The Great Proto-Indo-European Thread's Sequel

[Kuchigakatai]

But how ‘2P clitic’-like are these really?

Very, most of them.

Actually, now that you ask about it, I think -ve 'or' shouldn't count, as I think it's purely used to join nouns/verbs/adjectives within a sentence, in a way that's not at all necessarily in Wackernagel's position of course, like, say, senātus populusve.

=que, for instance, was inserted after the second conjuncted NP. (The famous example is of course Senatus Populus=que; 2P would be *Senatus=que Populus.)

Oh, that's the "que" that joins phrases within a sentence. I'm thinking of another -que here, the one that joins whole clauses. A rather famous example:

1 Gallia est omnis divisa in partes tres, quarum unam incolunt Belgae, aliam Aquitani, tertiam qui ipsorum lingua Celtae, nostra Galli appellantur. 2 Hi omnes lingua, institutis, legibus inter se differunt. Gallos ab Aquitanis Garumna flumen, a Belgis Matrona et Sequana dividit. 3 Horum omnium fortissimi sunt Belgae, propterea quod
a cultu atque humanitate provinciae longissime absunt,
minimeque ad eos mercatores saepe commeant atque ea quae ad effeminandos animos pertinent important,
4 proximique sunt Germanis qui trans Rhenum incolunt, quibuscum continenter bellum gerunt.

1 Gallia is divided in three parts, one of which is inhabited by the Belgians, another of which by the Aquitani, and the third part by those who in their own language call themselves "Celts", or "Galli" in our language. 2 These peoples differ in language, institutions and laws. The Garonne river separates the Galli from the Aquitani, and the Marna and the Seine rivers separate them from the Belgians. 3 Amongst these, the Belgians are the strongest, because
they are furthest from civilization,
["and" because] merchants very rarely travel there for them to import the things that turn spirits effeminate,
4 ["and" because] they're closest to the Germans who live beyond the Rhine, who they're in constant war with. (Julius Caesar, On the Gallic War 1.1)

Here, proptereā quod 'because' takes three clauses (which I tried to creatively separate with line breaks and a little teal colour on the first two letters). The three subclauses are coordinated by two 2nd-position (Wackernagel-position) -que that attach to whatever the first stressed word is. First you see it with minimē 'very little' (part of the adverbial minimē saepe 'very rarely', literally "very-little often"), then with proximī 'closest' (proximī sunt Germānīs, close.SUPERL.NOM.PL be.3PL German.PL.DAT 'they're closest to [the] Germans [who...]').

As for the rest, though they’re certainly 2P, the orthography seems to indicate phonologically independent words rather than clitics. (The situation seems reminiscent of Tagalog, which also has 2P ‘clitics’ which are really separate words.)

Conversely, I'd say you shouldn't let yourself be misled by the orthography. Does what I said about them being able to interrupt even proper names not surprise you?

Marcus Tullius Cicerō 'Mark Tully Chickpea'
Marcus enim Tullius Cicero vēnit. 'This is because M. Tullius Cicerō came by.'

They regularly break whatever phrase they can break if needed in order to be attached as enclitics:

Tribus igitur modis video esse a nostris de amicitia disputatum. (Cicero, On the Ends of Good and Evil 1.66)
'So I see that we have discussed friendship in three ways.'

Tribus modīs, three.PL.ABL way.PL.ABL, 'in three ways', is one noun phrase, but igitur needs to be attached at the end of the first stressed word (tribus 'three'), and so happily interrupts it. This is so pervasive with all the words I listed above (-ne, enim, autem, igitur...) I'd say it's evidence they're enclitics, even if written separately.

Also, for the most part, these pseudo-conjunctions/adverbs actually exclude each other. Basically you don't normally have two of them, like -que + enim, or enim + autem, or vērō + igitur. However, as an exception, -ne can indeed be combined with the others... You can have "Tribusne autem modīs [nōs...]" 'But is it that in three ways [we...]?'

In Romance, every single one of them was abandoned. And mihi/tibi lost their ability to be in Wackernagel's position too.

Interesting! So, if the clitics were abandoned, then what constructions arose to replace them?

In a nutshell, the same constructions you find in modern English.

[bradrn]

=que, for instance, was inserted after the second conjuncted NP. (The famous example is of course Senatus Populus=que; 2P would be *Senatus=que Populus.)

Oh, that's the "que" that joins phrases within a sentence. I'm thinking of another -que here, the one that joins whole clauses.

Interesting, I hadn’t known about that one. (I don’t know Latin, so perhaps I shouldn’t be making these statements…)

As for the rest, though they’re certainly 2P, the orthography seems to indicate phonologically independent words rather than clitics. (The situation seems reminiscent of Tagalog, which also has 2P ‘clitics’ which are really separate words.)

Conversely, I'd say you shouldn't let yourself be misled by the orthography. Does what I said about them being able to interrupt even proper names not surprise you?

No, not in the least. In fact this seems to be a rather common behaviour of 2P clitics. (Well, in IE at least; I’m not quite sure about other families.) Anyway, proper names aren’t even single words phonologically.

They regularly break whatever phrase they can break if needed in order to be attached as enclitics: … Tribus modīs, three.PL.ABL way.PL.ABL, 'in three ways', is one noun phrase, but igitur needs to be attached at the end of the first stressed word (tribus 'three'), and so happily interrupts it. This is so pervasive with all the words I listed above (-ne, enim, autem, igitur...) I'd say it's evidence they're enclitics, even if written separately.

Ah, and here we start to encounter the messy question: what exactly is a clitic, anyway? The standard definition seems to be that it’s a phonologically bound entity which is positioned according to syntactic rules — but this seems to be the exact opposite, namely a phonologically independent element which is positioned according to non-syntactic rules. Nonetheless, people seem to call both of these things ‘clitics’.

[KathTheDragon]

"Clitics" are an ill-defined and not really definable category. The best answer is that they're any (language-dependent) category intermediate between affix and independent word. Obviously further details would be added on a language-specific basis.

[bradrn]

True, I had forgot about Hittite (which indeed has a rather elaborate clitic system). I’m similarly unsurprised that Vedic Sanskrit has them as well — most of the conservative IE languages do. But I hadn’t known about clitics in Old Irish, which is rather divergent; do you have any more information on this?

This paper by John Koch addresses early Celtic syntax more broadly, though quite thoroughly, including the development of preverbal elements and the emergence of the absolute/conjunct verb distinction, but does address the treatment of Wackernagel enclitics as well.

Only just got around to reading (well, skimming) through this, and it looks pretty interesting — thank you! But given that I know nothing of Celtic diachronics, and the paper looks somewhat dense, do you think I would be able to read it just now, or would I need to look at some more introductory stuff before I can understand it properly?

[Znex]

Only just got around to reading (well, skimming) through this, and it looks pretty interesting — thank you! But given that I know nothing of Celtic diachronics, and the paper looks somewhat dense, do you think I would be able to read it just now, or would I need to look at some more introductory stuff before I can understand it properly?

I think it's more just making it through the prose than needing background knowledge on Celtic diachronics, but if you get stuck on anything, feel free to pm me about it. I've been studying Celtic on and off for my conlangs and for understanding pre-Roman Celtic.

[Zju]

"Clitics" are an ill-defined and not really definable category. The best answer is that they're any (language-dependent) category intermediate between affix and independent word. Obviously further details would be added on a language-specific basis.

Isn't the very concept of what constitutes an independent word not well defined in the first place? Gotta love the unsolved problems in linguistics.

[KathTheDragon]

"Clitics" are an ill-defined and not really definable category. The best answer is that they're any (language-dependent) category intermediate between affix and independent word. Obviously further details would be added on a language-specific basis.

Isn't the very concept of what constitutes an independent word not well defined in the first place? Gotta love the unsolved problems in linguistics.

Well this is why I said "language-dependent". It's easier to answer in any single language than for all languages simultaneously (and in fact I'm of the opinion that there is no universal definition of "word")

[Nortaneous]

would have to dig more than I have time to rn to find it but I think I've seen it claimed that some languages (Coptic?) can be adequately described without reference to the category of "word"

[Zju]

Is stress the only difference between syntax and purely concatenative morphology anyway? Syntax already places significant restrictions on the positioning of some elements, e.g. clitics, but also determiners, and, in some languages, adverbs.

[Talskubilos]

Although Yahoo! Groups closed some time ago, I've found out an archive of the IE group cybalist by Richard Wordingham.

[WeepingElf]

Yes, that's a valuable resource.

[missals]

Not a real scholarly observation to make here, more of a gripe, but I've been looking at a lot of PIE etyma lately due to the class I'm teaching, and I feel like the reconstructed PIE roots without at least two consonants (with solidly regular reflexes in multiple branches) are kind of fake... The numerous reconstructed "suffixes" with vague meanings attached liberally to all kind of roots also seem kind of suspect—a very convenient way to make cognates "fit"— but of course, modern languages such as English have many such artifacts, so it's no surprise that PIE would have such artifacts too.

But there's also the thing where laryngeals and ablaut seem to allow scholars to take considerable liberties with vowel correspondences—"Oh, well it must have been a different ablaut grade"—which again, I do get, since English has things like food/feed and old/older/elder/elderly/*olderly/*eld that are not synchronically explainable and may appear as mysteries to a scholar in the distant future trying to reconstruct Proto-English—but, nonetheless, when Indo-Europeanists have both a suite of suffixes and various kinds of ablaut at their disposal, seeing a "root" like h₂eǵ- with such a general meaning, appearing in such a diverse range of postulated derivations with extremely semantically-divergent reflexes... it makes me doubt it somewhat.

We all know what a joke it is when proponents of various "macrofamilies" reconstruct "words" like *sV that play fast and loose with consonants and seemingly regard vowel correspondences as optional... It almost appears to me that Indo-Europeanists may have a backdoor way of doing this. This is not, of course, to doubt the existence of Indo-European or even to cast doubt on the core elements or broad strokes of reconstructed PIE... but among the thousands of roots and words reconstructed for PIE, I somewhat want to suspect that quite a large proportion are simply junk.

[Ketsuban]

The main thing that's suspicious about *h₂eǵ- to my eyes is that there's no reflexes (e.g. Anatolian) listed which indicate it's not just *aǵ-. Indo-Europeanists may be a little overzealous in reconstructing a laryngeal when we can reasonably suspect that PIE had some unalloyed /a/ (an example might be *albʰós "white", Hittite alpas) and it's only paucity of data that prevents us identifying them.

[hwhatting]

but, nonetheless, when Indo-Europeanists have both a suite of suffixes and various kinds of ablaut at their disposal, seeing a "root" like h₂eǵ- with such a general meaning, appearing in such a diverse range of postulated derivations with extremely semantically-divergent reflexes... it makes me doubt it somewhat.

That's why it's important to look at the exact argumentation for each root and form. We know quite a lot about which ablaut grades were used for various formations, what ablaut grades went together with which suffixes; what kind of semantic relationship were established by various foms of derivation; what paths different inflection classes took in the individual branches, etc. Anlauting laryngeals left traces in compounds, reduplicated and prefixed forms, and in the accent system of Balto-Slavic. If you want to know how solid a reconstruction is, you have to check the sources given in the reference works, and often to go back to the original articles themselves.
Concerning *h2eg'-, there is evidence for zero-grade forms with -ij- (< *-H2g'-) in Vedic, so I don't see much reason to doubt this reconstruction.

[WeepingElf]

Certainly, there are many dubious forms in circulation. Especially Pokorny's dictionary, on top of being more than 60 years old and thus to a great part outdated (yet everyone still uses it because there simply is no more recent one), heavily over-reconstructs - there just can't have been that many homonyms and synonyms - and contains many items with skewed distributions, which means that at least some of these were loanwords from unknown languages into the IE languages (or PIE dialects) where they occur. (This, of course, makes the dictionary more useful for those who hunt such substratum words!)

[Richard W]

There seems to be some active (or unresolved) debate over the origin of Sanskrit jñāti 'kinsman'. The two candidate roots are *ɡ̂enh₁ 'to ɡenerate' and *ɡ̂neh₃ 'to know'. Semantically, a shift from 'kith' to 'kin' does not seem difficult. How does one derive jñāti from *ɡ̂enh₁?

[Kuchigakatai]

There seems to be some active (or unresolved) debate over the origin of Sanskrit jñāti 'kinsman'. The two candidate roots are *ɡ̂enh₁ 'to ɡenerate' and *ɡ̂neh₃ 'to know'. Semantically, a shift from 'kith' to 'kin' does not seem difficult. How does one derive jñāti from *ɡ̂enh₁?

Maybe such people have words like Latin cognātus 'related by blood, kindred; sibling' in mind.

[Richard W]

But that is traced back to zero grade formation *ǵn̥h₁tós, which is supposed to yield Sanskrit jāta, via Proto-Indo-Iranian * *ȷ́aHtá-.

[hwhatting]

There seems to be some active (or unresolved) debate over the origin of Sanskrit jñāti 'kinsman'. The two candidate roots are *ɡ̂enh₁ 'to ɡenerate' and *ɡ̂neh₃ 'to know'. Semantically, a shift from 'kith' to 'kin' does not seem difficult. How does one derive jñāti from *ɡ̂enh₁?

Two possibilities - a contamination of *jāti- with other forms where jñV- is regular, or assuming a variant *ɡ̂neH1-ti-. In my undertanding the discussion is about whether such forms are acceptable and a besser explanation than dervivation from ɡ̂neH3-.

[abahot]

Hey everyone! I'm very new here and am thrilled that there's a place where people not only discuss Proto-Indo-European, but also their own personal theories about it. I've been reading all the posts (both here and on the original thread) and see a bunch of things I've thought of before and many new things as well.

(Crackpot theory incoming)

I do have one question, though -- on the Wikipedia page for "Indo-European migrations" there's a graphic for the locations of the earliest Indo-European river names. Am I misinterpreting this graphic, as one would expect the earliest names to be around the Pontic-Caspian steppe? If not, perhaps this region represents the homeland of pre-Proto-Indo-European (and related "para-Indo-European" languages), whose speakers would have left southeastwards for the Pontic-Caspian steppe a millennium or two before the days of PIE. The remaining para-IE languages in the region would have been erased by later Indo-European and/or Uralic migrations. If so, para-IE languages are a particularly interesting explanation to at least some of the claimed Finno-Volgaic substrate words, like *täštä "star", or *kümmin "ten", which appear suspiciously similar to the equivalent PIE words.