One of our blog followers, Tom, encountered the following situation, which, really isn’t so uncommon.
“I started up a Y-DNA surname project and recruited my only three living male 1st cousins who carry that name.
The first set of 67 results have recently been posted and within a day an exact match appeared with an individual who had only tested to 37 markers back in 2008. Apparently the individual has had very few close matches and never a perfect match like my cousin’s. But the individual has a different surname.
Is it possible to have an identical match as a random event? How common are such occurrences? What possible explanations could there be?”
Let’s look at Tom’s situation from different perspectives and see what we can find.
When I do DNA Reports for people, I still find people who don’t have any matches. It’s not as unusual as people think. In a way, it’s a blessing as compared to people who have so many matches that they can’t even begin to sort through them. But to the person who doesn’t have any matches, it surely doesn’t seem like there’s a positive side to the situation.
First of all, remember that mutations can happen at any time in any generation….or not. In the Estes line, Abraham Estes, one of two Estes immigrants to colonial America was born in 1647. He had 8 sons. We had DNA from the descendants of all 8 sons. We reconstructed Abraham’s DNA using triangulation, so we know what his original genetic “signature” looked like. One of those sons’ lines has 4 mutations in 8 generations, and one line has none. The rest fall in the middle someplace.
I only mention this to illustrate that mutations are truly random events. We use statistics to look at the “most likely” scenario, based on averages, but mutations are personal events and while they, as a whole, fall nicely into statistical models, individually, they happen when they happen. You can see that the mutation rate can vary quite a bit, even within families. Keep that in mind during the rest of this discussion.
Family Tree DNA gives us some tools to work with these kinds of situations. The TIP calculator, available for every match by clicking on the little orange TIP button, tells us, statistically, how likely people are to match at which generational level. This is called the time to the most recent common ancestor, or MRCA. I did an earlier blog about this.
Comparing two exact 37 marker matches, below, we see that, statistically speaking, on the average, these two people are most likely to share a common ancestor about the second generation, meaning grandparents. Again, word of caution, these are averages, which is why you have a range shown here. DO NOT TAKE THEM LITERALLY. I can’t tell you how many people obsess over these numbers and think that these numbers are telling them exactly when they share a common ancestor. They don’t.
So let’s answer the questions that Tom asked.
Is it possible to have an identical match as a random event?
No, it’s not. These men share a common ancestor at some point. The question is, when and where. However, it is possible to match on many markers, and then not on others. I would suggest that these men upgrade to 111 markers and see how closely they match at that level. I have seen at least one instance where 2 men matched at 37 and then had 5 or 6 mutations at the 67/111 marker level. Unusual? Yes. Impossible? No.
How common are such occurrences?
This isn’t as unusual as you would think. I see this fairly often. I always tell people to do four things.
First, upgrade the people you have to 111 markers. If they continue to match at 111 markers, exactly, you probably have a very close match genealogically.
Second, find a second person to test from each line, as far back as possible. In other words, if you’re testing the Abraham Estes line, you would want to find another son from Abraham to test to see if the DNA of the two sons match. If they do match, then you know you have the lines proven back to 1647. If not, then you know you have a non-paternal event (NPE) of some type, otherwise known as an undocumented adoption. I call them undocumented adoptions because everyone knows what that means, and regardless of how it happened, it’s “undocumented” because we didn’t know about it. In Tom’s case, he already has his 3 cousins, so his line is proven back to the common ancestor of those men. Hopefully the person with the other surname can also find someone else from his line to test.
Third, enter the results into Sorenson at www.smgf.org and also into www.Ancestry.com for Y-line results to see if you come up with any other people who also match with that surname. This is especially useful if you are having difficulty finding people to test.
Note: Neither SMGF nor Ancestry’s Y data base are available.
Fourth, look around the neighborhood – genealogically. Are there reports of the two families in question being allied or intermarried in some location? Were they neighbors in the same county? In many cases, once you figure out that an undocumented adoption occurred – you can figure out in which generation through selective DNA testing, and often, which families were involved through DNA combined with historical and genealogical records.
In essence, to solve this type of puzzle, you need to become somewhat of a genealogical detective.
Tom’s last question was what kinds of situations could explain these results.
In some cases, especially where there are some mutations involved (meaning not exact matches), suggesting some time distance between common ancestors, matches between surnames occur because the families involved simply adopted different surnames. When surnames were adopted varies dramatically by the location of the families and the circumstances involved. For example, in the US, some Native American families were still using Native names in the 1880s. Freed slaves adopted surnames upon obtaining their freedom in the late 1860s and early 1870s, and sometimes changed that surname at will. In the Netherlands, some families didn’t take names until in 1811 when Napoleon mandated that they do so. In England, some wealthy families had surnames by 1066, but peasants didn’t adopt surnames, for the most part, until in the 1200s and 1300s. Jewish families in some parts of Europe didn’t adopt surnames until in the early 1800s.
Sometimes surname changes that look to us like undocumented adoptions occurred not at birth, but later in life. Some people simply changed their names for a variety of reasons.
In one case, a man named John though he killed a man in Tennessee, ran off to the frontier which was at that time in Texas and changed his name, only to discover years later that not only had the man he shot not died, but that man had then married John’s wife he abandoned when he left. Hmmm….karma at work.
One of the most common reasons for ‘undocumented adoptions’ is that a step-father raised the children and the kids simply used his name….forever. So maybe John’s kids, above, took the surname of the man John shot. Now this is getting interesting!!! No wonder we have trouble figuring these things out retrospectively.
Another reason, of course, is that illegitimate children took the mother’s surname, but carry the father’s DNA signature. In Native American cultures, matrilineal naming was very common, as is it in the African American culture, especially immediately after the end of slavery. Children took whatever surname their mother adopted at that time.
Of course, the one thing we haven’t mentioned is the obvious….where someone was unfaithful. Generally, that’s the first thing people think of…but it’s really not the most common reason. But sometimes, indeed, it appears that Granny might have been a “loose woman.” Don’t judge Granny too hardly though, because you really never know what happened in Granny’s life. She could have had no choice in the matter, or her husband could have been abusive. We often see these conceptions during periods of war, especially the Civil War. In one case, we know that a woman exchanged sexual favors for food for her children. It’s really hard to be critical of that woman.
So the real answer for Tom is that there is no cut and dried answer, but lots of possibilities to explore. You’re going to have to get out your Columbo tools and sleuth away….
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- Legacy Tree Genealogists for genealogy research
Thanks Roberta for your insight into this issue. I have a similar and maybe more complex issue.
My Murtishaw line has been fairly well documented in the United States back to enlistment in the Delaware Regiment for the RW, but we have found no one of this name elsewhere – except a place called Murdishaw Wood in Cheshire England.
I started a project with my brother’s test, hoping to entice some men named Mottershaw in England to join – no luck as of yet. But he matched 67/67 with a man named Wood from PA. We also match 66/67, 65/67, and 64/67 with men named Barlow; 65/67 and 64/67 with Crooks; 65/67 with Vonada; 63/67 and 62/67 with Ainsworth; 63/67 with two men named Hollingsworth, one named Gregory, and one named Middleton. Some of these men match each other just as closely and some are somewhat more distant. We (Wood and Murtishaw) appear to be the modal. We are searching for places we may have intersected and some of us think that may have been in Cheshire/Lancashire England area .
Other than Jewish ancestry, have you seen this many close matches with different surnames?
Yes, I see that often, especially in British Isles ancestry. When I do DNA reports for people in this situation, I create a chart for them with their rare and very rare markers based on a frequencey chart that I’ve created over the years for each marker in each haplogroup. Then you can use that rare and very rare frequency “profile” to determine who is actually closer and further away from you.
What’s the chance that there was a sort of “John of Murtishaw Wood” naming going on, and it got simplified on both trees, just differently? I have a line where the surname is Richmond Webb, but it gets written as Richmond-Webb, Richmond, Webb and Richmond (Webb) – I think someone thought it was a maiden name in that last one….
Roberta, I’m very confused by what you said about the comparison chart.
“Comparing two exact 37 marker matches, below, we see that, statistically speaking, on the average, these two people are most likely to share a common ancestor about the second generation, meaning grandparents.”
Can you explain that? So far I have no one that is a perfect match, but I have found 2 people that match 36/37, and a couple that match 66/67. I thought the common ancestor would be closer to 100%. What am I missing?
It is 100% certain that you share a common ancestor. But the MRCA percentages are showing you the most likely percentage that you share a common ancestor at any particular given generation.
As for multiple surnames, don’t leave out the Scandinavians. The traditional naming pattern was first name plus patronymic. There was no surname. This pattern continued until the late 1800’s in rural Norway. I have one family of 5 brothers and a sister who emigrated and took several different surnames in America. The sister, Synnove Johannesdatter (something close to Mary Jones for 19th century western Norway) married, so she doesn’t really count with this. The 5 brothers took 3 different surnames: Johnson, Foley and Ferley. The latter 2 names are Americanizations of the name of the farm where they lived in Norway. This story is also a lesson in not jumping to conclusions. Foley is a good Irish name but these Foleys are of Norwegian descent. Without knowing the story, who would guess that John Johnson and Peter Foley were brothers, and Norwegian immigrants?
This is an interesting discussion and it has been very helpful.
I have a YDNA 37 marker test that has a difference of 4 from another person’s 37 marker test. Does that meam there are 4 markers that are different by 1 or one marker that is different by 4 or can it be any combination of marker differences that add up to 4? His surname is not the same as mine.
It means there are a total of 4 mutation steps. All 4 could be on one marker, but that would be extremely unusual, especially with no other mutations. Generally, they are on different markers, but not always.
Your story about John in Tennessee who shot a man and ran to Texas could be my GGgrandfather. We tested as Andrews, but grew up as Johnsons. Is your story true? If you don’t mind, what were the last names? Karen (Johnson) Bernard. Thanks.
Yes, it’s true but the last name is Estes.
Here’s a question from our recent YDNA testing on an Anderson line. The cousin tested (DYS37) matches no one with the Anderson surname at 12 markers, at 25 markers or at 37 markers. But in the Anderson surname project his test doesn’t look very different from other Anderson men in his haplogroup. If we upgrade to 67 or 111, is it possible that we might find an Anderson match at that level?? It would be just the opposite of your example, “. . . where 2 men matched at 37 and then had 5 or 6 mutations at the 67/111 marker level.” I had thought if he didn’t match at 12-37, he never would. So we did the FamilyFinder test instead of upgrading to 67 or 111. What do you suggest?
Hi Janet, Yes, it’s possible. If all of his mutations fell in the first few panels, instead of the last several. I have seen that happen, but it is rare. Count the mutations in the first panels. The maximum at 111 markers is 10 that would she shown as a match. So if he has more than 10 in the first several panels, then, no, he won’t match at 111 either.
My paternal half-brother, Ryan Hughes, matches 37/37 to two different men with the McMurray surname and matches 25/25 to a third McMurray surname, but not a single match to anyone with the Hughes surname (he also matches 36/37 and 35/37 to MANY McCracken surnames). After his results were posted it really confused me so both he and I both did the autsomal Family Finder and we matched as half-siblings. I also matched to a confirmed Hughes cousin going back to our 3rd g-grandparents, Franklin Hughes, but my brother does not. I have had no luck finding the parents of Franklin Hughes. Since my brother and I match as half-sibling and I also match a Hughes cousin going back to 1813, I assume the NPE could have been with Franklin Hughes or a generation prior? I have searched the area where I think Franklin Hughes was born, but have not found any McMurray neighbors. According to FTDNA’s TIP calculator he should have a common ancestor with a McMurray at 83% within 4-generations and 89% within 5-generations. I am so interested in discovering where this possible NPE may have taken place, although it doesn’t seem very likely that I will be able to determine where/when it occurred.
It’s also possible that you and your brother did not both inherit the same “Hughes” gene. It would be good if you could find a closer Hughes cousin and ask them to test, the closer, the better.
Thank you for the suggestion. Unfortunately, there are only two older Hughes’ left in the family. One is my father, the other is a male cousin – neither of them will test. Oh well….
I have Y-67 and AT results from FTDNA. Repeated attempts to access Sorenson and Ancestory have failed. I have a pwd to Sorenson and it recognizes me but I don’t get any further, except to learn that the database is closed to new data and contains one surname. Have an account with Ancestory but cannot find a database. Any further suggestions?
I still access the Sorenson free data base at http://www.smgf.org regularly. Regarding Ancestry, I would call them on the phone. Good luck.
Pingback: Projects, Administrators and Expectations | DNAeXplained – Genetic Genealogy
What dose a ydna test resuilt of 19 out of 20 mean
Generally it means they match someone at 19 of 20 markers, but I don’t know of any companies that test at 20 markers, so this is a rather odd comment. Given that, I’m not positive I’m interpreting the question and information correctly.
My surname is Jackson, took ftdna 37 ytest match with only one Jackson 35-37 , but one. And Houghton surname 36-37, and match 35-37 with 2 other Houghtons and 34-37 with four other Houghtons what does this mean?
It means that sometime in the past that the Jackson and Houghton families are related.
Pingback: The Orphan Train and the Mystery of William Jennings Duckett | DNAeXplained – Genetic Genealogy
i dont know much about my fathers family. I am born in Bulgaria. I did recently my 12 y test and found over 300 exact matches and they are all in uk, scotland and ireland. Can that be a mistake and how can i take things further. Kras
A high number of matches at 12 markers means that you probably don’t have any rare markers in the first 12, so you match a lot of people who are related a long way back. You need to upgrade to at least 37 markers. You’ll have fewer matches and they will be more relevant. You may also need to upgrade to 67 eventually, depending on what your 37 marker matches show.
Thanks for a great article. I want to mention that also the Portuguese have somewhat unpredictable surnames and can have similar results as those from Norway. I found someone with a perfect 12 marker match to my father who has done 111 marker testing. I wouldn’t have been so interested in that except that this person’s oldest known ancestor was from the same general area (island of Terceira, Azores). So I wrote to ask more details and found that this person now uses the surname in the USA of Lawrence. It was Lourenço in Portugal. They had enough information for me to trace this family back and I discovered we have the same direct paternal line with the most recent common ancestor born in 1751. Some of his children and grandchildren honored him by taking his given name and tacked it onto their surname. This resulted in many with the surname combo Rocha Lourenço. While mine continued with the surname Rocha.Homem. This same common ancestor, Lourenço da Rocha Homem added to the confusion by adopting the surname Homem, possibly because he might have worked for a very wealthy family named Homem da Costa e Noronha (the godparents for all his children). So I have a paper trail that says we are related, but the upgrade to their 12 marker test has not happened. So my question to you is should I expect a perfect 37/37 match with this person? I read that mutations can happen in any generation, so I guess it doesn’t matter what you guess. Only the test will give me the answer.
This is an interesting outcome due to DNA testing. I recently had a 111 DNA test done at FamilyTree DNA and it brought back two 67 marker matches (that’s all they tested) to two different individuals. All three of us have different surnames. FamilyTree states that we have a 95% common ancestor at level six. We can go back five to our grandfather Albert Francis but hit a roadblock. One of the 67 marker matches has a direct line to Joseph Johns, founder of Johnstown PA (he is level six). Their are no records in the Johns line to indicate Albert existed. Would it be that Albert was an undocumented adoption and Joseph Johns was his father? Also, would these two 67 marker matches be 1st cousins six times removed? Thanks for all the great information.
You are taking the MRCA and TIP numbers way way too literally. Remember, mutations can happen, or not, at any time. The RMCA and TIP numbers are estimates constructed from averages, only.
Are you saying we aren’t related in a genealogically time period of approximately six years? If DNA results cannot be taken literally why test? My understanding of mutations is based on them occurring over several generations based on average data obtained in an extensive data base such as FamilyTree DNA’s. I know that DNA science is not 100% due to mutations and generational time periods can vary, but our results indicate that we are related we just don’t have a 100% accuracy on when, where and who. I also know what time period to look in 1812/1813. The search continues. Thank you for your all your information.
No, I’m not saying that. What I’m saying is that you can only use those as guidelines. I look at the 50th percentile, but you could also be in the 25th or 75th. Here’s an example. http://dna-explained.com/2012/08/06/what-does-mcra-really-mean/
This next article speaks directly to your situation. http://dna-explained.com/2012/11/29/averages-tip-calculator-and-one-size-fits-all/
Doing research on random mutation events, I came across this new video (http://wn.com/are_mutations_random) regarding the randomness of gene mutations and may not be as random as past predictions. New discoveries happen all the time now and do you think this may lead to a better understanding of what gene mutations are?
Our family had this occur recently when a relative had a y-dna testing done. We had hit the proverbial brick wall in our research on our shared ancestor so a male descendant (cousin) was tested hoping to learn something new. The cousin was surprised when the test results came back matching us with a group of men with a totally different surname.
The donor matched the modal for this grouping perfectly at 111 markers.
Our shared ancestor did live closely by people with the same last surname so it was suspected the NPE may have been around this place/era with someone from this line as the paper trail ends and no parent has neen identified. But so far all males tested from this line living near our ancestor seem to carry a mutation that our ancestor doesn’t. The one direct descendant from the matching surname line who has upgraded to 111 markers had the same mutation but matches the donor on all the rest.
Needless to say but i will anyway.
We are currently looking for more descendants to test.
It happens. At least you know now and can deal with it. Nothing like doing the wrong genealogy for who knows how long.
I posted this before, but don’t think it posted….
Yes, and there are also the Welsh patronymics, which throw a couple of different monkeys into the barrel. Child ap Daddy sometimes changed his surname before he left Wales, and sometimes he waited until after the move to America. So if you have Welsh in your paternal line, don’t automatically assume there’s an Oops.
Our line here is Thomas. The great majority of our Y matches are Thomas, Griffith and Pugh, which tells me that any of the following could have moved here and been the American patriarch: 1) There was a daddy named Pugh in Wales; 2) there was a daddy named ap Hugh in Wales; 3) there was a daddy named Thomas in Wales; 4) there was a daddy named ap Thomas in Wales; 5) there was a daddy named Griffith in Wales; 6) there was a daddy named ap Griffith in Wales. So let’s look at the possibilities that I could be looking for – Thomas Pugh, Thomas ap Hugh, Hugh Thomas, Thomas Hugh, Thomas Hughes, Thomas (ap) Thomas – Thomas Griffith, Thomas ap Griffith, Griffith Thomas, Thomas Griffiths…. and we don’t know whether they were here or there, so we have to look at all the names in both places, PLUS all over Wales. Because, as you know, Thomas being the 7th most popular surname puts it in the “There are Thomases who left for American from every Welsh province” category. And in PA, we get to have German Thomases, too; luckily their main choices for given names are not John Seth William David Robert or Edward.
Oh, and since any of the aps *could* have taken the name of their parent, it makes it worse – John ap Thomas’s children were all named Jones, not Thomas.
There’s more, but I’m dizzy just from reading this.
WordPress.com / Gravatar.com credentials can be used
Pingback: DNAeXplain Archives – Basic Education Articles | DNAeXplained – Genetic Genealogy
I had my son y-dna tested. He is of Italian descent with his line paper documented back to the 1700’s. The test came back and I am told he is of Irish, Swedish, English and United Kingdom descent also with surnames that match my ancestral surnames. When I questioned this I was told that my lineage crossed path with my sons and had a relationship causing the surname matches. I can see where the surname would be different from an affair because of the mother and adoption but not matching my line.
I have had 3 males do the ydna for my side of the family only to have each one come back with different surnames. I can deal with this as I do believe my great grandfather was adopted or illegitimate. Having my son’s ydna come back and match my father’s side of the family is just too much for my understanding when his relatives are all from Italy.
Could you make a comment on this?
Well, we did our Y-DNA for our Hughes family and it did not match ANY Hughes, or anybody in the Hughes surname project on Family Tree DNA…. we match to German names. ?????
Pingback: Concepts – Undocumented Adoptions vs Untested Y Lines | DNAeXplained – Genetic Genealogy
I recently did a y-DNA test from FT DNA. I had a paper trail for my paternal line back to the late 18th Century. There was good evidence from the 1880 census that this ancestor was born in Germany. So, I was surprised when I got my results and matched perfectly (37/37) with someone who had a very different (English) last name.
I fretted and wondered if my love of Bratwurst and the language (I’m a German speaker) was somehow all a lie. I tried to think of improbable ways that the same paternal line could have existed in Germany and England.
When my match finally wrote back he explained that he was adopted and that court papers revealed his birth father and I shared the same last name.
So, to reiterate what this article says: don’t jump to conclusions!
This would be my line …. looking for the NPE event
I agree strongly that an exact Y DNA match at 37/37 or higher is no certainty of a recent or close kinship. I have a 67/67 exact match and also another 111/111 exact Y match. I searched for two years to find this “guaranteed” close match but nothing connected or even came close to a paper trail that led the families within a thousand miles of one another. I also confirmed the lack of a paper trail with a family finders test which showed no relationship with either exact Y match. Hope this helps someone!
Through DNA testing around 3.5 years ago, I learned that my birth father was almost 100% Finnish, with a small amount of Volga-Ural Russian. I have recently identified my birth father through a DNA match. He was born in the US and both parents were born in Finland. He was also a married father of 4 children when he raped my 14-year old birth mother, who was baby-sitting his youngest daughter – now my half-sister. My DNA match is her daughter. My family and I are adjusting and my eldest son said, “It is what is is and we wouldn’t be here, otherwise.” Both my birth mother and birth father are now deceased.
I have around 3,000 DNA cousins in Finland. When I notified one of my closest DNA cousins, he was able to quickly identify my birth father’s paternal/maternal Finnish ancestry. My dilemma, which is very upsetting to me, is: In order to identify my birth father in Ancestry and other genealogical family tree programs, he has to be listed as a spouse of my birth mother, which he clearly was not! I’m not sure how to proceed. What do you suggest?
First of all, I am truly sorry for the circumstances in which you were conceived. Having said that, I realize how upsetting that is to you, but your only other alternative is to create two different trees, one for your mother and one for your father, and attach your own DNA to both parents. I realize this would make all of the estimates and relationships a generation “off.” However, that is one way around the problem. You could also simply think of this as your biological tree and nothing more.
Thank you, Roberta, for your thoughtful reply and your suggestion is something I’ll definitely consider. I think I’ll also consider contacting Ancestry because as soon as I identified him as my birth father, Ancestry automatically made him a spouse of my birth mother. In all of recorded history, there have been millions of children who were born to individuals who were unmarried to each other and in a consensual relationship or who were not in a relationship and the conception wasn’t consensual, as in my case. Plus, my biological father was also an adoptee like me. Who knew creating a family tree could be so darn complicated?!! With all kind wishes to you…