Re: TECH: se, te, & lujvo

[Logical Language Group <lojbab@GREBYN.COM>]

You cannot assume that seljerna is semantically identical to se jerna, as was
pointed out.  You have given an example of where the English tanru may be
taken more broadly than the English lujvo, even though >I< would never guess
at the tanru interpretation without a lot of supporting context.  Lojban
allows any lujvo to take a more restricted meaning, and seljerna could
similarly take a more restricted meaning; it >need< not but it >may<.  There
is no equivalent mechnaism for the unconverted place, so if you want to
restrict jerna, you must to so by providing some modifier that indicates
what kind of restriction is intended.  Zipf's Law, which Lojbanists may seem
to worship, requires us to allow seljerna to be used in lieu of a longer
lujvo that is more explicit as to the restriction, because we allow seljerna
as a lujvo in the first place, and there is no justification in having the
longer form when the short form exists, UNLESS it could carry a somewhat
different meaning.  We are required to have "sel-" forms in lujvo, because
a three-part tanru-based lujvo may depend on using the "se" conversion, and
could not be made into a lujvo without a rafsi for "se".  Ina ddition,
as was also pointed out, in some cases, the two-part lujvo based on "sel-"
will be shorter than the tanru, and in this case people will tend, because
of Zipf 9so we argue) to use the shorter form.

lojbab