Re: Language Change

raccoon@ELKNET.NET wrote:
> 1) Are there any trends or statistical norms with respect to how long it
> takes certain aspects of language to change (morphology, syntax, phonology,
> etc.)? My plan for Dhak is to start with one language that branches out into
> daughter languages, and perhaps even they will branch out. I'm wondering
> what time spans would be realistic for different kinds of changes. (Granted,
> Dhak is spoken on another planet, but I want to make the years there to be
> pretty close to the length of Earth years. However, the speakers tend to
> live longer than humans, so perhaps language change would be slower than it
> is on Earth?)
The speed of linguistic change is quite variable.  Icelandic, for
instance, has hardly changed at all in the past thousand years, while
I've heard of Australian languages which have changed dramatically in
just 20 or so years, to the point where a man who'd left his tribe for
20 years had slight difficulty being understood.

As a general rule, languages which are isolated from other languages
tend to change more slowly, since contact with other languages tends to
cause features and words to be borrowed between them.  Also, there's a
rough correlation (tho only a very general rule) that social changes
tend to be accompanied by linguistic changes.

Also, certain features have a higher chance of changing, e.g., irregular
verbs tend to be regularized.  I have a suspicion that certain
characteristics or combinations are inherently more "stable" or
"unstable", but I don't have any data to back that up.

> 2) I want to start with at least three numbers: unmarked/unnumbered,
> singular, and plural. Unmarked number can refer to either singular or
> plural, but is usually used when the number doesn't really matter. Would it
> make sense for the old singular number to be replaced by the unmarked
> number, which would then become singular? The unmarked number will probably
> be the most common in speech in the parent language.
Depends.  Is the unmarked literally unmarked, in the sense that there is
no affix for it?  If so, it is quite probable.  But, having a marked
"unnumbered" number seems quite unlikely to begin with, and unlikely to
be retained.

> 3) Would it make sense for stative verbs to evolve into a perfect aspect,
> and eventive verbs into imperfect? (Stative meaning verbs which describe a
> pretty much unchanging state, including lots of what English would call
> adjectives; eventive meaning verbs which take place at a certain time and
> describe an event)
Sure.  Things like "He is dead" --> "He has died" seem like very
probably developments to me.

--
"Old linguists never die - they just come to voiceless stops." -
anonymous
http://members.tripod.com/~Nik_Taylor/Conlang/W.html
http://members.tripod.com/~Nik_Taylor/Books.html
ICQ: 18656696
AIM Screen-Name: NikTailor