From lojbab Tue Jul 12 00:36:51 1994
Date: Tue, 12 Jul 1994 00:36:44 -0400
From: Logical Language Group <lojbab>
Message-Id: <199407120436.AA14136@access1.digex.net>
To: jorge@PHYAST.PITT.EDU
Subject: Re: a simple question...
Cc: lojbab@access.digex.net, lojban@cuvmb.cc.columbia.edu
Status: RO

JL> Could you give an equally explicit rule for no'a and the vo'a, vo'e, ...
JL> series? It would be nice that they all refer to the same level of selbri
JL> and arguments, but from the last discussions it's not clear to me to which
JL> selbri/arguments
JL> they refer. Is it the first selbri up from where they appear, the main
JL> selbri of the sentence, or something else? (I'd prefer first selbri up.)

Let's try two sentences

> le plise cu cpana le tanxe noi [] cpana le karce
> The apple is on the box, which incidentally is on the car.

At location [], 
ri = le plise
ke'a = le tanxe
vo'a = le plise
vo'e = le tanxe noi cpana le karce
no'a = cpana, thus giving access only to the same sumti given by vo'a
no'axino = isn't formally defined - I would say cpana, but there is
   justifiable argument for tanxe, without which I don't think you can
   access specified description places of tanxe, if they were present.
   (e.g. "le tanxe be le plise be'o noi [] cpana" has no way that I see
   to access "le plise", unless we use "no'a" or some as-yet-undefined
   subscripting or something based on "ke'a"

> le karce cu bevri le dasni be le [] mapku le zarci
> The car carries the wearer of the hat to the store.

At location [],
ri = le karce
   (but if it were le karce pe la djan. or le la djan. karce, ri = la djan.
    and ra = le karce)
ke'a is undefined
vo'a = le karce
vo'e = le dasni be le mapku
vo'i = le zarci
no'a = dasni
no'axire = bevri

This is not exactly the most tested feature of the language, so arguments
for change can be considered (though we better hurry %^).

lojbab