To puzzle-solving aficionados, the allure of an uncracked code can be overpowering. Some have dedicated their whole lives to solving famous cases such as the Beale Ciphers, the Dorabella Cipher, and the Tamám Shud murder. But cracking such codes brings more than just a chance at possible wealth or personal satisfaction. Codebreakers like Alan Turing and Elizebeth Smith Friedman have been catapulted to the hero status of Indiana Jones crossed with Robert Langdon, their names writ large in the global history of brilliant minds, and for some, joining this star-studded constellation is the only worthwhile reward. One way to guarantee meteoric ascendance would be to solve one of the most alluring cases of all: the Voynich manuscript.
Less glamorously known as MS408, this roughly 234-page codex is illustrated with plants, women, and astrological signs, but the real magic is in the writing. The tantalising script, which looks like a hybrid of elfish and Arabic, is thought to be the only known sample of this writing system. British intelligence, the FBI, and countless others have spent more than a hundred years trying to decode it, but so far, no one has succeeded.
- Medieval manuscripts and the First World War (podcast)
- Cuneiform: 6 things you (probably) didn’t know about the world’s oldest writing system
- The evolution of British code-breaking
- Unusual things found in medieval manuscripts
Then, on 15 May 2019, like a bolt from the blue, Bristol University heralded that their academic Dr Gerard Cheshire had bested the brilliant minds of the past century and deciphered the Voynich manuscript in a mere two weeks. The media reaction was spectacular. Countless outlets feverishly reproduced the claims in dozens of countries, seemingly ending one of the golden mysteries of our time.
But then the internet – that fastest and most brutal of critics – responded. Intrigued academics in fields such as codicology, mediaeval studies, comparative linguistics, and cryptography were quick to read the underlying research, and even quicker to tweet their dismay that the claims were being taken seriously. In one Twitter thread, Liverpool classicist Ben Cartlidge described some of the worst problems, and the blog Language Log published a post that pointed out other issues besides. But if the claims are so dubious, how did this story ever make headlines in the first place? To answer this, we must go back to how research is accepted for publication.
OK so are you ready for a live-tweeted reading of the new #Voynich MS article?— Ben Cartlidge (@benjcartlidge) May 16, 2019
Once an academic has written an article, they typically find a relevant journal. Relevance is a crucial first step. Choose an off-topic journal and the editor(s) should instantly reject the work as being outside of their expertise. In this case, Cheshire submitted what we might call a linguistic analysis to Romance Studies, a “fully refereed journal devoted to the study of the Romance literatures and cultures”, and for whatever reason, Romance Studies deemed it relevant enough to progress to peer-review. (NB. Literature and linguistics are very different!)
In the humanities, articles are typically peer-reviewed by two or three specialists in the topic, and the author and reviewers should have no knowledge of each other’s identities. This double-blind system theoretically stops people from scuppering enemies and promoting friends. However, peer-review also has plenty of problems. One is that this highly skilled, protracted job is unpaid, so many academics simply decline all requests. Academics are also notoriously swamped, so even those that do accept may squeeze a review into a spare hour, and then rush through it with only half of their brain engaged. So much the worse if the paper is on a topic outside of their scope.
This takes us to Cheshire’s defence as the discontent about his work grew ever louder. When challenged by the Guardian, he replied: “The journal paper has been blind peer-reviewed and verified by other scholars – that is standard confirmation in the scientific arena.” But as we’ve already seen, the choice of journal was problematic, and this casts doubts over the reviewers’ expertise. Also, peer-review is not standard confirmation in the scientific arena. Far from it, in fact. In academia, the idea is that a scholar works on a theory or method, and then, as the very first step towards getting their work judged by the wider academic community, they publish it in a (relevant!) peer-reviewed journal. That community is then well within its rights to celebrate the work as a masterpiece, or burn it to the ground as a monstrosity. In fact, scientific claims being fiercely repudiated by the academic community is nothing new. One need only look to cases like Andrew Wakefield’s infamous links between the MMR vaccine and autism, or the harrowing stories of Roy Meadow’s “statistics” about sudden infant death syndrome. Instances like these are precisely why sites like Retraction Watch exist.
The problem from the media’s perspective, however, is that formal rebuttals can take months or years to be published, and a breaking story like cracking the Voynich manuscript has a shelf-life of maybe hours. Also, as mentioned above, for most non-academics, publication in an academic journal stamps the work with a kitemark of innate validity. From there, all we can do is speculate about the ensuing events. Perhaps Cheshire took his newly published article to Bristol University’s media team himself. Whatever the case, they put out excited press releases in good faith. Even if the first critical arrows had shot back through the internet only seconds later, and even if Bristol had instantly clawed back their press releases, it was already too late. In moments, the story had raced round the globe through news wires and social media. In the end, for Bristol, there was only one way forward: the excruciating public statement. Predictably, their acknowledgement that “concerns have been raised about the validity of this research” triggered an inevitable cascade of follow-up stories as the media exuberantly documented the embarrassing climb-down.
So what are we left with? Well, the Voynich manuscript has not, as yet, given up its secrets, but it has still taught us something about the way that scientific claims can make their way into the media. In this case, it shows us the vital importance of peer-review, but the problems inherent in that system may prove just as intractable as any ancient, mysterious codex.
Think you can crack the Voynich manuscript? You can download a fully digital image-based version at Yale University’s Beinecke Rare Book and Manuscript Library, here. Good luck!
Dr Claire Hardaker is a senior lecturer in Forensic Corpus Linguistics at Lancaster University. Her podcast, en clair, is all about forensic linguistics, literary detection, language mysteries, cryptography, ancient undeciphered languages, and more. You can follow her on Twitter at @DrClaireH.