Science

Algorithm Virtually Unfolds a Historical Letter without Unsealing It


In 1697 a legal professional in the French city of Lille wrote to his cousin in the Netherlands to request a relative’s death certificate, possibly in order to finalize an inheritance. He folded and sealed his mundane missive so it would hold together without an envelope and mailed it off to the Hague. For some reason, it never reached its recipient and remains sealed today. Yet a team of historians and scientists have still read it—thanks to high-resolution imaging and a “virtual unfolding” algorithm.

The scanning-and-unfolding technique could have a variety of applications, from revealing text on artifacts too delicate to touch to reverse engineering historical origami procedures. But the researchers behind the new study, published in Nature Communications, are most excited about using it to examine the art of “letterlocking.” This practice of intricately folding, tucking, cutting and sealing letters to make them function as their own tamper-proof containers was used widely in the era before mass-produced envelopes. Almost two decades ago it captured the interest of Jana Dambrogio, a researcher and conservator at Massachusetts Institute of Technology Libraries and a co-author of the new paper.

Dambrogio had noticed minuscule, apparently intentional cuts and creases in a number of historical documents and eventually guessed their purpose. “It looked like locks, and security was a natural leap—that it was built in on purpose,” she says. “The [reaction] for me was, ‘Oh my gosh, we need to let people know not to remove this evidence, because the little tiny slits and holes and folds are the evidence of this security tradition that is understudied.’”

For the past seven years, she has collaborated in this quest with Daniel Starza Smith, a lecturer in early modern English literature at King’s College London and a co-author of the paper. The two researchers and their team have catalogued more than 60 letterlocking methods. “Before 1850 there weren’t really any such things as modern gummed envelopes,” Smith says. Paper was an expensive commodity not to be wasted on a separate envelope. “So if you wanted to send a letter, you had to use letterlocking—the letter had to become its own envelope or sending device.”

Some letterlocks merely hold pages together. Others are designed to rip when opened, thus revealing any attempts to peek at a letter’s contents. Telltale signs of letterlocking can be seen in the creases and tears of archived material left by many historical figures, including England’s Queen Elizabeth I and her spymaster Francis Walsingham, Italian diplomat and author Niccolò Machiavelli and English poet John Donne—as well as less-famous correspondents such as the average Joe (or rather average Jacques) in Lille whose message was the focus of the new study. The team wanted to keep the Renaissance-era letter sealed to avoid erasing the key signs they would use to reconstruct its folding pattern. For example, a tuck (a subtle manipulation of paper often used in letterlocking) “is like a sneeze,” Dambrogio explains. “It’s ephemeral.”

To view letters without opening them, she and Smith teamed up with M.I.T. computer scientists. “The first step is to create a really high-resolution scan of these letters,” says Amanda Ghassaei, an independent researcher and a co-author of the study. With help from an imaging lab at Queen Mary University of London, the team used a technique called x-ray microtomography, which Ghassaei compares to a medical CT scan—but much more detailed.

“The scans are incredible,” she says. “You can see individual layers of paper.” Once the images revealed the internal folds of the letter packet, she and Holly Jackson, an M.I.T. undergraduate student and a co-author of the paper, wrote code for an algorithm to “open” it—while keeping it locked. “We developed this process for virtually unfolding that data,” Ghassaei says, “so we’re actually figuring out how you go from this folded geometry, that it starts in, to a flat state.”

Then, by mapping the ink on the paper to the virtually “unfolded” version, the algorithm can reconstruct the letter’s text. In the Lille case, it dealt with rather banal family matters. But the algorithm also showed exactly how the letter had been folded 300 years ago. “Along with generating an image of the text, we also create an image that shows what the crease pattern would look like,” Ghassaei says. “You can see the direction that they fold; you can see which [creases] are sharper and which ones are more gentle. So we get a really nice geometric type of analysis of what this folded packet looks like.”

“How people fold things—or the ways in which they close a letter until it arrives and it’s ready to be opened—that’s not a subject that people talk about a lot,” says Paula Findlen, a Stanford University history professor who was not involved in the new study. “Maybe they should, and this … will certainly get people thinking more about looking at all these different folding techniques.”

Findlen suggests the study could also have implications beyond letterlocking. For example, it could hone techniques that help historians study extremely delicate documents, which may be rolled up and stuck together because of water damage, for example. She compares the new study to efforts using x-rays to view the original, worn-away text on recycled documents called palimpsests or to see an early version of a painting hidden under layers of oils. Such techniques offer “a really rich, multilayered history of that artifact by combining traditional scholarly skills with new technology and new ways of peeking under the hood,” Findlen says.
 
The processes developed by Ghassaei and Jackson might also help learn and preserve secrets from the famously intricate art of origami, adds Erik Demaine, an M.I.T. computer scientist and a co-author of the study. “There are a lot of older origami models that were never diagrammed. It was never documented how they were made,” he says. “There’s always a risk that they might get damaged—you could just sit on them, and [the knowledge that made them is] gone. I think this could be a way to conserve those works and reconstruct how they were made so we can understand better how the early origami masters worked.” In addition to preserving ancient techniques, he says, the unfolding algorithm could help modern researchers more easily document each step as they develop a new design. As Demaine notes, “we often fold things in order to check that it is possible to contort paper in a particular way. This would let us analyze from the folded state instead of taking it apart and trying to remember how we put it together in the first place.”

For Dambrogio and Smith, much of the algorithm’s value is in its ability to expand their historical study of letterlocks and how the objects evolved over time. The letter from Lille, for example, is part of the Brienne Collection—a trunk containing 2,600 undelivered messages collected by 17th-century postal employees and eventually bequeathed to a Dutch museum. “One of the interesting things about the Brienne trunk is that over time … there have been distinct changes in fashion, distinct changes in technology, and letters start to look much more like the modern envelope—they’ve got a kind of closing flap, like you’d expect if you bought some envelopes at the shop,” Smith says. “You can see a real story emerging about how communication technology develops over 100, 200 years.”

The trunk offers “a random snapshot of humanity that got trapped in time,” Findlen says. “We have a whole big box full of what a letter looks like when it has been sealed but not opened. This is definitely allowing us to open it further.”


Source link

Show More

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button