Re: Conlang labels (wasR: Futurese, Chinese, Hz of NatLangs, etc.)

And Rosta writes:
 > Tim May:
 > > Anyway, I mentioned in my post that it might be worthwhile to create
 > > two parallel triangles to describe a conlang (parallel in a semantic,
 > > rather than geometric sense - although one could also say that they're
 > > seperate because their subjects are orthogonal, which sets up a nice
 > > contradiction).  Anyway, one of these would describe purpose, the
 > > other features - or rather, aims.
 >
 > It is useful to make a distinction between (A) a system that classifies
 > conlangs according to their properties, so as to insightfully
 > capture the nature of the conlang and of its relation to other
 > conlangs, and (B) a system that classifies conlangs into the
 > natural macroclasses that conlangs tend to cluster into. Both
 > are interesting and worthwhile, but while (my revision of) the
 > Gnoli triangle usefully accomplishes (B), you are aiming to do
 > (A), and my point is that to do (A) is not to discard (B).
 >
If we're talking about classification systems in general, I agree.
Most conlangs are classifiable as loglangs, auxlangs or artlangs.  But
if we're using a triangle, I find these poles suboptimal.  A triangle
should ideally encompass all possible conlangs within its area, and
it's my belief that there are conlangs which cannot be described
through any combination of auxlang, loglang, artlang (unless you
stretch the definitions of these such that they become
counterintuitive).

 > > The latter I haven't worked out
 > > yet, but I've come up with two possible models for the first.
 > >
 > >
 > >                         Experimental
 > >                              ^
 > >                             / \
 > >                            /   \
 > >                           /     \
 > >                          /       \
 > >                         /_________\
 > >                  Practical       Artistic
 > >
 > > Experimental conlangs are those created to explore some feature or
 > > other.  Most loglangs would tend towards this point, as would
 > > ideological languages, and any language created simply to see how
 > > certain linguistic features might interrelate.
 > >
 > > Practical conlangs are those intended to be used in the real world,
 > > including IAL's, machine translation interlinguas, and, well not a lot
 > > else, but it provides a place for certain things not covered by a
 > > strict definition of "auxlang".
 > >
 > > Artistic conlangs are those created for aesthetic reasons.  I think
 > > we're all familiar with the general concept of an artlang.
 >
 > Speaking as an engelanger and loglanger, this classification doesn't
 > resonante with me. If anything, my purpose (and that of most other
 > engelangers) would fall at the Artistic apex -- the conlang is an
 > end in itself, just as someone might design a new can-opener simply
 > for the sake of designing a new can opener.
 >
 > Accordingly, 'Artistic' needs to be replaced by 'Own sake' or the
 > like.
 >
Well, I'm not convinced.  If you design a can opener, you hope to
improve on existing models on aesthetic or utilitarian grounds, or
perhaps if you're an inventor to experiment with a radical innovation
and see how it works.

Your language may be "for its own sake", but if you're anything like
me, this can really be seen in terms of providing you with a) an
aesthetically pleasing conlang, and b) an opportunity to experiment
with various linguistic features.  Perhaps your motivation is
radically different to mine, but I find it hard to imagine anyone
producing a personal conlang for reasons other than a combination of
aesthetics and personal research.  (Note that these may apply equally
to the process of creation as to the finished product, if that's
causing any misunderstanding).


 > > What these three poles are to be called, I don't know.  Lablang,
 > > Praclang and Artlang are the best I can think of at present.  (I'll
 > > post seperately on what I think of the various
 > > category-name-suggestions floating around.)
 >
 > It can't be 'artlang', because that term is already in use with
 > another meaning. 'Lablang' and 'praclang' work.
 >
May I understand what you consider to be the essential definition of
"artlang"?  Perhaps I have misunderstood how the term is generally
applied.

 > > I think that's a pretty good triangle for purpose, but as I said
 > > before, "artlang" or "artistic" covers a lot of ground.  Some might
 > > therefore prefer
 > >
 > >
 > >                           Abstract
 > >                              ^
 > >                             / \
 > >                            /   \
 > >                           /     \
 > >                          /       \
 > >                         /_________\
 > >                  Practical       Fictional
 > >
 > > which shifts artlangs which are not part of a fictional construct
 > > (with imaginary speakers, etc) into the same pole as Experimental and
 > > calls it Abstract.  I think this makes some sense, although I myself
 > > prefer the first variant.  It's a matter of taste, I guess.
 >
 > The first triangle distinguishes, according to their purpose, two types
 > of purposeful conlang from the purposeless (own sake) conlang. The second
 > triangle distinguishes from practical conlangs two types of nonpractical
 > conlang, according to the independent criterion of fictionality. The
 > first triangle is therefore more coherent.
 >
Well, as I said, I do prefer the first triangle (see also below).  But
I'm not sure that I agree with you that the distinction between
experimentation and aesthetics is more a matter of purpose than the
distinction between abstraction and fictionality.  Fictional languages
must in some sense aim to emulate natlangs, and their purpose is to
form part of a work of art that extends beyond the language.  Abstract
languages, whether indended for aesthetic enjoyment, a learning
exercise, or both, have no purpose beyond themselves.

Possibly "purpose" is not the best word to describe the differences
between these apices, but I still feel that we're comparing like with
like on both these triangles to a greater extent than on the Gnoli
triangle.

Interestingly, it's also possible to divide the second into languages
without speakers (abstract) and those with real (practical) or
imaginary (fictional) speakers.

I've decided, now, that the reason I dislike this triangle is that
fictionality is more or less a binary quality.  A language can be
fictional and also practical or abstract, perhaps, but it can't really
be "somewhat fictional" - you've either defined fictional speakers or
not.  This isn't true of practicality, as many people have made
conlangs designed for optimal usefulness as an IAL without seriously
intending such use.  Aesthetics and experimentality will have some
impact on almost any conlang.


 > > I don't suggest that these make the categories Artlang, Loglang and
 > > Auxlang obselete, by any means, I just think they capture the spread
 > > of possible conlang purposes rather better.  Few languages will be
 > > well described by any one of these labels, but this is desirable for
 > > this purpose, as it means they'll be spread out more across the
 > > triangle.
 >
 > You are right. The terms 'conlang', 'artlang', 'auxlang', 'loglang'
 > and 'engelang' all have established definitions, so cannot be
 > rendered obsolete. But there is still room for a full descriptive
 > cross-classificatory system, though I don't think special names are
 > required for all categories defined by such a system.
 >
 > --And.

I agree, particularly as the poles of the triangle are intended less
as categories than as reference points.


Incidentally, it struck me that if the extent to which a conlang is
described by a pole is expressed as the distance from that pole, these
shouldn't really be triangles at all.  They should be... there must be
a name for this shape, but I don't know it.  The points of an
equilateral triangle, with each two joined by a circular arc with
origin at the third point.  An interesting shape - constant diameter
without constant radius.  Thing's'll roll quite smoothly on them.  But
as they're a little tricky to draw in ASCII art, I'll stick with
triangles in mail. :)

Replies