Re: [tied] *ekwos and friends

----- Original Message -----

From: george knysh

To: cybalist@yahoogroups.com

Sent: Wednesday, October 17, 2001 5:31 PM

Subject: Re: [tied] *ekwos and friends

*****GK: I agree. On the other hand I don't think you can "demonstrate" (I use this in the strong sense) your point of view either, since there is no evidence that what linguists reconstruct as ancient IE was actually ever spoken by anyone. It's internally
consistent. It's very useful. But it could be utterly unreal.******

 

Still, it sums up what _can_ be learnt about the ancestral language of the family -- it gives us a rough but reasonably accurate idea of what it was like. Of course much of the relevant information has been lost forever and can never be recovered. We reconstruct the shadow of PIE, not the language itself -- that goes without saying, but then science in general is not about the ultimate reality but about models that work. I've never seen an electron, but I understand the reasons that make it a necessary entity, and the same is true of PIE. The logic of the comparative method is in fact quite strong. Of course I can't present a native speaker who would testify to the reality of, say, Nom.sg. ek^wos, Acc.sg. ek^wom, Loc.sg. ek^woi, etc., but these forms are "real" in the sense that they result from principled comparison -- a strict method that can be used for making predictions and verifying them. Feed it with good data, apply it well, and it will yield interesting results. Feed it with garbage or misapply it, and it will yield garbage. There is no other way to work, no shortcut -- except, perhaps, some kind of paranormal revelation or field research using a time-machine, but I don't happen to trust such methods :).

*****GK: But you can't really prove this. I don't see why we have to wait a few thousand years for hypocorisms, "baby talk" et sim. And what you've just said can apply equally well to some ancient type calling out "kos' kos'" to a horse. You have no way of demonstrating that this is/was impossible except in terms of a science which while internally consistent cannot really be verified in all of its assumptions and claims. I have a high degree respect for its accomplishments but also a high degree of skepticism
as to the universal existence of a "normative" ancient IE language, given the reality (verifiable today) of so much dialectical variety within existing normativized languages. Dialectologists seems continually amazed at the variety of terms and words they discover "in the field". It takes a while for Academies of Science and school systems to
evolve.*******

I don't have to demonstrate that a proposal is impossible. If something is "not impossible" but at the same time unverifiable or unreconstructible, it cannot be a valid part of the accepted model of PIE. If you claim that the PIE horse-call was *kos-kos, the burden of the proof is on you. Why *kos-kos rather than *hop-hop, *hei-ja-hou, *[whistle-whistle] or *[click-click]? I do not doubt that a whole range of such or similar interjections were used by the "real IEs", but it's not something we can reconstruct. Surely, [kos'kos'] is not a universal horse-call (no, we don't use it in Poland, and as far as I'm aware, such "calls" are many and various), and no facts known to me guarantee its historical continuity into the distant past. Here I rest my case.

 

Piotr