Re: WG: average word length in VMS

[Jorge Stolfi <stolfi@xxxxxxxxxxxxx>]

    > [stolfi:] But even if we were to use the true Voynichese
    > alphabet, the KMC strucure would probably prevent [Sukhotin's]
    > algorithm from finding enough CV and VC transitions to call home
    > about.

    > [Jacques:] Not sure. I once tried the algorithm on a word list
    > of a language of Vanuatu which was 90% consonants -- but had 10
    > vowels. Every vowel identified, none misidentified.

What I meant is that the KMC model allows at most one CV and one VC
transition per "word", for any plausible C-V partition of the
alphabet; so all partitions would look equally wrong to the algorithm
(and to us). Except for the partition V = {a o y e} --- which probably
gives enough transitions to satisfy Sukhotin, but doesn't seem to be
right for other reasons.

    > If gallows are tonal marks they won't occur alone. Ditto if they
    > mark vowel/consonant length or/and aspiration. Ever  seen a
    > spiritus asper occur alone in Greek, and a shaddah without
    > a letter underneath in Arabic? (Not to mention a Portuguese
    > tilde without an a, an o, or an e, but not an n!)

Right...

    > [stolfi:] Curiously, the tones [of Tibetan] are denoted
    > (inconsistently) by prefixing certain dummy consonants to the
    > syllable, like b and r in "'byung rtsis" d ("astronomy").

    > [Jacques:] No, the tone is determined by the final consonant, which
    > is most often lost in pronunciation.

Oops, thanks for the correction.

    > [stolfi:] the pleiades are called "sMen-du's" in Tibetan.
    > Can we match that to EVA <doaro>?
    
    > [Rene:] But doaro (or doary) is not a _very_ typical Voynichese
    > word and fits the KMC model only with a stretch (unless I am
    > mistaken, please correct). Thus, it could be a 'foreign' word in
    > whatever language the MS is mainly written in.

Indeed, because of the <oa> group (two consecutive circles) which is 
very rare. 

All the best,

--stolfi