Hallo!

On Sat, 16 Feb 2008 00:59:46 -0800, Sai Emrys wrote:

> I'm considering a topic for a talk at a future LCC, about conlanging
> with constraints.
> 
> A literary allusion that comes to mind is that of the Abbe in Count of
> Monte Cristo, responding to the future Count's suggestion that as a
> free man is inventiveness might have known no bounds, to say that it
> were the bounds themselves that made him inventive.
> 
> Drushek, Kēlen, and Toki Pona are some examples that come to mind
> offhand as being in some sense formed by the constraints within which
> they flourish - voicelessness, verblessness, and complicatedlessness
> (hee).
> 
> What are other examples?
> 
> How have you experienced your conlanging as being influenced (for
> better or for worse) by constraints imposed upon it, of whatever
> source? What constraints do you have, and whence derived? Why have you
> imposed them? What constraints have you considered trying?
> 
> Please consider this a completely open-ended question (i.e. pretend I
> asked you the right question to elicit the most interesting answer
> :-P).

Each time an artist creates a work of art, he has to decide what
to do and what NOT to do.  So you always have "constraints" of
some sort.  In a conlang, you usually start with setting up a
phonology - by which you get a set of constraints determining
which word shapes may occur in your conlang and which may not.
You get further constraints when you decide on the morphology,
the syntax, etc.

My main conlang projects are meant to be naturalistic, and thus
naturalism is the main constraint for them.  So, I would not have
a stack-based syntax, an oligosynthetic structure, a taxonomic
vocabulary, or anything else one would not expect to meet in a
human natlang.

Further constraints result from the intended position of the
language in question in the human language tree.  Germanech,
for instance, is meant to be a Romance language that underwent
similar sound changes as German (High German sound shift) - this
means that it is guided by the application of a Grand Master Plan
derived from the historical phonology of German to Vulgar Latin.
That actually is quite much of a constraint that leaves little
freedom in the construction of the language.

I am working more freely in my Albic languages, but even there,
I have set myself similar constraints.  The individual Albic
languages are linked to each other by a system of regular sound
changes, so as soon as I determine the shape of a word in one
of them, the shapes of its cognates in the other Albic languages
fall out from those rules automatically.  Of course, a cognate
may have been lost and replaced by a word of different origin
in a particular language, or its meaning may shift, but most
of the words are determined by the interplay of the Proto-Albic
word form with the sound changes of the various Albic languages.

In my experimental engelangs, I use different constraints from
language to language.  X-1, X-2 and X-3 all have (different)
self-segregation rules.  In X-1, the valency of a predicate word
is equal to its length minus 2: a triliteral word is unary, a
quadriliteral word is binary, etc.  In X-3, all native morphemes
are exactly one phoneme long (the language is oligosynthetic,
so I get by with a large but not utterly unmanageably large
phoneme inventory).  However, none of these projects have
progressed much beyond the basic idea, as I find the constraints
rather difficult to follow through, and I feel that naturalistic
languages work better for me than engelangs.

... brought to you by the Weeping Elf