The incompatibility of Voynichese with natural human language

[bi3mw]

On the original topic of this thread, I can only pose this rhetorical question: is the Voynich text compatible with a medieval cipher?

This is a good (rhetorical) question. As far as I know, before Alberti there were only rudimentary approaches ( simple substitution ) and even an Alberti-like code would probably have been decoded already. So one can only assume that, if there is an encryption, it is based on a completely unknown,  unique system. This would not surprise me. So the question is what non-anachronistic method(s) would be possible.

[Aga Tentakulus]

The nice thing about mating signs, I can put them together however I want.
For the writer, it doesn't really matter as long as the order is correct. For the reader it looks different optically, but it is actually the same.
Simple technique and used in the Middle Ages. ( combination )

[Torsten]

The text in the VMs behaves unusual on character and also on 'word' level. On character level it is possible to describe certain rules for the usage of the glyphs. Some of the rules apply for the whole manuscript whereas other rules only apply for some parts of the manuscript. Sometimes a rule only applies for a certain page, paragraph or 'word'. On the other hand, there are no similar rules on the 'word' level. Every 'word' can follow every other 'word'. 

For a text using a natural language grammatical rules on word level and only a few rules on letter level are expected. Words are used in the context of other words and if a word is modified this usually happens in a consistent way (in a  alphabetical or syllabic writing system for instances by changing prefixes or suffixes). The rules for using the words are the same for the whole text.

In summary a text in natural language is using words to form grammatically correct sentences whereas the text in the VMs is using glyphs to form groups that only look like 'words'.

[Aga Tentakulus]

@Torsten
Maybe you're right, sometimes it does seem that way.
Based on my work, and the corresponding results, VM is based on a natural language.
But it seems that the place of origin also plays a role. ( adding romanic dialects, which makes it difficult or impossible to work without reference material ).

[Koen G]

As I see it, there's a level playing field.
Yes, unaltered language is statistically improbable.
But complex ciphers and algorithmic nonsense generation are anachronistic, even at more modest scales. As far as we know, it just didn't happen.

So far, we know of no linguistic text which behaves like the VM. But I don't see why any alternative with just as little parallels in reality should be preferred.

[Mark Knowles]

As I see it, there's a level playing field.
Yes, unaltered language is statistically improbable.
But complex ciphers and algorithmic nonsense generation are anachronistic, even at more modest scales. As far as we know, it just didn't happen.

So far, we know of no linguistic text which behaves like the VM. But I don't see why any alternative with just as little parallels in reality should be preferred.

Some people say the Voynich could not be in cipher as we can break all the ciphers of the time. However there were ciphers at that time and major cipher advances throughout the 15th century, it was quite a dynamic time in cipher development, so the idea that a Voynich cipher could not be produced during that period is, I think, false. In fact I could produce a cipher with a few small modifications to early 15th century cipher techniques that would be very difficult to crack; just throw in a sizeable amount of null/filler/cover text and a large table of glyph pair homophone lookups and you have something very hard to break. The author would not need to be a genius to produce such a cipher and I believe such a cipher could display the statistical and other phenomena that others point to and that I am pretty familiar with.

I tend to view it as though linguists like to view Voynichese as a language and cipher people as a cipher, it could be argued that we all come with our own biases.

[Mark Knowles]

I think one problem that people, who claim it is all nonsense, have is that you can't prove it is nonsense. For example, it could be 99.9% nonsense and 0.1% meaning and that is fundamentally difficult if not impossible to prove. More realistically it could be 80% filler nonsense text and 20% real text, which is still difficult to disprove.

---

I find the idea that it is an artificial language a catch all which for that reason is difficult to disprove. An artificial language in its broadest sense it seems to me could have almost any properties that one can imagine. In fact a cipher and a straight natural language could arguably be a subset of that.

[Alin_J]

However, if one just made the though experiment that e , ee , eee are equivalent, then the number or repeated sequences will increase.

Yes, a bit, but only for shorter sequences up to 4 words in length. Sequences of 5 words or longer will still be absent (if you don't count the one-letter "word" sequences in the circular diagrams, f57v).

[elieD]

I talk about this hypothesis in my paper: You are not allowed to view links. Register or Login to view.
Basically, the VMS is an outlier, but other documents in a lot of other languages are also outliers for good reasons. It's interesting to know why these documents are outliers: mainly they are (sometimes romanized) retranscriptions of oral languages

[-JKP-]

Here's the Wiki on the preprint server:

You are not allowed to view links. Register or Login to view.