Do the divergent name-rosters reveal independent Greek and Latin transmission from a common archetype?

Type: supporting argument (off-timeline) · Assessment: sound — the interpretation the rosters serve is well-grounded and effectively uncontested: the surviving legend descends through separate Eastern-Greek and Latin lines that each go back to a common Eastern archetype, not through a single copying-chain. The name-sets corroborate this (a shared core points to one source, the divergent tails to two branches), though they are a noisy witness in their own right and are not the load-bearing proof · the framework of the recension scholarship (Koch, Huber, Brock).

The claim

The legend carries not one list of the sleepers’ names but several, and the two best-known pull cleanly apart. The polished Byzantine Passion of Symeon Metaphrastes names them Maximillian, Jamblichos, Martin, John, Dionysios, Exakostodianos, and Antoninos; the Latin Gregory of Tours — and the Western tradition that flows from him into the Roman Martyrology and the Golden Legend — names them Maximianus, Malchus, Martinianus, Constantinus, Dionysius, Johannes, and Serapion.¹ The positive literary-history claim built on this is that the divergence is a transmission fingerprint: two rosters this different, yet sharing a recognisable core, are the signature of two lines that descended independently from one common (Eastern) source, rather than one line copying another. It is a claim about the shape of the stemma — that it branches — and is distinct from the priority question of where the branching tree is rooted (in the Syriac, or in a lost Greek original).

Who advances it

The reading is the working assumption of the recension scholarship that mapped the legend’s spread — John Koch (1883) and Michael Huber (1910), who sorted the more than two hundred surviving manuscripts (some 104 Latin, 40 Greek, 33 Arabic, 17 Syriac, and the rest scattered across Ethiopic, Coptic, Armenian, Old French, Old English, and Middle Irish) into their language-families — and it is the ordinary tool of textual criticism, where a shared but imperfectly-shared onomastic core is read as evidence of common descent with branch-drift. S. P. Brock’s survey of the recensions treats the Syriac, Greek, and Latin as related but distinct lines in just this way.²

The evidence for

The core is shared; the tails diverge. Line the two lists up and the pattern is unmistakable. Four of the seven match or are plainly cognate — Maximian(us)/Maximillian, Martin(ianus)/Martin, John/Iohannes, and Dionysios/Dionysius — while the remaining three differ wholesale: the Greek carries Jamblichos, Exakostodianos, and Antoninos where the Latin carries Malchus, Constantinus, and Serapion. This is precisely the profile a stemma predicts. A single copying-chain would tend to preserve the whole roster intact; wholly independent invention would share no names at all. A shared core with divergent tails is what you get when two scribal traditions each inherit one archetype and then, over centuries and across a language boundary, each drift, drop, and refresh the harder-to-remember names in their own way.

The branching is independently documented. The rosters do not have to carry the thesis alone. Gregory states in his own colophon that he did not work from the Greek Passion at all but from a Syriac exemplar read to him by an interpreter — “which, interpreted by a certain Syrian, I translated into Latin” (Siro quodam interpretante in Latino transtulimus), the interpreter named in the Passio as John the Syrian (interpretante Iohanne Syro). That is hard, non-onomastic evidence that the Latin line reached the West through the Syriac branch, on a track separate from the Greek — exactly the independence the divergent rosters imply.³

A third branch confirms the shape. When the tale passes into Arabic, the names al-Ṭabarī and the later commentators record — Yamlīkhā, Maksalmīnā, Marṭūs, and the rest — are transparently Arabized from the Greek set (Yamlīkhā from Iamblichos/Jamblichos, Marṭūs from Martinos), not from the Latin. The Islamic tradition thus visibly hangs off the Eastern-Greek line rather than the Western one — a second, independent case of a whole onomastic sub-tradition tracking one branch and not the other, which is just what the “independent transmission” reading predicts.⁴

Adversarial assessment

What holds. That the surviving forms descend through separate Greek and Latin lines from a common Eastern archetype is not really in dispute, and it does not rest on the names: Gregory’s colophon, the geography of the witnesses, and the shared narrative spine settle the branching on their own. The rosters fit that picture well — the shared core argues for one source, the divergent tails for two branches, and the Arabic list’s descent from the Greek set corroborates it a third time.

Where the argument is overstated. The weak link is treating the rosters themselves as a precise stemmatic instrument. Names are the most volatile element in hagiography — freely coined, latinized, graecized, and re-spelled — so they drift within a single tradition as readily as between two. The point is easy to see: the standard reference article on the legend confidently assigns Gregory a completely different set (Achillides, Diomedes, Diogenus, Probatus, Stephanus, Sambatus, Quiriacus) than his critical text actually carries, so even the secondary Latin transmission scrambled the “Latin” roster.⁵ A tool that noisy cannot bear fine-grained weight.

And the names may be a late layer. The two earliest witnesses barely fix the roster at all: Jacob of Serugh’s homily and the Qur’an tell the story without the canonical seven names, and the Qur’an even leaves the number openly disputed (18:22). If the names crystallized only after the archetype had already split, then divergent rosters date the naming layer of each branch, not the branch-split itself — still evidence of two lines, but weaker and later evidence than the clean “fingerprint” framing suggests. The honest form of the claim is therefore modest: the rosters corroborate a branching that is established on firmer ground; they do not, by themselves, prove it.

Assessment

Sound. The interpretation the argument serves — that the legend reached us through independent Greek and Latin lines descending from a common Eastern archetype — holds, and is close to consensus among the scholars who have worked the recensions. The divergent name-rosters are real and genuinely corroborative: a shared core with divergent tails is the classic signature of common descent with branch-drift, and the Arabic tradition’s descent from the Greek set confirms the pattern a third time. What must be resisted is the stronger, tidier claim that the rosters are a decisive fingerprint: names are the noisiest data hagiography offers, they scatter within a single line as well as between lines, and they appear to be a secondary layer laid over an already-split tradition. So: the branching stemma is sound; the rosters are good corroboration of it, not the proof that carries it. This is a question about the shape of the tree, and is left open only at its root by the companion priority question.

Sources