Previously in Everybody has Two Family Trees…which in hind sight, I should have called “Revenge of the Undying Machine”. I went through some of the back story of how, while I was hunting down the Thompson Elmores, I was also trying to hunt down a Thompson…(fill in the blank) with my aunt’s own NPE. Her search was much more painful than the search for my grandfather’s paternal family because I didn’t have Y DNA to help me out, it’s so close in time that there are living people involved and my aunt has a time limit imposed by the late stages of kidney disease. Those additional stress elements made the usual suspects that get in the way of a genetic genealogy search a lot more disconcerting.

I’ll pick up here, nearer in time, with a helpful and unexpected helping hand.

Winters Gives me some Direction

First off, although my aunt’s top match at 23 and me (possible second cousin of some sort) was silent, he did leave some clues in his profile. He left a partial tree for his mother’s side of the family. I think his father’s side ended with his father or grandfather. That is okay. Not everyone is interested in every branch of their tree. He was more focused on his maternal Italian relatives. So I assumed he was trying to build out their tree and didn’t want to be bothered with his paternal side.

Although I hate using them, the ethnicity estimates actually helped me out there. Some of my aunt’s other matches had Italian heritage at the grandparent level and I was able to see what that looked like. It definitely showed up. Comparing that to my aunt, who showed no Italian ethnicity and to Winters who had close Italian relatives, led me to believe I needed to follow up on Winters’ absent paternal family tree.

So I built a family tree for him and found that it was invested in Iron Mountain, Calumet, Houghton and other areas of the U.P. It was also heavily English and Scottish with a branch from Austria. Again, my aunt’s ethnicity estimates also ran high in the Isles. Particularly with Irish populations. She was somehow more “British isles” than my dad and I were. She also showed a substantial input from continental Europe…a blob that represents France and Germany, etc.

Winters also had Hungarian ancestry, and I pursued that because my aunt showed several matches who were also recently Hungarian or Austrian. So I thought maybe Austrian mixed with some French Canadian might explain her continental ethnicity blob.

I built a huge “Winters” related tree to explore all the relationships. What I struggled with was connecting Winters to other matches at 23 and me, FTDNA and AncestryDNA (through Gedmatch.com). Without sharing genomes, your 23 and me research is crippled because you cannot see exactly where someone matches anyone else and so you cannot triangulate segments to find common ancestors.

Michigan’s upper peninsula saw a lot of recent immigration from all over the world. So there is a very recent mix of Italians, French, Irish, Germans etc. Ironically, the variety wasn’t as helpful as it might be because it seemed everyone was the same mix of Italians, French, Irish, Germans etc. So that information did not help me sort matches and nothing was as obvious as the Quebec matches (which are the bulk of my aunt’s matches).

I really needed to know how Winters matched my aunt. So I periodically begged for access using examples from his family tree as possible directions for research.

After a year or more of silence, Winters quietly shared his DNA. The family tree I have for him, again, contains no real genetic input for him from Quebec but large amounts of input from the U.P. where my aunt has many matches. I expected his DNA segments to show the same and that he would fill in different parts of my aunt’s genome than those dominated by the French Canadians. Possibly matching other British or Hungarian families. I actually expected the Hungarians to shine through because I had some good Hungarian/French Canadian cross overs.

I was wrong.

Not only did he match the French Canadian matches, he matched the biggest of them in multiple places. In fact, he ONLY matched the French Canadians. His genetic family tree did not match the paper trail family tree I had for him. Winters (apparently judging from the triangulated segments I did have) was also a Robert/Robar descendant. All the research I had put into his extended paternal family looking for French Canadians who may have married in, was useful to me, but more as a process of elimination than of discovery.

Without any real conversation to give further direction, I had to think back to the his profile. It did not contain his paternal family beyond a certain generation. Perhaps it was not a lack of motivation in his family research, but a genetic brick wall like I had in my own family with my grandfather that had kept him from filling in his tree.

That sounds bitter, but really, it’s helpful to know when you’re barking up the wrong tree and I like to think that maybe he silently shared his DNA so I could figure that out. It’s nice to know that you’ve been trying to put the square peg in the round hole because you can quit trying it.

I went back to mapping the Robert family, but paid special attention to the branches that moved from Quebec to the U.P. assuming the unknown connection for my aunt and Winters (and other matches who also had ties to the Houghton area) most likely happened there.

I informed him of my findings and thanked him for sharing. No response was returned. That could be because 23 and me was in transition and my message didn’t go through, or it could be because he was just giving me a boost and was not really interested. Either way, I will not know what he knows about their shared family.

Dumb Luck Strikes Again

I should probably make “Dumb Luck” a category of this site. Hopefully, if you’ve read any of my other posts, you’ll know that I spend a lot of time doing my (and a lot of other people’s) homework. I really do put hours and hours into this, but often the real breakthroughs are driven by the unseen hand of fate. All my hard work, ends up giving me insight into what chance provided.

Here again, evidence falls into my lap.

Although I have struggled to get basic information from many of the top matches for my aunt at Ancestry.com, and I routinely throw my hands up in disgust at further road blocks and dead end leads, this past week, I got a solid.

A brand new 2nd cousin level match. I will call him “PureLuck” for having fallen into my lap.  In his family tree, I see a recent “Roberts”. Thanks to Ancestry’s addition of Michigan death records, I find out that John Roberts from Ontonagon Michigan is really Jean Robert from Quebec. He and his parents already exist in my research tree for the Robert/Robar family although I had no idea he changed his name and moved to Michigan or that he was married and had children.

John Roberts fits perfectly into the genetic picture with descendants in the U.P. that could coincide with Winters family (possibly adding the Robert DNA to them) and sons that move to Flint Michigan, which is where my grandmother worked and my aunt was born.

John’s wife Mary Bayhan is the daughter of two Irish immigrants, which brings me back to my aunt’s penchant for Isles ancestry. John’s father Amable Robert is married to Melanie Ledoux which coincides with Ledoux related matches at AncestryDNA and 23 and me (finally making sense of some of the earliest matches my aunt had).

Overall, John Roberts fits very well and layers into the existing evidence in the right spot.

He has two sons that move to Flint. One son appears to have left before 1940, but the other (Irving Roberts) is there in Flint and I can find his place of employment in 1946, about a mile from Hurley Medical Center where my grandmother is in school.

Irving also has a son, who would be a good candidate and is 10 years older than my grandmother. I still suspect Irving though because being born around 1890, as our unhelpful relative pointed out, he would be “some old man” in 1946. He’s in a position to spend time within proximity of my grandmother.   My aunt’s match to her brand new second cousin level person, PureLuck, is bigger than you’d expect for a second cousin at 344cM. If Irving is her biological father, then that would make my aunt a 1st cousin 1 time removed for PureLuck (one generation closer to their shared common ancestors). I also cannot find a lot of compelling evidence for a connection to Irving’s wife who is the daughter of German immigrants who live in Idaho.

After a process of elimination for other Roberts descendants related to John, I think Irving is the best possible person given the evidence I have. I can’t write off his son, or his brother for that matter, but he seems like the most reasonable choice.

What could go wrong?

Well, for one thing, this is about confidence levels not absolute proof. I’m not getting a lot of good information on these matches that are close to home. Simple conversation might turn up some interesting facts and possibilities I’m not considering. I don’t know what information I’m missing because all of the closest matches (so far) are completely mute.

I feel really confident about Irving because a lot of pieces fit, but there are some caveats. That new PureLuck match is about 100cM bigger than ISOGG lists as a rough guide to second cousins, but he’s also about 80cM smaller than what ISOGG lists as a rough guide to a first cousin 1x removed (which is where I place him in the family tree).

Irving ties up the loose ends on many Robert/Robar matches and Ledoux matches, but he doesn’t (so far) help me tie up all of the matches my aunt has from the U.P. Including some pretty big ones.

Probably the most troubling of those are large Winters related matches. As of this writing (part way through actually), she now has two Winters family matches. They are first cousins of each other and given the triangulation I’ve done on the one at 23 and me (as I wrote earlier), most likely very recent Robert family members. I just have no clear idea how. These two cousins share a male ancestor born around 1900 in Calumet Michigan and that man’s wife who is the daughter of Austrian immigrants, also born around 1900 in Iron Mountain.

Because of the strong Robert family ties in their DNA, neither of those ancestor’s family trees fits. I suspect that one of those ancestors is really a Robert, but I couldn’t say which one. I have read a lot into the original Winters match’s missing paternal tree and, for the record, the new Winters related match chose to leave his Winters family off his tree as well.  On a gut check I think it’s the Winters patriarch whose story is more complicated, but again…what do they know that I don’t I know?

These Winters matches bookend my aunt’s brand new match PureLuck with the documented Robert family. One Winters sits at 280cM and the other at 366cM, you could expect them to have nearly the same relationship level as PureLuck. which would mean they most likely tie up to the Robert family right there around 1900.

Comparing the tests at AncestryDNA, I know the Winters there shares PureLuck as a match.  Since I’m dealing with AncestryDNA, I don’t really know how big of a match they are to each other. I only have the perspective of my aunt, which makes these matches practically equidistant. I would have to suspect that everyone in the scenario is descended from John Roberts somehow.

Also eerily, among these three matches, each man is larger than you’d expect for a second cousin and smaller than you’d expect for a first cousin one time removed. Making me question whether I’ve got my aunt situated in the family tree correctly. Being off a generation here or there can be a big deal.

If I was dealing with PureLuck alone, my confidence level would be high given our dumb luck and triangulating homework done specifically with the Robert family. I just can’t seem to escape the gravity of the Winters family. I am certain this uncertainty around them will continue to haunt me and leave doubts to gnaw at my confidence until more second cousin level relatives outside the Winters family begin showing up in my aunt’s matches.

What has gone right?

Even with those nagging doubts, I think I’ve got the right family and I’m most confident that I have the right man in that family.

Come Hell or high water, at the cost of not being invited to family reunions anymore and alienating my own relatives as well as total strangers, I feel I have answers for my aunt’s basic questions.

I used four different databases of autosomal matches and spent three…nearly four years putting together clues and triangulating DNA segments with matches and their family trees. I had luck on my side, but luck favors the prepared. I feel like I beat the clock of kidney disease to provide the best answers I can, which is much better than the answers she had before.

I can’t say that I never gave up, because I gave up all the time, but I do feel like I kept giving it chances to work (once enough time…and bitterness had passed). Now, I think I can take a step back and wait for yet more confirming matches to roll in (the way they did with my Elmores).

Maybe, I can treat this more as an enjoyable hobby now rather than the painful dismantling of my recent ancestor’s great and terrible machines.