[tied] Re: Automatic clustering of languages

> Not all regularities are a result of genetic descent. See p. 89
>
> http://books.google.com/books?

id=QBFPiT4dUQQC&pg=PR11&lpg=PR11&dq=kruskal+dyen+black&sig=yOpWhN34c7
1kXn1jusmvSqxmt3k#PPA89,M1

>

Please use eg. tinyurl.com in the future to reduce the length of
your references so we won't have to cut-paste to reconstruct them.

> English street and German strasse exhibit "regular" correspondence
> but they are both loans from Latin.

No need for the quote marks, they are regular.

> This kind of loaning is
> possible to detect only when written histories of langauges are
> available.
> If PIE is assumed to be spoken 6000 years ago "regular"
> correspondences could come from 100% genetic decent or 0% genetic
> descent.

In the reconstructed language, yes.
Applying the comparative method to 'street' and 'Strasse' tells us
that the ancestor of that word would have occurred in Proto-
Germanic, which was spoken 2000 years ago, so it's a loanword from
Latin into Proto-Germanic. Of course with the reconstructed language
you would have to use other methods to find out what might be
original and what was loans in that language, the same methods one
must use to discover what is original and what is loaned in eg.
Sanskrit.


Torsten