Just how did we do that?
Recently, after centuries of silence, we heard from beyond the grave, from the bio-father of my Grandma's grandma, Ellen Bagshaw (1846-1901). Ellen has been dead a long time and was a tough cookie. There was some kind of encounter nine months prior to her birth, most likely off the market place in a town like Buxton in early Spring after a cold winter. The protagonists were foolish, fecklish and delirious youth of 23 and 20, intent on embarking on a bit of comfort in the sun, which the sands of time would forget. Something from a Hardy novel. Ellen herself was the antitheses of these qualities and devoted serious time to ensure her own family's future. We had never considered her biological father to be a real breathing person, but he was.
So, here is the news:
My DNA matches screamed Staffordshire, but I didn't have any Staffs ancestry? Piecing together trees of varied 20cM matches led me through new surnames to the Turnock family of Leek and thus an unknown burglar 3xgreat-grandfather (had fling in 1845- Derbyshire).
Let me say, there was no papertrail at all. This 'father' just vanished on arrival. DNA did resolve this, but I definitely could not have predicted this would happen, in advance. So, just how did we do that?
The wind was just 'in the right direction', and a number of factors lined up in making this possible. I am listing them here, and may revise this over time*, and after reflection:
- We would have a surname of the babyfather that is very rare: there are 36 times as many "Mortons" around than these Turnocks, for example.
- The quality of parish register and census data for the area where this group lived, North Staffordshire and southern Cheshire, was excellent, which combined with a rare surname made family tree reconstruction easy.
- We had chosen to test on Ancestry which has a very large database of testers and a very user-friendly interface. I also had a current Ancestry subscription which would help when it came to looking at the trees of matched people.
- I knew the rest of the
family tree very well, so as researcher I could eliminate lines that had
nothing to do with it, and could also identify
an 'alien' group of distant cousins as worthy of exploration.
- I had no other known ancestry in Staffordshire: that would have muddied the waters considerably.
- Ancestry was adamant that we had ancestry in the Potteries, Staffordshire. This meant that I had to take the information seriously. (I had been seeing Staffordshire-based people appear as matches for months and had ignored them.)
- Close relatives of the 'babyfather' (his siblings) had 'umpteen' descendants; and unbeknownst to me, a large number of them had tested (at least 40 I'm thinking) of whom a high percentage shared portions of DNA with us (20 people and rising).
- My relatives unwittingly 'favoured'
this ancestor rather than other ancestors of the same generation - one
does not inherit equal amounts of DNA from grandparents, still less from
those in previous generations. (Reference to 'sticky' DNA removed.)
- A member of an earlier generation had tested - and this increased the number
and quality of matches by an order of magnitude. Without this, I may not have established a connection and the last point would not apply.
- DNA matches themselves were largely co-operative and moderately chatty, enabling a few wrinkles to be smoothed or removed in the family tree.
- I also had three days spare and some experience of this work already, which meant clues were not overlooked but rather exploited, pet theories were ruthlessly demolished and had trained myself to keep going even when there was no obvious path to success.
- I had some experience of tracing families which meant those folks with blank trees,
limited trees or wrong information could still be identified as part of
the family. This was necessary as only 4 of our matches had the name Turnock in the tree.
- I had access to a clustering tool which 'flagged up' groups
of living cousins that were connected to each other by DNA. In fact, slowly working through this tool's output had me pause (for several weeks) as the 'flagged' group could no longer be ignored.
- I was familiar with the concept of 'shared matches' with a reasonable grasp of probability, kinship terminology, genetic inheritance, the 'ThruLines' software.
- And finally, a lone descendant of the 'babyfather' by a subsequent documented marriage was linked by papertrail to him, and had a demonstrably greater portion of shared DNA with us than anyone else from the line, meaning I could attribute parenthood to this gentleman rather than to any of his brothers/nephews.