I need some advice on PaleoCaspian

[hadad]

I've redone it a few times, and quite like it. Though I'm bothered by the fact that Nostratic and Amerind may not be valid language families with at least somewhat valid reconstructions. I'm tempted to start over and make a descendant of only ProtoIndoEuropean and go from there. Though in thinking of this, I wanted to know more about the internal interrelatedness of the indoeuropean languages. Here is what I already know, or assume to be true:

Italo-Celtic: Italic, Celtic
Germanic: Uncertain whether it'd be classified with BaltoSlavic
Balto-Slavic: Baltic, Slavic
Greek
Albanian
PaleoBalkan languages: Uncertain how they relate to eachother and their neighbors.
Indo-Iranian: Indo-Aryan, Nuristani, and Iranian.
Armenian
Tocharian
Anatolian

Aside from using glottochronology, what are some of the more deeper reconstructed stages? I found Glen Gordon's IndoAegean, but I'm not quite sure its accurate.

[Salmoneus]

I think it's generally thought to be wrong to force IE families into a definite tree model - there's not even consensus on Italo-Celtic. Instead, I think it's seen as more useful to consider the families as dialects of PIE that share certain areal features between them. Greek, for instance, is a language with augment, so groups with Indo-Iranian and Armenian. But it didn't have the Satem change, so in that respect it doesn't group with those two, whereas Balto-Slavic does, even though it lacks augment.

I think all that's agreed on is that Anatolian split before the other branches. Tocharian is often given as the next language to break away, although even then there are links to particular families.

Italic and Celtic are close, though this may be areal/borrowing/convergence rather than shared descent. Some people think there may be a particularly close link between Greek and Armenian, but Armenian diachronics are such a mess it's hard to say much for sure. Armenian also has many similarities to Indo-Iranian, although a lot of that is probably due to the extensive contact between them, and Greek and Balto-Slavic tend to be closer to Indo-Iranian, with Germanic, Albanian, Tocharian, Anatolian and Italo-Celtic further away from Indo-Iranian.

As for connections beyond PIE, nobody knows at all. Some people think some pre-IE European languages look like they're related to IE (the Old European Hydronymy). Of surviving languages, the best bet looks like a connection with Uralic, although I think some people still cling on to the older idea of a link to Afro-Asiatic. Somebody suggested a link to Tsimshianic in North America, but their methodology was found questionable.

[atman]

I think it's generally thought to be wrong to force IE families into a definite tree model - there's not even consensus on Italo-Celtic. Instead, I think it's seen as more useful to consider the families as dialects of PIE that share certain areal features between them. Greek, for instance, is a language with augment, so groups with Indo-Iranian and Armenian. But it didn't have the Satem change, so in that respect it doesn't group with those two, whereas Balto-Slavic does, even though it lacks augment.

I think all that's agreed on is that Anatolian split before the other branches. Tocharian is often given as the next language to break away, although even then there are links to particular families.

Italic and Celtic are close, though this may be areal/borrowing/convergence rather than shared descent. Some people think there may be a particularly close link between Greek and Armenian, but Armenian diachronics are such a mess it's hard to say much for sure. Armenian also has many similarities to Indo-Iranian, although a lot of that is probably due to the extensive contact between them, and Greek and Balto-Slavic tend to be closer to Indo-Iranian, with Germanic, Albanian, Tocharian, Anatolian and Italo-Celtic further away from Indo-Iranian.

As for connections beyond PIE, nobody knows at all. Some people think some pre-IE European languages look like they're related to IE (the Old European Hydronymy). Of surviving languages, the best bet looks like a connection with Uralic, although I think some people still cling on to the older idea of a link to Afro-Asiatic. Somebody suggested a link to Tsimshianic in North America, but their methodology was found questionable.

Everything you wrote looks to me. You summarized the consensus view of IE subgroups pretty well.

A couple of points:

Graeco-Armenian: Greek and Armenian really seem to share deep structural similarities; it's not just a matter of augments. However, Armenian has a well-known love for rare, typologically unlikely and at times plain weird sound changes, and this is a factor that helps obscure a possible close historical link. Another problem is that :hye: borrowed a lot, so much so that most of its inherited lexicon was already gone by the time Armenian was written for the first time. Bottomline: perhaps we'll never know whether and :hye: are really closely related, unless older (2nd-1st millennium BC) inscriptions in Armenian surfaced somehow.

Indo-Uralic: Long, long shot. A problem for this hypothesis could be that the speakers of Proto-Uralic (the Pit-Comb Ware culture) appear to have reached Northeastern Europe from the East, while the speakers of PIE (the Dnepr-Don/Yamna culture) look like they came from the West, from Europe. If the above is true, there seems to be no room for Indo-Uralic.

Old European Hydronymy/Europic family: This one looks more probable to me. In such a scenario, PIE would be the eastern-most language of a large family that would span the majority of Europe. Also, things like Pre-Germanic, Pre-Greek and perhaps also the Tyrrhenian languages (Etruscan, Rhaetian...) could also be related, more or less distantly, to this family.

IndoEuro-Tsimshianic: You serious??

[sangi39]

IndoEuro-Tsimshianic: You serious??

Maybe

[Salmoneus]

Indo-Uralic: Long, long shot. A problem for this hypothesis could be that the speakers of Proto-Uralic (the Pit-Comb Ware culture) appear to have reached Northeastern Europe from the East, while the speakers of PIE (the Dnepr-Don/Yamna culture) look like they came from the West, from Europe. If the above is true, there seems to be no room for Indo-Uralic.

I don't see why that's a problem. We wouldn't be talking about Uralic as a branch at PIE level, after all - we've got thousands of years of migration to play with. In any case, we don't really need it - PIE and PU are both supposed to have originated in exactly the same place, the Volga, only 2000 years apart (and with PIE moving downriver onto the step and PU mostly going north). At the time of PU, there were late PIE (or early IE) speakers just down the river from them. We could say that PIE is later than the archeological evidence suggests, and that early culture on the volga lead to both PU and PIE - or we could say that the timedepth is a little greater and they are two sister groups somewhere in the volga/ural/pontic-caspian area. Of course, this also means that borrowing may be a culprit rather than descent.

Old European Hydronymy/Europic family: This one looks more probable to me. In such a scenario, PIE would be the eastern-most language of a large family that would span the majority of Europe. Also, things like Pre-Germanic, Pre-Greek and perhaps also the Tyrrhenian languages (Etruscan, Rhaetian...) could also be related, more or less distantly, to this family.

Beeks is set against Pre-Greek being related to PIE, iirc. And I'm not sure there's much respectable evidence linking anything to Etruscan/Rhaetian.

IndoEuro-Tsimshianic: You serious??

Oh yes! In fact, it's not just one person. The parallels were noticed by Sapir in the early 20th century, and sporadically ever since. For instance, I'm looking at an article here by Tarpent in the 1990s where she semi-seriously suggests that IE looks a better fit for Tsimshianic than Penutian does, and in particular makes comparisons to Germanic. She clearly doesn't believe this, but says the links would normally be dismissed as coincidence but "could be suggestive to someone with and independent reason to believe in a genuine connection". But the famous thing along these lines is the paper by Dunn about ten years ago, where he 'proves' that there are systematic sound changes linking Tsimshianic with PIE (with particular affinity to Tocharian), in core vocabulary as well as in morphology, plus archeological evidence, and explaining in passing some loans into Uralic.

Of course, everyone but Dunn believes this must all just be a massive coincidence.

Oh, and one other connection: PIE-North Caucasian. I think it's Northwest Caucasian most often mentioned? But this is probably an areal thing: PIE just looks very typologically similar to caucasian languages, but I don't think there's been much said about specific soundchanges. Of course, they'd have to work out the diachronics of caucasian first. check back next century...

[hadad]

Alot of native american languages at first glance have similarities to IE. Though admittingly, I know nothing of that language. Kinda makes me want to revive my old idea for paleocaspian.
I have a hard time imagining languages that far afield being closely related to eachother. Then again, Malagasy is closely related to the Bornean austronesian languages.

Thanks guys!

[atman]

Oh yes! In fact, it's not just one person. The parallels were noticed by Sapir in the early 20th century, and sporadically ever since. For instance, I'm looking at an article here by Tarpent in the 1990s where she semi-seriously suggests that IE looks a better fit for Tsimshianic than Penutian does, and in particular makes comparisons to Germanic. She clearly doesn't believe this, but says the links would normally be dismissed as coincidence but "could be suggestive to someone with and independent reason to believe in a genuine connection". But the famous thing along these lines is the paper by Dunn about ten years ago, where he 'proves' that there are systematic sound changes linking Tsimshianic with PIE (with particular affinity to Tocharian), in core vocabulary as well as in morphology, plus archeological evidence, and explaining in passing some loans into Uralic.

Of course, everyone but Dunn believes this must all just be a massive coincidence.

Massive coincidence? I'm not sure. It's not that difficult to find correspondences between unrelated languages. It's even easier to find them when one postulates a collapse of the PIE stop system to just three phonemes like he does. And there isn't much core vocabulary in his lists, not to mention his ridiculously thorough morphology section...

Or maybe the Proto-Tshimshianic speakers just decided to buy from the internet a few thousand copies of Say It Yamna Style: Proto-Indo-European In Twenty (not so...) Easy Lessons

Beeks is set against Pre-Greek being related to PIE, iirc. And I'm not sure there's much respectable evidence linking anything to Etruscan/Rhaetian.

You remember correctly about Beekes's opinion on a possible PIE/Pre-Greek connection. On the other hand Etruscan seems to have several affixes in common with ancient IE languages, but that's not nearly enough to prove anything. These would have to be very "deep", very old connections anyway...

I don't see why that's a problem. We wouldn't be talking about Uralic as a branch at PIE level, after all - we've got thousands of years of migration to play with. In any case, we don't really need it - PIE and PU are both supposed to have originated in exactly the same place, the Volga, only 2000 years apart (and with PIE moving downriver onto the step and PU mostly going north). At the time of PU, there were late PIE (or early IE) speakers just down the river from them. We could say that PIE is later than the archeological evidence suggests, and that early culture on the volga lead to both PU and PIE - or we could say that the timedepth is a little greater and they are two sister groups somewhere in the volga/ural/pontic-caspian area. Of course, this also means that borrowing may be a culprit rather than descent.

I think I didn't explain clearly. I agree (most everyone does) that PU was spoken in the higher Volga/Kama river region, but I wanted to say that Pre-PUralic was most likely spoken in Asia, east of the Urals. The evidence for this would lie in the typological similarities between Uralic and other families of NE Asia (Yukaghir, Turkic, Altaic). After all Uralic and Altaic were even considered to be the same family until some decades ago. On the other hand, it appears to me that Pre-PIE speakers didn't come into the steppe from Asia, but from Europe (see the Old European Hydronymy, the possible links with Tyrrhenian languages). It seems to me (but I may be totally wrong ) that the further back you go, the further the histories of IE and Uralic diverge... if this is so, then the similarities between IE and Uralic may have to be explained with areal effects, borrowing and so on.

[Salmoneus]

Oh yes! In fact, it's not just one person. The parallels were noticed by Sapir in the early 20th century, and sporadically ever since. For instance, I'm looking at an article here by Tarpent in the 1990s where she semi-seriously suggests that IE looks a better fit for Tsimshianic than Penutian does, and in particular makes comparisons to Germanic. She clearly doesn't believe this, but says the links would normally be dismissed as coincidence but "could be suggestive to someone with and independent reason to believe in a genuine connection". But the famous thing along these lines is the paper by Dunn about ten years ago, where he 'proves' that there are systematic sound changes linking Tsimshianic with PIE (with particular affinity to Tocharian), in core vocabulary as well as in morphology, plus archeological evidence, and explaining in passing some loans into Uralic.

Of course, everyone but Dunn believes this must all just be a massive coincidence.

Massive coincidence? I'm not sure. It's not that difficult to find correspondences between unrelated languages. It's even easier to find them when one postulates a collapse of the PIE stop system to just three phonemes like he does. And there isn't much core vocabulary in his lists, not to mention his ridiculously thorough morphology section...

Yes of course, that's the longer version - it's never a matter of finding regular correspondences, it's a matter of finding sufficiently elegant ones.
But on the other hand, overly 'convenient' rules do happen sometimes. I think everyone now agrees that Tocharian is IE - but the example you give, the complete collapse of the fifteen-phoneme stop system into only three phonemes, is exactly what happens in Tocharian (aspiration is dropped, voicing is dropped, and palatovelars and labiovelars both merge with plain velars). [And then Tocharian completely reworks the vowel system, and deletes most of the vowels, and then inserts other ones. And then completely reworked all the morphology from scratch for good measure.]

I think I didn't explain clearly. I agree (most everyone does) that PU was spoken in the higher Volga/Kama river region, but I wanted to say that Pre-PUralic was most likely spoken in Asia, east of the Urals. The evidence for this would lie in the typological similarities between Uralic and other families of NE Asia (Yukaghir, Turkic, Altaic). After all Uralic and Altaic were even considered to be the same family until some decades ago. On the other hand, it appears to me that Pre-PIE speakers didn't come into the steppe from Asia, but from Europe (see the Old European Hydronymy, the possible links with Tyrrhenian languages). It seems to me (but I may be totally wrong ) that the further back you go, the further the histories of IE and Uralic diverge... if this is so, then the similarities between IE and Uralic may have to be explained with areal effects, borrowing and so on.

I don't see either line of argument as persuasive here. On the one hand, OEH and/or tyrrhenian could be related to IE without PPIE having come from the west. IE could represent the eastern branch and those others the western branch(es). Or they could have gone west and then gone east (or vice versa).
On the Uralic side, we a) don't know where those language were (and we know that Yukaghir and Uralic seem to have had early connections, but then Yukaghir also migrated from the west to its current location) and b) don't know what other languages there may have been around through which areal features may have spread. Areal features can, after all, spread over a very large area. Besides, west of the Urals and east of the Urals aren't far from one another - it's just the Urals between them. If Indo-Uralic were, say, from around the Volga area, pre-Uralic would have had at least two, probably four, maybe even more millenia to account for its whereabouts in before proto-Uralic came about - a quick hop to see what was on the far side of the Urals before coming back could easily be slotted into that time.

[You may want to discuss some of these issues over the the ZBB, btw]

[sangi39]

Yes of course, that's the longer version - it's never a matter of finding regular correspondences, it's a matter of finding sufficiently elegant ones.
But on the other hand, overly 'convenient' rules do happen sometimes. I think everyone now agrees that Tocharian is IE - but the example you give, the complete collapse of the fifteen-phoneme stop system into only three phonemes, is exactly what happens in Tocharian (aspiration is dropped, voicing is dropped, and palatovelars and labiovelars both merge with plain velars). [And then Tocharian completely reworks the vowel system, and deletes most of the vowels, and then inserts other ones. And then completely reworked all the morphology from scratch for good measure.]

That's why I freakin' love Tocharian! It's by far one of my favourite IE languages on the basis of phonology just because of how amazingly messed up it is.

You see so many (diachronic) conlangers either trying to split phonemes up or trying to maintain as much of the old phonological information as possible (so, say, if they drop coda consonants they might lengthen preceding vowels, so effectively the number of phonological contrasts is pretty much the same throughout the languages history). I've only seen a few conlangs where phonemes merged over the language's "historical" development but I've never seen anyone go nearly as far as Tocharian has and it's kind of a shame.

[atman]

I don't see either line of argument as persuasive here. On the one hand, OEH and/or tyrrhenian could be related to IE without PPIE having come from the west. IE could represent the eastern branch and those others the western branch(es). Or they could have gone west and then gone east (or vice versa).
On the Uralic side, we a) don't know where those language were (and we know that Yukaghir and Uralic seem to have had early connections, but then Yukaghir also migrated from the west to its current location) and b) don't know what other languages there may have been around through which areal features may have spread. Areal features can, after all, spread over a very large area. Besides, west of the Urals and east of the Urals aren't far from one another - it's just the Urals between them. If Indo-Uralic were, say, from around the Volga area, pre-Uralic would have had at least two, probably four, maybe even more millenia to account for its whereabouts in before proto-Uralic came about - a quick hop to see what was on the far side of the Urals before coming back could easily be slotted into that time.

[You may want to discuss some of these issues over the the ZBB, btw]

Well, yeah. Now I realize that the amounts of time (thousands of years) and space (millions of square kms) involved in these matters make almost any hypothesis feasible. All cultures in the North Eurasian lands had lots of time and room to migrate, split, unite and so on. Andthe fact that language families don't have to match archeological cultures exactly makes things even more complicated (and more fun ). Thanks!

[And then Tocharian completely reworks the vowel system, and deletes most of the vowels, and then inserts other ones. And then completely reworked all the morphology from scratch for good measure.]

I've used something like the new case innovation in Tocharian for my own Atlántiqa innovated cases...

And talking about that: what do you think of the backstory of the Atlántiqa speakers?

Atlántiqa is (would be?) the language of the descendants of Pre-Classical Greeks who (would have) settled in Southwestern Spain, Portugal and the islands of Macaronesia. Then, as the centuries go by, the continental population is gradually assimilated, while the insular group remains independent and fairly isolated from Mediterranean geopolitics. Then the various Hellenic towns in Macaronesia would form a united kingdom, and later still (after contact with the european naval powers in the Renaissance and some cultural europeization) a more and more modern republic (with UN, EU memberships and the likes). Ideas? Questions? Suggestions? Criticism?

[Salmoneus]

Oh, sorry, didn't see that PM.

I suppose I'd look at it bit by bit.

Where is 'macaronesia', in this case? That covers four archipelagos that are quite some distance from one another (the Azores I think are about 1500 miles from Cape Verde, or about the distance between London and Moscow...). Are you thinking of a culture spanning all of those islands, or just one area?

So then there's getting there. Canaries, no problem at all, indeed Greeks may well have been there (Carthaginians, Phoenicians and Romans all got there). Madeiras is a little more implausible, particularly for organised settlers, although not impossible, particularly if by shipwreck. Cape Verde is probably possible in terms of getting there, but implausible than anyone would bother, given how far it is from anywhere. The Azores are probably out - I don't think Greeks could even get shipwrecked that far from land.

Then there's what would happen to them. Unfortunately, we probably know the answer to this - I think the carthaginians are probably the source of the native canarian population (genetically they're closest to the berbers of the atlas mountains, but the berbers of the atlas mountains are not noted for their seamanship; however, they did culturally and genetically mix with the phoenician settlers who founded carthage. So I think that the guanches were probably carthaginian sailors originally, or at least their berber slaves/employees). And the problem is, they didn't form a united kingdom - they had half a dozen 'kingdoms' on each little island. There were nine kingdoms on the single island of Tenerife alone. In the absence of substantial contact with the outside world, I'm really not convinced that these islands have the resources necessary to sustain large complicated kingdoms - collapse into little local chiefdoms seems much more likely.

Then there's the problem of Europe. What's going to stop Portugal and Spain from conquering these people in the 15th century - as they did in reality?

So I don't think there's anything impossible here, but it does look super-probable to me either.

I suppose that if I were doing it, I'd have the greeks devolve to the level of petty tribes (which I guess isn't too much of a devolution if they're pre-classical to begin with), and then get conquered by the spanish or portuguese. Just have them be a little better at maintaining their language/identity than the guanches were, and eventually demand independence (like Cape Verde)

[loglorn]

And then Tocharian completely reworks the vowel system, and deletes most of the vowels, and then inserts other ones. And then completely reworked all the morphology from scratch for good measure.

Does anybody know why that happened?

[Lambuzhao]

Out of loneliness, boredom, and a handful of small but persistent aphasias/solecisms & speech impediments.

A little more seriously - Probably from various prolonged contacts with some very, very different languages.

[atman]

Probably from various prolonged contacts with some very, very different languages.

Very, very Turkic languages if you ask me. But in my opinion, the Tocharian languages don't seem to me much more innovative overall than their western contemporaries, like Old English, Old Church Slavonic...

[Salmoneus]

Probably from various prolonged contacts with some very, very different languages.

Very, very Turkic languages if you ask me. But in my opinion, the Tocharian languages don't seem to me much more innovative overall than their western contemporaries, like Old English, Old Church Slavonic...

You mean other than their entire consonant system, their entire vowel system and their entire verbal system?

[atman]

You mean other than their entire consonant system, their entire vowel system and their entire verbal system?

Well, at least Tocharian had many conservative features not often found in the Western IE langs: if I remember well it was still largely SOV, postpositional, very inflected and synthetic, kept the PIE stops as stops and didn't turn them into affricates/fricatives (even if they became just three, and unless they were palatalized later), kept the dual in verbs and perhaps also nouns...

And after all :hye: did even crazier things in the historical phonology department...