P52 (Rylands Papyrus P457, ~125 AD)

The verso carries 18:37–38: "To this end was I born," and Pilate's "What is truth?"

Both sides are written on. That single fact matters as much as the text. A scroll is written on one side; a codex — a bound book — is written on both.

Christians adopted the codex early and almost universally while the wider Greco-Roman world was still unrolling scrolls, and P52 is physical evidence of that habit in its first generations.

Provenance: Bernard Grenfell bought a lot of papyri on the Egyptian market in 1920 for the John Rylands Library. The scrap sat unsorted for fourteen years until Colin H. Roberts, going through the collection in 1934, recognized Johannine Greek. He published it in 1935 as An Unpublished Fragment of the Fourth Gospel in the John Rylands Library.

Roberts dated the hand by comparison with papyri that carry their own dates: P. Fayum 110 (AD 94), P. London 2078 (Domitian's reign, AD 81–96), the Berlin Iliad P. Berol. 6845 (c. AD 100), P. Oslo 22 (AD 127), and P. Egerton 2 (c. AD 150). He concluded the first half of the second century. Frederic Kenyon, Harold Bell, Wilhelm Schubart and Ulrich Wilcken — the leading papyrologists of the day — concurred.

Extrapolate from the fragment and you get a book. The reconstructed leaf was about 21 by 20 centimetres; a complete Gospel of John would have run to roughly 130 pages across some 33 folded sheets. Not a private extract or a lectionary snippet: a full-length codex of John, in provincial Egypt, in the second century.

That is what makes the fragment historically expensive. F. C. Baur and the Tübingen school had dated the Fourth Gospel to about AD 160–170, on the argument that its high Christology required generations of development after the apostles. P52 leaves no room for that timetable — composition, publication, copying and distribution up the Nile all have to fit in front of it.

The traditional date of AD 90–95 sits comfortably.

The text itself agrees with the later Greek tradition of John 18. There is no variant of consequence — spacing and line-breaks aside, the words are the words printed in critical editions today.

The limit is real and should be stated plainly: palaeographic dating of a literary hand gives a range, not a point. Brent Nongbri showed in Harvard Theological Review 98 (2005) that Roberts's own comparanda also admit parallels from later in the second century, and argued the honest window must extend into the later second and possibly early third century. Nongbri did not propose that P52 is late; he argued that the confidence interval is wider than apologists usually state.

P52 also no longer stands alone. P90 (Oxyrhynchus Papyrus 3523), carrying John 18:36–19:7, and P104 (Oxyrhynchus 4404), carrying Matthew 21:34–37 and 43–45, are both dated to the second century, and P66 and P75 bring substantial continuous text of John and Luke to about AD 200. The second-century evidence for the Gospels is a small cluster, not a single scrap.

The mainstream range has held. Pasquale Orsini and Willy Clarysse, working through the early Christian papyri systematically in 2012, placed P52 at AD 125–175. Even at the widest end of any proposed range it remains the earliest identified New Testament manuscript.

It is catalogued as Greek Papyrus 457 at the John Rylands Research Institute and Library, University of Manchester.