Re: [lojban] Re: ralju bangu be le gligu'e

[John E Clifford <clifford-j@sbcglobal.net>]

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

> On 5/4/06, John E Clifford
> <clifford-j@sbcglobal.net> wrote:
> > To answer the last (and central) question
> first:
> > {bridi} like logic assumes that the structure
> of
> > a bridi is (ignoring some irrelevant details)
> > [selbri]<sumti>, in the instant case [ralju
> > bangu]<le gligu'e>.
> 
> But {ralju bangu be le gligu'e} is a selbri,
> not generally a bridi.

No, {ralju bangu} is a selbri; the incidental
conversion required for fitting the arguments
into a sumti does not change their status as
arguments.
 
>  The Lojban grammar has it
> > ralju [bangu <le gligu'e>], which does not
> > translate conveniently into the logical form.
> 
> How does the "logical form" deal with tanru?
> Because this is
> only about how one tanru-unit modifies another.
> Consider
> for example the tanru {broda be ko'a broda be
> ko'e}, where
> the tanru-unit {broda be ko'a} modifies the
> tanru-unit
> {broda be ko'e}. Would it make more sense for
> it to parse
> as {((broda be ko'a) brode) be ko'e}? The
> grammar might have
> been defined that way, but it wasn't (and I
> suspect it would be
> more complicated doing it that way). But
> neither is more logical
> than the other. Similarly {broda brode brodi}
> was arbitrarily
> defined as {(broda brode) brodi} but that's
> neither more nor
> less logical than if it had been defined as
> {broda (brode brodi)}
> In fact, I suspect the latter would have been
> more convenient.
> But how do you get one being more "logical"?

Well, I suspect you are playing on an ambiguity
in "logical."  I mean (as does the "logical" in
"logical language") that the grammar is that of
First Order Predicate Logic -- as adapted.  Now,
FOPL doesn't have complex predicates directly,
but a large number of them can be constructed
within its framework and from that the pattern
emerges that all the predicates go together and
all the arguments together, however they may be
spread out eventually on the surface form.  The
underlying forms are of many sorts since there
are many ways that one predicate can modify
another (including some that cannot be done in
FOPL since they are essentially Second Order --
at least).  Lojban grammar tries to do the whole
thing with the surface structure rather than the
logical structure in this sense (and hence
supposedly the semantic structure) and so makes
life much more complicated at the interpretive
level.
 
>  It
> > is, of course, a possible logical form, but a
> > derivative one (it assumes the first form,
> note).
> 
> I don't see it. How is one more or less
> derivative than the other?

Well, as noted the one presupposes the other and
uses it in its own construction -- a reasonable
definition of derivative.
 
> >  Yet it is the unmarked form here and the
> basic
> > form is marked by a complex set of
> parentheses.
> 
> The parentheses are only needed if you want to
> link a sumti
> to a compound selbri, they are not needed for
> simple selbri.

Naturally, since there can only be a difference
if ther is more than one predicate involved.

> That would seem to be at least as "logical" as
> the other way,
> and more symmetric if we consider that the
> modifier tanru-unit
> can have its own arguments too.

As I said, this is just equivocation.  The other
way (i.e., as Lojban does it) might be as
sensible (though I don't think it ultimately is),
but it is not logical.
 
> > Whether this derivative form matches the
> meaning
> > depends upon what the meaning is, and it is
> not
> > clearly the one this form eventually matches.
> 
> Not sure what you mean by that, but the meaning
> of the Lojban
> structure does in this case match its form as
> defined by the
> grammar.

That is an assumption which is at least
controversial, so does not help this issue at
all.
As noted, some bits in CLL seem to imply that it
is the logical form that prevails, though others
point toward the Lojban one.  There is not enough
concrete discussion or clear cases to really come
down definitively on one side or the other.  The
syntactic point is clear, however: from the point
of view of logic, the basic form is marked and
the derivative one unmarked.  this may make
linear sense, but queers the deep structure.