[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lojban] zipf computations & experimental cmavo
At 02:09 PM 9/25/01 -0400, Pierre Abbat wrote:
On Tuesday 25 September 2001 10:57, And Rosta wrote:
> There still remains the question of how to zipf down long lujvo whose
> syllablecount is excessive relative to their frequency, but Type IV fuhivla
> might provide a solution here.
I agree (e.g. noltrube'a -> noltube, not that I think we should actually
shorten noltrube'a). The hard part (or one of the hard parts) about zipfing
lujvo is getting a representative sample of usage. Most of us have occasion
to use {la'edi'u} in whatever we write about; not many of us will need
{glaukysedyterjo'e},
Which is precisely why the former is justifiably shorter. Zipf's laws
don't apply to individual use necessarily, but across all use in a
language. In some scholarly writings, it seems stylistically preferable to
use polysyllabic verbosity and complexity so as to project the appearance
of sophistication. That a technical word is used a lot in such cases
doesn't belie the fact that across all language use it doesn't amount to
more than a blip on the word frequency counts.
I've been talking in terms of requiring 100 usages/mentions in order to
include a lujvo in the dictionary. Most technical terms won't hit that
threshold even in a couple of articles.
but if we do need to talk about them, we will use the
word a lot in one article.
(owlhead joint: a device that can turn through at least a complete circle,
but not rotate indefinitely, used for mounting radio antennas and the like)
Cowan suggested binding it to broda series, which is what it is there
for. In addition, if it is only used in sumti (likely in this case), you
can use gy. or ko'a.
lojbab
--
lojbab lojbab@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban: http://www.lojban.org