My proposed tree of Indo-European languages

[DuPidh]

This 1976 analysis is too old and wrong
Read this from latest haplogroup report
.
Iapypians are messapians........they are now know as East-Balkan people ....thracians............so clearly thracians are not illyrians and neither are messapians
.
https://macsphere.mcmaster.ca/handle/11375/22810
.
https://indo-european.eu/tag/iapygian/
.
Iron Age southern Italians likely descended from early to late Neolithic farmers from Anatolia and possibly as far East as the Caucasus, and from migrants arriving from eastern Europe around the late Neolithic/early Bronze Age. These findings support previous hypotheses that the ancestors of the Iapygians may have originated in the eastern Balkan region, or derive shared ancestry with a common source population from eastern Europe.
.
The Greek mythographers, as usual, derived the name from a hero, Iapyx, whom they represented as a son of Lycaon, a descent probably intended to indicate the Pelasgic origin of the Iapygians. (Anton. Liberal. 21; Plin. Nat. 3.11. s. 16.) For a further account of the national affinities of the different tribes in this part of Italy, as well as for a description of its physical geography, see the articles APULIA and CALABRIA.
.
so if albanians think they match messapian language , then they are not illyrian, but most likely thracian, Albania could have had thracians before others arrives ...............I mentioned this to Salento about 2 months ago.
.
To finalise, we have no illyrian script or language ...all we have is illyrian personnel names from Noricum ( east-austria)

Wrong, wrong!. You get too many things wrong! The most numerous inherited words from Illyrian are not personal names, but toponims. All toponims are still in use in Albanian language and have meanings.
Had Albanians been Thracians they would have today been genetically close to Bulgarian insisted of Toscans and South Italians. Also Albanians had not come from Illyrians should not have CELTIC dna which they have,. So genetics and linguistic combined puts Albanians firmly where they are, in Illyrian lands.
What amazes me is that you use very little to no reasoning in making your statements!

 

[ToBeOrNotToBe]

There are many Tarim mummies from different periods. The ones that were tested date from the Scythian period and were R1a. Modern Uyghurs have both R1b-L23 and R1a-Z93, and yet the Scythians hardly had any R1b (at least not as much proportionally as the Uyghurs), so it's likely that another, earlier migration brought R1b to the Tarim basin.

Not related to the quote but what do you think of this as an alternative explanation?

I genuinely think, through deduction, there's enough information to be very confident in the L51 Iberian Beaker hypothesis. Where the hell did the L51 come from basically! It can't have come from the Balkans as part of the Yamnaya expansion, which was so clearly Z2103 given the Z2103:L51 ratio of the Balkans, which discounts that as the source of Eastern Beaker Steppe. It can't have come from the Steppe through Europe via the Northern European plain with the Corded Ware culture, and it can't have come before it either (otherwise we would have Steppe DNA much earlier in Europe). It is also extremely unlikely it powered all the way to East-Central Europe from the Steppe through Corded Ware, firstly because the Steppe was dominated by Z2103 in the South and R1a in the North with no evidence of any L51 at all, but also because if it had done that it would have left virtually no trace, as we have none to follow - this is very unlikely. So it couldn't have come to Western Europe from the East.

If we imagine a mixture event, where the Steppe ancestry comes from blending with the Corded Ware folk, THEN we have a good picture. Here, the only plausible picture is that the R1b L51 comes from an Eastwards Beaker migration.

This, to me, is foolproof. There's lots of other small hints, such as with phenotypes (looking at Baskid+Corded=North Atlantid), but there are a few potential holes.

First is language, which can easily be resolved by saying that the Beaker folk who imposed themselves above CW did so as a minority (like we know the Beaker folk always were), and so simply adopted the language of the CWC. When there isn't large scale population replacement, this tends to be the case - for example in Britain the incoming group did replace the language, but not so much in Iberia, where the population was far larger. The second issue is the spread of Bronze metallurgy clearly being from East to West - this can only truly be explained by the Beaker folk learning it somewhere, and we know that the Beaker folk came into contact with the Balkan arm of Yamnaya, where being metallurgists themselves they probably picked up the techniques.

 

[APO276]

I agree with you that it is wider theme and we can discuss.

But it is not my opinion, you can see which languages are Satem and Centum in literature, for example in book of author Colin Renfrew, "Archaeology and Language: Puzzle of Indo-European Origins", page 107:

The satem/centum subdivision

Western group (centum)

Germanic
Venetic
Illyrian
Celtic
Italic
Greek

Eastern group (satem)
Baltic
Slavic
Albanian
Thracian
Phrygian
Armenian
Iranian
Indian

Look they reconstruct a root ḱm̥tóm.

I will start with my language which has a /k/
If that /m̥/ had shifted to /e/, in Modern Greek the word would have been /ece'to/ (Standard), /etɕe'to/ (Cretan), /etʃe'ton/ (Cypriot).
We have /k/ because in Greek, /m̥/ supposedly shifted to a schwa sound first and then to an /a/ (or /o/ in Arcado-Cypriot)

But the linguists think that there was no palatalization of this type in 'Ancient Greek'. Note also, though:
Italian cento is pronounced /ˈtʃɛnto/
French cent is prounced /sɑ̃/, (liaison)/sɑ̃t‿/
Spanish cient is pronounced /θjẽn/, /sjẽn/
Portugeuse cento is pronounced /sẽtu/

If we assume, that there was palatalization of some short next to front vowels, /k/ > /c/ or /kʲ/ (and further to [ts] or sounds in different dialects) in Late PIE, the difference between 'centum' and 'satem' languages wouldn't appear so significant.

The proto-Indo-Iranian word is reconstructed as *ĉatám. The proto-Indo-Aryan as *śatám.
But the short [a] sound that originated from /m̥/ (which is said that originally gave a 'schwa' /ə/ sound in Greek (later /a/ or /o/), ę in Old Church Slavonic, im̃ in Lithuanian), is thought to have been a near-open central vowel.

If it descended from a near-open front vowel /æ/, maybe the palatalization can be explained as palatalization next to front vowels (which apparently is a phenomenon common all over Europe).

Maybe I haven't considered some things, but I really believe the 'centum' [kentum] / 'satem' thing isn't really important.

 

[LATGAL]

This 1976 analysis is too old and wrong
Read this from latest haplogroup report

He surveys plenty of older views on the topic too before presenting his own and little has changed since then debate-wise on those little attested languages in linguistics. He can also hardly be "wrong" per se with his analysis of how Illyrian anthroponymy was distributed, unless newer epigraphy updates it, not DNA!

Regarding DNA, as I've said elsewhere, ancient findings are still very slim for firm conclusions in certain areas but linguists class Messapian as a close relative to Illyrian based on the few findings and with perhaps some other ties to Greek (it's not impossible that all three branches, along with some others had a more recent common ancestor within IE). And ancient DNA findings might never really shine much light on linguistic differentiation at the dialectal level necessarily which your question was about.

 

[Garrick]

Look they reconstruct a root ḱm̥tóm.

I will start with my language which has a /k/
If that /m̥/ had shifted to /e/, in Modern Greek the word would have been /ece'to/ (Standard), /etɕe'to/ (Cretan), /etʃe'ton/ (Cypriot).
We have /k/ because in Greek, /m̥/ supposedly shifted to a schwa sound first and then to an /a/ (or /o/ in Arcado-Cypriot)

But the linguists think that there was no palatalization of this type in 'Ancient Greek'. Note also, though:
Italian cento is pronounced /ˈtʃɛnto/
French cent is prounced /sɑ̃/, (liaison)/sɑ̃t‿/
Spanish cient is pronounced /θjẽn/, /sjẽn/
Portugeuse cento is pronounced /sẽtu/

If we assume, that there was palatalization of some short next to front vowels, /k/ > /c/ or /kʲ/ (and further to [ts] or sounds in different dialects) in Late PIE, the difference between 'centum' and 'satem' languages wouldn't appear so significant.

The proto-Indo-Iranian word is reconstructed as *ĉatám. The proto-Indo-Aryan as *śatám.
But the short [a] sound that originated from /m̥/ (which is said that originally gave a 'schwa' /ə/ sound in Greek (later /a/ or /o/), ę in Old Church Slavonic, im̃ in Lithuanian), is thought to have been a near-open central vowel.

If it descended from a near-open front vowel /æ/, maybe the palatalization can be explained as palatalization next to front vowels (which apparently is a phenomenon common all over Europe).

Maybe I haven't considered some things, but I really believe the 'centum' [kentum] / 'satem' thing isn't really important.

Reputation and it is very good for new thread to discuss.
...

Linguists say satem/centum subdivision is useful.

My intention was not to enter in discussion about satem/centum but to say that Albanian satem is not result newer contacts between South Slavs and Albanians, it is a much deeper and older story which links Balto-Slavic, Thracian, Albanian, and Iranian (Armenian too).

Linguist Mayer even thought that Pre-Slavic was a variant of pre-Albanian, and North Iranian influenced Slavic and South Iranian influenced Albanian.

He even highlighted that Slavic was Balticised Albanian. He claimed that Carpathians were ancestral home for Pre-Slavs and Pre-Albanians.

I don't say that Mayer is right but talking about this that someone understands much deeper links between Baltic Slavic and Albanian than newer contacts. Thracian is key for understanding this.

I will not further expand this discussion in this moment and quote other linguists and contemoprary studies with mathematical-computational methods.

Useful picture for geografic relations about previous discussion (keep in mind Albanian is one of variant of Thracian).

 

[Sile]

First of all, only some Iapygians are Messapians, as the latter was only one tribe among the group that formed the former.

All the other tribes also spoke messapic and all are noted into one group named Iapygians
Messapian (/mɛˈsæpiən, mə-, -ˈseɪ-/; also known as Messapic) is an extinct Indo-European language of southeastern Italy, once spoken in the region of Apulia. It was spoken by the three Iapygian tribes of the region: the Messapians, the Peucetians and the Daunians. The language has been preserved in about 300 inscriptions dating from the 6th to the 1st century BC.

Second, the study you mentioned has no connection to the Thracians. It shows that preLGM mtDNA haplogroups were replaced to a large extent by postLGM ones, mainly eastern farmers ones, and some from the Eurasian Steppe, maybe with some Armenian too. If you want a source of migration the paper mentions in different parts Central Europe, Anatolia, Armenia, Near East, Balkans, Greek Ionians, EEH, and many many more. You just selected the one that includes the Balkan reference, but most of the time the author talks about Central Europe, Near East and Armenia. It does not look like the mtDNA is the result of one migration.

east Balkans since 3200BC to roman times is always noted as thracian
west balkans are from south to north greeks, epirotes, then albania which had corinthians from 700BC then macedonians and last the romans from 196BC and to the north of this was illyricum starting at modern montenegro.

Third, it has already been shown that the expansion of the eastern farmers was relatively balanced sex-wise while the indo-european expansion was very heavily male-dominated. You cannot distinguish between indo-european migrations based on mtDNA alone.

which ?..........cetina culture in Dalmatia which originated in the Dunubian area of ancient thracia or you speaking of vudecol culture which headed into pannonia ( hungaria and beyond)

 

[Sile]

Wrong, wrong!. You get too many things wrong! The most numerous inherited words from Illyrian are not personal names, but toponims. All toponims are still in use in Albanian language and have meanings.
Had Albanians been Thracians they would have today been genetically close to Bulgarian insisted of Toscans and South Italians. Also Albanians had not come from Illyrians should not have CELTIC dna which they have,. So genetics and linguistic combined puts Albanians firmly where they are, in Illyrian lands.
What amazes me is that you use very little to no reasoning in making your statements!

as I said, do not match albanian with messapic because it is not illyrian ............which do you want
.
do you know what if means?

 

[Sile]

He surveys plenty of older views on the topic too before presenting his own and little has changed since then debate-wise on those little attested languages in linguistics. He can also hardly be "wrong" per se with his analysis of how Illyrian anthroponymy was distributed, unless newer epigraphy updates it, not DNA!
Regarding DNA, as I've said elsewhere, ancient findings are still very slim for firm conclusions in certain areas but linguists class Messapian as a close relative to Illyrian based on the few findings and with perhaps some other ties to Greek (it's not impossible that all three branches, along with some others had a more recent common ancestor within IE). And ancient DNA findings might never really shine much light on linguistic differentiation at the dialectal level necessarily which your question was about.

analysis of the samples states that the iapygians are an east-balkan people, so they cannot be illyrian .
.
They have no Illyrian language to match messapic with ............so messapic was assumed to be illyrian from linguistic scholars in the past since it could not be from italy and had to be a neighbour ......it was clearly an old assumption.
Currently there still is no script for illyrian.
.
maybe it was what the ancient greeks scholars stated that Iapygians are pelagsian and messapic is pelagsian

 

[LATGAL]

Useful picture for geografic relations about previous discussion (keep in mind Albanian is one of variant of Thracian).

Vladimir Orel also thought that within IE Albanian overall shared most lexical isoglosses with Balto-Slavic and secondarily with Greek (which he thought might imply cohabitation of the ancestral branches in some part of the Balkans before reaching their respective destinations), though this isn't necessarily the best criterion, especially since Albanian has been attested so late. Similarly he thought that Albanian was to be connected to Daco-Moesian in the area of Dacia Ripensis rather than Illyrian, which has been the main area of debate about Albanian's ancient associations since at least Weigand. Matzinger was brought up here the other day and he similarly thinks that Albanian is to be situated roughly in the Diocese of Dacia and a theory deriving them from the Eastern Balkans isn't particularly strong, though he also doesn't think that Albanian fits in with the Illyrian evidence we have.

You are right that this isn't the result of recent contacts with Slavic but, as far as I know, Albanian itself doesn't neatly fit into satem (same with RUKI where I think the more generalized s > sh doesn't make it fully clear how it operates). My general impression is that it might have separated from the remaining IE continuum after Greek but before Balto-Slavic and Indo-Iranian and shares a similar, "intermediate" position as Armenian within that area of the late PIE continuum. I'm not sure if we can argue its ultimate location based on that kind of argument though. Armenian moved a greater distance compared to Greek which stayed in the Balkans despite likely splitting later, for example. Albanian might have moved all the way to the Adriatic already in ancient times in that case.

But all these arguments have been going on for a very good while now. The solution will come in the form of Y-DNA, e.g. the eastern Balkans potentially showing a rather different profile compared to the western or if that doesn't work due to great overall similarity, haplotype sharing and the like before linguists could hope to agree. Something like Maciamo's OP tree seems good but within Balkanic, the specific position of Albanian is hard to pinpoint and, to reiterate your point, its partial satemization doesn't seem to be the result of later contacts (but considering that satem seems to be just a late areal feature, the argument is likely mostly about whether a language separated from the continuum earlier or later, per above; I'm also confused about what he's specifically referring to with the "partial satemization...of Greek" that he mentioned).

analysis of the samples states that the iapygians are an east-balkan people, so they cannot be illyrian

What they meant by "eastern Balkan region" was likely with reference to Italy, i.e. east of Italy. After all, it references Anatolia, the steppe and the Caucasus in respect to mtDNA.

"Phylogenetic analysis of 15 Iron Age mtDNAs together with 231 mtDNAs spanning European prehistory suggest that southern Italian Iapygians share close genetic affinities to Neolithic populations from eastern Europe and the Near East"

There's nothing that argues specifically about them being (linguistically) related to the historical eastern Balkan peoples, just that their matrilines likely came from a mix of steppe, Anatolia (and maybe extra Caucasus? not unlike the Mycenaean or Thracian samples we have, for example, which reflect the influence of the pre-IE Balkan substratum).

It's as Ownstyler said.

 

[Rizla]

There are many Tarim mummies from different periods. The ones that were tested date from the Scythian period and were R1a. Modern Uyghurs have both R1b-L23 and R1a-Z93, and yet the Scythians hardly had any R1b (at least not as much proportionally as the Uyghurs), so it's likely that another, earlier migration brought R1b to the Tarim basin.

Yes, of cause. It makes sense with the R1b since many linguists class tocharian as centum. I never tought to look at the modern uyguhrs Y-DNA today.

When I said tarim mummies I meant the oldest and most famous mummies, like beauty of Loulan, cherchen man, witches of Subeshi and the Xiaohe cemetery.

The Xiaohe cemetery is from the begining of the bronze age, and 11 out of 12 males tested were found to be R1a1a. They obviously predate the iron age scythians with some 1000 years. I thought I remembered cherchen man to be of the same haplogroup, but I can't find it know. Both he and the beauty of Loulan are just as old as Xiaohe cemetery. so likely of the same or related people. Witches of Subeshi are only from 2-300 BCE though, if I recall correctly.

Here's the study on the Xiaohe cemetery
https://bmcbiol.biomedcentral.com/articles/10.1186/1741-7007-8-15

To my knowledge, no mummies with R1b have ever been found. But on the other hand, the R1b in the uygurs today has to come from somewere. When and from were, do you think these R1b people would have come to the Tarim basin?

I just read this article by Mallory. I can highly recommend it if you haven't read it. Every solution to the tocharian problem seems to have something against it. It's kind of like the anatolian problem, there just seems to be some pieces missing from the puzzle.
http://sino-platonic.org/complete/spp259_tocharian_origins.pdf

it would be interesting to know what branch of R1a :
CWC R1a-M417, or Sintashta R1a-Z93 or the older Lake Bajkal & Combed Ware R1a-YP1272 ?

I don't know. R1a1a seems to be as deep as they went. I read somewere that the earliest mummies were very close to Andronovo according to Allentoft 2015, "population of bronze age Eurasia", but I just skimmed that paper again and didn't see any mention of it. Might have to read it more closely again if it's really in there.

 

[Ygorcs]

Look they reconstruct a root ḱm̥tóm.

I will start with my language which has a /k/
If that /m̥/ had shifted to /e/, in Modern Greek the word would have been /ece'to/ (Standard), /etɕe'to/ (Cretan), /etʃe'ton/ (Cypriot).
We have /k/ because in Greek, /m̥/ supposedly shifted to a schwa sound first and then to an /a/ (or /o/ in Arcado-Cypriot)

But the linguists think that there was no palatalization of this type in 'Ancient Greek'. Note also, though:
Italian cento is pronounced /ˈtʃɛnto/
French cent is prounced /sɑ̃/, (liaison)/sɑ̃t‿/
Spanish cient is pronounced /θjẽn/, /sjẽn/
Portugeuse cento is pronounced /sẽtu/

If we assume, that there was palatalization of some short next to front vowels, /k/ > /c/ or /kʲ/ (and further to [ts] or sounds in different dialects) in Late PIE, the difference between 'centum' and 'satem' languages wouldn't appear so significant.

The proto-Indo-Iranian word is reconstructed as *ĉatám. The proto-Indo-Aryan as *śatám.
But the short [a] sound that originated from /m̥/ (which is said that originally gave a 'schwa' /ə/ sound in Greek (later /a/ or /o/), ę in Old Church Slavonic, im̃ in Lithuanian), is thought to have been a near-open central vowel.

If it descended from a near-open front vowel /æ/, maybe the palatalization can be explained as palatalization next to front vowels (which apparently is a phenomenon common all over Europe).

Maybe I haven't considered some things, but I really believe the 'centum' [kentum] / 'satem' thing isn't really important.

The centum vs. satem thing definitely existed and mattered, but as you say it probably was overrated by some linguists of the past, in that they believed that it necessarily meant a common and closer descent of the centum dialects in relation to the satem dialects, but they hadn't considered the possibility that the satem change happened as an areal feature spreading gradually, not necessarily at the same time frame, in several Late PIE dialects, some of which may have been until then more closely connected to other dialects that somehow were spared of this phonetic trend.

In my opinion the centum vs. satem division may perhaps tell us more about the geographic distribution of the Late IE dialects/early IE proto-languages and how they were connected (more or less) with each other, than it can tell us about the chronological tree and clades of IE branches. If dialects A and B split from a common source, but only dialect A was in close contact with the dialect W, which underwent a series of regular satem changes, it's possible that you'd eventually have A and W as satem languages, but A and B remained as more closely related than A and W.

However, we should remember that the satem innovation was pretty regular and happened in virtually the same way across several IE subfamilies, so it probably involved a particular change that happened in some time and place and spread from there, and not a series of totally unrelated phonetic developments in individual IE branches without any influence onto each other (for example, it wasn't like the much later Romance palatalization and eventual fricativization, which affected only Latin [k] and [g] before [e] and ; besides, the centum vs. satem division IIRC involved a generalized merger of non-palatalized and palatalized [k] and [g], for example that distinction was lost in the centum IE variants, so its outcome was generally very distinct from other processes of palatalization or loss of palatalization that happened independently in some of the daughter/grand-daughter languages later).

 

[Ygorcs]

too many things speak for pre Steppe imo. And there is absolutely no archeological or genetic trail for a Balkan solution. Anatolian lacks linguistic features typical for Steppe Indo Europeans. It has no Finno_Ugric borrowings.

Finno-Ugric specifically or did you mean more generally Uralic? I mean, Proto-Uralic as a whole is considered to have been more or less contemporaneous with Proto-Indo-European, and Finno-Ugric is already a daughter branch of PU, certainly too late to have lent many words to a still undifferentiated Steppe IE.

 

[Ygorcs]

This 1976 analysis is too old and wrong
Read this from latest haplogroup report
.
Iapypians are messapians........they are now know as East-Balkan people ....thracians............so clearly thracians are not illyrians and neither are messapians
.
https://macsphere.mcmaster.ca/handle/11375/22810
.
https://indo-european.eu/tag/iapygian/
.
Iron Age southern Italians likely descended from early to late Neolithic farmers from Anatolia and possibly as far East as the Caucasus, and from migrants arriving from eastern Europe around the late Neolithic/early Bronze Age. These findings support previous hypotheses that the ancestors of the Iapygians may have originated in the eastern Balkan region, or derive shared ancestry with a common source population from eastern Europe.
.
The Greek mythographers, as usual, derived the name from a hero, Iapyx, whom they represented as a son of Lycaon, a descent probably intended to indicate the Pelasgic origin of the Iapygians. (Anton. Liberal. 21; Plin. Nat. 3.11. s. 16.) For a further account of the national affinities of the different tribes in this part of Italy, as well as for a description of its physical geography, see the articles APULIA and CALABRIA.
.
so if albanians think they match messapian language , then they are not illyrian, but most likely thracian, Albania could have had thracians before others arrives ...............I mentioned this to Salento about 2 months ago.
.
To finalise, we have no illyrian script or language ...all we have is illyrian personnel names from Noricum ( east-austria)

Well, you can't really use a source that talks specifically about population movements of the Neolithic and Early Bronze Age into Italy to support a claim about late Iron Age languages that were spoken in the Balkans during the early Roman era. There's a huge time gap between one event and the other. In my opinion there are some sensible indications about Albanian being closely related to Thracian (actually, I think the most plausible hypothesis is the one that links it specifically to some form of Dacian or Moesian whatever was the original branch those languages came from, since the Romanian substrate looks surprisingly close to Albanian vocabulary), but certainly not more evidences than those linking it to Illyrian, and in fact it is probable that what we're seeing, in all its confusion, is just an evidence that Illyrian and Thracian were no uniform language, but more like two distinct branches of several similar languages, and that they in fact had a much closer mutual relationship (either through immediate common descent or through a centuries-long Sprachbund) than is sometimes assumed.

 

[Ygorcs]

east Balkans since 3200BC to roman times is always noted as thracian

What's your source? How can anyone possibly know, with the evidences we have as of now, if the east Balkans were "always" Thracian since 3200 BC and had no changes until the Roman era, especialy when in fact in 3200 BC Late PIE was probably still even an undivided language, and it would've been pretty much impossible that such a thing as a distinctive "Thracian" language and ethnicity already existed?

 

[APO276]

However, we should remember that the satem innovation was pretty regular and happened in virtually the same way across several IE subfamilies, so it probably involved a particular change that happened in some time and place and spread from there, and not a series of totally unrelated phonetic developments in individual IE branches without any influence onto each other (for example, it wasn't like the much later Romance palatalization and eventual fricativization, which affected only Latin [k] and [g] before [e] and ; besides, the centum vs. satem division IIRC involved a generalized merger of non-palatalized and palatalized [k] and [g], for example that distinction was lost in the centum IE variants, so its outcome was generally very distinct from other processes of palatalization or loss of palatalization that happened independently in some of the daughter/grand-daughter languages later).

Well, I believe the palatalization before e, i, j existed in all late PIE dialects, at least.
People assume that in Attic Greek, for example, K was always pronounced as /k/.
But what if it had an allophone /c/ or /kʲ/ in complementary distribution before front vowels? Most likely native speakers wouldn't even be able to notice the difference.

In general, the reconstructed pronunciation of Greek and Latin are basically based on circular reasoning. On the other hand it is huge work to challenge it.

I like Kortlandt's work, by the way. I am not sure about his motivations and the methodology he follows but I still like it.

I like the following ideas (I'm leaving palatals and 'labiovelars' aside):
1) That reconstructed PIE 'voiceless stops' *p, *t, *k correspond actually to p:, t:, k:
2) That reconstructed PIE 'voiced aspirated stops' *bh, *dh, *gh correspond actually to p, t, k

And the following idea is interesting (I don't agree exactly)
"It is probable that the whole inventory of PIE stops and laryngeal consonants can be derived from the five Indo-Uralic stops * p , *t , * c , * k , * q with palatalization, labialization and uvularization under the influence of contiguous vowels"

Therefore, I would be in favor of a reconstructed PIE without 'palatals' and 'labiovelars'. Basically I would consider the possibility that 'labiovelars' actually corresponded to some other type of stops or are just the product of a velar next to a /w/

The second series of stops, the 'plain voiced stops' (*b), *d, *g are seen as ejectives or glottalized stops according to the Glottalic theory or pre-glottalized voiceless stops according to Beekes' version of the theory. But the first one likely didn't exist at all.

I am not sure about that, though. (I am not sure about anything to be frank. I would question the validity of the views of neogrammarians, the concept of 'sound laws' etc. Historical linguistics isn't an exact science)

[Off topic: I think philosophically I'm closer to Carneades really. I consider many things impossible to know. I find fascinating the way the word 'know' is used by people like Beekes, for example. Or how the word 'evidence' is used.]

 

[Ygorcs]

Well, I believe the palatalization before e, i, j existed in all late PIE dialects, at least.
People assume that in Attic Greek, for example, K was always pronounced as /k/.
But what if it had an allophone /c/ or /kʲ/ in complementary distribution before front vowels? Most likely native speakers wouldn't even be able to notice the difference.

It COULD have been this way, but I'd say that if it wasn't perceived as a distinctive phoneme by the native speakers themselves then it wasn't phonologically relevant, it was just an allophone under certain conditions. In the case of PIE I doubt that at least the first generations after this palatalization wouldn't have noticed the change, because PIE probably already had some distinction between two forms of stops (whether it was really through palatalization or not, because as you say there is the Kortlandt hypothesis, but nevertheless it still assumes that a certain distinction in stops, which eventually was generalized and resolved into just one common sound, existed originally).

What happened in centum vs. satem is just that there was a generalization and merger after the native speakers, as you say, ceased to perceive that differentiation, probably first considering them as allophones of the same phoneme and later merging them altogether into the same consonant: "centum" dialects assumed that [k] and [k']/[kj] were the same and merged them into one [k]; "satem" dialects generalized the palatalizing/fricativizing trend to most positions and thus also started to merge [k] and [k']/[kj] into just one phonologically distinctive consonant. I personally think that pretty mainstream explanation makes sense.

The specific ways that those trends and eventual mergers happened, with the simplification of the sound inventory in one direction or another, may explain the slight differences between IE families, and it is possible that some dialects, for some reason (later displacement, loss of contacts?), were influenced by only a small part of that process that spread in continuous waves of areal features, and that explains the so called "partial satemization" of some IE branches like Albanian.

 

[Ygorcs]

Therefore, I would be in favor of a reconstructed PIE without 'palatals' and 'labiovelars'. Basically I would consider the possibility that 'labiovelars' actually corresponded to some other type of stops or are just the product of a velar next to a /w/

Why? I mean, I always thought that the existence of palatalized consonants in PIE, with their eventual generalization to other originally non-palatalized but similar consonants (k vs. kj; g vs. gj) in a late "satem" continuum of dialects, was a very neat explanation to the eventual development of "s", "sh" and "ch" consonants where we would expect a stop/velar sound if we take into account the cognates in the "centum" languages. Similar processes of sound change happened not only in IE languages later, but also in several other language families, and it in fact looks like one of the most "natural" developments during a language's phonetic evolution, thus avoiding the need to find much more complex explanations for what may have been triggered by a simple change. A split between a generalized palatalization vs. a generalized loss of palatalization explains and fits at the same time all the many correspondences betweeen [k] in centum branches and [sh] [ch] in satem branches. It's also such a recurring change in languages that it could explain why this "satem" areal feature spread so widely and easily.

 

[APO276]

Why? I mean, I always thought that the existence of palatalized consonants in PIE, with their eventual generalization to other originally non-palatalized but similar consonants (k vs. kj; g vs. gj) in a late "satem" continuum of dialects, was a very neat explanation to the eventual development of "s", "sh" and "ch" consonants where we would expect a stop/velar sound if we take into account the cognates in the "centum" languages. Similar processes of sound change happened not only in IE languages later, but also in several other language families, and it in fact looks like one of the most "natural" developments during a language's phonetic evolution, thus avoiding the need to find much more complex explanations for what may have been triggered by a simple change. A split between a generalized palatalization vs. a generalized loss of palatalization explains and fits at the same time all the many correspondences betweeen [k] in centum branches and [sh] [ch] in satem branches. It's also such a recurring change in languages that it could explain why this "satem" areal feature spread so widely and easily.

If there were 'palatovelars' in PIE the generalized loss of palatalization should have happened independently in Hittite and in centum languages?

I'm thinking we can reconstruct a velar and something that triggers the palatalization. One of the triggers would be the vowel that follows.

Now, I understand at least some of the problems with what I say.

Leiden will publish at some point a 'new Pokorny'. When they do it, I will try to test some of the things I have thought.

For now, it suffices to say that I consider likely that the stop system of Hittite represents something closer to the original state of affairs.

If we take into account the Nostratic hypothesis, we see that traditional PIE *bʰ, *dʰ, *gʰ / *ǵʰ correspond to Hittite p, t, k and Uralic p, t, k (with fricatization between vowels)

(and sorry to say it, you see how closer phonologically Modern Greek -with its voiceless fricatives- is to Uralic than to traditional PIE which has voiced murmured stops that don't exist anywhere in Europe and don't even exist in all Indo-Iranian languages)

 

[Sile]

What's your source? How can anyone possibly know, with the evidences we have as of now, if the east Balkans were "always" Thracian since 3200 BC and had no changes until the Roman era, especialy when in fact in 3200 BC Late PIE was probably still even an undivided language, and it would've been pretty much impossible that such a thing as a distinctive "Thracian" language and ethnicity already existed?

I read it from
An Early Bronze Age pile-dwelling settlement of discovered in Alepu lagoon (municipality of Sozopol, department of Burgas), Bulgaria
and other web sites
.
If you want to call them Pelagsian, Greeks, Indians or black sea coastal villagers or whatever , so be it.
.
A new pile-dwelling settlement has been discovered during coring investigations on the shores of the Alepu lagoon (municipality of Sozopol, department of Burgas), on the western Black Sea coast, in Bulgaria. A multi-disciplinary methodology was applied to analyze the archaeological dataset, composed of wood piles, abundant charcoals and wood fragments, seeds, fish and shell remains, a few small bone fragments, some lithic fragments and potsherds. The piles were trimmed from oak trees and sunk into lagoonal muds, and currently lie 5.8 to 6.8 m below mean sea level. It highlights a wooden building at the edge of Alepu palaeo-lagoon. Charcoal remains confirm the use of oak tree as a dominant timber resource, consistent with pollen data for this period. Palaeo-botanic remains highlight gathering activities and the consumption of wild grapes, raspberries and figs. The herbaceous assemblage evokes deforestation activities. Exploitation of coastal resources is well attested by the great density of fish remains, dominated by anchovy (61%), highlighting possible preservation of fish products. Five radiocarbon dates constrain the age of the site to between 3350 and 3000 cal. BC.
The Alepu piles-dwelling settlement sheds new light on the very beginning of the Early Bronze Age in coastal Bulgaria.
.
The area as noted by these studies are called Thracia Pontica
.
https://journals.openedition.org/mediterranee/docannexe/image/8203/img-1.jpg
where is is situated

 

[Sile]

Well, you can't really use a source that talks specifically about population movements of the Neolithic and Early Bronze Age into Italy to support a claim about late Iron Age languages that were spoken in the Balkans during the early Roman era. There's a huge time gap between one event and the other. In my opinion there are some sensible indications about Albanian being closely related to Thracian (actually, I think the most plausible hypothesis is the one that links it specifically to some form of Dacian or Moesian whatever was the original branch those languages came from, since the Romanian substrate looks surprisingly close to Albanian vocabulary), but certainly not more evidences than those linking it to Illyrian, and in fact it is probable that what we're seeing, in all its confusion, is just an evidence that Illyrian and Thracian were no uniform language, but more like two distinct branches of several similar languages, and that they in fact had a much closer mutual relationship (either through immediate common descent or through a centuries-long Sprachbund) than is sometimes assumed.

link me a illyrian script and a thracian script