Scientific Progress

[Torsten]

Torsten: There are all sorts of possible methods of text generation such as pick a symbol at random according to probability distribution based on its word position. Is that more plausible than your explanation?

The burden of proof is on the side claiming that there is another text generation method, that is even more plausible. So maybe you should publish your "pick a symbol at random" text generation method along with some generated sample texts. Then it would be possible to check if a medieval scribe could have used such a method and if your method is able to reproduce features like the "binomial-like" word length distributions for word types as well as for word tokens, the shift from Currier A to Currier B, both of Zipfs-Law etc.
 

Fundamentally, your method it is too vague. If you could produce an algorithm which would exactly generate the Voynich text that would be something else.

The core instruction is simply to repeat parts of previously written word tokens with various changes and additions. The experiment by Gaskell and Bowern demonstrates that it is even sufficient to instruct volunteers to "write some 100 words with fake, meaningless text in a language that you make up as you go" [Gaskell & Bowern 2022]. Some volunteers will likely conclude that it is easier to repeat previously written words with various changes and additions than to invent new words. More detailed instructions are not necessary. The key point of the method is that someone who repeatedly copies words over and over again will develop their own style and habits in the process. Try it yourself: write some pages full of language-like gibberish or ask some volunteers to do so.

The whole point of writing gibberish is that it doesn't matter what you write. Therefore, if someone could present an algorithm that exactly reproduces the Voynich text, this would actually disprove my hypothesis.

I think the implications that the text is meaningless raises all sorts of questions with respect to the plausibility of that conclusion.

It only makes sense to discuss the implications of the self-citation method if you accept the idea that this method was used for writing the Voynich text. Therefore, it surprises me when people argue against a conclusion of a theory they claim they haven't considered.