Historical ciphers, when they were introduced and their effect

[kckluge]

Conclusion: all known ciphers that have actually been used before the late 15th century make the plaintext even less like Voynichese than would be the case with simple substitution. Any proposed cipher methods that do produce Voynichese-like properties are therefore hypothetical.

Or at the very least anachronistic.

There is undoubtedly utility (if only for the benefit of relative newcomers) to going through how the Voynich text doesn't look like what you get if you apply known varieties of early 15th century ciphers to a natural language, but in some sense (if the text is an enciphered natural language) I think that's looking at the question backwards. If the text is an enciphered natural language, then the text has to be engaged with on its own terms, no matter how anachronistic the result of doing that may be. I think an approach more along the lines of William Friedman's "Synoptic Tables for the Solution of Ciphers" (You are not allowed to view links. Register or Login to view.) is more likely to be productive, i.e. how do the properties of the text constrain what type of cipher (if any) we could be looking at? At some point going down that road gets you to "not like anything we know was in use in the early 15th century (at least in Western Europe)" as a side effect of the analysis.