[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lojban] Re: zo xruti xruti
At 10:01 PM 8/17/02 +0200, Adam Raizen wrote:
'sezyxru' (and 'vo'a zei xruti' and any other lujvo I can think of
other than 'zilpavyxru') often do not have the necessary meaning. One
often wants to say that something reverted to a former state,
fatne binxo? (fatybi'o)
Originally binxo/galfi cenba/stika
were intended to be the generic words for change, both agentive and
non-agentive. Regularization of place structures and sumti-raising made
the x1 of galfi and stika an event, allowing the use of gansu or zukte
for agentive and purposeful agentive action to modify those. fatne then
was to be used to indicate the reverse of any of these changes (or any
other reversal of process not easily indicated by na'e/to'e). xruti was an
overlapping word for what now is fatne ke galfi (gasnu) or fatne ke stika
(gasnu) because the latter were too long for their frequency per Zipf.
If you don't
think of lujvo components as anything more than mnemonics, then maybe
that is not a problem, but for me it is completely against the spirit
of lojban.
Your feeling, shared by many, seems to have been a later development in
Lojban (though perhaps old to recent arrivals since it emerged around
1993-4), since mnemonicity was the classic purpose (But even that was not
original. It was the 1982 GMR revision to Loglan that even gave uniquely
decomposable mnemonics, and TLI Loglan still uses madzo (lojban zbasu) for
all agentive causality patterned apparently after the English or Latinate
"make". Before that, lujvo were recognizable from their components the way
gismu are recognizable from their etymologies.)
Thus I can see it as being a dispreferred choice to go away from dikyjvo,
but it is hard to call the language "broken" when this isn't easy or
requires a longer word form than we would prefer/Zipf would suggest. (Part
of the reason that Zipf's Law is noteworthy as a language design tool is
the very realization that words are sometimes shortened with information
loss in order to be consistent with usage frequency; hence acronyms.
(I concede that 'zilpavyxru' would have the necessary
meaning, but it is unnecessarily long and complicated.)
Why not just zilxru? The pa seems to be the logical default for a deleted
place, since you can get the others with SE rather than numbering.
I think that best solution to this, rather than changing the place
structure of 'xruti', is to create a new gismu using the gismu
algorithm. This will preserve everyone's existing lojban and the
baseline, and should also satisfy those clamoring for the change,
since an extra gismu will not hurt anything, and it will be easier to
get people to start using a new gismu correctly, than to get them to
use 'xruti' in a way inconsistent with the baseline.
Adding new gismu is an even bigger change to the baseline. We truly want
the gismu to be suffering "hardening of the arteries", even if the results
are a bit clumsy, at this stage. Otherwise the long term prognosis is that
language evolution will be excessively rapid, with new roots popping into
the gismu list as rapidly as new roots appear in Esperanto. The goal was
to force people to use lujvo, even LONG lujvo, to build the lexicon up,
constrained between the rigid set of short gismu, and the very clumsy
artificially-lengthened Type 3 borrowing rules (I still believe that we
have no frequency basis for making Type 4 fu'ivla yet, and only the bias
basis for culture words that did not make the gismu list and can't be made
as lujvo).
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