Re: [lojban] Re: Loglish: A Modest Proposal

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

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

> 
> > The example, though I assume it is not meant
> > seriously, illustrates the problem that has
> been
> > found in these regimented English proposals:
> 
> Loglish is not fairly called "regimented
> English", because its syntax is 98%
> Lojban; only its vocabulary is English

This is definitional, of course, but, since, for
the most part, Loglan syntax is a subset of
English syntax (formally speaking, of course),
the  name seems to fit.  The point is that
keeping people to that subset is very difficult
if they are native speakers (or even very fluent)
in the full set.
 
> >the
> > whole range of ambiguity of the natural
> language
> > creeps in
> 
> I disagree -- Loglish uses Lojban syntax, so
> that most of the ambiguities of
> English do not creep in.  The only English
> ambiguities that creep in are
> ambiguities regarding
> 
> -- word sense
> -- argument position
> 
> but this is a small percent of all English
> ambiguities, and is resolved by
> proper use of qui and quu.

It is not clear what percentage of ambiguity is
which, especially since they often go together
(different syntactical structures often rely on
different readings of the same word -- or
conversely).  But English words are generrally
very ambiguous (even when we stick to a single
etymology for a phonemic sequence) and this will
carry over into Loglish to its disadvantage
(relative to Lojban at least).
  
> > Loglish also loses what is practically
> Lojban's
> > most significant feature for any computer
> use:
> > the unique decomposition and parsing.
> 
> That is sort of correct; however Loglish as I
> defined it will have *close to
> unique* parsing IF quu is used correctly, so
> that it's possible to resolve
> the intended argument position from the quu
> specifier using simple,
> automated semantic inference.

Which is only a gleam in some few technicians'
eyes.
 
> In nearly all cases it will be possible to
> achieve successful results via
> simple algorithms such as
> 
> "Resolve 'X qui Y' to the sense of X whose
> WordNet definition has the
> smallest semantic distance to Y."

Does wordNet really have quantized semantic
distances? Is it likely to soon? As of now, I
suspect this has to be done by hand for the most
part.  Not that this is necessarily a lethal
flaw, but it does cut back on the value of
Loglish relative to Lojban, which has unambiguity
buiolt in at almost all level (any solution for
Loglish would work even more effectively for
Lojban).
 
> "Given 'X quu Y' , assign Y to the argument
> position of X whose description
> in the Loglish dictionary has the smallest
> semantic distance to Y."
> 
> I'm quite confident these algorithms would work
> with 97%+ accuracy, and 99%+
> accuracy after some training and fiddling.
> 
> Of course, this quu algorithm requires a
> Loglish dictionary to be written,
> but this dictionary doesn't have to be complete
> because one can use another
> algorithm:
> 
> "Given 'X quu Y' , if X is not in the Loglish
> dictionary, find the
> semantically closest Z to X so that Z is in the
> Loglish dictionary, and
> assign Y to the argument position of Z whose
> description in the Loglish
> dictionary has the smallest semantic distance
> to Y, and then assign Y to a
> corresponding argument position for X"
> 
> I bet this will work with 90%+ accuracy.

If X is not in the Loglish dictionary, how do we
get a measure of semantic propinquity to X?  I am
not quite sure how well this will work even with
knowledgeable human speakers, let alone with
machines.


Obviously this is more complex and funkier than
> Lojban parsing, but OTOH
> having the full English vocabulary to use is a
> big thing...

And a source of many of the problems -- even
(maybe especially) if all the words are fully
defined.
 
> > I suppose that
> > "qui" -- and in another way "quu" -- would
> come
> > to function like this in Loglish, both
> > disambiguating simple expressions and
> > constructing new complexes. It seems a viable
> --
> > though remarkably messy and uninteresting --
> > idea.
> 
> I agree it's viable and messy, but I don't find
> it uninteresting.  But of
> course, this is a matter of taste ;-)
> 
> Tanru are also messy, and semantically
> underspecified.

Quite true.  I didn't say that Lojban was not
messy -- though we tend to talk about that as
"creative space' or the like.  The same would
apply to Loglish "qui" and "quu," to the
detriment of automatic interpreting, though
perhaps as an aid to literature. Still, the
problems with Lojban are inherently smaller than
those with Loglish, since the semantic range of
the items involved is smaller (and remember that
the most stunning uses of combo are precisely
those that involve the largest leaps over
semantic distance, not the ones that are
calculable as the shortest leaps).

> What I like is that this messiness of compound
> formation -- like the
> messiness of tanru -- occurs within a context
> of predicate-logic-based
> syntax/semantics (defined by the Loglish cmavo
> and sentence structure)

Yes, I too like (for professional reasons if no
other) the underlying FOPL.  But it seems to me
that, having started down that route, we should
follow it as far as feasible.  Lojban seems to me
to be pretty close to that goal (some further
steps have turned up in various proposals over
theyears that have not gotten incorporated into
the development of Lojban, but they are mainly
rather minor -- trimming, not the core.