Thursday, May 15, 2014

Analyze This! Pattern Recognition in Genealogy

As budding (or not-so-budding) genealogists, we're taught to ask specific questions that will guide our search for evidence. But on our hopeful journey to a conclusion, we may find ourselves surrounded with data, and looking for a pattern in a sea of (say) seven dozen deeds.

Now, we pulled those deeds because they might be relevant (right county or close, relevant surnames or close, right century). But which ones will actually help and how is not always so obvious -- especially since difficult cases may have us hunting for a pattern that does not appear in any particular record by itself.

Of course it's essential to be immersed in the subject and the families. Beyond that I like the "kaleidoscope" approach. How many ways can I rearrange the data? Table? Spreadsheet sortable on all different fields? Timeline? Color-coded list for particular properties? Maps? Compared to the nearest census, or church membership book?

How do you ferret out patterns in your work?

Illustration from "Rabbit-Duck Illusion," Wikipedia ( : viewed 11 May 2014), citing "Kaninchen und Ente" ("Rabbit and Duck"), Fliegende Bl├Ątter, 23 October 1892.

Harold Henderson, "Analyze This! Pattern Recognition in Genealogy," Midwestern Microhistory: A Genealogy Blog, posted 15 May 2014 ( : viewed [date]). [Please feel free to link to the specific post if you prefer.]


Judy G. Russell said...

Hard enough going back and forth between the rabbit and the duck. Now you want us to find a ferret??? (Sorry. Couldn't resist.)

Harold Henderson said...

You really can't see it?????

Geolover said...

"How do you ferret out patterns in your work?" is an extremely useful question. A coyote knows that the prairie dog town has a number of different access tunnels.

Patterns in one's work is not limited to patterns in documentation. Learning from others' methods includes evaluation of possible mistaken assumptions and applying them to one's own approaches.

A viewer of a list of debtors to the estate of a person who died in a city in Alabama may assume that the debtors also lived there -- but did the decedent just recently move there from (say) central New York?

A reader of a list of a decedent's heirs who were grantees of a KY deed asked, "who were the children?" not bothering to do the follow-up research as to identities of the listed heirs, and came up with a partially wrong answer. Thus a published article listed a nonexistent daughter rather than a widow who was second wife to the deceased. The author was not aware of the second wife and failed to do the research in 3 states to learn her identity.

Harold Henderson said...

Good points. Thorough research is necessary to bring out patterns; sometimes it's sufficient, sometimes something more is required.