Understanding how endogamy and pedigree collapse can affect DNA matches supports finessed fine-tuning with DNA triangulation.
The paternity of my 2x great-grandfather, Cornelius White, has been a mystery ever since I began my ancestral journey in 2010. All I had was the usual information that could be gleaned from online record sources. He was born about 1829 in Virginia, either in Wythe, Smyth, or Augusta County. He married Ann St Clair. Together, they raised a small family in Wytheville, Wythe County, Virginia.
The only census return I could positively associate with him was the 1880 Census, where he, Ann, and their small family were enumerated. I had hoped to find him in the 1865 Cohabitation Records for Wythe County. Neither he nor anyone else from his immediate family was listed in this invaluable African American genealogy resource. Nor could I find them in Smyth County, another central location for my extensive extended family. Frustratingly, similar records for Pulaski and Augusta, additional counties that feature largely in my southwest Virginia family’s history, have either been lost, destroyed, or undiscovered. So I put Cornelius on the back burner. I’d return to him from time to time – only to put him back on the back burner. I just couldn’t make any headway with him.
I continued my overall genealogy research on a county-wide level by adding more extended families to my tree. At this point, I have most of the late 18th Century to late 19th Century Wythe, Smyth, Pulaski, and Augusta County family groups in my tree.
Thanks to endogamy (where groups of people marry amongst themselves, creating one large extended family group over time), I’m related to most of the people in these counties – black, white, and Native American – with pre-1900 roots in these counties through a succession of cousin marriages from the early 1700s onwards.
This beautiful region of Virginia is nestled within the Blue Ridge Mountains. It’s sparsely populated even to this day. Before the automobile, it would take a day or more to walk from town to town in this region. So you tended to marry who you knew, which was going to be someone in the same community. Which meant you either married a cousin of some description. Or you didn’t marry at all. I’d imagine that newcomers, who mixed the gene pool up a bit, were feted. I went through something very similar when I moved to a fairly isolated part of Cornwall in southwest England. I was single at the time and invited to every manner of dinner party, church gathering, local dances, parties, and saint festival days you could imagine…with single daughters, grand-daughters and nieces being introduced to me left, right, and center for the first two years I lived there.
Around 18 months ago, an interesting picture was beginning to emerge where Cornelius was concerned.
Both Cornelius and his wife Ann had something to do with Colonel James Lowry White (1770 – 1838) of Staunton, Virginia. Ann, I believe, was enslaved by James White. James was the Rockefeller or Vanderbilt of his day. He was one of the richest men in America with vast business enterprises, land holdings, and slaves in Tennessee (Knox County, Ann’s place of birth), Alabama (Huntsville, Madison County), West Virginia, and Virginia. For now, Ann’s trail has gone cold.
Cornelius was a different prospect. I just kept returning to the notion that Cornelius and James were blood relations. James White fathered one known child by my enslaved 3x grandmother, Elsey George (wife of Jacob Sheffey). Could he also be the father of Cornelius? I wouldn’t have been surprised. I kept looking at the year Cornelius was born (1829) and the year James was born (1770)…and a father-son relationship just didn’t seem likely. I shouldn’t assume that, I know. I have distant relations who were still fathering children in their 60s, 70s, and 80s. And looking at his family tree below, he was clearly still having children by his wife at the time Cornelius was born.
Could these two men be a grandfather and a grandson? That seemed the most likely prospect. I can’t explain it. It felt right.
It was time to delve into the DNA matches I had on Ancestry, FamilyTree DNA, and Gedmatch.
Endogamy, endogamy, you will be the end of me!
The first hurdle I faced was this: a descendant of the old Quaker White family who had originally settled in Cumberland, Pennsylvania, James Lowry White was already my blood relation in 3 different ways:
My mother was a descendant of the same family via her Quaker Harlan lineage;
- My father’s maternal Roane ancestors shared common Parke, Dandridge, Henry,
and Carter ancestors with James’s maternal Lowry ancestors; and
- A marriage between James’s half-sister Margaret and my 5x great uncle, Major Henry Lawrence Sheffey, meant an entire Sheffey line also shared the same blood relation.
So, in his own right, James was already a cousin twice over – as well as my great-uncle. He was also a relation through marriage. Let that one sink in for a minute. That is the joy of endogamy. So, no matter how I looked at it, all of his descendants were going to be my cousins. So how was I going to crack finding Cornelius’s father if James and all of his sons were already my cousins?
All of their lines were going to be genetic matches to me.
DNA triangulation was going to be the key
DNA triangulation. So what’s that? In autosomal DNA testing, triangulation is the term used to describe the process of reviewing the pedigree charts of people who match the same autosomal DNA segment(s) to see if a common ancestor can be found. The technique is best used in conjunction with chromosome mapping. It is a long, long process requiring meticulous attention to detail, care, and copious notes.
Triangulation has helped me identify a number of white men who had children – and indeed whole second families- with enslaved as well as free women of color in my family.
This time around, I knew I couldn’t look at any of the men in James’s tree because they were all already related to me. I had to look at the women who married them and research their families.
Looking at the abridged family tree above…there were quite a few sons with wives who required researching. Triangulation was going to take some time. In this instance…18 months!
The reason why it has taken so long is I had to go back anywhere from 5 to 8 generations for each woman who married into the family in order to be certain that I wasn’t genetically connected to any of them. If I was related to any of these women, triangulation wouldn’t produce the result I needed. In other words, I’d get a false positive as a result.
So let’s start with James Lowry White II’s mother, Ann Marie Lowry.
I wanted to start with Ann Lowry to see if I had any matches on her maternal line. I couldn’t look at her paternal Lowrys. I already knew I shared their DNA. I had to look at her maternal Boggs line. As far as I am aware, I only have 1 line of Bloggs. Sure enough, there they were in my DNA matches: Boggs from her mother’s side of the family. This put all of Ann Lowry’s sons, including James Lowry White, in the frame. The only way I could have a combination of White, Lowry, and Boggs matches would be via a son, who would have passed DNA from both parents down to Cornelius, who passed enough of this DNA down to me for me to have strong autosomal DNA matches.
However, just to be certain that I should only be looking at the sons of James, I researched the families of Colonel James White’s sisters-in-law (James II’s aunts) and came up empty-handed. I didn’t share any matches with the names in their trees. Now, that could be because none of their descendants have taken DNA tests – or at least not with AncestryDNA. That’s always an option. Or they haven’t uploaded their results to Gedmatch or FamilyTree DNA. Or not enough of this DNA has been inherited for a positive result.
However, thanks to being active on numerous Virginia genealogy-based Facebook groups, I know of descendants from these allied families who have taken DNA tests. Armed with Gedmatch kit numbers to compare, we quickly confirmed that we didn’t share any DNA. I feel safe to say that while I would be a distant relation to these people via marriage, we are not blood relations. Not through their maternal lines, at any rate.
At this stage, I was confident that I had eliminated Colonel James White’s nephews from the list of paternal candidates for Cornelius.
Next, I began looking at Colonel James White’s sons. One of them would be the strongest candidate to be the father of Cornelius.
I eliminated half of them almost immediately. William Young Conn White I died in infancy, so it wasn’t going to be him.
James Lowry White II was a strong candidate, as were his brothers William Young Conn White II, and Francis Smith White. All of the remaining brothers would have been too young to father a child in 1828/29. Out of 9 brothers, I had whittled the list of candidates down to 3.
As soon as I began researching James Lowry White II, my heart sank. It was my worst nightmare. His wife, Margaret Rhea Preston, wasn’t just a cousin to me…she was a double cousin. I’m related to her on both her Rhea and her Preston lines.
Undaunted, I continued.
I began working on William Young Conn White II’s wife’s family. It wasn’t long before I hit shared families with her paternal and maternal lines in Pennsylvania, Ireland, and Scotland. She was another double cousin. I remember looking out my window and muttering “Are you kidding me?” I was seriously ready to walk away from the whole thing at this point.
I turned to Francis Smith White. He presented another kind of difficulty. I found very little information about him in the official records or the Virginia genealogy books that form the core of my trusted genealogy research resources. I wasn’t overly dismayed about a lack of results for Francis. Born in 1814, I felt that he too would have been quite young to have fathered a child in 1829. Not unheard of, but quite young nonetheless.
With two White family wives turning out to be my double cousins, I was going to have to tackle this from a different direction. I was going to have to compare degrees of genetic separation between me and the descendants of James White II and his brothers.
I began comparing degrees of estimated relatedness and the amounts and lengths of DNA segments that I shared between the descendants of James II and the descendants of his brothers. My matches are between 1 to 2 generations closer when it comes to James II’s descendants when compared to my matches with his brothers’ descendants. I share more, and longer, DNA segments with James II’s descendants.
The long and short of it is that James Lowry White II is my prime candidate. However, I have to acknowledge that his brothers William and Francis could also be Cornelius’s father.
I know, it seems an awful amount of work to do to not arrive at a definitive answer. Sometimes in genealogy – and especially genetic genealogy – there isn’t a clear-cut answer. Not when you have endogamy in just about every corner of your family tree. All you can do is eliminate the impossible and/or improbable and keep chipping away at the probable until you arrive at what will be the most likely result.
That’s all I can do until a death certificate surfaces for Cornelius. That is, if one exists. If he died before the turn of the 20th Century, there most likely won’t be one. The other possibility is that if a death certificate does exist for him, it won’t necessarily follow that the names of his parents were provided. I could be facing my even older nemesis: ‘parents’ name unknown’. It’s always worth remembering that such records are only as insightful as the information an informant provided at the time.
At least AncestryDNA offered a kind of consolation prize: 2 shaky leaf hints related to Cornelius. These appeared 48 hours after I placed James White II as his father. One hint shows that James II is a common ancestor between me and another of his descendants. The second showed James II’s father, Colonel James Lowry White, was the shared ancestor between me and one of his daughter’s descendants.
That’s about as good as it’s going to get for now!
This exercise is adding more information about the names freed slaves took after Emancipation. So far, the majority of my formerly enslaved ancestors took the name of their blood relations. They didn’t just adopt a name they liked. Or pull one from the galactic ether. This, of course, makes me wonder about the handed-down notion that freed slaves chose family names of owners they liked or felt had been kind to them. Or merely because they liked a name. If only a handful of my ancestors had randomly chosen names like that, I wouldn’t give it a second thought. My DNA results are suggesting something fundamentally different.
Interestingly too are the minority of my ancestors who could have taken a surname based on a blood connection to a family who owned them – and didn’t. A small percentage of those we’re aware of didn’t simply choose because they either didn’t like or didn’t want to be associated with the paternal European-descended side of their family. Instead, they opted for another kinship-based surname.
It’s an interesting area of research.