In November 2017 a new name appeared in my AncestryDNA matches. I’ll call him L. Ancestry rated him a first or second cousin with “extremely high” confidence, and my mother, a generation closer to him, shared even more DNA with him than I did. His shared matches were people on my mother’s Esterson side. That told me where to look. It didn’t tell me where he belonged. Nobody in the family had heard of him, and it took eight years, a lot of arithmetic and some patient new relatives to find his place in the tree.
This post is about how that worked. I’ve left out the names of the living people involved, and I’ve kept to what the DNA and the records show.
A match with no obvious place
The first question with any DNA match is which side of the family it comes from. Ancestry’s shared matches answer that: they list the people who match both you and the stranger. L. shared matches with my mother’s Esterson cousins and with a known cousin from her grandmother’s Weintraub family. That put him among the descendants of my great-grandparents, Sol Esterson and Anna Rae Weintraub, who raised six children in Chicago. Matching the Weintraub side as well as the Esterson side meant he probably descended from both of them, not just from one.
I knew of all six children and their families, or thought I did. L. didn’t fit anywhere. His wife managed his DNA test, and I wrote to her through Ancestry’s messages, on and off from 2017 to 2021. In my first message I said that these matches sometimes turn up things that need to be handled with sensitivity. Her reply held the most useful clue I would get for years: he knew his mother’s family in Chicago, but not much about his father’s side. His tree named both his parents and their parents, and none of those names were in mine. In 2021 I asked Howard Esterson’s daughter, my mother’s first cousin, whether the name meant anything to her. It didn’t.
Better tools, a clearer picture
By 2024 Ancestry’s tools had improved, and the picture sharpened. Among L.’s matches, Howard’s daughter now appeared as his likely half-sister or niece. Those are very different relationships, but they share about the same amount of DNA, so the amount alone can’t tell them apart. Dates can. For her to be his niece, L. would have to be Howard’s full brother, a son of Sol and Anna born when Anna was nearly sixty. Half-sister was the only reading that made sense.
In January 2025 another new match appeared, a young woman who matched me as a second cousin once removed. She turned out to be L.’s granddaughter, and she was happy to help. My first theory was about the surname. Her family’s patriarch was buried in the same Chicago cemetery as several of my mother’s Rietman relatives, and I wondered about a connection somewhere around 1900. It was the wrong tree. A week later she wrote back with the piece I was missing: L. had been raised in that family and taken its name, but he wasn’t related to them by blood. That explained why his surname meant nothing to anyone, and it fit what his wife had told me in 2017. The connection had to come through his father. My best guess that day was Howard, which would make L. Howard’s son, born more than five years before Howard married.
Writing to him
In January of this year I found an email address for L. and wrote to him. It is a hard letter to write. You are telling someone that the story they grew up with may not be the whole story. I tried to make it easy to say no. He wrote back the next day. His birth certificate named a different man as his father, one he had never met, and he had no reason to doubt it. He copied his half-brother on his mother’s side, who has been generous and careful in everything since.
I made a mistake at this point that is worth confessing. In my first replies I suggested Sol, my great-grandfather, as L.’s father. L.’s half-brother pointed out that Sol was nearly sixty in 1942, and that didn’t fit anything the family knew. He was right, and the DNA had said so all along. If Sol were L.’s father, L. would be a half-uncle to Howard’s daughter, and they would share about half as much DNA as they do. I had got ahead of the numbers.
What the numbers say
DNA matches are measured in centimorgans (cM). The more you share with someone, the closer the relationship. Each relationship has a typical range, and the ranges overlap, so one match is a clue, not an answer. What settles a question is several matches from different branches that all point the same way. Here are L.’s closest Esterson matches, described by their relationship to Howard, with the ranges Ancestry publishes for each relationship:
| Match | Relationship to L. if Howard is his father | Shared DNA | Typical range |
|---|---|---|---|
| Howard’s daughter | Half-sister | 1,754 cM | 1,160–2,436 cM |
| A son of Howard’s sister Jean | First cousin | 1,066 cM | 396–1,397 cM |
| My mother, daughter of Howard’s sister Faye | First cousin | 953 cM | 396–1,397 cM |
| A daughter of Howard’s sister Jean | First cousin | 854 cM | 396–1,397 cM |
| Two of Jean’s grandchildren | First cousins once removed | 556 and 520 cM | 102–980 cM |
| Me, grandson of Howard’s sister Faye | First cousin once removed | 445 cM | 102–980 cM |
Every one of these lands where it should. The half-sister match is close to the middle of its range. The three first cousins come from two different sisters of Howard, and the next generation down steps down the way it should. His granddaughter’s match to me, a second cousin once removed, is exactly what it should be. L.’s mother was not an Esterson, so the link had to come through his father, and his father had to be one of Sol and Anna’s children. Their other five children were daughters. Howard was their only son.
The part DNA can’t tell you
DNA can tell you who is related. It can’t tell you how two people met. I wanted a second line of evidence that put Howard and L.’s mother in the same place at the same time, in the spring of 1942.
Some of it is there. The 1930 census shows the two families living a short walk apart on Chicago’s West Side. His old high school’s 1944 yearbook, listing former students in the service, has Lt. Howard Esterson in Texas, and his daughter remembers that he trained bombardiers. L.’s family knew that his mother had spent part of early 1942 in Texas. That is suggestive, but it isn’t proof. Howard’s service records would help, but many Army records from that era were lost in the 1973 fire at the National Personnel Records Center in St. Louis, and I haven’t been able to get his yet.
Photographs were the other thing everyone wanted to see. Howard’s daughter sent pictures of her father in uniform. L.’s half-brother looked at one and said it was the image of L. as a young man. I also tried an AI tool to sharpen some faded photos for comparison. The results were interesting, but an enhanced photo is not a source, and I don’t treat it as one.
Where things stand
By March, L. and his half-brother had both accepted the DNA’s answer. L.’s family has updated their tree on Ancestry, and he and Howard’s daughter, half-siblings who have never met, have both said they’d like to talk. We’re still working on that. It takes time, and it should. In my files and in the Reunion family tree, L. is now Howard’s son.
What I’d pass on
- Test more than one relative, and the oldest generation first. My mother shared twice as much DNA with L. as I did, and it took matches from several branches to place him.
- Use shared matches to pick the side of the family, and cM ranges to narrow the relationship. Then use ages and dates to rule things out.
- Trust the numbers over your own old notes, and say so when you get it wrong.
- DNA answers “who” much better than “how.” Keep looking for records, but don’t expect them to be there.
- Behind every surprise match is a person with a family story of their own. Go slowly, make it easy to say no, and let them decide what they want to know.
Leave a Reply