Re: Divergence vs. convergence (was: Witzel and Sautsutras)

From: Tavi
Message: 70251
Date: 2012-10-24

--- In cybalist@yahoogroups.com, Jörg Rhiemeier
<joerg_rhiemeier@...> wrote:
>
> Just that. And the "rest of IE" changed between that early split
> and its final disintegration into the non-Anatolian IE languages.
> The standard model describes Late PIE, i.e. the latest common
> ancestor of the non-Anatolian IE languages, not perfectly but
> aptly enough; Early PIE, the common ancestor of Late PIE and
> Anatolian, was a quite similar language (it is only about 500 to
> 1000 years older than Late PIE), but different in some respects.
>
In my view, there's no need for this "Early PIE" to be the ancestor of
"Late PIE". This is where I disagree with you and Rodríguez Adrados.

> > A characteristically Tavi-esque non-answer, throwing around
> > a lot of terminology without actually saying anything
> > concrete. I believe that Jörg has had some experience of
> > them in the past on the Zompist board, so it probably won't
> > come as any surprise.
>
A nasty Brian-esque "ad hominem" attack. This is why I've chosen not to
reply to his posts.

> Certainly, there are overlapping isoglosses in the IE family.
> This is because before it finally broke up into distinct
> languages, PIE was a language that extended over a large area
> and consisted of many dialects through which innovations could
> propagate (the "wave model"). Only later those dialects became
> separate branches of Indo-European.
>
> The family tree model and the wave model do not contradict each
> other - they just apply to different stages in the process of
> diversification of a language lineage.
>
In my view, these "waves" are actually language contact and replacement
processes.

> > And your evidence for this unsupported opinion is?
>
> Waiting for that evidence is about as useful as waiting for
> Santa Claus, I wager.
>
On the contrary, this lexicon can be classified according to internal
sound correspondences, relative chronology, and so on.

> The 2000-something PIE roots follow a
> coherent set of root structure constraints, which shows that
> they were part of a *single* language, even if some of them
> may have been borrowed from other languages in an earlier
> stage.
>
Instead of "a single language", I'd rather say a cross-section through
the last stages of the IE family.

> > I'm afraid this is an over-simplification. Not only the
protolanguage
> > from which Anatolian "split" is earlier, but also different (in
> > morphology, lexicon, etc.) from the "late" PIE from which Indic
> > (Sanskrit) and other languages are supposed to derive.
>
> *Of course* it was different, in all these regards! Otherwise
> it would be meaningless to talk about an earlier split.
>
See above.

> > Other isoglosses
> > across non-Anatolian IE lead to scholars such as Rodríguez
Adrados to
> > propose a series of splits and branchings, which IMHO is better than
the
> > traditional tree model but still inadequate.
>
> Rodríguez Adrados is pretty much in agreement with my own
> thinking. What he calls "IE III" is what I call "Late PIE";
> what he calls "IE II" is what I call "Early PIE"; what he calls
> "IE I" is what I call "Proto-Europic", and it may have been the
> ancestor of other languages besides IE, such as that of the
> Linearbandkeramik culture which may have left traces in the Old
> European Hydronymy (but I am digressing into paleolinguistic
> speculations here - though unlike Tavi, I do not try to sell
> them as "facts").
>
Actually, I think your *theory* is more speculative than mine.
Apparently, you've got a preconceived idea of what the nearest relatives
of IE can and can't be.

> > In my own view, which is closer to the of the Spanish scholar
Francisco
> > Villar (a former disciple of R. Adrados), who has studied the
ancient
> > toponymy and hydronymy of Europe and SW Asia (I'd recommed his last
book
> >
<http://books.google.es/books?id=BAwzUADajUwC&printsec=frontcover&hl=ca&\
\
> > source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false> to skepticals),
the
> > IE family isn't the result of the spread of a single
(proto-)language at
> > a rather recent time (Chalcolithic-Bronze Age) but rather to a
series of
> > language spreads and contact/replacement processes over many
millenia:
> > 1) The repopulation of Europe from the Ice Age refuges in the
> > Mesolithic,
>
> Led to the spread of several non-IE families into northern Europe.
>
Some of whose remains still survive in the IE lexicon, as e.g. the
Germanic word for 'bear'.

> > 2) The immigration of Near East farmers in the Neolithic,
> > who introduced agriculture in Europe,
>
> Controversial. The "Danubian" culture complex (Linearbandkeramik,
> Vinča, Cucuteni-Trypillia etc.) is of uncertain origin, but it
> does not resemble the Anatolian Neolithic much. My pet theory is
> that these cultures were founded by refugees from the Black Sea
> Flood (which, however, is controversial). In the Mediterranean,
> at least, Neolithic agriculture seems to have spread by cultural
> influence without large-scale migrations, and we thus have lots
> of unrelated linguistic stocks, of which Basque and the three
> Caucasian stocks have survived until today.
>
The lexicon relative to farming, such as 'plough' (with Semitic
cognates) had to be introcuded by Neolithic farmers.

> > and 3) The Kurgan invasions of
> > nomadic pastoralists from the Pontic Steppes.
>
> "Kurgan culture" is an obsolete cover term for a whole bunch of
> different Late Neolithic/Chalkolithic cultures of the steppes
> which are characterized by a particular type of grave sites,
> and do not necessarily form a linguistic unit. *One* of these
> cultures, the Sredny Stog culture (in Ukraine ca. 4000 BC), may
> have been the Early PIE speakers, and a daughter culture thereof,
> the Yamna culture (in Ukraine and southern Russia ca. 3200 BC),
> Late PIE.
>
I'd call "Kurganic" the language of the steppe people, which would
*only* be the direct ancestor of the Indo-Greek group (= Adrados' IE III
A), with a variable influence over the rest of the IE family.

> Certainly, many loanwords from the Mesolithic and Neolithic
> languages ended up in the attested IE languages in the process
> of the latter replacing the former. You are battering an open
> door here. Yet, in their grammatical structures and the
> majority of their lexemes, the IE languages show such a degree
> of coherence that their descent from a single common ancestor
> remains the most plausible hypothesis.
>
If only we forgot about Anatolian as well as other "inconveniences".

> I admit that many
> Indo-Europeanists sweep the residue away into a dustbin labeled
> "Uncertain etymologies", while this "debris" may contain valuable
> information about the lost languages of pre-IE Europe.
>
Actually, many "PIE" etymologies are actually "pre-IE" in the sense you
employ this term.