HistoricalChristian.Faith

How early is the Georgian Balavariani — 7th century (Nutsubidze) or 9th–10th (Lang)?

Type: supporting argument (off-timeline) · Assessment: sound — the Georgian Balavariani, the stage at which the Buddha-tale first became a Christian saint’s life, belongs to the ninth–tenth century, the dating on which the three specialists who edited and studied it (Lang, Kekelidze, Abuladze) independently converged. The rival seventh-century thesis of Shalva Nutsubidze — that the romance is an original composition by the Byzantine writer John Moschus — is unsupported: it “failed to attract much support,” was refuted point-by-point by Devos (1957), and founders on three independent facts — the Georgian is downstream of an eighth-century Arabic book (so a seventh-century Georgian is chronologically impossible), its New-Persian loanwords bar a date before the ninth century, and its linchpin etymology (“Moschus = the Georgian”) is false. What stays genuinely open is narrow — ninth versus tenth century, and whether the long or the short recension is prior — all of it inside the mainstream window, none of it reopening the early date.

The claim

The relative place of the Georgian stage is not in dispute: it stands after the Arabic Kitāb Bilawhar wa-Būdhāsaf and before the Greek Barlaam and Ioasaph that Euthymius the Athonite made from it. What this page weighs is its absolute date — the moment the tale was first Christianized — where two genuinely-advocated positions have stood against each other:

“There is one entirely novel and individual conception of the origins of the Georgian version of Balavariani, which failed to attract much support among scholars — the theory of Professor Shalva Nutsubidze… According to his view, the Georgian romance Balavariani (The Wisdom of Balavar) is a literary monument of the seventh century A.D. — an original composition written, like the Georgian Limonarion or ‘Spiritual Paradise’, by the Byzantine writer John Moschus.”

— Ilia V. Abuladze, introduction to David Marshall Lang, The Balavariani (1966), reporting the thesis of Sh. Nutsubidze, K proiskhozhdeniyu grecheskogo romana “Varlaam i Ioasaf” (Tbilisi, 1956).¹ Against it stands the mainstream: Lang, Korneli Kekelidze, and Abuladze — the three scholars who edited and studied the Georgian texts — independently placed the long recension in the ninth or tenth century, as an original Georgian Christian adaptation of the Arabic book. The stakes are real: an early Georgian date would pull the whole Christianization of the legend back into the seventh-century, pre-Islamic–to–Umayyad Caucasus and make the Georgians the inventors of the Christian tale rather than the adapters of an Arabic one that reached them much later.²

Who advances it

The seventh-century dating is the work of Shalva Nutsubidze (1888–1969), the Georgian philosopher and Rustaveli scholar. It is one leg of a lifelong programme of “Georgian priority” in Byzantine letters: Nutsubidze is also the author (with Ernest Honigmann, who reached it independently) of the thesis identifying Pseudo-Dionysius the Areopagite with Peter the Iberian, a Georgian — the same instinct to find a Georgian hand behind a major Greek corpus. The Balavariani-as-John-Moschus claim belongs to that pattern, and draws on the earlier Caucasianist work of Nikolai Marr, who first printed and studied the Georgian Balavar texts in St Petersburg. On the other side, the later dating is not one scholar’s but a convergence: Lang (9th c.), Kekelidze (turn of the 9th–10th c.), and Abuladze (9th or 10th c.), reached from independent starting points.³

Steelman: the case for a seventh-century original

Nutsubidze’s reconstruction is not a vague “it feels older.” It has a specific mechanism. (1) The surname Moschus, in the Georgian form Meskhi, can be read as an ethnic label — “the Meskhian,” i.e. “the Georgian” — so John Moschus (author of the Pratum Spirituale / Leimonarion, d. c. 619) could be recovered as a bilingual Georgian author writing in both Georgian and Greek; on that footing the Balavariani becomes his seventh-century Georgian composition, a companion to the Georgian Limonarion. (2) There is a real, prior tradition of Georgian priority in this legend to build on: already in 1896 F. C. Conybeare, working from Marr’s edition, argued that the Georgian is the earliest Christian form of the story, ancestral to the Greek and (through it) the Armenian — a live scholarly current, not an eccentricity. (3) Nutsubidze could point to the short Wisdom of Balavar as the authentic, more primitive core (an oral-Indian narrative record), treating the long recension as secondary — a way to argue that behind the surviving texts lies something considerably older than their manuscripts. Taken together, these move the Georgian tradition back toward late antiquity.⁴

The case for the ninth–tenth century

The Persian loanwords bar a date before the ninth century. The dating rests first on language. Marr already noted that the Persian words scattered through the Wisdom of Balavar — susti (“weak”), plasi (“hair shirt”), changi (“lyre”), pasukhi (“answer”), chabuki (“youth, knight”), laghi (“bold”) — “seemed to rule out any attempt to assign the work to a period prior to the ninth century,” because New Persian vocabulary of this kind enters Georgian literature only from the ninth century on. Abuladze’s verdict is that it is “very difficult… to judge the Georgian version… to be a production of the earliest phase of Georgian literature, the period from the fifth to the eighth century.”⁵

The Georgian is downstream of an eighth-century Arabic book. The decisive constraint is the transmission chain itself. The Georgian Balavariani derives from the Arabic Kitāb Bilawhar wa-Būdhāsaf — it carries Arabic syntactic calques and a name-chain (Arabic Janaisar → Georgian Habeneser → Greek Abonner) that fix the direction of borrowing. But the Arabic book is itself an eighth-century product: its lost first rendering from Pahlavi is placed in the ʿAbbasid translation-milieu associated with Ibn al-Muqaffaʿ (d. c. 756–759), and it is first catalogued in Ibn al-Nadīm’s Fihrist of 987–88. A source cannot predate its own source: if the Arabic parent did not exist until the eighth century, its Georgian child cannot belong to the seventh.⁶

The manuscript floor is the eleventh century. The unique full-length manuscript Lang translated — Jerusalem Georgian codex 140 — was copied in the second half of the eleventh century; the short-recension witnesses are later still (Jerusalem codex 36, 13th–14th c.; complete Wisdom of Balavar texts only from the 16th–18th c.). No surviving Georgian manuscript is anywhere near the seventh century; the early date is an inference the manuscript record does nothing to support and the loanwords positively resist.⁷

Adversarial assessment

Why the seventh-century thesis fails. Its linchpin does not hold. John Moschus is, on the settled account, a Greek monk born at Damascus (surnamed Eucrates), not a Georgian; the “Moschus = Meskhi = the Georgian” etymology is a false equation of a Greek personal surname with a Caucasian ethnonym, and the rest of the identification hangs from it. Independently of the etymology, the Arabic terminus is fatal: a seventh-century Georgian Christianization would have to precede by a century the Arabic book from which the Georgian demonstrably descends. And the New-Persian loanwords converge on the same ninth-century floor from a third, unrelated direction. Three independent lines — philological (loanwords), stemmatic (the Arabic parent), and prosopographical (who Moschus was) — all point away from the seventh century, which is why the thesis “failed to attract much support” and drew a dedicated Bollandist refutation from Paul Devos in 1957. No scholar advocates it today.⁸

What Conybeare really shows — and does not. The steelman’s appeal to Conybeare is half-right and instructive. Conybeare (1896) does argue Georgian priority — that the Georgian is earlier than the Greek and Armenian and ancestral to them — and that relative priority is now the mainstream view (it is exactly what fixes Euthymius, working from the Georgian, as the author of the Greek). But relative priority is not absolute antiquity: “the Georgian comes before the Greek” is satisfied by a ninth-century Georgian feeding a tenth-century Greek just as well as by a seventh-century one. Conybeare wrote before the Arabic-terminus and Persian-loanword arguments were in focus, and he nowhere endorses Nutsubidze’s John-Moschus dating (which postdates him by sixty years). The early-dating case borrows his priority thesis but cannot borrow a conclusion he never drew.⁹

What remains genuinely open. Rejecting the seventh century does not pretend the date is fixed to a year. Two finer questions stay live, and neither reopens the early option. (1) Ninth or tenth century? — Lang leans ninth, Kekelidze to the turn of the ninth–tenth, Abuladze allows the tenth; the window is a century or so wide. (2) Long or short recension first? — Lang, Kekelidze, and Abuladze take the full Balavariani as original and the Wisdom of Balahvar as its abridgement (M. Tarkhnishvili dissenting), against Nutsubidze’s reverse ordering; but even reading the short version as primary places it in the same ninth-to-tenth-century Georgian milieu, not in the seventh. The upper bound is firm from the other end too: Euthymius the Athonite (d. 1028) had the Georgian in hand to turn into Greek by around the 980s, so the Georgian must predate him by enough time to have circulated — comfortably consistent with a ninth–tenth-century composition, and giving no purchase to a seventh-century one.¹⁰

What it does not touch. This is a question about when the tale became Christian, not whether Josaphat is the Buddha. Whether the Christianization happened in 830 or 980, the prince’s name still descends from bodhisattva and the plot is still the Buddha’s; the date only sets how long the tale circulated as Christian hagiography in Georgia before Euthymius carried it into Greek.

Assessment

Sound. The Georgian Balavariani is a ninth–tenth-century Christian adaptation of the eighth-century Arabic romance — the dating on which its three editors independently agree, anchored by New-Persian loanwords that bar an earlier date and by the Arabic parent that no Georgian version can predate. Nutsubidze’s seventh-century thesis, resting on the identification of the author with John Moschus “the Georgian,” is unsupported: the etymology is false, the Arabic terminus makes it chronologically impossible, and it was refuted in detail (Devos, 1957) and abandoned. The residue of honest uncertainty — ninth versus tenth century, and which recension is prior — sits wholly inside the mainstream window and leaves the early date dead. (Nutsubidze’s overreach is flagged on the Georgian detail page; the companion question of when the tale became Greek — Euthymius, c. 980s, not John of Damascus — is weighed at the Greek-authorship argument.)

Sources