Re: The philosophical language fallacy (was ...)

Hallo!

On Sat, 5 Jul 2008 16:31:06 +0200, Benct Philip Jonsson wrote:

> IMHO there is a distinction to be made between
> using a taxonomy to decide which meanings to
> provide words for and using a taxonomic morphology
> if the Wilkinsian kind.
Yes.  Actually, I use a general taxonomy of ideas to organize
the vocabulary of Old Albic.  The words are not derived in
form from the taxonomy - they are as arbitrary as in any
natlang - it is just that I set up several classes which
I fill with dictionary entries.  A great way to get a well-
rounded vocabulary; you instantly see which fields of
discourse need more work.  (Actually, the idea is not mine
but Mark Rosenfelder's:

http://www.zompist.com/thematic.htm

I also based the taxonomy I use on his, though I have modified
it in places.)

>       The latter clearly leads
> to the problem you describe, especially if the sub-
> distinguishing morphemes are only one phoneme
> long, which opens a whole other can of worms, viz.
> whether short morphemes are a virtue in themselves
> and how much redundancy is desirable (once one has
> decided that redundancy is desirable in the first
> place, which the likes of Wilkins seem not to have
> understood in the first place).
Yes.  Wilkins did not realize that redundancy in languages
is not a bug but a feature, and wanted to design a language
which avoids all "unnecessary baggage" of natlangs.
The result is an awfully nonredundant language.  This is
a problem which befouls not only Wilkins-style taxonomic
languages but "speedtalk"-style systems in general.

>        One strategy which
> might help would be to use morphemes which are
> meaningful in themselves for the sub-distinctions,
> a bit like is done in Chinese compounds (e.g. "pomacea-
> sweet" vs. "pomacea-sour" for 'pear' vs. 'apple'.
> That there are apples which are very sweet is
> beyond the point; the typical apple is sourer than
> the typical pear).
Yes.  Compounding is a very useful and also natural device.
However, if you use a closed set of roots, your compounds
soon get pretty long and clumsy.  Any taxonomic language
likely will need compounding beyond the taxonomic derivation
of the lexemes, I guess.

On Sat, 5 Jul 2008 12:57:48 -0400, Herman Miller wrote:

> Well, proper names aren't usually translated anyway,
I have seen at least one taxonomic language scheme that derived
place names from geographic coordinates!

>        but that could be
> an issue if the standard name for something is based on a proper name,
> like "hamburger" from Hamburg, or "watt" from James Watt. Some of these
> sorts of names could be fit into an oligosynthetic scheme, I guess. You
> could have "single reed conical metal wind instrument" for "saxophone"
> if naming it after Adolphe Sax doesn't fit well into your system.
Your 'saxophone' example illustrates the problem very well.
The compounds get very long and clumsy.  Your compound consists
of (at least) six morphemes, and in English, it is 12 syllables
long.  (In a speetalk-type language, this is not a problem though,
as the whole shebang will be just six phonemes.  As long as you
can pronounce it, of course.)

I have mentioned chemical nomenclature earlier in this thread.
The "rational nomenclature" of chemistry works quite much like
a taxonomic language: the "rational" names are build from
elements which represent the atoms and atom groups the compounds
are made of, and are compounded in such a way that you can
derive the structure of the molecule from the name.  However,
only with rather simple compounds, such rational names are
actually used - because those names quickly become very long
and unwieldy.

... brought to you by the Weeping Elf

Reply