The Great Proto-Indo-European Thread

[Chagen]

Anyway, the reason I came to this thread tonight was to post something amusing I found trying to find the PIE word for 'colo(u)r'. I did a quick search on this board and it doesn't look like anyone's mentioned it before.

http://www.academia.edu/3287583/PIE_Roo ... _Code_2.0_

This document posits the concept that all PIE roots were phonosthemic, i.e., that each phoneme in a word has an associated meaning. For example, initial /*p/ means 'body' / 'accumulation', and final /*n/ means 'complete' / 'end' / 'sound'.

Man as stupid as this sounds this would be an awesome idea for a robot conlang.

Also I like how each phoneme has a different meaning initially and finally. Sometimes with complete opposites for the same phoneme too!

[KathTheDragon]

I think conlangs are the only thing quackery is good for.

[Tiamat]

I'm not sure if this is exactly appropriate for this thread but does any know of any good resources on the development from PIE to Proto-Indo-Iranian as well as a dictionary for the latter? I have been looking and the best I could find are the relevant sound changes, some comments on morphological development, a dictionary of Proto-Iranian verbs (If anyone wants that I can send a link).

[KathTheDragon]

Does anyone know where I can find a electronic copy of the LIV? Failing that, somewhere where I can get it moderately cheaply.

[Sleinad Flar]

Here ya go: http://lib.freescienceengineering.org/v ... ?id=631214

(Check that site out, lots of goodies!)

[KathTheDragon]

Thanks a bunch!

*bookmarks*

Edit: Oh crap, now I have to brush up on my German linguistics terminology... Halp?

[Chagen]

I'm slightly reviving this thread to talk about a bizarre set of correspondences in Sanskrit and Greek.

Their words for "daughter" are, respectively, duhitar and thugater. They don't seem to match up. The Sanskrit word implies something like *dug^hh₂ter. But the Greek form points to something like *d^hugh₂ter. *dug^hh₂ter. should give Greek *dukhater. On the contrary, *d^hugh₂ter should give Sanskrit *dhugitar or even *hugitar! Grassman's Law would allow the Sanskrit reflex to come from *d^hug^hh₂ter, but the law also applies in Greek and would give *tukhater...? The English reflex suggests an aspirate for the first sound, though.

There must be some sound law I'm missing, or one or both of these were borrowed from a dialect. I'm not too familiar with how the Greek sound changes work together, I'll admit. This had just been bugging me for a while.

[Pole, the]

Their words for "daughter" are, respectively, duhitar and thugater. They don't seem to match up. The Sanskrit word implies something like *dug^hh₂ter. But the Greek form points to something like *d^hugh₂ter. *dug^hh₂ter. should give Greek *dukhater. On the contrary, *d^hugh₂ter should give Sanskrit *dhugitar or even *hugitar!

Does first, *g →*gʰ before a laryngeal, and then, dʰ →d / _VCʰ, make sense?

[Chagen]

Possibly but I dunno if that happened in PIE or not.

[KathTheDragon]

Those are developments in Sanskrit.

[sangi39]

Those are developments in Sanskrit.

Wikipedia and Wiktionary at least seem to back this up, with traditional PIE *g becoming *gʰ before a laryngeal.

[Salmoneus]

Yep, pretty straightforward, really.

[Pole, the]

And now for something different…

First, I have seen somewhere in this thread a link arguing that the ablaut *e : *o could come from an earlier *o : *ō alternation. Second, there is this Brugmann's law showing that one of the branches of IE had *o → ā / _$ — but only for the ablauting instances of *o.

Do you think these two ideas work together? I mean, the *e : *o apophony could just reflect something like earlier [ə] vs. [əː], with the length of the latter having a retention in the Indo-Iranian languages. The original pre-ablaut vowel inventory could be then plain [i u e ə o a] with an abundance of schwas. Is it plausible?

[TaylorS]

And now for something different…

First, I have seen somewhere in this thread a link arguing that the ablaut *e : *o could come from an earlier *o : *ō alternation. Second, there is this Brugmann's law showing that one of the branches of IE had *o → ā / _$ — but only for the ablauting instances of *o.

Do you think these two ideas work together? I mean, the *e : *o apophony could just reflect something like earlier [ə] vs. [əː], with the length of the latter having a retention in the Indo-Iranian languages. The original pre-ablaut vowel inventory could be then plain [i u e ə o a] with an abundance of schwas. Is it plausible?

I actually think the same thing.

[KathTheDragon]

I like this idea.

*Promptly goes to her notes and tries out this idea.*

[Chagen]

Those are developments in Sanskrit.

Wikipedia and Wiktionary at least seem to back this up, with traditional PIE *g becoming *gʰ before a laryngeal.

Well this is embarrassing. Then again, syllabic lanryngeals are rather uncommon in PIE so I don't have much experience with them.

They certainly had some bizarre distribution though. The laryngeals appear to be able to go ANYWHERE in a root so we have things like ghrebhH-, h₂ner, and h₃meigh-. The only other fricative, /s/, certainly couldn't appear in as many positions, such as next to voiced stops unless ablaut forced it to be. Never mind that all four were fricatives, but only the laryngeals could be syllabic, and their pronunciation as vowels, unless the syllabic liquids, had zero to do with the original consonants (being some kind of schwa, I think), though Greek suggests that they actually were never really true syllabics and just had a schwa in front of them (later colored in Greek).

Man PIE was just weird as hell in general.

[Tiamat]

though Greek suggests that they actually were never really true syllabics and just had a schwa in front of them (later colored in Greek).

I don't see how greek's reflex for the syllabics necessitates this, especially since other IE languages treatment points to them being true syllabics.

[Richard W]

I don't know where the idea that apophonic *o was originally long comes from, but Miguel Carrasquer floated it on Cybalist years ago and nobody shot it down.

[WeepingElf]

I don't know where the idea that apophonic *o was originally long comes from, but Miguel Carrasquer floated it on Cybalist years ago and nobody shot it down.

At least, nobody can say that this was not the case. Brugmann's Law may suggest that the *e/*o distinction once was a distinction of length, with *a > *e and *a: > *o. Why not?

[Pabappa]

Isnt /o/ generally associated with unstressed syllables though? It would be weird for there to be a rule that lengthens only the vowels in *un*stressed syllables. e.g. πατήρ "father" changes to πάτωρ with the change of stress causing the /e/ to change to /o/.

See also https://en.wikipedia.org/wiki/Kwetwores_rule

[Salmoneus]

It's possible that the alternating stem e/o 'o' had a different origin from the stress-related 'o', perhaps? The fact that the stem alternation e/o is seen alongside e/e: in other roots does led credance to the idea of a long 'o'; on the other hand, as you say, the fact that 'o' is usually the unstressed post-stress version of 'e' militates against it being long. So maybe it was both? And then there are roots with 'o' in them all the time, many of which look to me to be in plausible vowel-rounding/backing situations.

Perhaps then PIE /o/ has three pre-PIE sources?
If you start out with only /a/ (let's ignore the high vowels for now), it could be lengthened to /a:/ in some contexts. Meanwhile, it could be reduced to /@/ in others, and rounded to /Q/ in others. Then let /a/ raise to, say, /{/, but /a:/ raise to, say, /O:/, but remain /{:/ in certain contexts - all common enough. Schwa could then back (/@/ > /V/ > /O/), and /Q/ could rise, giving us /{/, /{:/, /O/, /O:/. Then merge /O/, /O:/, and we get the PIE system of *e, *e:, *o.

[2+3 clusivity]

PIE tech question, did pie have common terms for grapes/wine, grains/beer, honey/mead? And were the precoursers/products distinct if PIE had common terms?

[Sleinad Flar]

Grapes/wine, no: 'Wine' is THE proto-typical 'Wanderwort'; a loanword which is gained with the substance itself.

Grains/beer: yes. And distinct: *g^rh₂nó- and *h₂élut- or the like. I'm unsure how far-spread the 'beer' word is; it might be just European. Also: there were several 'grain' words, basically for each type of grain (*Hat-, *yewo-, *g^^hérd^h- etc.). Not surprising for an agricultural society.

Honey/mead: yes, semi-distinct: *mélit- and *méd^hu-. The latter both means 'mead and 'honey', while the former (AFAIK) is only 'honey'.

Adams and Mallory (Oxford Introduction to PIE or Encyclopedia Of Indo-European Culture) have the answer to these kind of questions. There's probably a digital copy at the link I provided earlier. Insta-edit: Well yes, there is: http://lib.freescienceengineering.org/s ... it=Dig+for

[2+3 clusivity]

Oh nice! Thanks so much.

[KathTheDragon]

Going back to the idea of ablauting *e ~ *o originally being something like **ə ~ **ə:, I've just worked through my notes, making the appropriate substitutions, and it turns out that much of the PIE ablaut system can be trivially derived from schwas everywhere. Even the ablaut of the thematic vowel in nouns becomes transparent.