Re : Root Structures

Dans un courrier dat=E9 du 17/09/99 05:14:52  , Ed a =E9crit :

> There are a few possible root structures for Proto-Indo-European roots
>  (variations on C-C or C-C-C with a vowel thrown in somewhere, or one
>  of the C's turned into a vowel).  I understand that virtually all
>  Semitic roots are triconsonantal.
> =20
>  What do roots look like in languages with overwhelmingly CV syllable
>  structure?  CV or CVCV?
> =20
CVC i think ;-)

ur-native japanese looks (C)VCV(CV)
but verbal roots are either (C)VC or (C)VCi/e
so you get :
kir-u / kir-a-nai / kir-i-masu etc.
mi-ru / mi-nai / mi-masu etc

another nice one is the khmer infix system :
tlay > t-um-lay
ket > k-omn-et

so pulling that rope you could have a conlang
with CVC roots that never appear as such :
let's imagine :
ket > k-omn-et / k-aps-et / k-irk-et
combining : k-omnapsirk-et
hehe.

or :

tam : to hold (physical)
t-o-am : to hold (social) =3D to run (business), to govern=20
t-i-am : to hold (psycho) =3D to behave
t-e-am : to hold (expression) =3D to have the attitude of
t-u-am : to hold (space, time) =3D to contain


>  Do people's conlangs have well defined possible root structures?
>=20
tunu is scientifically designed as V, VC, CV, VC,
CVC roots with class affixes. it happens that root+affix
makes all words CVCV.
but that's only a random side-effect.

for instance "tunu" is really root "-u-u" + prinfix "t-n-".
of course the meaning of the prinfix t-n- depends
on the root itself. below are listed various roots to which
t-n- prinfix may be associated :
1. big stuff
2. thin tax-return
3. banana shoot
4. airport lanes
5. rabbit and tuna fish

the affix mutates the root so it's difficult to
track roots and roots make affixes mutate so it's difficult
to tell affixes.

mathias