Williams' reconstruction is methdologically much more rigorous and its results much more radical. Williams allows that both Tertullian and Epiphanius are likely to quote accurately from their copies of Marcion's but demonstrates that nevertheless there are serious difficulties involved in trying to recover genuine readings from their testimony. Not only does Tertullian quote in Latin, (whereas Epiphanius does so in Greek), but also he often passes over material which stands in Luke. Therefore it is difficult to determine whether Tertullian does so because a passage was not present in his text of Marcion, or because it was a passage on which he had no comment to make. Epiphanius, by contrast, does refer specifically to Lukan passages which Marcion has excised. Tertullian also includes variant forms of the same quotations from the Marcionite Gospel, and accuses Marcion of omitting material which does not occur in Luke at all. Williams also cites an instance where each of Tertullian and Epiphanius refers to a passage as missing in Marcion's Gospel that the other appears to indicate is present, and these difficulties in the witnesses to Marcion's Gospel, together with the observation that Marcion's text appears to have been further revised by his followers, lead Williams to conclude that it is extremely difficult to be certain of Marcion's Vorlage. His case is compelling, and it leads him to offer a minimalist reconstruction of Marcionite readings which can be accepted with any degree of surety, in contrast to Harnack's maximalist reconstruction. There are 23 such readings, and only five, argues Williams, are such as to allow reasonable certainty about the words used in Marcion's Gospel, although not necessarily their order.
Central to the current discussion are the questions of whether Marcion knew and abbreviated a text more or less equivalent to Luke; whether or not it bore that name; and whether Marcion abbreviated it to the extent that is alleged. Williams' critique of the scholarly consensus on Marcion's Gospel does allow that modern scholars may with good reason be just as confident as were Irenaeus and Tertullian that Marcion's Gospel was similar to but shorter than Luke, but so too it demonstrates that modern readers have much less reason to be confident as to how much shorter it was. This is consistent with Williams' first conclusion which is that our detailed knowledge of the wording of Marcion's text is extremely limited, a limitation exacerbated because of continuing uncertainty as to the language in which Tertullian knew the Marcionite bible.
Williams' second conclusion, that Marcion's text would appear to have been based on a Gospel that was similar but not equivalent to Luke, comes with three qualifying features
This leads to Williams' third conclusion, that there are difficulties in the belief that Marcion edited his text on theological grounds. [Andrew Gregory, the Reception of Luke and Acts in the period before Irenaeus p. 178 - 180]
I do not at all agree with the author's consistent interpretation of Williams work favoring the Marcionite gospel resembling Luke. This is not what Williams says at all. I just used his work to provide an overview of what Williams says generally. More to follow ...
