Is Clement of Alexandria’s witness undermined by textual unreliability — that it survives only in an expurgated sixth-century Latin translation?

Type: adverse argument (off-timeline) · Assessment: weak — the objection concedes that Clement “appears to support” Mary’s virginity, then discounts him on the ground that “the Greek original is lost” and only a sixth-century Latin adaptation by Cassiodorus survives, “written specifically to correct offending sentiments.” It misfires by a category error: it hangs that caveat on Clement’s Stromateis VII.16, which survives in Greek and was never a Cassiodorus adaptation — the passage the objector cites is not the text his source is cautioning about. Juniper Carol’s “we cannot rely absolutely on this text” is a caution about Clement’s Latin-surviving Adumbrationes (his notes on the Catholic Epistles, the source of the remark on Jude), not about the Greek Stromateis; and even applied there it cuts the wrong way, since a Latin editor purging error and holding the rival Hieronymian view would have removed Clement’s Epiphanian line, not inserted it (J. B. Lightfoot) — its survival is a mark of fidelity, not corruption. Carol, a Catholic Mariologist affirming the doctrine, is quoted against his own conclusion · pressed by Thorncrown Ministries

The claim

— Thorncrown Ministries, “Roman Catholics and Their Queen, Part 4” (2017), quoting Juniper B. Carol (ed.), Mariology, vol. II, p. 271.¹ The move is a familiar one: grant that a father “appears to support” the doctrine, then neutralize him by casting doubt on the text that carries the witness. Here the objection turns on two facts about transmission — that Clement’s Greek is “lost” and that what survives is a sixth-century Latin adaptation by Cassiodorus made expressly to purge unsafe matter — from which it infers that Clement’s apparent support cannot be trusted.

Who advances it, and what Carol actually said

The caveat is quoted, accurately, from a Catholic authority: Fr. Juniper B. Carol’s Mariology (1957), the standard pre-conciliar English handbook. But Carol is himself defending the perpetual virginity, and his caution is about a particular text. Cassiodorus (c. 485–585), in his Institutiones, records that he had Clement’s notes on the Catholic Epistles rendered into Latin and the “certain offensive matters” in them purged — the fragmentary Latin work known as the Adumbrationes, Clement’s comments on 1 Peter, Jude, and 1–2 John.² That is the “translated adaptation… expurgating anything offensive” Carol names — the Latin notes on the epistles, whose Greek original (from the lost Hypotyposeis) is indeed gone. It is the same text from which Clement’s clearest brethren-statement comes: the note on Jude, that Jude was “the brother of the sons of Joseph… (the son) of Joseph,” weighed on Clement’s detail page.

The evidence for

The kernel is real and should be granted. Clement’s Adumbrationes really do survive only in a sixth-century Latin abridgement that Cassiodorus, on his own account, edited to remove what he judged doctrinally unsafe; so on any point drawn from that Latin a critic is entitled to ask whether the wording is Clement’s or the translator’s. And it is fair to note that Clement’s Hypotyposeis as a whole is lost, preserved only in fragments (Greek, in Eusebius and others; Latin, in Cassiodorus). A witness surviving at one remove, through an editor who admits to pruning, is not the strongest kind of witness, and the objection is right to say so.

Adversarial assessment

What holds. The Adumbrationes are a Latin, expurgated, secondhand text; Carol’s caution about relying on them “absolutely” is sound source-criticism, and the Hypotyposeis is genuinely lost. On the narrow point that this particular Latin fragment is an edited translation, the objection states a fact.

Where it fails. Three things sink it as an argument against Clement’s witness.

First, it is aimed at the wrong text. The objector attaches the caveat to Stromateis Book VII, chapter 16 — but Book VII of the Stromateis survives in Greek, and every English rendering of the VII.16 passage (Ante-Nicene Fathers and the rest) is made from that Greek, not from any Latin of Cassiodorus.³ Cassiodorus never adapted the Stromateis; his Latin covers only the notes on the Catholic Epistles. So “the Greek original is lost, and the Latin version dates to the 6th century” is simply untrue of the passage cited. Carol’s caveat about the Latin Adumbrationes has been transferred onto a Greek text it does not touch — a conflation of two different works of Clement, one Latin-surviving and one Greek-surviving.

Second, even where the caveat does apply, it cuts the other way. Grant that Clement’s brethren-note on Jude reaches us only through Cassiodorus’ expurgated Latin. That note is Epiphanian — the brethren are Joseph’s sons by a former marriage, not Mary’s children. As Lightfoot observed, a Latin churchman like Cassiodorus “would naturally hold the Hieronymian hypothesis” (the brethren as cousins), so the statement about Jude “must be Clement’s own and cannot have been inserted by Cassiodorus.” An editor purging error to fit his own view would have struck such a line, not added it; that it survived his pruning is evidence it stood firmly in Clement, not evidence of tampering.⁴ The expurgation-caveat, pressed honestly, strengthens the note’s claim to be Clement’s.

Third, the Greek passage says exactly what the objector concedes — and more. Stromateis VII.16 attests that Mary’s virginity was held to be unimpaired through the birth (the virginitas in partu): “many even down to our own time regard Mary… as having been in the puerperal state, although she was not. For some say that, after she brought forth, she was found, when examined, to be a virgin”; and “‘She brought forth and yet brought not forth,’ says the Scripture.”⁵ This is a Greek, firsthand report that the belief was current and reverently received “down to our own time” — the very early currency the objector needs to deny. It is not the whole doctrine: the in partu claim (intact childbearing) is distinct from the post partum claim (no marital relations or later children) that the brethren-question contests, and Clement carries the latter through the separate note on Jude. But the objection cannot both cite VII.16 as Clement’s apparent support and disown it as unreliable Latin; the passage it names is the one piece of his witness that is neither Latin nor lost.

Assessment

Weak. The textual-reliability objection does not hold as leveled. Its one sound element — that Clement’s Adumbrationes survive only in Cassiodorus’ edited Latin, so cannot be trusted “absolutely” — is a caution about a text the objector is not actually quoting: he cites the Greek Stromateis VII.16 while borrowing a caveat that Carol aimed at the Latin notes on the epistles, conflating two distinct works. Where the caveat does bite (the Jude note), it favors authenticity rather than undermining it, since an expurgating Hieronymian editor would have deleted Clement’s Epiphanian line, not planted it. And the Greek passage the objection names is itself firsthand evidence that Mary’s virginity in the birth was believed and honored “down to our own time” — corroborating, not dissolving, the doctrine’s early currency. One honest qualification keeps the verdict off unsupported: the Hypotyposeis is genuinely lost, and a witness surviving through an admitted editor is worth weighing with care — but that care, applied correctly, leaves Clement’s witness standing. Whether Clement settles the brethren-question is a separate matter, weighed on his detail page; the charge that a bad text has smuggled his support in is weak.

Sources