Re: Yitzik and Basilius met in Kiev

Herman Miller wrote:

> Isaac Penzev wrote:
>
> > I would appreciate if you ppl can share your ideas wrt creating a_priori
> > vocabulary without using software (that I usually feel inadequate to
> > satisfy
> > my creativity).
>
> I had some success in the early days of Tirelat with my "eight words a
> day" project -- think of eight related ideas (names of trees, vehicles,
> parts of the body, or some such thing) and assign names to them.
I guess I'm a little more obsessive :-))) Once I'd used the old Langmaker
program to generate several thousand possible Kash forms (it helped that the
phonotactics were pretty simple), I'd do 25-30 words at a time. My
experience devising a wordlist for research in Indonesia helped-- it was
categorized by Numbers/Quantifiers--the Household--Body part--Kin
terms--Nature---then mixed lists of Verbs--Adjectives and finally
Pronouns/Sentential adverbs. Having that in the back of my mind really
helped; also the fact that the derivation systems of Kash and Indonesian are
rather similar. That wordlist BTW is available online--
http://cinduworld.tripod.com/wordlist.txt (thanks to BP Jonsson for
html-izing it!!) and might be helpful to others. It happens also to contain
all the words of the 200-word Swadesh list, though they aren't marked (they
were in the original).

(In the 20 yrs or so before I got computerized, I'd created only a handful
of words with pen and paper; most of them were discarded. ;-(((
>
> Another thing if you have lots of derivational morphology would be to go
> through from time to time and look for new derivations that might be
> useful -- pick one affix and try it with lots of different roots.
That's de rigeur with Kash; any new verbal/adj. base immediately leads to
multiple derivations, reduplications etc.

OTOH vocab.creation in Gwr is a slow and haphazard process since it's
monosyllabic and isolating (with minimal fossilized derivation that helps at
least to form verb/noun alternants). In this case I created by hand a list
of ALL the protoforms (again simple phonotactics helped, but still, there
are 30,000+); I pick a handful of them, run the sound-changes and assign
what seems a suitable meaning.

Reply