WeepingElf's Europic thread

[WeepingElf]

If there are similar sounding 'wheel' words in NWC and NEC languages, they are certainly borrowings from Indo-European languages for two reasons:

1. The etymology of the PIE word is so clear that it can't be borrowed from any language.

PIE *kWel- could be a loanword or Wanderwort prior to the invention of the wheel. I don't say it's necessarily so, but a priori nothing prevents it from happening.

2. The people in the Pontic steppe probably had wheeled vehicles before the people in the valleys of the Caucasus had them.

This is at least questionable. If I'm not mistaken, the root was invented in Mesopotamia, where neither North-Caucasian nor PIE were presumably spoken.

You mean, the wheel was invented there. Of course, neither IE nor anything related to "North Caucasian" was spoken there. (Some scholars maintain the hypothesis that Hurrian-Urartean was related to NEC, but that is not widely accepted.) That does not necessarily mean that the word came from there.

AFAIK, PNC has 3 roots for wheel and wheeled vehicles:

1) *hwǝlkwē 'carriage, vehicle; wheel'. Found in Basque orga 'carriage' and possibly connected to PIE *H4welk- 'to pull'. I associate it with solid wooden-wheeled carts driven by oxen.

Basque orga doesn't look much like *hwǝlkwē, does it? PIE *H4welk- does indeed look quite similar, though.

2) *ʡwilʡa (˜ -ʕ-,-ǝ,-ɨ) 'wheel' (only NEC). Found in Basque -bil 'round' and possibly connected to PIE *wel- 'to turn, wind, roll' (Latin volvō). I associate it with spinning wheels (millstone, spindle).

I wouldn't say that this etymology was impossible; the forms are indeed not dissimilar, and this time you adduce only NEC, which is an accepted entity. But the question of the direction of borrowing remains.

3) *tɬwɨ̄ri / *rɨ̄tɬwi 'wheel, vehicle' (only NEC). Possibly connected to PIE *retH2- 'to run' ~ *rótH2o/eHa- 'wheel'. I associate it with spoked-wheeled carts driven by horses.

Once again, the resemblance is poor.

[Octaviano]

2. The people in the Pontic steppe probably had wheeled vehicles before the people in the valleys of the Caucasus had them.

This is at least questionable. If I'm not mistaken, the root was invented in Mesopotamia, where neither North-Caucasian nor PIE were presumably spoken.

You mean, the wheel was invented there. Of course, neither IE nor anything related to "North Caucasian" was spoken there. (Some scholars maintain the hypothesis that Hurrian-Urartean was related to NEC, but that is not widely accepted.) That does not necessarily mean that the word came from there.

Yes, I think Hurro-Urartian is Vasco-Caucasian, but I don't think PIE *kWekWlo- was borrowed from that family.

1) *hwǝlkwē 'carriage, vehicle; wheel'. Found in Basque orga 'carriage' and possibly connected to PIE *H4welk- 'to pull'. I associate it with solid wooden-wheeled carts driven by oxen.

Basque orga doesn't look much like *hwǝlkwē, does it?

First-sight resemblance isn't relevant, but regular sound correspondences.

PIE *H4welk- does indeed look quite similar, though.

I see it as a possibility.

2) *ʡwilʡa (˜ -ʕ-,-ǝ,-ɨ) 'wheel' (only NEC). Found in Basque -bil 'round' and possibly connected to PIE *wel- 'to turn, wind, roll' (Latin volvō). I associate it with spinning wheels (millstone, spindle).

I wouldn't say that this etymology was impossible; the forms are indeed not dissimilar, and this time you adduce only NEC, which is an accepted entity. But the question of the direction of borrowing remains.

I think this is out the question. IE borrowings into NEC are easily detectable thanks to phonetics. This isn't the case of *ʡwilʡa, with no less than two epiglottalic stops!

3) *tɬwɨ̄ri / *rɨ̄tɬwi 'wheel, vehicle' (only NEC). Possibly connected to PIE *retH2- 'to run' ~ *rótH2o/eHa- 'wheel'. I associate it with spoked-wheeled carts driven by horses.

Once again, the resemblance is poor.

Once again, it's regular sound correspondences what matter.

[Morrígan]

First-sight resemblance isn't relevant, but regular sound correspondences.

How very reasonable of you to say that. But please, show, don't just tell.

[Octaviano]

PNC *hwǝlkwē > orga:

PNC *h > Basque h or zero
PNC *wǝ > Basque o
PNC *lk > Basque rg
PNC *wē > Basque a

====================================
PNC *rɨ̄tɬwi > PIE *retH2-

Although at first sight it seems a reasonable hypothesis, there're some problems here. Firstly, one should expect a long vowel on the PIE side (e.g. *reH1- instead of *re-). Secondly, the correspondence between PNC *tɬ and PIE *tH2 is "irregular" in the sense that it doesn't belong to the set of the ones proposed by Starostin.

However, regional IE *reidh- 'to ride' fits nicely. Thus the question of a possible relationship between this root and *retH2- remains open.

[alice]

PNC *hwǝlkwē > orga:

PNC *h > Basque h or zero
PNC *wǝ > Basque o
PNC *lk > Basque rg
PNC *wē > Basque a

How many other words do you have for each of these sound changes? Are they securely attested, i.e. regular and typical, throughout the evolution of PNC to Basque? In particular, how does the last one work - surely there are intermediate stages?

Without such reassurances, its about as helpful as saying that etymology comes from entomology through loss of the first nasal and change of /o/ to /i/ by dissimilation, and some sort of semantic explanation based on the fact that insects don't have noses.

[Octaviano]

How many other words do you have for each of these sound changes? Are they securely attested, i.e. regular and typical, throughout the evolution of PNC to Basque?

The most important is the progressive labialization of a vowel due to a preceding labial glide *w (*CwV > *Co). There's also exists regressive labialization due to a following labial glide (*CVCw > *CoC).

I attribute this evolution to the group/branch I call Tyrrhenian. Only part of the Basque lexicon is of that origin.

In particular, how does the last one work - surely there are intermediate stages?

I think there's a backing of the former *e into a. But in general, last vowel isn't significative for comparative purposes, as there're other competing factors like analogy with the surrouding Romance languages, for example.

[WeepingElf]

How many other words do you have for each of these sound changes? Are they securely attested, i.e. regular and typical, throughout the evolution of PNC to Basque?

The most important is the progressive labialization of a vowel due to a preceding labial glide *w (*CwV > *Co). There's also exists regressive labialization due to a following labial glide (*CVCw > *CoC).

I attribute this evolution to the group/branch I call Tyrrhenian. Only part of the Basque lexicon is of that origin.

That looks like another attempt at overriding etymological problems to me. Some Basque words show the "wrong" sound correspondences, so they are declared loanwords from yet another branch of "Vasco-Caucasian". Easy. Hell, you can "prove" anything with such leaps of logic! And you haven't answered bricka's question yet. He asked for more data which shows that the change is regular.

In particular, how does the last one work - surely there are intermediate stages?

I think there's a backing of the former *e into a. But in general, last vowel isn't significative for comparative purposes, as there're other competing factors like analogy with the surrouding Romance languages, for example.

Never heard of "analogy with surrounding languages". What do you mean by that? Analogy happens between elements of the same language. Languages usually don't care a heck about other languages in their diachronic phonological development.

The sound change itself looks reasonable, but as long as you don't present at least a dozen other words displaying the same change, it remains an ad-hoc change which doesn't prove anything.

Oh, and let's not forget that this is not Octaviano's Vasco-Caucasian thread. This is WeepingElf's Europic thread, and I don't see in which ways this is relevant to the Europic hypothesis.

[Octaviano]

I attribute this evolution to the group/branch I call Tyrrhenian. Only part of the Basque lexicon is of that origin.

That looks like another attempt at overriding etymological problems to me. Some Basque words show the "wrong" sound correspondences, so they are declared loanwords from yet another branch of "Vasco-Caucasian".

I redirect you to the relevant thread.

My interest here is primarily about 'wheel' words and other linguistic devices which can be used to date PIE and hence prove or disprove your hypothesis.

Never heard of "analogy with surrounding languages". What do you mean by that? Analogy happens between elements of the same language. Languages usually don't care a heck about other languages in their diachronic phonological development.

Basque has been immensely influenced by its Romance neighbours, as in the antiquity it was by Celtic and Latin.

[WeepingElf]

I attribute this evolution to the group/branch I call Tyrrhenian. Only part of the Basque lexicon is of that origin.

That looks like another attempt at overriding etymological problems to me. Some Basque words show the "wrong" sound correspondences, so they are declared loanwords from yet another branch of "Vasco-Caucasian".

I redirect you to the relevant thread.

My interest here is primarily about 'wheel' words and other linguistic devices which can be used to date PIE and hence prove or disprove your hypothesis.

Sure. In my opinion, the words in question show that PIE did not break up before the invention of the wheel. I repeat that the PIE etymology of *kWekWlos is so secure that it certainly wasn't borrowed from anywhere, unless of course the verb root *kWel- is such a borrowing, which is unlikely - the meaning is too basic and there appear to be cognates in Uralic and Altaic, if these families are related to Indo-European (which is not proven yet, but likely).

Never heard of "analogy with surrounding languages". What do you mean by that? Analogy happens between elements of the same language. Languages usually don't care a heck about other languages in their diachronic phonological development.

Basque has been immensely influenced by its Romance neighbours, as in the antiquity it was by Celtic and Latin.

It borrowed many words from them, yes. But it did not undergo the same sound changes.

[Octaviano]

My interest here is primarily about 'wheel' words and other linguistic devices which can be used to date PIE and hence prove or disprove your hypothesis.

Sure. In my opinion, the words in question show that PIE did not break up before the invention of the wheel.

I agree if you mean "Post-Anatolic PIE", that is, the proto-language after the split of Anatolian.

Basque has been immensely influenced by its Romance neighbours, as in the antiquity it was by Celtic and Latin.

It borrowed many words from them, yes. But it did not undergo the same sound changes.

Not only that, but they have also influenced morphology.

[WeepingElf]

My interest here is primarily about 'wheel' words and other linguistic devices which can be used to date PIE and hence prove or disprove your hypothesis.

Sure. In my opinion, the words in question show that PIE did not break up before the invention of the wheel.

I agree if you mean "Post-Anatolic PIE", that is, the proto-language after the split of Anatolian.

I know that many items of Indo-European reconstructed core vocabulary are missing in Anatolian, but I don't think it split off earlier than about 4500 BC. Hittite at least has yukan < PIE *yugom 'yoke'. What did the Proto-Indo-Europeans do with yokes before they had wheeled vehicles? I doubt that *yugom is older than the invention of the wheel.

Also, Black Sea Flood refugees from the area where now is the Bay of Odessa are not very likely to end up in Anatolia. They couldn't travel into the Aegean by boat because the Bosporus (which is still difficult to navigate today) was either a huge waterfall or at least an enormous torrent they would have had to navigate against: impossible. They would have seen to get away from the rising sea as fast as they could, which would have meant boating up the great rivers that flow to the Black Sea. Not even the most foolish of them would have thought to sail against the very torrent that was destroying their homeland.

Of course, it is possible that later, descendants of Black Sea Flood refugees might move south from the lower Danube.

But the strongest argument against a separation of Anatolian and the rest of IE at the Black Sea Flood is in my opinion that Anatolian is not quite archaic as the language of the Old European hydronymy. Anatolian already has ablaut; the OEH shows no trace of it. Of course, one could imagine that Proto-Europic was yet earlier, but I consider a connection of the western branch of Europic with the Central European Neolithic the likelier option.

Basque has been immensely influenced by its Romance neighbours, as in the antiquity it was by Celtic and Latin.

It borrowed many words from them, yes. But it did not undergo the same sound changes.

Not only that, but they have also influenced morphology.

I don't see how that strengthens your point. You claimed that some Basque words converged with their Romance counterparts by way of "analogy" between languages - and that is and remains to be nonsense. Or did I misunderstand?

[Octaviano]

I know that many items of Indo-European reconstructed core vocabulary are missing in Anatolian, but I don't think it split off earlier than about 4500 BC. Hittite at least has yukan < PIE *yugom 'yoke'. What did the Proto-Indo-Europeans do with yokes before they had wheeled vehicles? I doubt that *yugom is older than the invention of the wheel.

It's always possible that it originally referred to another thing and then shifted its original meaning to be applied to a new invention. Meaning shifts of this type (e.g. 'log' > 'pole') are quite common in ancient languages. I'm not saying this actually ocurred, I only mention it as a possible scenario.

But the strongest argument against a separation of Anatolian and the rest of IE at the Black Sea Flood is in my opinion that Anatolian is not quite archaic as the language of the Old European hydronymy. Anatolian already has ablaut; the OEH shows no trace of it.

I broadly agree with you.

Of course, one could imagine that Proto-Europic was yet earlier, but I consider a connection of the western branch of Europic with the Central European Neolithic the likelier option.

Unfortunately, this is utterly unprovable (by now).

I don't see how that strengthens your point. You claimed that some Basque words converged with their Romance counterparts by way of "analogy" between languages - and that is and remains to be nonsense. Or did I misunderstand?

Yes, you did. My point is that final vowels in bisyllablic roots (although not all the Basque words have this structure) are of little significance for comparative purposes, among other things because this slot corresponds to the gender morpheme o/a in Romance languages.

[phoenix]

6 Pages and nobody has pointed out yet that a Uvular Fricative shifting to a alveolar fricative is more exotic than anything Armenian ever came up with?! Come on!

Sure s > h shift is fine. And I'm wiling to believe that every so often an h > X shift occurs. But this does not happen the other way around.

I can't think of a single language that ever had this happen that is on a more substantially proven level than Octaviano's hypotheses.

Anyone have any examples of this? (And no, anything Starostin came up with does not count as an example. At best that's proof that it's extremely unlikely like most of his reconstructions)

And even if someone finds an example of this, can said person find 2?

A shift so rare is extremely silly to propose as a shift in a yet unproven proto-language is insane.

Yes there's a paradox here, it may have occurred somehow somewhere and we wouldn't be able to reconstruct it. Therefore there's a little but:

Sure you can reconstruct such a shift in a proto-language if it's EXTREMELY well substantiated in the data. In this case it is not.

Why am I even spending time on this?

I guess it's to teach anyone silly enough to read through this hijacked thread how comparative linguistics is supposed to be done.

[Nortaneous]

Could be hypercorrection. There have apparently been /h/ > /p/ shifts or something like that because of it.

[Octaviano]

6 Pages and nobody has pointed out yet that a Uvular Fricative shifting to a alveolar fricative is more exotic than anything Armenian ever came up with?! Come on!

Sure s > h shift is fine. And I'm wiling to believe that every so often an h > X shift occurs. But this does not happen the other way around.

I suppose you refer to this:

This case is very illustrative of Paleo-Eurasian *H2- > PIE *s- in words like

*H2elA 'to dwell, live' > PIE *selo- 'dwelling, settlement'
*H2EmV 'warm' > PIE *sem- 'summer'

The above reconstructions are essentialy Dolgopolsky's, not Starostin's. They usually differ on the choice of cognates, phonetical reconstruction and so on.

The root *H2elA (ND 2579) is found in Semitic, Kartvelian and Uralic and *H2EmV (ND 2586) in Afro-Asiatic (Semitic, South Cushitic) and Altaic (Turkic).

Some time ago, my fellow Arnaud Fournet proposed PIE *H2 could hide several phonemes, among them /z/, so he proposed a substrate language he called "Z-retaining block" (ZRB) which kept this phoneme, lost in common PIE (see his essay for more details). Although I disagree with him in most respects, he inspired me the above etymologies (I even named this sound shift as "Fournet's Law" after him).

[jal]

What did the Proto-Indo-Europeans do with yokes before they had wheeled vehicles?

Plough?


JAL

[Octaviano]

I see. This is the same isogloss *(H)al-/*sal- we can see in some OEH roots and it's important indeed fot it gives us insight into dialectal differentiation.

The meaning 'salt' is very unsuitable to a river, so I think this root corresponds to PIE *sel- 'to move quickly; to jump' (eg. Latin saliō 'to spring, leap, jump, bound'). Its counterpart without initial *s- is *Hael- 'to well up, to flow'. Dolgopolsky reconstructs *dz|dʑæLV 'to swing, to wave, to be shaky; (?) to spring, to leap' (ND 2720).

I then suppose some "Europic" varieties (i.e. the ones from which late/tradtional PIE descends) retained the former voiced affricate as *s while in others it became *X.

*H2EmV 'warm' > PIE *sem- 'summer'

The s-less variant is found as PIE *H2em- 'to mow' (that is, 'harvest' ~ 'warm season' ~ 'summer').

*H2elA 'to dwell, live' > PIE *selo- 'dwelling, settlement'

I'm not 100% sure, but this could correspond PIE *H2elwo- 'elongated cavity, hollow' (e.g. Latin alvus), as in some cultures dwellings were built by carving a hollow in the ground.

A word like *H2em- can be readily associated to a farming culture like LBK. If we could also establish they made dwellings with a hollow in the ground, we'll be pretty sure they were also responsibles for the OEH thanks to circumstancial evidence.

[sangi39]

I see. This is the same isogloss *(H)al-/*sal- we can see in some OEH roots and it's important indeed fot it gives us insight into dialectal differentiation.

The meaning 'salt' is very unsuitable to a river, so I think this root corresponds to PIE *sel- 'to move quickly; to jump' (eg. Latin saliō 'to spring, leap, jump, bound'). Its counterpart without initial *s- is *Hael- 'to well up, to flow'. Dolgopolsky reconstructs *dz|dʑæLV 'to swing, to wave, to be shaky; (?) to spring, to leap' (ND 2720).

I then suppose some "Europic" varieties (i.e. the ones from which late/tradtional PIE descends) retained the former voiced affricate as *s while in others it became *X.

*H2EmV 'warm' > PIE *sem- 'summer'

The s-less variant is found as PIE *H2em- 'to mow' (that is, 'harvest' ~ 'warm season' ~ 'summer').

*H2elA 'to dwell, live' > PIE *selo- 'dwelling, settlement'

I'm not 100% sure, but this could correspond PIE *H2elwo- 'elongated cavity, hollow' (e.g. Latin alvus), as in some cultures dwellings were built by carving a hollow in the ground.

A word like *H2em- can be readily associated to a farming culture like LBK. If we could also establish they made dwellings with a hollow in the ground, we'll be pretty sure they were also responsibles for the OEH thanks to circumstancial evidence.

If I remember my Neolithic Europe lecturer correctly, he said that the predominant dwelling associated with the LBK Culture was the (Neolithic) Long House which was typically built on the surface rather than having a dug-out floor.

[sangi39]

What did the Proto-Indo-Europeans do with yokes before they had wheeled vehicles?

Plough?

JAL

I'd say this is pretty likely given the evidence for a non-regional PIE word for "(to) plough" and "(a) plough" as presented by Adams and Mallory who also note the suggestion that the yoke was originally used for pulling ploughs and non-wheeled vehicles before the invention/adoption of the wheel.

[WeepingElf]

What did the Proto-Indo-Europeans do with yokes before they had wheeled vehicles?

Plough?

JAL

I'd say this is pretty likely given the evidence for a non-regional PIE word for "(to) plough" and "(a) plough" as presented by Adams and Mallory who also note the suggestion that the yoke was originally used for pulling ploughs and non-wheeled vehicles before the invention/adoption of the wheel.

True, I have overlooked those uses. Indeed, the Indo-Europeans may have had yokes before they had wheeled vehicles.

[Octaviano]

If I remember my Neolithic Europe lecturer correctly, he said that the redominant dwelling associated with the LBK Culture was the (Neolithic) Long House which was typically built on the surface rather than having a dug-out floor.

IMHO, this type of building would fit nicely PIE *dom(H)- 'house' (*t´om(H)- in glottalic PIE), which is cognate to Altaic *t`āma 'wall, roof'.

This would imply Altaic is actually a branch of Europic, and it would also explain those Altaic-like substrate items in NW Europe as remnants of the Europic languages spoken there before IE.

[WeepingElf]

If I remember my Neolithic Europe lecturer correctly, he said that the redominant dwelling associated with the LBK Culture was the (Neolithic) Long House which was typically built on the surface rather than having a dug-out floor.

IMHO, this type of building would fit nicely PIE *dom(H)- 'house' (*t´om(H)- in glottalic PIE), which is cognate to Altaic *t`āma 'wall, roof'.

This would imply Altaic is actually a branch of Europic, and it would also explain those Altaic-like substrate items in NW Europe as remnants of the Europic languages spoken there before IE.

While I consider it likely that the Altaic languages are distantly related to Indo-European, I would not go as far as to consider them Europic, and I don't think there ever were Altaic languages in Northwest Europe.

The LBK people indeed did not live in dug-out dwellings. They built, as sangi39 correctly said, long houses with a floor level with the ground around the house.

[Octaviano]

This would imply Altaic is actually a branch of Europic, and it would also explain those Altaic-like substrate items in NW Europe as remnants of the Europic languages spoken there before IE.

While I consider it likely that the Altaic languages are distantly related to Indo-European, I would not go as far as to consider them Europic, and I don't think there ever were Altaic languages in Northwest Europe.

Ok, let's put aside Altaic for the moment. Perhaps I've runned a bit too fast.

The LBK people indeed did not live in dug-out dwellings. They built, as sangi39 correctly said, long houses with a floor level with the ground around the house.

I hope you can see the linguistic implications of this.

[WeepingElf]

The LBK people indeed did not live in dug-out dwellings. They built, as sangi39 correctly said, long houses with a floor level with the ground around the house.

I hope you can see the linguistic implications of this.

Which linguistic implications? I never claimed that any IE word meaning 'house', 'home' or `dwelling' has anything to do with a verb meaning 'to dig', and I am not aware of any such connection existing in PIE. It seems to me that you are beginning to confuse your critics with each other.

EDIT: Checking back in the thread, I see that it was you who brought up such a connection between *selo- and *h2elwo-, but I don't consider these two lexemes related. In my opinion, *s and *h2 are not freely interchangeable. They are entirely different phonemes, as are, for instance, *t and *k.

I also don't think that the speakers of Proto-Europic lived in dug-out dwellings - neither the LBK people nor the Sredny Stog people (the speakers of PIE) did.

[Octaviano]

Which linguistic implications? I never claimed that any IE word meaning 'house', 'home' or `dwelling' has anything to do with a verb meaning 'to dig', and I am not aware of any such connection existing in PIE. It seems to me that you are beginning to confuse your critics with each other.

You used "linguistic archaeology" to date PIE. Well, the weak point of your hypothesis is the lack of linguistic evidence which links LBK farmers to OEH. Now I'm trying to build such a evidence (either positive or negative).

In my opinion, *s and *h2 are not freely interchangeable. They are entirely different phonemes, as are, for instance, *t and *k.

I think you didn't understand my point. As yourself stated some time ago, there're 2 OEH roots, al- and sal-, with different phonetics but similar meaning. They correspond respectively to PIE *Hael- 'to well up, to flow' and *sel- 'to move quickly; to jump', both possibly being different reflexes of the same Paleo-Eurasian/Nostratic root.

If Dolgopolsky's correspondences are right, then the form with *s- would be PIE-native while the one with *H- would be a borrowing from a sister (i.e. "Europic") language, although it also could be the other way around (as Mr Fournet proposes), so the sound shift *DZ- > *H2- would be PIE-native.

In that case, it's also reasonable for *H2- > *H2- to be PIE-native, so words with *H2- > *s- like *selo- or *sem- would be loanwords from an "Europic" sister language to PIE.

My question is then: is there any chance we can adscribe the language of the s-words to a particular archaeological culture such as LBK or not?