Artificial Intelligence May Have Cracked the Code of the Voynich Manuscript: Has Modern Technology Finally Solved a Medieval Mystery?

What is it about the Voyn­ich Man­u­script—that cryp­tic, illus­trat­ed 15th cen­tu­ry text of unknown ori­gin and meaning—that has so fas­ci­nat­ed and obsessed schol­ars for cen­turies? Writ­ten in what appears to be an invent­ed lan­guage, with bizarre illus­tra­tions of oth­er­world­ly botany, mys­te­ri­ous cos­mol­o­gy, and strange anato­my, the book resem­bles oth­er pro­to-sci­en­tif­ic texts of the time, except for the fact that it is total­ly inde­ci­pher­able, “a cer­tain rid­dle of the Sphinx,” as one alchemist described it. The 240-page enig­ma inspires attempt after attempt by cryp­tol­o­gists, lin­guists, and his­to­ri­ans eager to under­stand its secrets—that is if it doesn’t turn out to be a too-clever Medieval joke.

One recent try, by Nicholas Gibbs, has per­haps not lived up to the hype. Anoth­er recent attempt by Stephen Bax, who wrote the short TED Ed les­son above, has also come in for its share of crit­i­cism. Giv­en the invest­ment of schol­ars since the 17th cen­tu­ry in crack­ing the Voyn­ich code, both of these efforts might jus­ti­fi­ably be called quite opti­mistic. The Voyn­ich may for­ev­er elude human under­stand­ing, though it was, pre­sum­ably, cre­at­ed by human hands. Per­haps it will take a machine to final­ly solve the puz­zle, an arti­fi­cial brain that can process more data than the com­bined efforts of every schol­ar who has ever applied their tal­ents to the text. Com­put­er sci­en­tists at the Uni­ver­si­ty of Alber­ta think so and claim to have cracked the Voyn­ich code with arti­fi­cial intel­li­gence (AI).

Com­put­er sci­ence pro­fes­sor Greg Kon­drak and grad­u­ate stu­dent Bradley Hauer began their project by feed­ing a com­put­er pro­gram 400 dif­fer­ent lan­guages, tak­en from the “Uni­ver­sal Dec­la­ra­tion of Human Rights.” While “they ini­tial­ly hypoth­e­sized that the Voyn­ich man­u­script was writ­ten in [ancient] Ara­bic,” reports Jen­nifer Pas­coe, “it turned out that the most like­ly lan­guage was [ancient] Hebrew.” (Pre­vi­ous guess­es, the CBC notes, “have ranged from a type of Latin to a deriva­tion of Sino-Tibetan.”) The next step involved deci­pher­ing the manuscript’s code. Kon­drak and Hauer dis­cov­ered that “the let­ters in each word… had been reordered. Vow­els had been dropped.” The the­o­ry seemed promis­ing, but the pair were unable to find any Hebrew schol­ars who would look at their find­ings.

With­out human exper­tise to guide them, they turned to anoth­er AI, whose results, we know, can be noto­ri­ous­ly unre­li­able. Nonethe­less, feed­ing the first sen­tence into Google trans­late yield­ed the fol­low­ing: “She made rec­om­men­da­tions to the priest, man of the house and me and peo­ple.” It’s at least gram­mat­i­cal, though Kon­drak admits “it’s a kind of strange sen­tence to start a man­u­script.” Oth­er analy­ses of the first sec­tion have turned up sev­er­al oth­er words, such as “farmer,” “light,” “air,” and “fire”—indeed the sci­en­tists have found 80 per­cent of the man­u­scrip­t’s words in ancient Hebrew dic­tio­nar­ies. Fig­ur­ing out how they fit togeth­er in a com­pre­hen­si­ble syn­tax has proven much more dif­fi­cult. Kon­drak and Hauer admit these results are ten­ta­tive, and may be wrong. With­out cor­rob­o­ra­tion from Hebrew experts, they are also unlike­ly to be tak­en very seri­ous­ly by the schol­ar­ly com­mu­ni­ty.

But the pri­ma­ry goal was not to trans­late the Voyn­ich but to use it as a means of cre­at­ing algo­rithms that could deci­pher ancient lan­guages. “Impor­tant­ly,” notes Giz­mo­do, “the researchers aren’t say­ing they’ve deci­phered the entire Voyn­ich man­u­script,” far from it. But they might have dis­cov­ered the keys that oth­ers may use to do so. Or they may—as have so many others—have been led down anoth­er blind alley, as one com­menter at IFL Sci­ence sug­gests, sar­cas­ti­cal­ly quot­ing the wise Bull­win­kle Moose: “This time for sure!”

You can find the Voyn­ich Man­u­script scanned at Yale’s Bei­necke Rare Book & Man­u­script Library. Copies can be pur­chased in book for­mat as well.

Relat­ed Con­tent:

An Ani­mat­ed Intro­duc­tion to “the World’s Most Mys­te­ri­ous Book,” the 15th-Cen­tu­ry Voyn­ich Man­u­script

Behold the Mys­te­ri­ous Voyn­ich Man­u­script: The 15th-Cen­tu­ry Text That Lin­guists & Code-Break­ers Can’t Under­stand

1,000-Year-Old Illus­trat­ed Guide to the Med­i­c­i­nal Use of Plants Now Dig­i­tized & Put Online

Josh Jones is a writer and musi­cian based in Durham, NC. Fol­low him at @jdmagness


by | Permalink | Comments (5) |

Sup­port Open Cul­ture

We’re hop­ing to rely on our loy­al read­ers rather than errat­ic ads. To sup­port Open Cul­ture’s edu­ca­tion­al mis­sion, please con­sid­er mak­ing a dona­tion. We accept Pay­Pal, Ven­mo (@openculture), Patre­on and Cryp­to! Please find all options here. We thank you!


Comments (5)
You can skip to the end and leave a response. Pinging is currently not allowed.
  • Nikolai says:

    There is a key to cipher the Voyn­ich man­u­script. The man­u­script was not writ­ten in Hebrew.
    The key to the cipher man­u­script placed in the man­u­script. It is placed through­out the text. Part of the key hints is placed on the sheet 14. With her help was able to trans­late a few dozen words that are com­plete­ly rel­e­vant to the theme sec­tions.
    The Voyn­ich man­u­script is not writ­ten with let­ters. It is writ­ten in signs. Char­ac­ters replace the let­ters of the alpha­bet one of the ancient lan­guage. More­over, in the text there are 2 lev­els of encryp­tion. I fig­ured out the key by which the first sec­tion could read the fol­low­ing words: hemp, wear­ing hemp; food, food (sheet 20 at the num­ber­ing on the Inter­net); to clean (gut), knowl­edge, per­haps the desire, to drink, sweet bev­er­age (nec­tar), mat­u­ra­tion (matu­ri­ty), to con­sid­er, to believe (sheet 107); to drink; six; flour­ish­ing; increas­ing; intense; peas; sweet drink, nec­tar, etc. Is just the short words, 2–3 sign. To trans­late words with more than 2–3 char­ac­ters requires knowl­edge of this ancient lan­guage. The fact that some sym­bols rep­re­sent two let­ters. In the end, the word con­sist­ing of three char­ac­ters can fit up to six let­ters. Three let­ters are super­flu­ous. In the end, you need six char­ac­ters to define the seman­tic word of three let­ters. Of course, with­out knowl­edge of this lan­guage make it very dif­fi­cult even with a dic­tio­nary.
    If you are inter­est­ed, I am ready to send more detailed infor­ma­tion, includ­ing scans of pages show­ing the trans­lat­ed words.
    And most impor­tant. In the man­u­script there is infor­ma­tion about “the Holy Grail”.
    Niko­lai.

  • Giuseppe BIANCHI says:

    nor­mal intel­li­gence is enough if you use it:
    https://www.youtube.com/watch?v=yr9HM5Pi3wM

  • Валерий Дедух says:

    https://www.youtube.com/watch?v=y15cXoU_Tso The Voyn­ich man­u­script. The gold­en house of Nero. Watch my video.

  • Alan McGregor says:

    What we read via the media is like­ly to not be the whole infor­ma­tion avail­able.
    It’s sug­gest­ed that their was more than one ‘author’; the illus­tra­tions were be anoth­er at least. Hand­writ­ing ‘experts’ can say whether it was writ­ten by one, or maybe this per­son­’s per­son­al­i­ty or men­tal health etc.
    A per­son like Da Vinchi could be play­ing a ‘trick’, but also as myself, on the autis­tic spec­trum, it may just have been a per­son­al ‘cre­ation’, or amuse­ment.
    I read of inap­pro­pri­ate com­par­isons with present lan­guages, and the assump­tion that it is just one lan­guage and not a hybrid. It may be a per­son­al lan­guage or of a small group. With par­ents that use more than one lan­guage, it could be a nat­ur­al local hybrid.

    I find it strange though, that even as an arti­fi­cial lan­guage with quite repet­i­tive word forms, that there can’t be a cor­re­spon­dence between what is writ­ten and the illus­tra­tions.

    Per­haps there were men­tal health prob­lems involved. Why are words repeat­ed two or three times in suc­ces­sion?

  • M Lord says:

    Describes a real­i­ty in a dif­fer­ent time and place

Leave a Reply

Quantcast