From: Piotr Gasiorowski Message: 66498 Date: 2010-08-29
W dniu 2010-08-29 09:09, G&P pisze:
> And one root can participate in more than one correspondence.
It *has to* participate in more than one correspondence to count as a
reconstructible root in the first place. Typically, it will support a
large number of proposed correspondences, if the protolanguage
reconstruction is to be taken seriously.