Discussion of "A possible generating algorithm of the Voynich manuscript"

[MarcoP]

the full repeats are assumed to be full repeats of words, and in that case they must already exist as repeats in the source text. They would not be introduced in the process I was describing.

Thank you, Rene!
I find it difficult to imagine that Voynichese can be based on word-to-word correspondence with a common European language. The pure frequency of reduplication is something I have never seen in European texts. Also, the words that frequently reduplicate are the most frequent words in the language (see graphs You are not allowed to view links. Register or Login to view.). The hypothetical English text should have several occurrences of each of "the the" "and and" "to to"....

Something I like in Torsten's approach is that he treats exact reduplication and quasi-reduplication as related phenomena: my impression is that some kind of relation must exist. 

Also, I think that in Voynichese similar words often tend to behave similarly and are somehow interchangeable.

For instance (using Takahashi's transcription) the sequence qokaiin okaiin occurs twice, as does its reverse okaiin qokaiin.
Both words appear a few times before otaiin:
okaiin otaiin 7 times
qokaiin otaiin 3 times

All these words also appear in reduplicated form:
qokaiin qokaiin 2 times
okaiin okaiin 4 times
otaiin otaiin 3 times

In English, consecutive words tend to belong to different parts-of-speech, e.g. an adjective and a noun, or a verb and a preposition: one wouldn't expect consecutive words to be interchangeable. If similar words systematically corresponded to consecutive words in an English text, I would expect them to systematically behave differently.