Re: HELP: Vocabulary

Trebor Jung wrote at 2004-07-18 13:28:43 (-0400)

 >
 > But anyway, back to my problem. When I try coming up with
 > compounds, they all seem to be imprecise; when I try to make the
 > compound more precise, the result is something like 'to
 > continuously attempt to crush with the teeth' for 'to chew'. That
 > is far too long... Or maybe I need to add some grammatical
 > categories to my language?
 >
 > So, does anyone have/know of a scientifically sound list of root
 > concepts for a human language (one designed by trained linguists, I
 > mean)? Anna Wierzbicka's Natural Semantic Metalanguage comes to
 > mind, but the list
 > (<http://www.une.edu.au/arts/LCL/disciplines/linguistics/nsmpage1.htm#model>
 > ) doesn't seem sufficient to cover all the concepts I want Kel to
 > be able to express. Or maybe it is - it was created by a
 > linguist... Or maybe I should just wait till I have a linguistics
 > degree before I attempt something like this... Or maybe I should
 > just learn Ithkuil...
 >

I don't think you'll find any list that will satisfy your
requirements.  I don't think there's any complete consensus among
linguists that the concepts on Wierzbicka's list are genuinely
universal semantic primes, and the problem only grows if you try to
extend the list.

Anyway, if you want the meaning of every complex expression in the
language to be _precisely_ determined by its composition, I think
you're likely to have some problems.  Particularly if you limit your
lexicon to basic, universal conecpts.  There are a lot of things that
any reasonable language should be able to express, aren't really basic
concepts, and can't be precisely defined.  Take "dog" for example.
Clearly it's not really any kind of universal basic concept, but it's
very difficult to give a precise definition that covers the set of all
possible dogs and no possible non-dogs.  Certainly not one that you
can condense into a serviceable word.

So, I would suggest that if you want your language to be practically
usable at all, you may have to include some roots that aren't basic in
any fundamental sense, and derivational processes that don't
_precisely_ define the semantics of the resulting form.

Still, you can obviously get a lot closer to your ideal than
English... in selecting basic concepts the ULD is a useful start, as
would be the Lojban gismu list.  I'd suggest you work through these,
classifying concepts according to the kind of distinctions youu're
thinking of making, and altering or discarding anything you think
should be a derived form rather than a root.  Oh, you might take a
look at Jeffrey Henning's Dublex, which was (as I understood it) an
experiment in finding out which roots would be most useful for
compounding.

Incidentally, 'to continuously attempt to crush with the teeth'
doesn't sound totally unworkable to me - I can certainly imagine
certain American languages expressing such a thing with a verbal root
"crush", with an instrumental affix meaning "with the teeth", and
marked for repetitive aspect.

Reply