HistoricalChristian.Faith

The Biblical Canon

How the Church came to recognize which books are Scripture — the Old Testament (and the disputed deuterocanon) and the New — is a story of gradual reception, not a single decree. The earliest lists are partial and regional: the Muratorian fragment (late 2nd c.), Origen, Eusebius's sorting of the books into "acknowledged," "disputed," and "spurious" (Church History 3.25), and Athanasius's Festal Letter 39 of 367 — the first catalogue to match the present twenty-seven-book New Testament — the Roman synod under Damasus (382), to which a matching list is ascribed though only on contested grounds, and the African councils of Hippo (393) and Carthage (397), with the East long hesitant over the Apocalypse and the West and East still divided over the deuterocanon at the Reformation. Around this real history a popular legend grew: that the canon was decided in one stroke at the Council of Nicaea in 325 — voted into being, even sorted by a miracle — the claim this page first takes up below.

Related doctrines: Scripture's text & canon — see also the Old Testament canon · the Septuagint's origin · rabbinic corruption of the text

Timeline

The chronological witnesses to the canon's reception, earliest first. The cross-cutting legends and historical theses they bear on are weighed below under Arguments.

Marcion of Sinope (c. 144)

Before the church published any list of its own, a rejected teacher published his. Around 144, breaking with the church of Rome, Marcion of Sinope issued the earliest fixed, closed collection of Christian scriptures on record: a single Gospel — a shortened Luke, stripped of the nativity and of every tie to the Creator-God, and carrying no evangelist's name — together with ten edited letters of Paul (his Apostolikon), the whole Old Testament, the other three Gospels, Acts, and the Pastoral epistles cut away. His was a canon defined as much by deletion as selection: Tertullian charged that he "used the knife, not the pen," and Irenaeus that he "mutilates the Gospel which is according to Luke" and "dismembered the Epistles of Paul." On the classic thesis of Harnack and von Campenhausen this truncated list provoked the catholic church to fix and publish a broader one — a fourfold Gospel against his one, a full Paul against his ten, the Old Testament retained — so Marcion answers the "why a list at all?" question the timeline otherwise opens after (how decisive the provocation was is debated). Everything known of him survives only through the opponents who wrote to refute him.¹

Justin Martyr (c. 155)

A decade after Marcion, the earliest witness on the catholic side attests not a list but a function: at Rome around 155, Justin Martyr already treats the Gospels as Scripture. He calls them the apostles' "memoirs composed by them, which are called Gospels," citing them for the words of institution (First Apology 66), and reports that in the Sunday assembly "the memoirs of the apostles or the writings of the prophets are read, as long as time permits" (67) — the apostolic records read aloud beside the prophets, the practical mark of a book held to be Scripture. The boundary is not yet drawn, though: Justin never enumerates the four, never names the evangelists, and his Gospel citations sometimes harmonize the texts — so he marks the moment the Gospels function as Scripture, a generation before Irenaeus fixes their number and the Muratorian list first counts the books.²

The Muratorian Fragment (c. 170)

A generation after Justin, the boundary is first drawn as a list. The Muratorian Fragment — an eighty-five-line Latin catalogue, mutilated at both ends, the oldest surviving enumeration of New-Testament books — already knows most of the future twenty-seven: the four Gospels (Matthew and Mark assumed in the broken opening, since it numbers Luke "the third book" and John "the fourth"), Acts, the thirteen letters of Paul ("to seven churches"), Jude, two epistles of John, and the Apocalypse of John, some twenty-two of the eventual books. But its edges are still soft in both directions: it omits James, both epistles of Peter, Hebrews, and 3 John; it counts the Old-Testament Wisdom of Solomon among the received books; it accepts the Apocalypse of Peter while noting "some of us are not willing that it be read in church"; and it allows the Shepherd of Hermas to be read privately but not publicly "among the Apostles, for it is after [their] time" — apostolic age the criterion. It closes by rejecting forgeries: two epistles "forged in Paul's name" for the Marcionites (to the Laodiceans and Alexandrians) and the books of Valentinus, Basilides, and the rest, "for it is not fitting that gall be mixed with honey." So the earliest list is a canon firmly taking shape yet not closed — received-and-disputed at once, the fuzzy boundary the later witnesses harden. (A minority of scholars, following Sundberg and Hahneman, redate the fragment to the fourth-century East; the internal notice that the Shepherd was written "very recently" under bishop Pius of Rome, c. 140–154, anchors the majority second-century date.)³

Irenaeus of Lyons (c. 180)

A decade after the Muratorian list, Irenaeus hardens the boundary from the center out. He is the first to argue that the Gospels are exactly four — "neither more nor fewer in number than they are," matched to the four zones of the world and the four faces of the cherubim, the fourfold "Tetramorph" (Against Heresies 3.11.8) — and he narrates each evangelist's apostolic origin: Matthew among the Hebrews, Mark "the interpreter of Peter," Luke "the companion of Paul," John at Ephesus (3.1.1). Across his five books he quotes as authoritative Scripture some twenty-one of the eventual twenty-seven New-Testament books, the same shorter epistles (Philemon, 3 John, Jude) still uncited.⁴ Yet his canon stays soft at the edges: he treats 1 Clement as a "most powerful" apostolic authority (3.3.3) and cites the Shepherd of Hermas as "Scripture" (4.20.2, though the word may mean simply "writing") — both later excluded — while on the far side he files the tale of the boy Jesus and his alphabet-teacher among the Marcosians' forged "apocryphal and spurious writings" (1.20.1) and brands the Gnostic Gospel of Judas a "fictitious history" (1.31.1). He thus attests both faces of one boundary at once: the Gospel core already closed, the forgeries pushed out, a middle ring (Clement, Hermas) still inside the line — the gradual, edges-last reception the later witnesses harden.⁵

Clement of Alexandria (c. 200)

Between Irenaeus and Origen stands Origen's own teacher, Clement of Alexandria — the clearest witness that at Alexandria around 200 the category "Scripture" still had porous edges. In the Stromata he quotes the Epistle of Barnabas as "the Apostle Barnabas," the Shepherd of Hermas as words spoken "divinely" by revelation, the letter of Clement of Rome as "the Apostle Clement," and the Preaching of Peter for the risen Lord's own commission — all books later left outside the canon — and draws freely on the Traditions of Matthias and the Gospel according to the Egyptians. Eusebius, reading Clement's now-lost Hypotyposeis, reports abridged commentary on "all canonical Scripture, not omitting the disputed books… Barnabas and the so-called Apocalypse of Peter." Yet Clement already privileges "the four Gospels that have been handed down to us" (Stromata 3.13), rejecting a saying precisely because it stands outside them — the fourfold core fixed, the outer ring still soft. He is the taproot of this page's fuzzy-edges thesis.⁶

Tertullian (c. 207)

Between Irenaeus and Origen, and now in the Latin West, Tertullian carries the same picture forward. Against Marcion’s single, unattributed, mutilated Gospel of Luke he defends the fixed fourfold Gospel on the Irenaean scheme — “the evangelical Testament has apostles for its authors” (Matthew, John) and “apostolic men” who wrote with apostolic authority (Luke the companion of Paul; Mark “whose interpreter Mark was” of Peter), Marcion convicted by the lateness of his one edited book against the older four (Against Marcion IV.2, 5). His criterion of authenticity is reception and provenance: go to “the apostolic churches … in which their own authentic writings are read” — Corinth, Philippi, Ephesus, Rome (Prescription 36). Yet his edges stay soft in both directions: he throws the Shepherd of Hermas out as “apocryphal and false,” hardening what Irenaeus had cited as Scripture, while still defending the book of Enoch as inspired (Apparel of Women I.3) and ascribing Hebrews to Barnabas (On Modesty 20) — a closed Gospel core with a boundary still forming at the rim.⁷

Origen of Alexandria (c. 248)

Two generations on, Origen attests both faces of the boundary at once. On the inclusion side he is among the earliest to enumerate the whole New Testament from within: casting the apostles as priests sounding the trumpets of Jericho, he runs out the four Gospels, the Acts, "two" epistles of Peter, James and Jude, the epistles of John "[and Revelation]," and Paul's "fourteen" letters (the thirteen plus Hebrews) — the present twenty-seven, soft only at the bracketed Apocalypse that some manuscripts of the homily omit (Homilies on Joshua 7.1), with a terser version naming the evangelists and apostles as the diggers of "the wells of the New Testament" (Homilies on Genesis 13.2) and the one doubt settled where he affirms that John "wrote also the Apocalypse" (Commentary on John 5.3). On the exclusion side he is the first to name the apostolic-pseudonym and infancy gospels and set them outside the four: reading Luke's "many have taken in hand to draw up a narrative" as a veiled criticism, he ranks the rival gospels as mere "attempts": "those who wrote the Gospel according to the Egyptians and the Gospel according to the Twelve 'attempted' to write … for there is also the Gospel according to Thomas and the Gospel according to Matthias and several others. These are attempts; the church of God has approved only the four" (Homilies on Luke 1).⁸ And in the Commentary on Matthew he can cite "a tradition in the Gospel according to Peter … or The Book of James" — the Protoevangelium — for the view that the Lord's brethren were Joseph's sons by an earlier marriage, reporting only what "some say" and never granting the book the standing of Scripture (X.17, on Matt 13:55).⁹ His boundary is drawn by reception and authority — the church has "approved only the four" — not, in these passages, by a charge of heresy; the sharper "forgeries of heretics" language waits for Eusebius. And because his list of titles is bare, it cannot by itself fix which "Gospel of Thomas" he meant, the Gnostic sayings gospel or the Infancy text (see the argument below).

Eusebius of Caesarea (c. 324)

On the eve of Nicaea, Eusebius draws the most deliberate early sorting of the books — accepted, disputed-but-widely-recognized, and rejected — and then sets the apostolic-pseudonym gospels of the heretics a step below even the rejected. It is here that the backbone names of the New-Testament apocrypha stand together: among "those that are cited by the heretics under the name of the apostles" are "the Gospels of Peter, of Thomas, of Matthias, or of any others besides them, and the Acts of Andrew and John," books that "no one belonging to the succession of ecclesiastical writers has deemed worthy of mention" and whose style and matter "clearly show themselves to be the fictions of heretics," "to be cast aside as absurd and impious" (Church History 3.25.6–7). His is the sharpest pre-Athanasian language of exclusion — and a contemporary's testimony cutting against the later legend: present at the council of 325, he still records the canon itself as disputed, not decreed, which is why this same chapter recurs among the Nicene-canon arguments below.¹⁰

Cyril of Jerusalem (c. 350)

A generation after Eusebius, and within his own province, Cyril of Jerusalem turns the sorting into a pastoral rule for the newly baptized. Catechizing the candidates, he tells them to learn the canon "from the Church," read "the twenty-two books of the Old Testament" and the recognized New, and "read none of the apocryphal writings … for why dost thou, who knowest not those which are acknowledged among all, trouble thyself in vain about those which are disputed?" (Catechetical Lecture 4.33). His New-Testament list runs the four Gospels — "for the rest have false titles and are mischievous" — Acts, the seven Catholic Epistles, and "the fourteen Epistles of Paul" (Hebrews folded into the Pauline corpus, as Origen and Athanasius also count it): twenty-six of the twenty-seven books, conspicuously stopping short of the Apocalypse (the Eastern hesitancy over Revelation), and naming a single apocryphal example: a "Gospel according to Thomas" he ascribes to the Manichaeans (4.36). The boundary, as with Eusebius, is drawn by church reception and the register is repudiation, not destruction: the apocrypha are to be "put aside in a secondary rank," and "whatever books are not read in Churches, these read not even by thyself" (4.35–36).¹¹

The Council of Laodicea (c. 363)

A dozen years after Cyril, and from the same Eastern church, the omission he drew as one catechist's list becomes a synod's. A regional council at Laodicea in Phrygia rules that "only the Canonical Books… of the New and Old Testament" may be read in church (canon 59), and — in the much-disputed canon 60 — appends a catalogue whose New Testament is the present twenty-seven minus the Apocalypse: four Gospels, Acts, the seven Catholic Epistles, and "fourteen Epistles of Paul" (Hebrews folded in, as Cyril and Athanasius also count it), closing at Titus and Philemon without the Revelation of John.¹² It is the sharpest conciliar instance of the Eastern hesitancy over Revelation — a whole synod, not one bishop, drawing the New Testament at twenty-six books, four years before Athanasius would seal the list with the Apocalypse. Canon 60's own genuineness is doubted (Percival heads it "of most questionable genuineness," and it is absent from several early collections of the synod's acts), so it may be a somewhat later hand; but canon 59's boundary is secure, and even as received the list stops short of Revelation.¹³ (Its Old Testament — the Hebrew twenty-two, the deuterocanon absent — is treated on the Old Testament canon page.)

Athanasius of Alexandria (367)

A generation after Cyril, and from the same Eastern church, Athanasius draws the line that has held: his thirty-ninth Festal (Easter) Letter of 367 gives the first surviving catalogue whose New Testament is exactly the present twenty-seven — the four Gospels, Acts, the seven Catholic Epistles, "fourteen Epistles of Paul" (Hebrews folded in, as Origen and Cyril also count it), "and besides, the Revelation of John" — sealed with the closure-formula "let no man add to these, neither let him take ought from these" (Festal Letter 39.5–6). Where Cyril a generation earlier had stopped at twenty-six, omitting the Apocalypse, Athanasius closes the list with Revelation, settling on the inclusion side the one book the Greek churches longest doubted. Yet he frames the catalogue not as a decree but as his own undertaking, written "of necessity" and urged by "true brethren," of the books "included in the Canon, and handed down, and accredited as Divine" (39.3) — a bishop who had sat at Nicaea (325) still, forty-two years on, listing the canon himself because it had never been formally enacted, the contemporary's testimony against the legend that the council fixed it. Alongside the New Testament he sets a twenty-two-book Hebrew Old Testament and, in a middle rank, books "not indeed included in the Canon, but appointed by the Fathers to be read" — Wisdom, Sirach, Esther, Judith, Tobit, the Teaching of the Apostles, and the Shepherd (39.4, 39.7) — ranking the deuterocanon below the canon a generation before the African councils would list it among the Scriptures, the seed of the later Old-Testament dispute.¹⁴

The Council of Rome under Damasus (382)

Fifteen years after Athanasius, and now in the West, a canon-list matching the present twenty-seven-book New Testament is ascribed to a Roman synod under Pope Damasus I — but on contested grounds. A synod did meet at Rome in 382, in the aftermath of Constantinople (381) and occupied with the Antiochene schism, with Jerome newly arrived as Damasus's secretary; and the "Damasine list" credited to it runs out the full New Testament (Hebrews folded into "fourteen" of Paul) and the longer Septuagint Old Testament, the Western canon as the late fourth century received it. Yet that list survives in no fourth-century document: it is preserved only as a chapter of the Decretum Gelasianum, the sixth-century compilation this timeline meets again below as the rejection-lists — a work whose opening chapter quotes Augustine (416), so it cannot as a unit be a decree of 382. The attribution of its canon-chapters to Damasus is itself modern (Arévalo, 1794; C. H. Turner's "Roman Council under Damasus"), and Jerome — present at Rome, the age's foremost authority on the canon, and the very writer the list depends on — nowhere mentions a Roman conciliar canon, ranking Judith, Tobit, and the Maccabees below the canon in his own prefaces. So Rome (382) belongs on the timeline as the probable Western reception of the canon a decade before Hippo, not as the decree the popular claim makes of it; whether the council actually promulgated the list is weighed at the argument below (assessment: weak).¹⁵

The African councils: Hippo (393) and Carthage (397)

A generation after Athanasius, and now in the West, the boundary is for the first time drawn by a council. The North-African synods gathered around Augustine and Aurelius of Carthage — Hippo in 393, Carthage in 397, reaffirmed at Carthage in 419 — each enact the rule "that besides the Canonical Scriptures nothing be read in church under the name of divine Scripture," then list the books. The New Testament is exactly the present twenty-seven (Carthage still setting "one epistle … to the Hebrews" a half-step apart from Paul's "thirteen," the lingering Western doubt over its authorship that Athanasius's "fourteen of Paul" had dissolved), Revelation now securely in. But these are regional African synods, not ecumenical councils, and their canon is framed as a rule for what may be "read in church" — the conciliar ratification of a reception already settled, the Western echo of Athanasius's Eastern list, not the moment the canon was created; the note appended in 419, deferring the list to Boniface of Rome and "the Church beyond the sea" for confirmation, shows a boundary acknowledged, not legislated. On the Old-Testament side the lists place Tobit, Judith, the two books of Maccabees, and the Wisdom books flatly among the Scriptures, where Athanasius had ranked the deuterocanon a tier below — the seed of the later Old-Testament dispute. (This same African list, Hippo's canon 36, is the one a lone later codex would mislabel "of Nicaea"; see the argument below.)¹⁶

Jerome (c. 394)

Between the African councils and Innocent I stands the age's foremost biblical scholar. Jerome — secretary to Damasus, translator of the Vulgate that would carry the Latin New Testament for a thousand years — both transmits the twenty-seven-book canon and preserves the memory of its disputes. Revising the Gospels for Pope Damasus he sets them in the order "Matthew, Mark, Luke, John" (Vulgate preface, 383), and in Letter 53 to Paulinus he runs the whole New Testament book by book — the fourfold Gospel, Paul "to seven churches," the seven Catholic Epistles of James, Peter, John, and Jude, and the Apocalypse. Yet he records every doubt his sources carried: Hebrews denied to Paul in the West (Tertullian ascribing it to Barnabas), 2 Peter and James suspect, the two shorter Johannine epistles assigned to "John the presbyter," the Apocalypse doubted in the East — the old antilegomena. In the end he receives them all: "yet we receive them both," he tells Dardanus of Hebrews and Revelation, on "the authority of the ancient writers," not any council. The scholar who knew every dispute ratifies the full canon — and, present at Damasus's Rome in 382, nowhere cites a Roman or Nicene conciliar canon to do it, the silence that weighs against the Rome-382 and Nicaea-Judith claims.¹⁷

Innocent I of Rome (c. 405)

By the turn of the fifth century the Western boundary is drawn by a sitting bishop of Rome. Answering an inquiry from Exuperius of Toulouse, Innocent I appends a canon-list matching the present twenty-seven-book New Testament, then turns to the apocrypha: "the rest of the books, which appear under the name of Matthias … or under the name of Peter and John (which were written by a certain Leucius) … or under the name of Andrew … or under the name of Thomas, and whatever others there may be," he writes, "you should know they are not only to be rejected but also condemned" (Ep. 6, 405). It is the sharpest Western papal exclusion before the Gelasian Decree, naming the same apostolic-pseudonym gospels and acts Eusebius had bracketed and tagging them to their reputed forgers (the Petrine and Johannine acts to "a certain Leucius," the Andrean to two named philosophers). Yet his verb tells against the legend: the books are damnanda — condemned, repudiated — not gathered and burned; this is exclusion by formal repudiation, the documented register of the late rejection-lists. (The "apocrypha" he condemns are the pseudo-apostolic New-Testament writings, not the Old-Testament deuterocanon, which his list keeps.)¹⁸

The rejection-lists (c. 500–810)

By the sixth century the boundary is drawn not book-by-book but as a formal catalogue of rejection — a list of titles "not to be received." The Western Decretum Gelasianum (De libris recipiendis et non recipiendis, c. 500) is the one early list that names the infancy texts by name: among some sixty rejected apocrypha it registers "the book on the infancy of the saviour" (the boy-Jesus material of the Infancy Gospel of Thomas) and "the book of the nativity of the saviour and of Mary or the midwife" (the Protoevangelium of James), alongside — and kept distinct from — "the Gospel in the name of Thomas which the Manichaeans use," each tagged "apocryphal" and "to be avoided by catholics." The Greek catalogues run a parallel Eastern record but a narrower one: the List of Sixty Books (c. 650) and the Stichometry of Nicephorus (c. 810) reject a "Gospel of Thomas" and a "Circuit of Thomas" without ever naming the infancy gospels or the Protoevangelium — and the "Gospel of Thomas" they list, over-long by its own line-count, is most likely not the infancy book at all. Two things follow. The exclusion of the infancy gospels as such is, on the surviving lists, a Western datum; and the register on both sides is repudiation — books "to be avoided," "not received" — not the "banned and burned" of the legend. (The Decretum's separation of the Manichaean "Gospel of Thomas" from "the book on the infancy of the saviour" is itself evidence on the which-Thomas argument below.)¹⁹

Martin Luther (c. 1522)

A thousand years after the rejection-lists, the New Testament's edges are reopened one last time on this timeline. Translating the New Testament into German in 1522, Martin Luther read it as Eusebius had — a core of "true and certain chief books" and a disputed remainder — and printed the distinction into the volume. He translated and bound all twenty-seven books, but moved the four he most doubted, Hebrews, James, Jude, and Revelation, to the end and left them unnumbered in a table of contents that numbers the other twenty-three 1–23, prefacing the group with the note that these "have from ancient times had a different reputation." His prefaces give the reasons — James "an epistle of straw… for it has nothing of the nature of the gospel about it" and "not… the writing of an apostle"; Jude "an extract or copy of St. Peter's second epistle"; Revelation, in 1522, "neither apostolic nor prophetic" — the criterion being apostolicity and whether a book "preaches Christ." It is a canon within the canon, the ancient antilegomena revived as the layout of a page; but it is demotion, not deletion. Luther excised nothing: all four stood in his New Testament and in every Lutheran and Protestant Bible after it, and he softened his own verdicts — the "epistle of straw" line dropped from later editions, Revelation given a warm new preface in 1530. Whether the popular "Luther removed four New Testament books" follows from any of this is weighed at the argument below (assessment: weak).²⁰

Arguments

Sub-claims marshaled for (or against) a reading, which cut across the timeline rather than sitting at one date — each weighed adversarially on its own page, with an assessment of the interpretation it serves.

Related arguments — the Nicene-canon legend

The popular legend that the biblical canon was settled in one stroke at the Council of Nicaea (325) — its miraculous mechanism, and its New-Testament-apocrypha face. Three facets of one story, weighed together: the general thesis that Nicaea fixed the canon, the altar-ordeal by which the legend has the books sorted, and the twin claim that the excluded gospels were actively suppressed — banned, gathered up and burned.

Did the Council of Nicaea (325) fix the biblical canon? — the popular legend (assessment: unsupported)

The Nicaea-fixed-the-canon legend, in its popular modern form — the claim, bestselling in Dan Brown's The Da Vinci Code (2003) but Enlightenment-rationalist in origin (Paine, Voltaire), that the contents of the Bible were decided at Nicaea: a canon voted, a "new Bible" commissioned and financed by Constantine "which omitted those gospels that spoke of Christ's human traits," the losing gospels "outlawed, gathered up and burned." The interpretation is unsupported on every part. The council's twenty surviving canons and acts say nothing of the canon of Scripture; Eusebius, who was present, records the canon as still disputed in his own day (Church History 3.25); Athanasius, also present, in 367 still frames listing the books as his own undertaking of "necessity," the canon "handed down" rather than decreed — from which Breen rightly infers no Nicene catalogue was ever promulgated; and Constantine's one real Bible-project, the fifty copies ordered from Eusebius (Life of Constantine 4.36), was the transcription of an existing collection, not its selection, with no trace of suppression or burning. The legend distorts a real but smaller history — an emperor who convened a council that legislated on doctrine and discipline, and who later paid for Bibles. (A far narrower, genuinely scholarly question — whether Nicaea used or recognized particular disputed books in passing — is separate, and even its defenders deny the council drew up an official list.)²¹

The altar-ordeal miracle — were the inspired books sorted onto the altar at Nicaea? (assessment: unsupported)

The miraculous "book-ordeal" — the strand of the legend that supplies its mechanism: the disputed writings laid by the altar, the council at prayer, and the inspired books found on top while the spurious sank beneath. The interpretation is unsupported. The tale rests on a single source — §35 of the anonymous Synodicon Vetus, compiled around AD 900, nearly six centuries after the council — a work crowded with details found nowhere else and recounting synods scholars doubt ever met; no contemporary witness and none of the council's own acts or twenty canons know of it. Its result, a canon closed by divine sign in 325, is contradicted by Eusebius and Athanasius, both present at Nicaea, who go on treating the canon as still open. The anecdote entered Western letters only when its Greek manuscript was carried west by Andreas Darmasius and printed in 1601 by Johannes Pappus (hence later writers' "Pappus, in his Synodicon"), whence Voltaire relayed it — the source of the modern embellishment that the canon was at Nicaea "even sorted by a miracle."²²

Were the infancy gospels suppressed — banned, voted out, gathered up and burned? (assessment: unsupported)

The same legend in New-Testament-apocrypha dress — the popular framing that the gospels left out of the New Testament, the infancy gospels above all (the Infancy Gospel of Thomas, the Protoevangelium of James), were not merely passed over but suppressed: outlawed, voted down, "gathered up and burned." The interpretation is unsupported. The documented history is gradual non-reception: from the second century on the proto-orthodox never widely received these texts as apostolic, and the formal exclusion-lists, when they come, speak in the language of books "not to be received" — Innocent I (405) pronounces the pseudo-apostolic gospels "rejected and condemned," and the Decretum Gelasianum (c. 500) is literally titled On books to be received and not to be received — not destroyed. The one real decree-and-bonfire near Nicaea burned a heretic's doctrinal treatises — Constantine's order to torch Arius's writings on pain of death — not the apocryphal gospels, and even that suppression failed; meanwhile the excluded gospels went on circulating freely for a thousand years, the Protoevangelium surviving in 140-odd Greek manuscripts and a dozen languages, the very footprint of free circulation rather than eradication. A milder, true claim survives: they were never received as Scripture. The "banned and burned" image is the Nicaea-fixed-the-canon legend on the New Testament side.²³

Related arguments — the scholarly case for a Nicene canon-action

The sober, scholarly cousin of the popular legend: not a voted "new Bible," but the claim — built by the Jesuit introduction-scholar Rudolph Cornely on three textual props — that the council of 325 genuinely took up the canon and enumerated the sacred books, deuterocanon included. All three of his props are weighed here; they run from his strongest to his weakest.

Did Jerome's Preface to Judith show the Council of Nicaea counted Judith among the Scriptures? (assessment: weak)

The one early text adduced for a real Nicene action on the canon — set apart from the popular legend above, and Cornely's strongest prop. Jerome (c. 405) writes that Judith "is found by the Nicene Council to have been counted among the number of the Sacred Scriptures," and Cornely reads this as proof that the council of 325 enumerated the sacred books in a decree, deuterocanon included. The interpretation is weak. Jerome's hedged "is found to have been counted" attests at most a viva voce use of Judith in the council's deliberations — a book quoted as Scripture in argument, as Ephesus later leaned on Hebrews without decreeing a New Testament canon — not a promulgated list; he never cites a Nicene catalogue, the council's twenty surviving canons say nothing of Scripture, and Jerome himself, in the same prefaces, expressly rules Judith out of the canon ("not in the canon … among the Apocryphal writings"), which he could not do had he taken Nicaea to have decreed it in. Athanasius — present at Nicaea — likewise excludes Judith in 367, and Hefele treats a Nicene canon-decree as a later misattribution. The modest reading (Nicaea cited Judith in argument) is plausible; the decree-reading does not hold.²⁴

The Council of Hippo's canon 36 under a "Nicaea" heading — does the Vallicellian codex prove the council drew its canon from Nicaea? (assessment: unsupported)

Cornely's third prop. A single eighth/ninth-century codex (Vallicelliana B.58, fol. 77) prefaces the canon-list of the Council of Hippo (393) — "that besides the canonical Scriptures nothing be read in church" — with the rubric "Chapter XXIV of the Council of Nicaea," then runs out a catalogue intermixing the proto- and deuterocanonical books; Cornely concludes the Fathers of Hippo drew their list from the decrees of Nicaea (carried to Africa by Caecilian of Carthage), so that Nicaea decreed the wider canon. The interpretation is unsupported. The list beneath the rubric is the securely-dated African canon of Hippo — reaffirmed at Carthage in 397 and 419 — not a Nicene one; the "of Nicaea" heading survives in this lone late manuscript and is a textbook case of the mislabelling Hefele documents, later conciliar canons drifting under Nicaea's name because the genuine Nicene canons always headed the early corpus (the Rome copy welding Nicaea and Sardica sub uno titulo; Pope Zosimus quoting a Sardican canon as Nicene). The very Nicene exemplar Cornely invokes — Caecilian's copy at Carthage — is on record as having held no such canon when the African bishops searched it in the Apiarius affair, forcing them to send to Alexandria and Constantinople for authentic texts. Breen grants the prop "what force it may have" yet still refuses to "refer the Council of Nice among the proofs of the Canon," noting it "exists but in one sole codex"; the council's own twenty canons are silent.²⁵

Cassiodorus's Institutiones I.14 — does listing Nicaea among the dividers of Scripture show the council took up the canon? (assessment: unsupported)

Cornely's weakest prop. Cassiodorus (c. 543) names "the Councils of Nicaea and Chalcedon," with Hilary, Rufinus, and Epiphanius, among those who "by their divisions, fitted the sacred books to the appropriate mysteries"; Cornely infers the Nicene Fathers must have drawn up a division of the canon. The interpretation is unsupported. Cassiodorus's chapter is about how the already-received books are divided and numbered — reconciling Jerome's, Augustine's, and the Septuagint's schemes — not about which books are Scripture; his sentence yokes Nicaea to Chalcedon, which decreed no canon, so the inference proves too much; and as sixth-century, second-hand testimony to 325 it is best read as an instance of the later drift of canon-material under Nicaea's name — the very drift Hefele documents, and Hefele is the source from which Cornely (Breen notes) borrowed the citation. Breen's verdict — "obscure words" that "form no forcible proof" — stands; the council's own twenty canons are silent.²⁶

Related arguments — the shape of the canon's formation

The doctrine's positive backbone, told from both sides: not the legend the arguments above dismantle but the history the timeline positively shows. The same Irenaean proof-text (Against Heresies 3.11.8) anchors both — the canon's edges stayed soft for centuries, while its core, the four Gospels, was effectively closed early. Two complementary theses, weighed together: core fixed early, edges contested late.

Was the New Testament canon settled gradually, contested at its edges — not delivered whole? (assessment: sound)

The doctrine's positive backbone, and the corrective to the Nicaea-decreed-it legend above. The timeline's own witnesses show a boundary that formed by gradual reception and stayed fuzzy at its edges for centuries: books once received as Scripture or near it were later dropped (Irenaeus's 1 Clement and Shepherd of Hermas; the Muratorian list's Apocalypse of Peter and Wisdom of Solomon; the Barnabas and Shepherd bound into Codex Sinaiticus, the Clementine epistles into Alexandrinus), while seven that made it in were long disputed — Eusebius's antilegomena (James, Jude, 2 Peter, 2–3 John) plus Hebrews (doubted in the West) and the Apocalypse (doubted longest in the East: questioned by Dionysius of Alexandria, omitted by Cyril, absent from the twenty-two-book Syriac Peshitta). The early lists are partial and regional, the first exact twenty-seven coming only with Athanasius (367). The interpretation is sound — mainstream canon history, attested from within by Eusebius's threefold sorting and articulated by Metzger and Bruce — with one load-bearing qualification: the fuzziness belongs to the edges. The fourfold-Gospel-and-Pauline core was effectively fixed by c. 180 (Irenaeus's closed Tetramorph), so the thesis underwrites neither the "decreed whole at Nicaea" legend nor its mirror image, the "late and arbitrary" debunking.²⁷

Was the fourfold Gospel already closed by c. 180? — Irenaeus's Tetramorph (assessment: sound)

The early-fixed-core half of the same picture. By c. 180 Irenaeus defends the Gospels as exactly four — "neither more nor fewer in number than they are" — matched to the four zones of the world, the four winds, and the four faces of the cherubim, the fourfold "Tetramorph" (Against Heresies 3.11.8). Two claims must be kept apart. The argument for the number four — the cosmic symbolism of zones, winds, and living creatures — is an a priori fitness-argument, not evidence, and as proof of why there must be four it is weak (Skeat showed Irenaeus even borrowed the living-creatures scheme from an earlier source). But the historical datum the passage presses — that the fourfold Gospel was a fixed, closed collection by the late second century, a century before the rest of the New Testament settled — is sound: it is the mainstream of canon scholarship (Stanton, Hengel, Hill, Kruger), and it rests not on the symbolism but on independent evidence — Justin's "memoirs … called Gospels" (c. 155), Tatian's harmony of the four (the Diatessaron, c. 170), the Epistula Apostolorum (c. 140s), and the early four-Gospel codices. Irenaeus is the witness to a settled fourfold Gospel, not its inventor; his weak argument for the number no more unsettles the fact than a poor proof unsettles a true theorem. This is the firm center the soft-edges thesis is measured against.²⁸

The Council of Rome (382) and the "Damasine list" — did Damasus promulgate the canon? (assessment: weak)

The documentary basis of the popular claim that "the canon was fixed at the Council of Rome in 382 under Damasus." A real synod met at Rome in 382, and a canon-list matching the present twenty-seven books and the longer Septuagint Old Testament — the "Damasine list" — is ascribed to it; a respectable minority (Arévalo, who first proposed the attribution in 1794; Turner; Jurgens) holds the canon-chapters authentically Damasan. But the interpretation is weak. The list survives in no fourth-century document — only as a chapter of the Decretum Gelasianum, which von Dobschütz showed to be a single sixth-century pseudonymous compilation (its opening chapter quotes Augustine's work of 416), and which this timeline dates c. 500; the 382 attribution is itself an eighteenth-century conjecture unsupported by any contemporary witness; and Jerome — present at Rome, the age's foremost authority on the canon, and the very writer the list depends on — nowhere mentions a Roman conciliar canon, ranking Judith, Tobit, and the Maccabees below the canon in his own prefaces. What the evidence bears is the modest, plausible claim that Damasus's Rome received the canon the West then held, a decade before Hippo put it into conciliar form; the "fixed at Rome in 382" framing does not hold — a settled boundary read backward onto a council that left no such act, the close cousin of the Nicaea-counted-the-canon error.²⁹

Which "Gospel of Thomas" did the Fathers condemn — the sayings gospel or the Infancy Gospel? (assessment: weak)

A two-text problem on the New-Testament side. One title, "Gospel of Thomas," belongs to two unrelated works — the Gnostic sayings gospel (114 logia, recovered at Nag Hammadi) and the Infancy Gospel of Thomas (the miracle-working boy Jesus) — so the early condemnations of "the Gospel of Thomas" (Hippolytus, Origen, Eusebius, Cyril of Jerusalem, the Gelasian Decree) are routinely read as the rejection of the infancy text. That equation is weak. Where the target is identifiable it is the sayings gospel or a Manichaean book of that title — Hippolytus quotes its saying 4 and ties it to the Gnostic Naassenes; Cyril names a "Gospel of Thomas" written by "one of the three wicked disciples of Manes"; Eusebius brackets it among the heretics' apostolic-pseudonym forgeries (Peter, Thomas, Matthias). Only Origen's bare list and Eusebius's catalogue-name are loose enough to admit the infancy text, and an undescribed title identifies nothing. The witness who does attest the infancy material — Irenaeus, on the boy-Jesus alphabet tale (Against Heresies 1.20.1) — never calls it "Thomas" or a "gospel"; and the Gelasian list itself registers the Manichaean "Gospel of Thomas" and "the book on the infancy of the saviour" as separate entries. The infancy gospel was excluded too — but on its own, partly different evidence; the famous "Gospel of Thomas" rejections should not do double duty.³⁰

Did Luther remove four New Testament books — Hebrews, James, Jude, and Revelation? (assessment: weak)

The Reformation-era trope that Martin Luther "took out four New Testament books… and put them in an appendix without page numbers," which were "later put back into the New Testament by other Protestants." The interpretation is weak for that "took out… put back" framing: it confuses demotion-and-reordering with excision, and its "put back" half is false. Luther's 1522 German New Testament contains all four — translated, printed, prefaced, merely set last and unnumbered — so the books were doubted and demoted, not removed; they then stood in his own New Testament and in every Lutheran and Protestant Bible since, and he softened his verdicts (the "epistle of straw" line dropped from later editions, Revelation given a favorable new preface in 1530). The charge that the demotion was doctrinal is on this much fair — Luther's criterion was theological — but it tells against excision as much as for it, and the four were exactly the church's ancient antilegomena, questioned in Luther's day by Erasmus and Cajetan too. What survives is the narrower, sound claim: Luther doubted these four and ranked them below the rest. The argument is the New-Testament twin of the Old-Testament "Luther removed the deuterocanon" trope, and over-reads in the same way.³¹

Sources