(Moved to CONLANG from conlang_learners because it's more meta and
less project-oriented than it started out.)

[old Subject: Re: [conlang_learners] Comments on Alurhsa]
On Mon, Aug 31, 2009 at 12:43 PM, Amanda Babcock
Furrow<[log in to unmask]> wrote:
> On Mon, Aug 31, 2009 at 12:10:29AM -0400, Jim Henry wrote:

>> Alurhsa is a complex language, perhaps one of the more difficult
>> ones nominated, but there are certainly plenty of materials to practice
>> with.

> I'm surprised to hear it called complex, but I guess complexity is in
> the eye of the beholder.  To me, a language like Kelen or Ilaksh, where
> I would have to compose my thoughts in a completely different manner,
> would be complex; unfamiliar word orders, unfamiliar parts of speech,
> unfamiliar mappings of cases to roles, a different idea of how big a chunk
> of meaning a "word" consists of, or very different semantics so that
> most words did not map one-to-one from English to the conlang, would be
> complex.
>
> Granted, the verb paradigm of Alurhsa is very elaborate, and we'd
> have to work to be sure we were taking full advantage of it instead of
> recoding English ways of expressing things, but I note that the verbs
> do come in one unified conjugation, excepting the handful of irregular
> ones.  And it was very easy to use for the Inverse Relay, since there
> appear to be close correspondences between the meanings of the Alurhsa
> words and the English glosses.

Interesting.  What you're calling "complexity" I would call
"exoticity" (or "exoticness"?  or simply "strangeness" or "weirdness",
with no pejorative (rather the opposite) intent).  A language is
complex, in my 'lect, if it has many grammatical categories words
inflect for, or the grammatical categories have many .... hm, what's
the word?  Subcategories?  I mean, other things equal, with two
languages that both inflect nouns for case and number and nothing
else, one is more complex if it has the same cases as the other plus
an instrumental and locative, and the same numbers as the other plus
dual and trial.  Or if two otherwise similar languages inflect nouns
for case on the one hand, and definiteness on the other hand, I might
call them equally complex if the number of cases in the one is equal
to the number of kinds of definiteness in the other.

On the other hand, if two languages have the same number of categories
and subcategories, but the categories of one are case, number and
definiteness and the categories of the other are beauty, truth, and
goodness, then they're equally complex but the second language is more
exotic.  (Unlesss your native language inflects for those categories
but not for case, number, etc.)  Irregularity also increases
complexity.

Re: phonology, I'd say complexity increases with the number of
phonemes and with the number of possible syllable structures, but I'm
not sure offhand of an objective way to compare the overall complexity
of a language with a small phoneme inventory and complex phontactics
vs. another with a large phoneme inventory and simple CV syllables.

Re: semantics, it seems that a language with more polysemy on average,
and more idiomatic phrases and compounds whose meaning can't be
predicted from the meaning of their component words or morphemes, is
more complex than one with less polysemy or fewer idiomatic
constructions.  E.g., comparing English with E-Prime, the latter is
less complex semantically because of its devices to disambiguous
polysemous English constructions.  Basic English is probably more
complex semantically than most oligosynthetic engelangs with a
superficially similar vocabulary size, because it smuggles in English
phrasal verbs and the like.

Exoticness is in the eye of the beholder, consisting in degrees and
kinds of difference from one's native language and other languages one
has studied deeply.  Complexity on the other hand is, I would argue,
something we could agree in principle on an object way to calculate,
at least re: inflectional morphology, syntax, and phonology.

Toki Pona is simple and moderately exotic; Ithkuil is very complex and
very exotic; Alurhsa is fairly complex and slightly exotic; Sambahsa
is moderately complex but not exotic.  gjâ-zym-byn was originally
designed to be simple and exotic, but it's gotten more complex over
the years, though (except perhaps in its phonology) I don't think it's
as complex as most naturalistic artlangs.

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