Re: Most developed conlang

On 4/18/07, MorphemeAddict@wmconnect.com <MorphemeAddict@...> wrote:
> In a message dated 4/18/2007 12:26:51 PM Central Daylight Time,
> markjreed@MAIL.COM writes:
>
>
> > > Esperanto prefixes and suffixes can be applied by rule, so they wouldn't
> > > count in counting words either.
> >
> > It depends.  Some of the meanings are not terribly predictable from
> > the base, either... words with -ajxo come to mind...
> Then speaking theoretically, all of Esperanto's words would be transparent,
> although in fact, some are not.
Even theoretically, there is no expectation that words derived with
-um  will be transparent.

Generally speaking I would say that non-transparent compounds and
derived words should count as distinct words while predictably derived
words should not count.  But this is a continuum, not a binary property
(with "eldoni" near the opaque end, words like "vortaro" somewhere in
the middle,
and obvious things like "belsona" at the transparent end), and it's
not something you could mechanize (unless you had previously marked
up a large E-o dictionary assigning a subjective transparency value to
each compound word).  The same is probably true of any non-isolating
language; even in engineered languages which aspire to
avoiding idiomatic compounds, probably some compounds and
derivations have a more transparent meaning than others.



On 4/18/07, Dirk Elzinga <dirk.elzinga@...> wrote:
> > Harold wrote:
> > <<
> > 1. Lexicon must be greater than 2000 (and contain essential words)
> > 2. Lexicon must be "a priori". (some loans are acceptible)
> > 3. Lexicon must be accessible on the internet.
> > 4. Grammar must be complete.
> > 5. Grammar must be able to be learned on the internet.
> > 6. Language must be speakable.
> >  >>
> I understood Harold's inclusion of these two items (internet presence,
> speakability) to be purely practical. The first because it is
> difficult to find descriptions of constructed languages otherwise, and
> the second I suppose because we are all used to "speaking", as opposed
> to emitting odors, or flashing light patterns, etc. I would hate for
> this thread to devolve into a "gee, that's a stupid requirement" kind
> of discussion.
I understood his point 6 to mean something like "the language must
be usable by humans in real time".  That is, ruling out languages
that are too alien for human brains to parse in real time, like Fith,
or that are probably too complex for most humans to learn to
speak fluently, like Ithkuil, but maybe not ruling out sign languages
or iconic languages on principle.  Harold can speak to this, and
anyway it is only relevant to his own project.


There was a discussion on Wikipedia some time ago about
evaluating the completeness of conlangs (among other
issues), some of which might be relevant here; see:

http://en.wikipedia.org/wiki/Wikipedia:Conlangs

and especially

http://en.wikipedia.org/wiki/Wikipedia:Conlangs/Notability%2C_verifiability%2C_merit%2C_completeness


--
Jim Henry
http://www.pobox.com/~jimhenry