The script of the Voynich manuscript: an alphabet without a key

In Yale University's Beinecke Library, under the shelfmark MS 408, is kept a small parchment codex that for more than a century has resisted every attempt to read it. It is the Voynich manuscript, named after the antiquarian bookseller who bought it in 1912. Its pages are illustrated with unknown plants, astronomical diagrams and female figures immersed in basins, and are covered with an elegant, regular script that no one has managed to decipher convincingly.

A codex with a history

Wilfrid Voynich, a bookseller of Polish-Lithuanian origin working in London, bought the manuscript in 1912 from the Jesuits of Villa Mondragone, near Frascati. With the codex was kept a letter of 1665 or 1666 from the Prague scholar Johannes Marcus Marci, addressed to the Jesuit Athanasius Kircher in Rome. Marci reported a story according to which the Emperor Rudolf II had paid six hundred ducats for the book, believing it to be the work of the English philosopher Roger Bacon.

On the first page, under ultraviolet light, one can also read the name of Jacobus Horcicky de Tepenec, a chemist and apothecary in the service of Rudolf II. These traces place the manuscript in Prague at the beginning of the seventeenth century. In 2009 a radiocarbon analysis carried out at the University of Arizona dated the parchment to between 1404 and 1438, ruling out both the attribution to Bacon, who lived in the thirteenth century, and the hypothesis of a modern forgery on new parchment.

What the script is like

The text is made up of a limited number of signs, between twenty and thirty depending on how the variants are counted, written from left to right with a fluid stroke and without hesitation. Some signs recall Latin letters or the abbreviations of medieval manuscripts, others have no evident parallels. To study it by computer, scholars have developed transcription systems, the best known of which is called EVA, the European Voynich Alphabet.

Statistical analyses have shown that the text displays regularities similar to those of natural languages, such as the repetition of frequent words and the different distribution of signs at the beginning and end of words. At the same time, some features are unusual: very similar words recur at short intervals, and the variety of the vocabulary is lower than expected. These peculiarities feed very different hypotheses.

Hypotheses and attempts

During the twentieth century the manuscript attracted the attention of professional cryptographers, among them William Friedman, who had worked on breaking Japanese codes during the Second World War. Neither he nor his working groups managed to read it. Since then numerous solutions have been proposed, seeing in the text a natural language written in an invented alphabet, a cipher, an artificial language or a meaningless text.

None of these proposals has won the consensus of the scholarly community. The manuscript remains open to investigation, and its digital reproductions, made available by the Beinecke Library, allow anyone to examine its pages. Newspapers periodically announce a definitive decipherment, but so far every proposal has come up against the same difficulty: producing a coherent reading of whole pages, verifiable by other scholars, and not merely of a few carefully chosen isolated words.

Among esoteric alphabets the Voynich holds a place apart, because we do not even know whether it is an esoteric alphabet. Perhaps it holds a treatise on medicine, perhaps a cipher system, perhaps a riddle with no solution. Precisely this uncertainty has made it, for over a century, the very symbol of secret writing.