Print

Print


--- On Sun, 6/28/09, <deinx nxtxr> <[log in to unmask]> wrote:

> This is one of the reasons I really don't like Lojban or
> its cousin Loglan.  I like the idea of predicate logic
> but see the biggest fault here being the use of too many
> arguments.  This means having to learn the argument
> list for each word and having to put a lot of thought into
> formulating a sentence so that the arguments come out in the
> correct order, complete with placeholders for those that are
> not present.
> 
The number of arguments isn't a problem; it's predictability that's tricky. My system allows users to predict the structure of any predicate.
 
> > The odd thing is that a truly human-parsable loglang
> could be
> > done. Loglan was the first attempt, so of course it
> was buggy.
> > The Lojbanists, instead of correcting the bugs,
> faithfully copied
> > them. When the Loglanists realized that there were
> fatal flaws in
> > the original design, they should have begun over, but
> they wanted
> > backwards compatability. The Lojbanists had no such
> excuse. At
> > some point I hope to field a proper project, but for
> the time
> > being it's a low priority.
> 
> I have a loglang that I've started.  It uses predicate
> logic too just like Lojban but I've the is nothing with a
> valency beyond 2, subject and object.  Where I'm stuck
> though is figuring out how to make it so the predicate logic
> can be easily parsed.  This approach involves a lot of
> nesting of phrases within phrases so I'm not sure it would
> be workable with complex sentence structures.

I'm not convinced that the Loglan compulsion to see how many modifiers can fit in a phone booth is useful. For one thing, it usually works better to focus on one thing at a time, so instead of having a lot of heavily modified arguments, for example, introduce and describe each argument individually. But in any case, if it takes a program to parse utterances, you're no longer dealing with a real language.

Steve