Julian Lord:
> > > Anyway, it's very difficult to apply modern linguistics to this
> > > problem without assuming the existence of such an ur-language, because
> > > one of the very foundations of linguistics is that all speech evolved
> > > from a single root.
> >
> > Hrm. Last I heard even Nostratic was regarded as an interesting
> > hypothesis, but somewhere between "unproven" and "unprovable". I'm not
> > aware of any such assumption about *all* language being widespread among
> > linguists, much less it being the 'foundation' of anything. What'd "go
> > wrong" in linguistics if this were falsified? Functionally nothing,
> > surely.
>
> Er, yes.
I feel one of my periodic anti-ums-and-ers rants coming on... Must... resist...
> The basic premise and more importantly _methods_ of linguistics
> assume that there has been an historical evolution from a common core.
>
> This permeates every single part of linguistics, but to give a simple example
> it's assumed that the Romance languages evolved from Latin, which in turn
> evolved from an earlier group, itself derived from Indo-European etc.
Well, I can hear the "linguistic is a respectable science, honest" mob reaching for their revolvers up and down the land. These aren't _assumptions_, these are hypotheses that can be verified to some reasonably high degree. (Very high in the case of Latin, and PDG in the case of I-E.) But that's at the level of a linguistic (sub-)family, and has exactly zilch implications for a global ur-language. Rather it'd imply anything up to about 250 "ur-languages". (Just counting surviving ones, obviously.)
> Some linguists attempt to find the Origin of Speech in our genes,
> which are they themselves the product of an evolutionary (and
> cybernetic) process ; others prefer the idea that the Origin is
> linguistic in nature (not incompatible with Christian theology BTW).
>
> However, the very concept of linguistic shift and evolution et cetera
> _requires_ (from our limited POV) that the model postulate an Origin,
> or the model would simply not function.
>
> The notion of Multiple Origins is pretty much an anti-linguistic one.
Evidence, please. Not for the belief in a single origin for Indo-European, or even of the Nostratic super-family (which has the status of an unproven but more or less respectable theory), but of a _fundamental_ assumption or reliance on any single origin for _all_ language. Everything I've read on the subject suggests that this is waaaaaay beyond anything that's considered testable in linguistics research, and would thus have no business being such; nor am I aware of any methodological reason for having it in linguistics.
Please distinguish this from some linguist saying something to the effect of "I reckon there's probably a single origin of language, but it's completely unprovable", which is common enough, but almost the exact opposite of your assertion.
--__--__--
Powered by hypermail