The Great Proto-Indo-European Thread

[kanejam]

@Zaarin: You need the labial component so the first one is kinda out how else are you going to account for rounding. The second does not account for pharyngeal/lowering effects, which is why /ʕʷ/ is more likely. Other possibilities that I would speculate on would be /qʷʼ/ /xʷ/ or things like that, but I am pretty sure there are reasons why this is not so because I couldn't fathom why I would be the first one to think up of them.

Not at all; the labiovelars are pretty uncontroversially reconstructed with labialisation but didn't cause rounding at all. If, as in a few theories, /o/ and /e/ came from the same source, then chances are that /o/ was originally unrounded. Now factor in the point that /o/ and h₃-coloured /e/ didn't merge in Indo-Iranian (and thus can't have in PIE), h₃'s main effect seems to be backing or lowering rather than rounding.

Having said that, I do like the idea that h₃ was labialised, as it serves as an easy point of distinction from h₂ and I'm not convinced by the voicing evidence. I also like to think that they come from the same source and line up with the velar stops (with h₂ corresponding to the plain and palatovelars). My point is more that you probably shouldn't talk with authority about things you don't know, especially if you don't even believe in them.

But you know what I honestly believe is the case?

Yes, because you mentioned it before. And it is irrelevant. No-one is claiming what you are claiming is claimed. Since "strawman" seems to be your middle name, I'm not surprised.

I think JAL might be taking his disliking of you a little far (and should probably start ignoring you) but he definitely has a point. If you had a crackpot theory then you would be labelled a crackpot; as it is, you're simply bad at arguing.

Back on track, I've been reading a lot of Kortlandt and Kloekhort's work, mainly because there's a lot of it online and in English and thus accessible. While I wouldn't call myself an ardent follower of the Leiden school, they are doing a lot of work on internal reconstruction and Indo-Uralic stuff (WeepingElf, have you seen this article?).

Edit: is there any evidence that h₂ had or didn't have the voicing/glottalising effect of h₃?

[sirdanilot]

If you have a question to one of the people authoring those Leiden articles, I could ask them for you as I see these people in real life a lot . But I think most of them have their contact info available online and aren't afraid to answer questions by e-mail.

[jal]

I think JAL might be taking his disliking of you a little far (and should probably start ignoring you)

I don't dislike dani, I just can't stand strawmanning :). But yeah, I probably should shut up now, and let others handle it. Thanks for reminding me.


JAL

[WeepingElf]

Back on track, I've been reading a lot of Kortlandt and Kloekhort's work, mainly because there's a lot of it online and in English and thus accessible. While I wouldn't call myself an ardent follower of the Leiden school, they are doing a lot of work on internal reconstruction and Indo-Uralic stuff (WeepingElf, have you seen this article?).

Yes, I have read it - and many others from Kortlandt and Kloekhorst. They are interesting. The late, lamented Jens Elmegard Rasmussen also wrote lots of interesting stuff on internal reconstruction in PIE - too bad that he was called off to another world before he could come up with a comprehensive summary. So all we have are the articles strewn out over various journals, festschriften and other volumes.

[masako]

Relevant article from NATGEO

[jal]

Relevant article from NATGEO

Quite interesting, thanks.


JAL

[WeepingElf]

Relevant article from NATGEO

Quite interesting, thanks.

Yes. However, such attempts at finding genetic evidence for language spreads always have to be handled with care, as has already been discussed here. But the article shows that the "kurgan scenario" makes sense, and the ex negativo argument of some supporters of the Anatolian hypothesis is untenable.

[jal]

Yes. However, such attempts at finding genetic evidence for language spreads always have to be handled with care, as has already been discussed here.

Indeed. Though what has been discussed here, iirc, is that a language can spread faster than one would expect actual people to be displaced. In this case, we see that people can spread very, very quickly, and in this case have genetically displaced the original population. The hypothesis that the language they spoke prior to invading is equal to the language after invading is imho simpler than one that has the invaders adopt the local languages without any retention of their own languages, whilst displacing the entire original population.

What I'm missing from the article though is comparison of both the old and the new populations with modern populations. Also, if we can find Neanderthal DNA in our genes, I would find it highly suspect if no trace of the DNA of the original inhabitants can be found.


JAL

[kanejam]

What would be nice is a massive coherent tome on the subject, rather than needing to rifle through various semi-relevant articles. But I realise how much work that would be so for now I'm happy to read what is available.

Here's a go at a reconstruction based very closely on their material (not a lot of originality).

Fortis: /p t k kʷ ʔ/
Lenis: /b d g gʷ/
Glottalic: /ɓ ɗ ɠ ɠʷ/
Fricatives: /s x̠~ħ x̠ʷ~ħʷ/
Resonants: /l r w j m n/
Vowels: /a i u/
The fortes and lenes don't necessarily need to be voiced/unvoiced at an earlier stage of PIE, and the glottalics could well have been ejectives or some other glottalic consonant. I follow Kümmel that implosives explain the various daughter languages the best.

Most words are made up of a root (usually verbal; can be stressed or unstressed) plus a suffix. Nouns are either common or neuter gender (basically animate vs inanimate) and either have fixed or mobile stress (determined by whether the root is stressed or not). Some common nouns, one static and one mobile:

Code: Select all

nom ... ħáʔstar   ... ħáʔstars   ... máħtar   ... máħtar
acc ... ħaʔstárma ... ħaʔstármas ... máħtarma ... máħtarmas
loc ... ħaʔstár   ... ħaʔstári   ... máħtar   ... máħtari
abl ... ħaʔstarás ... ħaʔstarjás ... máħtaras ... máħtarjas
ins ... ħaʔstarát ... ħaʔstarját ... máħtarat ... máħtarjat

And some neuter nouns, one static and one mobile:

Code: Select all

nom ... jákʷan   ... jákʷans   ... míħwan ... míħwans
acc ... jakʷánma ... jakʷánmas ... míħwanma ... míħwanmas
loc ... jakʷán   ... jakʷáni   ... míħwan   ... míħwani
abl ... jakʷánas ... jakʷánjas ... míħwanas ... míħwanjas
ins ... jakʷánat ... jakʷánjat ... míħwanat ... míħwanjat

The difference between the neuter and common mobile nouns is that the common nouns like to stress the ending, whereas the neuter nouns stress the suffix. The words are star (PIE h₂stḗr, literally glow-er), mother (méh₂tēr), liver (yḗkʷr̥) and time (méih₂un). To get to PIE, simply delete all unstressed vowels, juggle the paradigms with analogy, introduce unstressed /a/ again and then /á/ becomes /e/ and /a/ becomes /o/.

I haven't got to verbs yet; the multiple full grade vowels are difficult to explain without resorting to analogy or saying that verbs were a series of separate words. I'm still toying with the idea that the vowels had a length distinction with /a:/ becoming /o/ but it leaves just as many gaps and questions as the above theory.

[Pabappa]

I love your reconstruction. It's nice to see people who still believe in the two-dorsal theory. But mostly I just like how much more beautiful it makes PIE words look when laryngeals are treated as normal consonants and the vowels are not pushed to the extreme of monotony (no pun intended).

That PDF looks familiar; youre probably teh same one who posted it last time but I remember reading it although I dont remember much of what was there.

[Fixsme]

Yes. However, such attempts at finding genetic evidence for language spreads always have to be handled with care, as has already been discussed here.

Indeed. Though what has been discussed here, iirc, is that a language can spread faster than one would expect actual people to be displaced. In this case, we see that people can spread very, very quickly, and in this case have genetically displaced the original population. The hypothesis that the language they spoke prior to invading is equal to the language after invading is imho simpler than one that has the invaders adopt the local languages without any retention of their own languages, whilst displacing the entire original population.

What I'm missing from the article though is comparison of both the old and the new populations with modern populations. Also, if we can find Neanderthal DNA in our genes, I would find it highly suspect if no trace of the DNA of the original inhabitants can be found.


JAL

The main problem I have with that genetic study is that it was specifically designed to show the dispersion from the Kurgan Area. They over sampled the Corded Ware individuals and the Yamna individuals to show that they are genetically related. The other populations are just to show that it was different.
For instance why isn't there comparison with Hittit individuals or with Tocharian individuals? Do Indian peoples have also Yamna genes? This would give a good point of view. The only point of comparison are with modern populations, with all the bias there is with newer migrations (Turkic...)
Moreover if I read correctly the graphs, the Yamna individuals are already a mixture between Yamna and local populations.

[KathTheDragon]

Returning to the putative voicing effect of *h₃, I've started reading Kortlandt's From Proto-Indo-European to Slavic, and he cites Russian pilá "(she) drank" < *ph₃iléh₂. If correct, that's evidence against *h₃ having a voicing effect on preceding voiceless stops, and evidence for a devoicing of word-initial *b (which Kortlandt assumes in the paper).

[hwhatting]

Returning to the putative voicing effect of *h₃, I've started reading Kortlandt's From Proto-Indo-European to Slavic, and he cites Russian pilá "(she) drank" < *ph₃iléh₂. If correct, that's evidence against *h₃ having a voicing effect on preceding voiceless stops, and evidence for a devoicing of word-initial *b (which Kortlandt assumes in the paper).

That's either a typo or Kortlandt is assuming some kind of inversion here - Common Slavic /i/, which you have here, goes back to Balto-Slavic /i:/ or /ei/, and to have a long /i:/ here, you need PIE *piH₃-leH₂. I think I have a print-out of that paper somewhere at home, but can you check - how does he explain the length if his starting point is *ph₃i?

[KathTheDragon]

He doesn't. The derivation appears exactly once in the paper, in a discussion on Hirt's Law. It's very clear from the context that the form he gives is exactly what he means.

The stress was not retracted if the laryngeal followed the second component of a diphthong, as in Latvian tiêvs ‘thin’ < tenh₂uós, or preceded the syllabic nucleus, as in Russian pilá ‘(she) drank’ < ph₃iléh₂.

Edit: Turns out he does. I just didn't find it because the transcription has changed.

Loss of the laryngeals in pretonic and post-posttonic syllables with compensatory lengthening of an adjacent vowel, e.g. ... pīlòH < pHilàH ‘(she) drank’

[hwhatting]

Edit: Turns out he does. I just didn't find it because the transcription has changed.

Loss of the laryngeals in pretonic and post-posttonic syllables with compensatory lengthening of an adjacent vowel, e.g. ... pīlòH < pHilàH ‘(she) drank’

Thanks! But that seems to be his own hypothesis, I haven't seen that assumed before. The usual assumption is metathesis of *pH₃-i- to *piH₃- in zero grade already in PIE in order to explain the forms with long /i:/ in the daughter languages (see e.g. the reconstructions in LIV pp. 462-463). Does he have more examples of this lengthening? And does he assume that for PIE generally or only for Balto-Slavic or Slavic? If the latter, it still wouldn't explain the long /i:/ in the other daughter languages.

[KathTheDragon]

He lists it as Early Slavic, and all his other examples are *VH. Note, however, that according to him, laryngeals in tonic and immediately post-tonic syllables are not lost at this point.

[hwhatting]

He lists it as Early Slavic, and all his other examples are *VH. Note, however, that according to him, laryngeals in tonic and immediately post-tonic syllables are not lost at this point.

OK, so it doesn't explain the other languages. And yes, I know that Kortlandt assumes late survival of the laryngeals (actually until about the break-up of Common Slavic), because he doesn't believe in tonal features on non-stressed syllables. That's why I'm always cautious if it's only him reconstructing a form, like in this case - he has his fixed ideas.

[KathTheDragon]

I'm cautious with the Leiden IEists in general, hence why I said "if correct" when I mentioned this word. Although, it seems like the only real evidence either way for this development is to be found in derivatives of this one root.

[hwhatting]

Although, it seems like the only real evidence either way for this development is to be found in derivatives of this one root.

Well, the other case is *ap-H3on- > *abH3on-, which is disputed as well (I linked to that before). A sound law based on two cases will always be shaky.

[sirdanilot]

I'm cautious with the Leiden IEists in general, hence why I said "if correct" when I mentioned this word. Although, it seems like the only real evidence either way for this development is to be found in derivatives of this one root.

Yes you should be cautious. They are good, but they are also very conservative and traditional in their ways of thinking and do not like to diverge from the way of thinking they have acquired at some point in time. This really goes for many groups at Leiden, really, as Leiden is tradition-wise the most conservative university of the country by far. For example promotion ceremonies are still held in the ancient traditional way (which is good, but being overly conservative also has some downsides).

[jal]

Conservatism in science is dangerous in general, as it easily leads to dogmatism.


JAL

[Pogostick Man]

Does s-mobile occur with verbs or is it restricted to nouns?

[Chagen]

Yes, cf. Latin speciō and Sanskrit paśyati. Actually, I'm pretty sure s-mobile is mostly found on verbs, but given that verbal roots are basically the bedrock of a vast amount of PIE vocab, seeing it on nouns shouldn't be too surprising.

[Terra]

Does s-mobile occur with verbs or is it restricted to nouns?

It can occur on either.

For example promotion ceremonies are still held in the ancient traditional way

What is the ancient traditional way?

***

Btw, speaking of Leiden, does anybody have Leiden's Greek etymological dictionary?

[Sleinad Flar]

Yes, http://www.uz-translations.net/ has got it. Just search for "Beekes" and it'll show up. You need to log in to see the download link. However, I cannot guarantee that the link is still working (it's been a while since I got it there).