Book Technology and the Synoptic Problem

In the ancient world there were two forms of books: the scroll and the codex.

• As everyone knows, scrolls were long rolls that you had to roll and unroll to read. They had the pages attached side by side to make a long, continuous strip.
• Codices, by contrast, were like modern books. They had the pages attached at a spine, allowing you to flip from one passage to another.

These two types of books amounted to different forms of “book technology.” They worked in different ways, as the ways of accessing the material (rolling vs. flipping) indicates.

Before the rise of Christianity, scrolls were by far the most popular format for books. We have almost no references to pre-Christian books being sold in codex form, and pagans and Jews used scrolls almost exclusively when they had scribes copy books for them.

By contrast, Christians were enthusiastic users of codices. This is clear from the surviving second and third century Christian manuscripts, the large majority of which are in codex form. (See Larry Hurtado’s catalogue of early Christian manuscripts.)

Scholars have debated why the codex became so popular among Christians, and they have proposed many possible reasons. However, we don’t know for sure. Codices have some advantages, but they aren’t decisive. (See Larry Hurtado, The Earliest Christian Artifacts, and Harry Y. Gamble, Books and Readers in the Early Church).

The primary reason was likely cultural: Somebody in the Christian community started producing books in codex form—perhaps a very influential edition of a major Christian text (likely something that’s now part of the Bible)—and this became the expected form for books among Christians.

However that happened, the trend must have started in the first century, because it was clearly in place by the second century.

We can’t be certain, because there are other possibilities, but my guess would be that the influential codex that started the trend was one of four things:

1. The first collection of Paul’s letters, which would have included Romans, 1 Corinthians, 2 Corinthians, and Galatians (see David Trobisch, Paul’s Letter Collection)
2. The Gospel of Mark (the first Gospel to be written)
3. The Gospel of Matthew (the most popular of the four Gospels in the early centuries)
4. A bound edition of two, three, or four of the Gospels (something too long to fit in a single scroll)

Is there any way we can shed light on this question?

 

Book Technology and the Synoptic Problem

British scholar Alan Garrow has done a lot of work on the Synoptic Problem—the question of how Matthew, Mark, and Luke are related to each other.

He advocates the “Matthew Conflator Hypothesis” (sometimes called the Wilke Hypothesis), which holds that Mark wrote first, Luke used Mark, and then Matthew conflated Mark and Luke (as well as other sources).

This has been my preferred view for a long time, though I’m open to arguments for other positions.

You can watch Garrow’s videos arguing for this view here.

In his third video, he makes an argument that involves the scroll vs. codex issue. He points out, as have others, that when people were copying from a scroll, they tended to do so in a different way than if they were using a codex. This was because of the physical nature of the book, the ease of maintaining eye contact with the text being copied, and the ability to easily move between passages.

Scroll-users have a greater tendency than codex-users to:

• Paraphrase rather than copy word-for-word
• Keep the material they are copying in the same order
• Switch between sources less often

By contrast, codex-users have a greater tendency than scroll-users to:

• Copy word-for-word
• Change the order of the material they are copying
• Switch between sources

In light of this, what can we say about the Synoptic Problem?

 

What If . . . ?

See Garrow’s third video for the details, but we can say the following:

• Luke seems to have been using a scroll of Mark
• Matthew seems to have been using a codex of Mark
• If Luke used Matthew, then he seems to have been using a codex of Matthew
• If Matthew used Luke, then he seems to have been using a codex of Luke

Garrow argues that these (and other) considerations give us reason to prefer the Matthew Conflator Hypothesis because it allows Matthew to behave consistently: He operates like a codex-user when dealing with both Mark and Luke.

However, if the situation were reversed (a view known as the Farrer Hypothesis) then Luke would be inconsistent: He would operate like a scroll-user with Mark but a codex-user with Matthew.

These facts are certainly consistent with the Matthew Conflator Hypothesis, but I don’t think they provide a particularly strong reason to favor it. This seems to be the weakest part of Garrow’s case, and the arguments he advances in the other videos are much stronger. (There are also arguments that he doesn’t go into in the videos.)

 

Why Not?

So why don’t the above facts give us strong reason to favor the Matthew Conflator Hypothesis?

The reason has to do with the availability of books in scroll and codex forms. Scribes were overwhelmingly used to producing scrolls, and it is unlikely that the codex trend began with the very first Christian books and then instantly dominated the Christian book world. Indeed, Hurtado’s list shows that Christians were still using scrolls for centuries, even after the codex form became dominant.

It is thus likely that the first Gospel to be written—Mark—was originally published as a scroll, something supported by the fact that Luke seems to have used a scroll of Mark.

Further, books in this period were fantastically expensive due to the costs of materials and the hand copying that was involved. This means that, if you were an Evangelist, you would be incentivized to use whatever copy of a prior Evangelist’s Gospel you had—whether it was a scroll or a codex.

You wouldn’t be likely to undertake the expense of having the earlier Evangelist(s) re-copied into your preferred format. And even if you had a rich patron, he might not be inclined to go along with what he would see as a frivolous expense.

Nor would you be likely to slice up a scroll and convert it into a codex. That would produce a very damaged copy as you would be slicing through the joins where the individual sheets were attached, it would be hard to effectively bind them to a single spine, and the book would be extra thick since scrolls were usually written only on one side of the page.

The probability is that you would use the prior Evangelist(s) in whatever format you had.

If Luke was the last of the Synoptic Evangelists to write, the reason for his inconsistency in how he treated Mark and how he treated Matthew thus might simply be due to the fact that he had a scroll of Mark but a codex of Matthew—and he didn’t bother having Matthew recopied as a scroll before he set to work.

I’m not saying that this possibility deprives Garrow’s argument of all force. There is still some value in a scenario that allows the final Synoptic Evangelist to use his sources in a consistent manner. However, I do think the possibility substantially weakens this particular argument for the Matthew Conflator Hypothesis.

 

What Was the Influential Codex?

Can we learn anything from this about what the influential book may have been that kicked of the Christian codex trend?

It seems that we can.

Earlier I proposed four possibilities for what this book may have been:

1. The first collection of Paul’s letters
2. The Gospel of Mark
3. The Gospel of Matthew
4. A collection of more than one Gospel

Since it would have taken time for the codex trend to become established in Christian circles, it is likely that Mark—the first Gospel written—would have initially appeared as a scroll. This is supported by the fact—as Garrow points out—that Luke seems to have used a scroll of Mark. So option 2 is less likely than the others.

The facts we’ve seen also lend some extra probability to the idea that Matthew may have been the influential codex:

• If Luke used Matthew then he apparently did so in codex form, indicating that the most popular of the four Gospels was already circulating as a codex.
• If Matthew used codices of Mark and Luke then he may have been such a codex fan that he did have copies of them made in this form—or he may have had scrolls of them sliced up and re-bound. Either way, he would have been such a codex superfan that he likely then published his own Gospel as a codex.

Further, if Matthew was using codices of Mark and Luke, then—unless he were the kind of codex superfan we’ve just described—copies of Gospels in codex form were already in circulation, and they were probably separate copies—i.e., not bound together as a single volume. This would remove a degree of probability from option 4.

Of the four options, then, Garrow’s analysis causes both Mark and a multi-Gospel collection to lose probability as the influential codex—and Matthew to gain it.

However, Paul’s initial letter collection (Rom.-Gal.) is still a strong possibility.

And there are other options. We are not locked into these four. It could be that Mark was initially a scroll but—by Matthew’s time—it was available as a codex. Or that it was initially a codex but some people made scrolls of it because these were the more familiar book form. The same possibilities are true of Luke.

Either of these thus could have been the book that kicked off the codex trend—and there are other possibilities yet. The key work even could have been an influential work from the Old Testament (a copy of the Pentateuch?) or even an unknown work, though these possibilities are less likely.

Unless dramatic new evidence emerges, this matter will retain its mystery.