Decode the Voynich Manuscript in Only Five Years…

(Here’s a guest post by Voynich researcher Marke Fincher that I’ve edited to Cipher Mysteries house style – I hope you enjoy it!)

Let’s say you want to do something really crazy, like decode the Voynich Manuscript. What… you do? Well, good luck to you! And to help on your way, here are some handy (but hard-earned) rules of thumb that might just help you achieve what nobody else has managed in five centuries…

(1) Don’t waste time on very well-trodden ground

• Don’t make assumptions. Don’t make assumptions about what it can or can’t be. One of the reasons the Voynich Manuscript remains unsolved despite all the brilliant minds that have worked on it is probably due to similar assumptions. Keep your mind open – there has to be a first for anything which is by definition unprecedented.
• Get more facts, as well as different types of facts. It is always a good use of time to go out and “get more facts”, by which I mean: get more detail, qualify (and quantify) what you know / expect / suspect / predict, study entirely new areas and learn more about what the VMs actually is.  If you try to link the different areas of your knowledge together, in due course better fitting hypotheses will emerge naturally.
• Strong disproof is more useful than weak proof. It is far easier to prove what the Voynich Manuscript isn’t than to prove what it is. Keep a list of “it can’t be X because of Y” statements,  but check through them periodically. Don’t spend ages studying what it “could be”,  just study what it is (and what it is not) i.e. don’t spend too much time trying to fit your pet hypothesis to the facts.
• The drawings aren’t just window dressing. Study the images as much as you study the transcriptions.  You should check and relate what you learn from the transcription back to the real thing as much as possible.  Check any suggestive text results manually against the images & vice versa.

(2) Appreciate the rational patterning underlying ‘Voynichese’

• Get a good working grasp of Voynichese. One good pragmatic way to achieve this is (a) to learn the EVA transcription system and (b) to transcribe an entire A-language page and an entire B-language page for yourself. Once you’ve done this, you should start to see that the countless patterns at play within the text are not (despite what some linguists like to claim) really like the ones you find in normal languages.
• Look for differences in system between pages. As far as these patterns go, some pages show similar rules at work, while others exhibit important differences – be warned, ignoring these differences between different system variants is undoubtedly a Very Bad Idea. Do your best to categorise the VMs pages into what you think are consistent sections (hint: these may not actually be on currently adjacent pages). Futhermore, it’s another Very Bad Idea to analyse large sets of pages without first satisfying yourself that the same system is at work throughout those pages.
• Determine which letters are usual and which are unusual. Take the time to compile a study of the unusual and rare symbols, especially the unusual gallows variants and “bench” symbols. Might these shapes hold clues to the function of the regular symbols of which they appear to be variants?

(3) Look at the patterns inside words

• Study symbol order preferences and rules within words.  There is more evidence of the nature of the system at work here than anywhere else.   It is one of the best places to start to get a feel for the thing.
• Study repeating sequences (but NOT in terms of whole words).  Introduce some flexibility/fuzzyness to your matching of sequences or you’ll not see it all (but not too much, or you’ll go crazy).
• Study the extent of “anagrams” within the vocabularies and contrast with known languages.
• Study the extent of “words within words”. Similarly, look at the hierachical nature of Voynichese word formation.

(4) Look at the patterns between words

• Study the relationships between words.  They are not like “real” words. Don’t just study vocabularies, word-lists and generative grammars in isolation.  Relate those structures and rules to what appears on actual pages (i.e. study them in context) and you will see a lot more of what is going on.
• Be wary of coincidental relationships.  Contrast the real patterns with coincidental ones by randomizing the word order and repeating the analysis.
• Adjacency relationships exist across spaces.  Study these and always be aware of them. Be extemely wary of spaces and assumptions of what they are there for.
• Study the ‘families’ of similar words. Do they just look the same or do they operate similarly as well? There are also vital things to learn here about variation in Voynichese.

(5) Look at the positions of words inside paragraphs, lines, and pages

• Study the position within the line and paragraph that ‘words’ appear.  There are plenty of unusual happenings here that any reasonably comprehensive explanation would need to cover.
• Repeat any analysis that you do on the VMs on real-world language samples. Contrast the results and try to understand the differences you find. Many “unusual properties” of the VMs turn out to be properties of known language as well!  🙂
• Pick some of the hardest properties to explain and target those first. Systems which match the easier properties of the VMs to recreate whilst ignoring the more perplexing and problematic properties aren’t really very helpful.  The more unusual the property, the more it probably has to say about the actual system at work.

(6) Transcriptions can hide problems, not just expose them

• Check your transcriptions. There are errors (of course) but also many situations where the transcribers were uncertain of what to record, or glossed over unusual or awkward details in the actual text.  There are also details not captured in any of the current transcriptions which could well turn out to be crucial.
• Repeat any analysis methods you use on multiple transcriptions. Compare and contrast the results.

(7) Be patient – this whole process takes an awfully long time!

• Once you get going, five years can pass by in a flash. So be warned!