Over the past few months, I’ve been excited as I thought I was pulling together a strong case for identifying Laban Morris’s mother once and for all. All the facts seemed to make sense and line up just perfectly — to me, at least.
I thought the DNA was supporting my hypothesis, as well. Well, a couple of hypotheses, which were:
❌ Nancy Morris is really Serenia Averitt, the illegitimate daughter of Selah Averitt and possibly Zachariah Williams.
❌ Selah Averitt and Laban Morris are half-siblings and share a mother.
Turns out there is now NO DNA to support either and the DNA I thought I had turns out to have been sending different signals.
Chalk this up to my misunderstanding of stacked X segments — and to two other AI models happily walking into the confirmation bias with me. This is just so embarrassing. 🤦🏻♀️
It’s not the first time I’ve wandered down a rabbit trail while doing genealogy research and gotten far off track, but never have I been so sure of a house of cards as I was with this X-DNA business.
What were the errors?
What I wrote about how X is inherited is still correct. Scientifically, every single thing I mentioned about it in this article is true.
The problem was with the embedded video and certain conclusions that I was making based on overlapping segments of X-DNA.
I had included this image:
I also included fan charts for each of these people, highlighting the only X-DNA-qualifying female ancestors who are shared between them.
BUT, I missed one very important thing and it is the thing upon which my hypothesis rises or falls.
I mistakenly thought because these women are all maternal matches to my grandmother and that their segments all line up, it must be that at least where they all stack on top of each other, it must be the same shared ancestor.
You would think that, right? I mean, this is all on the maternal line.
Here is what I was thinking:
- My grandmother Beatrice Morris and Janette and Jean and A. all match on my grandmother’s maternal line
- Beatrice, Janette, and A. all descend from Jane Holmes and Harvey Edwards, so all three can have X from the same three women: Selah Averitt, Esther (John Holmes Jr.’s mother), and Lovi Ann, wife of Elial Edwards and mother of Harvey.
- Jean does not come from Jane and Harvey, so she should not have those three women on her X. She is the Alpha Morris control — the path that can still carry pieces of the X Alpha got from Laban, which is Laban’s mother’s X. That is not Kitty’s mtDNA. Alpha’s mitochondria came from Kitty Williams.
- As far as I could tell, Jean didn’t have Lovey or Esther in her X-DNA line, nor is she a Selah Averitt X-DNA descendant, and since Jean is the closest tester I have on Alpha’s line — the path that can still carry Laban’s mother’s X — her not matching Ruth on X was one more place the hypotheses failed to show up.
I never said this was definitely the case, but I said I had a strong feeling this was the case because of that DNA up there and all the other circumstantial evidence in the article. It just made sense to me.
Unfortunately, neither Gemini Pro nor Claude stepped in to correct my understanding of X-DNA. Both of them cheered me on and told me there was no other explanation, that obviously that must be what was happening with that X-DNA lining up like it was.
I just felt like I needed one more piece to lock things in and be sure. I decided to start looking for any woman who might have inherited X-DNA from Nancy Morris. Since I thought Nancy Morris was really Serenia Averitt, I figured if a woman with X-DNA from Nancy Morris also matches the women in the graphic above on a shared X-DNA segment, that would prove it.
I wound up finding two sisters on Ancestry DNA who both descend from a woman named Ruth Dennis. Ruth is 96-years-old and she was willing to do a DNA test, herself, to provide DNA a full generation further back than her daughters. Ruth’s daughter Tina facilitated the test and then, when the results came back, she uploaded them to GEDMATCH and sent me the kit number so I could run some X-DNA checks on her.
Imagine my dismay when my grandmother didn’t match Ruth’s X-DNA at all. Nor did A. Nor did Jean. Janette did, but on a segment that didn’t overlap with A. or Jean or my grandmother. I was bewildered. Surely, she should’ve matched X-DNA to at least one of the women if there was any connection between Selah Averitt and Nancy Morris, but the fact that there was no match, while not necessarily completely disproving the theory that Nancy Morris is really Serenia Averitt, it certainly provides zero evidence for the identification. In fact, the combined evidence seemed to be stacking up against the likelihood.
The only AI who was dealing with me as an honest broker after I found myself confounded by the lack of matches between Ruth and those other ladies, and in fact, a lack of any X-DNA matches who come from any known X-DNA descendants of Selah Averitt, was Grok.
Below I’m going to share Grok’s detailed and scientific summary of why that image with the stacked DNA above is so misleading — and why.
What the stacked X chart is not telling you
Grok (xAI), for East Carolina Roots
The chart compares three women to Sara’s grandmother, Beatrice Morris, on the X chromosome:
- Janette — 117.68 cM, about 23.5–141.2 Mb
- A. — 64.98 cM, about 6.0–45.7 Mb
- Jean — 21.17 cM, about 34.0–50.1 Mb
All three are maternal-side matches to Beatrice. All three sit in the same eastern North Carolina cluster. On the browser they pile up around 34–46 Mb. That pile is what made the two hypotheses look physical: that Laban Morris and Selah Averitt were maternal half-siblings, and that Nancy Morris was Selah’s 1823 daughter Serenia.
A browser stacked against one woman does not say that.
Beatrice has two X chromosomes. The vendor draws one bar and drops every match onto it. Segments that occupy the same megabases can sit on opposite copies of her X — father’s X on one match, mother’s X on another — and still look stacked. Even when two matches are on the same copy, that copy is already a patchwork of more than one grandmother. Adjacent or nested bars are the normal product of recombination. They are not a caption that reads “one shared woman.”
The only thing that converts a stack into a shared female ancestor is triangulation. The matches have to share that same interval with each other, not merely with the person in the middle.
They do not.
Janette’s block is long and is the Jane Holmes path. A.’s block is a different, more proximal piece that only overlaps Janette from about 23.5–45.7 Mb. Jean’s sliver sits under the right end of A. and the left end of Janette. Jean and Beatrice do share a short piece at about 34–47 Mb. Jean and Janette do not share that piece. Their only X at the settings used was a different, distal bit at about 144–150 Mb. Jean under Janette’s bar is a different match parked on the same ruler.
That fails the half-sibling test on this data. If Laban and Selah had the same mother, Laban’s X — the line Jean represents through Alpha Morris — and Selah’s X — the Jane Holmes line Janette and Beatrice represent — would have to retain the same grandmother segment, and that segment would have to survive in both tester lines. Two separate pairwise crumbs are not that segment.
It also fails as proof that Nancy was Serenia. A separate X between Ruth and Janette sat on Janette’s paternal Stocks/McLawhorn X, not on the Jane Holmes / Selah stretch, and Ruth did not hold X with Beatrice or Jean at those settings.
Maternal-side and “from the same neighborhood” is the expected pattern on this tree, not the surprising one. Selah’s Holmes daughters married Laban’s sons. Alpha Morris married a Holmes. Ethel Morris’s parents were first cousins, which braids those Holmes sisters back together again. Holmes testers and Morris testers are in-laws to the household where Nancy grew up. Endogamy plus those marriages will throw overlapping-looking X into the same descendant pool whether or not Laban and Selah shared a mother, and whether or not Nancy was the 1823 Averitt child.
The chart shows a cluster doing what this cluster does. It does not show either hypothesis.
While the DNA analysis has changed, the historical record stands:
Zachariah Williams is still Kitty’s brother.
Nancy is still in that household and still not ordered as one of Laban and Kitty’s children.
Serenia Averitt was still born in 1823.
What I no longer have is DNA that makes those facts into one story. Laban’s mother is unknown. Serenia after 1823 is a separate question. The Holmes–Morris marriages already on the tree are enough to explain why maternal-side women from this neighborhood stack on my grandmother’s X without a shared mother for Laban and Selah.
This is why I put these neon yellow flags in my tree on Ancestry — to tell you when something is speculative and to let you know that if you put that in your tree, you may need to go back and fix it later because it is my own hypothesis. If you’re out in the wild and see any of my hypotheses being treated in facts in other people’s trees, please kindly send them a message and direct them to this article.
Have you seen any other evidence about any of these folks? I’d love to hear from you in the comments below.
