Re: [tied] *pa(x)u- "small" and "fire"?

From: Glen Gordon
Message: 17390
Date: 2003-01-04

>I don't think the general theory of PIE extensions (if ever formulated)
>will be that simple, [...]

However, it's not meant to be a "general theory" (if that is a way of
saying "panacea") and it isn't that simple. The CVCC/CCVC thing only
accounts for the extension *-eu- and *-ei-, the s-aorist and the usual
accent alternating verb pairs (eg: *ters-/*tres-). (Perhaps it could also
explain *-em- if it can be satisfactorily related to nouns ending in
*-mo- or *-m-n.)

Other extensions like *-(a)x- [transitive] appear to be bonifiedly
non-denominal and ancient, even if their accent alternations are to be
explained by my above "general theory" (if that is to mean "an _overall_
theory" like that of Grimm's Law).

>but I find your idea that *pah2-u- might be a derivative of *pah2- really
>brilliant and I think I'll buy it if you don't charge too much

$200 and I take cash. Hey, I need new shoes :)

>I still don't think a derivation like *pah2u- + -r is viable; even in
>the Caland system you can have adjectival derivatives with *-u- or with
>*-ro- (*-no-), but less usually with both at the same time.

But I'm not trying to form an adjective with *paxwr -- it's a noun.

I'm not sure where you're going here. The ending *-ro- is different from
inanimate *-r. Now, *paxwr "fire" either can be divided as *pax-wr (as
you say) or *paxw-r (as I say), but in the end it appears to me that
either analysis would create the same basic meaning ("that which is
protected"). One is the combination of verb and passive *-wr, and the
other is an adjective plus active inanimate agent *-r.

The question between us now is, if *paxu- underwent a semantic shift from
"protected" to "small" during Late IE, did *paxwr form before or
after such a shift? If after, *paxwr could have meant "that which is
small" -- a respectful term for fire. If before, as you might suggest,
*paxwr was to mean "that which is protected". Granted, the "before"
scenario is more direct and Occam tells me that I should opt for that
idea before my own pet theory. However, the topic is still interesting
in terms of IE belief-systems and, I think, worth a thought if it can't
be ruled out morphologically.

On the topic of *-u-ro-, however, I'm not surprised that its attestation
is scanty, being that *-ro- was created very late. It is one of many
"thematicized" animate suffixes made by the addition of the thematic
vowel to a consonant-final suffix.

However, before this was a la mode, middle Late IE employed the thematic
vowel in a different way. This method dates before the time when the
thematic vowel (which was then pronounced *&) had changed to *e or *o,
depending on automatic lengthening before voiced segments. Thus, the
animate variant *-or- (< *-&:r < inanimate *-r), and others of its ilk
such as the later "feminine" animate *-ax (< *-&x), were produced at
this time.

What you state only shows that adjectives can indeed be given the same
endings as verbs. Plus even though *paxuros relates to *paxu- rather
than *paxwr, doesn't *paxwr nonetheless have the potential to form
a derivative like *pxuros in the same manner as *udros from *wodr?

- gLeN

Help STOP SPAM: Try the new MSN 8 and get 2 months FREE*