies open up a new window on the sexual behavior of our ancestors.
https://www.cell.com/current-biology/fulltext/S0960-9822(19)31305-3

Jenny
01-16-2020, 07:39 AM
I found my biological great grandfather through traditional genealogy.
Looking at migration records, the man called Matti, came to the US before my grandfather could have been conceived. But that very winter, Gustaf, was living with with his brother Matti's. wife.
What can I say, it gets chilly in Finland in January.
No one in my mother's generation ever suspected

Jenny
01-16-2020, 07:41 AM
Deleted

Jenny
01-16-2020, 07:46 AM
I remember some papers (perhaps 10 years ago) from medical studies about NPE. There is a social bias: the lower the social class is, the more higher is the probability of NPE, and reverse. At this point, statistics were saying that near 25% of the children in the lower classes are the results of NPE, versus less than 1% in the higher classes.

Number were revised because the study was made with a sample of divorced mothers (in England, in the early 2000's).

Instead of 25% in the lower classes, it's more 10% (in later studies).

But the conclusion is the same: social bias of NPE.

It's not surprising: a woman has far more to lose in the higher classes than in the lower with infidelity.

How convenient for our betters ;)

Baltimore1937
01-17-2020, 06:25 AM
Yeah, I came to that conclusion all by myself. That is, the lower social classes would more likely have more NPEs. Back when I was little (1940s, etc), but already interested in my family tree, the lower socio-economic classes (to which I belonged) thought family trees were for the snobby wealthy.

geebee
01-17-2020, 02:39 PM
The only thing I can really speak to is my experience with my own surname line. The immigrant who brought our surname to America -- though in its original German form -- was Johan Buchhammer. I have a perfect 46-marker match (GD 0) from Ancestry with the descendant of his son Johannes. I'm a descendant of another son, Jacob. My match is actually from my grandfather's generation, so two steps closer to our common Y-line ancestor. We're 4th cousins twice removed.

I have another match at FTDNA, at 37 markers. However, this match has a genetic distance of 2. The difference, I believe, is mainly due to FTDNA's test including a couple of markers not on Ancestry's test. Of course I have no way of knowing what the results would be with the same set of markers used by Ancestry. In this case, the match is from my father's generation, but our shared Y-line descent is a generation closer -- we're both descended from Jacob, but from different grandsons of Johan. This match and I are 4th cousins once removed.

Of course, this is only one of my lines, but it is also the only unbroken line of males. And I don't have to go very far back to find examples of misattributed paternity on other lines. In one case -- my maternal grandfather -- we actually seem to have a case of both misattributed paternity and misattributed maternity, in the form of an "informal" (and undisclosed) adoption.

razyn
01-17-2020, 06:01 PM
Back when I was little (1940s, etc), but already interested in my family tree, the lower socio-economic classes (to which I belonged) thought family trees were for the snobby wealthy.

Some of that was socially structured, too. For the propertied people, family trees were about how to retain and/or distribute it when somebody croaks. And the paper trails, both public and private, tended to facilitate that. More so for real estate than personal estate, but both are covered if there's a will, and decided by law if there isn't. DNA testing can be a fly in that ointment.

And the real paper trail may be unapologetic about its negligible relationship to the genetic facts of the case. The right to something like a castle may be passed around among distant relatives, in-laws and outlaws -- by naming cousins as heirs, adopting a wife's surname if she's the only heir, acknowledging bastards, disinheriting legitimate sons, and so on.

RobertCasey
01-18-2020, 01:15 AM
I would like to put a very practical and useful application of NPE rates with the following assumptions:

1) We use the widely acceptable overall NPE rate of 2 % over the last 1000 years. You need to adjust this rate for Cohen descendants which are well documented to be much less and probably some European royalty where NPE rates could be lower.
2) The time frame of the establishment of surname usage became widespread - For Ireland & Scotland - probably 1000 AD, for England around 1100 to 1200 AD and for Welsh much more recent. R-L226 is 90 % Irish, so I use 1000 AD.
3) The average number of years per generation is another important factor. Most research shows it is much higher than 25 years for the average years - 30 years is more acceptable (I use 33 years to be more conservative).
4) This results in 30 generations - at 2 % NPE rate (any scenario where surname does not track YDNA) - this would be a 54.5 % NPE rate. I round this down to 50 % for simplicity.

Why is this important - declaring surname clusters (at a YSNP branch or YSTR branch) and much more accurate TMRCA estimates based on surname clusters and variable rate years per YSNP which can be determined.

I declare a surname cluster when two criteria are met:

1) At least five testers with the same surname belong to any branch.
2) No more than 50 % NPE rate (believable amount which will vary).

With this definition, the 843 Y67+ testers under L226 have 25 surname clusters. This really helps charting to let people know if there testers share a common ancestor with the same surname. It also standardizes the methodology of when testers share enough testers with the same surname to be related in the last 1000 years (Irish only). With this information in hand, I actually calculate the years per YSNP mutation based on the number of mutations between L226 and the surname cluster. The variable years vary from a low of 48 years (500 years / 10.5 YSNP mutations = 48 years). It is 10.5 YSNPs as the surname cluster is a YSTR branch between two YSNPs and does not begin with any YSNP mutation. The maximum years per YSNP is 302 years (one YSNP is established to be 48 years, the other 1.5 YSNPs to get to the surname cluster. (500 - 48 / 1.5 = 302 years). The variable years per YSNP is only used for YSNPs between L226 and the 25 surname clusters. The average (now down to only 70 years per YSNP which continues to decline over time with larger sample sizes) is used for branches not between L226 and surname clusters and for YSNP branches below surname clusters. We already have one TMRCA of 1490 AD (7 YSNP levels below the YSNP where the surname was declared).

So there are two very practical usages of NPE rates:

1) Consistently declaring surname clusters (which L226 project members now embrace).
2) Assigning TMRCA dates which are much more accurate than mathematical models based on YSNPs or YSTRs. The variation of years per YSNP is accounted for vs. reporting 500 AD to 1500 AD for TMRCA estimates.

This is only possible with charting:

http://www.rcasey.net/DNA/R_L226/Haplotrees/L226_Home.pdf

Tolan
01-18-2020, 07:44 AM
I would like to put a very practical and useful application of NPE rates with the following assumptions:

2) The time frame of the establishment of surname usage became widespread - For Ireland & Scotland - probably 1000 AD, for England around 1100 to 1200 AD and for Welsh much more recent. R-L226 is 90 % Irish, so I use 1000 AD.



I do not think so!
The surnames (same as the father) must be fully generalized for the stats to be valid.
And I think we have to wait the 14th century!
The surnames may have existed before, but were not inherited from all the sons!

RobertCasey
01-18-2020, 04:40 PM
R-L226 is more aggressive with 1000 AD since it belongs to the royal line of King Brian Boru and his royal ancestors. This surname started around 1000 AD and includes several clans associated with his brothers as well. We have 92 Y67+ testers that belong to the O'Brien surname cluster (the O'Brien clan chief has Big Y tested as well). Kennedy (22) , Casey (23) and Ahearn (6) are known clans associated with Brian Boru's brothers are very genetically close as well. Of the 782 Y67+ testers that are charted, 143 (18 %) belong to surnames established around 1000 AD. Many other surnames may have started around 1100 AD but 100 years only affects the TMRCA dates which are a huge improvement over YSNP and YSTR TMRCA models. The more important issue is the formation of surname clusters when any particular surname for any branch (YSNP or YSTR) goes above 50 % (making the NPE rate fall below 50 %).

The NPE rates for the last 1000 years are pretty well covered at 2 % (where surname tracks YDNA). The TMRCA estimates from YSNP and YSTR models vary from 500 AD to 1500 AD, so even if some surnames did start at 1100 AD or 1200 AD, this would be a massive improvement in TMRCA dating. Even the NPE rate should be variable as well (there could be an early NPE within a surname cluster). I started adding surname clusters at 1050 AD where there are six testers with only two surnames. I could add a variable date for surname creation: 1000 AD for lines tied to the royal O'Brien that track genetically as well and then use 1100 AD for others. But this would add a lot of confusion just to gain a minor improvement in TMCRA dating - which is far superior to TMRCA estimates from YSNP or YSTR based models (for TMCRA dating below predictable haplogroups). The YSNP TMRCA methodology works well for predictable haplogroups (or older). I think the YFULL approach is the best for predictable haplogroups. Here is a summary of my TMRCA methodology:

http://www.rcasey.net/DNA/Temp/L226_TMRCA_Calculations_20191005B.xlsx

FionnSneachta
01-19-2020, 01:36 AM
Just as a comparison, the percentage of NPE Big Y testers in our surname group is 37.8% of our surname group.

RobertCasey
01-19-2020, 09:01 PM
Just as a comparison, the percentage of NPE Big Y testers in our surname group is 37.8% of our surname group.

I believe that any YSNP branch or YSTR branch that falls below 50 % NPE rate is reasonable. At a 2 % NPE rate, 30 generations is a 57 % cumulative NPE rate. However, since Big Y penetration of testing is only
around 20 % of the Y67+ marker testers, it is much better to include the other 80 % of testers that can be predicted via signatures and charting. Your branch FGC6550/FGC6545 is ideal for a predictable haplogroup:

1) The TMRCA is between 1500 and 2500 YBP (YTree has 1760 YBP).
2) Your haplogroup should be genetically isolated having 26 branch equivalents.
3) It is well tested for YSNP (42 NGS tests via BigTree and 27 branches below)
4) Assuming 20 % are Big Y tested, you should have around 210 Y67 marker testers.
5) Can be charted with SAPP (less than 500 testers).
6) Is dominated by one surname (Kelly) and four earlier branches are non-Kelly testers.

FionnSneachta
01-19-2020, 11:14 PM
I believe that any YSNP branch or YSTR branch that falls below 50 % NPE rate is reasonable. At a 2 % NPE rate, 30 generations is a 57 % cumulative NPE rate. However, since Big Y penetration of testing is only
around 20 % of the Y67+ marker testers, it is much better to include the other 80 % of testers that can be predicted via signatures and charting. Your branch FGC6550/FGC6545 is ideal for a predictable haplogroup:

1) The TMRCA is between 1500 and 2500 YBP (YTree has 1760 YBP).
2) Your haplogroup should be genetically isolated having 26 branch equivalents.
3) It is well tested for YSNP (42 NGS tests via BigTree and 27 branches below)
4) Assuming 20 % are Big Y tested, you should have around 210 Y67 marker testers.
5) Can be charted with SAPP (less than 500 testers).
6) Is dominated by one surname (Kelly) and four earlier branches are non-Kelly testers.

I actually only have 53 Y-67 STR matches. There are 71 in the DNA results page for the project but that also includes Y-37. If I include everyone in the group of a different surname to Kelly among the Y-67 matches, that gives a total of 22 which gives NPEs as making up 41.5%. If I exclude the Traynor surname from both NPEs and total (since I don't think that it's actually an NPE), that gives 31.1% (14/45). The reason that I did the percentage of NPEs from Big Y rather than at the STR level is because some Y-67 matches could be from before the introduction of surnames and the Block Tree displays if someone is likely an NPE or if the shared ancestor is before the introduction of surnames. I'm not at all disputing the statistics though. I was just giving the example of the NPEs in my own surname group.

FionnSneachta
01-21-2020, 12:03 AM
1) The TMRCA is between 1500 and 2500 YBP (YTree has 1760 YBP).
2) Your haplogroup should be genetically isolated having 26 branch equivalents.
3) It is well tested for YSNP (42 NGS tests via BigTree and 27 branches below)
4) Assuming 20 % are Big Y tested, you should have around 210 Y67 marker testers.
5) Can be charted with SAPP (less than 500 testers).
6) Is dominated by one surname (Kelly) and four earlier branches are non-Kelly testers.

Thanks for mentioning the SAPP tool. I had seen it before but never really tried to figure out how to use it. Between you mentioning it and it being mentioned in another thread, I decided to give it a proper try. I had fun creating a tree for all the members of our surname cluster using David Vance's version. It's interesting to see where there might possibly be new sub-branches for testers who haven't done SNP testing. I'll have to edit it and add names to make it easier for me to read.