That is embarrassing. But still, hardly a multitude of witnesses. 

Porphyry said
... First, as I indicated earlier, I don't think we have a multitude of manuscripts that give us Lk 3:22 prior to the 5th century. ...   

Correct. As I indicated above only four Greek manuscripts reliably dated prior to the 5th century contain Lk 3,22, namely two papyri (P4 & P75) and two codices (Vaticanus & Sinaiticus). Some might argue that Codex Washingtonianus could conceivably date to the late 4th century. Early attempts at dating it ranged from the 3rd to the 6th century, which will give you an idea of how fuzzy paleographic dating can be.

>> why do they outweigh the multitude of manuscripts contemporary with and prior that more closely align with Mark, on whom we know Luke relied.

First, as I indicated earlier, I don't think we have a multitude of manuscripts that give us Lk 3:22 prior to the 5th century. 

Second, I think named and known theologians are particularly compelling witnesses when they witness a variant reading that they were uncomfortable with; that gives us very good reason to think they really did think their reading was the only reading. 

But finally, I'm not here to argue that the variant was the original. I'm just saying that it is a mistake to pretend the case is open and shut and there is no case to make for thinking the variant was original. We are working with very thin evidence. Multiple, mutually exclusive theories can fit that evidence.

Whether one thinks the variant or the majority reading was original, one should hold one's conclusion with an appropriate tentativeness, given that the early record is both thin and ambiguous. 

And I agree, one needs to look at the data, not what one wants to be true. But I also think it is a mistake to conclude from a passing discussion in one of Bart's popular books that his conclusion is based on his desires rather than data. 

Sometimes Bart oversimplifies things, when engaging the public. Sometimes he seems to let his biases influence his judgement (but don't we all?). Despite all that we shouldn't forget that Bart is a very highly trained textual critic; he wrote his dissertation under Metzger. He spent a long career teaching at R1 universities. He shouldn't be easily written off as a half-wit who doesn't understand the basic principles of textual criticism. 

Porphyry said
>> I can imagine copyists during the 5th c. finding Psalm 2:7 attractive,
>> It demands that we imagine some sneaky little history whereby Luke’s original goes dark for centuries while some copyist’s edit proliferates and dominates the transmission history; and then, at certain points in the 5th c., the original just “pops up” in a few manuscripts, then goes dark again. 
But it is attested long before the 5th century.
Also, keep in mind, we have no complete copy of Luke from prior to the 5th century; prior to that we have a few fragments and incomplete copies; so while a scribe's edit may have dominated the tradition over that time, we can't say that the original "went dark" for centuries--the original reading may have survived as a minority variant that happens not to be attested in the few early MSS we have access to today. And that scenario is perfectly consistent with the data we have, showing that reading attested in a handful of patristic authors from the mid-2nd century to the 5th. 
But I agree with your point about lectio difficilior. We need to know what would have been the more problematic reading for the scribes in question before we can apply the principle. 
  

Long before the fifth century? Which manuscripts are you referring to, and why do they outweigh the multitude of manuscripts contemporary with and prior that more closely align with Mark, on whom we know Luke relied.

When I say "went dark" I mean went dark for us.

It all seems like special pleading to me: we WANT there to be adoptionist christologies, a desire that quickly becomes an assumption that there MUST have been adoptionist Christologies, and now we are reconstructing scenarios to explain away the only real data we have to pave the way for early adoptionist christologies.  surely this is not good historical practice.  We have data, real data: zero variants for Mark's baptism that quote Psalm 2 and scanty and late variants for Luke's gospel.  To take the route of imagining scenarios for why Luke's original was soon edited out by numerous scribes yet kept alive by a few is to indulge in conspiracy theories.  Science my friend, science.  And that means sticking to the data.  What does the data say?

P.S., I am not entirely speaking to just you.  Your final comment is sound and admirable.  

Oh, I think you are right about that. In the case of a patristic attestation, where we know the witness's theology, and can see his discomfort with the line, the argument is especially powerful. The fact Clement was uncomfortable with it--even as he addresses it--makes it pretty much certain that he really thought that was how the verse read, and didn't have any basis to dismiss it as a corruption. 

It is harder when we are dealing with unknown scribes: if you don't know (or can't reasonably surmise) their theology, the lectio difficilior argument really doesn't have legs.

Lectio difficilior argument is easiest when we are dealing with ungrammatical or obviously awkward constructions. If we have a clearly competent scribe (who doesn't make the dumb mistakes of an illiterate, who blindly copies without understanding) and he records a strange or ungrammatical line, that suggests what he records was actually what he had in front of him. 

But I agree with your point about lectio difficilior. We need to know what would have been the more problematic reading for the scribes in question before we can apply the principle. 

I take the point.  But it's not hard to see where an early Adoptionist reading would trouble folks like Clement after the community had largely accepted the doctrine of the Virgin Birth and an Incarnational understanding had become the measure of orthodoxy.  

brown.connor4 said
 
... And the data that we have before us is an overwhelming number of older manuscripts all in favor of no direct quote from Psalm 2 but matching more closely what we have in Mark, which we have good reasons to believe Luke consulted. 

Two older papyri and two or three older codices. Is that really overwhelming? Not when we have two older citations from Justin Martyr, a third from Clement of Alexandria, and a majority of manuscripts from the older Old Latin translation.

To favor the minority view is to indulge in conspiracy theories.  It demands that we imagine some sneaky little history whereby Luke's original goes dark for centuries while some copyist's edit proliferates and dominates the transmission history; and then, at certain points in the 5th c., the original just "pops up" in a few manuscripts, then goes dark again.

Disagree.

Have you looked at the witnesses behind the variants?  If not, you absolutely must.

Good advice:

τὸ πνεῦμα οὖν τὸ ἅγιον, καὶ διὰ τοὺς ἀνθρώπους, ὡς προέφην, ἐν εἴδει περιστερᾶς ἐπέπτη αὐτῷ, καὶ φωνὴ ἐκ τῶν οὐρανῶν ἅμα ἐληλύθει, ἥτις καὶ διὰ Δαβὶδ λεγομένη, ὡς ἀπὸ προσώπου αὐτοῦ λέγοντος ὅπερ αὐτῷ ἀπὸ τοῦ Πατρὸς ἔμελλε λέγεσθαι· “Υἱός μου εἶ σύ· ἐγὼ σήμερον γεγέννηκά σε·” τότε γένεσιν αὐτοῦ λέγων γίνεσθαι τοῖς ἀνθρώποις, ἐξ ὅτου ἡ γνῶσις αὐτοῦ ἔμελλε γίνεσθαι· “Υἱός μου εἶ σύ· ἐγὼ σήμερον γεγέννηκά σε.” (Trypho, 88)

ἅμα τῷ ἀναβῆναι αὐτὸν ἀπὸ τοῦ ποταμοῦ τοῦ Ἰορδάνου, τῆς φωνῆς αὐτοῦ λεχθείσης· “Υἱός μου εἶ σύ· ἐγὼ σήμερον γεγέννηκά σε,” ἐν τοῖς ἀπομνημονεύμασι τῶν ἀποστόλων γέγραπται (Trypho, 103)

For at the moment of the Lord's baptism there sounded a voice from heaven, as a testimony to the Beloved, You are My beloved Son, today have I begotten You. Let us then ask the wise, Is Christ, begotten today, already perfect, or — what were most monstrous — imperfect? If the latter, there is some addition He requires yet to make. But for Him to make any addition to His knowledge is absurd, since He is God. For none can be superior to the Word, or the teacher of the only Teacher. Will they not then own, though reluctant, that the perfect Word born of the perfect Father was begotten in perfection, according to œconomic fore-ordination? And if He was perfect, why was He, the perfect one, baptized? (Paedagogus, 1.6.25 ...)

Sorry, I don't have the Greek of Clement available at home, but I quoted more of the context to illustrate that Clement, more so than Justin before him, seems to struggle with the idea that Christ was begotten at his baptism. 

>> It demands that we imagine some sneaky little history whereby Luke’s original goes dark for centuries while some copyist’s edit proliferates and dominates the transmission history; and then, at certain points in the 5th c., the original just “pops up” in a few manuscripts, then goes dark again. 

But it is attested long before the 5th century.

Also, keep in mind, we have no complete copy of Luke from prior to the 5th century; prior to that we have a few fragments and incomplete copies; so while a scribe's edit may have dominated the tradition over that time, we can't say that the original "went dark" for centuries--the original reading may have survived as a minority variant that happens not to be attested in the few early MSS we have access to today. And that scenario is perfectly consistent with the data we have, showing that reading attested in a handful of patristic authors from the mid-2nd century to the 5th. 

But I agree with your point about lectio difficilior. We need to know what would have been the more problematic reading for the scribes in question before we can apply the principle. 

Stephen said
There is a working principle in textual criticism that the more difficult reading is the stronger. It's not like the laws of physics or carved on stone tablets by the finger of God. But like Occam's Razor, it makes a lot of sense.  
The problem with "begotten",  however you parse it, is that it implies there was a discrete moment when Jesus became the Son of God.  Before that moment, he was not.  The composers of the Nicene Creed certainly knew they had to address it. 
It makes more sense to think Christians went from an Adoptionist view to an Incarnational view than the other way round.  My comment about Mark was pure speculation. Scholars have to follow the data.
  

I acknowledge the principle of "the more difficult reading", but that is not the sole principle or even the most important. We might think a variant difficult to explain, but that is based on our reconstruction of a period, a reconstruction based on scanty evidence.  As I pointed out, I can imagine copyists during the 5th c. finding Psalm 2:7 attractive, since Christology had by this point developed the technical language of Trinitarian terms.  The copyist, who had theological inclinations which he inherited from his context, receives a manuscript with the words "you are my son"... and he thinks, this is just like Psalm 2:7! (obviously he did not think in terms of chapters and verses, but you get my point.)  He's not at all thinking about the words "today" or "begotten" or any kind of adoptionist christology; no, he's thinking about how he might supply a received text with a scriptural quotation that he thinks more fitting.  We might add to his context.  If he were living at a time when "begotten" language had achieved divine status, as it does in the earliest Trinitarian creeds, the copyist might have additional reason for inserting the Psalm. 

But this is all somewhat subjective. A creative mind will always find some reason why some copyist might change x to y.  And so the principle of Occam's Razor actually works against such fancies. It demands we focus on data. And the data that we have before us is an overwhelming number of older manuscripts all in favor of no direct quote from Psalm 2 but matching more closely what we have in Mark, which we have good reasons to believe Luke consulted. 

To favor the minority view is to indulge in conspiracy theories.  It demands that we imagine some sneaky little history whereby Luke's original goes dark for centuries while some copyist's edit proliferates and dominates the transmission history; and then, at certain points in the 5th c., the original just "pops up" in a few manuscripts, then goes dark again.   

Have you looked at the witnesses behind the variants?  If not, you absolutely must.

The problem with "begotten",  however you parse it, is that it implies there was a discrete moment when Jesus became the Son of God.  Before that moment, he was not.  The composers of the Nicene Creed certainly knew they had to address it. 

It makes more sense to think Christians went from an Adoptionist view to an Incarnational view than the other way round.  My comment about Mark was pure speculation. Scholars have to follow the data.

Porphyry said
N-A appears to give "D it; Ju (Cl) Meth Hil Aug". I think that means Codex Bezae; Old Latin; Justin Martyr; Clement of Alexandria (with variation); Methodius of Olympus; Hilary of Poitiers; Augustine. 
  

Stephen said
One interesting question is what we would find in a truly early version of Mark.  We have no variant readings of Mark 1:11 but if the variant in Luke was the original reading as seems likely, you can't help but wonder. Luke looks just like an originally Adoptionist text that was modified to an Incarnational view.  Mark seems clearly Adoptionist in outlook. As Prof Ehrman points out in his book, quoting Psalm 2:7 would have become increasingly problematic to later generations of Christians.
  

I just found a free source (VarApp Step) that lists all the witnesses, and viewing them makes me wonder why we think the variant of Luke containing the word begotten is likely to be original?  The witnesses behind the other variants are overwhelmingly stronger in number, date, and provenance.  And it is not hard to work out why some later copyists would prefer an actual quote from a Psalm rather than the vague allusions presented to them in the received manuscripts, especially as the church became more liturgical.  Let's remember that there are only so many verses in the Hebrew bible that have "Son" and "God" in proximity to each other.  Psalm 2:7 is the obvious choice.  To make "begotten" the operative word seems a bit forced.  If there were numerous other options that that combined son (in a christological sense) and God but lacked begotten, then the choice of Psalm 2:7 would be significant; it would suggest that the quotation was chosen precisely bc it contained the word begotten.  But there aren't.  In fact, I can't recall any passage that relates the Christ to God in filial terms except Psalm 2:7 (and its echo in Samuel).  Let us also remember that the term "begotten" would not have been nearly as problematic in the later centuries of the church.  By that time the term had acquired a Trinitarian meaning completely compatible with a high Christology (begotten not made).  I propose we are reading way too much into a single word of the one verse that they had available to them.