The Diatessaron, Canonical or Non-canonical? Rereading the Dura Fragment

Abstract

Among those texts that vied for a position as authoritative Scripture, but were eventually rejected by ecclesiastical authorities, was the so-called Diatessaron of Tatian. Having been compiled from the four canonical gospels, Tatian's work occupies a liminal position between the categories of ‘canonical’ and ‘apocryphal’, since the majority of its content was common to users of the fourfold gospel, though this content existed in a radically altered form and was tainted by association with an author widely accused of heresy. In order to demonstrate the originality of Tatian's gospel composition, this article gives a close reading of the only surviving Greek witness to it, a fragment of parchment found in excavations at Dura-Europos. Dura's very location as a borderland between Rome and Persia corresponds with the fact that in this outpost garrison city Christians were using a gospel text that would have appeared markedly strange to those in the mainstream of the Christian tradition. The wording that can be recovered from the Dura fragment shows how Tatian creatively and intelligently combined the text of the four gospels to produce a new narrative of the life of Jesus, choosing to leave out certain elements and to make deliberate emendations along the way. However, it was precisely such originality that made his gospel appear problematic, so in order to rescue his text from censure, later scribes had to domesticate it by making it conform throughout to the canonical versions. Comparison of the Dura fragment with the medieval Arabic gospel harmony and with the Latin version in Codex Fuldensis illustrates well this process whereby Tatian's gospel went from being a rival to the fourfold gospel to a designedly secondary, and therefore acceptable, work.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 29 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.