I've been writing about Ancestry DNA's new wrinkle on autosomal DNA matching--DNA Circles. To be considered a member of a DNA Circle based on a particular ancestor, you must have that ancestor in your tree on Ancestry and you must match DNA with at least one other person in the circle. For example, I belong to the Daniel Reed DNA Circle. Daniel Reed is my 4th great-grandfather. Out of 23 members of the circle, I share DNA with 5 members. The other 18 members share DNA with at least one other person in the circle. We all have Daniel Reed in our public trees on Ancestry.
Recently, Ancestry DNA has added another feature that they call "New Ancestor Discoveries." These are ancestors or relatives suggested by DNA evidence alone. In this post I want to look at the 8 "new ancestors" that Ancestry DNA suggests for me, based on my DNA test results and the test results of members who already have these ancestors in their family trees. (See Roberta Estes's blog, dna-explained.com, for her take on the ancestors/relatives Ancestry DNA suggested for her. She calls them Bad NAD's.)
NAD #1 -- Hannah Elizabeth Embry
The following facts were compiled from 595 trees on Ancestry, according to her NAD bio. Hannah was born 10 May 1815 in Garrard County KY. She married on 22 October 1833 in Butler Co. KY. She died 1 September 1877. Her parents were Isaac Embry (1786-1850) and Martha "Patsy" Jameson (1780-1882.) Ancestry says, "DNA evidence suggests that you're related to a group of Hannah Elizabeth Embry's descendants." I match 3 of the 11 members of the Hannah Elizabeth Embry DNA Circle.
Oh, how I wish I could figure out how I am related to Hannah Elizabeth Embry! I feel certain that I have a connection to the Embry and Pharis families of Butler Co. KY, but I have been unable to identify the exact ancestor that connects me to these two families. I suspect he is James Pharis of Jackson Co. TN, whose mother was Delilah Embry. (Many of the families of Jackson Co. TN had ties to Butler Co. KY. Delilah's father John is supposed to have died in Butler Co.) If James is the unidentified father of my 2nd great-grandmother, Elzina Huff, I don't know how I will ever prove it. Hopefully, future DNA test results will point definitively to James or some other Pharis/Embry ancestor. I will definitely keep an eye on future DNA matches to this NAD.
NAD #2 -- Christopher Columbus Snodgrass
Christopher Columbus Snodgrass (1801-1877) was the spouse of Hannah Elizabeth Embry. If, in fact, I am related to this couple, it's to Hannah and not to Christopher; however, since Ancestry relies on family trees to create DNA Circles, they have no way of showing from which spouse the DNA comes. Since my 3 DNA matches are to descendants of this couple, they both show up as my NAD's. Since Ancestry is suggesting relatives, not necessarily ancestors, I guess it doesn't really matter to them that Christopher is not my relative by DNA.
NAD #3 -- William Jackson
Ditto William Jackson. He is the spouse of my 2nd great-aunt, Sarah Elizabeth Wheat. Sarah was the sister of my great-great grandmother, Susana Wheat Ming. William was born in 1812 in White Co. TN, married Sarah in 1834 in Madison Co. AR, and died in 1879 in Pilot Grove, Grayson Co. TX. On the 1860 census of Grayson Co. TX, William and Sarah and 6 of their children are living next to William F. and Susana Ming.
According to Ancestry DNA, once I put William Jackson in my tree, he would disappear from my New Ancestor Discoveries, and that is exactly what happened. He and Sarah are now in my DNA Circles, the only circles I have that are not direct-line ancestors or their spouses. In the case of this NAD, Ancestry DNA did make a helpful suggestion.
NAD #4 -- Martha Jane Kendrick
According to her bio compiled from 104 trees on Ancestry, Martha Jane was born on 22 May 1854 in Kinderhook, Virginia. She is found on the 1860 census in Washington Co. VA. She married first in 1871 to Joel Kaylor and later to Wyndham Clark. By 1900 she was living in Madison Co. AR. She died in Pettigrew, AR in 1947. Her parents were John Kendrick (1831-1865), born in Washington Co. VA and Phoebe D. Morgan (1829-1882), born in Scott Co. VA.
Both the name Kendrick and her mother's birthplace in Scott Co. are clues as to how I could be related to Martha Jane. I do have a Kendrick in my family tree. She is Isabelle Kendrick, my 5th great-grandmother. She is the ancestor of Rachel Sargent, my 2nd great-grandmother, wife of Goldman Davidson Castle. Isabelle was born in 1754 in Stafford Co. VA and died in 1822 in Russell Co. VA. Her parents were Patrick Kendrick (1725-1803) and Jane Fox (1730-1805).
As you may have noticed, Martha Jane and Isabelle were born a century apart. Since there appears to be no direct-line connection between them, common sense would dictate that the connection between them has to be at least a generation or two before Isabelle. The way the NAD's work is that Martha Jane already has a DNA Circle. The fact that she shows up as a NAD to me is a suggestion that I would show up in her circle if I had her in my tree. Of the 4 members of Martha Jane's DNA Circle, I have DNA matches to two of them. Maybe one of their trees will give me a clue to how, or if, Martha and Isabelle are connected.
One of the two DNA matches, Azkjo55, has an extensive tree. In fact, we have
two "Shared Ancestor Hints," symbolized by Ancestry's little green leaf. One of our shared ancestors is, sure enough, a Kendrick--two generations back from Isabelle. He is Thomas Kendrick (1699-1770), my 7th great-grandfather. My tree only includes my direct-line ancestor, Patrick Kendrick, son of Thomas, and Azkjo55 only shows her direct-line ancestor, John Kendrick, son of Thomas. However, if Patrick and John were, indeed, brothers, that could definitely be my DNA connection to Azkjo55 and to Martha.
Here's where genetic genealogists (and to be honest, even serious paper genealogists) have problems with Ancestry.
1) There is little evidence on most people's trees that Thomas Kendrick had two sons named Patrick and John. That's not to say that there isn't evidence, but a lot of Ancestry users borrow names from other users without documentation, or with documentation for a completely different person with the same name. Just because 104 people say that Martha Jane Kendrick was a descendant of Thomas Kendrick through his son John doesn't mean that's necessarily true.
2) To really prove a DNA relationship, according to genetic genealogists, you must have triangulation. Three descendants must match on the same DNA segment and show paper evidence (wills, census records, etc.) for the same ancestor. First of all, Ancestry doesn't give segment information, and second, Azkjo55 and I have another possible shared ancestor couple, John Barker and Martha Snead. Without segment information, we will never know if our DNA match is through the Kendricks, the Barkers, or both.
I understand why people might prefer Ancestry DNA to one of the testing companies that show matches along with segment information. It's a lot of work to match up segments with an individual and then find your common ancestor, especially when many of the people who test on Family Tree DNA, for example, don't even upload their family trees. Ancestry does all of that for you. In fact, their viewpoint is that "it's almost impossible for you to find enough matching segments with other users to have confidence that you have a common ancestor. To solve this problem, we created DNA Circles, where we collect evidence across millions of trees and DNA from all Ancestry DNA members. Because of this power of numbers, the evidence that you really did inherit DNA from the same ancestor as everyone else in the circle can increase."
I'm not sure I completely agree with them, but I've decided I'm going to try to use the information they provide, with reservations, and hope that it can help me. I think they can make a unique contribution because they do have a great number of users who have contributed trees and DNA.
I could put Martha Kendrick in my family tree, as I did with William Jackson, and see her NAD disappear to be replaced by a Martha Kendrick DNA Circle; however, since I can't prove to my satisfaction that John and Patrick Kendrick were brothers, I'm going to hold off for now.
NAD #5 and #6
Based on almost 1200 trees, Benjamin F. Burden was born in 1783 in Baltimore, MD. He married Elizabeth Tully on 6 November, 1801, in Nicholas Co. KY. He appears in Butler Co. KY by 1810 and died there on 12 May 1862. Out of 23 members of the Benjamin F. Burden DNA Circle, Ancestry says I match 10 of them by DNA. I match 9 of 21 members of the Elizabeth Tully circle.
I have absolutely no idea how I am connected to Benjamin F. Burden and Elizabeth Tully. While many of the DNA matches to the Burdens include an Embry in their trees, not all of them do. I guess it's possible that I am picking up Butler Co. KY in each of the matches but don't have a true connection with the Burden/Tully descendants. With that many matches, it's worth trying to figure out how we are related.
NAD #7 and #8
My last two New Ancestor Discoveries, at least for now, are Sabrina Cutbirth and David Wilcoxson. Sabrina was born in 1800 in Maury Co. TN and died in 1877 in Farmersville, Collin Co. TX. David was born in 1796 in Rowan Co. NC and died in 1883 in Farmersville. They married in 1825 in Tennessee.
I match 4 of 23 descendants of Sabrina and the same 4 of 20 descendants of David. Two of the 4 matches are in the same family group and administered by the same person. They are both direct-line descendants of Sarah E. Wheat and William Jackson. K.D., the administrator, is a member of four of the same DNA Circles that I am a member of: Zachariah Wheat, Elizabeth Whitley, Robert Stephenson, and Samuel Wheat. Easy, huh? Our connection is with the Wheats or their ancestors--except that the other two matches don't have Zachariah or Samuel Wheat, Elizabeth Whitley, or Robert Stephenson in their trees.
Even Ancestry's DNA Circles and New Ancestor Discoveries don't always make things easy.