Hallo conlangers!

On 26/11/2016 21:35 Melroch wrote:

> Artlanging has always felt to me like a discovery process rather than an
> invention or design process.

Same to me, to some degree at least. When I work on Old Albic, it almost
feels as if I was exploring a lost reality rather than as if I was
making something new. Of course, I know that the language is really just
my invention, but I sometimes seriously ask myself whether I am not up
to something more than that.

Sometimes I make a discovery that confirms what I had decided on before.
An example is the West Hesperic (Hesperic is the family Old Albic
belongs to, and Old Albic belongs to the western branch of it) definite
article _a_. This started as a decision about Old Albic which just felt
right to me, but could have been something else. Later, I became aware
of the phenomenon which I call "alpha mobile": in substratum loanwords
in western European languages, there are sometimes doublets involving an
initial *a, such as German _Amsel_ 'blackbird' < *amasVla vs. Latin
_merula_ < *masVla. A possible explanation is that the source language
had a definite article *a - and that's just what I had decided before
that my family - which is a "re-creation" of that linguistic layer -
would have it!

> To the extent that I have made decisions they
> have not been of the "I'm going to make an ergative language" kind. Even
> though most projects have started out as experiments with some feature or
> other those that have survived have invariably taken on a life of their own
> -- even on the one occasion I did conlanging for hire on a small scale.
> Rather I have decided whether the stuff I'd been writing or mulling over
> felt right or not.

Yep. It is a matter of whether the stuff I am doing feels right or not.
Most of my conlangs are "lostlangs", i.e. languages that represent
continuations of extinct or hypothetical languages in a version of the
consensus reality with some extra languages surviving. This involves
contemplating about what these languages would look and sound like, what
fits into the linguistic landscape of the relevant location, etc. I ask
myself questions such as "What could the substratum that caused the
typological aberrancy of Insular Celtic have looked like?" (Old Albic)
or "What would an IE language of the Volga region, intermediate between
Slavic and Iranian and influenced by Uralic, look like?" (Zamanic), and
try to find answers that feel "correct".

> Discovering the right shape of a conlang or story has
> always been a slow and somewhat painful process for me. An initial period
> of great creativity followed by long periods of hesitation and creative
> draught punctuated by bursts of creativity.

I feel just the same. When I begin a project, I am full of ideas, but
later there is a lot of hesitation and contemplation and slow progress.

> It has not been periods of
> disinterest. I've been shaking the creative ketchup bottle more or less
> constantly but only occasionally has anything which felt right come out of
> it.

Concurred once again. It is a matter of progress in bursts. Every now
and then I come up with a new idea; but I can't force them out. This is
especially true about lexicon. The phonology and grammar of a conlang
are usually fixed quite early, but words that feel right and have
interesting etymologies take forever to make.

> As a consequence it has been hard to write any finished descriptions,
> let alone texts. I have keenly felt that much of what I had was 'made up'
> scaffolding rather than the 'reality' which was still to be discovered. As
> a consequence I have not felt very much need to work the old abandoned
> stuff into the new once my creative juices actually have flown. On the
> other hand those 'truths' about my conlangs which I have discovered have
> tended to stick around.

There is a small number of obsolete Old Albic texts, mostly relay
torches, in circulation. The oldest show a different morphology; most
show the "final" morphology that is still valid, but are full of words
that I now feel obsolete. Even the latest torch (of Relay 23) contains
at least one word which I now know to have been formed wrongly. Some of
these "mistakes" may find a new life as dialectal variations, earlier or
later stages of the language, or whatever.

Most of my conlangs are also in a disorganized state. I find it hard to
put anything in a finished state. The files I have are partly out of
date, and many things I have found out haven't even been written down.
At least, that means that I don't have much written stuff to correct
when I hit upon a revision.

> The exception is Classical Sohlob and Chidilib. The
> latter started as a rewrite of the former but I let both stay around as
> closely related sister languages, and CS has since lost some of its
> roghness, nudging closer to its true distinctive character. Since I work
> diachronically I have sometimes discovered new historical stages of my
> langs. The most thrilling to date was when I discovered that what I had
> thought was the Sohlob protolanguage in reality only was an early ancestor
> of one side of the family. As a consequence I discovered that much of what
> had felt wrong with Classical Sohlob was because it was a close cousin of
> Chidilib rather than a sister.

I also work diachronically, and sometimes make similar discoveries.
Albic started as a small stand-alone family, though with a distant
relationship to IE, and later I found out that it was just one branch of
a larger family, Hesperic. Some features of old versions of Old Albic
and Proto-Albic have already resurfaced in Proto-Hesperic, but
Proto-Hesperic itself is a moving target about which I still find out
new things. And sometimes I make unexpected discoveries that change the
picture substantially.

> Then I found that CS actually was kind of a
> conlang within the conworld -- a literary language which was either an
> early form of Chidilib influenced by the ancestor of Hïlïb or the other way
> around. It did not make everything I had previously written about CS right,
> but it did make some of its rough bits fit into the picture. It made things
> simpler in that some of the inconsistencies of CS could be explained by its
> mixed, artificial origin. It made things more complicated in that the real
> common ancestor was and largely still is to be discovered, but a corner of
> the puzzle had fallen in place.

The examples of Old Albic I have presented here and elsewhere are mostly
actually from a "classical" version of the language that is based on a
particular dialect but in some ways artificial and incorporating a few
things from other dialects. Kind of an "Albic Sanskrit". Indeed, I
wonder myself sometimes how artificial this Classical Old Albic is
within its own conworld. But I try to find diachronic solutions to
problems I run into, and take inworld artificiality only as a last
resort when everything else fails.

--
... 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