Sunday, May 31, 2020

Early Bronze Age Ancient DNA From Anatolia and Adjacent Regions

Overview and Analysis

A new paper on ancient Anatolian genetics is not open access so I have to rely on the blog of Bernard, a French anthropologist, and the abstract, for the details. But, this is enough to see that what is really notable.

What is most striking is the apparent absence of steppe ancestry in any of the 89 new samples, in an overall region in which speakers of the Hittite language and another representative of Anatolian branch of the Indo-European language family were attested historically from around 1800 BCE in central Anatolia, and eventually more widely in the region. 



But, this is somewhat less notable than it seems, because all of the samples reviewed in the paper from 2000 BCE to 1000 BCE come from either Alalakh (26) or Ebla (12), both in the Northern Levant, and all of the Ebla samples, as well as several of the Alalakh samples, come from before 1800 BCE. 




To my knowledge, there is no actually Anatolian, let alone definitively Hittite DNA from the 2000 BCE to 1000 BCE time period from any source. 



If my hypothesis is correct, people with ancestry derived from the Hittite founding population, including in particular, most nobles and elites, would show significant steppe ancestry in that time period.



There is one woman, from the twenty-six individuals from Alalakh whose DNA was sequenced, who is an outlier with ancestry characterized as Central Asian/Iranian and dated to about 1540 BCE (a few decades after the city of Alalakh was sacked by the Hittites). She may have arrived there when the city was under Mittani rule or influence, as it was at the about that time. The Mittani kingdom was a historically attested Northern Iranian kingdom, whose territory overlapped with the geographic region of this woman's likely genetic origins (notable, in part, because it appears that its elite spoke Sanskrit) that grew to rule Northern Syria. Eurogenes discusses at greater length that possibility that this is a Mittani woman and one other individual in the study. In regard to this paper Davidski also notes that:

Note that one of the Bronze Age females from Alalakh, labeled ALA019, appears to have ancestry from Turan and the Eurasian steppe. She may well have been a Mitanni of Indo-Aryan origin. 
Interestingly, a Copper Age [a.k.a. Eneolithic a.k.a. Chalcolithic era] male from Arslantepe [in eastern central Anatolia a.k.a. Melid], ART038, belongs to Y-haplogroup R1b1a2 aka R1b-V1636. This is an unusual find, because R1b hasn't yet been reported in any Copper Age or earlier samples from outside of Europe and the Eurasian steppe. 
As far as I can tell, this individual doesn't harbor any genome-wide ancestry from north of the Caucasus. However, R1b-V1636 is a rare lineage that is first attested in Eneolithic samples from the North Caucasus Piedmont steppe, so ART038's Y-chromosome might be the first evidence of the presence of steppe ancestry in Copper Age Anatolia.
All of the samples in this paper from Arslantepe are from between 4000 BCE and 3000 BCE. Wikipedia describes the history of this location as follows:


Late Chalcolithic 
Earliest habitation at the site dates back to the Chalcolithic period. 
Aslantepe (VII) became important in this region in the Late Chalcolithic. A monumental area with a huge mudbrick building stood on top of a mound. The building had a large building with wall decorations and its function is uncertain. 
Early Bronze 
By the late Uruk period development had grown to include a large temple/palace complex. 
Culturally, Melid was part of the "Northern regions of Greater Mesopotamia" functioning as a trade colony along the Euphrates River bringing raw materials to Sumer (Lower Mesopotamia). 
Numerous similarities have been found between these early layers at Arslantepe, and the somewhat later site of Birecik (Birecik Dam Cemetery), also in Turkey, to the southwest of Melid. 
Around 3000 BCE, the transitonal EBI-EBII, there was widespread burning and destruction, after which Kura-Araxes culture pottery appeared in the area. This was a mainly pastoralist culture connected with the Caucasus mountains. 
Late Bronze Age 
In the Late Bronze Age, the site became an administrative center of a larger region in the kingdom of Isuwa. The city was heavily fortified, probably due to the Hittite threat from the west. It was culturally influenced by the Hurrians, the Mitanni and the Hittites. 
Around 1350 BC, Suppiluliuma I of the Hittites conquered Melid in his war against Tushratta of Mitanni. At the time Melid was a regional capital of the Land of Isuwa at the frontier between the Hittites and the Mitanni loyal to Tushratta. Suppiluliuma I used Melid as a base for his military campaign to sack the Mitanni capital Wassukanni. 
Iron Age 
After the end of the Hittite empire, from the 12th to 7th century BC, the city became the center of an independent Luwian Neo-Hittite state of Kammanu. A palace was built and monumental stone sculptures of lions and the ruler erected. 
The encounter with the Assyrian king Tiglath-Pileser I (1115-1077 BC) resulted in the kingdom of Melid being forced to pay tribute to Assyria. Melid remained able to prosper until the Assyrian king Sargon II (722-705 BC) sacked the city in 712 BC. At the same time the Cimmerians and Scythians invaded Anatolia and the city declined. 
As explained in more depth below, despite their proximity of Anatolia to these sites in the Northern Levant, neither of the sites in this study that are the source of all of the samples in the time period from 2000 BCE to 1200 BCE were colonized or occupied by the Hittite people for any extended length of time and were never places where Indo-European Anatolian languages were spoken.

Two Explanations

There are basically two ways to resolve this conundrum. 

I believe that the most plausible of the two is that Indo-Europeans arrived in Anatolia mostly after 2000 BCE, and that the apparent time depth of the Anatolian languages relative to other Indo-European language families is due to the very different substrate influences the Indo-Europeans encountered and to a fairly elite driven language shift process. 

But, I will first present the other possibility, before justifying my preferred hypothesis in greater depth.

The Middle Neolithic Anatolian Languages Split Scenario

One narrative hypothesizes that the Anatolian languages have a much greater time depth of differentiation from proto-Indo-European relative to other branches of the Indo-European language family, because the Anatolian languages are so strongly differentiated from the other Indo-European languages. 

But, this study does fairly definitively establish the absence of steppe ancestry in the time period from 4000 BCE to 2000 BCE which would have been the early phase of Anatolian language development in this scenario. 

So, in this narrative, Indo-European languages might originate in Anatolia, and spread to the nearby Pontic-Caspian steppe via an elite with little demic impact (for there is stunningly little Anatolian or Caucasian ancestry among the people of the steppe from ca. 4000 BCE to 1200 BCE), where it made a secondary expansion that was dominated by people with steppe ancestry. 

There is really no archaeological indication, however, of an elite led cultural revolution on the steppe dramatic enough to lead to language shift from a South Caucasian or Anatolian source at about the right time period. And, there is no attestation of anything strongly suggestive of an Indo-European culture (not even in proper names) from neighboring Egyptian, Semitic and Sumerian civilizations, that were literate from 3500 BCE on, about 1650 years before the earliest attested evidence of the Hittites or other linguistically Anatolian peoples in the writings of these peoples, many of whom were engaged in wide ranging trade that extended into Anatolia.

Eurogenes quotes some of the analysis from a 2017 paper in support of this hypothesis and then goes on the criticize it.
Apparently, Mathieson et al. 2017 aren't comfortable with putting the PIE homeland on the Pontic-Caspian Steppe because they can't find any evidence in their ancient DNA dataset of a significant migration through the Balkans that would potentially bring Anatolian languages from the Pontic-Caspian steppe to Anatolia. From the paper:

"One version of the Steppe Hypothesis of Indo-European language origins suggests that Proto-Indo European languages developed in the steppe north of the Black and Caspian seas, and that the earliest known diverging branch – Anatolian – was spread into Asia Minor by movements of steppe peoples through the Balkan peninsula during the Copper Age around 4000 BCE, as part of the same incursions from the steppe that coincided with the decline of the tell settlements. If this were correct, then one way to detect evidence of it would be the appearance of large amounts of characteristic steppe ancestry first in the Balkan Peninsula, and then in Anatolia. However, our genetic data do not support this scenario. While we find steppe ancestry in Balkan Copper Age and Bronze Age individuals, this ancestry is sporadic across individuals in the Copper Age, and at low levels in the Bronze Age. Moreover, while Bronze Age Anatolian individuals have CHG/Iran Neolithic related ancestry, they have neither the EHG ancestry characteristic of all steppe populations sampled to date [20] , nor the WHG ancestry that is ubiquitous in southeastern Europe in the Neolithic (Figure 1A, Supplementary Data Table 2, Supplementary Information section 1). This pattern is consistent with that seen in northwestern Anatolia [11] and later in Copper Age Anatolia [23], suggesting continuing migration into Anatolia from the East rather than from Europe." 
And this... 
"On the other hand, our data could still be consistent with the Steppe-Balkans-Anatolia route hypothesis model, albeit with constraints. It remains possible that populations dating to around 1600 BCE in the regions where the Indo-European Luwian, Hittite and Palaic languages were spoken did have European hunter-gatherer ancestry. However, our results would require that such ancestry was not ubiquitous in Bronze Age Anatolia, and was perhaps tightly linked to Indo-European speaking groups. We predict that additional insight about the genetic origins of the potential speakers of early Indo-European languages will be obtained when ancient DNA data become available from additional sites in this key period in Anatolia and the Caucasus."
I am agnostic about the route by which the early Anatolian speakers made their way from a possible steppe origin to Anatolia, although an Eastern route along the Black Sea coast or over the Caucasus mountains is at least as plausible as a Balkan route.
The Bronze Age Anatolian Language Origins Scenario

The other hypothesizes that Indo-Europeans with steppe ancestry first arrived in Anatolia around 2000 BCE to 1750 BCE when they were first attested in writing by neighboring literate cultures (as newly arrived peoples found in just a few places), around the same time as Indo-European expansion into the Balkans, the Aegean, Iran, and South Asia. 

In this second scenario, steppe ancestry was absent from Anatolia in earlier time periods (before 2000 BCE) because no Indo-Europeans had arrived there yet. 

Steppe ancestry is not seen in Anatolia proper in the Bronze Age after 2000 BCE, because we only have one sample in the area and time period, and that may not be ethnically a Hittite or one of some other Anatolian population. 

In this narrative, the linguistic distinctiveness of the Anatolian languages could also be explained not by time depth, but (1) by stronger and qualitatively different substrate impacts and contact effects in a place where the substrate language may have been very different from the former LBK culture and Harappan culture's languages (substrate influences that are identical across a language family are hard to distinguish from features of a proto-language of the superstrate language itself), and (2) because the Anatolian language speakers who may have been merely a ruling elite, leading to lots of substrate influence from language learners adapting to the language of their new elites gave rise to more substrate influence, relative to other places where a much larger percentage of the population was replaced by expanding Indo-Europeans.

Why would the substrate be different in Anatolia from that of the Anatolian derived early European farmers (EEF) who expanded into Europe in the European Neolithic revolution?

Because, the major change in the Anatolian gene pool documented in this new paper which probably occurred ca. 6500 BCE (at or shortly after the time the first early European farmers left), in which South Caucasian/Iranian farmers admixed with Western Anatolian farmers who were ancestral to both of the main branches of the first farmers of Europe (the LBK by land, and the Cardial Pottery culture along the Mediterranean coast or at least further South, possibly by sea). Indeed, mass migration from the East could have been a push factor driving Western Anatolian farmers to migrate into Europe.

The language family, although perhaps not actual language, shared by these two main European Neolithic cultures of the early European farmers, would have provided a very similar substrate language for all of the Indo-Europeans who subsequently expanded into Europe in the Bronze Age. 

But, the admixture seen across Anatolia was probably accompanied by a language shift to a Caucasian/Iranian farmer language from a different language family entirely from that of the language family of the Early European Farmers. 

This would give the Anatolian languages a very different substrate than the Indo-European languages encountered in Europe. And, it also appears that while the Neolithic and Eneolithic civilizations of Europe and South Asia had almost completely collapsed by the time that the Indo-Europeans arrived, Anatolia was less hard hit, and thus, its existing civilization, while ultimately conquered by the Indo-Europeans, were able to put up more of a fight and avoid massive population replacement to the extent seen, for example, in Britain and Ireland. A disconnect linguistically and ethnically between a ruling elite and ordinary people in a contemporaneous kingdom nearby was seen among the Mittani whose elite may have been Sanskrit speaking Indo-Aryans, but whose ordinary people where non-Indo-Europeans who spoke the non-Indo-European Hurrian language.




Image via the Wikipedia page for the Mittani

I have long favored this latter scenario, which is not inconsistent with the data in the new paper.

Alalakh and Ebla Were Linguistically Semitic In The Time Periods For Which We Have Samples And Were Not Occupied By Significant Numbers of Anatolian Language Speakers For Extended Periods Of Time

Ebla was never ruled by the Hittites, who destroyed the city about three hundred years after the last sample in the city in this study was taken, so no steppe ancestry would be expected there. As Wikipedia explains:
Starting as a small settlement in the Early Bronze Age (c. 3500 BC), Ebla developed into a trading empire and later into an expansionist power that imposed its hegemony over much of northern and eastern Syria. Ebla was destroyed during the 23rd century BC; it was then rebuilt and was mentioned in the records of the Third Dynasty of Ur. The second Ebla was a continuation of the first, ruled by a new royal dynasty. It was destroyed at the end of the 3rd millennium BC, which paved the way for the Amorite tribes to settle in the city, forming the third Ebla. The third kingdom also flourished as a trade center; it became a subject and an ally of Yamhad (modern-day Aleppo) until its final destruction by the Hittite king Mursili I in c. 1600 BC. 
Ebla maintained its prosperity through a vast trading network. Artifacts from Sumer, Cyprus, Egypt and as far as Afghanistan were recovered from the city's palaces. The kingdom had its own language, Eblaite, and the political organization of Ebla had features different from the Sumerian model. Women enjoyed a special status, and the queen had major influence in the state and religious affairs. The pantheon of gods was mainly north Semitic and included deities exclusive to Ebla. The city was excavated starting in 1964 and became famous for the Ebla tablets, an archive of about 20,000 cuneiform tablets found there, dated to around 2350 BC. Written in both Sumerian and Eblaite and using the cuneiform, the archive has allowed a better understanding of the Sumerian language and provided important information over the political organization and social customs of the mid-3rd millennium BC's Levant.
We also have extensive contemporaneous written accounts from Alalakh, well preserved in cuneiform that allow us to discern the former city-state's history and linguistic situation in a manner that wouldn't be reliable from other archaeological evidence. 

This city was founded by Semitic Amorite people until the Hittites destroyed it, ca. 1587–1560 BC. Less than a century later, people writing in Semitic cuneiform resided there again in a rebuilt city, which, while it may have owed allegiance to the Hittite empire from the mid-1300s BCE,  as spoils involved in the defeat of the Mittani rulers of the city (whose elites may have spoken Sanskrit), it does not appear to have ever been colonized by the Hittites and the city was abandoned within half a century of be acquired by the Hittite empire.




Alalakh was founded by the Amorites (in the territory of present-day Turkey) during the early Middle Bronze Age in the late 3nd millennium BC. The first palace was built c. 2000 BC, contemporary with the Third Dynasty of Ur
Middle Bronze II 
The written history of the site may begin under the name Alakhtum, with tablets from Mari in the 18th century BC, when the city was part of the kingdom of Yamhad (modern Aleppo). A dossier of tablets records that King Sumu-Epuh sold the territory of Alakhtum to his son-in-law Zimri-Lim, king of Mari, retaining for himself overlordship. After the fall of Mari in 1765 BC, Alalakh seems to have come under the rule of Yamhad again. King Abba-El I of Aleppo bestowed it upon his brother Yarim-Lim, to replace the city of Irridu. Abba-El had destroyed the latter after it revolted against his brother Yarim-Lim. A dynasty of Yarim-Lim's descendants was founded, under the hegemony of Aleppo, that lasted to the 16th century [BCE]. According to the short chronology found at Mari, at that time Alalakh was destroyed, most likely by Hittite king Hattusili I, in the second year of his campaigns. 
Late Bronze 
After a hiatus of less than a century, written records for Alalakh resume. At this time, it was again the seat of a local dynasty. Most of the information about the founding of this dynasty comes from a statue inscribed with what seems to be an autobiography of the dynasty's founding king. 
According to his inscription, in the 15th century BC, Idrimi, son of the king of Yamhad, may have fled his city for Emar, traveled to Alalakh, gained control of the city, and been recognized as a vassal by Barattarna. The inscription records Idrimi's vicissitudes: after his family had been forced to flee to Emar, he left them and joined the "Hapiru people" in "Ammija in the land of Canaan." The Hapiru recognized him as the "son of their overlord" and "gathered around him"; after living among them for seven years, he led his Habiru warriors in a successful attack by sea on Alalakh, where he became king. 
However, according to the archeological site report, this statue was discovered in a level of occupation dating several centuries after the time that Idrimi lived. There has been much scholarly debate as to its historicity. Archeologically-dated tablets recount that Idrimi's son Niqmepuh was contemporaneous with the Mitanni king Saushtatar. This seems to support the inscription on the statue claiming that Idrimi was contemporaneous with Barattarna, Saushtatar's predecessor. 
The socio-economic history of Alalakh during the reign of Idrimi's son and grandson, Niqmepuh and Ilim-ilimma, is well documented by tablets excavated from the site. Idrimi is referred to rarely in these tablets. 
In the mid-14th century BC, the Hittite Suppiluliuma I defeated king Tushratta of Mitanni and assumed control of northern Syria, then including Alalakh, which he incorporated into the Hittite Empire. A tablet records his grant of much of Mukish's land (that is, Alalakh's) to Ugarit, after the king of Ugarit alerted the Hittite king to a revolt by the kingdoms of Mukish, Nuhassa, and Niye. The majority of the city was abandoned by 1300 BC.  
Alalakh was probably destroyed by the Sea People in the 12th century BC, as were many other cities of coastal Anatolia and the Levant. The site was never reoccupied, the port of Al Mina taking its place during the Iron Age.

The Wikipedia entry for Hattsusili I (the Hittite King who destroyed Alalakh before it was rebuilt) is also informative:
Hattusili I (Ḫattušili I) was a king of the Hittite Old Kingdom. He reigned ca. 1586–1556 BC (short chronology). 
He used the title of Labarna at the beginning of his reign. It is uncertain whether he is the second king so identified, making him Labarna II, or whether he is identical to Labarna I, who is treated as his predecessor in Hittite chronologies. 
During his reign, he moved the capital from Neša (Kaneš, near modern Kültepe) to Ḫattuša (near modern Boğazkale), taking the throne name of Ḫattušili to mark the occasion. 
He is the earliest Hittite ruler for whom contemporary records have been found. In addition to "King of Ḫattuša", he took the title "Man of Kuššara", a reference to the prehistoric capital and home of the Hittites, before they had occupied Neša. 
A cuneiform tablet . . . written in both the Hittite and the Akkadian language provides details of six years of his reign. In it, he claims to have extended the Hittite domain to the sea, and in the second year, to have subdued Alalakh and other cities in Syria. In the third year, he campaigned against Arzawa in western Anatolia, then returned to Syria to spend the next three years retaking his former conquests from the Hurrians, who had occupied them in his absence.
Wikipedia tells us this about the Mittani Kingdom:
Currently there are two hypotheses regarding how Mitanni was formed: 1) Mitanni was already a powerful kingdom at the end of the 17th century or in the first half of the 16th century BC, and its beginnings are from before the time of Thutmose I, so dated to the time of the Hittite sovereigns Hattusili I and Mursili I. 2) Mitanni came to be due to a political vacuum in Syria, which had been created first through the destruction of the kingdom of Yamhad by the Hittites and then through the inability of Hatti to maintain control of the region during the period following the death of Mursili I. In this case Mitanni (c. 1500 to 1300 BC) could have come to be a regional power after the Hittite destruction of Amorite Babylon and a series of ineffectual Assyrian kings created a power vacuum in Mesopotamia. 
While the Mitanni kings were supposedly Indo-Aryan, they used the language of the local people, which was at that time a non-Indo-European language, Hurrian. Their sphere of influence is shown in Hurrian place names, personal names and the spread through Syria and the Levant of a distinct pottery type.
The Paper

The paper and is abstract and a couple of its illustrations are as follows: 
Highlights

• Genome-wide analysis of 110 ancient individuals from the Near East
 
• Gene pools of Anatolia and Caucasus were biologically connected ∼6500 BCE 
• Gene flow from neighboring populations in Northern Levant during 3rd millennium BCE 
• One individual of likely Central Asian origin in 2nd millennium BCE Northern Levant 
Summary
Here, we report genome-wide data analyses from 110 ancient Near Eastern individuals spanning the Late Neolithic to Late Bronze Age, a period characterized by intense interregional interactions for the Near East. 


Fig. 1A  showing the new samples examined. The differently colored outlier sample from ca. 1500 BCE is also from Alalakh. 
We find that 6th millennium BCE populations of North/Central Anatolia and the Southern Caucasus shared mixed ancestry on a genetic cline that formed during the Neolithic between Western Anatolia and regions in today’s Southern Caucasus/Zagros. 
The "Graphical Abstract" from the paper below.
During the Late Chalcolithic and/or the Early Bronze Age, more than half of the Northern Levantine gene pool was replaced, while in the rest of Anatolia and the Southern Caucasus, we document genetic continuity with only transient gene flow. 
Additionally, we reveal a genetically distinct individual within the Late Bronze Age Northern Levant. 
Overall, our study uncovers multiple scales of population dynamics through time, from extensive admixture during the Neolithic period to long-distance mobility within the globalized societies of the Late Bronze Age.
Keywords 

human population history: ancient DNA: Near East: Eastern Mediterranean: genome-wide data: admixture: genetic continuity: archaeogenetics: Ubaid: Uruk: Kura-Araxes
Eirini Skourtanioti, et al., Genomic History of Neolithic to Bronze Age Anatolia, Northern Levant, and Southern Caucasus 181(5) Cell 1158-1175.e28 (May 28, 2020). Hat tip to Bernard's Blog.

Bernard reports (in Google translate with minor modifications of my own from the French original) that:
Early genetic studies have shown that the first farmers in Anatolia, the southern Levant and northwest Iran are the descendants of local hunter-gatherers. Thus the transition from the Mesolithic to the Neolithic took place in the region without significant movement of populations.  
However two millennia later the situation changes. In fact, the Chalcolithic and early Bronze Age populations show less genetic differentiation suggesting that this latter transition was accompanied by large population movements. . . . Archaeological evidence in the region shows many connections between different regions of the Near East at different times. Thus at the end of the Neolithic the north of Mesopotamia (cultures of Halaf and Samarra) is connected with the east of Anatolia. During the second half of the 4th millennium BCE the Kura-Araxes culture originating from the south of the Caucasus, extends west and east. Are these different connections the result of movement of populations or only of ideas? . . . 
The authors eliminated 16 samples due to the poor quality of the genome obtained and 5 samples closely linked in the first or second degree with other samples. The remaining 89 genomes have been added to around 800 previously published genomes, including 17 Anatolians from the archaeological sites of Tepecik-Ciftlik, Barcın, Gondürle-Höyük, Topakhöyük and Kaman-Kalehöyük and dated from the same period. . . . 
[T]he individuals of the end of the Neolithic or the beginning of the Chalcolithic . . . are located on a gradient which extends along the second component between the individuals of the West Anatolian Neolithic (Barcin) . . . and ancient individuals from Iran and the Caucasus . . . . However, the samples from the early Chalcolithic from the Tell Kurdu site in the northern Levant are slightly deviated towards the individuals from the southern Levant. . . .
The qpAdm software makes it possible to model the individuals of Büyükkaya and the South Caucasus as coming from a genetic mixture between a West Anatolian population (respectively 76% and 69%) and a population of northern Iran (respectively 24% and 31%), and individuals from the North Levant of Tell Kurdu as coming from a genetic mixture between a West Anatolian population (47.9%), the Levant (36.6%) and a population from northern Iran (15.5%)[.] . . . 
Unlike the individuals of the previous period, the individuals of the end of the Chalcolithic and the beginning of the Bronze Age all regroup in the middle of the previous gradient. These results are confirmed by the f3 and f4 statistics. There has therefore was significant genetic admixture in the region which homogenized the genetic profile of the populations located between western Anatolia and the southern Caucasus. The authors estimated the date of genetic admixture at around 105 generations or 3000 years before the time of these individuals, corresponding to a date of genetic mixing around 6500 BCE. This genetic mix varies between 21 and 38% of Iranian ancestry and the rest of West Anatolian ancestry. 
The individuals of the North Levant genetically most resemble the other Anatolian populations. However, the qpAdm software and the f4 statistics suggest that these individuals still differ from other populations due to their greater genetic affinity with the populations of the southern Levant.

Among the individuals of 
Alalakh, there is one that differs from all the others. This individual is a woman who was discovered at the bottom of a well, dated around 1540 BCE, and her skeleton shows traces of old wounds. She is located on the PCA near individuals from Central Asia and eastern Iran.
As an aside, this paper is the first published journal article that I have seen with both a "video abstract" and a video accompanying it in the supplemental materials.

No comments: