Theiling Online    Sitemap    Conlang Mailing List HQ    Attic   

Re: Concept_sitting

From:Sai Emrys <saizai@...>
Date:Friday, January 16, 2009, 3:23
If I may put a cogsci twist on the discussion...

It seems to me that words *cannot* ever be fully defined concepts. For
this reason, they cannot ever be atomic; there is always some further
division of meaning that can be made, as there is always further
definition that has not yet been elaborated and excised in previous
cuttings.

The reason is that all concepts, indeed all communication, depend on a
shared experience between the people talking.

Gödel, for example, proved that any (mathematical) system necessarily
has certain axioms that cannot be proved within that system. They must
simply be accepted, or not; if one does not accept them, then no
fruitful discussion can be had - they're not things one can argue to
be correct without going into a homunculus fallacy.

This is true of languages as well. Any "atomic" idea that one might
want are necessarily not truly atomic; calling them so is, at best, an
axiom that one may or may not share with others.

For this reason, any ontology of language - any list of semantic
primes - is at best a list of axioms. Someone else can always come to
that list and say, "I view this as actually a combination of things".

If this were not true, it wouldn't be easily possible to define the
word. Definitions are, as it were, a tweezing apart of the meaning in
the word.

So, IMHO, any ontology is doomed that does not acknowledge this, and
does not acknowledge that:
a) choices of how to divide or define a concept are necessarily arbitrary; and
b) choices of what concepts to adopt as 'atomic' are equally arbitrary.

Being arbitrary is OK. Conlangers do that all the time; eventually one
decides what goals one has, what one considers to be aesthetically
pleasing, how to balance choices that require tradeoffs. These things
cannot be really justified more than, at root, they feel right.

So yes, you can create a language based on semantic "primes". Indeed,
I think it's a useful idea; it gives rise to elegance like Arabic's
triconsontal semantic roots.

It simply will not ever be universal, and chasing universality -
chasing some sort of Truth of semantic primacy - will only lead one
into yet another form of qabbalah or OTO.

The world has enough of those, IMO.

- Sai

Replies

Erbrice <erbrice@...>
Mark J. Reed <markjreed@...>
Lars Finsen <lars.finsen@...>