Researchers say the centuries-old Voynich manuscript was likely bound with its pages in the wrong sequence, a disorder that has long blocked any real chance at deciphering the text.
Two scholars now argue the famous coded book at Yale University sits out of order, and that fixing the sequence could finally give linguists and codebreakers a fair shot at reading it.
Dr. Lisa Davis of the Medieval Academy of America and Simmons University, working with Dr. Colin Layfield of the University of Malta, used statistical analysis of the text. They found that pages now sitting next to each other show very low similarity, while some widely separated chunks share striking statistical matches. Their paper appeared in the journal Digital Medievalist.
Daily Mail reported the pair’s conclusion that a later binder who could not read the script simply assembled the sheets wrong.
Davis told interviewers the physical clues line up once the binding error is considered. Handwriting shifts, illustrations that vanish into the gutter and reappear pages later, and mismatched water stains all make more sense under their model.
"It doesn't get us there [to a full translation], but it moves the ball because you can't make sense out of a book where all the pages are mixed up."
She added that the odd physical structure itself raises questions.
"We don't know if it's a code, but there are some odd characteristics of the Voynich manuscript in terms of its physical structure."
Davis pointed to drawings that run off one page and continue on another that is no longer adjacent. In a normal bound book that should not happen.
"There are several examples where the image goes off the page in the middle and comes out on the other side, even though there are a bunch of pages in between."
She said the same logic clears up other headaches that have dogged Voynich researchers for decades.
"When you start to think about it in the context of the other strange features of the manuscript, all of those problems are resolved."
One more concrete gain: the pattern of different scribes.
"It solves the problem of each scribe writing one sheet. It solves the problem of the images crossing the gutter."
Medieval books were commonly built from nested bifolia, large sheets folded once and stacked five deep into a quire. Davis and Layfield propose the Voynich was never meant to work that way. They suggest it was produced as separate folded bifolia, or “singulons,” meant to be read one pamphlet at a time like a stack of short booklets. A later owner or binder who could not decipher the script forced those separate units into a single conventional binding and scrambled the order.
The manuscript itself runs about 240 pages and holds roughly 38,000 words written in the still-unread script called Voynichese. Carbon dating places the parchment in the early 15th century. For more than 600 years it has defeated historians, botanists, and professional codebreakers.
Earliest firm records put the book in Prague. Alchemist George Baresch owned it by 1639 and sent a sample to the Jesuit scholar Athanasius Kircher in Rome, hoping for a translation. Baresch refused to surrender the volume itself. After Baresch died, his friend Jan Marek Marci of Charles University forwarded the full manuscript to Kircher with a letter dated August 19, 1665. That letter claimed Holy Roman Emperor Rudolf II had once paid 600 gold ducats, about four and a half pounds of gold, for it more than a century after the pages were made.
The book apparently stayed among Kircher’s papers in Rome until Italian troops under Victor Emmanuel II annexed the Papal States in 1870. Polish-American book dealer Wilfrid Voynich bought it in 1912, put it on public display in 1915, and gave the manuscript its modern name. After his death the volume passed to his wife Ethel, then to dealer Hans P. Kraus, who donated it to Yale’s rare book library in 1969. It remains there today.
Serious cryptanalytic efforts began early. William Friedman, the American code expert, corresponded with Voynich in the 1920s and later floated the idea of a synthetic language. British intelligence officer John Tiltman examined the text for Friedman and judged it too complex for a simple cipher. German code-breaker Erich Hüttenhain later compared notes with the U.S. National Security Agency. During the Cold War the FBI even looked at the book under a theory it might be communist propaganda. No one produced a working reading.
Priest Theodore C. Peterson finished a painstaking hand copy in 1944 so others could study the pages without handling the original; he kept at the problem until he died. Danish botanist Theodore Holm used that copy to offer tentative identifications of 16 plant drawings. None of it cracked the script.
More recent claims have not aged well. University of Bristol researcher Dr. Gerard Cheshire announced a proto-Romance decipherment that drew sharp criticism. The university publicly distanced itself, stating the work was “entirely the author's own work and is not affiliated with the University of Bristol, the Faculty of Arts or the Centre for Medieval Studies.” Davis herself had earlier dismissed that effort as gibberish.
Stories that Alan Turing tried and failed circulate, yet the document notes contain no concrete evidence he ever examined the pages.
Davis and Layfield are careful about what their statistical work actually delivers. They do not claim a translation. They claim the current page order is wrong and that the original format was probably a set of independent folded sheets. Correct the sequence, they argue, and earlier linguistic and cryptographic studies become worth another look.
They close their paper with the hope that linguists and cryptologists will revisit past work in light of the singulon hypothesis and open a new chapter in Voynich studies. Davis has said her next target is the large water stains, which may supply still more physical evidence of how the sheets once sat.
The book has survived emperors, alchemists, Jesuit scholars, World War code rooms, and modern universities. Every flashy announcement of a full solution has collapsed. A quiet statistical case that the binder simply got the order wrong is less dramatic, and far more useful. If the pages truly sit scrambled, then six hundred years of effort have been spent reading a deck that was never shuffled back into sequence.
Careful measurement beats another round of unearned breakthrough claims. Get the physical book right first; the text can wait its turn.
