Latest language family proven?

[Atom]

In general, loanwords are very poorly controlled for by most comparative linguists working with East Siberian languages (of whatever stock or mesh) -- IMO at least!

This is a matter which bugs me about many long range comparison attempts. Most of these people don't really compare languages; they merely compare dictionaries. (For an especially nutty case, see Octaviano.) It is easy to do that - but can be thrown lightyears off the scent by layers of loanwords. And massive lexical borrowing happens FREQUENTLY. Languages such as English, which have borrowed almost the entire cultural vocabulary, plus a sizeable part of the basics, from other languages, are the norm, not the exception.

Which is why it's important to look at morphology, all the decent evidence for a relationship between PIE and Proto-Uralic come from morphology.

Just that. There aren't many lexical cognate candidates, and most of those rather look like loanwords. Yet, the morphological systems of IE and Uralic look so similar that one cannot easily dismiss those resemblances as accidental, and languages usually don't borrow entire morphological paradigms.

(I hate to turn the conversation, but I really want to know) how widely accepted is the PIE/Proto-Uralic connection in academic linguistics? Like, a familial connection.

[Drydic]

It's not proven because most people concerned with it are still trying to figure out exactly what the connection is. Lots of grammatical similarities but few-ish lexical cognates*. Is it just a simple brother-sister relationship in the greater Nostratic framework or is there something more than that, a possible shared period of development? They're related, a good portion of people who aren't dyed-in-the-wool splitters agree on that, but the how isn't as clear.


*I just KNOW this part of the statement is going to bite me in the ass somehow (lexical).

[WeepingElf]

It's not proven because most people concerned with it are still trying to figure out exactly what the connection is. Lots of grammatical similarities but few-ish lexical cognates*. Is it just a simple brother-sister relationship in the greater Nostratic framework or is there something more than that, a possible shared period of development? They're related, a good portion of people who aren't dyed-in-the-wool splitters agree on that, but the how isn't as clear.

Yes, that is pretty much true. Few linguists would opine that IE and Uralic couldn't be related, and many feel that the morphologies are similar enough to assume a relationship, but the sound correspondences are difficult to reconstruct because of the small amount of cognate material, and it's sound correspondences where it is at in historical linguistics.

[linguist.in.hiding]

It's not proven because most people concerned with it are still trying to figure out exactly what the connection is. Lots of grammatical similarities but few-ish lexical cognates*. Is it just a simple brother-sister relationship in the greater Nostratic framework or is there something more than that, a possible shared period of development? They're related, a good portion of people who aren't dyed-in-the-wool splitters agree on that, but the how isn't as clear.

Yes, that is pretty much true. Few linguists would opine that IE and Uralic couldn't be related, and many feel that the morphologies are similar enough to assume a relationship, but the sound correspondences are difficult to reconstruct because of the small amount of cognate material, and it's sound correspondences where it is at in historical linguistics.

I'm really going to regret coming out of hiding... And for this...

Now, quote _even one_ contemporary Uralicist that believes in a real connection between IE and Uralic. Vague hand waiving doesn't count. BTW, I'm really astonished about the amount of people here giving any serious thought about Nostratic.

[Tropylium]

Eugene Helimski would be one. In particular, see the last paper in the link'd section of his homepage:
"Polemics: The kinship theory (Nostratic theory) explains the cumulative evidence of the most ancient Indo-Uralic parallels better than any contact theories. Issues of direct and lateral (areal) kinship, gradual loss of kinship transparency, etc."

Anyway, an IE-Uralic connection pretty much requires a major collapse of the vowel and sibilant systems on the IE side + an equally major collapse of the stop system on the Uralic side. This makes the probativ value of even regular correspondences fairly weak: a root like Uralic *pala "bit" could be counted as a match for IE roots of at least shapes *pel *bʰel *bel, perhaps *pol *bʰol *bol; likewise, an IE root like *sem "1" could be counted as a match for dozens of Uralic roots, of shape *{ć, č, ś, š, s}{a, e, ë, i, o, u, ü, ä}m{a, i}. Without an internal explanation or external corroboration (say, if Altaic is also related) for these contrasts, I think it's unlikely for any Indo-Uralic theory to progress beyond "probable but still unproven".

[WeepingElf]

Eugene Helimski would be one. In particular, see the last paper in the link'd section of his homepage:
"Polemics: The kinship theory (Nostratic theory) explains the cumulative evidence of the most ancient Indo-Uralic parallels better than any contact theories. Issues of direct and lateral (areal) kinship, gradual loss of kinship transparency, etc."

Anyway, an IE-Uralic connection pretty much requires a major collapse of the vowel and sibilant systems on the IE side + an equally major collapse of the stop system on the Uralic side. This makes the probativ value of even regular correspondences fairly weak: a root like Uralic *pala "bit" could be counted as a match for IE roots of at least shapes *pel *bʰel *bel, perhaps *pol *bʰol *bol; likewise, an IE root like *sem "1" could be counted as a match for dozens of Uralic roots, of shape *{ć, č, ś, š, s}{a, e, ë, i, o, u, ü, ä}m{a, i}. Without an internal explanation or external corroboration (say, if Altaic is also related) for these contrasts, I think it's unlikely for any Indo-Uralic theory to progress beyond "probable but still unproven".

Yes, the IE and Uralic phoneme inventories are structured very differently; if there was a Proto-Indo-Uralic at all, both PIE and PU must have gone through very sweeping changes of their phonologies, and the sound correspondences are certainly complex and not easy to track. Yet, I would not consider the endeavour hopeless, and maybe we'll see the proof being delivered during our lifetimes (i.e., before 2050).

[Drydic]

BTW, I'm really astonished about the amount of people here giving any serious thought about Nostratic.

Not everyone is a dyed-in-the-wool splitter, as I said. If the sound correspondences work out, why wouldn't we? I do of course admit Greenberg's left a bad taste in the mouth for any long-range comparisons. I really do want to know what in fuck's name he was thinking when he proposed Amerind.

[TaylorS]

It's not proven because most people concerned with it are still trying to figure out exactly what the connection is. Lots of grammatical similarities but few-ish lexical cognates*. Is it just a simple brother-sister relationship in the greater Nostratic framework or is there something more than that, a possible shared period of development? They're related, a good portion of people who aren't dyed-in-the-wool splitters agree on that, but the how isn't as clear.

Yes, that is pretty much true. Few linguists would opine that IE and Uralic couldn't be related, and many feel that the morphologies are similar enough to assume a relationship, but the sound correspondences are difficult to reconstruct because of the small amount of cognate material, and it's sound correspondences where it is at in historical linguistics.

I'm really going to regret coming out of hiding... And for this...

Now, quote _even one_ contemporary Uralicist that believes in a real connection between IE and Uralic. Vague hand waiving doesn't count. BTW, I'm really astonished about the amount of people here giving any serious thought about Nostratic.

I'm a Nostraticist of a sort, but I think most stuff out there is pure fantasy. A big problem, IMO, is that they try to force Afro-Asiatic and Dravidian into it.

IMO there is, at the very least a grouping WeepingElf has called "Mitian", so-called because so many of the languages in Northern Eurasia have M-root 1st Person pronouns and T-root 2nd Person pronouns.

[WeepingElf]

I'm a Nostraticist of a sort, but I think most stuff out there is pure fantasy. A big problem, IMO, is that they try to force Afro-Asiatic and Dravidian into it.

IMO there is, at the very least a grouping WeepingElf has called "Mitian", so-called because so many of the languages in Northern Eurasia have M-root 1st Person pronouns and T-root 2nd Person pronouns.

I agree with you that trying to force Afro-Asiatic and Dravidian into a relationship with the Mitian languages burdens the matter with problems and gets in the way of finding solutions. For instance, Bomhard, by insisting on matching the IE laryngeals with the Semitic ones, misses the probable identity of PIE 1sg. stative *-h2e and Uralic *-k. His work certainly is not all rubbish, but at least some parts of it are highly questionable.

The "Mitian" pronoun roots are not easy to dismiss. Sure, two pronoun roots can match by accident, but we are dealing with a large, contiguous area here where almost every language family has these pronouns - that very much looks like an ancient language family.

[linguist.in.hiding]

Tropylium: "Eugene Helimski would be one."

No he wouldn't. I was afraid something like this would happen. Uralicists were baffled by how could Helimski really believe in Nostratic. Anyway, some people did ask. Helimski confessed that he didn't really believe in Nostratic. He held it as a fruitful idea to toy with. Quite like taking Romance and Hellenic loanwords in English. Loanwords, areal connections, but a far cry from genetic connections. This is how I recall it explained to me. I know it goes opposite of what he wrote there but such is life.

[gsandi]

I'm a Nostraticist of a sort, but I think most stuff out there is pure fantasy. A big problem, IMO, is that they try to force Afro-Asiatic and Dravidian into it.

IMO there is, at the very least a grouping WeepingElf has called "Mitian", so-called because so many of the languages in Northern Eurasia have M-root 1st Person pronouns and T-root 2nd Person pronouns.

I agree with you that trying to force Afro-Asiatic and Dravidian into a relationship with the Mitian languages burdens the matter with problems and gets in the way of finding solutions. For instance, Bomhard, by insisting on matching the IE laryngeals with the Semitic ones, misses the probable identity of PIE 1sg. stative *-h2e and Uralic *-k.

That's assuming that PIE *-h2e was in its origins the marker of the 1st person and not of the stative. The second person singular of the perfect (which is somehow related to the stative) may well be *-th2e, which looks like the combination of the second-person marker -t + -h2e.

If we wanted to match up Uralic *k, we may as well go for the PIE subject pronoun *eg(h)-.

[TaylorS]

The PIE 1SG Nominative pronoun seems to have originally been a verb meaning something like "I-here"

[TaylorS]

To elaborate further, I suspect the Middle PIE verbal system looked something like this:

Active Punctive: >>> Aorist, Preterite
1SG: -m
2SG: -s
3SG: -t
1PL: -men
2PL: -ten
3PL: -nt

Active Durative: >>> Present, Imperfect
1SG: -mi
2SG: -si
3SG: -ti
1PL: -meni
2PL: -teni
3PL: -nti

Stative Punctive: >>> Perfect
1SG: -h2e
2SG: - t(h2)e (intrusive h2 by analogy???)
3SG: -e
1PL: ???
2PL: ???
3PL: ???

Stative Durative: >>> Mediopassive
1SG: -h2or
2SG: -t(h2)or
3SG: -or
1PL: ???
2PL: ???
3PL: ???

[Sleinad Flar]

^ This cannot be right, as the secondary markers (-m, -s, -t) marked both the imperfect and the aorist. In fact, a root aorist is formally indistinguishable from an imperfect to a root present. The primary markers (-mi, -si, -ti) seem to have been used only in the present and (optionally) the subjunctive. The -i was possibly a locative/temporal (here-and-now) marker. The categories punctual-durative were marked lexically or by stem formation.

As for the middle, in fact two sets of endings are reconstructed (both with primary and secondary forms). The first (-h2o(r),-th2o(r),-o(r)) looks suspisiously like the perfect (stative) endings, the second like a mixture of active and middle/perfect endings (-m-h2e(i), -s-o(i)/-s-h2e(i)), -t-o(i)). The -r extension seems to have been a primary marker, just like -i, and not a mediopassive marker (IIRC sanskrit still has aorist mediopassives without -r or -i). The difference between perfect and middle I seems to have been -e vs. -o.

Then of course there's the enigma of thematic 1 sg -o: (<-o-h2?) (and possibly 3 sg -e-i), which seems to connect with the stative rather than present endings...

[jal]

that very much looks like an ancient language family.

Or a very early pronoun borrowing, of course.


JAL

[kuroda]

Wow. There really isn't any thread topic here that doesn't degenerate into IndoEuropeanist squalor, is there?

;p

Kim

[WeepingElf]

I'm a Nostraticist of a sort, but I think most stuff out there is pure fantasy. A big problem, IMO, is that they try to force Afro-Asiatic and Dravidian into it.

IMO there is, at the very least a grouping WeepingElf has called "Mitian", so-called because so many of the languages in Northern Eurasia have M-root 1st Person pronouns and T-root 2nd Person pronouns.

I agree with you that trying to force Afro-Asiatic and Dravidian into a relationship with the Mitian languages burdens the matter with problems and gets in the way of finding solutions. For instance, Bomhard, by insisting on matching the IE laryngeals with the Semitic ones, misses the probable identity of PIE 1sg. stative *-h2e and Uralic *-k.

That's assuming that PIE *-h2e was in its origins the marker of the 1st person and not of the stative. The second person singular of the perfect (which is somehow related to the stative) may well be *-th2e, which looks like the combination of the second-person marker -t + -h2e.

It's not impossible that your analysis is correct; however, that doesn't disprove the equation because that may already have happened in Proto-Indo-Uralic.

If we wanted to match up Uralic *k, we may as well go for the PIE subject pronoun *eg(h)-.

Which may be from a verb meaning something like 'I am here'.

[TaylorS]

^ This cannot be right, as the secondary markers (-m, -s, -t) marked both the imperfect and the aorist. In fact, a root aorist is formally indistinguishable from an imperfect to a root present. The primary markers (-mi, -si, -ti) seem to have been used only in the present and (optionally) the subjunctive. The -i was possibly a locative/temporal (here-and-now) marker. The categories punctual-durative were marked lexically or by stem formation.

As for the middle, in fact two sets of endings are reconstructed (both with primary and secondary forms). The first (-h2o(r),-th2o(r),-o(r)) looks suspisiously like the perfect (stative) endings, the second like a mixture of active and middle/perfect endings (-m-h2e(i), -s-o(i)/-s-h2e(i)), -t-o(i)). The -r extension seems to have been a primary marker, just like -i, and not a mediopassive marker (IIRC sanskrit still has aorist mediopassives without -r or -i). The difference between perfect and middle I seems to have been -e vs. -o.

Then of course there's the enigma of thematic 1 sg -o: (<-o-h2?) (and possibly 3 sg -e-i), which seems to connect with the stative rather than present endings...

This is a rough sketch, admittedly. It's inspired by some of the ideas from the the blog Paleoglot, but I disagree with Glenn Gordon's hypothesis about the semantics of the personal endings of PIE prior to the split-off of Anatolian. PIE verbs are a mess of crazy irregularities and it's hard making sense of it all!

[Tropylium]

No he wouldn't. I was afraid something like this would happen. Uralicists were baffled by how could Helimski really believe in Nostratic. Anyway, some people did ask. Helimski confessed that he didn't really believe in Nostratic. He held it as a fruitful idea to toy with. Quite like taking Romance and Hellenic loanwords in English. Loanwords, areal connections, but a far cry from genetic connections. This is how I recall it explained to me. I know it goes opposite of what he wrote there but such is life.

If you're after someone thinking that a reasonably complete Indo-Uralic theory exists, then yes, I'm pretty sure not even most of those who "subscribe to" Nostratic would sign anything like that.

[Richard W]

[quote="Tropylium"]Eugene Helimski would be one. In particular, see the last paper in the link'd section of his homepage:
"Polemics: The kinship theory (Nostratic theory) explains the cumulative evidence of the most ancient Indo-Uralic parallels better than any contact theories. Issues of direct and lateral (areal) kinship, gradual loss of kinship transparency, etc."[/home]
To find this, one now needs to use the way-back machine, which got me to http://web.archive.org/web/201107111618 ... /2.171.PDF .