The Voynich manuscript is dated to the early 15th century.

The Voynich manuscript (MS)

It is a medieval handwritten book of almost 250 pages, and even today the text cannot be understood. It has become quite famous, and it is recognised as one of the main unsolved problems in the history of cryptography.

Can we tell whether it is written in a natural or constructed language?

  • 2
    There are other possibilities: It can be a cyphertext, or it can be no language at all (being an elaborate hoax). Feb 8, 2018 at 17:24

3 Answers 3


There's no evidence to say that it is a constructed language.

The idea was investigated in detail by John Tiltman, a British army officer who specialized in cryptography. The possibility of the Voynich manuscript being a constructed language was brought to his attention by William Friedman, an expert on the manuscript. Friedman confided to Tiltman, after decades of work on it, that he thought the writing could be a form of "synthetic universal language", and so the cryptographer expanded his investigation in that direction, writing a paper on the theory.

The basic premise behind a "synthetic universal language" is to use characters to represent certain logical ideas; by chaining together letters, you can create "words" that describe more complicated principles. By adding the appropriate prefixes and suffixes, characters can be quite powerful. Tiltman noticed that some of the patterns in the words fit this type of design. However, the evidence for this sort of construction was not strong enough, and Tiltman's analysis of the frequency of the occurrence of certain letters did not support Friedman's hypothesis.

Tiltman also traced the historical origin of a "universal language", to determine if such an idea had been bounced around early enough. However, this proved difficult, and Tiltman found no concrete examples prior to the mid-17th century. This is two centuries after the Voynich manuscript was written - sometime in the 15th century - and so if the manuscript was indeed written in a universal language, it would predate these universal languages by two centuries.

Tiltman noted

It was clear that the productions of these two men [Wilkins and Dalgarno, two early proponents of universal languages] were much too systematic, and anything of the kind would have been almost instantly recognisable. My analysis seemed to me to reveal a cumbersome mixture of different kinds of substitution. When I was attempting to trace back the idea of universal language, I came upon a printed book entitled The Universal Character by Cave Beck, London 1657 (also printed in French in the same year). Cave Beck was one of the original members of the British Royal Society and his system was certainly a cumbersome mixture.

Beck's writing claims that the idea of a universal language goes back to the 16th century, but this still places it many years after the manuscript's origin. The idea of synthetic universal languages, though perhaps relatively widespread in the 17th century, simply was unknown in the 15th. This was a strike against Friedman's theory. If the Voynich manuscript was a synthetic universal language, it would have borne no relation - in time of conception of structure - to later examples.

The universal language route is of course not the only possibility; constructed languages can take many forms. However, it was possible the most promising option, and research into it went nowhere. Friedman's claim remains unsupported, and, to my knowledge, no further academic analyses have been done. We certainly can't rule out the theory that the Voynich manuscript is written in a constructed language, but we have no reason to believe that it is.


We don’t know. It is written in an unknown script, by an unknown author. We don’t even know whether it is even in any language at all, or just random scribbles. Any attempts to connect it to a known language (such as Latin) have failed thus far.

  • 6
    You're absolutely right, but this answer would benefit from some citations or links to further reading, just so that it can be easily verified. Feb 6, 2018 at 22:40

Hope springs eternal. This article points to one recent lead in computational linguistics. original article here

I've read that some work has been done on the script, indicating that there are patterns, even if we don't know what they are referring to.

I have a copy of V.M. (a book èvery glossopoet should have!). I'm frankly of the opinion that, considering the size of the book and the cost of the materials in it and the cost in time to produce the work, it can't just be 240some pages of random nonsense. It's too expensive and would be an incredible, inconceivable waste of resources if it were just a meaningless & private toy of some overly rich maker.

Each page of vellum requires a relatively labour intensive process to make. Inks have to be made. Each page has to be hand written and hand drawn. Some pages are really large fold-outs. Almost every page is decorated in some way. That's a lot of work for it be nothing at all but a pretty conversation piece!

  • 1
    This doesn't really address the question of whether the manuscript is written in a constructed language or a natural language.
    – HDE 226868
    Feb 7, 2018 at 17:58
  • The linked article indicates that it's Hebrew. Clarified links.
    – elemtilas
    Feb 7, 2018 at 19:56
  • 1
    In fact, the link indicates that the manuscript is written in an anagram cypher that they claim to have decrypted into Hebrew. But they only give some short text bites, no longer passages yet. Feb 8, 2018 at 17:19
  • 1
    BTW, I have some experience with the kind of statistical methods they use, and those methods tend to produce seemingly good text from utter gibberish, e.g., doing OCR correction of French texts into English. Feb 8, 2018 at 17:44
  • 1
    Interesting source, but the article itself notes that "according to a native speaker of the language, this is not quite a coherent sentence." It really says something more like and the priest made for her man to him to his house and upon me his men the commandments (which is unsalvageable gibberish). The methodology is also flawed since it relies on a "hypothetical non-standard orthography of Hebrew" which has never existed.
    – b a
    Apr 4, 2018 at 12:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.