[lojban] Re: A Proposed Explanation of {gunma}

[John E Clifford <lojban-out@lojban.org>]

--- Jorge Llambías <jjllambias@gmail.com> wrote:

> On 12/15/05, John E Clifford
> <clifford-j@sbcglobal.net> wrote:
> > After (probably immediately) first arguments,
> > immediately before others but also in
> compound
> > predicates at the appropriate places around
> the
> > predicate (hard to state clearly which
> probably
> > means it still has bugs in it).
> 
> Yes it does. You would basically have the same
> situation
> as now, except that you would put the mark at
> the back
> of the sumti instead of at the front.
> 
> Consider a simple example:
> 
>   ko'a ko'e ko'i broda gi'e brode vau fo'a fo'e
> fo'i
> 
> Suppose you want to mark broda as distributive
> for ko'e
> and non-distributive for fo'e, and brode as
> non-distributive
> for ko'e and distributive for fo'e. Where do
> the marks go?

Yeah, this is pretty convincing that any complex
case will be a mess.  It does, however, point to
another place where the gadri solution does not
work -- and a much more common one than the
previous peculiarities: when the sumti is not a
description at all.  As always, we can say that
context solves it or that we can always find some
sort of paraphrase, maybe even find a rule for
constructing such, but none of these are very
elegant -- or vewry Lojbanic -- suggestions.  If
the distributive-collective distinction is to
really be functional in Lojban (and that is an
open question with several presuppositions about
what terms mean and the like) it should have a
clear, universally applicable, representation
(that is part of the design criteria for
Loglan/Lojban from the get-go).  It doesn't now.

> One solution would be to have the marks on the
> selbri, with
> an indication of which place it applies to,
> either a subindex
> or a different word for each place.
> Alternatively, you can
> mark it on the sumti with an indication of
> which selbri it
> corresponds to. Either way is cumbersome.

The immediate idea was using something like the
{fV} series to show places (mark on the predicate
as it were,)so 
{ko'a ko'e ko'i broda fe d fu c gi'e (I suspect
it has to be a different connective at this
point) brode fa c fu d vau (maybe this will have
to change as well) fo'a fo'e fo'i}

> >  I suspect that any solution is going
> > to have some problems but overall the UIish
> one
> > seems simplest and is more uniform (and has
> fewer
> > odd ontological suggestions).
> 
> Ulish?
> 
Able to occur in a wide variety of contexts, even
if not literally after (or before) absolutely any
word.

Back to something you said earlier, namely that
gunma1 stands for a single thing and gunma2
stands for the several things that comprise it. 
While you are objecting bunches (I think -- you
may just be pointing out that the bunches theses
so far are not quite right for plural
quantification/reference), I wonder what one
thing you have in mind for gunma1.  Given that I
also take you as saying that {lo broda} and {loi
broda} are the same thing only with difference in
declared distributivity, I suppose (given your
views on what lo broda is) that loi broda -- and
hence gunma1 generally -- is some sort of Urgoo,
at least in many cases.  In which case I suppose
that gunma2 is a list of
avatars/slices/exemplars/... .  Do I have any of
this close to right?




To unsubscribe from this list, send mail to lojban-list-request@lojban.org
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
you're really stuck, send mail to secretary@lojban.org for help.