Re: VMs: RE: wet blacnket?/ (was that "blanket"?)

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

On Thu, 26 Aug 2004, Jacques Guy wrote:
> >For example, "Japanese" or "Hawaiian" words could be generated from a table
> >of valid syllables, or "Chinese" words could be generated from a table of
> >onsets and rhymes.
>
> Or syllables. There are only 400 possible syllables in Mandarin.

Syllable = onset + rhyme, so this is saying the same thing, really.  I
think the terminology is just different.  Of course, it's my understanding
that most working lexical entities ("words" in one sense) in Mandarin are
compounds.  Because Chinese writing represents the basic constituent
elements of these compounds, and many basic words are uncompounded
monosyllabic roots, there is a tendency even by educated native speakers
to think in terms of one sign = one word = one syllable and ignore the
compounds or treat them as phrases, which is at best an
orthographically-based simplification as I understand it.

A lot of languages around the world are more or less analyzable into
collections of monosyllabic or bisyllabic elements - roots and affixes and
enclitics - with strong limits on their shape or canonical form.  That is
to say, the "words" of these languages are decomposable in this way.  It
may require a certain amount of specialized knowledge to do the
decomposition, depending on how productive the derivational processes
involved are.  An unproductive process is one that a contemporary speaker
can't (normally) use to make new forms, e.g., -kin suffixation or be-
prefixation or shilly-shally reduplication in English.

A language with a lot of vocabulary using non-productive derivational
processes may not be analyzable by the average speaker, even a
well-educated one.  English tends to fall into this category because it
has an enormous amount of borrowed vocabulary from a variety of sources
and the derivational processes in this borrowed vocabulary are mostly
moribund in English.  So, English has a lot of long words with opaque
innards.  I think this is actually fairly atypical.

> >If you tuned your algorithm well enough a high
> >percentage of the words you generate would be real words in your target
> >language.
>
> In fact, the best illustration of all this is Lewis Carroll's
> "Jabberwock".

Lewis Carroll was a highly educated individual with a delight in wordplay
and was able to tease out the various strands in English vocabulary and
manipulate the derivational processes in each.  He also had a significant
advantage in being able to use actual English functional words freely.
In the case of the VMs we aren't dealing with, say, Latin function words
and inflections attached to nonsensical roots.  So, if the VMs is intended
to be pseudo-language the constructor was facing a problem of a different
order or sort than Carroll was.

> > Rugg has spent some time tuning his algorithm
>
> Lewis Carroll tuned his "algorithm" to perfection. The result
> looks and sounds like credible English. It follows, in
> Rugg's line of argument, that Shakespeare's sonnets are
> meaningless gibberish, a hoax, a slithy tove.

I think there may be some fallacies in that analogy.  Native speakers have
different reactions to the sonnets and the Hunting of the Snark.  And,
unfortunately, we have nothing comparable to native speakers on which to
test Voynichese.  Aren't we dealing with apples and oranges here?  Rugg
would be demonstrating that the sonnets were gibberish only if we were
able to put forth a Voynichese text that was known not to be gibberish.

> I cannot understand, for the life of me, why that grotesque claim was
> not laughed out of court. It appears that it is not only that Rugg
> cannot string two coherent thoughts together, but that the editors of
> Cryptologia and Scientific American cannot tell incoherent gibberish,
> which that is, when they see it.

I don't believe Rugg is right.  He admits he hasn't worked out all the
kinks, as I recall, and so is only making a suggestive argument.  It
doesn't seem to me to be particularly incoherent.  So, I'm not sure I'd go
so far as to say he was putting forward a grotesque claim.  One could
certainly write a more or less comparable article arguing the reverse.
But pending a solution we can all believe in, it's a bit hard to reject
any approach out of hand.

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list