According to sociolinguistically minded
scholars like William Labov (who has studied many an ongoing sound
change and measured its progress), sound change is typically initiated
as a gradual transformation of the phonetic features of a phoneme without regard
to lexical or grammatical factors (thus displaying mechanical "Neogrammarian"
regularity). It usually occurs below the level of social awareness (and so
is "imperceptible"). In its later stages, however, sound change often develops a
social significance and spreads by lexical and social diffusion -- word by
word and speaker by speaker, prone to lexical, grammatical and socio-stylistic
conditioning. It is no longer imperceptible, may affect salient distinctive
features simultaneously and consist in abrupt phonemic replacement. Lexical
diffusion often fails to affects all the target words and leaves messy
exceptions even in the long run.
What you call shibbolethisation is possible
at that late stage. What I mean is that no-one ever starts a sound change
deliberately, but a change in full swing creates forms that may be utilised as a
means of social bonding and regarded as prestigious or deprecated. There may
have been a time when pre-Proto-Indo-Iranians who still labialised their /kW/'s
were scorned by their neighbours. "For Diwos sake, what do you mean by
speaking like that? Did you hear him, lads? He said [k_W_etwores],
like a bloody Proto-Greek. [kWetwores], would you believe it? He's a refayned
person. He thinks he's smarter than us Aryas."
Piotr
----- Original Message -----
Sent: Saturday, July 21, 2001 12:10 PM
Subject: [tied] Re: Satem shift