Full Contact Christianity

If you look at the end of the Gospel of Mark in any Bible that has some footnotes, you’ll notice something odd: there will be a note about 16:9-20. Depending on which translation you’re looking at, it will either say that verses 9-20 are part of the original text, but some manuscripts don’t have it, or it will say that verses 9-20 are not original, but some manuscripts have them. How do we tell which is right? And why do scholars differ on something like this?

First the good news: we have the original text of the New Testament established with a very high degree of confidence. The amount of manuscript evidence is extraordinary; other than the Hebrew Bible, no other ancient document even comes close. To give you an idea, we have around 25,000 total manuscript witnesses for the New Testament. The next most common ancient document is Homer’s Iliad, with under 1800. From there, it drops sharply: Plato’s works (about 250), Sophocles’ tragedies (under 200), Herodotus (just over 100), Aristotle (around 50), and so on.

In fact, the plethora of evidence is both a problem we have to navigate, and a source of solutions. That 25,000 manuscripts breaks down into 5,800 Greek manuscripts, plus the translations–10,000 in Latin and another 9,000+ in other languages. Every single one of them was hand-copied, and so of course there are scribal mistakes. Because we have so much evidence, most of those mistakes are pretty obvious. If one manuscript leaves out a phrase, but all the rest have it, we can be pretty sure which reading is right.

Most of the variations are nowhere near as consequential as the ending of Mark. For example, the single most common variation is a feature called the “movable nu.” It’s the Greek letter “n”, sometimes added to the end of a word that ends in a vowel. English does this too, with the indefinite article: a vs. an. We say “a house,” but “an apple,” because the latter sounds better to us. Greek has that same feature, but uses it with more words. That’s the movable nu. It doesn’t change the meaning of the word at all; it’s purely there for the sound. That is the single most common textual variant in the Greek New Testament, and it makes no difference at all.

Even most of the variants that are a different word usually don’t change the meaning of the sentence they’re in. What those variations can do, though, is help us establish how different manuscripts are related to each other. Two manuscripts that exhibit the same pattern of variants are very likely to be related to one another somehow. This is now we identify families and subfamilies of manuscripts. We’ll come back to that in a moment.

Now the bad news: there are a small minority of variants that do make a difference, for instance the question of whether Mark 16 ends after verse 8 or continues through verse 20. The art and science of discerning which reading best represents the original text is called textual criticism, and it’s not an easy pursuit. My goal here is to acquaint you with the basics of the discipline, and help you understand why there are different schools of thought within it.

Textual criticism uses a set of criteria to try to sift the available evidence and reconstruct the original text. Along the way, we’ve developed a series of guidelines that help. For example, given two different readings, we generally prefer the shorter one, because we find that scribes tend to err by adding words more often than deleting them. We prefer the harder reading, because scribes tend to simplify the text, not complicate it. We prefer the reading that reasonably accounts for the other readings. We prefer the reading from the best quality of manuscripts, not just the reading we find in the most manuscripts. And so on. Of course this is an art as well as a science, and scholars differ in how they apply these principles to particular cases; hence the differences.

Now obviously none of those guidelines would matter if we could hop into a time machine and go take pictures of the original scroll of Mark’s gospel, straight from his hand, with the ink still wet. If we had a photograph of the ending of Mark’s original draft, then it wouldn’t matter if it was shorter or longer, harder or easier, etc. But we don’t have that, and we’re using these guidelines to try to figure it out from the evidence we do have.

Failing the original, we’re playing a written version of the Telephone Game: we’re looking at copies of copies of copies. What we’re trying to do is establish which copies are the best witnesses to the original. Again, remember the good news: we have a ton of evidence, and overwhelmingly, the manuscripts we have agree with one another.

But what about where they don’t? Of course, if one single manuscript has a reading that none of the other 5,800 manuscripts have, we can pretty confidently chalk it up to scribal error. That’s not really a hard problem. The harder problem is when we have a bunch of manuscripts that have one reading, and a bunch more that read differently, as we do at the ending of Mark.

When you look at a series of such differences, you quickly realize that what you’re seeing is “families” of manuscripts. So now the question becomes, which family or families are more reliable? How do we know?

Remember, we’re playing a giant, centuries-spanning game of Telephone. So all other things being equal, the older a family is, the better–it’s fewer generations away from the source, and therefore fewer opportunities for errors to creep in. If it were that simple, we’d all just take the oldest manuscripts we could get, and call it a day. But age is not the only criterion that matters.

We also want widespread evidence. If we have a family of manuscripts that all come from one specific place, that doesn’t tell us a lot about what they were reading in other places. The manuscripts copied in that one place might have some idiosyncrasies that aren’t representative of everywhere else. So ideally, we want manuscripts from many places that all agree with one another.

We also want the most careful copying we can get, and not every family shows equal care. How can we tell? By tracking how much variance there is within the family. Families where the manuscripts disagree with one another a lot (relatively) show that the scribes weren’t as careful in their copying. Families where the manuscripts disagree with one another very little show that the scribes were very careful.

So putting it all together, what we want is the oldest family, the most geographically widespread family, and the family that shows the least internal disagreement. If we could get all of those in one manuscript family, there wouldn’t be much debate about textual criticism. Alas, that option is not on the menu.

The oldest manuscripts we have all come from Egypt, because papyrus was by far the most common writing material throughout the Roman world, and papyrus doesn’t tolerate moisture well, especially over the long term. It has a tendency to just fall apart if it gets wet at all. So the Sahara desert did us a huge favor by preserving manuscripts that couldn’t have survived anywhere else. These are collectively known as the Alexandrian family of manuscripts. Unfortunately, the Alexandrian manuscripts also exhibit a relatively high amount of internal disagreement within the family.

The other major family of manuscripts is known as the Majority Text. (The use of “majority” here is not tactical; even its ardent opponents like Kurt Aland use the term, because it clearly does encompass the majority of Greek manuscripts.) The Majority family is younger than the Alexandrian, but it is geographically widespread and exhibits far less internal disagreement than the Alexandrian; it gives evidence of more careful copying. In fact, there’s so little disagreement within the family, and it’s so widespread, that Wescott and Hort (early advocates of the Alexandrian texts) felt the need to literally invent a historical event to account for it. They suggested that Lucian of Antioch revised the New Testament text in the early 4th century and forced everyone to accept his version — a ridiculous slander for which there has never been a shred of historical evidence. That idiocy has long since been debunked , and the better explanation is as simple as it is disagreeable to Wescott and Hort’s thesis: the Antiochenes were careful about their copying, and the Alexandrians were notably sloppier. (Just like the Antiochenes were more careful with their hermeneutics, and the Alexandrians more sloppy — but that’s another post.)

Scholars who prefer the Alexandrian family are betting that the older manuscripts of Egypt are a representative sample of the whole Roman world. They don’t have any evidence that’s the case (since we don’t have manuscripts that old from anywhere but Egypt), but that’s the bet they’re making. The hard pill they have to swallow is that the far majority of manuscripts, and the more carefully copied family of manuscripts, disagree with them.

Scholars that prefer the Majority text are making a different bet: that the more carefully copied and more widespread family of manuscripts are better evidence, even though the copies we have are younger. They are betting that the Alexandrian family is a regional text that existed because of sloppier copying practices, and the witnesses we find throughout the Mediterranean world (although they’re younger) are representative of what was available in those places earlier on. The hard pill they have to swallow is that they have no early manuscript evidence, because there aren’t any manuscripts that old in the rest of the Mediterranean world.

So you can pick older, but local and sloppier, or you can pick younger, but more widespread and more careful. Not surprisingly, people disagree about which option is better. There’s also a school of thought that you should take it on a case-by-case basis using internal evidence. Dangerous ground, that, since the operative definition of “internal evidence” is sometimes “what I think it oughta say.” But there’s a case to be made for it nonetheless.