>I don't see why one of the two has to be earlier.
It's far likelier that one is older than the other rather than dating from
exactly the same time - that's just common-sense probability. Since the
vrddhi-collective can be shown to stem from MidIE accentuation (*nam�nxe)
while the simple suffixing of *-x cannot (*n�mn-x(e)), it would appear to me
that vrddhi-collectives are more ancient than the latter, even if by only a
few centuries between 5500 and 4000 BCE.
Even with the loss of *-x in these pseudo-vrddhied forms, the collective
suffix still survived to Late IE just as nominative *-s similarly survived
despite its occasional loss. Most vowel-final stems retained *-x for
instance. We might count *nomn as one of these vowel-final stems :) The *-x
here was certainly applied a little later than in the "vrddhied" variant but
*nomn ends in a nasal vowel producing a new form *nomnx (CVCV-x structure),
contrasting with the earlier form *nem�nx (CVCVC-x structure). Early Late IE
*nem�nx becomes *nemo:n but *nomnx remains unchanged.
>Perfect sense. Skt. -i in the collective derives from *-h2, whether
>it's added to a vrddhied form (nama:n-i) or directly to the stem
>(adant-i).
Sanskrit /na:ma:n-i/ does NOT represent an IE form! It may represent the
vrddhi phenomenon but is a terrible example of an attestation of IE *-x that
never was part of this particular vrddhied stem! Yes, /i/ does often derive
from *x (*H2), so what? This isn't an example of the proper use of the
collective *-x in IE. Give it up.
>*dkontx
What a coincidence - I was just thinking about this blessed little numeral
:) Your reconstruction is wrong. It's *kontx at best. There was never a *d-
here. It's not even convincingly attested but rather assumed on the part of
many IEists. Only in the singular do we find *d- (*dekm). This is because
the *de- in *dekm is the ancient numeral for "one". ProtoSteppe *t:u-kam is
attested in IE and Altaic (Turkish on; OJap to"wo"-). The dental-initial
numeral "one" is preserved clearly in Etruscan /tHu/ "one" but also in IE
*dwo: & Etruscan /zal/ (from IT *t:Wei, *t:We "two" < Steppe *t:u-i)
In IndoTyrrhenian, "ten" is *kem or *dekem and "tens" is *kemec with plural
*-ec (IE *-es). The plural stem when declined with non-accusative case
endings is *kemec- but it is *kemet- in the accusative (all from Steppe
*kWumit). The loss of labiality in *kW- is admittedly unexpected and
irregular.
*k�mec oblique (nom.,loc.,voc.)
*k�metem accusative
*k�mec-e-se genitive
*k�mec-e-ta ablative
The *-x is, yet again, secondarily introduced. We should have expected
**k�ms-x but it would seem that, by Late IE, *kont- was the new regularized
stem form, disassociated from the plural in *-es, and the Early Late IE form
before *e/*o ablaut was *k�mt(-x). The corresponding Mid IE form should be
reconstructed as *k�mec without *-xe:
Mid IE E.Late IE Late IE
*k�mec > *k�m(e)t > *k�mt-x > *k�ntx
>If I'm not mistaken, you started out by saying that PIE *o was the >result
>of Tonal-pPIE **e ([@]) when unstressed (> *e when stressed). >Then I
>objected that we also find *o a lot in stressed position, even >in
>monosyllables. Then you said that *a too could give *o, which you >had
>failed to mention earlier.
Admittedly, I was a little confused before about how MidIE *a evolves to
LateIE and whether *e becomes *o, *a becomes *o or both become *o. So it
turns out an earlier *a almost totally becomes LateIE *o (even when
accented) and some LateIE *e's merge with this *a when affected by labial
phonemes. Conjugational vowel harmony explains the rest. I've adapted and
should be able in the future to respond better. I'm in a process of
evolution over here...
>So then I asked you for *your* examples of *e -> *o due to lack of >stress
>(the canonical examples being <poim�:n> and <d�imo:n>, where >we indeed
>have stressed -m�n- vs. unstressed -mon-, but that's >"classical" *e, not
>your schwa).
You're not talking about IE anymore. You're talking about the later
utilisation of *e/*o ablaut in Greek. I'm not reconstructing "pre-Greek". I
still don't see your point. Please stick to IE and its grammar.
>There are also PD -s, -n ~ -r/-n stems (mainly neuters). I suppose
>your infix -e- takes of that. But what about -i-/-u- HD stems, then?
HD -i-/-u- stems are a late development, of course. They should not have
been HD originally but the process of analogy can be seen to have affected
such previously PD stems as *doru. The ancient PD-style genitive is *drous
but we later find a new HD form *dorw-�s, obviously assimilated to a much
more common HD declension.
>Well, something along those lines would seem to be the answer. There
>are of course a lot of details to be filled in. For instance,
>Rasmussen has proposed that the static inflections can be explained >by an
>initial long vowel (HD **CV:CV'R(V)-, **CV:CVRV's(V)), which
>attracted the accent after the working of zero grade and the
>shortening of unstressed long vowels (**C�CR, **C�CRs).
That sounds really lame to me, if I may be so blunt. Why on earth would a
long vowel "attract" the accent. What's the linguistic process involved
here? Or is this another ad-hoc fantasy again? Are there real-world examples
of this "attraction"? At least the solution that I provide doesn't require
obscure accent attraction. The accent just stays where it is and the vowels
just vanish. Seems like my solution is more Occam-compatible.
>>>*g^ was most certainly not caused by *e. I, and many others with >>>me,
>>>consider it to be a separate PIE phoneme.
>>
>>Who exactly?
>
>To name a few, Brugmann, Beekes (with reservations), Gamqrelidze and
>Ivanov, Pedersen, Rasmussen, Illich-Svitych, Watkins.
On the subject of Nostratic, it is better to trust the opinions of the
much-alive like Allan Bomhard over people who died years ago like
Illych-Svitych whose vague linguistic musings like *Ku"jnA are now far too
antiquated to be of much modern-day benefit. You've already presented some
examples of Rasmussen's madness. Gamqrelidze & Ivanov have their own
problems upholding some of their theoretical monstrosities to the tide of
criticisms that fundamentally challenge their credibility.
>An interesting argument from G & I: as is known, PIE does not allow
>roots of the shape TET, DHEDH, PEP, BHEBH etc. [...]
Look, if I recall, G&I are the crackpots that posit an Armenian homeland for
IE. That in itself says it all about their quality of research which might
almost smell of sensationalistic marketing if I didn't know better. I've
never trusted them. Do you swallow their looney reconstructions for "monkey"
or the kooky impossible IE phonemes like *s^ or *q too? I dare not ask this
question. Sorry, I've just stabbed my ears with some cotton swabs and won't
be hearing the nonsense you're telling me.
>I think you're seriously confused. The law of the palatals (if >that's
>what you mean) has nothing to do with the question at hand.
Yes, they do indeed, but you're probably to busy reading G&I to notice.
>*Lack* of stress. **wa:dn -> **wodr, *wa:dn�:s -> *wedn�s. For the
>first couple of centuries, read the *o as in Sanskrit, always long.
>
>Then all *o's (there being no short *o) became phonologically short.
>In Indo-Iranian, in open vowels, they remained phonetically long long
>enough to merge with the new laryngeal and vrddhi long vowels. >That's
>Brugmann's law explained.
Which "first couple of centuries"???... Both long AND SHORT *o _DO EXIST_ in
IE (without a doubt, most definitely)... That is, unless you're planning on
wildly re-reconstructing common IE to suit your so-called needs. Haven't you
noticed that IEists (aside from Rasmussen, let's say) never write **wo:dr or
**okto- interchangeably with the actual forms *wodr and *okto:-? You
obviously need to get acquainted with how *wlkWos is declined because you
must not be aware that the gen.pl. is *wlkWo:m while the acc.sg. is *wlkWom,
bluntly showing the important contrast between *o and *o:!
To quote an old black-and-white sitcom, "Lucy? You goddaloddav esplainin'
too doo."
>>>That in any case. E.g. Slavic c^asU < *ke:s-.
>>
>>You forgot the palatal diacritic. I'm sure you mean *k^e:s-.
>
>Tsk, tsk.
Hey, are you tsking me? I wouldn't tsk. This is not your moment of tsk,
here. You can tsk all you want but it's you who's pretending that short *o
doesn't exist and supporting G&I and Rasmussen with a straight face. Tsk,
tsk :P
- gLeN
_____________________________________________________________________________________
Get more from the Web. FREE MSN Explorer download :
http://explorer.msn.com