> Evolution is still understood as a process in which random changes,
> which have survival value, are maintained.
> There are many examples of living things like, for instance, the
> crocodile, which apparently have not changed much over great periods
> of time. Should there not also be languages which have remained the
> same over great periods of time? And if not, why not?
> Let us take an example of phonological change which is quite commonly
> observed, the change of (aspirated) stops into their equivalent
> spirants, for example /p(H)/ into /f/.
> Presumably, in your evolutionary scenario, a segment of a language
> population would incorrectly repliacte /p(H)/, and this would
> eventually spread to the larger language population. What 'survival
> value' could possibly be attributed to this change?
Biological evolution and language change are both characterized by
variability, but beyond that the analogy is tenuous. Why would
language change have to be adaptive to an environment or
characterized by natural selection? Why does it have to possess
I think Meillet gave the best answer to the "why" of linguistic change a
long time ago (1906):
"From the fact that language is a social institution, it follows that
linguistics is a social science, and the only variable element that we
can resort to in accounting for linguistic change is social change, of
which linguistic variations are only consequences, sometimes
immediate and direct, more often mediated and indirect...We must
determine which social structure corresponds to a given linguistic
structure, and how in general changes in social structure are translated
into changes in linguistic structure."
The work I've taken this quote from, Labov's _Principles of Linguistic
Change, vol. 2: Social Factors_ attempts--to a limited degree--to do
just this, i.e, correlate contemporary linguistic change in the U.S.
(mainly) with social variables. Of course, it's one thing to examine
sound shifts in modern communities which can be subjected to fine-
grained sociolinguistic analysis, and quite another to understand
language change that took place centuries ago. We still have a
primitive understanding of what constitutes "changes in social
structure" that might be universally relevant to speech communities,
beyond very obvious historical factors like migration or conquest.