Hallo!

On Wed, 13 May 2009 10:52:29 +0200, Benct Philip Jonsson wrote:

> On 2009-05-12 Daniel Bowman wrote:
> > While I don't ever think my language will be "complete" 
> > in the sense that I'll stop developing it, 
> 
> It has always been my feeling that a conlang which
> one can not or will not develop any further is a
> dead conlang, or to put it in positive terms: if
> 'alive' for a natlang means 'still used for
> communication'[^1] then 'alive' for a conlang
> means 'still tinkered with'. 

Yes.  A conlang is never really "finished", and be it just because
there are sooooo many things to be named in either this world or
the fictional world the conlang is spoken.  A conlang on which its
author no longer works is indeed essentially a dead conlang.

>      To my mind at least
> conlanging is -- in spite of its solipsism -- a
> performing art. I'm not nearly as thrilled by
> reading a grammar of a conlang as of hearing the
> stories of how the conlanger discovered what the
> grammar looks like, and why. That's why to me the
> appropriate term for the art must be 'language
> construction', and the true sense of 'constructed
> language is 'language which is being constructed'
> rather than 'language which has been constructed',
> since the thrill is in the journey, not in the
> destination, or even in the arrival.

Indeed, "the road is the destination".  The fun in conlanging lies
in *making* it, and as long as you are *using* it, you are still
making it, as you will probably run into blank spots in the lexicon
or the grammar even if your language is highly developed.  Writing
down a grammar sketch and a few hundred words can be done in a few
days, but that will only be the *foundation* of the conlang: you
*will* run into *lots* of blank spots when you start using your
conlang.

>      This is not 
> to say that it is not a good thing to have some
> idea of the destination, but there should be room
> for changing the destination, or rather the
> destination is best not too sharply defined, and
> one must always be open to the possibility that
> one was mistaken about the destination, or about
> its true identity. (I guess I'w talking like a
> true Maháyánist here!)

I can only confirm that, looking at the history of my Albic project.
It started as Nur-ellen, a descendant of Tolkien's Sindarin; but it
soon developed into a completely different direction and is now no
longer related to Tolkien's languages (a few ideas, such as the
active/stative alignment with degree-of-volition marking, have
survived, but it tells a lot that the surviving features were mostly
my own invention even in early Nur-ellen).

>       Surely a language 
> continually worked on through decades will
> eventually change or expand less and less, but I
> think that if it still holds any interest at all
> one will continue to chip on it here and there at
> least occasionally.  What we can glean of
> Tolkien's case history seems to bear this out.
> What do you living practitioners who have worked
> on the same languages for decades say?

I haven't worked on the same language for decades yet, but even
when looking back onto the nine years I have been working on what
is now known as Albic, I recognize that the changes have become
less and less, with one total start-over (when I decided to sever
the diachronic link to Sindarin) and several major revisions in
the early years, while the grammar has been quite stable (but
constantly becoming richer and more detailed) since about 2005
or so.  But I still expect things to change in the future: there
is still much to be discovered in the grammar and the diachronic
development, and the lexicon badly needs more exploration.  And
then there are daughter and sister languages of Old Albic to be
discovered ...

> > I was thinking that one major milestone would be
> > the ability to define the language in its own
> > terms. 
> 
> I think a more useful, and truthful, term when
> talking about conlangs would be _mature_, in the
> sense of being capable of performing at least a
> semblance of the communicative functions of a
> natlang, most notably to be used to compose texts
> -- on any subject --, and thus also at least
> potentially being used for conversation, or
> possessing most of the characteristics of a
> 'natlang', not only in terms of vocabulary but
> also in terms of grammatical categories, or
> historical depth if you are of that bent.

Right.  I am aiming at just this kind of maturity with Old Albic;
perhaps some of its daughters will also be developed to maturity
later, but Old Albic goes first, and I still feel that is has come
quite close to maturity but not fully reached it yet.

>       This has 
> nothing to do with reaching a 'final form'; IME
> some conlangs achive maturity almost at once --
> even coming out of the egg virtually mature in
> some cases --,

One example of an "instantly mature" conlang is perhaps Germanech,
in which I could start composing whatever text I liked just right off
in the beginning, as I could make up words "on the fly" just by
applying the GMP to Latin words and all that.  Yet, Germanech lies
quite neglected, and I haven't done much work on it or with it for
a few years.  It is one of those side projects that are waiting
indefinitely in my mind to be picked up again.

>       while others stubbornly resist 
> achieving maturity, but are very much alive
> because they keep attracting their creators'
> attention and interest.  However it is probably
> true that the more stubborn perpetual fledglings
> may be more vital in the sense of maintaining
> interest and attention: there is indeed a threat
> of death in maturity, unless the language enters a
> phase of being used expressively, since using a
> language for communication inevitably entails
> tinkering with it; this is true of natlangs as
> well!

Yes.  Germanech is almost dead because it is essentially *finished*;
there is not much left to explore (other than running yet more words
through the GMP, but that is more of an automatic procedure rather
than a creative act of exploration).  What is left to do on it is
essentially bookkeeping work.

The Albic family offers much, much more uncharted territory to
explore (and then there is the notion of "Hesperic" languages, a
larger family of which Albic is just a branch, and which I am not
sure about whether those languages will exist at all - but I already
have some ideas about them, and will probably do them later); if I
compare Old Albic to Germanech, I must say that Germanech is more
*mature*, but Old Albic is definitely more *alive*!

> > This means being capable of writing the grammar
> > down actually in the language in question, and
> > also being able to define every word in the
> > language in terms of other words from the same
> > language.
> 
> I guess I could do that with Mærik, but then Mærik
> is possibly the deadest of all my conlangs: coming
> almost mature out of the egg in a flash of
> inspiration, and assigned the rôle of a Relay
> language from the start.

I could write a grammar of Germanech in Germanech with few
difficulties; however, I am much more interested in writing a
grammar of Old Albic in Old Albic - but for that, Old Albic
is not yet mature enough.

>       Yet there are big white 
> areas on the map, particularly in the area of of
> syntax, and they periodically cry to be unveiled,
> or spontaneously unveil themselves, and the
> largely unanswered question of *who* spoke that
> language still makes itself heard whenever I look
> at the language, and that too has repercussions on
> the language

Indeed, to feel really "alive" a conlang needs some sort of
concultural background.  This is fairly well-developed for
Old Albic, though there is still much to explore yet in that
domain; it is more sketchy with Germanech, but even there I
know who speaks the language.

>       -- particularly the vocabulary, which 
> was created  with computational assitance, meaning
> that I frequently feel that a given word-meaning
> collocation is not correct.

I do not use computer-generated words at all in Old Albic for
just that reason: the words have to feel *right*, and that is
only possible, it seems to me, if I create the words manually
in a creative process.

>       (Which however does 
> all too seldom mean that I know which collocation
> would be correct: that -- the central part for
> Tolkien -- is the hardest part for me, while I can
> almost effertlessly paint canvases of (historical)
> phonology, and with grammar falling somewhat in
> between.  The problem is that my spontaneous 
> vocable creations tend to be too alike, lacking in
> indviduality, more expressive of phonotactics than
> of wordness.)

Same to me!  I can easily make up interesting sound changes, and
grammar is also rather easy, but finding words that sound right
is a problem, and it is that aspect that is most lacking in Old
Albic.  (With Germanech, this problem is much smaller because the
words fall out ready-to-use, modulo some minor semantic shifts,
from the Grand Master Plan.)

>       Also Mærik has, 
> uncharacteristically for me, almost no
> con-history: almost the only thing I know is that
> here has been i-umlaut and apocope, but that's
> only because of the synchronic reflexes of those
> processes are visible.

I feel that a conlang, in order to feel alive and "real", needs
an elaborate con-history.  With Germanech, the diachronic
development from Latin codified in the GMP provides such a
history; with Albic, I am working on Old Albic and its ancestor
Proto-Albic (intrafictionally, Old Albic is the oldest attested
Albic language, and Proto-Albic the reconstructed common ancestor
of all Albic languages), and even on the prehistory of Proto-Albic
all the way back to Proto-Europic, the common ancestor of Albic and
Indo-European which I reconstruct internally from PIE (this involves
some actually comparative reconstruction work, as I am reluctant to
change Albic more than necessary in order to make it fit with
Indo-European).

> Granted the writing of a Mærik grammar would have
> to be _in character_: a medieval monk working
> through his Donat and giving the Mærik equivalents
> of its categories, with native speaker intuition
> breaking the mold only occasionaly.  I wonder what
> he'd make of inclusive and exclusive plural?  What
> would he call them in Scholastic Latin?  I guess
> he'd seize on the fact that the inclusive sort of
> refers to both the first and second person and
> call it _pluralis utrarumque personarum_.
> In fact I doubt he'd write a grammar *in Mærik*.

He'd probably write it in Latin.

What regards the Old Albic grammar I am intending to write in
Old Albic itself, I am going to write it in character: what I
already know is that there was a grammarian in the classical era
of the Elvish civilization who indeed wrote a grammar of his
native language.

... brought to you by the Weeping Elf