HistoricalChristian.Faith

Is the second-century stepbrother tradition genuine historical memory or a construct protecting Mary’s virginity?

Type: historical-value argument (off-timeline) · Assessment: contested — the Epiphanian reading (the brethren are Joseph’s sons by a former wife) is the only solution positively attested in Christian tradition before the late fourth century, surfacing around the middle of the second century in the Protevangelium of James, the Gospel of Peter, and the Infancy Gospel of Thomas; that early and unrivalled attestation is real weight. But the same tradition clusters, in early-second-century Syrian works, with two other non-canonical beliefs about Jesus’ origins — Mary’s Davidic descent and the miraculous (painless) birth — all three tending to glorify her virginity, which raises the live possibility that the stepbrother claim is a piece of the same Marian development rather than an independent memory. On the evidence the two readings cannot be separated: the second-century tradition could preserve an accurate historical memory, or it could be a legendary construct in favour of the perpetual virginity. Richard Bauckham, himself an Epiphanian, weighs both and concludes “we cannot be sure” · argued from Richard Bauckham, with the underlying data from J. B. Lightfoot and the second-century apocrypha

Related arguments: the Epiphanian (stepbrother) reading — see also The Epiphanian reading on its merits · Does Joseph’s reputed fatherhood license the brethren-title?

The claim

The wiki’s timeline shows the Epiphanian reading — the “brethren of the Lord” are Joseph’s sons by an earlier marriage, no blood-kin of Mary — as the earliest attested solution to the brethren-question, running from the Protevangelium of James and the Gospel of Peter (c. 150) through Clement, Origen, and Eusebius. This argument asks the sharper question that the timeline’s bare chronology cannot settle: does that early attestation amount to evidence that the brethren really were Joseph’s sons by a former wife — a memory of the family’s actual make-up — or only to evidence that the belief was in circulation early, as one strand of a growing devotion to Mary’s virginity?

The strongest form of the pro-memory case is Richard Bauckham’s, and it is worth taking from a scholar who himself favours the Epiphanian reading. He grants at the outset that direct proof is thin on both sides: the Gospel references to the brethren “are perfectly consistent with the view that they were his step brothers,” while “positive evidence that they were the sons of Mary, as well as of Joseph, is lacking” — so the Helvidian reading rests on it being “the most natural assumption,” not on a proof-text. In that vacuum the one serious datum for the Epiphanian side is its antiquity: it “is the only view positively attested in Christian tradition before Tertullian… or before the late fourth century,” appearing “around the middle of the second century in the Protevangelium of James (9:2; 17:1–2; 18:1), in the Gospel of Peter… and in the Infancy Gospel of Thomas (16),” and passing thence to Clement, Origen, and later writers.¹

Why the antiquity carries real weight

Two features lift the second-century attestation above a mere late guess. First, it is early enough that good tradition was still available: Bauckham insists that “good historical tradition certainly was still available in the early second century, even if mixed with much legendary material,” so that a claim of this date about the composition of Jesus’ household cannot be dismissed out of hand simply for the company it keeps. Second, the earliest witnesses take the stepbrother relationship for granted rather than arguing it: in the Protevangelium of James Joseph is already a widower with sons before he is betrothed to Mary, and in the Infancy Gospel of Thomas (16) the point is assumed “as though the readers are already quite aware of it” — which suggests the belief is older than the books that carry it, not their invention.¹ That the Protevangelium itself is, as J. B. Lightfoot judged, “a purely fictitious… narrative” does not by itself defeat the point: a legendary frame can still transmit an older report it did not create.²

(The double-edge must be kept in view. The uniform Epiphanian shape of this whole apocryphal stream — every downstream infancy narrative makes Joseph an aged widower with sons of his own, none reaching for a cousin-relationship — does firmer, one-directional work against the rival Hieronymian cousin-theory: as Lightfoot pressed, “the fabrications would scarcely have taken this form, if the Hieronymian view… had been received or even known when they were written” (weighed at “Was the cousin theory a novelty without earlier witness?”). But that finding is separable from the one this page turns on: showing the cousin-reading was unknown at c. 150 does not show the stepbrother account is history — the apocrypha are fabrications either way, and their unanimity is that of one dependent tradition, not independent memory.)

The Syrian cluster — why it may be a construct

Against this stands Bauckham’s own most telling observation. In the Protevangelium of James the stepbrother idea does not stand alone: it is “combined with two other beliefs concerned with the human origins of Jesus, which are not found in the New Testament: the Davidic descent of Mary (10:1) and the miraculous birth of Jesus (19:2–3), whereby Mary’s virginal condition was unimpaired by childbearing (the virginitas in partu).” And “clearly all three notions are connected with the virginity of Mary” — regarding the brothers as stepbrothers, tracing Jesus’ Davidic descent through Mary, and telling of a miraculous birth all “ensure that Mary’s virginity was preserved.”¹

None of the three, Bauckham argues, originates with the Protevangelium; all were “probably already traditional” when it took them up — Mary’s Davidic descent surfaces in credal formulations in Ignatius (Eph. 18:2; Trall. 9:1; Smyrn. 1:1) and elsewhere, and the miraculous birth in a simpler, earlier form in the Ascension of Isaiah (11) and the Odes of Solomon (19). Comparing every second-century work that carries either the miraculous birth or the stepbrother view yields a cluster of six — Ignatius, the Ascension of Isaiah, the Odes of Solomon, the Gospel of Peter, the Infancy Gospel of Thomas, and the Protevangelium of James — and “all six of these works are either certainly or quite probably of Syrian origin.” The result is “a complex of traditions which dates from the very beginning of the second century in Syria.”¹

This is what cuts two ways, and Bauckham lays out both edges without deciding. On the one hand, “the tradition about the brothers of Jesus may… seem condemned as unhistorical by association with the other two traditions: it looks like part of a doctrinal or hagiographical development of the theme of the virginity of Mary.” On the other, “it is possible that the brothers of Jesus were correctly remembered not to have been sons of Mary and that this made possible the development of the idea of Mary’s perpetual virginity as a result.” A true memory that the brethren were not Mary’s could be the seed of the whole development, rather than one of its products.¹

The decisive test: is the tradition Palestinian?

Bauckham identifies what would settle the matter: “This view of the brothers of Jesus would have more probable historical value if it could be plausibly traced to Palestinian Jewish Christian tradition” — a memory rooted in the Lord’s own homeland and kindred would outweigh a Syrian devotional cluster. “None of these six Syrian works can be very closely associated with such tradition,” so he tests several further candidates for a Palestinian root. Three, taken as direct witnesses, come up short:

• Sedulius Scotus’ “Gospel according to the Hebrews.” The ninth-century Irish scholar, in his Commentary on Matthew, quotes an infancy scene at the Bethlehem cave in which Joseph, watching the magi approach, turns and speaks to his son Simon — clearly one of the brethren — and ascribes it to a “Gospel according to the Hebrews”; a second Irish witness, in a twelfth-century manuscript, cites the same passage under the same title, though perhaps only copying Sedulius. If that title were exact, this would be a Jewish-Christian (Nazarene) notice of Joseph’s sons present at the Nativity — a Palestinian root for the Epiphanian reading. But the passage is in fact an extract from chapter 89 of the Latin Infancy Gospel (the compilation now catalogued as the Book of the Nativity of the Savior) published by M. R. James — a probably Carolingian conflation of the Protevangelium of James, Pseudo-Matthew, and an otherwise-unknown ancient source, one in which “Joseph’s son Symeon is prominent.” That unknown source “could well be as early as the second century” and may be the lost Gospel of the Nazarenes; but Sedulius may instead have loosely called the whole compilation by that title, “sometimes used inaccurately in the medieval period.” So the datum “deserves further study,” and “in the present state of investigation it would be hazardous to use this material as evidence of early Jewish Christian tradition” — the “Simon son of Joseph” detail cannot presently be dated firmly to the second century, and it is set aside here as too uncertain to weigh.³
• The Second Apocalypse of James (Nag Hammadi). It has Jesus call James “my brother” and his mother explain that “both of you were nourished with the same milk… he is your [step-brother].” But the “quite unparalleled name Theudas” it gives James’ father shows the author “either depended on a very eccentric tradition… or had no traditional information… at all”; the step-brother gloss is probably his own Gnostic attempt to explain the address, not an independent memory.
• Hegesippus. The one witness who really would give “our best chance of access to Palestinian Jewish Christian tradition” — but “from the fragments of Hegesippus extant in Eusebius it is not possible to be sure whether he held the Helvidian or the Epiphanian view.” Neither “brother… according to the flesh” nor “called” (of Jude) settles it. The most that can be said is that had Hegesippus plainly taught the Helvidian view, later writers who leaned on him — above all Eusebius, who “takes the Epiphanian view for granted” — would be expected to follow, and none do; but that is an argument from silence, not a positive attribution.⁴

(That Hegesippus tells against the rival Hieronymian cousin-theory is a separate and firmer finding — see “Does Hegesippus tell against the cousin-theory?” — but on the Epiphanian-vs-Helvidian choice his fragments are silent.)

There is, however, one route by which the specific Epiphanian claim — James as Joseph’s son by a former wife — may reach back to Hegesippus after all, even though the extant fragments do not state it outright. Besides his fullest account in Panarion 78, Epiphanius states the same point more briefly in Panarion 29, his chapter on the Nazoraeans: “this James was Joseph’s son by Joseph’s [first] wife, not by Mary… he was of Davidic descent because of being Joseph’s son, and… he was born a nazirite—for he was Joseph’s first-born, and (thus) consecrated.” Bauckham judges this passage “much more plausibly dependent on Hegesippus… than on apocryphal infancy literature” — and its setting bears him out: the “firstborn of Joseph” note stands inside an unmistakably Hegesippan block (James the first bishop of Jerusalem, his priestly office, his once-a-year entry into the Holy of Holies), which Epiphanius expressly credits to “Eusebius, Clement and others” — the same Palestinian channel Eusebius drew on at H.E. 2:1 and 2:23, not the Syrian infancy gospels. If that source-judgment is right, the concrete Epiphanian datum does have a Palestinian pedigree behind it, which is exactly what the test asks for.⁵

But this strengthens the memory reading without settling it, and the discipline of the three failed candidates still governs. It is a plausibility about Epiphanius’s source, not a proof: Hegesippus’s own words on the point are lost, so one still cannot show from the surviving fragments that he taught the Epiphanian rather than the Helvidian view (that was the burden of the third candidate). What the Panarion 29 datum adds is that the earliest identifiable form of the specific claim more likely travelled through the one genuinely Palestinian collector than through the Syrian devotional cluster — enough to lift the Epiphanian view to “a better claim to serious consideration than is often nowadays allowed,” not enough to turn plausibility into demonstration. Bauckham, having weighed it, still lands on “we cannot be sure.”

A name in the roll that will not bear weight: Hippolytus

Bauckham’s sketch of the tradition’s spread names one more early authority alongside those on the timeline: the stepbrother reading, he writes, was “probably through the influence of the first two of these works” — the Protevangelium of James and the Gospel of Peter — “then accepted by Clement of Alexandria, Hippolytus, Origen and many subsequent writers.”⁶ Clement, Origen, and Eusebius already stand on the timeline; Hippolytus of Rome (d. c. 235) does not — and on inspection the reason is that no genuine statement of the Epiphanian view survives from him. The point is worth making because a bare roll of names can read as cumulative attestation when one of the names is doubtful.

What travels under Hippolytus’ name are two short apostle-catalogues, On the Twelve Apostles and On the Seventy Apostles, which the Ante-Nicene Fathers editors already flag as “probably spurious, but yet antique.” They do carry an Epiphanian tendency, but only of a negative kind: they list “James the Lord’s brother [ἀδελφόθεος], bishop of Jerusalem” among the Seventy while separately placing “James the son of Alphæus… stoned to death” among the Twelve — that is, they decline to identify the Lord’s brother with the apostle, the Eastern instinct that underlies the Epiphanian reading (and cuts against the rival Hieronymian cousin-theory). But refusing that identification is not the same as asserting that the brethren were Joseph’s sons by a former wife, and these lists nowhere say so.⁶

The one text that does give a full Epiphanian account under his name reaches us only at third hand and very late. The fourteenth-century Byzantine historian Nicephorus Callistus “cites as from Hippolytus of Portus an elaborate account of our Lord’s brethren following the Epiphanian view.” Lightfoot, running the attribution down, judged the two apostle-catalogues spurious outright — the first “a blundering treatise… certainly not written by the bishop of Portus,” the second “manifestly spurious” — and traced Nicephorus’ elaborate account “either [to] Hippolytus the Theban” (a compiler of perhaps the tenth century) “or [to] some forged writings which bore the name of the older Hippolytus.”⁷ On either reading the account is centuries too late to be the Roman Hippolytus’.

None of this touches the core question — the second-century apocrypha, not the third-century roll, carry the antiquity that gives the Epiphanian reading its real claim — but it does trim the roll: Hippolytus adds no independent weight to it, and a survey that leans on the length of the patristic list should not count him among the sound entries.

Assessment

Contested. The question is genuinely disputed, with a strong case on each side and no decider in the evidence — the defining shape of a contested rating. For the memory reading: the stepbrother tradition is the earliest attested solution by a wide margin, it is early enough that reliable tradition survived, and its earliest carriers assume rather than argue it, as if drawing on something older than themselves. For the construct reading: it travels, in early-second-century Syria, in a tight cluster with two other extra-biblical beliefs whose common tendency is to magnify Mary’s virginity, so it may be one more product of that devotion rather than a datum behind it; and the test that could break the tie — a plausible Palestinian Jewish-Christian pedigree — is met only in part: the three candidates that would trace the tradition directly to Palestine fail, while the one indirect route that holds (Epiphanius’s “firstborn son of Joseph” more plausibly drawn from Hegesippus than from the infancy gospels) yields a plausibility about a source, not a proof. Bauckham, weighing it as a friend of the Epiphanian reading, states the impasse exactly: “The second-century tradition could preserve an accurate historical memory. On the other hand, it could be a legendary construct in favour of the perpetual virginity of Mary. We cannot be sure.”¹

Two clarifications keep the finding honest. First, an attestation is not thereby a proof: the timeline’s early Epiphanian witnesses establish beyond doubt that the belief circulated by c. 150, but the belief’s early date does not settle whether it records the family’s real make-up or codifies a growing piety — the two are, on this evidence, empirically indistinguishable. Second, contested here does not mean weak: the tradition is not to be judged unhistorical merely because it serves a doctrine (a legend can carry a memory, and a memory can seed a devotion), and its antiquity gives the Epiphanian view “a better claim to serious consideration than is often nowadays allowed.” What the evidence withholds is certainty either way — which is why the choice between the Epiphanian and Helvidian readings, as Bauckham notes, finally “make[s] no difference” to the historical role of the brethren in the early church.¹

Sources