On 29 July 2011 06:03, David Peterson <[log in to unmask]> wrote:

> >
> > One artistic possibility that's been around for a while but hasn't
> > gotten much attention in conlang-classification debates is Gary
> > Shannon's method of "using a language into existence" -- where
> > experimenting with different methods and processes of creating
> > conlangs is as much a focus of his art as the resulting conlang, and
> > where using the language (starting to write the corpus) is
> > simultaneous with or prior to many design decisions which, in the
> > traditional naturalistic school, usually long precede writing any text
> > in the language.  I've done some conlanging in that mode myself,
> > though I've never documented my efforts in that line the way Gary has.
>
> I'm not sure I'd call this a school, so much as a method. After all, one
> could very well do this and use it as an auxlang—that is the overall purpose
> of the end result isn't defined by the means of construction. It's just a
> different way of attacking the problem of creation.
>
>
Some schools of art are defined more properly through the methods by which
works of art are created, rather than the nature of the works of arts
themselves. So both are not exclusive :) .


>
> Beyond that, there are design elements or ideas that could be applied to
> language that have endless possibilities. Here are a couple I've been
> thinking of.
>
> == Decoder Ring Language ==
>
>
*snipping liberally*


>
> And, of course, one is free to make things less than regular, if one
> wishes. The ultimate goal, I think, would be to create poetry or stories
> that could be read multiple ways—something like the Oulipists (cf.
> http://en.wikipedia.org/wiki/Hundred_Thousand_Billion_Poems ). It would
> require a monumental effort on the part of the conlanger, but such languages
> would have the potential to be quite interesting.
>
>
This looks indeed interesting! You could even complicate it more by giving
the various keys a syntactic function as well (_ki_ could be affirmative
sentences, _te_ would be interrogative, _pa_ could indicate doubt, etc.) and
make the lexemes' meanings overlap each other, only with different keys. So
for instance:

Ki mana tilak sini. = The man sees the woman.
Te tilak sini mana. = Does the man love the woman?

With:

mana: man (ki); woman (te)
tilak: see (ki); man (te)
sini: woman (ki); see (te)

I especially like that _mana_ means "man" in affirmative sentences but
"woman" in interrogative ones! :)

You could become especially evil by making the overlaps only partial (so you
don't have true synonyms) ;) .

Add free word order and you have the potential to create true chaos :P .


> == Morpheme Language ==
>
> So many conlangers are so enamored of the fake "morpheme" concept that I'm
> surprised no one's taken it to its natural conclusion. Namely:
>
> kan-di = brick
> kan-na = bird
> kan-to = to be red
> kan-ze = ubiquitous
> tre-di = silly
> tre-na = young
> tre-to = lamp
> tre-ze = corridor
> di-na = man
> to-ze = hat
> di-kan-tre = rabbit
>
> Etc.
>
> To an extent, we've seen this with languages like Ro or Wilkins'
> philosophical language, but they take things *much* too seriously. So, true,
> while a land-mammal-quadruped-small-"big feet"-"long ears" is a rabbit, a
> conlanger has the chance to remake the world in whatever way they see fit.
> Perhaps there is a systematic relationship between bricks, birds, the act of
> being red and the adjectival notion of ubiquity. But that's up to the user
> to figure out (or the creator to imply). There's endless mischief to be had
> with a language like this! (You might need to borrow some self-segregating
> tricks to get it to work, but that's not so bad.) Anyway, while it can be
> random (as it is in the list above), there could be a logic to it that the
> conlanger works into the language. It need not be a realistic way of
> representing the world: It just need not be random (true randomness would
> probably be boring, though it might be fun to generate random languages like
> this, and see what comes o!
>  ut the other end—like looking at clouds or at the patterns that emerge
> from the stucco).
>
>
Yeah, true randomness is very difficult to achieve. We are hard-wired to see
patterns (even where there are none!) and to create patterns. This is the
very hard-wiring I have to go against when I'm working on Maggel. No wonder
it's taking ages to develop that language!


> Anyway, those are just a couple ideas I've had. Each one can be the basic
> schema for thousands and thousands of languages, and there's probably an
> infinite number of other ideas like these two. I mean, there are no limits
> here: the conlanger can do whatever they can imagine.
>
>
True, but at the same people shouldn't go *too* far. Art is *not*
"creativity gone wild". True art usually follows very strict rules, even
when those aren't made clear (even when the author themself doesn't
consciously know them!). This is especially true in conlanging: since
languages are complex systems with my facets, the temptation is big to use
various methods of creation, ideas and/or principles for different areas of
the language. In general I believe the result would be no different than any
other Kitchen Sink language.

Have one overarching principle that permeates the whole artlang, maybe with
one subprinciple as counterpoint, and run with it. Otherwise, you'll risk
creating something that doesn't feel like a whole.
-- 
Christophe Grandsire-Koevoets.

http://christophoronomicon.blogspot.com/
http://www.christophoronomicon.nl/