The Nart Sagas of the Caucasus

The Caucasus mountains run twelve hundred kilometers from the Black Sea to the Caspian. They contain more languages per square mile than any other region on Earth, more than forty, from a half-dozen unrelated language families. In high villages along their flanks, four peoples still tell variant versions of the same epic.

The Ossetians, the Adyghe Circassians, the Abkhazians, and the Karachay-Balkars all preserve a cycle of stories called the Nart Sagas. The Narts were a race of giant heroes who lived before ordinary humans and have died out. They were forged from steel by smith-gods. They drank from a cup that lifted itself only to the lips of the truthful. They rode immortal horses that knew the road to the underworld and wept when their riders died.

Almost nobody in the West has heard of them.

Three of these four peoples speak languages that are not Indo-European at all. The fourth, the Ossetians, speak Ossetic, the only surviving Northeast Iranian language and the direct linguistic descendant of Scythian and Sarmatian. Georges Dumézil put it this way in 1978: “the two dialects of Ossetian, Iron and Digor, are Scythic in the same way that Italian and Spanish are Latin.”

The Narts of the Ossetians are the last living mythology of the Scythians.

This article is the long version of what they tell. It pulls in Bronze Age comparative philology, the Sarmatian heavy cavalry that Marcus Aurelius sent to Roman Britain in 175 AD, a 1969 article in the journal Romania that compared the death of Arthur to the death of the steel hero Batraz, and the four-volume Ossetian etymological dictionary that one Soviet linguist worked on for thirty-one years.

There is a real chance, contested but not impossible, that some of what follows is the half-remembered original of King Arthur.

The Caucasus and what it preserved

The Caucasus is a refuge zone. When Indo-European steppe peoples swept across Eurasia from the third millennium BC onward, the Caucasus mountains caught them. When Turkic invasions arrived from the east in the medieval period, the Caucasus caught those too. When the Mongols destroyed the Alan kingdom in the thirteenth century, the survivors fled up into the high valleys and the mountains caught them. The result is a linguistic preserve unlike anything else on the planet.

The four Nart-bearing peoples illustrate the pattern. The Ossetians in the central Caucasus speak an Iranian language. The Adyghe Circassians and the Abkhazians in the Northwest speak languages of the Northwest Caucasian family that has no demonstrated relatives anywhere else on Earth. The Karachay-Balkars in the central Caucasus speak a Turkic language brought by later invaders. Four entirely unrelated language families, one shared mythology.

That fact alone is worth pausing on. The Nart cycle is older than any of the languages in which it is now told. It survived the linguistic turnover of two and possibly three migrations, carried forward by mothers telling children stories at the hearth. The forms differ. The names differ. The core figures and episodes do not.

When Georges Dumézil began publishing on the Narts in 1930, the comparative-mythology world recognised quickly that something extraordinary had been preserved. Vasily Abaev, the Soviet Iranist who became Dumézil’s lifelong correspondent, gave the formal verdict in his introduction to the Princeton edition of the Ossetian cycle: “the genetic ties of the Ossetian epic lead to north Iranian Scythian-Sarmatian tribes, occupying Southern Russia in the first millennium before our era.” The mountains had become a time capsule.

Who the Narts were

The Narts were a race of heroes who lived before ordinary humans. They were larger than us, stronger, longer-lived, harder to kill. They drank a fermented mead called rong that survives among the Svans of Georgia today and ate a hot cheese-and-cornmeal porridge called dzykka. They rode immortal horses that spoke and knew the road to the underworld, and that wept when their riders died. Their feasts lasted a solid week. Nobody rose from the table.

In the most famous closing-cycle, God offered them a choice between long life with mediocrity or short life with eternal fame. They chose fame. They are gone now, and only their stories remain.

The Ossetian Narts were organised into three clans. The Æxsærtægkatæ were the warriors, the line of Akhsartag, to which Batraz and most of the great heroes belong. The Borætæ were the rich herdsmen, the wealthy commoners. The Alægatæ were the wisdom-clan, the keepers of ritual and song. Dumézil saw in this tripartite division a textbook survival of the Indo-European three social functions, parallel to the Hindu varṇa structure of brāhmaṇa, kṣatriya, and vaiśya. The warrior-clan name itself preserves the relevant root: Æxsærtæɡ from Proto-Indo-Iranian kṣatriya-ka-, “of the warrior caste.” The Circassian and Abkhaz versions of the cycle lack the three-clan structure entirely, which is strong evidence that the tripartite scheme is an Iranic import laid over older material, not a Caucasian universal.

The Narts gather to feast at a long hall. They debate at council. They go on raids against neighbours and demons. They quarrel and reconcile. They slip into the underworld through caves or river-passages and come back changed. They love unwisely and produce children who become the next generation of heroes or the agents of their parents’ downfall. The structure is not unlike Homer.

What follows is the people of that hall.

The wolf-ancestor and the underwater bride

The line begins with Wærhæg, whose name in Ossetian means wolf, from Proto-Iranian warxa-ka- and ultimately from the same Proto-Indo-European root that gives English wolf and Russian volk. Wærhæg is the Nart Romulus, the founding ancestor whose blood runs through every later hero. In old age he and his wife Sainagon had twin sons, named by the heavenly smith Kurdalægon: Akhsar, “Valiant,” and Akhsartag, “Stronger than his brother.” The twins were tasked with guarding a Nart garden in which a miraculous apple tree grew, its golden fruit ripening once a year and being stolen each year by sky-doves.

Akhsartag tracked the doves to the river and shot one. The doves turned out to be the daughters of the sea-god Donbettyr. The wounded daughter healed by the river and Akhsartag fell in love with her. He followed her down into the underwater realm, where Donbettyr received him and gave him his daughter Dzerassae in marriage. The two returned to the surface world and Dzerassae bore Akhsartag the next generation: the senior Nart Uryzmæg and his lesser brother Khamyts.

The line did not end there. After both his sons died through a tragic mutual misunderstanding, the aged Wærhæg himself married his daughter-in-law Dzerassae, dying within a year. The whole Ossetian Nart line therefore funnels through one woman from the underwater clan of Donbettyr. The waters and the wolf are the two parents of the cycle.

The river-and-wolf ancestry is recognisably Indo-Iranian. The Iranian Achaemenids traced their line to a wolf. The Vedic Vṛkadvaras is “wolf-doored.” The Roman Lupercalia honoured a wolf-suckled origin myth. The Caucasus version locates the wolf at the head of the genealogy and threads the entire descent through the daughter of the river-god, an arrangement that turns out to matter when we get to Satanaya.

Satanaya, mother of all Narts

She is the figure every Nart cycle agrees on. The Ossetians call her Сатана or Satana. The Circassians call her Сэтэнай or Setenay-guasche. The Abkhazians call her Sataney-Guasha. Etymologically she has been linked to the Alanian-Armenian queen Satenik mentioned by Movses Khorenatsi in his fifth-century History of the Armenians, in connection with the Alans.

She is wise woman and witch, prophetess and mother of all the Narts, and at the same time the wife (and in some Ossetian variants also the sister) of Uryzmæg. In the Ossetian tradition she is the daughter of the river-god Uastyrdzhi, syncretised under Christianity with Saint George, and Dzerassae from the Donbettyr clan. In the Circassian tradition the river-bathing scene that produces Sosruko makes her the literal mother of the greatest hero of the cycle. Functionally she mothers them all. She holds the cup of Uatsamonga. She knows when famine will come and stores grain. She advises the Nart council on matters that the men cannot resolve.

Her wisdom-tales make her the most original character in the entire cycle. In one episode she defeats the sun in a weaving contest. In another she discovers, by watching a single flower wither in dry soil, that water is the secret of life. In a third she learns by careful observation that the Narts will face famine and arranges to set aside grain in jars sealed with beeswax to feed her people through the bad year.

The freedom of her position is the part that most surprises Western readers. North Caucasian women historically held unusually high social status compared to lowland neighbours, and Satanaya is read by ethnographers as a mythic projection of that social fact rather than a generic mother goddess. She gives orders. She withholds counsel from the unworthy. She has lovers besides her husband when she chooses. She is closer in spirit to the Norse Frigg than to the Greek Hera, and closer to neither than to a particular kind of Caucasian mountain matriarch one can still meet today in an Ossetian village kitchen.

The deep reading of her is sovereign-mother. Iranian Anāhitā in her wisdom-and-water aspect is probably the closest comparative figure on the Indo-Iranian side. The Christianised version of Donbettyr, syncretised with Saint Peter the fisherman, accidentally reinforces the etymology: don in Ossetic means “river,” from Proto-Indo-European deh₂nu-, the same root that names the Don, the Dnieper, the Dniester, the Donets, and the Danube. Satanaya descends from the river of rivers. Every great Nart hero descends from her.

The forging of Batraz

This is the most-told story in the cycle, and the most quotable.

Khamyts, the lesser of the two brothers Uryzmæg and Khamyts, was a mediocre warrior. The sources say so directly. He owned an extraordinary sword and a magical tooth, the Tooth of Arkyz, given to him by his aunt Kyzmyda after a quarrel with his elder brother. The tooth compelled women to love him. It was both blessing and curse.

He married a tiny woman from the Donbettyr clan named Bytsenon, who in eight of the eighteen recorded variants of the story transforms into a frog at night and travels in his pocket. She was beautiful. She was powerful. She was hidden. The trickster Sirdon noticed and one day publicly mocked Khamyts in front of the assembled Narts, revealing the frog-bride in his pocket. Bytsenon, shamed before the Narts, left him. Before she left, she breathed the embryo of their unborn son onto Khamyts’s back, where it became a glowing tumour.

For nine months the tumour grew on his back. The Narts watched it swell and harden. When the time came, Satanaya took a knife and surgically opened it. The white-hot infant Batraz emerged.

Sirdon, in some variants, threw the burning infant into the sea, where Batraz passed his childhood. In others Satanaya raised him in secret. When he was old enough, he was taken to Kurdalægon, the heavenly smith, whose forge sat in a cave at the edge of a great forest where mortals were forbidden to go. The smith heated Batraz to white heat and quenched him in the sea. From that day his body was steel. He was invulnerable, except in some Ossetian variants in his intestines, which the smith had not been able to temper.

His weapons struck lightning when swung. He smashed fortress walls with his body. From the west, when the sun-flash came off the sea, the Ossetians read the gleam as the sword of Batraz hurling itself out of the sea against the heavens to destroy evil powers and devils.

His revenge was brutal. He avenged his father, who had been killed by the Lord of the Black Mountain in a complicated affair involving the Tooth of Arkyz and seductions that Khamyts should not have committed. The killing-blow that ended Khamyts shattered the magical tooth, and a fragment flew into the sky. The Ossetian saga says that fragment is the moon, hanging in heaven still.

Batraz did not stop there. He turned on the Narts themselves. He turned on the heavenly spirits whom he blamed for his father’s death. He killed with such cruelty that God finally had to intervene. God told him he could not die until his sword had been thrown into the sea. The dying Batraz ordered the surviving Narts to throw it. The surviving Narts could not lift it. The few thousand still alive after his rampage finally dragged the sword to the shore and cast it in. When the sword struck the water a great storm rose, the sea boiled and turned red, and Batraz died.

This is the episode the Dumézilian philologist Joël Grisward wrote about in 1969 in the journal Romania: the death of Batraz beside the death of Arthur in the Old French Mort Artu. Three attempts, sword in water, a supernatural response, the king’s death following. We will return to it.

Soslan born from a stone

The other most-told hero of the cycle has the widest cross-cultural distribution of any Nart figure. The Ossetians call him Soslan. The Adyghe Circassians call him Sosruko. The Kabardians call him Sosryqwa. The Abkhaz call him Sasərqwa. The Adyghe folk etymology breaks the name into sa (sword) and wa (hit) and sər (heat) and qʷa (son): “the son of the fiery sword-strike.”

His birth is the most extraordinary in any mythology I have read. Satanaya was bathing naked at the river. A shepherd on the far bank, watching her, ejaculated onto a stone across the water. Nine months later the stone groaned. Satanaya, who had marked the stone, returned and split it open and lifted out the glowing-hot child. She carried him to the smith, who tempered his body in a trough of wolf milk (or a hundred goatskins of milk in another version). The trough was too short. Soslan had to bend his knees. The knees alone remained untempered, the only vulnerable spot on his entire body.

This is, structurally, Achilles. The mother tempers the child against death and misses the one small place that will kill him.

His career runs parallel to Prometheus. He stole fire from the one-eyed giant Inyzh, slipping a smouldering log out from under the sleeping ogre and accidentally dropping a coal that burned Inyzh’s eyebrow as he escaped. The Greek punishment never came. Inyzh just woke up confused.

His death is the great set piece of the cycle. The Wheel of Balsæg, in some variants the Wheel of Ojnon (a syncretic Christian figure) or the wheel sent by the Daughter of the Sun whom Soslan had spurned, descended from the sky. The Wheel was a sentient automaton, a razor-sharp metal disk with steel teeth and flames bursting from it. It first attacked his magical horse, striking the horse’s vulnerable hooves. Then it flew at Soslan and sliced through his unprotected knees. He died on the steppe.

Scholars read the Wheel as a solar symbol: the sun-disk that cuts down the solar hero, possibly a survival from solstice ritual. The most explicit Indo-Iranian parallel for Soslan is Mithra, who in the Roman mysteries is petrogenitus, born from the rock. The Hellenized commentators Justin Martyr and Firmicus Maternus noted the Mithras rock-birth motif in their critical writings on the cult. Some scholars, including Littleton and Malcor, have argued that the Roman Mithras’s rock-birth iconography itself derives via Sarmatian intermediaries from the Sosruko stone-birth narrative. The transmission line is contested. The structural identity is striking either way.

The Soslan cycle also contains the great Caucasian katabasis. To win the Daughter of the Sun he had to fetch leaves from the aza tree, which grew only in the Land of the Dead. He forced his way in alive, met his deceased first wife who helped him navigate the underworld, saw the sufferings of the dead as requital for their deeds during life, took the leaves from the underworld king Barastyr, and was harassed on the way back by the shape-shifting Sirdon, who learned the secret of his immortality. A 2022 article in Brill’s Iran and the Caucasus compared the Soslan descent directly to Dante’s Divine Comedy.

Sirdon the trickster

Dumézil identified Sirdon (Syrdon in some transliterations) as the most exact known Indo-European parallel to Loki. Closer than any other figure in any other tradition. The 1948 monograph Loki makes the case in detail.

Like Loki, Sirdon provokes the death of the solar hero. Loki engineers the death of Baldr. Sirdon engineers the death of Soslan. Like Loki, he is a shape-shifter, becoming a werewolf, an old man, a young girl. His weapon, in the saga’s own description, is “his tongue, sharp, poisonous and pitiless.” Like Loki, he is paradoxically a culture-hero. He invents the fændyr, the twelve-stringed Ossetian harp, in the most famous and most disturbing episode of his cycle.

The story is this. Sirdon had stolen something from the Narts, a cow in some versions, a magical object in others. The Narts in revenge captured his sons, killed them, boiled them, and served them to him at a feast. When Sirdon recognised what he was eating, he did not weep or rage. He went home, gathered the bones of his sons, stretched their dried sinews across a frame of bone, and played the first music ever heard in the world. The Narts heard the song from his hut and recognised that something new and unbearable had entered the world. They forgave him. They invited him into their council as a Nart of the highest rank. The harp survived. The harp survives.

The figure is half-Nart and half something else. In some variants his mother is a daughter of Donbettyr. In others she is a witch. In others she is a demon’s daughter. He is an outsider whose wisdom the elder Narts cannot dispense with. The Vilnius philologists Eloeva and Sausverde in a 2015 article in Literatūra traced a general evolutionary pattern through the figure of the trickster: chthonic deity becomes trickster becomes culture-hero. Prometheus, Loki, Sirdon, Coyote. Sirdon is the Caucasian datum point on that line.

The pantheon

The Narts share a world with several greater beings. Most of them retained their pre-Christian names and functions even after Christian syncretism arrived in the medieval period.

Tlepsh in the Adyghe tradition, Kurdalægon in the Ossetian tradition, is the heavenly smith. He forged Batraz in his sea-quenching furnace and tempered Soslan in the trough of wolf milk. His Caucasian name is cognate with Greek khálups, “iron, steel,” which entered Latin as chalybs and thence European languages. The Ossetian name Kurdalægon contracts to Kurd Alæ Wærgon, “Smith of the Alans.” His forge is in a cave at the edge of a great forest where mortals are forbidden to go; the Narts call him three times when they need him. He made the first sickle. He made a self-propelled sword that hunted and killed cholera personified, called Bearded Yamina. He is the direct counterpart of Hephaestus, Vulcan, and Norse Wayland the Smith.

Donbettyr is the god of all waters. His name combines don, “river,” from PIE deh₂nu-, with bettyr, a syncretism with Saint Peter the fisherman that came in with Christianity. The Iranic don is the river-name root that gives the Don, the Dnieper, the Dniester, the Donets, and the Danube. His daughters are the Caucasian analogue of the Slavic Rusalki. Through Dzerassae, his daughter, he is the maternal grandfather of every major Nart.

Tutyr is the lord of the wolves. His name was syncretised with Saint Theodore Tiron of the Eastern church. Ossetian belief holds that no wolf attacks livestock without Tutyr’s leave, and that calling on Tutyr makes the wolves flee. His cosmic enemy is Falvarra, lord of sheep, named after Saints Florus and Laurus, who is one-eyed because Tutyr struck out the other eye in their endless quarrel.

Æfsati in the Ossetian tradition, Apsat in the broader Caucasian, Mezitha in the Circassian, is the antlered lord of game and the hunt. He is old and bearded, often one-eyed or blind. He lives in a forest hut with a wife and daughters and sometimes lets a lucky hunter marry one of them. Ossetian hunters call game animals “Æfsati’s cattle.” On a successful kill they roast the heart and liver as offering. He was syncretised with Saint Eustace, the Roman general converted by a vision of Christ between the antlers of a stag.

Uastyrdzhi is the god of men, warriors, oaths, and travellers. He is forbidden to women, who must call him by paraphrase. He was syncretised with Saint George. He is the most-toasted figure at every Ossetian feast and the father of Satanaya in the Ossetian variant.

Barastyr is the ruler of the Land of the Dead, the gatekeeper of the underworld in Soslan’s descent.

There are giants, the Inyzh or Waeyug, some stupid and cyclopean (the one from whom Sosruko stole fire), some formidable warriors, a few friendly. Many die humiliating deaths to Nart trickery rather than Nart strength. There are mountain spirits called dauags, guardian-deities of particular peaks and passes. There are dragons, multi-headed, that guard treasures and princesses and occasionally swallow the sun.

This is the divine company of the Narts. Smith, river, wolf, hunt, war, death, and the giants and dragons that fill the spaces between.

The cup that knows when you lie

Of all the artefacts in the cycle, the most beautiful is the cup.

In the Ossetian sagas it is called Nartamongæ in the older pre-Christian usage and Watsamongæ or Uatsamongæ in the Christianised form. The linguistic shift is precise. Nart-amon-gæ is “Nart-pointing-diminutive.” Wæc-amon-gæ is “holy-pointing-diminutive.” The pre-Christian cup that pointed at Narts became the Christian cup that pointed at the holy. Same artefact. Different vocabulary.

It is a magic cup. At the great feasts, when each Nart in turn rose to recount his greatest deed, the cup raised itself to the lips of any Nart whose telling was true. It stayed motionless before braggarts. It contained rong, the heroes’ fermented mead. It was held in some variants by Satanaya and in others by Dzerassae. It polices boasting in real time.

The Russian Iranist Vsevolod Miller in 1881 first compared it to the gold goblet of Herodotus’s Scythians. Herodotus 4.66 describes a once-yearly Scythian feast at which each district chief mixed wine and water in a goblet that only Scythians who had killed an enemy were permitted to drink from. Aristotle in the Politics mentions the same restriction. Dumézil in 1978 placed the Uatsamonga in the Indo-Iranian ritual-cup family with Vedic soma in its golden chalice and Avestan haoma in its sacred vessel. These were not table cups. They were sacred objects through which the divine entered the human world.

The Scythians took the comparison further. Herodotus 4.59 lists their gods. Herodotus 4.62 describes their war-god, identified with the Greek Ares but in fact an Iranian war-deity, almost certainly Vərəϑraγna. The cult was iconic.

“An antique iron sword is planted on the top of every such mound, and serves as the image of Mars: yearly sacrifices of cattle and of horses are made to it, and more victims are offered thus than to all the rest of their gods.”

Abaev drew the line directly. The ceremony of casting Batraz’s sword into the sea compares with the Scythian and Alanic cult of the sword. The bonfire of a hundred wagonloads of coal into which Batraz strode to be tempered before the trembling Narts recalls the grandiose annual construction of a fire of a hundred and fifty cartloads of logs that served the Scythians as a pedestal for their sword-god. Batraz was not invented by the Caucasus. He was inherited from the steppe.

Littleton and Malcor in From Scythia to Camelot (1994, revised 2000) added one more step. The Sarmatian-Alanic ritual cup, they argue, is the structural ancestor of the Holy Grail, transmitted west via Sarmatian and Alanic contingents in the late Roman army. The linguistic part of that argument is real. The historical-transmission part is contested. We will get to it.

Dumézil and the trifunctional reading

Georges Dumézil began publishing on the Narts in 1930 with Légendes sur les Nartes. He was working at the Institut français de Leningrad and had access through Vasily Abaev to the Russian-language ethnography that French scholars otherwise could not read. By 1948 he had published Loki, which set up the Sirdon-Loki comparison that has held up for seventy years. By 1968 he had folded the Narts into his great synthesis Mythe et épopée. By 1978 he had distilled the case for non-specialists in Romans de Scythie et d’alentour.

The signal contribution was the trifunctional hypothesis. Dumézil argued that Indo-European societies preserved a tripartite ideology of three social functions: sovereign-priests, warriors, and producers. The hypothesis is contested in detail. It is undisputed in broad outline. The Indic varṇa system codes the same three: brāhmaṇa, kṣatriya, vaiśya. The Roman religious-civic structure codes them too: flamen, miles, quirites. The Norse Æsir code them in different garb: Odin and Týr for sovereignty and law, Thor for war, Freyr and Freyja for fertility and wealth. The Ossetian Nart cycle codes them with diagrammatic clarity: Alægatæ for ritual, Æxsærtægkatæ for war, Borætæ for wealth.

The Narts are the cleanest surviving Indo-European trifunctional epic. Cleaner than the Iliad. Cleaner than the Mahābhārata. Cleaner than the Norse sagas. The reason is that the Narts had no state. The cycle was preserved by mountain villagers who never had a court poet to systematise the variants. What survives is closer to the deep structure than any of its more famous Indo-European cousins.

Dumézil’s other Nart-specific contribution was the comparative reading of Batraz as Indra. Both are warrior heroes. Both have unusual births. Both wield weapons that produce lightning. Both are afflicted in their final phase by something approaching madness and ultimately need divine intervention to be ended. The Encyclopaedia Iranica places Batraz in the same Indo-European cluster as Indra, Achilles, Prometheus, the Georgian Amirani, and the Armenian Artavazd. The cluster is real. Dumézil was the first to map it.

The Sarmatian hypothesis: documented vs conjectured

Now the Arthurian question.

This is the part where popular accounts overstate. Scholarly accounts are more honest. The Position Three method we used for Adam and Eve and for the Jesus DNA piece applies here too: separate what is documented from what is conjectured, present both honestly, let the reader decide.

What is documented is real, and worth pausing on.

In 175 AD, the closing phase of the First Marcomannic War, the Roman emperor Marcus Aurelius defeated a Sarmatian people called the Iazyges and dictated peace terms. Cassius Dio in Book 72 of his Roman History (chapter 16, in the surviving epitome by Xiphilinus) records the settlement: the Iazyges returned a hundred thousand Roman captives still in their hands, agreed to live twice as far from the Danube as other defeated tribes, and “promptly furnished as their contribution to the alliance eight thousand cavalry, fifty-five hundred of whom he sent to Britain.”

Five thousand five hundred Sarmatian heavy-armoured cataphract cavalry. Sent to Britain. In 175 AD. By Marcus Aurelius. This is not a conjecture. It is the Loeb edition of Cassius Dio.

Their unit, an ala equitum Sarmatarum, is epigraphically attested at Bremetennacum, the modern Ribchester in Lancashire. The fort took the suffix Veteranorum, “of the veterans,” indicating that retired Sarmatian troopers and their descendants settled around the fort and stayed there as a community.

The Sarmatians worshipped, per Herodotus’s account of their Scythian relatives, a war-god whose image was a sword stuck upright in a mound. They threw their dying heroes’ swords into water at death. They drank from a sacred cup. They preserved a heroic mythology that we can now read in the Ossetian Nart Sagas.

That is the documented base. From here the popular case overreaches and the scholarly case stays cautious.

The popular case, made most vividly in C. Scott Littleton and Linda A. Malcor’s From Scythia to Camelot (1994, revised 2000) and Howard Reid’s Arthur the Dragon King (2001), is that the historical King Arthur was the Roman officer Lucius Artorius Castus who commanded the Sarmatian cavalry in Britain. The name Artorius gave rise to Arthur. The Sarmatian veterans of Ribchester transmitted their oral mythology to British folklore. The mythology resurfaced a thousand years later in the 12th-century Continental romances of Chrétien de Troyes and the Vulgate Cycle. The death of Batraz is the original of the death of Arthur. The cup of Uatsamonga is the original of the Holy Grail. Lancelot is “Alan-of-Lot,” named after the Alans who settled the Loire valley.

The scholarly case is less generous on every specific point.

Two surviving Latin inscriptions from Podstrana in Croatia, Corpus Inscriptionum Latinarum III no. 1919 and III no. 14224, document Lucius Artorius Castus’s career. Centurion of four legions, primus pilus of the Fifth Macedonian, prefect of the Sixth Victrix at York, and at one point dux of British detachments “adversus ARM[…]S.” The decisive disputed letters are those last six in the dux phrase. Linda Malcor reads them as ARMORICOS, “against the Armoricans,” meaning a campaign against Brittany. Britain’s leading Roman epigrapher, Roger Tomlin of Wolfson College Oxford, reads them as ARMENIOS, “against the Armenians,” meaning service in the Parthian-Armenian war of Lucius Verus around 161 to 166 AD. Tomlin’s argument is epigraphic: the seven letters required for Armoricos “cannot be fitted into the space available,” and the term Armorici is not attested in any other Latin inscription, while Armenios fits and Roman victory titles like Armeniacus were standard.

Three competing chronologies for Castus’s career exist in modern scholarship. Željko Miletić places him roughly 121 to 166 AD, dying before the Sarmatians ever reached Britain. Nenad Cambi places him in the late second century. Xavier Loriot places him in the early-to-mid third century, well after the Sarmatian arrival. None of these places him in command of Sarmatians. The inscription nowhere mentions Sarmatians, Iazyges, or Bremetennacum. The Castus-Sarmatian connection is entirely circumstantial.

The Lancelot-as-Alan-of-Lot etymology has not been accepted by mainstream Romance philologists. Four other etymologies compete for the name, none requiring Sarmatian transmission. Lancelot’s first appearance is in Chrétien de Troyes’ Le Chevalier de la Charrette around 1177, where he is already styled Lancelot du Lac, “of the lake,” referring to his fairy upbringing by the Lady of the Lake.

The strongest piece of the case is also the oldest and most independent of the others. Joël Grisward in 1969, working in the Dumézilian comparative-mythology tradition, published a long article in the journal Romania titled “Le motif de l’épée jetée au lac : la mort d’Artur et la mort de Batradz.” He set the death of Arthur in the Mort Artu (the dying Arthur orders his sword cast into the lake; on the third attempt the knight obeys; a hand rises from the water to receive it) beside the Ossetian death of Batraz (the dying iron-bodied hero orders his sword cast into the sea; on the third attempt the deed is done; the waters boil). He argued the structural parallel is too specific to be coincidence and reflects a shared Indo-European mythologem. Crucially, Grisward’s argument is not the Sarmatian-cavalry-in-Britain story. Grisward argued for shared deep Indo-European inheritance, not for second-century military transmission. Littleton and Malcor adopted his philological parallel and bolted it to a different historical claim. The parallel can be real and the cavalry-transmission story still wrong.

What does the responsible reading look like?

The cavalry transfer is real. The Castus-as-Arthur identification is much weaker than popular accounts suggest and is rejected by mainstream Arthurian scholarship (Higham 2002, Padel 2000, the Kenneth Jackson Welsh-Brythonic consensus, Tomlin on the inscription). The Batraz-Excalibur parallel is real but most economically explained as common Indo-European inheritance from a deep mythologem. The Lancelot etymology is wishful. The cup-as-Grail comparison has linguistic teeth (Nartamongæ → Watsamongæ is a documented Christian shift) but the historical-transmission story is conjectural.

The honest summary: there was a deep Indo-European mythological substrate that the Scythians preserved, that the Sarmatians inherited, that the Alans carried into the Caucasus, and that the Ossetians kept alive in their epic until our own century. Some of that substrate may have been carried west by Sarmatian and Alanic troops in the late Roman army. It very likely surfaced in different forms in the Welsh and the Continental Arthurian traditions. The exact pipeline from Ribchester to Camelot is harder to draw than the most romantic case suggests. The exact pipeline does not need to exist for the comparative-mythological parallels to mean what Grisward said they mean.

The Narts and the Arthurian cycle are cousins. They may not be parent and child.

How the cycle was preserved

We have these stories because of three generations of scholars working under three different political regimes who insisted the material mattered.

Vsevolod Miller (1848-1913) was the first. A Russian Iranist who published Осетинские этюды / Osetinskie Etyudy (Ossetian Studies) in three volumes between 1881 and 1887. Volume I gave the Ossetian-language folklore texts. Volume II gave the phonetics, the grammar, and a chapter on Ossetian religious beliefs that is the first systematic etymological analysis of the Ossetian pagan pantheon. Volume III gave the history, the ethnography, and the proverbs. Miller was the first scholar to argue Scythian-Sarmatian-Alanic continuity into Ossetian, the first to identify the Iranian nar root in the name “Nart,” and the first to read the Nart corpus as a national epic on the level of the Iliad or the Mahabharata.

Vasily Ivanovich Abaev (1900-2001) is the second. Born in Kobi, Georgia, on December 15, 1900; died in St Petersburg on March 18, 2001, at the age of 100. His career was largely at the Iranian Studies institute in Leningrad. His magnum opus is the four-volume Историко-этимологический словарь осетинского языка (Historical and Etymological Dictionary of the Ossetian Language), published in Moscow and Leningrad between 1959 and 1989. He drew on material from 190 languages and dialects. He received the USSR State Prize in 1981. He was Dumézil’s lifelong correspondent and friend, providing the Russian-language linguistic spine on which Dumézil built the comparative-mythological argument that the Nart cycle preserves Scytho-Sarmatian Indo-European material in fossilised form.

Asker Hadaghatla is the third, the equivalent figure for the Adyghe Circassian tradition. Soviet expeditions of the 1950s collected the Adyghe material across the Northwest Caucasus. Hadaghatla edited the seven-volume Нартхэр (Narts) published in Adygea between 1968 and 1971. The edition contains 705 texts and 40 musical notations recorded in the field. It is the largest published Nart corpus in any language.

The English-language reception did not catch up until the 21st century. John Colarusso, a linguist at McMaster University, published Nart Sagas from the Caucasus: Myths and Legends from the Circassians, Abazas, Abkhaz, and Ubykhs with Princeton University Press in 2002. Ninety-two tales. The first compendium in any language of Nart sagas from all four Northwest-Caucasian-speaking peoples. He followed it in 2016 with Tales of the Narts: Ancient Myths and Legends of the Ossetians, edited with Tamerlan Salbiev, with Walter May’s translation and with Abaev’s classic introduction translated into English for the first time. Eighty-nine tales in eight sections. The two Princeton volumes are now the standard English-language base. Colarusso 2002 for the Northwest Caucasian variants. Colarusso and Salbiev 2016 for the Indo-Iranian Ossetian variants. Together they are the only way an English reader gets the full divergence between the cycle’s two great branches.

The cycle is not preserved only in books. It is preserved in the kitchen. Every Ossetian household feast still uses the three pies. Every toast still names Uastyrdzhi. The cup of Uatsamonga survives metaphorically in the strict etiquette of who speaks first and last and how much one is allowed to claim of one’s own deeds. The mythology never died. It just waited for outsiders to learn the language well enough to write it down.

The mountain that kept what the lowlands lost

The Caucasus is what survives when civilisation cannot reach you. Imperial Persia could not get into the high valleys. Imperial Rome could not, the Mongol horde could not (though they devastated the foothills), and the Ottoman empire could not. The Russian empire took two and a half centuries of war to subdue the lowland Adyghe and never fully subdued the Chechens or the Daghestanis. Through every wave, the high mountain villages kept telling the old stories at the hearth.

What they kept turns out to be much older than any of them. The Ossetian Nart cycle preserves Indo-European mythological material in clearer form than any other surviving epic. The cycle as a whole records a Bronze Age Eurasian heroic ideal that has otherwise vanished: smith-gods who forge their warriors from steel, cups that police boasting at the feast, three social functions diagrammed into three clans, wolves at the head of the genealogy and rivers at the head of the matriline, a trickster who invents music from the bones of his murdered children, a hero born from a stone after a shepherd’s seed strikes the rock across a river, and another hero who can only die when his sword is thrown into the sea.

Some of these motifs may have been carried into Britain by Sarmatian cavalry in the second century and resurfaced a thousand years later as Excalibur and the Round Table and the Grail. The case is real and the case is contested. Either way, what the Caucasus preserved is older than the controversy. The Narts are the Bronze Age of Europe still speaking.

If you visit an Ossetian village in summer and are invited to a feast, you will be served three pies stacked one on top of the other. The elder will slide them slightly apart so the Most High can see there are indeed three. The first toast will be to the Great and Only God. The second will be to Saint George, who is also Uastyrdzhi, who is the father of Satanaya, who is the mother of all the Narts. The third toast will be to whatever has brought you to the table. Somebody will eventually tell a story about Batraz being forged from steel by Kurdalægon, or about Soslan stealing fire from the one-eyed giant, or about the Sirdon stringing the harp from his sons’ sinews.

You will be eating a meal that is also a cosmology. You will be hearing a story that is also a mythology. You will be a guest at a Bronze Age feast that has been continuously running for three thousand years, and that nobody in the West is paying attention to.

That is the article. The mountain kept what the lowlands lost. The Narts are still here.

The companion piece on universal-First-Couple mythology and Western theology is Why Are Adam and Eve Not Holy?, which uses the same Position Three method on the genetic Adam and Eve and the universal cultural pattern. For the related question of how a single mythological tradition can encode the worldview of an entire civilisation, see Would This Prove That God Exists? What Jesus’s DNA Would Actually Look Like.

Sources

Bibliography organised by the section it belongs to. The full list is also held in the article’s frontmatter for the citation tools that read it programmatically.

Primary saga collections

• Colarusso, J. Nart Sagas from the Caucasus: Myths and Legends from the Circassians, Abazas, Abkhaz, and Ubykhs. Princeton University Press, 2002 (paperback reissue 2016, ISBN 9780691169149)
• Colarusso, J., and Salbiev, T. (eds.), Walter May (trans.). Tales of the Narts: Ancient Myths and Legends of the Ossetians. Princeton University Press, 2016. ISBN 9780691170404
• Hadaghatla, A. (ed.). Нартхэр (Narts), 7 vols. Maykop: Adygea, 1968–1971
• Miller, V. Осетинские этюды / Osetinskie Etyudy, 3 vols. Moscow, 1881–1887

Ossetian philology and lexicography

• Abaev, V. I. Историко-этимологический словарь осетинского языка / Historical and Etymological Dictionary of the Ossetian Language, 4 vols. Moscow–Leningrad: USSR Academy of Sciences, 1959–1989
• Abaev, V. I. “Introduction: The Ossetian Epic Tales of the Narts,” in Colarusso & Salbiev (eds.), Tales of the Narts, Princeton, 2016, pp. xxix–lxviii
• Kim, R. “Ossetic,” in G. Windfuhr (ed.), The Iranian Languages. Routledge, 2009
• Kim, R. “Case Marking from Old Iranian to Ossetic,” Transactions of the Philological Society 123:2 (2025)

Dumézil and the trifunctional comparative reading

• Dumézil, G. Légendes sur les Nartes, suivies de cinq notes mythologiques. Paris: Honoré Champion, 1930
• Dumézil, G. Loki. Paris: G.-P. Maisonneuve, 1948
• Dumézil, G. Mythe et épopée, vol. I. Paris: Gallimard, 1968
• Dumézil, G. Romans de Scythie et d’alentour. Paris: Payot, 1978

The Sarmatian hypothesis: Arthur, Excalibur, the Grail

• Grisward, J. H. “Le motif de l’épée jetée au lac : la mort d’Artur et la mort de Batradz,” Romania 90 (1969), pp. 289–340 and 473–514
• Littleton, C. S., and Malcor, L. A. From Scythia to Camelot: A Radical Reassessment of the Legends of King Arthur, the Knights of the Round Table, and the Holy Grail. Garland, 1994; revised Routledge, 2000. ISBN 0-8153-3566-0
• Malcor, L. A. “Lucius Artorius Castus, Part 1: An Officer and an Equestrian,” The Heroic Age 1 (1999); “Part 2: The Battles in Britain,” The Heroic Age 2 (1999)
• Reid, H. Arthur the Dragon King: The Barbaric Roots of Britain’s Greatest Legend. London: Headline, 2001. ISBN 0-7472-6225-X
• Higham, N. J. King Arthur: Myth-Making and History. Routledge, 2002. ISBN 0-415-21305-3
• Padel, O. J. Arthur in Medieval Welsh Literature. Cardiff: University of Wales Press, 2000. ISBN 0-7083-1682-4

Ancient and classical sources

• Herodotus, Histories, Book IV, esp. 4.59–62 and 4.66 (Loeb edition)
• Cassius Dio, Roman History, Book 72 (formerly 71), ch. 16. Loeb Classical Library Vol. IX (Earnest Cary, trans.). Harvard University Press, 1927
• Corpus Inscriptionum Latinarum III, no. 1919 (= 8513 = 12813); CIL III 14224. Berlin-Brandenburg Academy of Sciences
• Movses Khorenatsi, History of the Armenians (5th century)

Modern journal scholarship

• Eloeva, F., and Sausverde, E. “Culture Hero’s Intrepid Past: Prometheus, Loki, Syrdon, Coyote.” Literatūra 57, no. 3 (2015), pp. 98–115. DOI 10.15388/Litera.2015.3.9880
• Ognibene, P. “Escape from the Land of the Dead: Nart Sagas, Divine Comedy, and the Journey Through the Afterlife.” Iran and the Caucasus 26, no. 3 (2022), pp. 201–216. Brill. DOI 10.1163/1573384X-20220302
• Nartamongæ. The Journal of Alano-Ossetic Studies, vols. I–II. Vladikavkaz / Paris, 2002–2003