Properties etc. and the Spatial Metaphor

[Ivan A Derzhanski <iad@MATH.ACAD.BG>]

We seem to have a problem here.  There is a bunch of semantic research
(I'm thinking primarily of Jackendoff's, because that's whose work I am
most familiar with) that shows that the Spatial Metaphor (which means
talking of being at or moving (away) from/to(wards) non-spatial areas,
eg possession, properties, existence etc.) is a very pervasive thing in
language.  That is, it is not at all malglico, nor even mal-[Standard
Average European], to think and talk of a house as being/staying in or
coming into someone's possession, an object moving from one size to the
other, someone keeping something in existence, and so on.

Now it could be said that Lojban isn't _a priori_ going to reflect all
this, and if there is something about the mind that absolutely requires
it, it will make itself known eventually.  Fine.  But as Jorge pointed
out, some gismu already do reflect it, and it appears most illogical --
to'e logji, I'm tempted to say -- for the rest not to.

A note about the extra argument places.  I don't see that as a problem.
Spatial motion doesn't always involve a route or a means of conveyance.
Think of teleportation; that doesn't happen via anything.  Or think of
the final phase of Three Men's Morris (the only phase in some variants
of the game, I think): any stone can be moved to any vacant field, so
a move has no route, and all fields are equally close to one another,
so there is never any `away from' or `towards' either, just `from' and
`to'.  Yet the stones are involved in genuine spatial motion.  Does it
matter that it is not the 3d continuous space we operate in?  I don't
think so.

On the other hand, Lojbab is right that letting every spatial word
in Lojban have a non-spatial interpretation can lead to confusion,
what with the general law that pretty much everything is elidable.
A natlang could say that a spatial word has its principal spatial
meaning unless one or more of the argument places are occupied by
non-spatial arguments.  But in Lojban there is no way to enforce
the filling of any argument place, so things like {le zdani cu
klama} will end up meaning nothing at all (they'd be trivially
true, since everything moves all the time, say, towards a state
of being older), and we don't want that to happen.

--
<'in tarY ad-dunyA 'a.gArat    wa-nu^gUmu as-sa`di .gArat
 fa-.surUfu ad-dahri ^sattY,   kullamA  ^gArat,  'a^gArat.>
                                               (Nasr Ibn Hasan  Marginani)
Ivan A Derzhanski                              <iad@banmatpc.math.acad.bg>
H: cplx Iztok bl 91, 1113 Sofia, Bulgaria  <http://www.math.acad.bg/~iad/>
W: Dept for Math Lx, Inst for Maths & CompSci, Bulg Acad of Sciences