Re: Hercynian (again)

--- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy <bhrihstlobhrouzghdhroy@...> wrote:

>
> > IMHO the root *perkW- 'oak, pine' isn't a native IE word but

rather a substrate borrowing (call it "Paleo-

> > European" or whatever

else), and its similarity to the name of a thunder god in some cultures is purely

> > coincidential. There's no need to imagine implausible semantic

shifts and the like.

>
> @Octavià: (we had already discussed the topic in other lists)
>

Sure? Under whose alias? I can't remember you.

> a Non-Indo-European substrate can be alternative to an Indo-European
> etymology only if one has documented Non-Indo-European languages with
> assuredly known diachronic phonology and where those lexical items are
> precisely attested in the very expected form; otherwise it's just
> possible, but always less probable (because far less economical) than
> a plain Indo-European etymology. A diachronic phonology based simply
> on comparisons between words attested in Indo-European languages only
> and in unexpected mutual phonological relationship runs the risk to be
> based on simply wrong comparisons
>

As pointed by F. R. Adrados et al. (Manual de lingüística indoeuropea, 3 vols.), the IE lexicon has been rather USED for stablishing sound correspondences than studied by itself. Do you really believe the hundreds of supposed PIE roots one can find in etymological dictionaries such as Pokorny's or Mallory-Adams' actually belong to a single (proto-)language? My educated guess is NOT.

For example, using the comparative method I can reconstruct perfectly valid IE etymologies of the Germanic words for 'bear' and 'horse', whose reconstructed meanings are respectively 'wild animal' and 'to run'. As they're different from the ones found in "common IE", the traditional explanation is the latter were replaced in Proto-Germanic by tabooistic reasons. However, I think it's more likely the Germanic words were the ones not replaced by the common IE words and not the other way around.

In chronological terms, such semantic shifts usually indicates an older age, so the roots corresponding to the Germanic words must be OLDER than the common IE ones (in fact, the terminus post quem of common IE *h1ek´w-o- 'horse' is the domestication of the animal in the Pontic-Caspian Steppes around 4,000-3,500 BC.)

By this and others reasons I consider the traditional PIE model as inadequate. This is why I replace it with a multi-layer and multi-tree model. "Multi-layer" means that IE languages are the result of one or more language replacement processes, which caused a superposition of lexical layers, and "multi-tree" means there was more than a protolanguage these layers have originated from. So IMHO they weren't a single but several PIEs.