Karen Stein <
comca...@gmail.com> writes:
> Adding to the end of a gismu is much less of a change than changing it in
> the middle. When we were originally working on place structures for the
> gismu long ago we knew there would be some that needed to be expanded (and
> shortened).
I was going to point this out... that adding places at the end of a
gismu's existing tersu'iste causes less semantic impact/damage/confusion
than re-ordering or changing the meaning of places with existing
definitions. That's because seeing an x3 for a gismu which only has 2
defined places automatically implies that a new, different, definition
is being used.
But there is another way to signal that a gismu has new meaning: change
the final vowel of the gismu. So, instead of re-defining the places of
{kelci}, create a new gismu, {kelco}, and declare {kelci} deprecated.
That way, existing uses of {kelci} would retain their meaning. Anyone
seeing {kelco} would immediately know that either (1) the writer made a
typo, or (2) some new variant of {kelci} is being used, and would
immediately recognize it as a derivative of {kelci}. Because only the
final vowel is changed, minimal damage would be done to existing {kelci}
lujvo. (The places of a lujvo don't necessarily match those of the
tertau.) Of course, any new lujvo would have to stop using using the
CVV-form rafsi {kei}, because the final vowel would no longer be {i}.
But that seems manageable.