Martin Shedd
My talk outlines manuscript evidence and internal authorial references to argue against the currently dominant, unitary authorship theory for the Historia Augusta. In 1889, Hermann Dessau published a thesis arguing that the Historia Augusta was not the work of the six authors whose names appear over various of the short biographies in our extant manuscripts, but rather a single writer, masquerading as a collective (Dessau, 1889). This argument relies on several passages within the work where a writer given one name claims to have composed a life attributed to a different one of the six. Sir Ronald Syme championed the unitary authorship theory, following Dessau’s principles, and brought the theory to general acceptance (Syme 1971, 1983). Computer studies since have alternately decided in favor of unitary authorship and the uncertain possibility of six authors (Marriott 1979, Gurney & Gurney 1998). Each return to the question of authorship considers only the possibilities of one author or the six as named. I show that by removing the six names to which the lives are attributed as permissible evidence, the case for unitary authorship becomes weaker than a continuation model of authorship.
I begin by setting out the internal evidence for the six authors as they appear in our manuscripts for the Historia Augusta. The author never once references another writer as a collaborator or even gives direct indication that a multi-authored work should be considered probable. On the contrary, when three of the six attributed names appear within the work, it is in contexts no different from the falsified sources that the author cites. Proceeding to the manuscript tradition itself, I show that there is disagreement within the manuscripts about the authorship. I will make clear that the only instance in which the lives attributed to an author perfectly match the works of theirs cited in the Historia Augusta is a construction of later scholarship, not based on the manuscript evidence.
I then turn to other evidence for later invention of authorial attributions in the Middle Ages. Using collections such as the Greek Anthology, I discuss how unattributed works were frequently ascribed to authors either historical or fictitious. The presence of falsified names within the Historia Augusta, particularly those for whom no works are specified or those whose works closely aligned with the short biographies, provides ample material for making such ascriptions.
Returning to the Historia Augusta itself, I set out the conclusions that were drawn about authorship solely on the assumption that the six names were an original part of the work. When that assumption is removed, the data supports the conclusion that a group of historically well informed lives within the work, often called the Primary Lives following Mommsen (1890), were written separately from the rest of the work as it survives. Though the secondary lives frequently refer to material in the primary lives and reference them as sources, the primary lives never directly reference the secondary lives. Stylometric analyses aiming to show either unitary or multiple authorship have only been able to show results tending toward their conclusion by suppressing data.
Accepting the premise that the six names are not original to the work leads to the conclusion that the Historia Augusta represents a literary method in vogue at the end of the fourth century AD: continuation. From the continuator of Cassius Dio to Jerome’s continuation of Eusebius’s Chronicon, continuation of historical works was becoming a common endeavor in the period in which the Historia Augusta was written. Understanding the work in this context changes our understanding of it from being an unprecedented literary farce to being an example of current literary taste. I close with evidence from medieval book catalogues for unattributed collections of imperial biographies that may describe the work now known as the Historia Augusta in a fuller, continuous form that begins with Julius Caesar, as the author claims his work should.