Re: KMC & syllables

["Gabriel Landini" <G.Landini@xxxxxxxxxx>]

On 27 Sep 2000, at 7:31, Rene Zandbergen wrote:

> Stolfi wrote
> > As far as I know, there is no letter that is
> > constrained to occur at most once in each Latin word (and yet occurs
> > on every other word!). 

The "space" character does. Do you think that some of the vms 
letters are in fact nulls and the visual spaces are irrelevant?
Would that keep Zipf's law?

> > I seem to recall that Sukhotin's algorithm applied to
> > Voynichese produced only a few unconvincing results that led
> > nowhere (probably echoes of the OKOKOKO model). Part of the
> > problem may have been the multiletter Voynichese->EVA encoding

> Jacques ran his tests using the Currier alphabet, IIRC, and while
> O, A and 9 were identified as vowels, the confidence levels of
> these identifications were lower than for Latin. 
 

I did this some time ago.


Language       characters   Vowels
English          21360    e,a,o,i,u,y
Latin		         22608    e,i,a,u,o,y

English + ?      21397    e,a,o,i,,u,y,?
dain daiin Latin 23726    i,e,d,o,f,y,h

EVA              32000+   o,c,a,y,n,e,s,g
CURVA            31582    o,y,a,e,Ee,p,g
GAVA             24508    o,y,a,e,Ee,g
FSG              26967    o,y,a,e,Z,g,u
Currier          26134    o,y,a,e,g,u
Frogguy          32000+   e,o,a,y,p,n,j,g

In the vms o, y, a, e and g seem to appear in the different 
alphabets, but one has to remember that the alphabets are not that 
different from each other. I may have the output tables somewhere 
at home.

> Here are a few other options.
> In various representations, numbers show this behaviour too.
> Roman numbers (MCLXI) have a strong positional behaviour.

Yes, but then if the letters were numbers, there are several patterns 
that do not appear even once in the whole vms. I did some pattern 
analysis trying to match Roman numerals abut 2 years ago. I can 
dig it up if anybody wants to look at that.

I seem to barely remember that a bar on top of the number is 
indicative of some factor (*10000 or so?).
There is 1 single instance of the character &173; which looks like 
eva y with a bar on top. I remember that during the Teddington 
meeting we discussed whether this was a numeral.


> > It is known that the Portuguese arrived in that region a few decades
> > before landing in Macao (~1510), a date that could take some strain
> > away from the chronology. Unfortunately I could not find any
> > information about those early contacts.
> 
> The silk route connection should not be ignored either.
> 
> >  (According to one source, the pleiades are called "sMen-du's" in
> >  Tibetan. Can we match that to EVA <doaro>?)

My favourite reading is "Touro" which is "bull" in Portuguese (the 
pleiades are in the Taurus constellation).

Cheers,

Gabriel