Hallo conlangers!

On Monday 27 August 2012 15:08:43 R A Brown wrote:

> I've changed the subject line as we seem to have moved away
> from Brithenig and gone back The Box.

Fair.

>       (Has this got a name,
> like Gnoli triangle has?)

One could call it "Mil's Box".  I don't know the real name of
"Mil", though.  I hate it when people don't disclose their real
name in situations where it would do no harm to do so.
(I understand well that people prefer to stay anonymous when
they speak out against injustice and would risk their life or
freedom if they disclosed their identity, but in an innocent
discussion about conlang classification systems, I consider
this unwarranted.)

> On 27/08/2012 03:51, Jim Henry wrote:
> > On Sun, Aug 26, 2012 at 2:58 PM, Padraic Brown
> [...]
> >> Then perhaps a different term would work better. I
> >> don't really want to get deeper into thát line of
> >> argument! I didn't take pragmatic to be so narrowly
> >> defined
> 
> Yes, I agree that would be a very narrow definition of
> 'pragmatism'.  I understand "pragmatism" to mean in this
> context "concern for the practical rather than for theories
> or ideals" [Chambers English Dictionary].

Sure.  A "pragmatic" conlang would be one where practical
usability is a design goal.  A good auxlang should thus be
"pragmatic", but practical usability is of course also an issue
with many other kinds of conlangs, including - as I laid out in
the other post - fictional languages!

> I would question very much whether Ro, for example, could be
> so described.

Indeed not!  I consider it, like all "philosophical" languages,
unwieldy and impractical.  There really is nothing to be won in
terms of usability with a taxonomic vocabulary.

One could for instance construct a language in which 'arm',
'foreleg', 'wing (of a bird or bat)' and 'pectoral fin' are the
same word, as they a homologous to each other, but that would
obscure the different purposes of these limbs, and having a
*different* word for 'wing of an insect' would be pretty
inconvenient.
 
> > Yes, on further thought I'm not sure art and pragmatism
> > are in opposition, such that increasing one
> > automatically decreases the other (though they may be
> > negatively correlated).
> 
> I do not think they are in opposition.  I would expect
> 'pragmatism ~ idealism'; but then to me the opposite of
> 'abstract' is 'concrete'.  In what sense are Ro and Volapük
> _abstract_?

I think the word _abstract_ does not apply well to conlangs
that aren't artlangs; it is a term of art theory after all.
I'd say that "abstract" is a meaningful category only in
artlangs: an "abstract artlang" is one that does not try to
represent a language of a fictional reality.  Calling an
auxlang or a loglang "abstract" is as meaningless as calling
a machine "abstract".
 
> As for "artism", it is a neologism and IMO requires
> definition.  "Artist" means someone who practises or is
> skilled in an arts, particularly one of the fine arts.  So
> shouldn't "artism" mean the practice of art?

Probably.
 
> I Google "artism" and find it seems to be used as a
> portmanteau word to mean "art of autism", i.e. works of art
> produced by people with various degrees of autism.  I assume
> that is not the meaning here!

Certainly not ;)
 
> [snip]
> 
> > Pragmatism is probably also negatively correlated with
> > naturalism --
> 
> Possibly - tho I would contrast naturalism and artificiality.

Yes.  "Artificial" is a good antonym for "naturalistic" (in the
artlang sense).  And I have heard it in that sense before.
I remember an e-mail from Andrew Smith in which he criticized
an artlang as "too artificial" - he meant "not naturalistic
enough".
 
> > I think most of the above-mentioned languages are too
> > regular to be naturalistic, because their creators want
> > them to be as easy to learn as is consistent with their
> > other design goals.  But I don't think one can situate
> > pragmatism at the opposite end of a spectrum from
> > naturalism, either.
> 
> Indeed not.

Exactly.
 
> > Naturalistic is the opposite of not naturalistic, and
> > nothing else.  (Some promoters of Interlingua and similar
> > auxalngs call the opposite of auxlang-naturalism
> > "schematic", but I think that's pejorative and probably
> > to be avoided for a classifying model like this.)
> 
> Indeed not.  My one-time sojourn in Auxlandia has led me to
> be wary of all such labeling.  While a_posteriori, a-priori
> and 'mixed' can IME be fairly safely given,

Sure.

>        when we come to
> other auxlang labeling we are liable to fan flames.

Yep.  

> > Pragmatic is the opposite of... not pragmatic?
> > Idealistic?  I think there's a word for it, but it's not
> > "artistic".
> 
> It is _not_ artistic.  I have opted for 'idealistic' above.

An apt suggestion.
 
> > And as Ray pointed out, a priori - a posteriori isn't a
> > single variable either; there are conlangs which are a
> > priori in vocabulary and a posteriori in grammar, and
> > others vice versa, and lumping them together in the
> > middle of an a priori - a posteriori spectrum can be
> > misleading.  We need at least two variables there, if
> > we're going to use a priori - a posteriori specta at
> > all.
> 
> Yep - I think applying the a_posteriori ~ a-priori labels to
> artlangs & engelangs is not helpful at best and, indeed, is
> often misleading.

It is *technically* always possible, though in some cases, the
origins of the words are obscured to such a degree that they
are no longer obvious to the learner (as in Volapük or Lojban);
but it most useful with auxlangs.  With artlangs, this
distinction is secondary to the "naturalistic/artificial"
distinction, and with engelangs, secondary to the design goals
of the languages.
 
> > And yet, to squeeze all the variables by which conlangs
> > can vary into a three-dimensional visualizable cube
> > model, rather than a many-dimensional unvisualizable
> > hypercube, maybe it's useful to conflate multiple
> > variables into one if they're sufficiently correlated.
> > That seems to be what Mil did with the cube model under
> > discussion.
> 
> As I wrote before, I have experimented I the past with 3d
> models of classification but not found any satisfactory.
> Clearly this one needs rethinking as to the definition of
> its axes.

Certainly, an opposition between "artism" and "pragmatism"
makes no sense.
 
> For the past 14 years or so the Gnoli triangle has served us
> well as model.  If one wants to go into a more detailed
> classification then I think it has to be in systems such as:
> http://steen.free.fr/classificatie.html

The Gnoli Triangle has indeed served us well!  Of course,
subdivisions of the three main types of conlangs are useful,
but within the different realms, very different distinctions!

Let me repeat the system I posted on August 16 in updated form,
just as a contribution to this discussion and not as a "final
word":

1.1. Naturalistic
1.1.1. A priori
1.1.2. Mixed
1.1.3. A posteriori
1.2. Non-naturalistic
1.2.1. Non-human languages
1.2.2. Abstract artlangs
2. Auxlangs
2.1. A priori
2.1.1. Taxonomic
2.1.2. Non-taxonomic
2.2. Mixed
2.3. A posteriori
2.3.1. Autonomous
2.3.2. Naturalistic
3. Engelangs
3.1. Philosophical
3.2. Logical
3.3. Experimental

Of course, these categories are all more or less fuzzy, and may
overlap!

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1
FREE PUSSY RIOT