Proto-Gospel Theories 39
Proto-Gospel Theories
Partly in order to fill in some of the gaps left with the two-source hypothesis, partly because of early Christian testimony, and partly because of internal indications, various scholars have posited the existence of an earlier edition of each of the synoptic Gospels. Lachmann, one of the first proponents of the two-source theory, worked from the assumption of an original gospel, arguing that Mark was the closest to that original. Some modern scholars, noting the problem of the minor agreements and some elements in Matthew and Luke that are difficult to explain if these evangelists were using the canonical Mark, have suggested that one or both may have used an earlier edition of Mark. 52This hypothesis must remain doubtful. The minor agreements are not all of the same kind; many cannot be explained by positing dependence on an "Ur-Mark.” 53More basically, we must question the assumption that dependence on a different source must be used to explain all the changes Matthew and Luke have made in their Markan source. Source criticism takes too much on itself when it presumes to explain every line in Matthew and Luke with reference to a written source. The influence of eyewitness accounts, various oral traditions, and the evangelists’ own theological purposes must be allowed. When these factors are taken into account, the need for an Ur-Mark disappears.
Much more popular has been the thesis that Matthew wrote an earlier edition of his gospel. In this case, however, the motivation is only partly a more satisfactory explanation of synoptic relations; more important is the apparent reference to such an earlier edition in the second-century remark of Papias (quoted by Eusebius, H.E. 3.39.16): "Matthew collected the oracles [τα λόγια (ta logia)] in the Hebrew language, [ Έβραΐδι διαλέκτω (Hebraidi dialekto)] and each interpreted [ηρμήνευσεν (hermeneusen)] them as best he could.” 54If Papias is referring to a gospel written in Aramaic or Hebrew, he must be referring to an earlier, Semitic edition of our Greek Matthew, since later church fathers appealed to Papias to prove the priority of canonical Matthew. It has been popular, then, to suppose that a Semitic Matthew was the first gospel written; that Peter, or Peter and Mark together, used that edition in composing Greek Mark; and that Greek Matthew then made use of Mark. 55The stubborn tradition that Matthew was first written in Aramaic or Hebrew, along with the widespread belief in the early church that Matthew was the first gospel, renders the hypothesis of a Semitic "first edition” of Matthew attractive.
Clearly, however, if such an edition existed, the canonical Matthew is not simply a translation of this Semitic original. Matthew does not read like "translation Greek”; more important, Matthew has probably, as we have seen, used Greek Mark in composing his gospel. And there are other problems for the supposition that Mark has used a Semitic-language Matthew. Strong early tradition views Mark as composing his gospel on the basis of Peter’s preaching (see the introduction to Mark’s gospel). But then it is hard to imagine how Mark could also be using an earlier edition of Matthew. Moreover, Papias may not be referring to a gospel at all (see the discussion in the introduction to Matthew). All in all, the hypothesis of an earlier, Semitic-language edition of Matthew cannot be certainly proven or disproven.
The evidence for a proto-Luke comes from within Luke itself and rests on three considerations: (1) the greater amount of special material in Luke in comparison with Matthew and Mark; (2) Luke’s tendency to "go his own way”; even in material shared with Matthew and Mark (esp. in the passion narrative); and (3) the fact that Luke includes material from Mark in blocks rather than scattered evenly throughout the gospel. These phenomena have suggested to many scholars that Luke had composed a first edition of this gospel with the use of Q and L (his special material) and then later integrated Mark into this initial work. 56While the case remains unproven, 57the hypothesis is an attractive one (see discussion in chap. 4).