Germanic lineage arrived from the east?

[Andour]

@Jaska

Do linguists have any (rough) idea when the process of satemization occurred in the languages concerned ?

Most (all?) satem-language cultures seem to be either R1a dominant or deriving from the arrival of R1a groups. CWC was essentially an R1a culture. If we could posit that CWC was already a satem culture or in the process of palatalizing its own version of PIE, then that would strike off one option for the emergence of Germanic.

I1 being an unlikely source, it would leave U106 as the most probable vector.

I am aware that we are dealing with time depths that leave us little chance to know for sure. But asking doesn't hurt...

[Vinitharya]

Satemization most likely started in the Indo-Iranian community, as it is realized most fully in that family. For instance, the word for stone in Persian and in Sanskrit is asman, while in Lithuanian it is akmuo, and kámen in Czech. The Balto-Slavic languages retain the k, showing incomplete satemization, while the Indo-Iranian ones complete the satemization. Balto-Slavic and Indo-Iranian had a land boundary, while Germanic was separated by the Baltic, so it was unable to participate in satemization. Satemization traces in Daco-Thracian and Armenian could also be a result of Indo-Iranian neighbors.

[Awood]

Could this mean that Germanic was originally a R1a and I1 language, and that U106 was picked up later?  With a focus in the Netherlands, U106 could be a Nordwestblock language marker.

This is what I thought initially from the extract, but the Bell Beaker group is separate from north Germanic/ south Scandinavian group. The R-U106, R-L11 (CTS4528/S1194?) group in the aDNA record seems to be associated with north Germany / Denmark / S. Sweden.

The paper also suggests the R1a-Z284 reached Norway very early, with Battle Axe and "steppe" ancestry I suppose, but is not directly related to the late expansion of the East Scandinavian ancestry.  

What I find interesting is that many of the R1a graves in Norway are from the same massacre grave which might be skewing the numbers a little bit. Nonetheless it may have been a majority R1a if not exclusively so in the Battle Axe period.

[Jaska]

@Jaska

Do linguists have any (rough) idea when the process of satemization occurred in the languages concerned ?

Most (all?) satem-language cultures seem to be either R1a dominant or deriving from the arrival of R1a groups. CWC was essentially an R1a culture. If we could posit that CWC was already a satem culture or in the process of palatalizing its own version of PIE, then that would strike off one option for the emergence of Germanic.

I1 being an unlikely source, it would leave U106 as the most probable vector.

I am aware that we are dealing with time depths that leave us little chance to know for sure. But asking doesn't hurt...

When satemization is defined as containing two changes – (1) palatovelar stops to affricates or sibilants, and (2) merger of labiovelar stops to plain velar stops – it only concerns Indo-Iranian and Balto-Slavic. In Armenian and Albanian palatalized stops have also changed to sibilant-like sounds, but the second change did not occur: these branches are therefore officially neither satem- nor centum-branches, but they have at least partially preserved three distinct series of velar stops. It is possible that Armenian and Albanian change was also affected by (Indo-)Iranian or Balto-Slavic contacts, but such a change is also typologically expected and can occur independently (“easier to pronounce” -principle).

Moreover, satemization is not the earliest sound change even in Indo-Iranian or Balto-Slavic, but it seems to have spread secondarily by contacts. It is probably earlier in Indo-Iranian, because (1) Indo-Iranian sound changes seem to have begun earlier, and (2) satemization is not as exceptionless in Balto-Slavic as it is in Indo-Iranian.

As to “when”: Late Proto-Indo-Iranian is dated around 2000 BCE and located within the Sintashta Culture, so satemization occurred some centuries before that. Into Balto-Slavic it apparently spread during the early 2nd millennium BCE. In Proto-Balto-Slavic it is reconstructed as *ḱ > *ś and *ǵ > *ź, which would still require the presence of Proto-Indo-Iranian *ć and *dź and not the Proto-Iranian depalatalized *c and *dz as the original model.

[Andour]

Thanks, guys! Fascinating.

"As to “when”: Late Proto-Indo-Iranian is dated around 2000 BCE and located within the Sintashta Culture, so satemization occurred some centuries before that."

Sintashta is thought to be derived from the CWC, if I remember correctly. So "some centuries before that" could make the whole difference - depending on how many centuries we are talking about. Guess we'll never know, unfortunately - much would be solved if we did.

[Rodoorn]

Quint:

Pre-Germanic in the sense of the various stages preceding Proto-Germanic which haven't gone through all the sound changes defining Germanic and which also were in contact with Western Uralic languages.

OK, thanks.
In either way, Pre-Proto-Germanic came from the south or from the east, so I do not understand why you would require such loanwords all over Europe in only one of the options.

Quint:

I read you in a way as if you're suggesting that the NBA was linguistically on the Germanic lineage from the start. That would not exactly be what the authors of the new paper seem to imply here:
“Linguistically, the Late Bronze Age is the period during which Palaeo-Germanic donated vocabulary to Finno-Saamic in the east and adopted vocabulary from Celtic in the south, suggesting that it was spoken widely among East Scandinavians distributed between Sweden and Denmark, and possibly also in the Nordic Bronze Age communities in Finland and Estonia.” (Page 34)
They suggest that Germanic languages only arrived during the Late Bronze Age with the newly detected East Scandinavian cluster.

Actually, they do not: here they only write that around 1000 BCE (their beginning of Late Bronze Age) Paleo-Germanic was ALREADY “spoken widely” by people sharing traces of this East Scandinavian IBD cluster. They do not exclude the earlier presence of Germanic language lineage there, nor do they claim that it only spread during the Late Bronze Age. This becomes obvious when we look at the whole picture. The total chronology can be described like this:

1. In the late 3rd millennium: the arrival of Baltic Hunter-Gatherer ancestry in Southern Sweden, and following that, the formation of the East Scandinavian IBD cluster. Germanic language lineage might have arrived from the east or from the south, or it was already there since the CWC; in any case, it was already there (at least in Southwestern Sweden) during the early 2nd millennium BCE.

2. Towards 1600 BCE: admixture of the East and West Scandinavian IBD clusters in Southern Norway. Beginning of the Scandinavian Bronze Culture. Pre-Proto-Germanic was already spoken here among the “EScan cluster people”, irrespective of where it arrived from and whether it had arrived only recently or a long time ago.

3. Towards 1400 BCE: admixture of the East and South Scandinavian IBD clusters in Denmark. Population in Denmark either adopted the Germanic language only now, or they had adopted it already earlier. Spread of Scandinavian Bronze Culture in Southwestern Finland and Coastal Estonia.

4. Around 1000 BCE: Proto-Celtic influence in Peninsular Southern Scandinavia and Celtic loanwords into Paleo-Germanic. In Southwestern Finland and Estonia: Paleo-Germanic loanwords were borrowed into Early Proto-Saami and Early Proto-Finnic, just recently arrived in the respective regions.

5. Around 600 BCE: sound changes leading to Proto-Germanic. Still a persistent genetic border between Southern and Eastern Scandinavians (although the former was admixed with the latter).

6. Around 1 CE: West Germanic people spread from Southern Jutland (South Scandinavian cluster) and East Germanic (> Gothic) people spread from Sweden (East Scandinavian cluster). North Germanic (Scandinavian) remains in Northern Jutland and/or Southwestern Sweden?

7. Around 400 CE: people from Germany move to Denmark, virtually replacing the earlier population in the Danish islands. However, no new language replaces Proto-Scandinavian (see my earlier messages).

and

8. Around 400 CE West Germanic people (=South Scandinavian cluster, Danish Isles IA related) spread around the North Sea and replaced (mostly) the East North Sea cluster.

Kind of carroussel.

And about 7. Orentil already mentioned the -leben -lev impact. 

And see this:
"The Danish Isles ancestry that was widespread on Zealand from 2200 BP disappears from ~1600 BP (>towards the North Sea area!- Rodoorn). For the few samples between 1600 BP and 1230 BP we find instead a variety of ancestries, Swedish Iron Age, Celtic Iron Age (sic! -Rodoorn), Norwegian Iron Age, and Jutlandic (check) Iron Age. In Northern Jutland, this additional resolution reveals a transition within the constant proportion Bronze Age Southern Scandinavian ancestry. Prior to 1600 BP it is modelled as North Jutlandic IA ancestry, which gradually shifts to become primarily modelled as North German IA ancestry."

I guess this replacement and amalgam of ancestries must have had more impact on the language than only the -leben -lev thing! But what?

[Orentil]

I wonder if what they describe as Celtic Iron Age ancestry in the post above and in another statement as: "Our results demonstrate the arrival of a strong component of North German IA ancestry, in combination with a series of ancestries previously associated with Celtic-speaking groups and populations carrying European Farmer (in addition to GAC) ancestry from north-western Europe." could also stem from the Przeworsk culture / Vandals from the east/Poland. I'm pretty sure that Przeworsk culture will turn out to be Celto-Germanic. We know unfortunately close to nothing about the language of the Vandals, but I could imagine that it was not as close to Gothic as often assumed.
Time to get Przeworsk culture samples...

[Ebizur]

Satemization most likely started in the Indo-Iranian community, as it is realized most fully in that family.  For instance, the word for stone in Persian and in Sanskrit is asman, while in Lithuanian it is akmuo, and kámen in Czech.  The Balto-Slavic languages retain the k, showing incomplete satemization, while the Indo-Iranian ones complete the satemization.  Balto-Slavic and Indo-Iranian had a land boundary, while Germanic was separated by the Baltic, so it was unable to participate in satemization.  Satemization traces in Daco-Thracian and Armenian could also be a result of Indo-Iranian neighbors.

OT tangent: why does the Germanic form of this etymon (e.g. English "hammer") have an /r/ at the end of the word instead of an /n/? An /r/ is also seen in e.g. the name of the city Samarkand (Ancient Greek Μαράκανδα) in Central Asia. The adjectival form of the word for "stone, rock" in Sanskrit also has an /r/ instead of an /n/ (i.e. aśmara "stony, rocky; pertaining to stone" vs. aśman "stone, rock"). Is this /n/ vs. /r/ variation a systemic phenomenon?

[Anglesqueville]

^^ 

If it is Indo-European, it could be either a secondary thematicization of an old *-mer/*-men heteroclite, or the thematic suffix *-ero- added to a root ending in *-m-.

(Hyllestedt)
In his thesis Hyllestedt proposes an alternative hypothesis: it may be loan from the Balto-Fennic *hamara (> Fi. hamara ‘back of an axe’, Est. hamar, hammar ‘back of a knife’). See "Word Exchange at the Gates of Europe Five Millennia of Language Contact", Hyllested, Adam.

[old europe]

we should go back to the original meaning of hammer that is likely stone.


hammer (n.)

Old English hamor "hammer," from Proto-Germanic *hamaraz (source also of Old Saxon hamur, Middle Dutch, Dutch hamer, Old High German hamar, German Hammer). The Old Norse cognate hamarr meant "stone, crag" (it's common in English place names), and suggests an original sense of the Germanic words as "tool with a stone head," which would describe the first hammers. The Germanic words thus could be from a PIE *ka-mer-, with reversal of initial sounds, from PIE *akmen "stone, sharp stone used as a tool" (source also of Old Church Slavonic kamy, Russian kameni "stone"), from root *ak- "be sharp, rise (out) to a point, pierce."

As a part of a firearm, 1580s; as a part of a piano, 1774; as a small bone of the ear, 1610s. Figurative use of "aggressive and destructive foe" is late 14c., from similar use of French martel, Latin malleus. To go at it hammer and tongs "with great violence and vigor" (1708) is an image from blacksmithing (the tongs hold the metal and the hammer beats it). Hammer and sickle as an emblem of Soviet communism attested from 1921, symbolizing industrial and agricultural labor.

[Jaska]

The question of Germanic *hamara-z ‘hammer’ is indeed difficult to conclude, as the Indo-European etymology requires some weird changes (according to Hyllested). On the other hand, the Uralic explanation is also difficult, because the change *š > *h was one of the latest towards Late Proto-Finnic and only occurred ca. 200 CE. At that point Finnic had contacts only with Proto-Scandinavian, not any longer with Proto-Germanic. The Finnic word could be a borrowing from Scandinavian, but then Saami and Mordvin cognates would not be real cognates (but they might not be in any case, see below).

West Uralic cognates are also highly irregular. Late Proto-Finnic *hamara points to original *šamara or *še̮mara, and Mordvin *šov would go back to original *šumV or *šuvV, matching neither Finnic nor Saami regularly.

Different Saami languages point to Late Proto-Saami *še̮mmērē (South Saami), *še̮mērē (Skolt Saami), or *šimērē (North Saami etc.), and this secondary LPSa *š is not possible in old Uralic words, because the primary *š > *s in Saami (although Hyllested proposes another word with assumedly preserved *š before nasal consonant). LPSa *e̮ comes from earlier *i or *ü, and LPSa *i comes from earlier *ej.

To me, the only way to bridge the Uralic vowels by even remotely credible way would be to assume some unknown Paleo-European? *šümara, borrowed independently as such in Pre-Proto-Saami and as *šuma(ra) in Pre-Proto-Mordvin (there are some apparently parallel borrowings showing Finnic *ü ~ Mordvin *u > *o). This still would not explain the Finnic word and the irregular variation in Saami.

In any case, the Finnic *hamara probably is more likely borrowed from Germanic than the other way round. But semantically the Finnic word is closer to the Saami and Mordvin words (‘back of an axe or knife’) than to the Germanic word. Germanic meaning ‘hammer’ itself would not be that close to the IE ‘stone’ without the variety of meanings in Old Norse: hamarr ‘Stein, Fels, Hammer, Haiart’ = ‘stone, rock, hammer, ???’. So, the Germanic word could also come from Proto-Indo-European, although requiring weird consonant metathesis just like in Slavic.

One more possibility could perhaps be an Early Proto-Slavic borrowing into Pre-Proto-Germanic or Paleo-Germanic before the Grimm’s Law. Hyllested writes:
“Slavic *kamy/*kamen- is mysteriously distorted. For those who reconstruct a laryngeal in the PIE word it looks like a late metathesis: *h₂áḱ-mon- > PBSl. *Hák-men- > *káH-men. However, the vowel PGmc. *hamara- is short, which rules out the metathetic explanation that works for Slavic.”

I believe the problem of the short Germanic *a would be solved, if the word was borrowed from Early Proto-Slavic. All we need is more EPSl loanwords into PaG to prove the contacts. It would also help us to solve the persistent problem concerning the whereabouts of Pre-Proto-Germanic. (Edit: the problem here still remains that in Germanic we have -r instead of -n.)

I think there are possibly two different words:
1. Indo-European 'stone' > Slavic with consonant metathesis --> Germanic 'stone, hammer' --> Finnic *hamara
2. Unknown 'back of axe or knife' > Saami and Mordvin --> perhaps the Saami word was similar looking enough to spread its meaning to the Finnic word? (A sort of contamination of two different words.)

P.S. There is a younger Scandinavian loanword in the northern dialects of Finnish spoken in Sweden: hamari 'mechanical hammer'.
https://kaino.kotus.fi/sms/?p=qs-article...ord=hamara

[NewEnglander]

Could this mean that Germanic was originally a R1a and I1 language, and that U106 was picked up later?  With a focus in the Netherlands, U106 could be a Nordwestblock language marker.

This is what I thought initially from the extract, but the Bell Beaker group is separate from north Germanic/ south Scandinavian group. The R-U106, R-L11 (CTS4528/S1194?) group in the aDNA record seems to be associated with north Germany / Denmark / S. Sweden.

The paper also suggests the R1a-Z284 reached Norway very early, with Battle Axe and "steppe" ancestry I suppose, but is not directly related to the late expansion of the East Scandinavian ancestry.  

What I find interesting is that many of the R1a graves in Norway are from the same massacre grave which might be skewing the numbers a little bit. Nonetheless it may have been a majority R1a if not exclusively so in the Battle Axe period.

I would really like to see much more Swedish samples from this period, to see what the relative frequencies of R1a and R1b in Sweden are before and during the Bronze Age. Of course I am biased due to my own haplogroup, but it looks like I1 and R1b have relatively similar frequencies in many of the regions sampled by this paper, besides in Jutland. Naudigastir also mentioned in the other thread that most of the R1b-Z18 in Denmark by the Iron Age was associated with the East Scandinavian cluster. This makes me wonder if there was contact in Sweden, and then subsequent spread together back into Denmark and Norway. I think more testing of Sweden, and in multiple regions, is key here.

[Ebizur]

The question of Germanic *hamara-z ‘hammer’ is indeed difficult to conclude, as the Indo-European etymology requires some weird changes (according to Hyllested). On the other hand, the Uralic explanation is also difficult, because the change *š > *h was one of the latest towards Late Proto-Finnic and only occurred ca. 200 CE. At that point Finnic had contacts only with Proto-Scandinavian, not any longer with Proto-Germanic. The Finnic word could be a borrowing from Scandinavian, but then Saami and Mordvin cognates would not be real cognates (but they might not be in any case, see below).

West Uralic cognates are also highly irregular. Late Proto-Finnic *hamara points to original *šamara or *še̮mara, and Mordvin *šov would go back to original *šumV or *šuvV, matching neither Finnic nor Saami regularly.

Different Saami languages point to Late Proto-Saami *še̮mmērē (South Saami), *še̮mērē (Skolt Saami), or *šimērē (North Saami etc.), and this secondary LPSa *š is not possible in old Uralic words, because the primary *š > *s in Saami (although Hyllested proposes another word with assumedly preserved *š before nasal consonant). LPSa *e̮ comes from earlier *i or *ü, and LPSa *i comes from earlier *ej.

To me, the only way to bridge the Uralic vowels by even remotely credible way would be to assume some unknown Paleo-European? *šümara, borrowed independently as such in Pre-Proto-Saami and as *šuma(ra) in Pre-Proto-Mordvin (there are some apparently parallel borrowings showing Finnic *ü ~ Mordvin *u > *o). This still would not explain the Finnic word and the irregular variation in Saami.

Could this word not have been loaned from Para-Indo-Iranian *śmara (basically, a form identical to Sanskrit aśmara "stony, rocky; pertaining to stone or rock" minus the initial vowel) independently to both Proto-Finnic and to Proto-Samic? Or perhaps to Proto-Finno-Samic with the preservation of *ś in Samic attributable to the (unusual for Uralic?) *śm sequence in the hypothetical source form (assuming that the vowel between *š and *m in the Proto-Samic form has been subsequently added through epenthesis, likewise for the vowel between *š > *h and *m in the Proto-Finnic form)?

I must admit that the apparently shared semantic development among the Western Uralic words is troubling for this hypothesis. I have almost no knowledge of the typical phonological development of Mordvin, Mari, Udmurt, or Komi, so I cannot tell whether or not the attested Mordvin form also could be descended from a hypothetically borrowed *śmara.

[Jaska]

Ebizur:

Could this word not have been loaned from Para-Indo-Iranian *śmara (basically, a form identical to Sanskrit aśmara "stony, rocky; pertaining to stone or rock" minus the initial vowel) independently to both Proto-Finnic and to Proto-Samic? Or perhaps to Proto-Finno-Samic with the preservation of *ś in Samic attributable to the (unusual for Uralic?) *śm sequence in the hypothetical source form (assuming that the vowel between *š and *m in the Proto-Samic form has been subsequently added through epenthesis, likewise for the vowel between *š > *h and *m in the Proto-Finnic form)?

I must admit that the apparently shared semantic development among the Western Uralic words is troubling for this hypothesis. I have almost no knowledge of the typical phonological development of Mordvin, Mari, Udmurt, or Komi, so I cannot tell whether or not the attested Mordvin form also could be descended from a hypothetically borrowed *śmara.

An interesting proposal.
According to Holopainen 2019, epenthetic vowel in connection with syllabic resonants in Indo-Iranian loanwords was Pre-Proto-Mordvin *e, so the round back vowel in *šov does not fit. There are no clear examples in Finnic and Saami about such a vowel, so such a development is in any case difficult to support.

Mordvin has also preserved the quality and distinction between *s, *š, and *ć > *ś, so it allows no other initial consonant than *š. Proto-Indo-Iranian had *ć, Proto-Indic *ś and Proto-Iranian *c, none of which is regularly suitable here, although there are some irregular examples showing West Uralic *š as the reflex of Iranian *c/*dz.

Hyllested speculated with a Balto-Slavic original with *š, an even though Baltic has both ašmuõ and akmuõ (Lithuanian), it lacks the metathesis which Slavic has. But because the early Balto-Slavic diversity appears to have been much wider than the present, it is hypothetically possible that there was a dialect with a word like *šaman. But it would need more words as evidence to grow to a credible option. And again, Germanic seems to be the only branch showing -r, although Hyllested mentions Slovincian kamor. I do not know what is the explanation for it.

[Parastais]

1. Indo-European 'stone' > Slavic with consonant metathesis --> Germanic 'stone, hammer' --> Finnic *hamara

Slavic consonant metathesis is quite young phenomenon. Likely late 1 M AD.
From wiki
"The liquid metathesis occurred in the Common Slavic era. It took place after or was still productive until the end of the 8th century. The name of Charlemagne, who died in 814, underwent the change:

Old High German Karl[note 1] > PSl. *karl′u[note 2] > Common Slavic *korl′ь > Russian koról′, Polish król, Slovak kráľ, Serbo-Croatian krȃlj
On the other hand, the change had already been completed in the earliest Old Church Slavonic documents. That implies that the change was completed, at least in the dialects of Bulgaria and of Macedonia, in no later than the 9th century, when the documents were written. There are, however, some attested unmetathised words in OCS such as ал(ъ)дии, a doublet of the metathised ладии."