The Great Proto-Indo-European Thread's Sequel

[Tropylium]

Anthony Yates recently argues that the reason *χ ≤ *h₂ was geminated but *s wasn't is that *s was unspecified for voice, while *χ contrasted with *ʁ ≤ *h₃. I suppose this could be taken to imply that there was an (allophonic) change *-s- > *-z- before the voice-to-gemination shift he defends.

(new notation coined: "*x ≤ *y" for "either *x = *y or *x < *y")

*h₂ = *q is maybe still possible (esp. given correspondences of the type PIE *h₂ ~ Uralic *k, or for that matter PIE *h₂ ~ Afrasian *x (I suspect because of reasons that Proto-Semitic *x *ɣ *ħ *ʕ do not come intact from Proto-Afrasian, but rather from PAA *q *ɢ *χ *ʁ)). I would however take this hypothesis to go with a reanalysis of the IE family tree as {{Core IE, Hittite}, Luwic} rather than {Core IE, {Hittite, Luwic}}.

[WeepingElf]

Anthony Yates recently argues that the reason *χ ≤ *h₂ was geminated but *s wasn't is that *s was unspecified for voice, while *χ contrasted with *ʁ ≤ *h₃. I suppose this could be taken to imply that there was an (allophonic) change *-s- > *-z- before the voice-to-gemination shift he defends.

(new notation coined: "*x ≤ *y" for "either *x = *y or *x < *y")

This requires that *h₃ was a voiced counterpart of *h₂, while it looks more as if the difference was one of labialization. (And don't get me started with Kath's "*h₂^w" as an entity distinct from *h₃, which I still consider entirely fictitious.)

*h₂ = *q is maybe still possible (esp. given correspondences of the type PIE *h₂ ~ Uralic *k, or for that matter PIE *h₂ ~ Afrasian *x (I suspect because of reasons that Proto-Semitic *x *ɣ *ħ *ʕ do not come intact from Proto-Afrasian, but rather from PAA *q *ɢ *χ *ʁ)). I would however take this hypothesis to go with a reanalysis of the IE family tree as {{Core IE, Hittite}, Luwic} rather than {Core IE, {Hittite, Luwic}}.

1. I see no reason to assume a relationship between IE and Afrasian. These language families look utterly different and seem to have utterly different origins, and the only motivation to connect them to each other was 19th-century typology ("both are inflected"), which is simply obsolete: there are plenty more "inflected", i.e. fusional languages which only lunatics would connect to IE or Afrasian (e.g., in the Americas), and the inflectional systems of IE and Afrasian are so completely different that no connection can be drawn between them.

2. The proposal that Hittite was more closely related to Core IE than to Luwian is a very bold one. Why do you think that way?

[Travis B.]

Think about how English speakers tend to pronounce a name like Bach as [bæk].

Actually, we pronounce Bach like /bɑːk/, to nitpick.

[Pabappa]

I think he meant loans, not genetics. Though few words with laryngeals are thought to be loans.

[jal]

for instance, they didn't take part in the root structure constraints that govern the distribution of stops

Isn't this a bit circular reasoning though? There could just well be two classes of stops, one having a constraint, the other not.

Note that I, like I said, have no knowledge of PIE other than the bits I pick up here, so this is not an attempt to dismiss or approve anything, just pointing out what could be faulty reasoning :).


JAL

[KathTheDragon]

It's ad-hoc to posit a second unconstrained class of stops, given that it would contain only the laryngeals.

[WeepingElf]

I think he meant loans, not genetics. Though few words with laryngeals are thought to be loans.

Fair. Indeed, I think there are quite a few Neolithic Wanderwörter in both IE and Semitic. I don't know, though, how many of those include laryngeals.

[WeepingElf]

It's ad-hoc to posit a second unconstrained class of stops, given that it would contain only the laryngeals.

Yes, while the laryngeals appear to be in a class of their own (*s behaves very differently, but sibilants often behave differently than other fricatives), it is very ad-hoc to declare this class to be stops. That's just circular reasoning. So far, the (difficult and somewhat ambiguous) evidence from Lycian notwithstanding, we have no really compelling reason to assume that the laryngeals were stops in either Early PIE or Late PIE.

[Tropylium]

Anthony Yates recently argues that the reason *χ ≤ *h₂ was geminated but *s wasn't is that *s was unspecified for voice, while *χ contrasted with *ʁ ≤ *h₃. I suppose this could be taken to imply that there was an (allophonic) change *-s- > *-z- before the voice-to-gemination shift he defends.

(new notation coined: "*x ≤ *y" for "either *x = *y or *x < *y")

This requires that *h₃ was a voiced counterpart of *h₂, while it looks more as if the difference was one of labialization. (And don't get me started with Kath's "*h₂^w" as an entity distinct from *h₃, which I still consider entirely fictitious.)

We can continue debating this for PIE, but I gather everyone agrees that Hittite had four laryngeals /χ χʷ ʁ ʁʷ/ ‹ḫḫ ḫḫu ḫ ḫu› (though with the voice contrast only medially), with the voiced ones often derived from *h₃. People like Melchert take this as the Proto-Anatolian state too, with the introduction of labialized laryngeals from *H+w as a defining Anatolian innovation. Kloekhorst would change these to PAn *qː *qːʷ *q *qʷ and shuffles the development of *h₃ somewhat but does not get rid of the four-way contrast either.

— Another possible reconstruction that could be entertained is *kʰ for *h₂, which might work better with his "Achaean" argument (5.3 §) (but it would also raises questions about whatever happened to other voiceless aspirates, & about what to do with *h₃; perhaps *kʷʰ??).

2. The proposal that Hittite was more closely related to Core IE than to Luwian is a very bold one. Why do you think that way?

There are already known areally shared innovations across Anatolian (e.g. stop gemination, stressed vowel lengthening, a tendency for vowel system simplification), plus really not that many Proto-Anatolian innovations, including a few that I think can be archaisms rather than innovations (*h₂ʷ, most issues of vocabulary). I suspect no-one has so far considered this idea seriously enough. For another Core IE = Hittite ≠ Luwic isogloss, there's the merger of *eh₁ into *ē (and in Luwic rather *ē *eh₁ > *ī *ā).

Of course this runs into some problems due to the k/h₂ isogloss being not even Luwic/Rest but rather Carian–Lycian/Rest.

[Tropylium]

I think he meant loans, not genetics. Though few words with laryngeals are thought to be loans.

Fair. Indeed, I think there are quite a few Neolithic Wanderwörter in both IE and Semitic. I don't know, though, how many of those include laryngeals.

Bomhard has the following examples for *χ ~ *h₂ (note how almost all of them have poor distribution across AA):
PS √lχp-, √lχp-, √lχm- 'to beat' (~ Berber, Cushitic) ~ PIE *lah₂w- 'to beat'
PS √šχn- 'to be warm' ~ PIE *sh₂wen- 'sun'
PS √χlkʼ- 'to wear down' ~ PIE *h₂al- 'to grind'
PS √χnb- 'to grow' ~ PIE *h₂andʰ- 'to bloom'
PS √χtʼtʼ- 'to carve' ~ PIE *h₂ad- 'to cut'
PS √χlʕ- 'to pull out' ~ PIE *h₂wel- 'to pull'
Arabic √χṭb- 'to preach', √χṭl- 'to prattle' (analyzed as extensions of *χtʼ-) ~ PIE *h₂wedH- 'to speak'
Egyptian ḫnt 'face' ~ PIE *h₂ant- 'front, etc.'
(also Egyptian ḫm 'to be wild' ~ PIE *h₂nēr 'man', bridged by supposed Dravidian cognates meaning 'male animal', but I really would not buy this)

Though for sure they could also be later than any possible *q > *χ shifts in Semitic. I'd also have to check if any of these rather have the newly proposed *x̣ (which gives West Semitic *ħ ~ East Semitic *x) which to me seems the most likely to have been a stop *qʼ or affricate *qχʼ the longest.

There's just one example for *ʁ ~ *h₃: PS √ʁrb- 'to abandon' ~ PIE *h₃orbʰos 'orphan'. This little could be coincidence too for sure.

[mae]

-

[mèþru]

Do we really know if it was geminate? Could the hittites have been sloppy spellers, like us, where "egg" has 2 g's,"bee" & "be" rhyme, etc? They used double letters for voiceless single stops elsewhere, right?

This made me think of sloppy spelling as in common inaccurate spellings. What if they have infected the Hittite corpus?

Also, much of those "sloppy spellings" actually have sound etymological reasons - egg is a borrowing from Old Norse, which had phonemic gemination. English "bee" and some forms of "be" had already merged in pronunciation and spelling back in Old English, where they were "bēo". The invention of different spelling was probably not a mistake but a purposeful and useful invention.

[Nortaneous]

The invention of different spelling was probably not a mistake but a purposeful and useful invention.

Yes, function words need at least three letters even when the pronunciation alone doesn't demand it - bee, egg, eye, axe, tie...

[Zaarin]

The invention of different spelling was probably not a mistake but a purposeful and useful invention.

Yes, function words need at least three letters even when the pronunciation alone doesn't demand it - bee, egg, eye, axe, tie...

In the US, "ax" is also an acceptable spelling of the noun, so...

[Kuchigakatai]

In Latin, all the correlative pronouns for motion-towards look like ablatives. Does anyone know how that came about? Are these, somehow, maybe remnants of the Indo-European instrumental?

I'm talking about quō 'where to?', aliquō/quōquam/quōpiam 'to some place', eō 'to there', eōdem 'to the same place', quōquō/quōcumque 'no matter where [you go] to', aliō 'to somewhere else'.

[vegfarandi]

I think he meant loans, not genetics. Though few words with laryngeals are thought to be loans.

Fair. Indeed, I think there are quite a few Neolithic Wanderwörter in both IE and Semitic. I don't know, though, how many of those include laryngeals.

Bomhard has the following examples for *χ ~ *h₂ (note how almost all of them have poor distribution across AA):
PS √lχp-, √lχp-, √lχm- 'to beat' (~ Berber, Cushitic) ~ PIE *lah₂w- 'to beat'
PS √šχn- 'to be warm' ~ PIE *sh₂wen- 'sun'
PS √χlkʼ- 'to wear down' ~ PIE *h₂al- 'to grind'
PS √χnb- 'to grow' ~ PIE *h₂andʰ- 'to bloom'
PS √χtʼtʼ- 'to carve' ~ PIE *h₂ad- 'to cut'
PS √χlʕ- 'to pull out' ~ PIE *h₂wel- 'to pull'
Arabic √χṭb- 'to preach', √χṭl- 'to prattle' (analyzed as extensions of *χtʼ-) ~ PIE *h₂wedH- 'to speak'
Egyptian ḫnt 'face' ~ PIE *h₂ant- 'front, etc.'
(also Egyptian ḫm 'to be wild' ~ PIE *h₂nēr 'man', bridged by supposed Dravidian cognates meaning 'male animal', but I really would not buy this)

Though for sure they could also be later than any possible *q > *χ shifts in Semitic. I'd also have to check if any of these rather have the newly proposed *x̣ (which gives West Semitic *ħ ~ East Semitic *x) which to me seems the most likely to have been a stop *qʼ or affricate *qχʼ the longest.

There's just one example for *ʁ ~ *h₃: PS √ʁrb- 'to abandon' ~ PIE *h₃orbʰos 'orphan'. This little could be coincidence too for sure.

What does the √ signify?

[Salmoneus]

root.

[Xwtek]

Do anyone has free resource of PIE morphology? I only found two book. One is Late PIE. And the others concerns more of culture than morphology

[WeepingElf]

In Latin, all the correlative pronouns for motion-towards look like ablatives. Does anyone know how that came about? Are these, somehow, maybe remnants of the Indo-European instrumental?

I'm talking about quō 'where to?', aliquō/quōquam/quōpiam 'to some place', eō 'to there', eōdem 'to the same place', quōquō/quōcumque 'no matter where [you go] to', aliō 'to somewhere else'.

Are you sure they aren't datives?

[Howl]

In Latin, all the correlative pronouns for motion-towards look like ablatives. Does anyone know how that came about? Are these, somehow, maybe remnants of the Indo-European instrumental?

I'm talking about quō 'where to?', aliquō/quōquam/quōpiam 'to some place', eō 'to there', eōdem 'to the same place', quōquō/quōcumque 'no matter where [you go] to', aliō 'to somewhere else'.

It is a remnant of an old allative case ('where to'?) in -o which has been attested in Hittite ( ne-e-pi-ša 'to heaven' ).

Do anyone has free resource of PIE morphology? I only found two book. One is Late PIE. And the others concerns more of culture than morphology

https://en.wikipedia.org/wiki/Proto-Ind ... n_language is a very good starting point. See also the list of subtopics at the bottom there.