Dating and Authorship: How Old Is the Gospel of Thomas?

No question about the Gospel of Thomas has produced more argument than its date, and none is less settled. The range under serious discussion runs from roughly 50 CE to roughly 140 CE — the difference between a text older than Mark and a text a generation younger than John.

What the manuscripts actually tell us

The physical evidence sets only an outer limit. The Coptic manuscript in Nag Hammadi Codex II was copied around 340 CE. The three Greek fragments from Oxyrhynchus date to around 200 CE. Since a Greek copy was circulating in provincial Egypt by 200, the composition must be meaningfully earlier — but how much earlier the papyri cannot say.

External references push it back a little further. Hippolytus of Rome quotes a “Gospel according to Thomas” around 225 CE, and Origen mentions it slightly earlier. Both treat it as an established text used by groups they oppose.

The case for an early date

Advocates of a first-century Thomas — associated particularly with Helmut Koester, Stephen Patterson and the Jesus Seminar — argue several points.

• No dependence on canonical order. The sayings shared with the synoptics appear in a completely different sequence, and are often linked by catchwords rather than by narrative. A writer copying Matthew would be unlikely to scramble the material so thoroughly.
• Parables without their allegories. Thomas gives the sower, the wicked tenants and the great banquet without the interpretive framing the synoptics attach. Since allegorical explanation is generally a sign of later reflection, the barer version may be the older one.
• No passion, no resurrection. A text written after those had become the centre of Christian preaching would be strange indeed to omit them entirely.
• The James saying. Saying 12 directs the disciples to James the Just, who led the Jerusalem community until his death around 62 CE. That looks like a fossil from a period when his authority was live.

The case for a later date

The counter-case — argued by Nicholas Perrin, Simon Gathercole, Mark Goodacre and others — is equally serious.

• Traces of redaction. Some sayings appear to reflect wording that is distinctively Matthew’s or Luke’s rather than shared tradition, which would mean Thomas knew the finished gospels.
• Syriac patterns. Perrin argued that catchword connections between adjacent sayings work best if the text was composed in Syriac, pointing towards Edessa around 140 CE and towards familiarity with Tatian’s Diatessaron.
• Developed theology. The emphasis on secrecy, on the pre-existent self and on the solitary monachos reads to some scholars as second-century, closer to the world of Valentinus than of Paul.

The compromise most scholars actually hold

The likeliest picture is that both camps are describing real features of a text that grew over time. Thomas contains obvious doublets — sayings 56 and 80, 87 and 112, 3 and 113 — which is what accumulation looks like. Many now favour a rolling corpus model: an early core of sayings, possibly from the middle of the first century, expanded and re-edited over several generations, reaching something like the form we have by the early second century.

The visible difference between the Greek fragments and the Coptic text supports this. The text was still being altered after 200 CE, so there is no reason to think it was fixed before it.

Authorship

The attribution to “Didymos Judas Thomas” is not accepted as historical by any significant body of scholarship. The doubled name — Didymos is Greek for twin, Thomas Aramaic for the same — points to the Syrian tradition, where Judas Thomas was venerated as the apostle to the east and appears in the Acts of Thomas and the Book of Thomas the Contender. The name is a claim to authority made by a community, not a signature.