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

From: Tavi
Message: 70272
Date: 2012-10-25

--- In cybalist@yahoogroups.com, Jörg Rhiemeier <joerg_rhiemeier@...> wrote:

>
> Indeed. Tavi likes to speak a lot about "preconceived ideas"
> which to fit others bend the facts, but it is actually *him*
> who does that. Of course, handbook knowledge can be wrong
> and it has been amended several times in the past and will
> continue to be amended in the future, but it is meaningless
> to claim bullshit about it the way Tavi does.
>
> > I didn't mean anything of that kind.
>
> So if you don't *mean* "anything of that kind", why are you
> *doing* that all the time? If you suffer from a compulsion
> and do that against your will, you should consult a psychologist.
>
You attribute me things I never said. What I think is the classical genealogic tree model, where all branches stem from a single parent node (PIE), is too simple and inadequate for the IE family.

> It is certainly not impossible that Basque, NWC, NEC and an
> unknown number of extinct languages of Europe are all related
> to each other. But the evidence I have seen so far did not
> convince me, and people who are more qualified than me to make
> statements about this matter tend to reject the proposal as well.

Well, many substrate loanwords in European languages (my own speciality are the ones spoken in the Iberian Peninsula) can be explained thanks to the Vasco-Caucasian hypothesis. But I can't say the same of, say, "Indo-Uralic" or "Nostratic".

>  Of course, the original meanings of the names are *unknown*,
> and can only be recovered where a particular name element
> correlates with a salient feature of the objects named thus.
> This is the case with about a dozen central European place
> names which contain the element */hal-/ - and all denote
> places where salt is produced or has been produced in the
> past. This seems to indicate that we are dealing with the
> LBK people's word for 'salt' here. (And it looks quite
> similar to PIE *sh2al-,
> >
> > Actually, this is *saH2l-, so please don't
> > *cheat* with data.
>
> AFAIK, there is evidence for *both* *sah2l- and *sh2al- in PIE.
> It seems to be a case of Schwebeablaut.
>

IMHO this word is cognate to NEC *q?eh\l(\)-  'bitter', with initial s- coming a post-velar (probably uvular) fricative. I know you favor *sH2al- because it fits your hypothesis, but IMHO there's no need for it, as the isogloss *s- ~ *h- has a better explanation as different reflexes of a former post-velar fricative. Look for example at *sam-/*sm-ro- 'summer' ~ *H2e:m-r- '(heat of) the day'.

> I haven't read Villar's latest book yet (none of the libraries
> I have access to carries it, and I am quite short on money, so
> I don't want to buy it), so I have nothing to say on it. What
> regards your points, I have said what I have to say about them
> often enough, and there is no use repeating that once again here.
> There really is not much to say on them in absence of supporting
> evidence.
>

You can't keep saying "there's no evidence" while at the same time you refuse to look at it!

What Villar and his team show is there was a very ancient dialectal fragmentation in paleo-IE before "Kurganic" (i.e. the language(s) of the Kurgan people) swept in.  

> It is true that, for
> instance, Pokorny reconstructed PIE roots for some words which
> have areally skewed distributions and are thus uncertain to
> actually be of PIE vintage; but I *did* work my way through
> the items limited to "western" IE languages (Italic, Celtic,
> Germanic, Baltic, Slavic), and found no phonological
> peculiarities (such as unexplained */a/-vocalism) that would
> point at substratum loanwords in most of them. They probably
> were dialectal formations in Late PIE.
>
> > I think you didn't a good work, because there're quite a few of these
> > /a/ words, e.g. 'summer'.
>
> Sure. There are quite a few words with */a/-vocalism, but
> many of the restrictedly attested items in Pokorny have
> "ordinary" */e~o/-vocalism, and some words with */a/-vocalism
> have, according to Pokorny, cognates in eastern IE languages,
> but many of those eastern cognates are IMHO poorly founded.

Exactly. In Pokorny's, the word 'summer' has a Sanskrit cognate, but IMHO it's doubtful. This' is a paleo-IE word.

> Anyway, the dictionary is not worth much because the phonology
> Pokorny uses is utterly out of date, and many items have
> semantic problems. He evidently tended to hammer things into
> place that actually did not belong there, and to contrive PIE
> etymologies for items that cannot be ascribed to PIE by any
> reasonable method. (Possibly in an reaction on the harsh
> criticism he earned earlier with his hypotheses about a Semitic
> substratum in Celtic.) It is widely recognized that Pokorny's
> dictionary has many problems, and that there indeed is a
> pressing need for a more modern PIE etymological dictionary.
>

You've got Mallory & Adams (2006).