Re: creating words (was Re: "Language Creation" in your conlang)

Hallo!

On Thu, 13 Nov 2003 18:48:05 -0800,
JS Bangs <jaspax@...> wrote:

> Jörg Rhiemeier sikyal:
>
> > > > [trouble finding words for my conlangs]
> > >
> > > This is also my problem. Yivrian has been around for nearly a decade now,
> > > and is yet to breach 1000 vocabulary items.
> >
> > And yet you managed to compose a creation myth in it.  I was deeply
> > impressed when I saw it!
>
> Thank you! Writing texts, of course, has been my other major impetus to
> create words. [...]
Oh yes, it is!

> > Finding roots is the most difficult part of word-creation, I think.
> > I have ample ideas for interesting etymologies that fail to materialize
> > because I simply can't settle on those goddamn *roots*!  I guess that
> > once I have a list of several hundred lexical roots covering most
> > realms of discourse, things will start moving more swiftly.
>
> Indeed. What are your sound-changes like?
The sound changes are not entirely fixed yet, but my idea about them
is that Nur-ellen has undergone similar changes as Welsh or Brithenig,
while Macaronesian has developed into another direction I also have
quite good ideas about.  So I can tell what a Nur-ellen or Macaronesian
word looks like when I have the Proto-Elvish form.

>      It used to be that the hardest
> part was finding words that fit with the Yivrian sound, but oddly enough
> settling on my sound changes made this easier, because I could make up
> just about any root in the world and have it turn out okay once I put it
> through the wringer :).
Yes.  I know what a Proto-Elvish root looks like, and what happens
to it in the history of the two main daughter languages, so the
phonological
"styles" of the languages are already rather well-defined.

> > My conlanging also involves historical development.  I am currently
> > working on Proto-Elvish, from which the individual Elvish languages
> > will be derived.
>
> I did my first language backwards: Yivrian existed long before
> Proto-Yivril, and I mangled the language backwards to find the
> proto-language. Doing it forwards (as I will with future languages) leads
> to more consistency, but can be more time-consuming.
Yes.  Actually, I am working on both the protolanguage and the daughter
languages in parallel, using a kind of "pipelining" technique:
when I know what something looks like in the protolanguage, I explore
what it has changed into in the daughter languages.  For example,
I already know the personal endings of the Proto-Elvish verb,
and have already determined what they look like in Nur-ellen,
while the tense/aspect/mood markers haven't stabilized yet and
I still don't have too many roots to try out the forms on.

And there is also some degree of working backwards involved.
Sometimes, I reconstruct a root such that it yields the Nur-ellen
(or Macaronesian) word I want; and determining the sound changes
implies answering the question, "How must the sounds change such
that the modern language sounds like that?".

And to make things even more complex, I fancy that Proto-Elvish
is remotely related to Indo-European, and explore the common
ancestor of both, which involves some rather speculative
internal reconstruction work on Proto-Indo-European!
I found the monumental book by T.V. Gamkrelidze and V.V. Ivanov,
_Indo-European and the Indo-Europeans_, very useful for this
purpose.  They reconstruct an early stage of PIE that has many
typological similarities with my Proto-Elvish language.

> What's the time-depth between Proto-Elvish and the projected daughter
> languages?
About 3000 or 3500 years.  Proto-Elvish was spoken in southern Britain
in the Bronze Age, and I plan to follow the evolution of the languages
all the way down into modern times.

On Fri, 14 Nov 2003 12:29:16 -0800,
JS Bangs <jaspax@...> wrote:

> Andreas Johansson sikyal:
>
> > [...]
> >
> > I've several times attempted to do language families the right way around,
> > but
> > it has always failed because it requires making the proto-language
> > vaguely "complete" first, whereas my real interest is always in the daughter
> > languages.
I know what you are talking about.

> Emoi de topw. Working "forwards" is rather more boring and less fun that
> doing things "horizontally" or backwards as you mentioned.
Yes.  See my comment on my "pipelined" approach above.

>      Even when I
> work forwards, however, I don't feel the need to make the protolang
> particularly complete. All I have for PY is a list of roots, a vague idea
> of morphology and the barest hint of syntax.
Yes.  The protolanguage needn't be much more than a skeleton of a
language.  Leaving it at that also makes sense intrafictionally,
at least in the case of my Elvish where the protolanguage is not
attested in writing, and thus only what can be reconstructed
is known about it.

>      Having so *little* material
> actually helps make the daughter languages interesting--in order to have
> anything remotely complete, I have to elaborate a lot, and I can always be
> sure to elaborate in a different direction than the other existing langs.
I can guess that.  Determining too many details in the protolanguage
may mean that the daughter languages come out all to similar.

> > What I have pulled off is "horizontal expansion" - from Tairezazh I've
> > figured
> > out the basics of the ancestral language, Classical Klaish, and from that
> > derived Tairezazh's sisters; Steienzh, Telendlest and Searixina, in order of
> > increasing sketchiness.
>
> As I have developed Tzingrizhil, Praci, and Keluril in increasing order of
> sketchiness.
Besides Nur-ellen and Macaronesian, I currently have the "sketchlangs"
Caledonian and Iverinian in the pipeline, and the fictional background
of my language family allows me to add about as many more languages
as I like to.  The Elves had set up *many* tradeposts in their heyday,
from Norway down to Dakar (or even farther down) and from Newfoundland
to the northeastern tip of Brazil, and while most of them were abandoned
or absorbed by the local population when the Elvish ascendancy fell,
a yet unspecified number of them may have survived.

Greetings,

Jörg.

Reply