previous research on Arabic and the Voynich ms?

[geoffreycaveney]

If we can assume it is written left to right, then it is safe to assume we have to read it in the same direction.

Although l completely understand this argument and would’ve agreed with it without other statistics, l think it is plausible that some sort of rearrangement of the glyphs within words may be necessary to get the rigid structure exhibited.  With those kinds of steps being seriously considered, putting an “al” at the end of a word rather than the beginning seems to be a pretty mild step.

Of course this messes with the VCI approach - but independently, Arabic words with “al”s placed at the end of the word doesn’t seem out of the realm of possibility and certainly can be accommodated.

I absolutely agree that we should be open to considering all possibilities. With that said, we have to be careful about what type of limitations we may place on the concept of the rearrangement of the glyphs within words. Completely random rearrangement of glyphs within words is the same as the concept of an anagram. But anagrams are notorious for their ability to rearrange almost anything into almost anything else. There would have to be some strict and logical rules about such rearrangement of glyphs. 

I wouldn't say that the consideration of any of these possibilities messes with the VCI approach. Strictly speaking, VCI is a transcription system, nothing less, nothing more. Yes, it was based on some hypotheses I had about logically plausible letter values for particular glyphs and multi-glyph sequences. But no such hypothesis at this stage is likely to be correct about all such details or even most such details. Its main value is to present a visual representation of the ms text that is more in line with the intuition behind Koen's verbose cipher analysis. It can serve that purpose regardless of how accurate or inaccurate its guesses about actual letter values prove to be. The same would apply to the rearrangement of glyphs within words, however arranged and organized. 

I should take this opportunity to note that it may seem strange that a transcription system based on a hypothesis of a substantial series of palatalized consonants could possibly have a connection with Arabic, which has no such series and only a couple palatal consonants. One possible explanation is that the distinction that had been tentatively identified as plain vs. palatalized may possibly rather instead represent a distinction of plain vs. "emphatic", which is indeed persistent throughout the Arabic consonant phoneme system, in particular in alveolars, sibilants, liquids, and gutturals. (For the latter phonemes, the famous Arabic uvular "q" can be considered the emphatic counterpart of velar "k".) Naturally this will require some adjustments where the VCI values will not correspond precisely with the Arabic values, if that is indeed what they represent. But again, the VCI transcription system can continue to be used as such, regardless of the actual phonemic values that the VCI values may or may not represent.