Re: [tied] Re: Dating PIE

From: Piotr Gasiorowski
Message: 10431
Date: 2001-10-19

----- Original Message -----
From: "george knysh" <gknysh@...>
To: <>
Sent: Friday, October 19, 2001 2:34 AM
Subject: Re: [tied] Re: Dating PIE

> *****GK: Yahoo can be mysterious also. Did it render
> your word correctly? I notice that when
> M-E-D-I-E-V-A-L
> is typed it comes out as M-E-D-I-A-R-E-V-I-E-W (at
> least for me and some others). Would you kindly retype
> the Tocharian cognate in split capitals? I'll make the
> adjustments. So Tocharian has one of the usual IE
> terms for "wheel" plus this one?*****

The Tocharian word is <wArkAnt>, where <A> = "a umlaut", really a kind of
schwa in Tocharian.

> >.... but [<hurki>] has
> > also been suspected of Hattic origin.
> *****GK: Is the proof for that as problematic as for
> it being a possible Tocharian cognate?*****

It's a fifty/fifty sort of balance, I'd say. There's an article by C.
Justus, in which <hurki> is quite convincingly related to words with clearly
Hattic morphology.

> *****GK: ... Interesting though that these should have borrowed the
> term while Hittite did not. I suppose the L/L
> borrowing substituted for something else since these
> languages would not have waited for the Mitannians to
> give them a word for designating horses. I'm wondering
> if the Hittite Sumerogram and the replaced L/L horse
> word might have been some kind of poetic
> circumlocution (in a substrate non IE language?).
> Seems like a lot of letters for a "basic" animal.
> "swift as the wind" or something like that.******

The point is that the horse became a familiar animal in Asia Minor only when
the Indo-Aryans introduced it. We don't know that the Hittite words was NOT
*asuwa-, *azuwa- or the like. The Sumerogram masks the actual Hittite term,
just like the use of "&" for "and" or "14" for "fourteen" in English
conceals the alphabetic spelling. It isn't "a lot of letters" -- it's just
three cuneiform characters.

> *****GK: Are there lexical substitutes from other
> (identified or not yet)non-IE languages? After all the
> Hittites had a marvelous war chariot machine. They
> must have had words for its components "spare parts"
> etc..?*****

You must have heard of Kikkuli's horse-training manual. It's in Hittite, but
the technical jargon is all Indo-Aryan. I don't know the Anatolian chariotry
terminology by heart. If I have a chance to check those terms, I'll share
them with you.

> *****GK: I think this is entering very slippery
> ground. Are you saying that many of the "mysterious"
> terms in Hittite etc. which some deem non IE could
> actually be IE words lost in every other IE language?
> The reverse seems far more likely. One of the problems
> (if I am not misreading you)would be to explain why
> the other IE languages should have deemed it necessary
> to replace the words kept by Hittite etc. with new
> ones. And are Anatolian terms always related in their
> narrower context? Not always from what I've seen.*****

It's just a matter of relative probability. If PIE split into two dialects
(let's call them Proto-A and Proto-B) at a very early date, the chances that
a term would survive in Proto-A or Proto-B alone were roughly equal. Now
let's imagine that Proto-A is Proto-Anatolian and Proto-B is the parent of
all the non-Anatolian languages. A term surviving in proto-B would be quite
likely to survive in several lower-order branches, since its total
extinction would require several independent losses. So, on the whole, a PIE
word would be more likely to survive only in Anatolian than, say, only in
Iranian or only in Celtic. This reasoning is intuitive rather than strict,
but I hope you see what I mean. The earlier a language branched off, the
greater its relative "weight" in comparison. Hittite and the other Anatolian
languages are not exceptional in having a rich non-IE substrate: Indo-Aryan
has plenty of "para-Munda" and Dravidian borrowings, Tocharian has numerous
loans from hitherto unidentified sources, and Armenian, Greek, and to a
lesser extent Germanic, Celtic, etc. also have their substratal components.

Strictly speaking, if a word does not survive in Anatolian, we should
refrain from labelling the reconstruction "PIE". IE scholars habitually sin
against methodological caution in this respect, thus tacitly treating
Anatolian as if it were just one branch of many (which it most likely
isn't). We lack a good name for the non-Anatolian part of the family
("non-Anatolian" itself is clumsy), which may partly account for this
ambiguous (and, in fact, biassed) use of the term PIE. I mean that if a word
is attested in, say, Celtic, Baltic, Greek and Iranian we call it "PIE", but
if it's attested in Hittite, Luwian and Palaic we call it "Anatolian". I
don't like this confusing state of affairs, but what can I do on my own?

> ****GK: Very "favourable" assumptions indeed. On that
> basis what would be your estimate as to Hittite. I.e.
> % of certain IE terms, % of certain non-IE terms, % of
> "unclear" terms. As I remember my Mallory his source
> for Old Greek contended that in that language 50% of
> its lexical fund was certainly IE. 8% was not, and 42%
> was unclear.== But I must say that for me "unclear"
> would just be a euphemism for "non-IE items we can't
> bring ourselves to admit being such"...****

There's nothing shameful about a word being non-IE. "Unclear" simply means
that it *might* be non-IE but we can't identify a plausible source, or that
it *might* be IE but no-one has proposed a convincing derivation so far. I'm
not going to venture any estimates concerning the origin of Anatolian
vocabulary. I don't specialise in Anatolian and my knowledge of the ancient
non-IE languages of Asia Minor and the Near East is very scanty. Even people
who know much more about those things could only give you the roughest