Talk:Europic

Europic and Afrasian

If the speakers of Europic were Neolithic farmers who came to Europe from the Near East, we should expect a close relationship between Europic and Afrasian (aka Afro-Asiatic), as implictly suggested by Vennemann, who proposed an "Atlantidic" (aka "Semitidic") substrate to explain some Germanic words of non-IE origin.

There's also the fact most Afrasian languages have a 3-vowel system like the one proposed for Europic, as the result of a collapse like the one proposed by Orel and Stolbova (1995), in which vowels e, o developed into i, u or the corresponding glides j, w. Then I assume the GVC was like this and not the way described by Jörg.

I agree with Jörg in the development from Europic to PIE, except in that Europic *a didn't developed into IE Ablaut vowel *e ~ *o in OEH and Indo-Iranian. So there's actually no need for a second vowel collapse. Also the reason why i, u didn't appear before resonants (in that case, the resonant shifted to the syllable onset) is that they were actually semivowels, so there's also no need for the RCL. Talskubilos 05:48, 17 July 2012 (PDT)

---

There are also (rare) cases of alternation of type *e ~ *i and *o ~ *u. I would imagine the Pre-PIE system of a type:

High	i	ə ~ 0	u
Low	e	a	o

With a height opposition and a schwa being allophone of the zero grade. Then they merge: [e] with [ə] to *e and [a] with [o] to *o. How much probable would that be? MilyAMD 13:03, 16 July 2012 (PDT)

---

@Talskubilos:

I cannot say you are wrong, but I don't consider Afroasiatic a likely candidate for the next closest kin of Europic.

@MilyAMD:

No. You'd get */i/ ~ */e/ and */u/ ~ */o/ "ablaut" alternations that aren't observed in PIE, at least not in the same function as the familiar */e/ ~ */o/ ablaut. These alternations are a myth spread by the late Joseph Greenberg, nothing else. He tried to connect IE ablaut to vowel harmony systems found in languages of eastern Siberia, but this is impossible, especially considering that IE ablaut is not a vowel harmony system of any kind (you'd expect all morphemes to be in the same grade if it was, which is not the case).

--WeepingElf 13:22, 16 July 2012 (PDT)

@WeepingElf:

I think you've got some pre-conceived ideas about the subject. Perhaps if you knew more about Afrasian, you'd change your opinion. Talskubilos 15:16, 16 July 2012 (PDT)

---

@Talskubilos:

That's your opinion. --WeepingElf 06:10, 17 July 2012 (PDT)

---

Which is based on actual data. You also might notice I changed somewhat my views since yesterday. Talskubilos 08:02, 17 July 2012 (PDT)

---

So is mine. The difference is that yours is based exclusively on lexicon, which admits the hazard of being misled by loanword layers, while mine is based on morphology, especially matches between entire morphological paradigms, where loaning is unlikely. But it is just opinion vs. opinion, I have to admit. --WeepingElf 08:45, 17 July 2012 (PDT)

---

Ah! but these "loanwords layers" must be studied anyway, something which very few comparative linguists have done.

Also the problem with morpohology is that it isn't very stable over long periods of time (i.e. millenia), so it's of little value in long-range comparisons. Talskubilos 12:41, 17 July 2012 (PDT)

---

Sure, morphology can change a lot over long periods of time - but so does lexicon. I wouldn't say that morphology was "of little value" in long-range comparisons, to the contrary: it often provides the best evidence. Many language families were first established by morphological comparison; Indo-European is an example. WeepingElf 13:04, 17 July 2012 (PDT)

---

Ah, but the example you quoted isn't precisely long range. We're talking about language families seprated by millenia. Talskubilos 13:54, 17 July 2012 (PDT)

---

I don't see why a greater time depth requires a different methodology, though of course things bceome more difficult the deeper one is trying to look. --WeepingElf 02:21, 19 July 2012 (PDT)

---

This is precisely why morphology is quite useless on time depths of 10,000 years or more. Also regarding the GVC, I almost forgot about NWC (Abkhaz-ADyghe), which has a 2 vowel system a, @ together with an exteremly rich consonant inventory. At an early stage, front vowels caused the preceding consonant to be palatalized (Ce/Ci > Cj@) while back vowels labialized it (Co/Cu > Cw@).

It's conceivable some similar processes took place in Kurganic (the language of Kurgan people), which after all was in contact with NWC (leaving loanwords such as the numeral '2'), leading to the formation of so-called "labiovelars" and "palato-velars", although the latter could also arise (by contrast to Europic) from palatal consonants. Also possibly the IE Ablaut vowel e ~ o evolved from @. Talskubilos 03:40, 19 July 2012 (PDT)

---

Two points, one in which I disagree with Talskubilos, one in which I basically agree.

On the alleged unsuitability of morphology to long-range comparison:

Morphology is not quite useless on time depths of 10,000 years. It was morphology by which Afroasiatic was established; the lexical evidence actually doesn't really look that good (there are two mutually incompatible reconstructions - one by Ehret and one by Orel and Stolbova - which cannot both be right and may both be wrong, and indeed have similar problems as the two major reconstructions of Nostratic, or, for that matter, the Etymological Dictionary of the Altaic Languages by Starostin, Dybo and Mudrak). Lexical resemblances can easily arise by chance, and allow people like Arnaud Fournet to "prove" that Basque and Hurrian are Indo-European languages! In both cases, it is the morphology which shows the wrongheadedness of these attempts by utterly refusing to match.

On the GVC and the NWC developments:

Yes, this kind of thing apparently happened in pre-Proto-Europic. A velar next to a front vowel became a front velar ("palatovelar"), a velar next to a rounded vowel became a labialized velar. This is indeed the most likely origin of the three velar series (which Uralic and other Mitian languages lack; all three IE velar series appear to correspond to Uralic velars), and preserves information on the pre-Proto-Europic vowels.

Here is an example: Proto-Mitian *kulV- 'to turn' (with apparent cognates in Uralic and Altaic) > *kolV- (RCL) > *kʷolV- (velar affection) > *kʷala- (GVC) > PIE *kʷel- (ablaut).

An areal or ad/sub/superstratal connection with the developments in NWC is not at all unlikely. These languages were neighbours.

--WeepingElf 09:45, 19 July 2012 (PDT)