When I discovered that the books On Christian Instruction were not completed I chose to complete them rather than to leave them as they were and go on to the re-examination of other works. Accordingly I completed the third book. . . . Then I added a new book, and so completed the work in four books. (Retractationes 2.30; cited in Gamble, 136)
How interesting that Augustine can say he "discovered" his work was not complete; he should always have known this since he, apparently, never completed it. (I'm reminded of a certain breed of student who might claim not to have known that his or her research paper had been left undone on their harddrive.) It looks like Augustine released an early, unfinished version of De doctrina into circulation, and only later decided to return to his earlier project and finish it. From this, Gamble offers the following analysis:
The question whether a text of De doctrina in only two books constituted its first edition cannot really be answered. On the one hand, from the start the scope of the work was intended to be larger, so that a text in only two books would always have been incomplete; but, on the other hand, the work had apparently been corrected and given out for circulation. Hence it was not an edition in the modern sense—a complete, definitive text—but functionally it was an edition, for the work, though incomplete, was allowed to circulate and to be copied. This case shows how misleading the term edition can be when applied in the conditions under which texts circulated in antiquity. A text qualified as an edition only when it had been emended and released by the author for copying. (137)
The issues swirling around Augustine's De doctrina and Retractationes 2.30—and Gamble's analyses of those issues—raise interesting questions about the earliest texts of the synoptic gospels (Matthew, Mark, and Luke).1 The dominant view of the gospels identifies Mark as the earliest written gospel, with Matthew and Luke dependent on Mark. One of the questions gospels scholars regularly raise is how (or why) the church preserved Mark once it had Matthew and Luke. This question is exacerbated by the fate of Q (if Matthew and Luke are independent), which was (apparently) allowed to return to dust.
What value—if any—might there be for thinking of Mark in terms similar to the incomplete, two-volume "edition" of De doctrina, and the other two synoptics as "completions" of the former? On the one hand, given Mark's abrupt beginning in 1.1–13 and—even more dramatically—the sudden ending at 16.8, the gospel of Mark does bear every indication of being an incomplete work that always anticipated "the rest of the story." On the other hand, however, Mark is its own carefully crafted, artistic, sophisticated account of the life and death of Jesus of Nazareth, and if Matthew and Luke originally intended their gospels to displace Mark they clearly failed.
I suspect these types of questions have already been set to Mark, Matthew, and Luke. But has anyone broached these questions from the perspective of book-production and -dissemination in the early Roman empire (and/or in late-Second Temple Judaism)?
1 Of course, De doctrina, the earliest books of which were written c. 397 and was completed c. 426, is over three hundred years later than the gospels. The dynamics of text-production—including its technologies and its cultural value—were certainly not identical in the first and fourth centuries. But according to Gamble, the testimonies of Augustine and Jerome "indicate that the traditional procedure of the first three centuries remained in effect" (132).