[lojban] Some possibly helpful notes from LoCCan3

[John E Clifford <kali9putra@yahoo.com>]

1.  On precision: no
LoCCan sentence can be more precise than its least precise brivla and will
usually be less so.
 
 2.  Simple bridi are all right as they stand,
simple arguments and simple predicate. This is the heart of the language of
Logic.
 
3.   Beyond that, as
& pointed out in a recent post, adhering to predicate logic, in any of its
various forms, in a spoken/written Language results in a hard to follow
structure full of useless words.  Not
what LoCCan was meant to be.  BUT all
that linguistic detritus is what allows predicate logic to be free of syntactic
ambiguity (parentheses of a variety of sorts variously realized in print).  
 
4.  So we need to make
all this more speakable and yet preserve the unambiguity.   Loglan tried one way to do this and Lojban
took that over for the most part and expanded it.  LoCCan3 should probably start by reexamining
the ways this was done and, hopefully, finding some other (better) ways to do
it.  Looking at Montague’s original
papers and the subsequent stuff on English and other languages in that
tradition may offer some clues.  So will
trying to figure out the back transformations from Loglan/Lojban to predicate
logic.  This needs to be unique (as it
may be, but is not obviously so).
 
5.  The semantics of
predicate logic, even of the wildly enhanced form that now underlies LoCCan, is
simple, reading directly off the syntactic structure of the sentences
involved.  LoCCan can have a simple
semantics, too, if this underlying syntactic structure is uniquely recoverable.  Again, Montague may be helpful.
 
6.  The transformation
of predicate logic to LoCCan seems to involve only a few thing: moving
quantifiers and negation from canonical position to “more natural” positions in
sentences, replaces variable by a range of variously ad hoc sorts of anaphora,
collapsing parallel sentences with overlapping content into single sentences
with the former parallelism now shown in joining the nonoverlapping parts
within the overlapping frame, contextualizing many quantifiers to some contextually
selected member of “natural” sets, and, possibly most significantly, dropping
RHE markers “where they are not needed.”
Are there other steps which might be easier or clearer and
yet preserve the languagey character of the resultant?
 
7.  Those contextually
selected items need some work (or rather need to be put back to an earlier
state).


      

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.