COLLECTIONS · COMMUNITIES · CENTURIES

The canon takes shape Gathered, read, listed

The writing stays the same. The claim changes.

Before the Bible had a table of contents, there were communities of readers. Jews and Christians copied writings by hand, read them aloud in worship and quoted them in argument. Over centuries, some drew up lists of the books they held to be Scripture, and those lists did not all agree.

This page follows that process through what survives: the manuscripts themselves, the writers who quoted and recommended books, and the councils and confessions that finally named them. Each list belongs to a place, a time and a conversation, and each piece of evidence can show only so much.

FIRST, WHAT “CANON” MEANS

The word comes from the Greek kanōn, a measuring rod or rule. Said of Scripture, a canon is the list of books a community receives as its Bible. There is more than one such list, and they still differ today.

The 24 and the 39 are the same Hebrew books, counted and grouped differently.

JEWISH · TANAKH

PROTESTANT · WESTMINSTER

ROMAN CATHOLIC

EASTERN ORTHODOX

ETHIOPIAN ORTHODOX

RECORDS

KINDS OF EVIDENCE

PERIODS

SPAN

132 BC – 1647

TRADITIONS

WRITINGS TRACED

SOURCES

Every list on this page is about books. Start with the books themselves, as they survive.

01 · THE SURVIVING BOOKS

Before any list, the books

Long before anyone wrote down a list, writings were copied by hand. A codex shows which writings were bound together; on its own it does not show which a community counted as canonical.

Pictures: public domain, from Wikimedia Commons; each is credited when opened.

A surviving book shows what was copied. Being copied, read, quoted, recommended or listed are different claims; keep them apart.

02 · READ THE EVIDENCE

Five claims that are not interchangeable

Here, canon means a recognized collection of scriptural books. Select a kind of evidence to see what it can establish.

WHAT DOES THIS ESTABLISH?

Jude 14–15 uses an Enochic passage.

A writer uses words from a work. The authority of this passage does not specify the boundaries of a whole collection.

Now hold one writing still and watch which of those claims each witness makes about it.

03 · FOLLOW A WRITING

Follow a writing. Watch the claim change.

The writing stays the same; the kind and reach of the evidence change. Open any stop to inspect its full record.

The same work can be copied, read and recommended without being placed in a writer's canon. These are different witnesses, not stages every community passed through.

Recommended; not public scriptural reading

Outside acknowledged books; public use reported

Copied; partly surviving

For instruction; outside this canon

Only records relevant to this trail are shown. Silence in another selected record is not counted as rejection.

Those witnesses are spread over eighteen centuries. Put each back in its own time.

04 · ENTER THE TIMELINE

Put each witness back in its own time

Explore eighteen records. Dates refer to the text, object or decision described; later copies and compilations are identified where relevant.

A year or short span A century or broad period Date debated Nothing is pinned more precisely than its record. The lanes are this page's grouping; each record's own kind is shown when it is opened.

Point at a record to read it here; click to open it.

And in its own place: the lists come from different cities and churches, not one centre.

05 · WHERE THE LISTS WERE MADE

Now put each witness back in its own place

Eighteen records, each placed only as precisely as it names its own place. Choose a period, or open any record.

Land outlines: Natural Earth (public domain). A soft area marks a region the record names, not a boundary.

PLACE NOT GIVEN

These records name no place, or say it is uncertain. They are listed, not guessed.

A named city or site A named region Place disputed

Point at a record to read it here; click to open it.

Use, copying and explicit boundaries overlapped. A council could regulate an inherited collection; a bishop could distinguish books for instruction from books in his canon. No single meeting in this record selected a Bible for every tradition.

Compare Athanasius, the African Code and Trent above. Greek biblical collections also require attention to the particular manuscript: “the Septuagint” should not be treated as one surviving ancient volume with a universal contents page.

Three writers within one generation show how far apart the boundaries could be.

06 · BEGIN WITH A COMPARISON · AD 367–C. 400

The same writings. Different boundaries.

Three selected passages. No universal baseline.

SELECTED WITNESS A

Athanasius

Alexandria · 367

SELECTED WITNESS B

Augustine

Latin North Africa · late 4th / early 5th century

SELECTED WITNESS C

Jerome

Latin translation programme · around 400

Schematic registers · not manuscript facsimiles or complete book lists. Each mark represents one writing or named pair in a selected passage.

Wisdom and Sirach enter Augustine's canon. Athanasius recommends them for instruction outside his canon; Jerome distinguishes their edifying use from doctrinal authority.

Athanasius and Augustine enumerate the same twenty-seven New Testament books, yet their Old Testament boundaries differ. Agreement in one collection does not establish agreement in another. “Apocrypha” also changes meaning across authors and later traditions; it is not a timeless shelf label.

Three cases, looked at closely, show where a boundary becomes visible and where the evidence runs out.

07 · THE EVIDENCE ROOM · THREE CLOSE INVESTIGATIONS

Where the boundaries become visible

A list can look decisive until you ask what its categories mean. A manuscript can look complete until you inspect its ending. Slow down at these three points.

JEWISH COLLECTIONS · AUTHORITY AND ARRANGEMENT

A collection before a contents page

A reader’s vocabulary

The translator distinguishes law, prophets and other ancestral books. He does not name the members of the third group. A threefold description is evidence of differentiation, not a recoverable final inventory.

An actual collected object

This scroll combines familiar psalms in a different sequence with other hymns and prose about David. Found in Cave 11 in 1956 and unrolled in 1961, it makes the shape of a psalm collection visible where a later printed contents page cannot.

CHANGE THE EXPLANATION

What kind of collection is 11Q5?

The object does not change. The inference does. Compare the assumptions each reading needs.

The collection witnesses a form of the Psalter before its contents and order became uniform.

It needs: Evidence that the arrangement functioned as an authoritative collection, rather than merely preserving selected songs.

What follows: Variation becomes evidence about the developing collection itself.

Neither option is proved merely by comparing this scroll with a modern Psalter.

Sanders’s edition (DJD IV, 1965) established the scroll’s publication. Lim traces the alternative-Psalter interpretation and objections by Goshen-Gottstein, Talmon and Skehan. The dispute concerns the collection’s authority, not merely its arrangement.

Mroczek’s 2016 programme questions projecting our bounded Bible onto ancient sacred writing. Lim objects that textual fluidity and canonical authority must be distinguished. Different arrangements do not by themselves establish an absence of authoritative writings.

And what about a council at Jamnia?

Mishnah Yadayim 3:5 debates whether Song of Songs and Ecclesiastes impart ritual impurity to hands. Akiva strongly defends the Song; the passage preserves differing voices. This is legal discussion about sacred writings, not an enumerated decree selecting an entire Bible.

Lim traces the “Council of Jamnia” narrative to modern scholarship and its critics. This source supplies neither a complete inventory nor a closure date.

EUSEBIUS · A CONDITIONAL CLASSIFICATION

When a writing appears in two categories.

A SOURCE THAT RESISTS ONE BOX

Revelation appears twice.

Eusebius’s Church History 3.25 reports reception as well as judging it. His categories resist a single yes/no column.

“if it really seem proper”

Revelation is placed conditionally among accepted writings.

A conditional placement is not an unqualified verdict.

THE MATERIAL WITNESS · SURVIVAL AND RECONSTRUCTION

The end of a codex is not the edge of a canon

Royal MS 1 D VIII · Codex Alexandrinus · 5th century. Inspect the final surviving works, then add what the contents list names. A dashed boundary marks a different kind of evidence—not a lower level of religious value.

A surviving page of the codex (013a Mc 6,27-54, from Mark), not one of the leaves below. Unknown author · Public domain · Wikimedia Commons

Its text precedes the Clementine writings. Physical sequence alone does not define the canonical status of the next work.

Schematic sequence, not a facsimile, page count or reconstruction of lost leaves. Labels follow the British Library catalogue. Its digitised images were unavailable at consultation.

ANOTHER LANGUAGE, ANOTHER COLLECTION

The Peshitta’s twenty-two

The Syriac Peshitta became the standard version around 400. Its New Testament collection does not contain the five writings shown here. This is a version’s contents, not a statement that every later Syriac church has the same boundaries.

Not represented in the twenty-two-book version

2 Peter
2 John
3 John
Jude
Revelation

Compare this with Athanasius’s twenty-seven, not with an assumed universal starting point. A translation’s history and a council’s jurisdiction answer different questions.

The question is still live: churches and Judaism today hold different collections.

08 · COMPARE TRADITIONS

Compare collections without a default Bible

Choose two named sources. These are selected writings, not exhaustive inventories. Read each source’s scope before comparing its entries.

Catechism §120 · contemporary official enumeration

46 Old Testament books and 27 New Testament books in the Catechism’s enumeration.

Confession 1.2–3 · OPC received text

39 Old Testament books and 27 New Testament books; the confession excludes Apocrypha from the canon.

WRITING / COLLECTION

ROMAN CATHOLIC

REFORMED · WESTMINSTER

Genesis–Deuteronomy

Canonical · Old Testament

Psalms

Canonical · Old Testament

Esther

Canonical · including Greek portions

Canonical · Hebrew form

Daniel

Canonical · including Greek portions

Canonical · Hebrew/Aramaic form

Tobit & Judith

Canonical

Excluded from canon

Wisdom & Sirach

Canonical

Excluded from canon

Baruch

Canonical

Excluded from canon

1–2 Maccabees (Greek tradition)

Canonical

Excluded from canon

Prayer of Manasseh

Not in this enumeration

Not in enumeration

1 Enoch & Jubilees

Not in this enumeration

Not in enumeration

Revelation

Canonical · New Testament

“Not specified” and “not individually named” describe this source’s limits, not a church’s rejection. “Greek portions” refers to material in Greek forms of Esther and Daniel beyond their Hebrew/Aramaic forms. No color marks one collection as correct.

Your turn. Weigh a piece of evidence and say how far it can take you.

09 · TRY THE HISTORICAL REASONING

How far can this evidence take you?

Choose the conclusion the source supports.

QUESTION 1 OF 10

Hermas appears in a biblical codex. What has that established?

Everything above rests on these sources. Check them.

10 · KEEP THE SOURCES

Keep the sources in the conversation

The links beside each record lead to the evidence. This directory records what was actually consulted.

Follow the arguments, not just the bibliography.

How do we read ancient lists accurately?

Original-language lists with translations and interpretive notes provide a route from a modern claim back to a particular witness. Use this for detailed textual work beyond the older translations linked here.

What did sacred writing look like before “the Bible”?

Reframes the question around ancient literary imagination: growing traditions, attributed authors and writings beyond later canonical boundaries. Its programme challenges treating every ancient collection as a preliminary modern Bible.

Does variation mean that there was no canon?

Separates changing textual forms from authority, while revisiting the Psalms-scroll debate and the limits of reconstructing closure from rabbinic evidence.

How early is the Muratorian evidence?

Dating changes how early this witness can be used. Hahneman argues for a fourth-century eastern setting. Guignard supports a Greek original, distinguishing a later fourth-century Latin text from a likely late second-century composition. A surviving copy’s date does not itself date the original.

What does an Ethiopian count of eighty-one mean?

Relates narrower and broader lists to counting and book identities. Meqabyan must not be equated with the Greek books called Maccabees. A number without names and grouping rules is not an inventory.

What remains open

The DJD edition and the earlier rebuttals were located through the institutional description and Lim’s bibliography, not independently read. No exact count of 11Q5’s biblical psalms is used: institutional and scholarly descriptions count the material differently.

No direct inspection or collation of Alexandrinus images; catalogue explicitly reports images unavailable. Gallagher–Meade, Mroczek, Hahneman and Guignard full works not accessed; their access limits appear on the page. Sanders’s 1965 edition and early responses are bibliographically located, not claimed as read. Syriac account concerns the historical Peshitta version, not a census of living churches. Samaritan, Armenian, Georgian and other reception histories remain outside this selected study. Comparison covers selected writings and named institutional descriptions, not exhaustive tradition-wide inventories.

Manuscript pictures are public-domain photographs from Wikimedia Commons, credited where shown; the diagrams are our own. Expanded research checked 14 October 2026.

Ask of every list: Whose collection? When? For what use?

Rooted in Scripture. Open for discovery.

A place to read slowly, ask deeply, and follow the connections. Scripture, study and the voices of Christian history, brought together for a lifetime of learning.

SPEND TIME IN THE WORD

Open the reader
Books & reading progress
Compare Scripture editions
Find a passage

FOLLOW YOUR QUESTIONS

Study collections
People & places
Faith & apologetics
Stories of faith

A library with a paper trail.

Meet the authors. Explore the collection. Trace each work to its source.

Download the code.

Open source on GitHub.