From: stlatos
Message: 68114
Date: 2011-10-19
>I said "no IE supports PIE reduplication of C-clusters", not that all IE show the same reduplication as PIE. The existence of a linguistic feature in an IE language doesn't support the existence of that linguistic feature in PIE if analogy can explain it (also, the examination of other IE is usually needed). Abnormal IE reduplication, such as in Sanskrit, is probably the result of analogy (some partly influenced by the deep structure of phonological patterning).
> W dniu 2011-10-18 22:08, stlatos pisze:
>
> > Lehmann rejected reduplication as the cause because he started w *
> > stestud+ and said, correctly, that dis. of st-st > st-s was unlikely. Of
> > course, there's no reason to start w * stestud+ since no IE supports PIE
> > reduplication of C-clusters.
>
> Gothic and Tocharian partly do (full reduplication *sTV-sT- is attested
> there), though *sV-sT- is probably the "original" IE pattern.
>