Date: Wed, 13 Jul 1994 04:58:36 -0400
From: Logical Language Group <lojbab>
Message-Id: <199407130858.AA09096@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
X-From-Space-Date: Wed Jul 13 04:58:42 1994
X-From-Space-Address: lojbab

JL> You say that in {le dasni be le mapku pe le no'a}, no'a = dasni.

I'm not sure.  I'm not even sure whether you might have omitted a be'o.
I will assume you did what you wanted and the pe phrase is intended to attach
to "le mapku" and not "le dasni be ..."

Actually, I didn;t say anything earlier about "pe" relative phrases; they are
in the same somewhat unconventionalized position as relative clauses.  Cowan
reminded be today that the use of a "ku" can make the difference between
whether a relative is attached inside the description or outside, which in
turn also may affect the interpretation.

I think this is something we had better talk out at length during LogFest.
Probablyanother cmavo is going to be needed to handle all cases clearly.

JL> Couldn't we have {vo'a}, {vo'e}, etc to be {le no'a}, {le se no'a}, etc?
JL> This would make reflexives easy:

Well, if we did that, the two would be redundant to an extent.

At one time the default for "vo'a" was "le go'i"; the referent was to the
PREVIOUS sentence by default, but people convinced me (by usage as much as
anything else) that the latter was sufficient.  When I put "vo'a" in the 
language, I intended it to be flexible, and to normally require restriction:
"vo'a pe di'u", "vo'a pe dai" etc.  (Actually, come to think of it, 
it originally had a syntax that did not have the "pe", but was followed by 
the bridi indicator immediately: vo'ago'i vo'afu'i (fu'i became no'a, if I
recall)).

One major purpose of "vo'a" is for explicitly dealing with "and vice versa"
which has a special metalinguistic syntax (soivo'evo'a).  It really WAS
intended to bounce you out to the main bridi, because I didn;t conceive of
the need to refer to other sumti at the subordinate level.  I think we still
need such a set, and vo'a, etc. was specifically chosen to assocaite with
fa/fe/fi, etc. and has seen a fair amount of usage.  "no'a" has seen little
if any usage, and must be though about in the context of its related words
also starting with 'n' if I recall.  But it is intended to be used more like
"go'i" - more flexible than the sumti set and taking less words.  There
is not a lot of need to devote separate words to each of the places of
"no'a" because it really isn't all theta often that one wants to deal with
oblique places of the next outer selbri.  We need the capability, but not
enough to devote a big chunk of words to it.

The question though is whether "no'a" really is intended to deal generally
with whatever we as natlang speakers consider 'reflexives', or whethe it
is to be used to access a particular selbri and its places (and if the
latter, which one).  It was born to the former purpose, but probably has only 
been thought about since in the latter context.

I'm sure this is a confusing answer.  Sorry about that.

lojbab