IE and Substrates and Time

[Anthony Appleyard]

  Someone wrote:-

> So can we not assume that somewhere between 1000 and 2000 years is required
> for communications difficulties to become strenuous, and more than 2000 years
> for them to be so large as to prevent communications ...

  Sheila Watts <sw271 at cus.cam.ac.uk> replied:-

> This seems to me to be too broad a generalisation to work in any useful way.
> ...

and others have written about the amount of change since dispersion in e.g.
Polynesian compared with Melanesian. I read a theory that there is one
mechanism whereby a massive change in a language, or even a completely
unrelated language, can develop quickly: in some natural environments where
living is easy: very rarely but in theory it could happen, some children too
young to have learned much of their parents' language could stray, or be the
only members of their tribe to survive an enemy attack in a tribal war, manage
to survive uncontacted to adulthood, and start a new tribe.