Some people eat, sleep and chew gum, I do genealogy and write...

Friday, March 9, 2018

The Impact of the MyHeritage Family Tree Study on Genealogical Research

https://blog.myheritage.com/2018/03/myheritage-science-teams-research-featured-in-the-prestigious-journal-science/
One of the common challenges of doing genealogical research is differentiating people with the same or very similar names. This problem is directly addressed by my Rule Nine: There are patterns everywhere. See "New Rules Added to the Old: The Rules of Genealogy Revisited." As the explanation for this rule indicates, computers are very good at looking for and finding these patterns. MyHeritage.com recently posted about the following scientific article.

Kaplanis, Joanna, Assaf Gordon, Tal Shor, Omer Weissbrod, Dan Geiger, Mary Wahl, Michael Gershovits, et al. “Quantitative Analysis of Population-Scale Family Trees with Millions of Relatives.” Science, March 1, 2018, eaam9309. https://doi.org/10.1126/science.aam9309.

This article is a masterpiece that fully supports Rule 9. The problem addressed by the findings of the study reported in the article and by Rule 9 is that many researchers focus on names to the exclusion of other important information about ancestral families. For example, I commonly find English families with children born in different locations around the country in the early 1700s. Here is one example.


In this family, the parents are shown to have been born and married in Sussex, England. Incidentally, the first three children lack birth information. The last child is said to have been born in Tenterden, Kent, England.

Thomas Byant, 1702-1770 christened in Beckley, Sussex, Kent.
Elizabeth Hovenden, 1700-1734 christened in Northiam, Sussex
daughter May Bryant, 1724 christened in Northiam, Sussex
daughter Elizabeth Bryant, 1727 christened in Tenterden, Kent
son John Bryant, 1730-1762 christened in Tenterden, Kent
daughter Rebecca Bryant, 1734-1801 christened in Tenterden, Kent

There is always an argument, that this was possible. But what about the sources? There is a marriage record for Thomas Bryant marrying Elizabeth Hovenden in Northiam, Sussex. There are also christening records for each of the four children in Tenterden with parents named Thomas and Elizabeth.

Here is a map of England showing where these two places are located.


Could the places in this family be possible? Possible yes, likely no. Why? Now, back to the MyHeritage study. Here is a quote from the MyHeritage.com blog about the study.
The team found that industrialization profoundly altered family life. Before 1750, most Americans found a spouse within six miles of their birthplace, but for those born in 1950,  [1850?] that distance had stretched to about 60 miles. Before 1850, marrying in the family was common — on average, fourth cousins married each other, compared to seventh cousins today. Curiously, they found that between 1800 and 1850, people traveled farther than ever to find a mate — nearly 12 miles on average — but were more likely to marry a fourth cousin or closer. Their hypothesis is that changing social norms, rather than rising mobility, may have led people to shun close kin as marriage partners. In a related observation, they found that women in Europe and North America have migrated more than men over the last 300 years, but when men did migrate, they traveled significantly farther on average.
Apparently, no problem here. Both the parents were supposedly born and married in Northiam. But what about the children? Would it help to know that this husband and wife were likely related? Let's see how likely it is that the children have the right parents.

Using another program, Findmypast.com, I can see how many Bryants there are in the millions of records on this huge website. Searches on Findmypast.com show that during the time period of 1702 plus or minus two years, there were 52,027 records of Bryants in England. By editing the search, I find that there are only about 503 records of Bryants in Sussex during that same time period and of those only 41 were named Thomas. Interestingly, none of these records show any births. However, there is a Thomas Bryant who dies in Northiam in 1762. However, the records on FamilySearch.org do show a marriage for Thomas and Elizabeth in Northiam in 1724.

Is this the same Thomas and Elizabeth that are shown as the parents in Tenterden? That is the question. The MyHeritage.com study would tend to make me consider the conclusion as unlikely.

Ignoring for the moment that we do not have a source for Thomas Bryant's birth, let's look at the Bryants in Tenterden, Kent about nine miles away. Back to Findmypast.com.

There are 1.106 records listed for Bryants in Kent in around 1702 and 68 of them are named Thomas. There are again no christening records for a Thomas Bryant but there is one death record in Tenterden. So, we have a Thomas in this time period who dies in Tenterden and another one who dies in Northiam, where Thomas and Elizabeth are married. A search in Ancestry.com shows that there are no marriage records for a Thomas and Elizabeth in that time period in Tenterden. Ancestry.com only has a death record.

Do we have the right parents for these four children? How does the information provided by the MyHeritage.com study impact your opinion as to whether or not we have found the right parents? Does the fact that we have no birth information for the father in either location and that there is a Thomas Bryant who dies in Northiam have any bearing on your conclusion?

I think that the study done by MyHeritage.com et al. is going to become an important factor in determining the reasonableness of many conclusions. 

2 comments:

  1. I think there's an instance of "1950" that should be "1850" in your article, James.

    ReplyDelete