The Great Proto-Indo-European Thread's Sequel

[Talskubilos]

Now we’re starting to get somewhere. Next question: why do you believe that *abVl- is not a native PIE word?

Mainly because:
1) it has a limited distribution, only found in a few branches.
2) it has a non-native phoneme *b.
3) it's seemingly related to Hittite šam(a)lu-.

[Ketsuban]

Mainly because:
1) it has a limited distribution, only found in a few branches.
2) it has a non-native phoneme *b.
3) it's seemingly related to Hittite šam(a)lu-.

Cool, we're all happy with the reasoning so far and I agree with the criticism that a *b makes a PIE reconstruction suspicious in particular. Next step: Fenwick (op. cit.) provides an explanation of how "abVl-" can still ultimately be related to *meh₂l-, via two irregular sound changes: a metathesis *meh₂l- > *h₂eml- motivated by an association with *h₂em-ro- "sour" which persists into Sanskrit and which (with s-⁠mobile) gives the Anatolian forms attested, and a fortition *h₂eml- > *h₂ebl- occurring after the breakup of Proto-Indo-European into its separate dialects (thus explaining the limited distribution).

This addresses all three of your criticisms without need for an unexplained loanword. As such, you need additional evidence to support your contention that "abVl-" is actually a loanword of some kind unrelated to *meh₂l-.

[Talskubilos]

Cool, we're all happy with the reasoning so far and I agree with the criticism that a *b makes a PIE reconstruction suspicious in particular. Next step: Fenwick (op. cit.) provides an explanation of how "abVl-" can still ultimately be related to *meh₂l-, via two irregular sound changes: a metathesis *meh₂l- > *h₂eml- motivated by an association with *h₂em-ro- "sour" which persists into Sanskrit and which (with s-⁠mobile) gives the Anatolian forms attested, and a fortition *h₂eml- > *h₂ebl- occurring after the breakup of Proto-Indo-European into its separate dialects (thus explaining the limited distribution).

This addresses all three of your criticisms without need for an unexplained loanword.

Sorry, but I disagree. Although I think *abVl- could be remotedly related to *meh₂l-, it can't be derived from it with regular/predictable sound changes (not ad hoc ones!), which also applies to the Anatolian word (there's no s-mobile here, for goodness sake!).

On a broader context, names of plants, fruits and trees (phytonyms) are often borrowed. To quote an example, Latin citrus and Greek kédros are instances of a Wanderwort of Semitic origin: *kˁtr- 'smoke, incense', referring to the aromatic characteristics of these trees, either the wood (cedar, thuja) or the fruit (citron).

[Ketsuban]

Although I think *abVl- could be remotedly related to *meh₂l-, it can't be derived from it with regular/predictable sound changes (not ad hoc ones!)

Not all sound changes are nice perfectly-regular neogrammarian ones; sometimes there's loans from related dialects (vixen) or folk etymology (eggcorn). Fenwick doesn't propose *meh₂l- > *h₂eml- > h₂ebl- as purely ad hoc changes; the association between "apple" and "sour" is backed up by evidence from Sanskrit and (as a semantic association) Germanic, and the irregular fortition m > b also occurred in Russian.

which also applies to the Anatolian word (there's no s-mobile here, for goodness sake!).

Based on what reasoning, the writing system representing it as š? S-mobile is pervasive in Indo-European, and its presence or absence is completely unpredictable.

On a broader context, names of plants, fruits and trees (phytonyms) are often borrowed. To quote an example, Latin citrus and Greek kédros are instances of a Wanderwort of Semitic origin: *kˁtr- 'smoke, incense', referring to the aromatic characteristics of these trees, either the wood (cedar, thuja) or the fruit (citron).

While this is true enough, "sometimes plant names are loanwords" isn't actually evidence to support your claim - sometimes they're inherited.

Also, stop saying "Wanderwort" when you mean "loanword".

[keenir]

Now we’re starting to get somewhere. Next question: why do you believe that *abVl- is not a native PIE word?

Mainly because:
1) it has a limited distribution, only found in a few branches.

isn't distribution a part of how we determine which branches are which?
(such as satem itself - thats not found in all the IE languages from the Canaries to Sri Lanka either)

2) it has a non-native phoneme *b.

Could the b be something that those few branches developed as they grew into their own branches?

3) it's seemingly related to Hittite šam(a)lu-.

Um...given that Hittite is a card-carrying member of the Indo-European family (at least, it was, when last i checked)...why would that stop *abVl- from being a native PIE word? your #3 folds into your #1, if I'm reading that correctly.

[Talskubilos]

Not all sound changes are nice perfectly-regular neogrammarian ones; sometimes there's loans from related dialects (vixen) or folk etymology (eggcorn). Fenwick doesn't propose *meh₂l- > *h₂eml- > h₂ebl- as purely ad hoc changes; the association between "apple" and "sour" is backed up by evidence from Sanskrit and (as a semantic association) Germanic, and the irregular fortition m > b also occurred in Russian.

Well, this is his theory. Mine is we're dealing with a Wanderwort with several reflexes in IE and elsewhere: Nakh-Daghestanian *mhalV- ~ *mhanV- 'warm', Uralic *omena ~ *omVrV 'apple', Basque udare, udari, madari 'pear', umo, umao (B) 'ripe, seasoned'.

which also applies to the Anatolian word (there's no s-mobile here, for goodness sake!).

Based on what reasoning, the writing system representing it as š? S-mobile is pervasive in Indo-European, and its presence or absence is completely unpredictable.[/quote]
There're some cases of Hittite š- corresponding to PIE *H₃-, namely šākuwa- 'eye' < PIE *H₃ekʷ- 'to see' and šankuwāi- 'nail; a unit of linear measure' < PIE *H₃n(o)gh- 'nail'.

Also, stop saying "Wanderwort" when you mean "loanword".

I say Wanderwort when I mean it.

[Talskubilos]

isn't distribution a part of how we determine which branches are which?
(such as satem itself - thats not found in all the IE languages from the Canaries to Sri Lanka either)

I'd suggest you read Mallory & Adams own list of "NW IE" words.

[keenir]

isn't distribution a part of how we determine which branches are which?
(such as satem itself - thats not found in all the IE languages from the Canaries to Sri Lanka either)

I'd suggest you read Mallory & Adams own list of "NW IE" words.

o-kay...and can you provide a title and-or a link so i can find said list?
thank you

for all i know, right now, you're agreeing with me

[Ketsuban]

Well, this is [their] theory. Mine is we're dealing with a Wanderwort with several reflexes in IE and elsewhere: Nakh-Daghestanian *mhalV- ~ *mhanV- 'warm', Uralic *omena ~ *omVrV 'apple', Basque udare, udari, madari 'pear', umo, umao (B) 'ripe, seasoned'.

This is quite a list.

• I can find a "Proto-Nakh" *mHaliᶰ but nothing for the entire Northeast Caucasian family. It's temptingly similar to Proto-Ugric *mälɜ, but I've got to call lack of evidence here since this is the stomping ground of the notorious Sergei Starostin.
• *omena, *omVrV is more Starostinian gibberish formed by smashing together Proto-Finnic *omëna and Hungarian alma (apparently assuming a metathesis from *omra with a magic vanishing vowel); both are thought to be loans from Indo-European, making using this as evidence of a Wanderwort circular reasoning.
• The only reason to mention Basque is madari, which has an entirely internal explanation as the expressive prefix ma- applied to Proto-Basque *udare. A claim of folk etymology for a loanword would require evidence to support it. (umo is unverified on Wiktionary; the usual Basque term for "mature, ripe" is zori. "Ripen" is heldu da.)

There're some cases of Hittite š- corresponding to PIE *H₃-, namely šākuwa- 'eye' < PIE *H₃ekʷ- 'to see' and šankuwāi- 'nail; a unit of linear measure' < PIE *H₃n(o)gh- 'nail'.

Yes, this is s-mobile; you can tell because the same roots are found elsewhere in Indo-European without an initial *s, as in Latin oculus and unguis. (There's also *h₂eḱru "bitter", found in Hittite with s-mobile as išḫaḫru and in Sanskrit without it as áśru, both "tear". Other branches seem to reflect a compound *dr̥ḱ-h₂eḱru "eye-bitter", hence Latin lacrima and the earlier incorrect reconstruction *dáḱru.)

I say Wanderwort when I mean it. ;)

Apparently not, since you just used it to describe a word loaned the incredible distance... across the Mediterranean.

[Talskubilos]

There're some cases of Hittite š- corresponding to PIE *H₃-, namely šākuwa- 'eye' < PIE *H₃ekʷ- 'to see' and šankuwāi- 'nail; a unit of linear measure' < PIE *H₃n(o)gh- 'nail'.

Yes, this is s-mobile; you can tell because the same roots are found elsewhere in Indo-European without an initial *s, as in Latin oculus and unguis. (There's also *h₂eḱru "bitter", found in Hittite with s-mobile as išḫaḫru and in Sanskrit without it as áśru, both "tear".

I disagree; s-mobile appears before consonants, which isn't the case. For the Sanskrit word you quoted, see Gamkrelidze & Ivanov (1995): Indo-European and the Indo-Europeans

I say Wanderwort when I mean it.

Apparently not, since you just used it to describe a word loaned the incredible distance... across the Mediterranean.

This is the precisely why I call it a Wanderwort.

[Talskubilos]

o-kay...and can you provide a title and-or a link so i can find said list?
thank you

I've uploaded a copy of the book here.

[bradrn]

I say Wanderwort when I mean it.

Apparently not, since you just used it to describe a word loaned the incredible distance... across the Mediterranean.

This is the precisely why I call it a Wanderwort.

I think you may have missed the considerable sarcasm in Ketsuban’s post. ‘Loanword across the Mediterranean’ is not an ‘incredible distance’ at all.

[WeepingElf]

o-kay...and can you provide a title and-or a link so i can find said list?
thank you

I've uploaded a copy of the book here.

Thank you - a valuable resource.

[Ketsuban]

I disagree; s-mobile appears before consonants, which isn't the case.

*h₃ekʷ-, *h₃negʰ- and *h₂eḱru all begin with laryngeals, which last I checked are generally considered to be consonants. (I am aware of the arguments for vocalic allophones; it seems a little odd to me that a language would be comfortable with *m, *n, *s, *r and *l as syllable nuclei yet require a vocalic allophone for dorsal fricatives, but if you want to argue the initial *h₃ in *h₃nṓgʰs is actually [ɵ] then more power to you. It's still a candidate for s-mobile, since it's not a vowel.)

For the Sanskrit word you quoted, see Gamkrelidze & Ivanov (1995): Indo-European and the Indo-Europeans

You could have summarised the argument rather than expect me to get a copy of a book, but it doesn't matter since I have more: Tocharian A ākär, Lithuanian ãšara. More fool me for thinking one example would suffice to show a pattern.

[WeepingElf]

For the Sanskrit word you quoted, see Gamkrelidze & Ivanov (1995): Indo-European and the Indo-Europeans

You could have summarised the argument rather than expect me to get a copy of a book, but it doesn't matter since I have more: Tocharian A ākär, Lithuanian ãšara. More fool me for thinking one example would suffice to show a pattern.

Well, Gamkrelidze & Ivanov's ideas are quite eccentric and not accepted by most Indo-Europeanists, so it is not very advisable to rely on their work. Their version of the PIE lexicon is full of spurious items and questionable reconstructions. Better use Mallory & Adams instead.

[Talskubilos]

*h₃ekʷ-, *h₃negʰ- and *h₂eḱru all begin with laryngeals, which last I checked are generally considered to be consonants. (I am aware of the arguments for vocalic allophones; it seems a little odd to me that a language would be comfortable with *m, *n, *s, *r and *l as syllable nuclei yet require a vocalic allophone for dorsal fricatives, but if you want to argue the initial *h₃ in *h₃nṓgʰs is actually [ɵ] then more power to you. It's still a candidate for s-mobile, since it's not a vowel.)

Sorry, I was in a hurry this morning.

In summary, Gamkrelidze & Ivanov posit a "compact fricative" sibilant *ŝ- in PIE which would account for Ø- in Sanskrit and š- in Hittite (e.g. šakuwa-) and t- in Luwian (e.g. tawi- 'eyes').

Well, Gamkrelidze & Ivanov's ideas are quite eccentric and not accepted by most Indo-Europeanists, so it is not very advisable to rely on their work. Their version of the PIE lexicon is full of spurious items and questionable reconstructions. Better use Mallory & Adams instead.

Still it's interesting to study, because they illustrate how part of IE lexicon relative to fauna and flora has been borrowed from other families.

[bradrn]

I disagree; s-mobile appears before consonants, which isn't the case.

*h₃ekʷ-, *h₃negʰ- and *h₂eḱru all begin with laryngeals, which last I checked are generally considered to be consonants. (I am aware of the arguments for vocalic allophones; it seems a little odd to me that a language would be comfortable with *m, *n, *s, *r and *l as syllable nuclei yet require a vocalic allophone for dorsal fricatives, but if you want to argue the initial *h₃ in *h₃nṓgʰs is actually [ɵ] then more power to you. It's still a candidate for s-mobile, since it's not a vowel.)

As it happens, the Manaster Ramer paper we were just talking about is precisely a suggestion that the laryngeals are vocalic.

[Ketsuban]

In summary, Gamkrelidze & Ivanov posit a "compact fricative" sibilant *ŝ- in PIE which would account for Ø- in Sanskrit and š- in Hittite (e.g. šakuwa-) and t- in Luwian (e.g. tawi- 'eyes').

And Fenwick, writing in 2016 and well aware of Gamkrelidze and Ivanov, points out that the existing well-documented process of s-mobile is a more parsimonious explanation than a magic fricative which turns into a pumpkin as soon as someone tries to write it.

Still it's interesting to study, because they illustrate how part of IE lexicon relative to fauna and flora has been borrowed from other families.

Unless they're not, which is the entire basis of contention here.

As it happens, the Manaster Ramer paper we were just talking about is precisely a suggestion that the laryngeals are vocalic.

I'm obviously a fan of a little more levity in academic discussion, but I couldn't actually figure out what Manaster Ramer was trying to say because they write like a conspiracy theorist.

[Talskubilos]

In summary, Gamkrelidze & Ivanov posit a "compact fricative" sibilant *ŝ- in PIE which would account for Ø- in Sanskrit and š- in Hittite (e.g. šakuwa-) and t- in Luwian (e.g. tawi- 'eyes').

And Fenwick, writing in 2016 and well aware of Gamkrelidze and Ivanov, points out that the existing well-documented process of s-mobile is a more parsimonious explanation than a magic fricative which turns into a pumpkin as soon as someone tries to write it.

Although I don't fully agree with Gamkrelidze and Ivanov, I think the mainstream theory doesn't explain the origin of s-mobile, which IMHO could be some kind of fossilized (i.e. no longer productive) prefix.

On the other hand, the main different between my and theory and Fenwick's is that, whereas he uses terms such as "Early PIE" and "NW IE", I think these words originated outside of PIE itself. As a matter of comparison, I still remember the desperate atempts of some Romanists to concoct Latin etymologies for Spanish Pre-Latin words such as e.g. tonto 'dumb', suposedly from Latin atonitus.

[keenir]

o-kay...and can you provide a title and-or a link so i can find said list?
thank you

I've uploaded a copy of the book here.

Thank you for that. This will help.