arbitrary case-marking?

[Noel Rude]

Greetings,

        When languages die--at least I see this here--there is the tendancy to
seize upon certain phonological and morphological features of the old
system as markers of "Indianness".  These languages have, for example, a
k - q as well as plain - ejective distinction, but younger speakers
often have only a k - k' distinction and will just arbitrarily throw in
the ejective k' (in place of any of the four k-like sounds and even in
place of tL and tL') for its cultural effect.  I have seen the same
thing with the ergative noun case, with -n at m arbitrarily thrown in
(suffixed to any noun) just for its effect.

        Obviously this is not the same thing as is going on in the vibrantly
alive language Tamil.  Or is it?  One wonders--where a case marking
system is breaking down (as in some German dialects?)--do we see a
similar tendancy?  Where some speakers no longer control a system, might
they try to exploit it for some other effect?  Perhaps this way (over
time) a structural feature could even change function, thus effecting a
"crazy" historical change.  Such might even wash back over speakers for
whom the system was not breaking down.

        Noel