Scientific Progress

[Torsten]

I agree that every argument should be judged by its own merit, regardless of the originator or the company it keeps. But I still think it is better, if at all possible, to approach the Voynich without forming a theory first. Usually, when someone thinks "maybe it's written in Old English", they will indeed find evidence that it is written in Old English. In the majority of cases, such approaches to the MS are not very productive.

Please don't attack the arguer instead of the argument. Believe it or not my approach was to research the Voynich without forming a theory first. Therefore the self citation theory is based on detailed observations of the Voynich text. I did already ask you, if you think my description of the Voynich text is wrong? Here I ask you again, where do you disagree with my description of the Voynich text and why?

Regarding your theory, it is not because you demonstrate that it could be done, that it was necessarily done that way. Theoretically possible (what you have shown) does not necessarily equal "likely". I believe that it is from this division between theoretical possibility and historical likelihood that our misunderstanding or miscommunication arises.

For example, one could probably devise a way for valid Roman numerals to be generated. But that does not mean that a medieval accounting book full of Roman numerals was generated by this method.

First, in my eyes scientific theories are not described as true or right, but as the best-supported explanation of the world based on evidence. Therefore a theory for the Voynich text should describe a large set of observations of that text, and provide a cohesive explanation for those observations.

Second, it is not theoretical, as the experiment by Gaskell and Bowern demonstrates. Just do the experiment yourself by writing a few pages of gibberish or ask some volunteers to do so. Then, consider whether it's easier to copy words or invent new ones. Isn't it plausible that people always try to do things in the easiest way possible? If it comes to historical likelihood, do you think that in medieval times students did behave different and never copied their homework? 

Third, if I understand you correctly, you are arguing that another unknown text generation method might exist which can result in exactly the same properties as the self-citation method. Moreover, you believe that this unknown text generation method is more likely than the known self-citation method. Can you name the additional criteria an alternative text generation method must fulfill? How do you evaluate an unknown text generation method? Do you suggest that we have to wait until someone publishes the "real" text generation? And what happens if at the day such a method is published someone argues that he believes that he believes that a third text generation might exist?