--- In cybalist@yahoogroups.com, "Brian M. Scott" <bm.brian@...> wrote:
>
> You really just don't get it, do you? PIE is by
> definition the stage immediately prior to that
> fragmentation. Anything earlier is not PIE. Anything
> ancestral to that stage is something else.
>
> > But Anatolian (and probably also Tocharian) is clearly
> > "ancestral to that stage", so an older "PIE" would be
> > needed as a common ancestor of Anatolian and "traditional
> > PIE". This is more or less which Adrados proposes with
> > their IE II and IE III. No, the traditional model is a
> > *clumsy* oversimplification.
>
> Are you really unaware that these issues are very familiar
> within the field? Just how out of date is your picture of
> PIE, anyway?
>
In my understanding, the majority *opinion* among IE-ists is that PIE was a real language spoken not earlier than the Chalcolithic, and this chronology also includes Anatolian. I've throughly read Mallory and Anthony, Brian.
> > But this "PIE" is only a comparatively small part of the
> > whole business, as a large part of the IE lexicon actually
> > predates it.
>
> Of course. And a large part of the PGmc. lexicon predates
> PGmc.; so what?
>
That it can't be "inherited" from "PIE" in the same way than the rest. A good example would be the 'bear' word we discussed earlier.
> Say rather that they've generally recognized that there's
> not much that can be said about it, at least until (a)
> some likely other branches have been identified, and (b)
> adequate reconstruction has been done in these branches,
> including IE. And obviously the first order of business
> for any sensible Indo-Europeanist is to work on setting
> his own house in order.
>
> > But "these other branches" doesn't survive anymore
> > (because they've being long replaced) but in the form of
> > *substrates* in IE languages, as well as in fossilized
> > toponymy and hydronymy.
>
> Which means that we may never know much about them, because
> the evidence is too skimpy and buried under too much noise.
>
On the contrary, I think there's still much work to be done from the available evidence.
> > And while Villar studies the latter, I study the former.
>
> You don't appear to do so. I've seen nothing in your posts
> like the various works identifying the characteristics of
> Old European hydronymy, or the work of Kuiper and Schrijver
> on characterizing specific substrates in (at least)
> Germanic. You go about it ass-backwards: instead of trying
> to identify a coherent substrate in an IE language or
> branch, you start with a source language, and one whose
> existence is doubtful at that, and then dig through the IE
> lexicon in search of words that possibly be borrowed from
> this source.
>
These specialists (to which you should add Villar) have their own method and I've got mine.
> Nor is there any reason to assume a priori that these
> substrates are detectably related to IE.
>
I disagree. For example, in etymological dictionaries such as Matasovic´s or De Vaan's you can find a significant number of Celtic and Latin words without a straight PIE etymology (i.e. derived with "regular" sound correspondences) which are yet of IE stock. For example, I've already mentioned Celtic *Flanda: 'heath, uncultivated land' and Latin fro:ns, fro:ndis 'foliage, leaves'. Clearly they're neither Semitic, nor Vasco-Caucasian, nor whatever else than IE.
> Similarly, it's entirely to be expected that a Germanist
> will be concerned primarily with the Germanic branch; the
> difference is that in this case we actually know what the
> sister families are and something about how they're
> related and how they descend from PIE, their nearest
> common ancestor. Despite some conjectures, some of them
> even quite plausible, we *don't* know what IE's sister
> families are, let alone what their common ancestor looked
> like.
>
> > Following this kind of logic, "PIE" would be also a
> > conjecture, although a bit more plausible.
>
> Of course, though it's considerably more than just 'a bit'
> more plausible. In broad outline it's an extremely
> well-supported conjecture, because the work required to
> support it exists. The same cannot be said for any of the
> hyper-taxa, macro-families, or whatever you want to call
> them.
>
We'll see.