[lojban] Re: Loglish: A Modest Proposal

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

--- Ben Goertzel <ben@goertzel.org> wrote:

> 
> Hi,
> 
> > On 8/16/05, Ben Goertzel <ben@goertzel.org>
> wrote:
> > > For instance, I was thinking about
> something like a "queh"
> > > reference-indicator, to be used as in
> > >
> > > "Lo man go lo playground.  It queh man kill
> lo dog.  It eat lo cat."
> > >
> > > ("The man goes to the playground.  He kills
> the dog.  He eats the cat.")
> >
> > I think you need {cu} in front of "go" and
> "kill", otherwise
> > you get tanru: "lo man go" would be "the man
> goer" and
> > "man kill" would be "man killer".
> 
> Correct, sorry...
> 
> > Also "go" could be confused with the cmavo
> {go}.
> 
> That is a more frustrating problem...
> 
> In writing one could solve it by some awkward
> mechanism like
> 
> "Lo man cu _go lo playground"
> 
> I don't think it occurs often because very few
> cmavo are English words
> 
> In speech I can't think of a non-annoying way
> to solve it... except by
> outlawing/replacing
> those few English words that overlap with
> cmavo, e.g. replacing "go" with
> "proceed"
> 
> "Lo man cu proceed lo playground"
> 
> or with "eng_go" (enggo?),
> 
> "Lo man cu enggo lo playground"
> 
> ;-p

Yuck indeed
 
> > > The idea here is that "it" is used as a
> generic referential
> > indicator, but
> > > "queh" is used to precede a qualitative
> indicator of what the
> > "it" refers
> > > back to.  Then, subsequent uses of "it" are
> assumed to keep
> > referring back
> > > to the same referent, until another use of
> "it" coupled with "queh"
> > > ooccurs...
> >
> > That's just how Lojban {goi} works.
> 
> Hmmm... I'm not sure it's exactly the same.
> 
> I think my example was not sufficiently
> evocative, though.
> 
> For instance, in my intended usage, you could
> say
> 
> "la Dr. Benjamin Goertzel cu enggo lo
> playground.  It queh man cu kill lo
> dog.  It cu eat lo cat.
> 
> In this case, "It" is defined to refer to "Dr.
> Benjamin Goertzel", not to
> "man" generically.

Well, this is a nice bit of editing, but is it
one that AI can universally do, that is, identify
what in the neighborhood falls appropriately
under "man" (in the sense "male" of course)and
what falls under female and so on. There are any
number of "genders" as the need arises --
including some that somehow cut across the
"natural" ones, to separate Flash from Ming in
the everpopular "Ming walked up to Flash.  He
struck him." Indeed, such a system of gendered
pronouns is one of the shorter solution to Lojban
problem with reference.  The use of letters as
pronouns (a nice old Logical habit that) is the
usual version of it, but runs afoul of 1) two
words in context with the same initial letter 2)
at a distance it is harder to remember how a
thing was referred to than what thing it was (for
both speaker and hearer).  these are relatively
minor problems compared to remembering the exact
position in the tree of a referrence three
sentences back -- or even on a different branch
in the present sentence, as the most precise
Lojban system would have it.  Of course, as the
old joke goes, "Repetition is also anaphora" but
we would like something at least as good as
English and better if possible.  Indeed, perfect
if possible: speaker could refer pronominally
uniquely to anything mentioned before and still
in his scenario and hearer -- assuming he has
been paying attention (kept up with the scenario)
could retrieve the correct referent.  Of course,
we can do this with forethought {goi}, but we
usually don't forethink and we often don't keep
enough track of what is said to get back to the
way the initial referent was made (did he call it
a dog or a mutt or...?), let alone exactly where
that reference fit into the sentence structure,
to reclaim things by these guidelines.  The
suggestion of using an approximation (well it was
called a canine of some sort so a pronoun flagged
"canine" is probably about it) is helpful, though
hardly an improvement on (partial) reptition (the
yellow-and-black dog" repeated as "the dog" or
even "the yellow thing").

> The "man" qualifier following the "queh" is
> intended to merely *guide the
> listener's mind* toward the right antecedent
> for the pronoun.  It's not
> intended to *explicitly define* the pronoun. 
> So, basically
> 
> "It queh man"
> 
> is the rough equivalent of the English "he",
> and
> 
> "It queh woman"
> 
> is the rough equivalent of the English "she"
> 
> Is there an equivalent of "queh" in this sense
> in Lojban?

Nope. But somehting like it might help.
 
> > Lojban also has {fi'o}, which is just like
> your "quu", except
> > for word order: "lo pliers quu weapon" =
> {fi'o -weapon lo -pliers}
> > (or rather {fi'o xarci lo cinza}.)
> 
> Yes, I guess that is right...
> 
>
http://www.lojban.org/en/publications/reference_grammar/chapter9.html
> 
> (I think I didn't encounter that one in Lojban
> for Beginners, or maybe it
> was there and I didn't remember it...)
> 
> > And while {zei} is not exactly the same as
> "qui", since you
> > don't have lujvo in Loglish you could use it
> for that.
> 
> That is a good suggestion.




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.