Re: VMs: Is it a human language? (was Moot points, getting long)

[Koontz John E <John.Koontz@xxxxxxxxxxxx>]

On Fri, 6 Aug 2004, Philip Neal wrote:
> Interesting. I take it that you are not saying that 'Voynichese' is a North
> American language:

Right, highly unlikely even if it's a "recent" forgery.  I just happen to
be better at fishing up examples from (some) such languages.

> you are saying that any spoken language has severe restrictions on the
> possible combinations of sounds and that this might be the reason for
> the apparent restrictions on the co-occurrence of 'Voynichese'
> characters.

More or less.  I wouldn't be quite so definite.  At least some languages
have severe restrictions of this character, and that could lead to
restrictions on co-occurrence, providing the text is not an encryption
designed to level such restrictions in the interests of obscurity.

Note that my example was for a language commonly transcribed ina more or
less phonemic orthography using more or less one letter per unique sound
(excluding use of diacritics, including things like following h to
represent aspiration of stops).

Gabriel Landini's example of qu plus certain vowels in Spanish is a bit
more complex.  It is an orthographic artifact.  Latin contrasts c + all
vowels, in which c represents a k sound, with qu + all vowels (but u?), in
which qu represents the cluster kw.  Spanish changes k (spelled c) in ce
and ci to theta (s in Andalusian and American dialects) (still spelled ce
and ci) and leaves it a k-sound in ca, co, and cu (still spelled ca, co,
and cu).  Spanish changes kw (spelled qu) to k in que and qui (still
spelled que and qui) and quo (now spelled co) and leaves it kw (spelled
cu) in qua (now spelled cua).

Latin                   Spanish (American)
Spelling Pronunciation  Spelling  Pronunciation

ca       ka             ca        ka
ce       ke             ce        se
ci       ki             ci        si
co       ko             co        ko
cu       ku             cu        ku

qua      kwa            cua       kwa
que      kwe            que       ke
qui      kwi            qui       ki
quo      kwo            co        ko

Or, looking only at Spanish

Spanish
Pronunciation Spelling

ka            ca
ke            que
ki            qui
ko            co
ku            cu

sa            sa, za
se            se, ce
si            si, ci
so            so, zo
su            su, zu (?)

Note that the c and z spelings in the second collumn of the second group
have theta, not s, in, e.g., Castillian Spanish.  I can't think of a case
of zu at the moment, but my Spanish is pretty limited.

It's worth going into an example like this because, if Voynichese is a
Romance language, one has to determine whether it represents (and/or
encrypts) the standard orthographic practice for the language or the
phonological reality (pronunciation) of the language, or some compromise
between these.  Of course, both Spanish orthography and Spanish spelling
have changed since, say 1400, and the same can be said of essentially any
European language.

> It is logically and historically possible. The theory has been floated on
> this list, notably by Jorge Stolfi, that 'Voynichese' might be Chinese. Not
> the Standard Northern Chinese of today as transcribed in books about modern
> Chinese but some different, older form of the language written in alphabetic
> script.

My reactions to this are, one, it's a very good idea to keep an open mind
about what is being reprented, and, two, if the writing system does use
systematic series of graphs to render the syllables of the language, then
a distributional study of these graphs (or some set of polygraph glyphs
deduced from them) might very well suggest series of series
syllables, and Chinese will come to the mind immediately.

> Chinese has a phonetic personality which comes out in transcription:
> so do Turkish, Arabic and Japanese.

All languages, or at least all regional assemblies of related languages,
have phonetic and grammatical personalities.

> I do not know these languages but it is possible to know many facts
> about a language which you do not understand.

As a linguist I make a principle of this.

> The question is this. Memorise any transcription and pronunciation of
> 'Voynichese' and read a page of the manuscript to yourself. Is it like
> memorising Arabic script or Japanese kana and trying to read a language
> which you do not know? Alternatively look at a descriptive grammar of some
> endangered language, written after years of hard work by a field linguist,
> and ask yourself if the structure of 'Voynichese' is really so complex.

I'll have to reserve judgement on this.  Voynichese is undoubtedly weird.
A sequence like Rugg's example of "qokedy qokedy dal qokedy qokedy"
doesn't sound very natural, however it might actually be pronounced.  On
the other hand, neither does "Come-a come-a down dooby doo down down," or
however that's spelled, and it's been pointed out that we can't tell the
difference between words and numbers, on the assumption that the text
contains both.
______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list