From @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET  Thu Oct 14 10:46:40 1993
Received: from ELI.CS.YALE.EDU by NEBULA.SYSTEMSZ.CS.YALE.EDU via SMTP; Thu, 14 Oct 1993 14:52:20 -0400
Received: from YALEVM.YCC.YALE.EDU by eli.CS.YALE.EDU via SMTP; Thu, 14 Oct 1993 14:49:32 -0400
Message-Id: <199310141849.AA19754@eli.CS.YALE.EDU>
Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2)    with BSMTP id 0513; Thu, 14 Oct 93 14:47:38 EDT
Received: from CUVMB.COLUMBIA.EDU by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with  BSMTP id 0808; Thu, 14 Oct 93 14:50:10 EDT
Date:         Thu, 14 Oct 1993 14:46:40 EDT
Reply-To: Jorge LLambias <jorge@PHYAST.PITT.EDU>
Sender: Lojban list <LOJBAN%CUVMB.bitnet@YaleVM.YCC.YALE.EDU>
From: Jorge LLambias <jorge@PHYAST.PITT.EDU>
Subject:      Re: TECH: Lean Lujvo and fat gismu
To:         lojban@cuvmb.cc.columbia.edu
Status: RO
X-Status: 

la djan spuda la lojbab di'e

> Thus "mibykulnu" surely means "kulnu be mi", my culture, not "kulnu be fa mi",
> which would lead to the place structure "I am a culture of nation x1", a
 rather
> useless lujvo.

I agree. This is how I'd interpret it. In Nick's notation:

mibykulnu  k1 (k2=mi)

But you change your mind in the case of {dunda}:

> Even so, there are possible cases where more than one
> interpretation might make sense.  Thus "dondunda" could be either "x1 gives x2
> to you" or "you give x1 to x2".  So the problem is not unique to "zi'o".

dondunda  d1 (d2=do) d3  is the most common construction, but
          d1 d2 (d3=do)  is the more reasonable semantically.

Any of the two are allowed by the be-lujvo construction. "You give x1 to x2"
is not, since {dondunda} should be a type of {dunda}, not a {se dunda}.

> Consider the notion that "don-" always means that "do" is in the x1 place,
> and we must say "terdon-" to put "do" in the x3 place.  Then the natural
> "mibykulnu" above must be "selmibykulnu", and we have no natural way to
> construct a lujvo with place structure "x1 is the gift you give to x2".

donseldunda  d2 (d1=do) d3

is how I'd make a lujvo for that.

> On my principles that is "seldondunda",

For me that gives "x1 is the gift x2 gives to you".

> or more precisely "seldonpavdunda",
> where the "-pav-" forces the "do" into the x1 place;

I don't like this subscripted rafsi anyway, but in the case of the {pa}
subscript, it goes against the principle that the x1 of the lujvo matches
the x1 of the last component. For the x2, it shouldn't be needed, unless
there is a more plausible interpretation that pushes it further on, as in
the case of {dunda}, where {do} is much more likelier to be a recipient
than a gift.

> on your principles,
> "seldondunda" means "x1 gives you to x2",

I don't like that method either.

> surely a less useful result
> (which on my principles can be expressed by "donreldunda").

This would be necessary to avoid the ambiguity between the possible
d1 (d2=do) d3, and the far more likely d1 d2 (d3=do). But this happens
in all lujvo. Why does {solgu'i} mean sunlight, rather than light
illuminating the sun? We don't need subscripts to choose between the
two, and neither should we need them for {mi}, {do}, etc.

For {zi'o} they are needed, because it is more difficult to see what
should be the eliminated place, but hopefully the concepts that are
obtained by such procedure and are not covered by other gismu will be
rare. In any case, any method to introduce zi'o in a lujvo will give
ugly results, and I almost prefer magical disappearance, to the
cumbersome zi'o.


Jorge