I am struggling with a celebrated New Testament passage, and I’ll use this blogpost as a way of working out my thoughts. It relates to the issue of the origins of the gospels as we have them, a topic about which I have blogged often enough in the past. As in most such cases, the issue at hand has a vast literature attached to it, and my main goal is not to be overwhelmed.
The text in question opens the Gospel of Luke, and in the NIV, this prologue reads “Many have undertaken to draw up an account of the things that have been fulfilled among us, just as they were handed down to us by those who from the first were eyewitnesses and servants of the word.” My question is: what did he mean by “many,” polloi, and how many of those accounts do we possess in any form? To put this in context, assume that Luke is writing around 90-95. What narratives were available at or before this time?
The word “many” is open to interpretation, and perhaps it is a rhetorical flourish. Quite possibly, Luke is using a standard rhetorical device in ancient history writing, as exemplified by a scholar like Livy. But as an exercise, assume he means it literally. So what could “many” mean? A dozen? Twenty? Could it just mean five or six? Based on what we have from that era, or what we plausibly can reconstruct, I can think of a fair number of accounts of Jesus that likely did exist by that point, and I will list them shortly. By the way, Luke says that many have put together such accounts, not that he has seen any or all of them personally.
Yet there is a problem. Luke speaks of to compile (anataxasthai) an account (diegesis), and that last word can be translated as narration or declarations. But diegesis is an unusually strong word, which means not just an account, but more like an account that is thorough or comprehensive. At first sight, that would mean something organized in the sense that we think of as a gospel, although with no necessary implication about length.
But what could that term cover? Might the term apply to something like the Gospel of Thomas, which is a collection of sayings without any particular sequence or organization? In fact, Thomas is nothing like this early, and I just use it as a hypothetical. To take an example that Luke assuredly did know, would he apply the term to the reconstructed Q gospel source, which lacks most of the narrative structure we know, not to mention not referring to the crucifixion? On the other hand, could it apply to accounts of the crucifixion, which might have circulated independently of other sayings or narratives? How about a source that described Jesus’s miracles, without much other narrative, or a crucifixion? The answer to any of those questions might be that yes, Luke was indeed applying the term diegesis to such items, but he would be stretching the language.
How Many Gospels?
Venturing into the realm of speculation, what might Luke have listed among the “many” in the 90s? Stressing that I don’t accept every theory on which these various claims are based, let me offer some plausible candidates:
–Gospel of Mark (definite).
–Gospel of Matthew (very debatable).
–Q Gospel (assuming that Luke’s terminology allowed for it).
–Gospel of the Hebrews. At least some sections of this appear to be strikingly early.
–A Syriac Gospel? Eusebius (iv.22) has an odd passage about the mid-second century historian Hegesippus, which (depending on how we construct the sentence) might be read as saying that he got some things from the Gospel of the Hebrews, and the Syriac [Gospel], Ebraious Euaggeliou kai tou Syriakou. I offer this for what it is worth, but have no sense of what exactly we are dealing with, or its date.
–The Egerton Gospel (Pap. Egerton 2), which includes an otherwise unknown Jesus miracle. We have no idea what this was originally called, but I would make a case that it is from either G. Hebrews or a closely allied text, but that is for a future blogpost.
*The “lost gospel” in Pap. Oxyrhynchus 840. Also perhaps part of G. Hebrews, or a related text?
–Cross Gospel or something like it, as hypothesized by John Dominic Crossan; although actually, not accepted by many others. Possibly some early version of the Gospel of Peter?
–The Logia or Sayings of Jesus, which according to Papias of Hierapolis (c.115) were originally in Hebrew, and which Matthew compiled (synetaxato). Then, everyone interpreted them as best they could. This is not, obviously, the same as the Gospel of Matthew we know.
–The L source, or the materials distinctive to Luke but not found in the other synoptics.
–Two or more components of the Gospel of John as we presently have it, including the Signs Gospel. John as we know it includes multiple earlier layers of composition and editing, and it is far from agreed exactly what different units might have been incorporated. Two is a minimum figure, three is more likely.
Based on texts we know or can hypothesize, we can get our list up to a dozen or so. Is that enough to be “many”?
To complicate the matter, it is wise not to assume that the texts that Luke knew in his day were identical to what we know from later generations. Some stories and sayings existed in a free floating form until they ended up either being incorporated into some canonical text, or else in an apocryphal or extra-Biblical writing. If not in a gospel, then they migrated into some other work, like one of the very extensive Acts associated with the various apostles – a vast literature.
To give one illustration, Papias recorded many intriguing remarks about the church of his time, and the texts they knew. In the fourth century, Eusebius tells us that Papias “also gives another story of a woman who was accused of many sins before the Lord, which is to be found in the Gospel according to the Hebrews.” Many scholars have suggested that this is the pericope adulterae, the famous story of the woman taken in adultery (not, in fact, of “many sins”) which was not found in early manuscripts of the Gospel of John, but which later ended up as our canonical John 7.53-8.11. Do note that Eusebius does not say this story came from the Gospel of the Hebrews itself, but rather that it was found there in his day. It might have come from another writing entirely, perhaps even a full gospel that is now lost, or a collection of miracles or sayings.
In a brilliant reconstruction, William L. Petersen demonstrated that something like this pericope originated in a very early Jewish-Christian gospel tradition before finding refuge in John, as well as the Gospel of the Hebrews, but was also referenced in the influential second century text known as the Protevangelium (c.170).
It is an interesting question how many other once free-floating sayings or passages from otherwise lost texts or gospels ended up in scriptures that we know well. Also, how in fact we would ever recognize such literary fossils. We recognize the pericope adulterae because it was added to John at a rather later date. But what about something that attached itself much earlier? How would we spot it? When we study the remnants of authentic lost gospels – such as the Egerton papyrus – we are struck by how close their style and approach echoes that of canonical texts, either John or the Synoptics. In that way, they would not stand out.
This has serious implications for tracing the early history of texts we know. We might for instance find some early writer like Ignatius, say, quoting a passage that we know from Matthew, and we assume that this proves that he knew Matthew at that date. It might. But it might also mean that the writer knew the passage from some other now-lost gospel, or collection of sayings or stories.
I wonder whether such examples help us understand Luke’s prologue. When he talked about a diegesis, an organized narrative, perhaps he was drawing a distinction between miscellaneous collections of a few such stories or pericopes, and more ambitious works that tried to synthesize them into a more substantial entity? In that case, he would certainly be including something like Q on his list, which helps us fulfill the criterion of “many.”
One implication of all this is how we describe fragments of ancient manuscripts. We often read that a particular papyrus, say, includes a portion of the gospel of Matthew or John. What we would better say is that the fragment includes text that as we have it, is preserved in the gospel in question, but without any particular conclusion about how it might have circulated at the time. Conceivably it wasn’t part of John or Matthew just yet. I confess that this was in my thoughts last week when I stood in the Rylands library in Manchester, England, gazing with awe at the legendary P52, one of the oldest of all surviving manuscript fragments of the New Testament. Actually, that is not a great example in itself, as recent research dates that fragment as closer to 200AD rather than 125, so it virtually certainly is part of a complete John codex. But the point remains. Fragments of text did migrate between gospels and gospel traditions.
How Gospels Went Missing
I have mentioned the gospels and other sources that we happen to know, to some degree. But what about those we simply do not know? How likely is it that a gospel could vanish without trace? Plenty of early Christians surely attempted to write something, and most of their efforts have vanished. But what about more substantial products like gospels, or “accounts”?
Just for the sake of argument, assume that around 80AD, there existed a hypothetical text that I am going to call the Gospel of Justus. Luke knows of it and refers to it among the “many,” but it was lost at an early date. I can’t prove that such an event never happened, but there are limits to the kind of oblivion we are suggesting. If Luke knew “Justus,” then other people presumably encountered it, and it was known in the Christian literary circles of – shall we say Antioch? From a variety of sources, we really do have a good sense of the gospel texts that enjoyed any circulation whatever, even if what we know does not go far beyond their title and very general character.
Even gospels that are “lost” in their entirety persisted for decades or centuries in different forms, as I explored in my 2015 book The Many Faces of Christ. If both church and state struggled to suppress a pseudo-gospel in the Roman Empire (say) then those efforts had no impact whatever outside that realm, and a great many suspicious or controversial texts flourished for centuries in the Persian or Islamic realms.
How Mark Almost Went Missing
So could “Justus” have vanished without trace? As a gospel text under that name, certainly. But it is unlikely that its content vanished wholly, as opposed to merging into some other text, possibly even a canonical one. The question is whether we would recognize it as a gospel if we saw it.
Again for the sake of argument, assume that Luke’s “many” accounts included more than I have listed earlier, say twenty, and we don’t have them today. A predictable modern response is to assume that these others were suppressed because they were subversive or heretical, and some might have been. But gospels also faded from use when they became superfluous.
To understand this, look at the Gospel of Mark, which was more or less entirely absorbed into Matthew and Luke. Of course, we know in retrospect that Mark also survived as an intact work, but it was a near run thing. Early manuscripts of Mark are far less common than those of the other canonical gospels, and it is not difficult to see how it might have vanished. This was not because it contained controversial content, but because any reader who wanted that content could find it easily elsewhere. So why bother to keep Mark? The main reason was the tradition linking it directly to Peter, which carried enormous weight. But assume that Mark actually had vanished, and some modern archaeologist dug up a portion of a manuscript in the sands of Egypt. Would they recognize it as an independent work, or just assume that it was a rogue fragment of Luke or Matthew?
Even better, assume that Mark had vanished, so that we had no separate attestation for Mark and Q. Would scholars just assume that both Matthew and Luke drew heavily on one lengthy common source, which in reality included two separate texts. Would they be willing or able to separate out Mark/Q?
Has a like fate befallen other early texts that have vanished, but left their ghosts in a canonical gospel? Are they hiding in plain sight? How many of those “many” accounts that he mentions did Luke himself cannibalize in his own gospel? To take a likely contender, the Mary material came from somewhere.
I end here with a note of some confidence. I believe that we can make a fair attempt to reconstruct roughly what Luke was talking about when he spoke of those “many” narratives or accounts that existed in his time. So no, I don’t think we have lost vast numbers of gospels that would tell stories too different from what we can piece together from other sources.
This goes beyond the specific issue addressed here, but on Luke generally, there is a fine commentary on The Gospel of Luke by Amy-Jill Levine and Ben Witherington (Cambridge 2018). Throughout the book, we see how these two top-flight scholars differ on specific questions and debates, such as the existence of Q, or whether the “Theophilus” of the Prologue was an actual person. They discuss whether the gospel’s first two chapters might have been a later addition, added perhaps to counter Marcionite ideas about Judaism and the Old Testament. (24-25: Ben Witherington rejects this idea, and I am with him on that one). You can, so to speak, eavesdrop on two brilliant people having a highly learned conversation, which never ceases to be mutually respectful. The whole book is a great exercise in seeing how serious scholarship works at its best. Spoiler: many academics often don’t work like that in practice.