When AI Connected a Dower Record to a PIece of Medical History

 

When AI Connected a Dower Record to a Piece of Medical History

Have you ever been working with a record for one reason only to discover that one of the people named in it had a much bigger story?

That's what happened while I was working on a recent Friday Find involving Thomas D. Crawford and his widow, Jane.

I started with a record, not with a famous person.

The Original Record

I have had a photocopy of a Jefferson County, Indiana, probate record in my files for years. I probably obtained the copy during a research trip to the Family History Library in Salt Lake City. More recently, I found the record again in the digitized images on FamilySearch and obtained a much better copy.

The record comes from the February 1836 term of the Jefferson County Probate Court and is headed:

Jane Crawford vs. The Heirs of Thomas D. Crawford Decd — On Petition for Dower

The proceeding involved the partition of land belonging to Thomas D. Crawford. His property was described as the Southwest Quarter of Section 26, Township 4 North, Range 10 East, containing about 113 acres.

Commissioners appointed by the court examined the property and set aside approximately 43¾ acres for Jane Crawford as her dower. Their description began at the northwest corner of the quarter section on the Michigan Road and followed a series of courses and distances around the parcel.

That land description was what interested me.

Asking ChatGPT to Help with the Record

After creating an edited transcription, I asked ChatGPT to pull the land description from it. ChatGPT separated the description of Thomas Crawford's entire 113-acre tract from the metes-and-bounds description of the 43¾ acres assigned to Jane.

I then provided another piece of information: I believed the Thomas Crawford in the record was the man identified as GVRH-QF4 on FamilySearch, whose wife was Jane Todd.

My intent was simply to give ChatGPT enough information to write a Friday Find post about the probate record and the partition of Thomas Crawford's land.

Instead, that little bit of identifying information opened another door.

Who Was Jane Todd Crawford?

While preparing the draft, ChatGPT searched for additional information about Thomas Crawford and Jane Todd.

It found something I didn't know.

The Jane Crawford petitioning for her dower appeared to be Jane Todd Crawford, a woman remembered in the history of medicine for surviving a remarkable operation performed by Dr. Ephraim McDowell.

Suddenly, Jane was no longer simply the widow named in an 1836 Jefferson County probate record.

According to historical accounts, Jane Todd married Thomas Crawford in 1794. The couple lived in Virginia and Kentucky before eventually moving to the Madison, Indiana, area.

While living in Kentucky, Jane became seriously ill in 1809. Her condition was initially thought to be a pregnancy, but Dr. Ephraim McDowell determined that she had a large ovarian tumor.

McDowell performed surgery to remove it.

Today, that operation is recognized as the first successful removal of an ovarian tumor. This was decades before modern anesthesia and long before surgeons understood antiseptic techniques.

Jane survived.

She went on to live for more than thirty years after the operation.

From a Medical-History Figure to a Widow Seeking Dower

What particularly interests me is the contrast between the historical accounts and the probate record.

Medical history remembers Jane Todd Crawford, surgical patient.

The Jefferson County probate record documents Jane Crawford, widow and landholder.

By 1835, Thomas Crawford had been dead for a number of years. Jane petitioned the probate court to have her dower formally set apart from his property.

On February 9, 1836, the court-appointed commissioners went onto the land and determined which portion Jane should receive. They considered the value of the different parts of the property and assigned her approximately 43¾ acres.

The following day, their report was presented for the court's approval.

The record doesn't mention Jane's surgery. There would have been no reason for it to do so. To the probate court, the important facts were that she was Thomas Crawford's widow and that she was entitled to dower from his land.

Yet knowing the larger story changes the way I read the record.

The woman seeking her share of this Indiana property appears to have been the same woman who, more than a quarter century earlier, had endured an operation that became a landmark in medical history.

An Example of AI Adding Context

This experience also illustrates one of the ways I'm finding AI useful in genealogy.

I didn't ask ChatGPT, "Who was Jane Todd Crawford?"

I was concentrating on the land.

I supplied the transcription, identified the probable Thomas Crawford on FamilySearch and mentioned that his wife was Jane Todd. ChatGPT used those clues to search for historical information that could provide context for the Friday Find post.

That search uncovered Jane's connection to Dr. Ephraim McDowell and her famous 1809 surgery.

ChatGPT then incorporated that information into the draft while keeping the probate record and its land description at the center of the post.

But AI Didn't Prove the Connection

There is an important distinction here.

Finding information about a Thomas Crawford and Jane Todd who lived in Jefferson County, Indiana, does not by itself prove that they are the Thomas and Jane Crawford named in my probate record.

The historical information provides a strong lead, but the genealogical identification still needs to be supported by records.

That's where my research comes back into the process.

I can compare the probate file with other records for Thomas and Jane Crawford, investigate the ownership history of the Southwest Quarter of Section 26, and look at the heirs named in Thomas Crawford's estate. I can also examine records for neighboring landowners—including John Brisben, whose property line appears in the dower description.

In other words, AI found the connection. It didn't finish the research.

A Different Kind of Friday Find

I started with an old photocopy and a question about a land description.

That led to a clearer FamilySearch image, a transcription of the boundaries of Jane Crawford's 43¾-acre dower parcel, and eventually a connection I had never made before.

The widow in my deed may also be a woman with a remarkable place in medical history.

And I might never have noticed that connection if I hadn't asked ChatGPT to help me write about an old probate record.

Sometimes the most interesting thing AI contributes to my genealogy research isn't an answer.

It's another question to investigate.