Use this site

Peter Gurry (NTS 2016) estimates on the order of ~500,000 readings across the whole Greek tradition — a extrapolation, not a hand count of every manuscript. Our site does publish a mechanical count, but only for word-aligned disagreements vs SR GNT in extant letters of our 1–400 CE CNTR witness slice — not the full ~5,700+ manuscript tradition.

Within our frozen window, substantial Greek material survives from the second and third centuries (papyri) plus fourth-century codices such as 01 and 03. That does not recover every verse in every book from every century — but it falsifies the blanket claim that nothing early exists.

Standard critical-edition framing (and our site’s own scope statement) treats the fourth-century major codices plus earlier papyri as the backbone for reconstructing the Greek text — while still not counting every later Byzantine variant or replacing NA28/ECM apparatus judgments.

In CNTR transcriptions in our corpus, that block is marked absent; narrative runs from 7:52 to 8:12 in those witnesses. Later Byzantine copies insert the pericope — a different story from our early slice.

Codex Sinaiticus (01) and Codex Vaticanus (03) end Mark at 16:8 in the main Greek tradition represented in CNTR; the longer ending 16:9–20 is a later supplement in the manuscript record. Textual critics still debate composition history — our data support the codex-level statement for 01/03.

1 John 5:7–8 is extant in early Greek witnesses with the short reading (Spirit, water, blood). The explicit heavenly Trinity wording of the Comma is not attested in Greek before the sixteenth century; it entered some Latin lines and late Greek retro-translations.

The Uthmanic Text Type is highly uniform, but published stemmatics (Cook; Sidky; van Putten) report on the order of ~35–40 regional rasm differences among the four ʿUthmānic exemplar traditions — not wholesale content rewrites, but not zero graphic variation. Early palimpsests such as Ṣanʿāʾ 1 preserve a lower text that is not UT.

Sidky and van Putten’s archetype framing treats UT as stable from the seventh century onward for the rasm tradition they model — while allowing later scribal noise and regional graphic differences, not a claim of byte-identical copies everywhere.

Optional intentional-vs-error tags on some Greek NT units are heuristic and model-assisted — useful for exploration, not proof of motive across the tradition. Most counted units are spelling, word order, or ordinary scribal difference vs SR GNT.

Out of scope by design: we mechanically align a bounded CNTR witness set to SR GNT. NA28/ECM/IGNTP record editorial judgments and thousands of witnesses beyond our 1–400 CE illustrative slice.

Second-Temple and pre-Christian Hebrew witnesses in our corpus (Dead Sea Scrolls, Nash Papyrus) show a largely continuous consonantal tradition alongside spelling diversity and occasional longer or shorter readings in some books. That supports substantial continuity for much of the Hebrew Bible — not byte-identical identity with every medieval Masoretic codex for every verse a first-century reader might have heard.

Jewish Greek Torah and prophetic books are attested on pre-Christian papyri in this site’s Hebrew/LXX slice — e.g. P.Ryl. 458 (Genesis), P.Fouad 266 (Deuteronomy), P.Oxy. 1007 (Exodus) — centuries before Christian recensions and church fathers. Later Christian copies and revisions exist; that is not the same as inventing the whole Greek Old Testament tradition.

The Great Isaiah Scroll agrees with the Masoretic book of Isaiah far more often than it disagrees — a key data point for stability. It also preserves thousands of spelling variants and documented textual differences elsewhere in the DSS corpus; “remarkable agreement in Isaiah” is not a license to claim zero textual history for the whole Hebrew Bible.

Within our Hebrew/LXX window, material survives from the second century BCE onward (DSS) and from pre-Christian Greek papyri. That does not recover every book on every leaf for every century — but it falsifies the blanket claim that no pre-medieval witnesses exist.