The Jar at Nag Hammadi

In December 1945, near the town of Nag Hammadi in Upper Egypt, a farmer named Muhammad Ali al-Samman was digging for sabakh, the nitrate-rich soil used as fertiliser, at the base of a cliff. He struck a sealed earthenware jar. What was inside was thirteen leather-bound papyrus codices containing some fifty-two separate texts, most of which no living person had read.

The find is unusual among manuscript discoveries because of what it does to a question that was previously unanswerable. Before 1945, almost everything known about the suppressed Christian literature of the second and third centuries came from the people who suppressed it. Irenaeus, Tertullian, Epiphanius and Hippolytus described these texts in order to refute them. Historians had the prosecution's summary of the evidence and nothing from the defence.

The jar contained the defence. Among the texts were the Gospel of Thomas, a collection of sayings attributed to Jesus with no narrative and no crucifixion; the Gospel of Philip; the Apocryphon of John, with its account of a lesser creator god who does not know what is above him; and the Gospel of Truth, which Irenaeus had named specifically as a work of the Valentinians. For the first time the refuted could be read directly.

The dating is where the story gets specific. The codices themselves were manufactured around the middle of the fourth century; the cartonnage, the waste papyrus used to stiffen the leather covers, includes dated receipts from the 340s. The texts they contain are copies of Greek originals considerably older, generally placed in the second century.

So the books were made in the fourth century and buried shortly after. The most widely accepted explanation for the burial turns on a document that survives independently. In 367 CE, Athanasius, bishop of Alexandria, issued his thirty-ninth Festal Letter. It contains the earliest surviving list of exactly the twenty-seven books of the New Testament as they now stand, and it instructs that writings outside that list are not to be read.

There was a large Pachomian monastery within a few miles of where the jar was found. The inference historians draw, and it is an inference rather than a documented event, is that monks in possession of a library that had just been declared illegitimate chose to bury it rather than destroy it. Someone sealed the jar carefully. Someone intended the contents to survive.

What matters here is narrower than the theological content, and it is the reason the find belongs in any serious account of how knowledge is sorted. The canon did not emerge by consensus over centuries of gentle convergence. At a specific moment, a specific bishop with specific institutional authority issued a list, and the material outside the list went out of circulation with sufficient thoroughness that most of it was unrecoverable for sixteen hundred years.

The recovery was also nearly incomplete in a mundane way. Muhammad Ali's mother is reported to have burned some of the papyrus in her oven for fuel. Codices were split up and sold separately. One was smuggled out of Egypt and offered for sale in the United States, where the Jung Institute in Zurich eventually acquired it. The full collection was not published in English translation until 1977.

The honest reading does not require deciding that the buried texts were true and the canonical ones false. They are a mixed body of writing of very uneven quality, and some of it is genuinely strange. The point is that they existed, that they had readers, and that the question of which of them counted was settled by an authority rather than by an argument anyone was invited to follow.

That is the thing the jar makes impossible to keep abstract. Someone decided. There is a letter with a date on it.