Re: Hercynian (again)

2012/2/29, Richard Wordingham <richard.wordingham@...>:

> --- In cybalist@yahoogroups.com, "Tavi" <oalexandre@...> wrote:
>
>> As I said before, it's extremely unlikle, if not possibiliy, that ALL
>> the reconstructed "PIE" roots could belong to a single protolanguage.
>
> Quantify 'ALL'. For example, it is seriously suggested that a lot of the
> roots in Pokorny are just coincidences or later or parallel loans. If one
> aims to list all PIE roots, one will list a lot of non-existent roots.
>
> Richard.
>
>

——————————————————————————————————————

If a root is actually a coincidence, then we have even more than one root.
If a root is a loan, it would be fair to detect the donor language
(= 1 find attestations of those very word in one or more non-IE
languages, and 2 make it more probable that the direction of the loan
has been from non-IE to IE and not vice versa), otherwise the loan
hypothesis is weaker than the hereditary one