Reid: Thanks for also raising the question of the defense of including phonoi.
I work on editing legal contracts (with an amateur interest in theology), so I figure I was hinting at a possible defense of inclusion all along. Little can be airtight in arguing these things. A case can be made for the defense.
First, the fact that Paul is written that way in Romans, with pthonou phonou in that order, means it's Pauline--his kind of language, if he wants to use it. The possibility of actually dating Paul's writing Galatians near Romans--for those who put James aside--means it's close to his mindset, too. I also introduced what, in the last analysis, we think Paul wrote. That should be broken into two parts: What he had to include or avoid, and what we think he usually includes or avoids.
I don't have my vonSoden, and just gave away my 1964 Merk, so I was checking NA27, Metzger and Aland. (Since I've been concentrating on New
Vulgate decisions, the UBS experts will have lots more detail than I recall.) So we do need Wieland's commentary on the rest of the New Testament!
Second, the fact that we can isolate early readings (before the standardization of the Byzantine text and those more critical students of Origen like Lucian, Hesychius, Jerome) means we can't look to early nonByz Greek to supply us with just one early reading. It's also good to admit that the early scholars did know more than we do about where they got their more trusted manuscripts, how old they were, and even who wrote them--so we don't know everything they knew. Just, we've learned their biases. This mitigates any argument that the shorter text is earlier because more frequent or favored in our early witnesses, as though the double phthonoi-phonoi did not yet exist (which I like to say means we can't give weighting the witnesses too much weight, once we're faced with the same choice the "best" witnesses
Third, a simple phthonou or pthonoi in each case could have been mistaken for phonou/phonoi, and then conflated. Or, in reverse, the double term could have just as easily been shortened to a single by h.t. Here is where I think of the later scribes as witness, which you asked about. This could only be used as a doubly negative criterion: We cannot argue the difficulty of a reading as a likely cause (it remains possible, of course) if the majority overall don't show the difficulty. Difficulty is best suspected as a cause of majority readings. To distinguish early and late difficulties--Marcionite worries, say, versus Arian--is good, however, in a case like this, where both readings existed early on. Usually, that applies when there was a mechanism (say, Alexandrian scholarship, or defensive episcopal authority) to create a standard that systematically thwarts simple scribal tendency.
And that's all the appeal to scribal difficulty
is worth in this case--except, as I say, that we should probably rule out the ideological difficulty in proscribing murder. Psychologically, the times before Constantine were less peaceful for Christians, perhaps. But the strong word, phonoi, could stick more with male scribes.
The textual question leaves us with drawing up a possible original (what we think Paul wrote) and testing it by a likely explanation of how it came out two ways. In some way, all the evidence can help, which is why I included the general scribal majority.