Re: POLL: lablang/engelang (clarification)

Garrett Jones:
> > Commenting not on the poll but on the definition: 'pot pourri lang'?
> > 'salmagundy lang'? I don't see what all these sorts of conlang
> > have in common. I mean, you're free to state a definition and seek a
> > label for it, but I myself don't perceive any utility in it.
>
> > Another comment: artlangs do not all have naturalistic design goals
> > and do not all emulate natlangs. Naturalistic artlangs form a
> > subgroup of artlangs; they were rather eloquently described by
> > Jesse Bangs in a recent message (of some weeks or months ago).
> > All conlangs to some degree emulate natlangs, because of course
> > natlangs serve to define language itself, as the prototype if not
> > as the limiting case. But what defines artlangs is that their
> > methods and purposes are solely artistic; that is, their raison
> > d'etre is that of art, and the way they set about achieving their
> > goals are guided solely by aesthetics.
>
> hmm, ok. I guess i've never actually known the precise definitions of these
> words in the first place, only had a generally good idea. Has anyone
> actually sat down and wrote out the precise definitions?
Words that evolve through cultural consensus don't really have
precise definitions. 'Auxlang' and 'artlang' were simply obvious
labels for preexisting natural kinds. 'Loglang' was coined to
define a hypothetical class of conlangs based on very much the
same principles as the Loglans, but not restricted to the Loglans
(Loglans being Loglan and its descendants/versions).
In the case of 'Engelang', I perceived another natural kind
(defined by me in previous messages in these threads), and
after much discussion eventually managed to articulate a
reasonably explicit definition. I initially applied the definition
to the term 'loglang', but found that popular consensus held to
the narrower definition of 'loglang' that I have given above,
so this gave rise to 'engelang', which follows the clipping-compound
pattern of the other terms and derives from the obvious and optimal
descriptive phrase 'engineered language'.

> I guess a better definition would be this:
>
> [language term] n. a language with one or more experimental features
> (optional: that do not occur in natural languages).
>
> the reason the strange languages would fit in this category is that they
> would have features that don't occur in natural languages.
In that case, 'does not occur in natural languages' is an integral
rather than optional part of the definition.

But the features 'experimental' and 'nonnatural' seem to me not
to define a natural class. Some naturalistic conlangs enjoy
typologically interesting experiments with natlang features.
Some conlangs with nonnatlang features are not in any meaningful
way engaged in experimantation. And I'm not conscious of any
marked tendency for the two properties to go hand in hand.

Please understand me to not be endlessly carping at your proposals;
rather, the taxonomy of conlangs is simply a matter of much interest
to me, and discussion helps shape one's own understanding.

> > Lastly, with regard to the question "'lablang' or 'engelang'", I
> > say "certainly not 'engelang'", for 'engelang' means an 'engineered
> > language', one with explicit design goals such that the degree
> > of success in achieving those goals is objectively assessable.
>
> to get a better idea of what exactly is meant by 'engelang', which languages
> on this list are categorized that way?
All those (quite a few) mentioned in the recent thread on
self-segmentation, that being a typical engelang goal. (Those
listed in Mike S's original post, plus Liva, Livagian, Dublex
& maybe others I forget.) AllNoun is a well-known engelang, though
its creator is no longer on the list. Likewise Yiklamu. Of the
conlangs currently represented on the list, I'm not sure,
because I'm rather out of touch with it -- can't keep up with
the rate of traffic & of newcomers, welcome though they are --
but I did get the impression that Henrik Theiling & Shreyas
Sampat do relatively engelangy stuff (I hope I'm remembering
the right people; if not, delete the last clause).

Besides self-segmentation, common engelang goals are:
* no grammatical ambiguity
* logical explicitness & no logical ambiguity
* no unwarranted complexity or irregularity
* concision
* compactness of grammar -- minimal rules
* minimal inventory of roots
* some variety of systematic form--meaning correspondence across
roots (so that roots with similar sound have similar meaning,
in some systematic way)

> and if you have a language that's an
> engelang, what's the website?
It is a matter of long-standing shame to me that my conlang doesn't
yet have a website. I have first to get the documentation into
publishable form, and then to face up to getting to grips with the
technical logistics.

--And.