The Great Proto-Indo-European Thread's Sequel

[WeepingElf]

If Grassmann's Law had been productive in PIE, why are the stops deaspirated by it voiceless in Greek? One would expect voiced stops instead. I think this is good evidence that this happened after the PIE breathy-voiced stops lost their voicing in Greek. Also, there are no traces of such a dissimilation in Germanic, Italic or Armenian.

[Zaarin]

Theyre the only branches that preserve the aspirates into written history*. It could have been once more widespread, but then survived in only these two branches because in the others it got moot. é.g. Slavic collapses all but one of the aspirated stops with its plain form. in fact it may be that it was productive in PIE too, since we wouldnt know, would we? that said, there are details in the two Grassmans' Lwas that suggest perhaps it was not completely developed in proto-Graeco-Aryan, or even that the original theory that it developed twice is the correct one. after all, a similar process appears in some non-IE languages too, so it could just be that languages with aspirates and variable stress tend to have laws like this no matter what the proto-lang was like.

*arm shifted, so the asps are not the original asps.

Germanic maintained all three as distinct with no trace of Grassman's Law. EDIT: As Weeping Elf said.

[Frislander]

On another note, I'm currently learning Ancient Greek over the summer (for an Indo-European philology paper I'm taking this year coming up when I go back to university), and I'm currently rather intrigued by the syntax and semantics of the participles. Quite often they appear to be functioning more as Altaic-style converbs than more traditional "participle-like" functions, so for example they can be used to denote events occurring simultaneously with (using the present participle) or before (using the aorist participle) the event denoted by the finite verb, in cases where English might use a straight up connective like "and" (e.g. ταῦτα εἰπὼν ὁ ναύτης ἐγέλασαν "the sailor said these things and laughed"). And this got me wondering whether the typical situation in modern-day IE languages (in Europe at least), where a wide variety of conjunctions do most of the subordinating work, may have developed from a system more resembling that found in Uralic or Turkic, where non-finite subordination is dominant. Is there support from other older IE languages for this proposal?

[Dē Graut Bʉr]

"Having said this, the sailor laughed" is syntactically and semantically pretty much the same as that Greek phrase, isn't it? It's just that Greek (and Latin, FWIW) uses such constructions much more freely than English does.

[Frislander]

"Having said this, the sailor laughed" is syntactically and semantically pretty much the same as that Greek phrase, isn't it? It's just that Greek (and Latin, FWIW) uses such constructions much more freely than English does.

I know that, but that's not a counter-argument to my point. I'm asking whether this greater freedom is indicative that perhaps it was used for all constructions that would in modern languages use conjunctions, if not in PIE then at some stage before that.

[hwhatting]

I think most introductions into IE historic linguistics that go into syntax will tell you that subordinating conjunctions are a recent development in all known families. (Enclitic coordinating conjuctions like *-k^we "and", *-we "or" are normally reconstructed for PIE.) Before the discovery of Anatolian, most scholars would have assumed that the participal constructions of Greek or Latin would also have been typical for PIE, but my understanding of Anatolian is that participles aren't frequently used that way in the Anatolian languahes, but that they use sentence-initial chains of clitics to express (among other things) relationships between clauses. Such clitics can also be found in Vedic and Homeric Greek, and become rarer in both Greek and Indic over time, as participal constructions and conjunctions take over. I think there is some consensus nowadays that clitic chains were the PIE method to express relationships between clauses, and that both participal constructions and subordinating conjunctions are younger developments.

[Salmoneus]

I think most introductions into IE historic linguistics that go into syntax will tell you that subordinating conjunctions are a recent development in all known families. (Enclitic coordinating conjuctions like *-k^we "and", *-we "or" are normally reconstructed for PIE.) Before the discovery of Anatolian, most scholars would have assumed that the participal constructions of Greek or Latin would also have been typical for PIE, but my understanding of Anatolian is that participles aren't frequently used that way in the Anatolian languahes, but that they use sentence-initial chains of clitics to express (among other things) relationships between clauses. Such clitics can also be found in Vedic and Homeric Greek, and become rarer in both Greek and Indic over time, as participal constructions and conjunctions take over. I think there is some consensus nowadays that clitic chains were the PIE method to express relationships between clauses, and that both participal constructions and subordinating conjunctions are younger developments.

Could you give an example of these 'clitic chains'?

[WeepingElf]

Could you give an example of these 'clitic chains'?

AFAIK, those "clitic chains" are a Hittite thing, more prominent in late than in early texts (which is a pretty sure sign of an innovation), with little evidence in favour of them in other IE languages, though Gamkrelidze and Ivanov reconstructed them for PIE.

[hwhatting]

Could you give an example of these 'clitic chains'?

Here is a description (scroll down to "17 The Enclitic String").

AFAIK, those "clitic chains" are a Hittite thing, more prominent in late than in early texts (which is a pretty sure sign of an innovation),

Do you have any sources on this? Up to now, I have only seen them referred to and dexcribed, but nothing on their internal history in Hittite. I'd also be interested in more information on their presence or absence in other Anatolian languages, especially cuneiform Luwian - I fear that we simply have too few continuous narrative texts in the other Anatolian languages to say how typical or untypical clitic chains would be there.

(which is a pretty sure sign of an innovation),

Not necessarily. It's only a sign if you take it together with evidence from the other PIE languages. Classical Sanscrit really went overboard with using compounds, but that doesn't mean that compounds are an innovation of Indic. The length of the chains and the slot-model of Hittite may be innovations, but the use of multiple clitics in Homeric Greek or Vedic suggests to me that PIE had something similar, which was further developed in Hittite, but reduced or abolished in other IE languages. Other features that continue this could be the combinations of what was originally adverbials and clitics that can be seen in the Old Irish pre-fixed verb.

[WeepingElf]

I see. So I was wrong - I thought that they were mainly a Hittite thing.

[WeepingElf]

I have a new idea about the sigmatic nominative. According to this idea, it is not related to the genitive. Rather, in late PIE1, animate subjects were reinforced by postposing the pronoun *sa (> PIE3 *so), in constructions similar to those found in colloquial German (Mein Freund, der singt im Kirchenchor), Unlike the genitive *-sa, this was a separate word and thus did not throw the accent rightward as the genitive (and other oblique cases) did. But it was itself unaccented, and when ablaut arose, it became the zero grade ending *-s familiar to us.

[KathTheDragon]

This makes a lot of sense, honestly. The consistent lack of a vowel probably implies the addition was very early, presumably before the mass syncopation of unstressed vowels.

[WeepingElf]

Thanks. I am probably not the first to come up with such a theory, though I arrived at it by my own means and don't remember any references.

[mèþru]

I like it too.

[Howl]

I think it is very likely that nominative -s had a pronominal origin.

In my own view early PIE had a fluid version of the system in Anatolian. It used ergative (-n/-nt?) -absolutive (-ø) case marking for passive/stative participants, while using nominative (-s) /accusative (-m) case marking for active/volitional participants. Such a morphosyntactic alignment system ('fluid SAP' ?) is not one of the usual morphosyntactic alignment systems you find in the handbooks. But it has a parallel in the way Yukaghir uses the morphosyntactic alignment of the case marking to mark focus.

Uralic also has clear traces of earlier ergativity but went through another path to nominative-accusative alignment, using the difference between intransitive/anti-passive conjugation and transitive conjugation of the verb to mark definiteness of the object. So I think Indo-Uralic must have been fully ergative. And the accusative -m must have originated from anti-passive constructions. (See also 'Proto-Uralic Ergativity Reconsidered' by Merlijn de Smit https://www.academia.edu/9845499/Proto- ... considered for the argument that Uralic was once ergative)

[mèþru]

Yeah, I definitely have the active-stative interpretation.
Now it sounds like I'm a broken record. I need something to disagree with.

[KathTheDragon]

Hittite ergativity in inanimates is secondary since the ergative "case" is a derivative in *-ent-, so it can't be projected back to PIE.

[mèþru]

That doesn't disprove an active-stative PIE, only that part of the idea. I mainly support the active-stative idea because of the arguments here: http://en.wikipedia.org/wiki/Active–sta ... _languages

Actually, what they describe sounds to me more like split-S than fluid.

[Howl]

Hittite ergativity in inanimates is secondary since the ergative "case" is a derivative in *-ent-, so it can't be projected back to PIE.

The ergative case of inanimates in Anatolian is a real ergative case. See 'The Problem of the Ergative Case in Hittite' by Craig Melchert http://linguistics.ucla.edu/people/Melc ... istext.pdf which dispels a few myths on this. Also, it's not sure that the comparison between Hittite ergative -anza to the -nt in participles etc. is valid. But even if the form may be innovative, that still does not prevent projecting split ergativity back to PIE.

[KathTheDragon]

It's a valid comparison since the original function of the *-nt- participles was not being participles, but rather deriving... I'm pretty sure it's something along the lines of agentive nouns, but whatever it is, it's backed up by direct continuations of that function in Tocharian, and it justifies the connection with the Hittite ergative. Besides, on the basis of what comparison do you reconstruct split-ergativity?