Baltic and Slavic are definitely close
relatives, I'd say closer than Indic and Iranian, and perhaps as close as
Goidelic and Brittonic Celtic. They are characterised by a sufficient number of
unique common innovations in phonology and morphology (also lexical ones, but
those are easier to borrow and consequently less probative) to be regarded as
members of one and the same branch. Remember that the striking similarity of
Sanskrit and Avestan is due to their great age -- their oldest known form was in
use barely more than a millennium after the breakup of Proto-Indo-Iranian.
Compare modern Bengali and Kurdish, or Oriya and Ossetic, if you want a really
fair parallel to the case of extant Baltic and Slavic. Sergei Tarasovas and
I discussed the position of Slavic some time ago and more or less concluded
that, technically speaking, Slavic is a highly innovative and therefore somewhat
aberrant-looking subbranch of Baltic (or even West Baltic, to be precise, which
would mean that reconstructible Proto-Baltic = Proto-Balto-Slavic). The
objection that common Balto-Slavic may not have been completely homogeneous is
hardly serious -- the same could be said of any language, dead or
living.
I sympathise with the view that Satem is
a genetically valid taxon, since the Satem innovation occurs in a
well-defined group: satemisation does not "spill over" into the surrounding
centum languages. There are sporadic "centum" forms in Baltic and Slavic (most
likely ancient loans, just as unpalatalised _get_ or _sky_ in English are
Norse-influenced northernisms), and perhaps cases of genuinely failed
satemisation (especially before liquids). They may reflect dialectal differences
within Proto-Satem, but are not really much more problematic than the
failure of *kv- and *gv- to undergo paltalisation before secondary front vowels
in East Slavic.
I'd (tentatively) remove Phrygian from
Sasha's list -- apart from geography, there's precious little connection with
Armenian. As far as I'm concerned, the allegedly Satem character of Phrygian has
not been demonstrated. Recorded Phrygian shows palatalised velars, to be
sure, but so do French, English, Greek, etc. in appropriate contexts --
conditioned palatalisations are not Satem. Phrygian may be closer to
Hellenic than to Thracian or Armanian.
Dacian and Albanian should certainly be
added (perhaps as a single branch). There's too little tangible evidence
to assess the relation between Thracian and Dacian. Outside the Balkan
area, where the ancient linguistic situation remains open to dispute, I'd say
that Slavic is most likely the youngest branch in the Satem
cluster.
Piotr
----- Original Message -----
Sent: Monday, January 21, 2002 9:31 PM
Subject: Re: [tied] Likely IE home: India
>>As far as I can see there was a real Common Satem
language but there were
no >"Common >Centum language" (if we don't want
to apply this term to PIE
proper).
>>That Common Satem produced
later Common IndoIranian, Common BaltoSlavic,
>>Common ArmenoPhrygian,
Common Thracian ... What else?
>Piotr, what should be put instead of
the very last group of dots in your
scheme?
>I mean, what language was
the _last_ who separated from the future Common
Satem?
Surely the idea
of a common BaltoSlavic is open to dispute? (I say
"surely"
because it must be open to dispute, since it is in fact disputed).
The whole
idea of a common Satem may also be unnecessary, if we go with a
wave theory,
which would explain some of the forms in Baltic and Slavic
rather more
easily.
A Common Satem would, I suppose, be characterised by
satemisation. We would
then have to suppose Baltic (and Slavic?) had
re-centumised the items which
do not show satemisation, which is a bit
unbelievable.
Peter
Your use of Yahoo! Groups is
subject to the Yahoo! Terms of
Service.