HistoricalChristian.Faith

Was “Theotokos” in use before the fourth century?

Related arguments: the origins of the title — see also Did “Theotokos” borrow the pagan title of Isis?

Type: historical-attribution argument (off-timeline) · Assessment: weak — that a specific pre-300 literary use of “Theotokos” can be pinned to Origen, Hippolytus, Dionysius of Alexandria, Pierius, or Peter of Alexandria (so that Alexander of Alexandria was not the first) does not hold: every one of the candidates rests on a text that is spurious, interpolated, or reported by an unreliable transmitter, and the earliest undisputed literary attestation remains Alexander’s anti-Arian encyclical of the 320s. A weaker, diffuse claim — that the word had some currency (probably Egyptian and liturgical) before the fourth century — is plausible and cannot be ruled out, but it is an inference from the term’s rapid diffusion, not a documented literary use.

The claim

The charge that Alexander was not the first to say “God-bearer” is itself ancient. When the Nestorian controversy broke, its defenders answered by marshalling patristic precedent. Nestorius’s own presbyter Anastasius had preached against the word — in the historian Socrates’s report, “Let no one call Mary Theotocos: for Mary was but a woman; and it is impossible that God should be born of a woman.”¹ Socrates rebuts the novelty by naming earlier users, and gives the citation on which the whole pre-fourth-century case turns:

Origen also in the first volume of his Commentaries on the apostle’s epistle to the Romans, gives an ample exposition of the sense in which the term Theotocos is used.

— Socrates Scholasticus, Ecclesiastical History VII.32.² On the strength of that sentence, and a handful of like reports, the title is routinely traced to the third century — to Origen (d. c. 254), Hippolytus (d. c. 235), Dionysius of Alexandria (d. c. 264), and Pierius of Alexandria (d. after 309) — with a fifth candidate, Peter of Alexandria (d. 311), resting not on Socrates but on a fragment of his own Paschal homily as it comes down to us. The interpretation this argument weighs is whether any of those attributions will bear a firm, documented pre-fourth-century literary use of the word. Taken one at a time, none does.

The attributions, weighed

Origen — the citation fails against the surviving text. The Commentary on Romans that Socrates points to survives, mostly in Rufinus’s Latin, with Greek catena fragments and a portion of the 1941 Tura papyrus — and at the place Socrates indicates, Book 1, it does not use the word. The modern critical translator notes plainly that the alleged quotation “does not correspond to” the extant commentary.³ The other “Origen” instances are catena extracts compiled no earlier than the sixth century, of doubtful authorship, assembled in the very era when using the word was a token of orthodox loyalty — one fragment cited as Origen’s proves in fact to be Eusebius. Origen may well have thought in these terms, given his high Mariology and Alexandrian setting; but that is an inference, and the specific attestation Socrates offers is not there to be checked.

Hippolytus — a spurious work. The word appears in a discourse On the End of the World (De consummatione mundi) circulated under Hippolytus’s name, but that piece is pseudo-Hippolytus, classed among the spuria and a much later compilation — it yields no third-century evidence.⁴

Dionysius of Alexandria — a forgery, and the clearest-cut case. The word occurs in a Letter to Paul of Samosata ascribed to Dionysius, but the letter has been recognized as spurious since about 1800. It refers to fathers who called the Son homoousios — the vocabulary of Nicaea (325), decades after Dionysius died — and Eduard Schwartz’s study (1927) exposed it as a fabrication, most likely an Apollinarist forgery of around 400, its clumsy author nothing like the celebrated stylist Dionysius was.⁵

Pierius — an unreliable transmitter reporting only a title. The one genuine-looking early trace is a report, preserved in fragments and generally credited to the church historian Philip of Side, that Pierius wrote a work On the Theotokos. Of the four traced to the third century, this is the strongest — the word stands in the surviving extract — but Philip of Side is a notoriously unreliable compiler, and even at best the notice attests a book title, not its contents, in an age when titles were freely “improved” by later copyists. Pierius, moreover, died after 309, so even a sound attestation would barely be pre-fourth-century.⁶

Peter of Alexandria — a Paschal homily reworked in Nicene idiom. The fifth candidate is the sturdiest as to the man — Peter was bishop of Alexandria and a martyr of the Diocletian persecution, and the word does stand in the surviving text — but it fails on the vocabulary. It occurs in a fragment of his treatise on the date of the Passover, where Christ is said to have been “born according to the flesh of our holy and glorious lady, Mother of God, and Ever-Virgin, and, of a truth, of Mary the Mother of God.” Peter died in 311, so even a genuine word from him would sit at the very edge of the fourth century, not before it. The decisive problem, though, is that the sentence carrying it is anachronistic: it calls the Son “co-eternal with the Father and the Holy Spirit, and of the same substance with them” — the homoousios of Nicaea (325), which Peter did not live to see — and wraps the doubled “Mother of God” in “Ever-Virgin” and a fully two-natures christology (Christ “of the same substance” with men “according to His human nature”). That is the idiom of Nicaea and of the fifth-century christological settlement, not of a bishop dead in 311; the passage as transmitted — in the paschal prolegomena to the Chronicon Paschale and Petavius’s Uranologion — has plainly been reworked in the later language, and modern editors print the whole Marian-and-consubstantial cluster among the Fragments (Pseudo-Peter).⁷ It is the same tell as ps.-Dionysius: a Nicene word in a pre-Nicene mouth. So Peter, too, does not unseat Alexander.

The earliest undisputed use, and why the disputes cluster before it

Strip away the contested attributions and the secure terminus is Alexander of Alexandria’s encyclical against Arius (the 320s), where the word does christological work — that Christ “in very deed, and not in appearance merely, carried a body, of Mary Mother of God.”⁸ It is not an accident that every disputed instance lies just before this line. Once the word became a badge of orthodoxy in the Arian and then the Nestorian crises, there was a standing motive to find it in the revered earlier fathers — which is exactly what forgers (ps.-Dionysius), pseudonymous compilers (ps.-Hippolytus), and catena-editors supplied. The pattern in the evidence is itself a caution: the attributions grow more, not less, doubtful the closer they sit to the moment the term became worth claiming.

The steelman: some prior currency, short of a literary text

The failure of each named attribution does not prove the word was coined ex nihilo in 320. Two considerations pull the other way. First, diffusion: by the 320s the term appears across Alexandrian writers apparently without needing introduction or defence, which is hard to explain unless it already had some devotional or liturgical life. That is a reasonable inference, but an inference — it locates a probable prehistory without producing a datable text. Second, the liturgical trace: the Sub Tuum Praesidium, an Egyptian papyrus prayer that invokes Mary as Theotokos, is on its oldest proposed dating a third-century witness — but that dating is itself sharply contested, palaeographers ranging from the third century to the eighth, so it cannot settle the question either way.⁹ The honest maximal claim is therefore modest: the word plausibly had some currency before the fourth century, most likely in Egyptian prayer, but no surviving pre-300 literary use can be securely attested.

Assessment

Weak as to the strong form — that a specific pre-fourth-century literary author (Origen, Hippolytus, Dionysius, Pierius, Peter of Alexandria) demonstrably used “Theotokos,” unseating Alexander as the first. Each candidate collapses on inspection: a citation that fails against Origen’s surviving text, a spurious “Hippolytus,” a forged “Dionysius,” for Pierius only a book-title reported by an unreliable hand, and, for Peter, a Paschal homily whose “Mother of God” comes wrapped in the homoousios of Nicaea he did not live to see. Plausible, by contrast, is the weaker and vaguer claim that the word had some pre-fourth-century currency — carried by its rapid, unremarked diffusion in the 320s and perhaps by early liturgy such as the Sub Tuum — provided one does not dress that inference up as a documented attestation. The upshot for the timeline is unchanged: Alexander of Alexandria remains the earliest certain literary witness, and the pre-history behind him is real but undocumented. For the word’s course once it was secure, see the Theotokos timeline; for the companion origins question, whether the title was borrowed from a pagan goddess, see Did “Theotokos” borrow the pagan title of Isis?

Sources