Return-Path: <jimc@math.ucla.edu>
	(Sendmail 5.61/1.07) id AA07300; Tue, 13 Apr 93 08:59:35 -0700
Message-Id: <9304131559.AA07300@julia.math.ucla.edu>
Subject: Re: lujvo paper, part 3.2. 
             <9304130453.AA24612@julia.math.ucla.edu> 
Date: Tue, 13 Apr 93 08:59:34 -0700
From: jimc@math.ucla.edu
X-Mozilla-Status: 0011

> For example, {posydji}, to want something for oneself, has places d1=p1, d3,
> p2. d2 is {lenu ponse}, and is thus redundant; we have decided that p3, the
> law of ownership, is an irrelevant detail in an expression of desire. 

In my gismu list (possibly old), djica = "desire/want...for..." where I
interpret d3 as some kind of goal that d2 will help bring about.  It
seems to me that p2 (the thing to be owned) belongs before d3.  

It also seems to me better to use a "lean" gismu definition in which
the goal is stuck on with mu'i/ mukti.  There are 22 gismu with places
keyed by "purpose" and (IMHO) well under half have any business having a
dedicated purpose case.  

With -gua!spi definitions at least, you get better automatic ordering
if you put the sub-word places (here ponse) before trailing main word
places, rather than after.

> An interesting example of a be-lujvo is {xande'icalku}, fingernail. 

This word illustrates your points better than jgalu would, but I think
jgalu is better semantically. It also has the "nail" keyword in my
gismu list.

			-- jimc