Was the New Testament canon settled gradually, contested at its edges

Type: supporting argument (off-timeline) · Assessment: sound — this is mainstream canon history: the New-Testament boundary formed by a long, regional reception, and it stayed fuzzy at the edges for centuries. Some books once read as Scripture or near it were later dropped (1 Clement, the Shepherd of Hermas, the Apocalypse of Peter, Barnabas), while seven that made it in were long disputed (Hebrews, James, 2 Peter, 2–3 John, Jude, and the Apocalypse — Eusebius's antilegomena). The caveat is that the core — the four Gospels, Acts, and the Pauline letters — was effectively fixed very early (by c. 180), so the contest was always at the edges, never the whole; this is the positive counterpart to, and corrective of, the legend that the canon was decreed whole at Nicaea · advanced by the modern study of the canon (B. M. Metzger, F. F. Bruce)

This argument is the affirmative backbone of the whole doctrine: where the neighboring arguments dismantle the legend that the canon was decreed in one stroke, this one states what the timeline positively shows — a canon received gradually, its edges soft for centuries.

The claim

— F. F. Bruce, The Canon of Scripture (Downers Grove: IVP, 1988), p. 22.¹ The interpretation the claim serves is that the New-Testament canon was not handed down whole and sharp-edged — neither delivered intact by the apostles nor cut to shape by a single council — but recognized over generations, book by book and region by region, with a genuinely fuzzy boundary: books later excluded were once received as Scripture or near it, and several books finally received were long disputed. It is the same history the timeline tells, stated as a thesis and weighed on its own.

Who advances it

This is the mainstream of modern canon scholarship — B. M. Metzger (The Canon of the New Testament, 1987), F. F. Bruce (The Canon of Scripture, 1988), Hans von Campenhausen (The Formation of the Christian Bible, 1972), and the recent source-edition of E. L. Gallagher and J. D. Meade (The Biblical Canon Lists from Early Christianity, 2017). But its root is ancient and internal: Eusebius himself, on the eve of Nicaea, sorts the books into the "acknowledged" (homologoumena), the "disputed" (antilegomena), and the "spurious" (notha) — a man describing a boundary still being drawn, not reporting a finished list.² The thesis stands between two errors. On one side is the popular legend that the canon was fixed in one imperial-conciliar stroke at Nicaea (325) — weighed, and found unsupported, among the arguments alongside this one. On the other is the opposite over-correction — that the canon was a late, arbitrary, top-down imposition with no basis before the fourth century — which the early stability of the core refutes (see the assessment below).

The evidence for

The boundary was soft in both directions, and the timeline shows it at every step. The data fall into three groups.

1. Books once received as Scripture, or near it, then dropped. Irenaeus (c. 180) treats 1 Clement as a "most powerful" apostolic authority (Against Heresies 3.3.3) and cites the Shepherd of Hermas as "Scripture" (4.20.2) — both later excluded. The Muratorian Fragment (c. 170) receives the Apocalypse of Peter (noting only that "some of us are not willing that it be read in church"), counts the Old-Testament Wisdom of Solomon among its books, and allows the Shepherd to be read privately. The great fourth- and fifth-century Bibles still carry such books bound in: Codex Sinaiticus appends the Epistle of Barnabas and the Shepherd of Hermas after Revelation, and Codex Alexandrinus appends 1 and 2 Clement. As late as 367 Athanasius keeps a middle rank of books "not indeed included in the Canon, but appointed by the Fathers to be read" — the Teaching of the Apostles (Didache) and the Shepherd among them.³

2. Books finally received, but long disputed. Eusebius's antilegomena — "disputed… yet known to most" — are James, Jude, 2 Peter, and 2–3 John; tradition adds Hebrews and the Apocalypse, the standard seven "disputed" books. Each has a documented history of doubt. Hebrews was accepted as Pauline early in the East but doubted in the West into the fourth century: the Muratorian list omits it, and Carthage (397) still sets "one epistle… to the Hebrews" a half-step apart from Paul's "thirteen." The Apocalypse was doubted longest in the East: Dionysius of Alexandria, Origen's pupil, argued on stylistic grounds that it was not the work of John the evangelist (preserved in Eusebius, Church History 7.25); Cyril of Jerusalem (c. 350) closes his New-Testament list at twenty-six books, conspicuously omitting it; and the Syriac Peshitta long carried a twenty-two-book New Testament, dropping 2 Peter, 2–3 John, Jude, and Revelation together.⁴

3. The lists themselves are partial and regional. No early catalogue is both complete and uncontested. The Muratorian list (c. 170) omits James, both epistles of Peter, Hebrews, and 3 John; Irenaeus quotes some twenty-one of the eventual twenty-seven; Cyril stops at twenty-six. The first surviving list to match the present twenty-seven exactly is Athanasius's, in 367 — and even he frames it as his own undertaking, written "of necessity," of books "handed down" rather than decreed. The conciliar ratifications (Rome 382, Hippo 393, Carthage 397) are Western and regional, "the conciliar ratification of a reception already settled," with the East still hesitant over the Apocalypse and East and West later divided over the Old-Testament deuterocanon down to the Reformation.⁵

Adversarial assessment

What holds. On the documented record the New-Testament boundary really was gradual and really was fuzzy at its edges — this is not a contested reconstruction but the plain reading of the witnesses, and the lists, the once-included books, and the long-disputed books all converge on it. It is the correct positive frame for the doctrine, and the direct corrective to the "decreed whole at Nicaea" legend: the canon was recognized, not voted; received, not delivered in one act.

What must not be overstated. "Gradual and contested at its edges" is exactly the right scope — and the word edges is load-bearing. The thesis does not warrant the stronger popular inference that the canon was wide open, late, or arbitrary. The core was effectively fixed very early: by c. 180 Irenaeus defends a closed, fourfold Gospel ("neither more nor fewer in number than they are," Against Heresies 3.11.8), and the four Gospels, Acts, and the thirteen Pauline letters — with 1 Peter and 1 John — are stable and undisputed across every early witness. What moved was a fringe of perhaps a dozen books: a handful of shorter epistles and the Apocalypse working their way in, a handful of near-canonical writings (Clement, Hermas, Barnabas, the Apocalypse of Peter) working their way out. The contest was real, but it was always at the margin of a settled center — so "the canon was a fourth-century invention" overshoots the evidence as badly as "the canon was decreed at Nicaea." The accurate picture is a firm core fixed by the late second century with soft edges that took another two centuries to harden.⁶

Assessment

Sound. That the New-Testament canon was settled gradually and stayed contested at its edges — books received then dropped, books finally received but long disputed, lists partial and regional well into the fourth century — is mainstream canon history, attested from within by Eusebius's threefold sorting and articulated by Metzger, Bruce, von Campenhausen, and the recent list-editions. The one qualification, which the best scholarship makes too, is that the fuzziness belongs to the edges: the fourfold-Gospel-and-Pauline core was effectively closed by c. 180, so the thesis underwrites neither the "decreed whole at Nicaea" legend nor its mirror image, the "late and arbitrary" debunking. Stated with that scope, it is the doctrine's positive backbone — the true history the legends distort.

Sources