I have disliked our citation system for a very long time. Whenever I have published on the non-canonical materials and tried to NOT italicize their titles (to match the convention of citation for canonical materials), the editors won't allow it. This is a terrific example of the way we "privilege" canonical materials. So much of "canonism" in our field is insidious (like many -isms in our world).
I want to push us to reflect further about the need to stop dividing our texts along canonical boundaries when writing our histories of early Christianity. I keep reading in remarks on this blog and in other blogs how much better the NT texts are for the reconstruction of Christian Origins. But this is not necessarily true. What texts we use will all depend on the question(s) we are trying to answer. For instance, the Gospel of Judas tells me a heck of a lot about what Christians were thinking about in 150 CE than any NT text can. And the Gospel of Judas is not a historiography, and has nothing whatsoever to do with the "historical" Jesus or the "historical" Judas if there was one. But it has a big story to tell about the choices Christians were making theologically and practically, and how they understood their own history.
I feel very sad that our field is so dominated by the "historical" Jesus - as if his quest IS the quest for Christian Origins. Frankly, he has less to do with Christianity's origins than most like to think. My scholarly investigations have led me to the point of concluding that Jesus had very little to do with Christian origins. The growth and survival of Christianity, in fact, probably happened in spite of he, his disciples and his brother James whose movement was Israel-centered, faithful to the Torah, and radically apocalyptic. The enmity that the author of Mark directs towards the Twelve apostles is startling. Have you ever really read what he has to say about them? He loathes them, while praising the same agenda as Paul.
So certainly we cannot a priori assume canonism. But even after we have investigated our texts for "age" and "historical reliability," we have to be very careful. Why? Because these should not be our determinants. Our questions should.