Your privacy setting should be public (not private)

[peterebay]

Unless you have the Atlantic Model Haplotype, I highly recommend that
you change your notification setting from the default of private to public. Doing so will only reveal your email to other Family Tree DNA customers you are a very close match with. Finding matches is the primary reason most people tested (correct?)

More precisely, in Update Contact Information in each person's myFTDNA there is a check box to "restrict match notifications to your surname project." This should be not be checked.

It appears that more than half of FTDNA customers keep this setting private. So, we have all spent a lot of money and we are not seeing the whole picture.

Through matches with different surnames I've been able to determine a likely area of origin (southern Ireland) for a Roberts family cluster in my Bahamas
DNA Project. Matches with McCarthy, Hinds, Owen, Donahue, and Meagher all point to the Eoganachts of Munster and matches with Dalton and Butler also
indicate that area. See http://home.comcast.net/~libpjr1/bahrobdna.htm

I strongly suspect we all could learn a lot more if additional people changed their default setting so it was not private. It would also be most helpful if you add your results to www.ysearch.org and www.base.org Ask your surname admin to add you to these databases if you have trouble. If you want to be private, ask your surname admin to be the contact person for you.

Sincerely, Peter

Peter J. Roberts
The Bahamas DNA Project

http://home.comcast.net/~libpjr1/bahamasdna.html

[bob_allison]

People changing to 'public' should be aware that any matches they find outside their surname are likely to be the result of convergence (i.e. they are not related within the last, say, 1000 years). This is particularly true if a person is close to the modal value of their haplogroup.

The other thing people should be aware of is FTDNA's policy of reporting matches, say 12/12, when each participant has taken the 25 marker test which shows a large genetic distance. It also reports close 25 marker matches when each has taken 37 marker tests which do not 'match'.

Regards,

Bob

[T E Peterman]

If people with different surnames match 12/12, disregard it. It is likely meaningless -just convergence.

If people with different surnames match 25/25, take careful note of it. Both should upgrade to 37 & see if the match holds. If so, they have most likely discovered a hidden case of surname switch.

I have already found a couple of these in people that I have had tested, but would have NEVER discovered these if I had left the results set to private.

My Wilson line appears to be McCown.

My Eagleton line appears to be Petticrew.

Timothy Peterman
Kansas City, MO

[EBurgess]

With regards ot Mr Peterson's comments.

I am unclear as to how a person can be certain of the "true" surname for an observed Haplotype. We must be careful when drawing these conclusions.

Haplotypes can converge as already mentioned, only combined with a fully tested haplogroup can we start to rule out a random match.

Historically, wasn't it the case in Scotland that Clan allegiance determined surnames of clan members? Even when no close blood relation existed between them?

Also unless the document trail leads back to the early 13th century, how can you be sure the switch hasn't happened twice! Especially with names that are side by side for centuries.

E.

[Jim Denning]

Also unless the document trail leads back to the early 13th century, how can you be sure the switch hasn't happened twice! Especially with names that are side by side for centuries.

E.

you can be sure by getting 20,000,000 people tested or more the more the merrier
the more matches the better the paper trails involved the more the pieces of the puzzle come together

[T E Peterman]

E. Burgess -actually no haplotype has a "true surname". The haplotype represents the scientifically valid identity here. The surname was merely attached to it a few centuries ago. If the haplotype holds together at really high resolution levels, I suggest that we are looking at actual kinship rather than merely convergence. Worded slightly differently, the odds of convergence become astronomically small when haplotypes are studied at higher resolutions. The likelihood of a surname switch is a lot greater than the chance of a 37 marker convergence.

If after, as Jim Denning suggested, 20,000,000 people have been tested, we could then look at a haplotype & note that say 75% have surname X, 20% have surname Y, and 5% have surname Z. I think it would be reasonable to assume that X came first, and that Y and Z are later branches.

A prominent genealogist once said that when paper trails get cold, the usual culprit is surname switch. Now, at the time, the cases that I thought he was referring to was when a young widow remarries & the children by her first husband grow up in a household with a different, undetermined, surname. But this could also pertain to changes in the patriline. Hidden adoptions, people taking on a new surname when they migrate.

High resolution y-DNA analysis will point to possible surname alternatives to investigate. With some good research (& a bit of luck), paper evidence of the adoption/ surname switch might be located.

Y-DNA analysis won't find the needle, but it could very well tell us which part of the haystack we need to be checking.

Timothy Peterman

[Jim Denning]

If after, as Jim Denning suggested, 20,000,000 people have been tested, we could then look at a haplotype & note that say 75% have surname X, 20% have surname Y, and 5% have surname Z. I think it would be reasonable to assume that X came first, and that Y and Z are later branches.

A prominent genealogist once said that when paper trails get cold, the usual culprit is surname switch. Now, at the time, the cases that I thought he was referring to was when a young widow remarries & the children by her first husband grow up in a household with a different, undetermined, surname. But this could also pertain to changes in the patriline. Hidden adoptions, people taking on a new surname when they migrate.
Y-DNA analysis won't find the needle, but it could very well tell us which part of the haystack we need to be checking.

Timothy Peterman

what is going to be important is when where
if i match west and i show we both have people in Bristol England in late 1500s the more people that come down the pike testing the more locations
whats needed besides dna testing is larger paper work studies like i am doing in chelsea ma. as a northerner i notice mentioning a southern line like my malones brings detailed answers with specific answers why?
because back in the 1800s they did this. when northerners cept for the mayflowers did care.
you can do genetic studies of mayflower people and say a small town in NC and get a alot of relavent info and matches .imagine if that type of stuff was available for new england cities. or bristol england. surnames shoot out in a straight line but its both lines that means something
imagine if your mtdna matches someone in bristol 1580 and the same person has the ydna . how do you know
by finding female decendent of the wife of a brother of your ydna
in chelsea many times brothers married the sister of their brother in law
especially in small towns like woburn ma.


add this to the pie and maybe we dont have to disregard any matches

[EBurgess]

Mr. Peterman thank you for your followup post I think I now better understand your previous post.

I agree with everything you said except the part about X,Y and Z. If I misread again feel free to correct me on this. I think this point is important for people to understand.

If people bearing the same surname present different haplotypes, you cannot assume that the most numerous haplotype is necessarily the oldest and therefore correct one ( I think we agree on this point). It may just be the case that through historical accident one line became

Likewise and this is for the X,Y,Z argument. If you have the same haplotype but different surnames, the smaller group may not necessarily be the youngest. I don't know how much statistical weight 20 000 tests would have. But, would you not agree that many male lines from the 14th century have died out, and some are far less numerous than others?

Now, the conclusions can change depending on what paper trail exists for each line and the type of surname whether locative or not.

My surname Burgess is rather generic and it would appear that genetic evidence, so far, bears this out. There are several haplotypes. Currently the Burgess population of the US is at least double what it is in the UK. Also I have a 23/25 match with a Stewart who is well documented for 200 years in the US. I have a 200 year old paper trail and a genetic match with a Burgess in the UK. What conclusions can I draw beyond that we may be related in the last 1000 years? If further testing shows that we indeed share a genetic relationship and he has 50 Stewarts that match and I have 10 Burgesses. What can be said of that result?

If I had a name like Wildgoose which is rare and Di Napoli which is a placename in Italy. I think you would be on safer ground drawing conclusions about genetic evidence. It would certainly be an interesting story to tell!

E.

[T E Peterman]

At the 25 marker level, I think different surnames matching on only 23/25 is likely convergence.

The really scientific approach would be to identify biological patrilines, first by the most high resolution SNP available. I would like to see genetic testing get several layers beneath merely R1b. This yields the haplogroup. Second, within each high resolution SNP clade, we could identify known high resolution y-str markers. As you know, this yields the haplotype. Finally, we would look at the surnames that testing has yielded within each haplogroup/ haplotype combo.

An important difference between genealogists & geneticists is that:
1. genealogists begin at the present time & work backwards.
2. geneticists begin at the earliest time (where all humanity was one haplogroup) & work forward (defining ever more refined haplogroups)

Surnames that are fairly generic may have had a number of independent beginnings. If people with the same surname belong to different highly refined sub-haplogroups, we know each person's earliest patrilineal ancestor obtained the surname independtly (ie, NOT as a result of kinship).

When it comes down to haplotypes within haplogroups, things can get a bit fuzzy. You have to rely on estimated likelihood of a common ancestor. 25/25 means a common ancestor is likely within the last 10 generations (300 years). 24/25 means a common ancestor is likely within the last 20 generations (600 years). 23/25 means a common ancestor is likely within the last 30 generations (900 years).

Surnames within some European countries remained unstable until just a few centuries ago (ie, they don't correlate well with actual patrilines). During the late middle ages, among landed Europeans, the surname sometimes went with the land. If a man's inheriting child was a daughter, the land would go to her & then her son, thus creating a surname switch in the youngster's patriline (or a patriline switch in the case of the surname). Then in some countries, Wales, the Netherlands, Scandinavia, etc., some surnames changed at every generation, since they merely indicated the father's first name.

Timothy Peterman

[EBurgess]

Excellent points Mr. Peterman

I very much look forward to the increased convergence between genetics and the study of surnames. The increased resolution of sub-clades beyond R1B will allow us to make better assumptions about our surname/genetic links.

I know there was more than one Burgess origin. I wonder how much of it is due to breaks in patrilineal descent vs. separate original surname lines (1000 year time-frame). This can be a sensitive topic for some people when their results come in. Personally I am not bothered by this because ultimately we are all related.

I hope that participation in our study will continue to grow and that eventually I can say with a reasonable amount of certainty that my Burgesses were from Cheshire and not Sussex for example. It puts a personal touch on history that I find very appealing.

Regards,

E.

[Jim Denning]

Sorenson Genomics

Sorenson Genomics is about to test tens of thousands maybe hundreds of thousands i bet they wind up as unrelaeased unknown origins but any ways they will be tested .i pray they at least put them in the database unnamed
stuff like this and the native american tribal membership and levite conection cohein that gets large numbers tested and us genealogy people.
whatever causes it people need to be tested err so i can get matches