Subartu: The Forgotten Northern Deep of Armenian History

Antiquities

A name from the world’s oldest archives

Long before Urartu, long before “Armenia” appears in any inscription, Sumerian and Akkadian scribes were already writing about a northern land they called Subartu (also rendered Šubartum, Subir, or Šubur). It shows up in cuneiform sources as early as the third millennium BCE, and the earliest clear attestation of the term dates to roughly 2400 BCE (EBSCO Research Starters). Geographically, Subartu was never a single fixed kingdom on a map — its borders shift from text to text — but most Assyriologists place its core in Upper Mesopotamia, along the upper Tigris and the Khabur and Balikh river basins, spilling into the southern edge of the Armenian Highland (Wikipedia: Subartu).

For the Akkadian Empire, Subartu functioned less like a country and more like a compass direction: exactly as Amurru meant “west,” Elam meant “east,” and Sumer meant “south,” Subartu simply meant “north” — the mountainous, unruly upland beyond the plains (Wikipedia: Subartu). Even a Sumerian mythological text, Enmerkar and the Lord of Aratta, lists Subartu among the lands of “confused languages,” alongside Sumer and Akkad — a mark of how foreign and distant it seemed to the scribes of the south (Wikipedia: Subartu).

Who were the people of Subartu?

The population associated with this region is sometimes called the Subarians (or Subareans), described by later scholarship as a people who inhabited the southwestern Armenian Highlands and northern Mesopotamia (Wikipedia: Subarians). Their language — “Subarian” or “Subarean” — is now generally treated by specialists as an older Sumero-Akkadian label for what we know linguistically as Hurrian (Encyclopedia.com: Hurrian). Early twentieth-century Assyriologists such as Adolf Leo Oppenheim and Arthur Ungnad already argued that “Subarian” was simply another name for Hurrian (Wikipedia: Subarians), and the World History Encyclopedia likewise notes that Hurrian speech and culture appear in ancient sources interchangeably as “Subartu” or “Subareans” — the same population the Bible later calls Horites (World History Encyclopedia: Hurrians).

Textual evidence of Hurrian names and places appears from around 2200 BCE, under the Akkadian king Naram-Sin, while archaeological indicators of Hurrian culture may reach back even further, into the mid-fourth millennium BCE (EBSCO Research Starters; World History Encyclopedia: Hurrians). The Hurrian heartland stretched from the upper Khabur and Tigris basins to the Taurus and Zagros ranges, with a strong presence around Lake Van (World History Encyclopedia: Hurrians) — precisely the lake that later sits at the center of the Urartian kingdom.

Why this matters for Armenian history

The chain of association runs like this: Subartu → Hurrians → Mitanni and the Nairi lands → Urartu → Armenia. It was from this Subartu-Hurrian base, according to standard historical reconstruction, that the powerful kingdom of Mitanni later emerged, and from which the warlike tribes of the Nairi lands (a precursor political landscape to Urartu) are thought to descend. Some scholars have even proposed that “Urartu” itself represents a later reconsolidation of the same Hurrian-speaking population base around Lake Van, though this remains a contested point rather than a settled consensus (Wikipedia: Hurrians). Linguistically, Hurrian is not isolated either: together with Urartian, it forms the Hurro-Urartian language family, whose deeper external relationships (to Northeast Caucasian, Kartvelian, or even Indo-European languages) remain actively disputed among comparative linguists (Wikipedia: Hurrian language).

In other words, Subartu is not a footnote — it is one of the deepest layers beneath the later civilizations of Hayasa, Urartu, and Greater Armenia, describing the same northern highland world under its oldest recorded name.

Martiros Kavoukjian and the “Subartu = Armenia” hypothesis

The idea that Subartu should be identified directly with the ancestors of the Armenians is most closely associated with Martiros Kavoukjian (1908–1988), an Armenian architect, self-taught Armenologist, and historian born in Niğde in the Ottoman Empire. After building much of his career as chief architect in Soviet Armenia’s postwar reconstruction, he turned to historical research and published his best-known work, Armenia, Subartu and Sumer: The Indo-European Homeland and Ancient Mesopotamia, in Armenian and English in 1987 (Wikipedia: Martiros Kavoukjian). The full text of the book’s early chapters is digitized and freely readable on the Internet Archive (Archive.org: Armenia, Subartu and Sumer).

Kavoukjian’s central claims, built up across chapters on Armani, Aratta, Erech, and the peoples of the Armenian Highland in antiquity, were threefold:

  • Linguistic continuity — he searched for direct etymological links between the name “Subartu” and later Armenian ethnonyms and place names.
  • Autochthony — he argued the Subarians were not merely neighbors of the later Urartians and Armenians but their direct biological and cultural forebears, native to the highland rather than migrants into it.
  • Cultural continuity — he held that Subarian mythology and craft traditions passed organically into Armenian tradition.

It’s worth being precise about where Kavoukjian sits in the wider field. His book is catalogued alongside the broader Armenian hypothesis of Indo-European origins associated with linguists Tamaz Gamkrelidze and Vyacheslav Ivanov, which argues for a Proto-Indo-European homeland on the Armenian Highland rather than in the European steppe (Wikipedia: Armenian hypothesis; Kiddle: Armenian hypothesis). Reader reviews of his book note both its appeal — a dense, source-driven argument for highland autochthony — and its controversial reception: one review remarks that despite his “in-depth analysis of linguistic elements,” mainstream scholarship largely rejected his conclusions (AllBookStores: Armenia, Subartu, and Sumer). Kavoukjian himself was blunt about this friction, writing in his preface that the European-homeland hypothesis had become so entrenched that “any proof or evidence that contradicts it is either rejected or ignored” (AllBookStores: Armenia, Subartu, and Sumer). His method leaned heavily on comparative etymology — for example, tracing Sumerian and Akkadian terms into Armenian — a technique that professional Assyriologists and Indo-Europeanists have generally treated with skepticism, since such comparisons are easy to construct and hard to falsify.

What ancient DNA and population genetics actually show

This is where the picture becomes genuinely interesting, because in the decade after Kavoukjian’s death, genome-wide population genetics gave scholars a tool he never had. The landmark study here is Haber, Mezzavilla, Xue, Comas, Gasparini, Zalloua, and Tyler-Smith, “Genetic evidence for an origin of the Armenians from Bronze Age mixing of multiple populations,” published in the European Journal of Human Genetics in 2016 (posted as a preprint in 2015) (Nature/EJHG; PubMed; full text via PMC).

Analyzing genome-wide data from 173 Armenians against 78 other world populations, the team found that:

  • Armenians form a distinct genetic cluster bridging the Near East, Europe, and the Caucasus (EJHG).
  • Their ancestry can be modeled as a mixture of several Eurasian source populations that combined roughly between 3000 and 2000 BCE (EJHG).
  • Crucially, the genetic signal of population mixing stops abruptly around 1200 BCE, precisely when Late Bronze Age civilizations across the Eastern Mediterranean collapsed (bioRxiv preprint).
  • After that Bronze Age collapse, Armenians appear to have remained a largely closed, endogamous population for roughly three thousand years, with internal genetic substructure only emerging much later — around five centuries ago — when the Ottoman–Safavid partition of historic Armenia physically separated eastern and western Armenian communities.
  • The study also found that about 29% of Armenian ancestry traces to a population best represented by Neolithic Europeans, giving Armenians a notably higher affinity to early European farmers than most other present-day Near Eastern populations (PubMed).

A follow-up physical-anthropology study by Movsesian and colleagues (2020, American Journal of Physical Anthropology) examined cranial traits from Bronze and Iron Age skeletal populations of the Armenian Highland specifically to test genetic continuity into the modern population, building directly on the Haber et al. genomic findings (Wiley Online Library).

More recent ancient-DNA work on the Kura-Araxes culture (also called the Early Transcaucasian Culture, c. 3625–2250 BCE), which spread across Armenia, Georgia, Azerbaijan, and into eastern Anatolia and northwestern Iran, has started to sequence genomes directly from Kura-Araxes burial sites such as Kaps, Karnut, Berkaber, and the well-known Shengavit settlement on the Yerevan plain (DNAGENICS: Kura-Araxes of Early Bronze Armenia). These analyses show a genuinely mixed ancestry profile — components related to Chalcolithic Armenia, the Chalcolithic Zagros, and steppe-adjacent groups from the North Caucasus — rather than a single uniform “steppe” or “local” origin (Eurogenes Blog; Eurogenes Blog: Armenian Highland population prehistory). Researchers studying these samples are careful to add a caution that is worth repeating here: with only a dozen or so genomes sequenced so far, the data offer a first glimpse, not a final verdict, on how much of the biological ancestry of modern Armenians actually descends in a direct line from Kura-Araxes populations as opposed to shared regional heritage (DNAGENICS).

Reading Kavoukjian against the genetic record

So where does this leave Kavoukjian’s Subartu hypothesis? The honest answer is: partially, and unevenly, vindicated.

What the genetics does support: the core intuition that the ancestors of modern Armenians were not late arrivals who displaced an unrelated highland population, but rather the product of population mixing that took place on the Armenian Highland itself, in the Bronze Age, followed by roughly three thousand years of genetic isolation. This is a strong, independently derived confirmation of highland continuity and autochthony — the very point Kavoukjian spent his career arguing on linguistic grounds decades before the genomic tools existed to test it.

What the genetics does not, and cannot, confirm: any specific identification of “Subartu” as an ethnic or genetic ancestor-state named in Sumerian texts. Population genetics can show that a region’s gene pool stabilized by a certain date; it cannot verify that a particular Bronze Age textual label — Subartu, Subir, Subarian — corresponds to a genetically coherent, self-identifying group, let alone one directly ancestral to Armenians in the way Kavoukjian’s etymologies assumed. The Hurrian/Subarian linguistic connection to Armenian remains, at best, an open question in comparative linguistics, distinct from the population-genetic question of highland continuity (Wikipedia: Hurrian language).

In short: modern DNA evidence has strongly reinforced the big idea beneath Kavoukjian’s work — that Armenians are a genuinely indigenous highland population rather than late migrants — while leaving his specific philological arguments about Subartu, Sumerian loanwords, and direct Hurrian descent exactly where they were: a bold, still-debated hypothesis rather than an established fact.

Sources

Artatsolum

You might also like:

Before the Sistrum: Tracing a Bronze Age Rattle from Anatolia to Ararat

Before the Sistrum: Tracing a Bronze Age Rattle from Anatolia to Ararat

The Goat on the Armenian Highland: A Symbol Older Than Writing

The Goat on the Armenian Highland: A Symbol Older Than Writing

"Summon Against Her the Kingdoms of Ararat": Urartu-Armenia in the Prophecy Against Babylon

“Summon Against Her the Kingdoms of Ararat”: Urartu-Armenia in the Prophecy Against Babylon

Leave a Reply

Your email address will not be published. Required fields are marked *