The Coast That Kept Its Name

In the year 137, a Roman official named Flavius Arrian sailed down the eastern shore of the Black Sea on the emperor Hadrian's business and wrote up what he saw. Among the peoples he passed were the Apsilae, whose king, Julian, had received the marks of royal office from Trajan a generation earlier. Just to the north lived the Abasgi under a king called Resmag, and beyond them the Sanigae under Spadag, both confirmed in their thrones by Hadrian himself. Pliny the Elder, who died at Vesuvius in 79, had already named the Apsilae and placed the Sanigia near the Roman base of Sebastopolis, where the modern city of Sukhum now stands.

Hold those names in your head, because the argument of this piece turns on them. The people Arrian called the Apsilae call themselves, to this day, Apsua. The exonym by which their neighbours and the wider world know them — Abkhaz — descends from his Abasgi. Place the ancient ethnonyms beside the modern ones and the correspondence is not subtle: Apsil- to Apsua, Abasg- to Abkhaz. A people recorded in the same stretch of coast, under recognisably the same name, across nearly nineteen centuries, is not a recent arrival. Yet a recurring claim holds exactly that: that the Abkhazians are "really" a Georgian (Kartvelian) tribe, and that the medieval Kingdom of Abkhazia was a Georgian kingdom flying false colours. Both claims dissolve on contact with the sources. Seeing why is worth the effort, partly because the evidence is genuinely interesting and partly because the honest answer is more complicated, and more credible, than either side's slogans.

EN — English   |   TR — Türkçe   |   RU — Русский   |   DE — Deutsch   |   KA — ქართული   |   FR — Français

The names that stayed put

Written attestation is one strand; the ground beneath it is another. Between roughly the second and seventh centuries, the valley of the Kodor and the hills around Tsebelda were home to what archaeologists call the Tsebelda, or Tsabal, culture, a dense web of settlements, hill-forts, churches and richly furnished graves. The famous burial at Taglan, near modern Gali, yielded gold deer-head ornaments, silver vessels and horse-fittings of the first and second centuries, and very probably belonged to the lineage of that same King Julian. What the spade does not find is the thing the "newcomer" theory needs: a horizon of destruction or replacement, a layer where one population stops and another begins. The line runs unbroken from antiquity into the Middle Ages. Oleg Bgazhba, summarising decades of excavation, describes a continuous local development, with these early polities sitting as a peripheral but organic part of the Roman–Byzantine world — their coastal castles, at Pitsunda, Sebastopolis and elsewhere, manned as part of the imperial frontier system, with a "First Cohort of the Abasgians" even serving in Egypt.

Then there is the language, which settles the chronology on its own. Abkhaz belongs to the Northwest Caucasian family, alongside Adyghe (Circassian) and the now-extinct Ubykh. Abkhaz and Adyghe are related, but the distance between them is enormous, the kind of gulf that opens up only over thousands of years of separate development. This single fact makes nonsense of the story, still occasionally repeated, that the Abkhazians descended from the mountains as latecomers in the seventeenth century. A language does not diverge that far in three or four hundred years. Whatever else is debated, the deep, in-place antiquity of Abkhaz speech is not.

What Agathias actually wrote

If the Abkhazians' ancestors were not a Kartvelian tribe, who were their kin? The sources answer plainly: each other. The clearest test case is a passage in the sixth-century historian Agathias, who continued the narrative of Procopius and described a revolt by the Misimians, a people of the foothills between the Kodor and the Ingur, against Byzantium in the 550s. It began when an imperial official, Sotterichus, tried to hand their fortress of Bukhlon to the Alans; his envoys were beaten, he was killed in a night raid, and a brutal war followed.

The detail that matters is small and grammatical. Agathias notes that the Misimians, though subjects of the king of the Colchians (the Laz), differed from them in language and in custom. He then groups the Misimians with the Apsilians as a kindred and neighbouring people, so much so that when the Byzantines wanted to negotiate, they sent Apsilian envoys, reasoning that the two peoples were close in speech and way of life. (The Misimians, undeterred by kinship, executed them.) Elsewhere Agathias states outright that among the Laz, the Svans and the Misimians "the language is different, as are their customs."

This is the kind of point that survives or dies on a single particle. The Greek of the crucial sentence uses the paired clitics men… de, whose whole function is to mark the two halves of a contrast. The contrast they mark is between the Misimians and the Laz, with the Apsilians appended as fellow-subjects of the same overlord. George Hewitt has shown that the Georgian and English translations of the passage blur precisely this, leaving it ambiguous whether it is the Misimians and the Apsilians, or the Misimians and the Laz, who are being distinguished, whereas the Greek admits no ambiguity at all. Read in the original, Agathias tells us that the Misimians were politically subject to the Laz but linguistically and culturally distinct from them, and akin to the Apsilians. The neighbouring Kartvelian Laz were the foreign overlords here, not the cousins. That is the opposite of the "Georgian tribes" thesis, and it comes from a contemporary who had no stake in the modern quarrel.

A kingdom called Abkhazia

Now to the kingdom itself, and the second claim, that it was a Georgian state under another label. Here the simplest evidence is the most decisive: the name. In the sources, in whatever language they happen to be written, the polity that emerged on this coast at the end of the eighth century is Abkhazia, and its rulers are kings of the Abkhazians. That is true of the Greek and Byzantine material, and it is just as true of the Georgian chronicles, the Kartlis Tskhovreba and the Letopis' Kartli. The dynasty, the Leonids, or Anosids, named for two of their kings because their family name is lost, built a realm that ran, at its height, across much of western Georgia, with its capital moved from Anakopia (today's New Athos) eastward to Kutaisi.

The word Sakartvelo, "Georgia," is younger than all of this. It appears only at the turn of the eleventh and twelfth centuries, to name the unified kingdom that had by then taken shape, and that kingdom, the sources are clear, grew out of the Abkhazian realm of the Leonids rather than swallowing it. A Georgian chronicler of the thirteenth century, looking back, called Abkhazia the Bagratids' "ancient kingdom." This is not an Abkhaz polemicist's gloss; it is a medieval Georgian writer conceding the genealogy of his own royal house.

That genealogy runs through a woman. In the second half of the tenth century the Leonid line ran out of sons: Leon III, Demetrius III and Theodosius III each died without a male heir, and Theodosius, blinded in a succession struggle, was a king in name only. The throne passed sideways and then down the female line, a recognised early-medieval custom, to Bagrat, son of the Abkhazian princess Gurandukht, Giorgi II's daughter, by her husband Gurgen, a Bagratid prince of Tao. Bagrat inherited Abkhazia through his mother around 978; through his father he later gathered the Kartvelian and Tao crowns as well, becoming the first ruler of the united kingdom, Bagrat III by the Bagratid reckoning, though the second of his name among the Abkhazian kings.

What he did with his titles is telling. Bagrat leaned hard on his Abkhazian descent, because that was the legal basis of his right to rule; the dynastic register compiled in his reign, the Divan of the Abkhazian Kings, carefully records that he was the son of the daughter of the king of the Abkhazians. His mother, styled "lady-mother," appears beside him in the building inscription of the Kutaisi cathedral as "Bagrat, king of the Abkhaz and the Kartlians, and his mother Gurandukht." And on his coinage the legend reads, in Georgian, "Christ, glorify Bagrat, king of the Abkhazians." Even after the Kartvelian title was added, "king of the Abkhazians" stayed first, and remained first for his successors, so much so that in shorthand they were simply "king of the Abkhazians," and foreign writers went on calling the whole sprawling Bagratid monarchy "Abkhazia" for centuries. A state does not put a conquered subject-people at the head of its royal style and keep it there for generations.

The Greek question

A careful reader will have spotted a difficulty, and it is the strongest objection to everything above. If early Abkhazia worshipped in Greek and answered to the Patriarch of Constantinople, doesn't that point away from any distinct identity, and certainly away from a Georgian one? The honest reply is that the Greek phase, properly understood, cuts the other way.

Start with what happened. When Justinian's Byzantium drew the Abasgi into Christianity around the 540s, it did so through an Abasg named Euphrates, sent from the imperial court to persuade his own people; a bishop arrived from Constantinople, and an imperial-funded church of the Mother of God was built. The liturgy, naturally, was Greek. Georgian as a church and chancery language came later, the scholarly consensus, accepted by Georgian historians such as Simon Janashia, is that it displaced Greek in the west only from around the turn of the eighth and ninth centuries onward, and the epigraphy bears this out. The earliest firmly dated Georgian inscription in Abkhazia is at Lykhny, and it was carved in 1066, it records the appearance of Halley's Comet. Before that, the writing on the ground here was Greek.

A liturgical language is not an ethnic passport. Latin was the language of worship, law and learning across the whole of Western Europe without turning Germans, Franks or Englishmen into Romans; Arabic script ran across Central Asia without making Persians or Turks into Arabs. In Abkhazia the sequence is Greek first, then Georgian, and the switch coincides not with any change in who the people were but with the kingdom's growing political pull toward the east, as its centre of gravity shifted to Kutaisi and its church was eventually detached from Constantinople and joined to the Catholicosate at Mtskheta. The Greek phase, far from embarrassing the Abkhaz case, is its best chronological anchor: it shows an early polity facing Byzantium across the sea, not Iberia across the mountains. The later Georgian phase is real, and it is part of the story, but it is a story about politics and prestige following the realm eastward, not about a population being discovered to have been Georgian all along.

"For history is the raw material for nationalist or ethnic or fundamentalist ideologies, as poppies are the raw material for heroin addiction. The past is an essential element, perhaps the essential element, in these ideologies. If there is no suitable past, it can always be invented.” 

— Eric Hobsbawm, The Invention of Tradition, Cambridge University Press (2012) (co-edited with Terence Ranger).

How the Georgian-tribe idea was built

The thesis that the medieval Abkhazians were a Kartvelian tribe is not an old tradition slowly eroded by modern sceptics. It is itself a modern construction, and a datable one. It was advanced in 1954 by the literary historian Pavle Ingorokva, in a book ostensibly about the tenth-century writer Giorgi Merchule, where he declared that the Abkhazians and their neighbours, Apsilians, Misimians, Sanigians, had been "purely Georgian tribes" speaking a Kartvelian dialect all along, and that today's Abkhazians were different people entirely.

It is worth saying clearly that the most damaging objections to this came from Georgian scholars. The linguist and Abkhazologist Ketevan Lomtatidze rejected it; so did the historian Nikoloz Berdzenishvili and the critic Giorgi Soselia; and it was taken apart in detail by the Abkhaz historian Zurab Anchabadze. Their case rested on the sources Ingorokva himself invoked. He claimed that classical Abkhazia had been inhabited by the Moschi, a Kartvelian people, and that "Abasgi" was simply "Moschi" worn down by time. But Strabo, his star witness, places the Moschi nowhere near the coast: their land lies beyond the Phasis (Rioni) and the Cyrus (Kura), divided among Colchians, Iberians and Armenians (Geography XI.2.18). Tacitus has the Moschi raiding the borders of Armenia, hardly the work of a people based in Abkhazia. And the supposed phonetic evolution from "Moschi" to "Abasgi" cannot have happened, because the two names appear in the record at the same time: Tacitus and the grammarian Aelius Herodian use "Moschi," while Arrian and Herodian alike describe the "Abasgi" as a settled people of the coast in the same era. A name cannot turn into another name that already exists alongside it.

+ Questions of Abkhazian history in the book by P. Ingorokva ‘Georgi Merchule - Georgian writer of the 10th century’, by Zurab V. Anchabadze

+ On some issues of ethnic identity and placement of the Abkhazians (Regarding the work of P. Ingorokva ‘Giorgi Merchule’), by Ketevan Lomtatidze

+ Giorgi Soselia’s Critique of Pavle Ingorokva’s ‘Giorgi Merchule’ and the Misrepresentation of Abkhazian History

+ The Ibero-Caucasian hypothesis and the historiography of Abkhazia, by Kevin Tuite

+ Rewriting History? A Critique of Modern Georgian Historiography on Abkhazia

The river-names tell against the theory too. The hydronymic element ps-/-psh-, "water," scattered through the toponymy of the western Caucasus and as far south as Guria, is Northwest Caucasian, as Janashia argued in 1940; Ingorokva's attempt to claim it for Georgian required exactly the kind of free-hand etymologising that Anchabadze, echoing the older charge against Nikolay Marr, called linguistic alchemy. And there is a quiet irony worth recording without making too much of it: well into the late twentieth century, leading Georgian academicians, Javakhishvili, Janashia, Chikobava, Japaridze among them, accepted, on toponymic grounds, that an Abkhaz–Adyghe linguistic layer once stretched across much of the western Transcaucasus. The "Georgian tribe" thesis was always a minority view, contested from inside Georgian scholarship as much as from outside it.

What, then, can honestly be conceded? A good deal. These were not nation-states, and "Abkhazian" and "Georgian" in this period are as much dynastic and political labels as ethnic ones. The Abkhazian kingdom did, over time, pursue an increasingly Georgian policy and shifted its capital toward the Kartvelian heartland; Janashia was right that, in its later development, it functioned within the emerging Georgian system, with Abkhaz and Kartvelian elements both at work in it. The ethnic frontier itself moved, running along the Ingur in late antiquity, the Kodor for much of the seventeenth century, and the Ingur again later, and the classical sources are sometimes vague about who lived where. Even the direction of events can be read more than one way: whether one says western Georgian lands were "annexed to Abkhazia" or that "Abkhazia was drawn into Georgia" depends partly on where one chooses to stand. The Abkhaz reading is the better evidenced of the two, but it is not the only conceivable one, and pretending otherwise would only weaken it.

None of those concessions touches the core. A people named on this coast by Pliny and Arrian, attested archaeologically without a break, speaking a language whose deep antiquity is written into its very distance from its nearest relative, and giving its name to a kingdom that everyone — Greeks, Byzantines, Georgians — called Abkhazia: that is not a tribe invented in 1954, nor a Georgian realm in disguise. It is what the evidence has been saying, quietly and consistently, since a Roman official wrote down the name of a king called Julian.