Gerry in reply to Dennis wrote
> Since for the moment Dennis I have no answer to your question
regarding
> which language Burushaski would be a dialect of, let me ask you
another
> question or two. How many word stems are needed to *prove* a
> correlation between two languages? Ten, twenty, several hundred?
And
> why can't Burushaski be correlated with one or several of the
American
> Indian languages?
Gerry, Glottochronology gives the following table
% of divergence between Minimum no. of centuries
any two languages since divergence occurred
100 0
90 3.5
80 7.4
70 11.8
60 16.9
50 22.9
40 30.3
30 39.9
20 56.6
10 102.6
1 225.0
Thus by knowing the difference in similarity you can read out a
minimum time of divergence. Since Amerindians have been living in
North America between 120-150 centuries, we are looking at Burushaski
having between 1-10% common vocabulary. This would rate as a
difference in language, Gerry, not a difference in dialect.
Nostratic is dealing with such long horizons. It is what makes the
field so contentious.
Hope this helps
Regards
John