N. T. WRIGHT ASKS:

Can We Trust the Gospels?
None of the Gnostic texts--or any other recently unearthed find--can trump the four canonical gospels.
By N.T. Wright
The key question for studying Jesus is: Can we trust the gospels? I am referring to the four books which are known by the names of Matthew, Mark, Luke, and John, and which are found in the "canon" of the New Testament--that is, the collection of books that the church, from early on, recognized as authentic and authoritative (hence the often-used phrase "the canonical gospels") ....

His answer to the question is a qualified yes, for which there is indeed a reasonable case.

He discusses the Gospel of Thomas as the prime contender for being a an additional useful source for the historical Jesus:

Take the best known, and one of the longest, of the Nag Hammadi documents: a collection of supposed sayings of Jesus known as the Gospel of Thomas. This is the book which, it has often been suggested, could and should be treated as at least equal, and quite possibly superior, to the canonical gospels as a historical source for Jesus himself. The version of Thomas we now have, like most of the Nag Hammadi material, is written in Coptic, a language spoken in Egypt at the time. But it has been demonstrated that Thomas is a translation from Syriac, a language quite like the Aramaic that Jesus must have spoken (though he pretty certainly spoke Greek as well, just as many people in today's world speak English as a second language). But the Syriac traditions that Thomas embodies can be dated, quite reliably, not to the first century at all, but to the second half of the second century. That is over a hundred years after Jesus's own day--in other words, seventy to a hundred years after the time when the four canonical gospels were in widespread use across the early church.

I'm not sure whose view he is citing here; I know there was work in the 1970s and 80s by Helmut Koester and others arguing for a Syrian provenance for the Gospel of Thomas and for all I know they're right. And it seems that April DeConick argues for a Syriac stage in the transmission of Thomas in her recent book. (Her book hasn't arrived in our library, so I haven't been able to read it yet.) But I have to say I'm very skeptical in principle of our ability to say that Thomas was composed (or transmitted) specifically in Syriac. (Syriac is the dialect of Eastern Aramaic that was spoken in Edessa and which later became an important church language.) It is possible that Thomas was composed originally in some form of Aramaic, although I'm skeptical that we have the data to establish even that. What we have is one complete Coptic manuscript from the fourth century and three very (very) fragmentary Greek manuscripts from around 200, and it's pretty clear that the Coptic must be a translation of the Greek. It would be very difficult to establish with any philological rigor on the basis mainly of one manuscript in a secondary translation, that the original was in a Semitic language, let alone Aramaic rather than Hebrew. But to suggest that it's been demonstrated that Thomas was written in one particular dialect of Aramaic rather than another goes well beyond what one could reasonably hope to establish on philological grounds. I have more on this problem of establishing that a Greek text was translated from Hebrew or Aramaic here (requires paid subscription to access) or here (an earlier and shorter draft, but free).

Wright continues:

What's more, despite efforts to prove the opposite, the sayings of Jesus as they appear in Thomas show clear indications that they are not as original as the parallel material (where it exists) in the canonical gospels. Sayings have, in many cases, been quietly doctored in Thomas to express a very different viewpoint. For instance, when Jesus says, in Matthew, Mark, and Luke, "Render to Caesar the things that are Caesar's, and to God the things that are God's;" the saying in Thomas has an extra phrase at the end: "and to me the things that are mine." What is going on here? In the worldview represented by Thomas, the word "God" denotes a second-rate kind of deity who made the present wicked world, the world from which Jesus has come to rescue people. Thomas and most of the other Nag Hammadi documents represent a worldview known as "Gnosticism,' in which the present world is a dark, evil place from which we need to be rescued by "gnosis," a special knowledge of hidden truth--a world quite different from the Jewish world of Jesus and the four canonical gospels.

I think what really needs to be said here is that this is one understanding of Thomas, but other experts on it (such as DeConick and Koester) place it much earlier, well within the time frame of the four Gospels and Q (if we want to accept Q). I can't see any particular "Gnosticism," as Wright defines it, in Thomas, and his example isn't especially compelling. Certainly such ideas are much more visible in other works from the Nag Hammadi Library. And I'm not a specialist in Jesus sayings, but for what my opinion is worth, I would say that Thomas has one or two noncanonical sayings that could well go back to Jesus (e.g., the Parable of the Assassin, #98) and some of the canonical parables in Thomas do look to me as if they preserve a more primitive version than in the canonical gospels (e.g., the Parable of the Vineyard, #65). In any case, the range of specialist views is considerably wider than Wright indicates here.

Wright concludes:

This doesn't mean, of course, that everything the gospels say is thereby automatically validated. Assessing their historical worth can be done, if at all, only by the kind of painstaking historical work which I and others have attempted at some length but for which there is no room in a book of the present kind. I simply record it as my conviction that the four canonical gospels, broadly speaking, present a portrait of Jesus of Nazareth which is firmly grounded in real history. As the late historian John Roberts, author of a monumental History of the World (1980), sums it up, "the gospels need not be rejected; much more inadequate evidence about far more intractable subjects has often to be employed [in writing history]." The portrait of Jesus we find in the canonical gospels makes sense within the world of Palestine in the 20s and 30s of the first century. Above all, it makes coherent sense in itself. The Jesus who emerges is thoroughly believable as a figure of history, even though the more we look at him, the more we feel once more that we may be staring into the sun.

Well, maybe. But see also my and Mark Goodacre's more skeptical comments (also in response to Wright) here.

UPDATE (14 September): Mark Goodacre has lots of good comments here. And, just to clarify, I don't dispute that Thomas has redacted some Jesus sayings for its own purposes. I just think that some of the material (not necessarily all) looks to have been transmitted independently of the Synoptic Gospels and some of it may give us new or better information about what Jesus said.