SV: Fortunatov's Law and tolkAppiyar's rules

[Miguel Carrasquer Vidal]

Lars Martin Fosse <lmfosse at ONLINE.NO> wrote:

>The fact that a certain phenomenon is shared by some languages does not exclude
>the possibility that they occurred independently in some of  - or all of - the
>languages. How do we know that the changes are not independent?

Indeed.  We know that for instance the change *k(w)e > c^e is
independent in Slavic (c^etyre), Indo-Iranian (catva:ra-) and
Armenian (c^'ork') [PIE *kwetwores "4"].  This is very common sound
change, which makes it unsurprising that it should have happened
independently many times over.  The "satem"-change (*k^ becomes some
kind of s(h)ibilant) is just as common and trivial, but there are
reasons to believe that this change was not independent, but affected
a single dialect of IE.  This has to do with the fact that the
"satem" languages have several other things in common, things that
less trivial.  One of them is the RUKI-rule (*s > /s^/ after i, u, r,
k (and H?)).  This makes it more likely that we are dealing with a
set of shared innovations, and that it is probably justified to speak
of a "satem" subgroup of IE (on the other hand, there is no "centum"
group: those are just the languages that fail to be "satem").

But there can't be 100% certainty: the satem-change might have
happened in all these languages independently.

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam