I indicated in my previous posts that there are serious methodological problems with using patristic evidence in NT textual criticism.   That was no less true for the church father I chose for my dissertation, Didymus the Blind, than for any other.   But there are always ways to deal with problems, and that proved to be the case here as well.

For one thing, I noted in my earlier post that scribes of the Middle Ages often changed the texts of the church fathers that they were copying in order to make their quotations of scripture coincide with the form of text known in their own (the scribes’) day.   With the OT commentaries of Didymus that was not so much a problem, because the manuscripts discovered in Egypt of his writings dated from the sixth century, not much later, and they may well have been copies of the originals, or of early copies of the originals.  And there is evidence that the quotations were not much changed: these quotations agree quite strikingly with the readings found in early manuscripts of the NT, not with later forms of the text.

I also noted that church fathers often quoted from memory without looking passages up in a manuscript before writing it down, and that since memory can be faulty, this could create problems.  In some cases, though, a father such as Didymus would quote a passage several times in different writings.  If these quotations are all the same in their wording, that would suggest that it was the form of the text that the author knew, as he had learned it, presumably from a manuscript.

Moreover, in Didymus’s case we are dealing with a unique situation.   He had been blind from a very young age, and so could not read a manuscript, or consult it the way sighted people could.   Metzger raised this with me as a very serious problem: if Didymus couldn’t read, then how could we know what his written manuscripts said?

I figured out a way around this problem.   For one thing, in antiquity Didymus was known to have a truly prodigious memory.  He memorized texts.  Lots of texts.  Including biblical texts.   The way he did  it, apparently, was to have a scribe read passages to him, and he would commit them to memory.   So when he quoted the Bible, it was always from memory.  On the other hand, since he quoted it in precisely the same way so many times, one could be reasonably assured that he quoted it as he had heard it.

But since he wasn’t actually seeing manuscripts, wouldn’t that suggest that we can’t know what his manuscripts looked like?  Moreover, if he dictated his writings to scribes (which he obviously had to do), isn’t it possible that they would write down the Scripture passages that he quoted in the form that they themselves were more familiar with, if his quotations differed from the ways they had learned the texts themselves by looking at manuscripts?

In my dissertation – then the book – I argued that even if this did happen, it would not have the slightest effect on the analysis.   What I was interested in was not actually the text that Didymus had memorized per se.  I was interested in knowing what manuscripts available in fourth-century Alexandria must have looked like, using the newly-discovered commentaries of Didymus as a way of getting to see which readings they contained.   Even if Didymus’s scribes/secretaries had substituted the form of text that they knew from their manuscripts for the form of the text that he dictated to them (either accidentally or inadvertently), by analyzing those quotations I would *still* know what fourth-century manuscripts in Alexandria looked like – since these scribes too were living and fourth-century Alexandria and had biblical manuscripts available to them in that time and place.

The way to proceed then was to isolate all the quotations found in the writings of Didymus and see what they could tell me about the development of the textual tradition in that lively center of intellectual life, both pagan and Christian, Alexandria at the time the best known – and best – manuscripts had been produced there.