Find potential siblings, check for DNA cousin matches

Historically, people rarely emigrated alone. Life was too precarious and difficult to handle without the support of a family network and this is especially true for populations undergoing group displacement, like the Loyalists who resettled in Canada after the US War of Independence. If you have a brick-wall ancestor, read up a little on the community they lived in and ask yourself, how likely is it that none of their relatives lived nearby?

For most North American cases, it is not likely your ancestor was a true "stray," unconnected to everyone else. In the past, many genealogists were not particularly focused on identifying collateral relatives because they were trying to get a generation further back, records tend to run out at specific points in time, and they assumed this stranded everyone in that generation without clues (not in itself a great strategy, as land records and the death registrations of the long-lived often prove). 

However, the advent of DNA testing changes the potential return-on-investment to be had from spending time identifying your brick-walled ancestor's siblings.

Firstly, if you know your ancestor's surname, it takes only a few minutes searching FamilySearch tree/records, FindAGrave, and Ancestry to locate people in the community of interest with that surname, born at the right time. If the name is rare the list will be short. Short lists and rare names are always easier to work with, so if your brick wall has that attribute, don't waste the opportunity to exploit it.

Identify the spouses of these potential siblings and check all their vital records and attached sources to see if other researchers have identified legitimate clues about the origins of these people. Then, add the potential siblings, their spouses, and their children to your DNA-test-connected family tree. Assuming your brick-wall is within about seven, maybe eight generations (4th/5th-great grandparents), if at least some of these newly-found siblings are in fact your biological relatives, some of their descendant testers (if they have descendant testers) should show up in your DNA matches. If you are using Ancestry, you would find new ThruLines suggestions in support of any correct theory after waiting a day or two for the tree additions to recalibrate with your match list (with the caveat that you need to check for additional relationships explaining the DNA shared).

Now, I'm writing this blog because last week someone submitted an incorrect FindAGrave memorial edit (thinking they had found an ancestor's burial, but cemetery documentation indicated it was an infant with the same name). Looking into the situation, I discovered a cluster of people with a moderately rare surname in the community who had been born from 1797 to 1811 (and in a consistent sequence of one born every year or two). They were all American and by checking my local genealogical history books for the area, I found three of them were closely associated (although not explicitly identified as family) in a chapter documenting one of their spouses. These three, plus a couple of others, had considerable name overlap showing in their children, as well as potential naming of children in honor of their married-in aunts and uncles. Looking at the documentation across the entire group, there were three vital registrations suggesting an origin in Dutchess County, NY. No online tree considered these people as siblings. 

In fact, a couple of them have been assigned to family origin groups that have nothing (else) to do with this particular community and there is a notable absence of name repetition across the generations in their current placements. None of these assigned family relationships have an evidentiary source to prove the connection (and none are from Dutchess NY) — in all cases researchers have simply selected a surviving American baptism record from roughly the right year of birth and decided it was for the person in question who ended up in Canada.

Having developed this tantalizing theory of relatedness, there is nothing else I can do because these are not my biological relatives. However, if descendants knew of this theory and have DNA tested, they could check its merits within 72 hours, just by following the steps above. If the theory of siblingship is supported by DNA, then they would have a specific place to go to for further records checks (Dutchess) and a list of significant personal (and potentially maiden surnames) to be on the look out for. They'd also have other unexplained DNA-group matches to potentially help find further back generations. This is a much better investment of time and resources, as compared to the ole' throw-a-dart-at-the-coincidentally-similar-baptism-record-and-pray method of the pre-DNA testing era.

Beware doppelganger places

There is a tranche of Ontario birth records on Ancestry with birth location indexed as "Prince Edward Island." PEI is a province. Ontario is also a province. They are both in Canada, a few hundred miles apart. Ontario did not register any births that happened in Prince Edward Island. It was not their business.

As it happens, Ontario has a county called Prince Edward. In the nineteenth century PEC (the county) had about a quarter the population of PEI (the province). If a Canadian moved to the US and told their children or a US registrar that they were from "Prince Edward, Canada" there was a one in five chance they meant the county, not the province.

About once a week I send in data corrections on FamilySearch or Ancestry for records indexed as "Prince Edward Island" when they should be Prince Edward County. Recently I was linking up FindAGrave memorials and hit this very problem, a birth location listed as PEI when it should be PEC. I sent a message to the memorial manager, who requested direct proof of the birth location. I explained that Ontario did not start registering births until 1869 but that (fortunately) this individual appeared on the 1861 Canada West (Ontario) census with his family group, birth location "CW" (Canada West). His US death registration gave his birth place as "West Prince Edwards, Canada," and I explained that had to be a mangled version of "Prince Edward, Canada West," which was the jurisdiction he left when he emigrated to the US in the 1860s (and which became Ontario in 1867, a name he was possibly never familiar with). 

That modest piece of evidence, coupled with the linkages to memorials for his ancestors who remain "permanent residents" of the county, convinced the memorial manager to process the update. I didn't begrudge him the delay — places are a tricky business, and not every genealogist has geographic aptitude. However, if you want to be a competent genealogist, geographic aptitude is definitely something you should develop.

The cardinal sin of genealogy is confusing two people with the same name and confusing two places with similar names is the next most serious problem on the list. 

To avoid both, you must do the same things: 

1) You must check for doppelgangers. Are there two people (or three, or four, or — if you are researching Ireland — fifty) people born at roughly the same time with the same name in the place of interest? Similarly, are there multiple places with the same or similar names in the country or region you have identified?

2) You must use contextual information and other clues (up to and including DNA matching) to pinpoint the correct person or place. Avoid making loose assumptions or guesses because if you make an incorrect choice it will probably become an apparent "fact" about the person you are investigating and this is the quickest and surest way to build yourself a brick wall.

Use atlases, google searches, regional and local archives and library content (in Ontario, this is Archives of Ontario), and published genealogy guidance for the region to identify place names and check to see if you could be confusing two similar-sounding places. You also need to tailor the search to the time period you are concerned with, as place names change over time. For researching Ontario in the 1860s, it would be important to review the census enumeration districts, any city or county directories, Wikipedia articles which include historic place names, the McGill County Atlases project and local historical society websites. Do not rely on Google Maps as it only reflects our current reality and is easy to misinterpret — it also skews to "most relevant" results, which tends to misdirect people to larger population centers.

Of course, the first step in this process is incrementally narrowing it down using documentation. There's no way someone should be indexing Ontario Birth Registrations to places in Prince Edward Island,  they shouldn't even be working with a list of places outside Ontario. Equally, for the FindAGrave memorial, he did appear on the 1861 census with birth location as "Canada West" and his descendants just needed to perform one key step (identifying that "CW" means Ontario) to be pointed — literally — in the right direction.