The Lost and Noncanonical Gospels

Noncanonical means a writing was not received into the canon — here, the twenty-seven-book New Testament. It describes a text's place in that collection, not a verdict that the text is a forgery or worthless. The canon formed over centuries, not in a single meeting or vote. Churches copied, read aloud, and argued over many writings; lists differed by place, and the edges were debated for generations. Athanasius's Festal Letter of 367 gives the earliest surviving list matching the twenty-seven exactly, and consensus kept maturing after it. Each noncanonical text must be weighed on its own date, author, and history.

In late 1945, near the Upper Egyptian town of Nag Hammadi, a sealed jar was found holding twelve leather-bound papyrus codices, plus a tractate from a thirteenth — fifty-two treatises in all. The books date to the third and fourth centuries and are written in Coptic, but the works are translations of Greek originals. The find changed scholarship for a plain reason: before 1945, many of these writings were known only through summaries and quotations in the works of their opponents, who quoted them to refute them. Afterward, scholars could read the texts themselves, in full, in their own voice — primary texts, not only their enemies' summaries.

The Gospel of Thomas, in Codex II, is the collection's best-known text, and the Nag Hammadi copy is its only complete one. It holds 114 sayings attributed to Jesus and nothing else: no birth narrative, no ministry, no crucifixion, no resurrection. Some sayings resemble material in the canonical Gospels; others are distinctive, and the last, saying 114, stages a dispute involving Peter and Mary that interpreters still debate. Dating is contested — some argue an early core reaches back further — but most scholars date the composition as we have it to the second century. The text presents itself as recorded by Didymos Judas Thomas; scholars do not accept the apostle as author, and the real author is unknown. Attributed to is not written by.

The Gospel of Mary did not come from Nag Hammadi. Its fullest copy is in the Berlin Codex, a fifth-century Coptic manuscript bought in Cairo in 1896 and published only in 1955; two Greek fragments survive too. Even it is incomplete: pages 1 to 6 and 11 to 14 are missing. Most scholars place composition in the second century. In the surviving pages, the disciples are distressed after Jesus departs; Mary comforts them, and Peter asks her to share teaching she received that the others did not. She recounts a vision, including the soul's ascent past hostile powers. Andrew and Peter object, questioning whether the Savior would have spoken privately with a woman over them. Levi rebukes Peter, defends her, and the disciples go out to preach. This Mary is usually identified as Mary Magdalene — the text names her only as Mary, so that identification remains an inference.

The Gospel of Philip, bound in Codex II with Thomas, is a Valentinian anthology whose main concern is sacramental theology: baptism, anointing, eucharist, redemption, and the bridal chamber as an image of spiritual unity. Most scholars date it to the third century. Its title appears only in a closing note; Philip is merely mentioned, not presented as author. Two passages draw outsized attention. One calls Mary Magdalene the companion of Jesus, using a form of the Greek word koinōnos, which ranges across associate, partner, and fellow worker — not by itself a word for wife; the text uses a different word where it plainly means a spouse. The other describes Jesus loving her more than the other disciples and kissing her often — then the manuscript breaks. A hole sits exactly where the sentence would say where he kissed her. Bracketed reconstructions are guesses; honest reading leaves the gap a gap.

The Gospel of Truth, in Codex I, is a Valentinian meditation — a sermon, not a story. Many scholars consider it the work of Valentinus himself; others assign it to his circle. In its telling, not knowing the Father produces fear and confusion, pictured as fog and troubled sleep, and knowing the Father dissolves that state. It narrates no events of Jesus' life.

The Gospel of Judas survives in one damaged Coptic copy in Codex Tchacos, radiocarbon dated to roughly 220–340 and itself translating a second-century Greek original. The codex resurfaced in Egypt in the 1970s, deteriorated through poor storage, and was first published in English in 2006 by the National Geographic Society. That translation presented Judas as the one disciple who understood — a hero who handed Jesus over at Jesus' request. The reading was quickly contested. April DeConick and others argued key choices reversed the sense: daimon, rendered as spirit, means demon in this literature, where another word serves for spirit; Judas was separated from the holy generation, not set apart for it; and a missing negative, later acknowledged, had flipped a statement that he would not ascend. The text is anonymous, and its Judas is still debated.

None of these writings is a first-hand report of Jesus. Each is a later composition, written decades to centuries after his lifetime, surviving in still later copies, in translation, sometimes with pages and words missing. A name in a title records an attribution, not verified authorship.

Read for what they are, these texts are first-rate evidence of the diversity of early Christianity. Second- and third-century communities asked real, differing questions — about knowledge and ignorance, body and soul, who could teach, what Jesus' death meant — and did not answer alike. These gospels are not transcripts of Jesus; they are records of the early Christians who argued about him.

This center therefore holds a simple rule and does not bend it: disputed sayings are never presented here as proven words of Jesus. They are introduced as the evidence allows — the Gospel of Thomas says, the Gospel of Mary portrays — not with a flat Jesus said. That is not a lack of reverence; it is the opposite. Words worth building a life on are worth labeling honestly.

This center reads them because their questions are real and have not gone away. These writings return again and again to inner knowledge over outward conformity, to light against a fog of confusion, and to self-knowledge as a beginning of freedom — documented themes of this study. A text can be a poor source for the events of Jesus' life and still be a rich source for the life of faith, if it is read with its labels on: second-century, damaged, disputed.

In the churches' tradition, the canon was received more than manufactured. Communities recognized in the four canonical Gospels — and the other twenty-three books — writings tied to apostolic witness, read publicly in worship across scattered churches, and consistent with the rule of faith they confessed. By the late second century Irenaeus was already arguing there were four Gospels, neither more nor fewer, and the core was broadly settled long before Athanasius named the twenty-seven in 367. On this view, the noncanonical gospels were later works the churches did not receive as Scripture. Holding that view still leaves room to read them historically, as this study does, without granting them canonical authority.