maryhelena wrote:I found a follow up post rather interesting - only available to subscribers though...
Bart Ehrman
External Evidence in Textual Criticism
The number of manuscripts that support one reading or another. You might think at the outset that this would be the most important factor – if lots of manuscripts have one of the readings and only a few have the other, then the majority should win, right? Well, not actually. Here’s why.THE REST OF THIS POST IS FOR MEMBERS ONLY. If you don’t belong yet, GET WITH THE PROGRAM!!!
Suppose the original copy of a book (say, Mark) was copied by two scribes, call them Scribe A and Scribe B. Both of them made some mistakes, but B more than A. And suppose the copy made by A is itself then copied by two other copyists, but copy B is copied by twenty. And then each of the two copies of A is copied three times, and each of the copies of B is copied ten times. What then? There are 9 copies with the A kind of text, but 221 copies of the B kind of text. And suppose then that A, B, and the first copies of each were all destroyed, so that all you have are 206 copies, 6 of them reading one way in one of the verses and 200 reading another way. Does the fact that far many more copies have the text of B make it better than the text of A? Quite the contrary, B was a worse copy than A. In this hypothetical case, the 6 manuscripts have a better text than the 200.
This is a purely hypothetical case, of course, but it shows clearly that because of the exigencies of copying, you can’t base a textual decision on the numbers of manuscripts that attest one form of the text over another.
The age of manuscripts that support one reading or another. This is almost always conceded to be a better criterion. If one form of the text is attested in a bunch of early manuscripts, nearer in time to the original, and the other is attested only in later manuscripts, then the later reading – even if it’s in the majority of surviving manuscripts – is more likely to be a later alteration of the text.
This rule cannot be applied woodenly, however. Suppose you have a hypothetical manuscript that we’ll call X and another that we call Y. Manuscript X is from the 5th century and manuscript Y is from the 7th century. You would think that Manuscript X is therefore innately better, since it’s 200 years earlier. But what if Manuscript X was itself a copy of a manuscript made in the 4th century, but Manuscript Y was a copy of a manuscript made in the 3rd century? In that case, the text of the 7th century manuscript is actually older than the text of the 5th century manuscript. Isn’t this fun?
The Geographic Distribution of Manuscripts. Suppose you have one form of the text that is found in manuscripts (and, say, versions and church fathers) that come from all over the place: Rome, Alexandria, Gaul, Spain, North Africa, and one that comes from only one place, say, Caesarea. Then it seems likely that the widely distributed reading is more likely original: the other could well be a local variant.
These are simply some of the external criteria that scholars use to decide which form of the text, if there are two or more forms available, is most likely the original one (well, two of the criteria, since the first is not of much use). There are other kinds of external evidence that are a bit more tricky, but they don’t matter much for my point here. If you have two forms of the text, and one of them, say, is found in almost all of the early manuscripts from a wide range of places, then that is more likely the original form over against the other, if it is found only in later manuscripts or only in manuscripts of a certain region.
But there is much more to a textual decision than looking at manuscripts. That’s what I realized full force early on when I decided that being a textual critic – at least a competent one – requires the scholar also to be an exegete. More on that in subsequent posts.
So, while the words are important - so too the necessity for interpretation...the forest is as important as any of the trees! The story, the context, has to be taken into account.
