Re: Is language cyclic?

En réponse à Christopher B Wright <faceloran@...>:

> Read the subject line.
>
> If language isn't cyclic, how can we have Hawai'ian and Georgian at
> the
> same time? How could any language still have cases, or have few cases?
> There has been sufficient time, I think, for languages to settle down
> as
> to basic grammar and syllable style.
>
Indeed, language history (because that's what you meant :)) ) is cyclic, and
proven so. It's not news :)) . In the 19th century grammarians were wondering
how it was possible for today's languages to still have cases while the current
of evolution seemed to indicate that cases tended to disappear with time. They
had all sorts of strange theories about that (most had to do with some
supremacy of a racial group over the other unfortunately :(( ). But now we now
that it's perfectly natural for an isolating language to create cases over
time. Language history is indeed cyclic. The most commonly described cycle is
the agglutinating-inflecting-isolating cycle. Of course, it's an approximation
of what really happens, but it's still interesting to explain. Let's take a
language which happens to be at an inflecting stage (cases, number marks on
nouns, verb tenses and conjugations, etc..). Over time, the endings lose
recognition through sound changes and the simple fact that most of them are
unstressed. They are slowly suppleted, than replaced, by separate words,
prepositions or postpositions. We get an analytic stage. in some cases those
appositions merge with the nouns and verbs, creating an agglutinating stage. In
other cases, they stay separate words which can even be used alone. It's an
isolating stage. But they still finally merge over the generations, and the
agglutinating stage is reached anyway. And now from this stage the affixes of
the agglutinating stage tend to blend with the roots and between each other,
until we reach again an inflecting stage. The circle is full, it all begins
again :)) .

> That leads me to another question. How old do linguists think PIE is?
> (Check how much green stuff is growing on it. Sorry, I had to say
> that.)
>
I've read figures going from 5000 to 8000 years old for the "oldest" form
reconstructed. Needless to say it's not something easy to know, since we don't
have any archeological evidence that would allow a datation. The comparative
method reconstructs a possible ancestor of the language compared, but it cannot
say anything about the age of the reconstruction (indeed, it often reconstructs
structures that didn't ever cohabit but are of different ages and formed and
disappeared at different times).

Christophe.

http://rainbow.conlang.free.fr

Take your life as a movie: do not let anybody else play the leading role.

Reply