Perfectly Clear — and Inscrutable

 

How can handwriting be so perfectly clear, and yet not communicate the information we need? That's what had me stuck last week as I puzzled over Rachel Riley Boothe's application for a widow's pension.

Granted, there were some rough spots in the several pages of her application, like the place recording her answer about her husband Alexander Boothe's place of birth. Nancyman? Perhaps that meant Nansemond, a county in Virginia at the time of her 1906 application.

But when it came to a series of questions about her own background, I was both excited to glean more information and equally stymied by the content of her response. The official completing the form used the clearest handwriting for question number three: "When and where were you born and what was your maiden name?" And yet, for the answer about her place of birth, I am at a loss to find the designated location.

While it is obvious that the answer was entered piecemeal, as each section of the information was added. The handwriting for the date of birth and Rachel's maiden name, for instance, seems different—more compressed and smaller than the handwriting that provided the beginning of the answer.

For one thing, while the state name seemed to be misspelled, it is obvious that the answer was South Carolina. However, in retrospect now that we have the resources to view each decennial census, I can't say I am confident of even that answer. For one thing, the earlier census records—1850 and 1860—reported Rachel's place of birth as North Carolina. While anyone's guess is good for the enigmatic entry in the 1870 census, the answer "South Carolina" didn't appear until the 1880 census.

Perhaps it was Alexander himself who provided the answer about his wife's place of birth for those earliest records. An answer like North Carolina would make sense for those living in the northeastern edge of Tennessee, where North Carolina was just over the mountain range. But with the answer Rachel provided on her pension application leading me to a dead end, I wondered whether perhaps even she had gotten the two states mixed up.

The handwriting was so clear on her answer. I enlarged the copy to make sure I was seeing it correctly, though I already knew what I was seeing.

The trouble was, there was no such county by that name in either South Carolina or North Carolina. Remembering that Virginia had some "extinct" counties—after all, that's the case with Alexander's own place of birth, Nansemond County in Virginia—I tried looking for historic names. Remembering "Nancyman," I tried imagining how a county's name might have been spelled phonetically. Still, no clue as to what she meant by her answer. Tobarrass County? Lobarrass? Fobarrass?

Over the weekend, I chatted back and forth with fellow blogger, Charles Purvis of Carolina Family Roots, who offered to help figure out the South Carolina location. Charlie has been experimenting with Artificial Intelligence, and offered to run a few questions through two different systems: Perplexity.AI and ChatGPT.

Based on input regarding the possibility of phonetic spelling coupled with variances in regional accents plus challenges with literacy levels, Perplexity read the entry as "Starrass" and offered a possible answer of Starr, South Carolina. ChatGPT, taking a different approach, noted that the optical character recognition result was distorted but managed to read the entry as "Spartansburg," which obviously would refer to present-day county of Spartanburg.

While further input and refinements would make AI assistance more helpful—not to mention, it's likely a tool which we all could develop skill in using—I can't help but go back to the original document and stare at that crystal clear handwriting, just wondering what the clerk meant when his impeccable handwriting penned that entry.