From: Piotr Gasiorowski
Message: 2100
Date: 2000-04-12
----- Original Message -----From: Glen GordonSent: Wednesday, April 12, 2000 2:23 AMSubject: [cybalist] The long awaited athematic answer to the athematic question... Oy veh.
Glen writes:
Early Anatolian and Tyrrhenian languages attest to the fact that the *-s nominative was not as widespread as it later would come to be. There is also the matter of IE compound word formations which attach a bare stem to another before terminating with suffixes, as if the bare stem was the earlier state of affairs. Inanimates also don't have this *-s marker and some have no endings at all. If we can accept that an earlier IE had no animate nominative *-s, then we must also accept that action nouns lied bare, identical to the root (ie: IE *geno-s < Old IE *k:ene, noun and 3ps).
Another reason to accept this is that the athematic Tocharian 3rd person appears to be /-a"s./, from a demonstrative *se according to Pedersen, attached to a BARE root! Comments?Glen,
Whoever told you that the animate *-s nominative wasn’t as widespread in the ‘early Anatolian’ languages as in any of the later branches was a bloody liar. You can do what you like (within the bounds of reason) when venturing a deeper reconstruction, but there are comparative constraints on what may legitimately be claimed for PIE. What you say is simply untrue of PIE as reconstructible with any methods known to me.
There’s a single termination in compounds just because they are compounds (combined stems), and not phrases (strings of words). It’s a universal tendency, and not an IE curiosity. The very existence of compounds is motivated by the wish to avoid redundancy in frequently occurring collocations. ‘As if the bare stem was the earlier state of affairs’ is just an ‘as if’ clause, not a legitimate conclusion.
What’s this *geno-s ‘action noun’? I hope you don’t mean the genus/genos/janas word, which is neither an action noun, nor animate.
I’m not sure what the Tocharian form is supposed to prove. It doesn’t even go back to Proto-Tocharian (Tocharian A pikäš ‘writes’ = Tocharian B pinkäM; A lukäš ‘illuminates’ = B lukšäM).
Second, a reduplicative like *bhi-bher-ti or even the late IE past tense in *e- is a derived form and therefore lacks thematic - so it is already accounted for. So is *wemh- which is not necessarily a derivative but rather assumed as such by IE speakers who analysed it as *wem-h- (also *leikW- misanalyzed as *lei-kW-). Theories that claim that the thematic is an ancient object marker don't even begin to explain this above pattern well at all. Why should the "object marker" be lost in the past tense or in these derived forms? Sounds far fishier than my idea. Let's not confuse the matter with inane side theories.I don’t support that theory; I just cited it amongst other historical attempts to ‘explain’ the thematic vowel. There have been quite a few of them, also along the lines you propose (though without the Indo-Tyrrhenian and Steppe backgroud, as far as I know). You’re very quick at calling other peoples’ theories inane or quack, without as much as having a look at what they actually wrote.
Third, there is a misunderstanding on your part and so I will try to clarify: The _athematic_ ACCENTUATION is the most ancient BUT the _thematic_ VOWEL is also most ancient. A verb lacking thematic vowel is NOT ancient
even though its accentuation is. In all, an athematic verb is either derivative in some form or fashion or is foreign. The adoption of athematic verbs would have occured before the regularisation of accent on thematic stems. Hope that makes better sense now.It does. Thanks.
The mobility of accent in *nebhes- is the proper result of my theory, by the way, since the stem is athematic and the idea that Semitic *napištu "soul" was loaned into early IE with the meaning of "cloud" or "mist" is hardly
speculative on both semantic and phonetic grounds. To boot, the strange nature of the root all the more cements the likelihood of this view (a disyllabic indivisable root -> rather long for an ancient word; a root form ending in *-s for no apparent reason; a stem that could be so well explained in Semitic just like *septm, if only those crusty IEists would accept the concept of foreign loans in early IndoEuropean, which surely must have occured.)*nebHes- itself isn’t mobile; I just said that there were OTHER archaic-looking mobile *-(e)s neuters, such as those functioning as infinitives in older Vedic, or as Hittite ais, Gen. issas ‘mouth’.
Not speculative on semantic grounds? You must be kidding, or have a very liberal attitude towards semantic derivations. This particular one is NOT IMPOSSIBLE, but still SPECULATIVE. The neuter *nebHos (*nebHes-) isn’t strange at all; we also have *gen(h)os, *nemos, *wekWos, etc., all with the same vocalism – a highly productive pattern, in fact. To be sure, there’s no *nebH- directly attested as a verb root, but this may be an accidental gap, since *nebhos is not alone in IE. We have *nebH-(e)lo- (as in nebula), and the zero grade in *mbH-ro- (Greek aphro-, Latin imber). The semantic range of all these words is clearly meteorological: cloud, sky (Hittite, Slavic), mist, foam, rain. Something like ‘mist/cloud over’ would work better as the prototypical meaning.
I'm not concerned with whether "eye" and "heart" were neuters in Late IE since IndoAnatolian didn't have a "neuter" gender, Piotr. We're talking about whether the forms are animate or inanimate in this stage of IE. I could swear I had a similar discussion with people on another IE list concerning the _animacy_ of "heart". I had assumed that "heart", at least, was "inanimate" because they showed up as neuters lacking accusative *-m in Late IE as you point out. To my surprise, I was hearing the direct opposite. I recall something like Hittite /karatan dai/ which seems to display an
_animate_ form with *-m. Is it possible that you are getting two stages of IE confused? What is your explanation for the infix -t- in Greek opt- if it isn't from IE? Certainly if *-t- did once exist, there would be phonetic motivation to lose it in the nominative and then subsequently throughout the paradigm but I don't see a motivation to add a -t- in Greek.
Here's my table again, just in case you "rejected it out of hand":
OIE *merec-/*merece "remember/(he) remembers"
MIE *merc-/*merc(e)
IE *mers-/*mers-t,*mers-ti
OIE *hWegWec/*hWegWet:em "eye (nom/acc)"
MIE *hWegW(e)c/*hWegWtem
IE *hWekWt-s/*hWekWtm
OIE *kerec/keret:em "heart (nom/acc)"
MIE *ker(e)c/*kert:em
IE *kert-s/*kerdmGlen, the IE part of these reconstructions is absolutely indefensible, pace Glen Gordon [above]. Consequently, it would take a leap of faith to trust the deeper reconstructions. Instead of changing or modifying your examples so that they can at least be seriously discussed you just give me the same old rubbish again. Is this an application of the Bellman’s ‘rule of three’? (‘I have said it thrice: / What I tell you three times is true.’) If so, I’m not going to wait for another presentation. I gave you a fair warning the first time, and feel free to reject the table out of hand, right now.
Also, Piotr, the perfect/imperfect thing that you're confused about is exactly what's found in Encyclopaedia Brittanica concerning the IE division of verbs into "imperfect" (*leikW-), "perfect" (*steH-) and "stative" (*es-,
*woid-). Stative verbs could neither be perfect nor imperfect. However, imperfects could be made perfects and vice versa. I often call the imperfect "active" and the perfect "stative" still acknowledging the special nature of
verbs like *es- and *woid-. The active/stative (or imperfect/perfect) mirrors the mi-class and hi-class system of Hittite and it's evident that the imperfects and perfects of late IE languages such as Sanskrit and Greek
are derivative from a system that classified verbs into such an active/stative system. This is nothing new so don't fly off the handle at me about it. Ask your colleagues.Piotr
I’m not confused, Glen. The EB presentation of the aspect system is deceptively short, but you didn’t read it with care. The author attempts to explain too much in three short paragraphs. He fails to draw a clear distinction between ‘aspect’ as an inherent semantic category (the nature of action associated with a given root), ‘aspect’ as expressed with lexical means (as in derived iterative, inchoative or ‘progressive’ stems), and ‘aspect’ as a paradigmatic category (the traditional present/aorist/perfect = imperfective/perfective/stative triad). At any rate, he is careful enough to emphasise the distinction between ‘perfect’ and ‘perfective’ (something YOU confuse), and besides he doesn’t classify *es- as ‘stative’ but as ‘imperfective’ (which it should be, the Hittite form being es-mi).