Hallo!

On Mon, 25 May 2009 16:15:41 -0700, Sai Emrys wrote:

> Regarding the 'hierarchy of aesthetics' and so forth: I've said my
> opinion on this before -
> http://archives.conlang.info/we/zhoenzhun/jolzeilbuan.html - but I'll
> reiterate as it seems apropos.
> 
> I believe that it is utter nonsense - and hubris - to talk of any
> objective valuations in any domain; valuations are axiomatically
> subjective, and thus must be agreed upon before the discussion begins,
> as with all axioms. (This applies elsewhere too;; eg coders
> differently prefer the three factors of elegance, power, and speed of
> reaching minimum utility needed for a job. None of them are
> objectively 'better' than the rest, and if two people disagree on what
> the order "should" be, then they won't have a very productive
> discussion of how to evaluate a particular project. [I happen to
> prefer them in about the order listed.])

Indeed.  Even in a "hard facts"-oriented discipline like engineering,
you get trade-offs all the time, and there is no single way of doing
it right.  Consider the endless debate whether front-wheel drive or
rear-wheel drive is better in automobiles.  Both have their advantages
and disadvantages.  It is the same with, for instance, auxlangs.  As
I have laid down in

http://www.joerg-rhiemeier.de/Conlang/auxlang-design.html

there are several design goals between which there are trade-offs,
and there are thus infinitely many ways of designing a "good" IAL.

> Aesthetics, in particular, are finicky things that people can't (or
> won't) even specify very well to each other, so it's rather hard to
> say whether two people who both claim to like "naturalism" and
> "elegance" will like the same things. This is why I proposed what I
> did in the link above, as a way to give a simple framework that would
> have relatively well-agreed-upon terms, so you can say *what* you
> like.

Amen.  If there is no objective way to tell which design is best in
engineering, there a fortiori is none in fine arts.  Criteria such
as "beauty" or "elegance" are even harder to operationalize than
"ease of learning" or "cultural neutrality" - tastes differ a lot,
and this ought to be honoured as a motivation for creating a vast
diversity of works.

> However, I also do not believe that it is helpful overall to simply
> brush aside all talk of evaluation and constructive criticism as
> mooted, primarily because we *do* have communally agreed upon values
> that can be (usually) accurately assumed tobe held by any particular
> conlanger, and we also have that person's own claims as to their
> desiderata for their language.

Sure.  There is certainly a difference in quality between Quenya
on one hand and the average teen-age beginner conlanger's cipher
of his L1 on the other.  But what is it that makes Quenya the
"better" conlang?  That is a question which is not as easy to
answer as it seems.

> If someone claims they want their language to be naturalistic, and you
> notice for example that it has only one click, then it is completely
> fair game (and IMO, a service to them) to point out that this is a
> very unlikely thing for a natural language to have, and thus is a mark
> against them under their own espoused valuation. (Likewise, if you
> have a language where all features individually are natural but rare,
> such as Klingon, it is legitimate to say that as a whole it is fairly
> unnatural.)

Yes.  But still, it is not always easy to say what is "natural"
and what is not.  There are surely things we can rule out as not
occuring in any natlang, such as oligosynthetic structure or
self-segregating morphology; but what would we think of e.g.
triconsonantal roots if the Afro-Asiatic family had not developed
them?  Such a device would probably seem "unnatural" to many
people - but the case of Afro-Asiatic shows that it *can* evolve
in human natlangs.

Yet, from examining human languages from all parts of the planet,
one indeed gets a workable idea of what kinds of phenomena occur
in natural languages and what kinds of phenomena do not, and
typological surveys such as the _World Atlas of Language Structures_
are highly useful and recommendable resources for the naturalistic
artlangers.

> The problem, thus, is simply one of first agreeing upon the terms of
> discussion. It would be silly to evaluate, say, Lojban or Solresol for
> their naturalism; they don't try to be, so complaining that they're
> not is fairly rude.

Indeed.  It is simply irrelevant, and different criteria are to be
called upon.  Naturalism is a relevant criterion only if the author
claims that the language was naturalistic or puts up some other
premise implying that (e.g., "This is a surviving pre-Indo-European
language spoken in some isolated farmsteads in a remote valley in
the Alps"); but most engelangs, for instance, are *meant* to be unlike 
natlangs in some ways, and a criticism of them as being not naturalistic
is meaningless.

>       It would likewise be silly to evaluate a newbie's 
> first conlang for its maturity, as they simply aren't at the stage yet
> where that's a relevant criterion. (Unless if course they claim
> otherwise, in which case one might want to gently point out how far
> they have to go before they really get there.)

Yes.  Some people (especially on the ZBB) tend to be snotty and
condescending towards beginners making beginner's mistakes.  Rather,
the experienced conlanger should be *helpful* and *constructive* in
his critique, pointing out problems in a polite and positive way,
and dwell on the strengths (there are few conlangs that do not have
some strong points, not even among newbie languages) more than the
weaknesses.  Nobody is born a master; and even the most coruscant
masters of our trade have cobbled together naive conlangs at the
beginning of their career.

> I believe that the question of *how* to assess languages publicly is a
> different question, and one more of social interactions and pedagogy.
> It is more useful to all concerned to give newbies a focus on what
> they do well, what they could change and what factors / options they
> might want to consider, than on what they did poorly. For someone more
> seasoned, however, I think it is perfectly legitimate to tell them
> what flaws you see in their creations more directly - possibly in
> private, depending on their personality. And for discussion in public,
> one tends to focus on the best features of the works one likes best.

Very much so.  Flaws indeed ought to be pointed out, because they are
something to learn from, but in a polite and positive manner, and
usually best in private.  Saying something like "Your conlang sucks!"
or "I see you have a LONG way to go to maturity" is not helpful.

> This is no different than any other creative field, really; the same
> discussion could be had about (say) architecture. It'd just be a
> different realization of most of the same underlying kinds of things
> (e.g. for architecture: elegance, cost to build, cost to run,
> enviro-friendliness, boldness, capacity, earthquake-proofness,
> lighting, bombproofness, etc etc).

Indeed.  The relevant criteria and, more so, their relative weights
in the evaluation process, differ from project to project, and it is
usually hard to say which solution is the best.

... brought to you by the Weeping Elf