Documenting mistakes in our documentation

If someone points out a typo in an online Finding Aid or a Hamnet catalog record, we gratefully say thank-you, fix it, and (usually) move on.¹ Sometimes, though, a big enough mistake has been around for a long enough time that we can’t just move on. We have to take extra steps to find the source of the mistake, and make sure its ghost doesn’t come back to haunt scholarship. Take “Edward Themilthorp” for example. This was, supposedly, a letter from him:

Folger L.d.586 (leaf 1 recto)

The letter is Folger L.d.586, dated 28 November 1616, and it had been described as “Letter from Edward Themilthorp to Nathaniel Bacon” in the online finding aid to the Papers of the Bacon-Townshend family of Stiffkey, Norfolk.

A while back, a researcher in England contacted reference@folger.edu to say that he had done extensive research on the Themilthorp family of Norfolk, and knows of no “Edward” who would fit the context. Nicholas Themilthorp, on the other hand, would make perfect sense. What was the evidence for this letter being from an Edward Themilthorp? We didn’t know, so we set to work.

Sometimes, a reported error can be traced to the fact that early modern handwriting has some very different letterforms than modern cursive. “Edward” and “Nicholas” are about the same length. Was it possible the current researcher or a long-ago cataloger had misread the first name? Nope. A quick look at the signature in LUNA, the Folger’s digital image collection, showed that the signature only had a first initial, not a full name:

“N. Themylthorpe” signature at the end of the letter (detail from Folger L.d.586 leaf 1 recto)

The signature had so many decorative flourishes that it was hard to untangle the capital letters at first, but there didn’t seem to be a capital E anywhere. It definitely looked like an elaborate N joined up with the T at the start of the surname—but maybe the writer just had a really strange way of writing the letter E? Maybe other instances of E (or N) in the letter would tip the balance? Nope. The handwriting in the body of the letter is completely different from the handwriting of the signature. Themilthorp’s secretary presumably wrote out the letter, then gave it to him to sign.

Handwriting sample from the body of the letter, immediately above the signature (detail from Folger L.d.586 leaf 1 recto)

Maybe the name “Edward” came from the brief summary provided by the recipient or his secretary, after folding the letter into a tall skinny rectangle for storage with other correspondence? Known as a “docket” or “endorsement”, these summaries at the top of the back of a folded letter sometimes give important clues about the letter’s content.² Nope again. The person who docketed the letter just wrote “Mr. Themelthorp” on the first line:

Docket on the back of the letter (detail from Folger L.d.586 leaf 2 verso)

In the end, the only evidence we could find for the sender being “Edward” was the sixth line on page 30 of the Calendar of the Bacon-Townshend collection of manuscripts in the Folger Library, a finding aid compiled by Folger curator Giles Dawson in 1958.

First six lines on page 30 of Calendar of the Bacon-Townshend collection of manuscripts in the Folger Library (detail from Folger Z6621.F61 B3 R.R.)

Now we had a new question. Where did Giles Dawson get “Edward” from? It seems quite extraordinary that someone with his expertise would have misread the first letter as an “E” when writing up the arrangement of the papers. Here’s where knowledge of past practice comes in handy: Giles Dawson would not have done his own typing. He would have written up his notes by hand, then given them to a typist. Sometimes the handwritten versions survive in the curatorial files, like this description from 1960.³

Example of a handwritten note by Giles E. Dawson, initialed and dated “GED 17 July 60” (image from Folger Handwriting Samples file)

Now that you know the original finding aid was handwritten, take another look at the entirety of page 30 of the typed version. See all those Edwards in a row?

Page 30 from Calendar of the Bacon-Townshend collection of manuscripts in the Folger Library (Folger Z6621.F61 B3 R.R.)

My best guess is that Giles Dawson correctly handwrote “N.” (or “Nich.” or “Nicholas”), but thanks to the typist’s eye-skip while transcribing, “Edward” landed there instead. I haven’t found the original handwritten version of the typed finding aid, so I can’t prove any of this, but I’m confident enough that I gave it as a possible reason when updating the online finding aid. It now has the note “Sender misidentified as Edward Themilthorp in Calendar of the Bacon-Townshend Manuscripts, compiled ca. 1958 (perhaps an error of the typist). Corrected 2018-10-15.”

Excerpt from updated online finding aid at https://findingaids.folger.edu/dfobacontownshend.xml

Given that “Edward Themilthorp” existed at the Folger for sixty years before we discovered he was a typo, we didn’t want to make a silent correction. It’s important that his name remain keyword searchable, alongside the date of his demise. If someone’s research notes refer to a 1616 letter by Edward Themilthorp at the Folger, and no one can find that letter, we want them to know that it’s Edward who disappeared, not the letter. We do the same thing with corrections in Hamnet records. If someone notices that Hamnet has “drawnigs” instead of “drawings” we just fix it. But if the record says “includes 36 original pen and wash drawings by Francis Hayman” instead of “31 by Francis Hayman, 5 by Hubert Gravelot,” we leave a trail of breadcrumbs.⁴

In conclusion, in case it wasn’t already obvious: if you find a mistake, please tell us! Catalogers are human, all humans make mistakes, therefore Socrates is a cataloger. No, wait… that’s a mistake.