to Ben,
Another point in favor of Marcionite priority, I think, is the infancy narrative in Luke. Not only is Luke 3.1 a fine way to start a gospel, as has been noted numerous times over the years, but John the baptist is introduced in Luke 3.2 as if for the first time: "the word of God came to John son of Zechariah". There have been no other Johns mentioned yet that would require this kind of distinction, and Luke 3-24 betrays no knowledge of the detailed events and family connections between Jesus and John in Luke 1-2. A proto-Luke lacking chapters 1-2 and beginning with 3.1 has been proposed many times.
I am aware of the problem. But there are some connections between Lk 1-2 and the rest: Lk 1:80 with Lk 3:2, and then, of course, "Zechariah" (not appearing in other gospels). I do not see why or where more should be expected.
Actually, I think Lk 1-2 is likely to have been composed before the rest of gLuke, as a complement to gMark, and by the same person, in view of the strong feminism in both parts. Then Lk 1-2 was incorporated to Lk 3-24 soon after, when the author decided to write her own gospel.
And, of course, the Marcionite gospel begins with Luke 3.1. So could the Marcionite gospel itself be proto-Luke? Well, it also skips the baptism, with the effect that John himself seems to come on the scene rather suddenly in Marcion, as the subject of an inquiry in Luke 5.33; so suddenly, in fact, as to draw criticism from Tertullian in Against Marcion 4.11.4: "Whence, too, does John come upon the scene? Christ, suddenly; and just as suddenly, John!" So this particular point seems to go in favor of Marcionite posteriority.
Very strong argument indeed in favor of Marcionite posteriority. I will be including that in one of my blog post.
Cordially, Bernard