The Roman Empire of Ammianus Marcellinus. With a New Introduction (review)

When it first appeared in 1989, John Matthew’s compendious book on Ammianus was greeted as the culmination of a quarter-century and more of renewed interest in the fourth-century ad historian, who had by that point become a poster-child of the new Late Antiquity. The present reprint edition adds nothing to the original except a five-page preface and “the correction of a few specific errors” (xii). The republication has been heartily welcomed, and rightly, since it is still the standard work on numerous fronts, and was becoming a little hard to find. Nevertheless, even if some twenty years are not nearly enough to render the work obsolete, they have allowed the “pause for digestion” that a prominent reviewer of the first edition prescribed.¹ In his new preface, Matthews only has space enough to touch upon a few points that have become controversial in the intervening years, as well as to mention some of the major contributions in Ammianean studies subsequent to his own. He is generous in his recognition of others, but also quite stubborn in cleaving to the conclusions he first promulgated in 1989. A brief review is no more a place to enter into a detailed examination than is the preface of a reprint edition, but there is space to mention some of the controversies provoked by Matthew’s book, as well as to situate its general approach to history and historiography within the subsequent debate.

Early reviewers were quick to point out that in summing up modern scholarship, Matthews chose to do so against a very broad backdrop (the author provides a charming anecdote of the circumstances of this decision in his preface, x). The book falls into two sections: Res Gestae, which in ten chapters provides a quasi-narrative modern account of the period covered by Ammianus’ extant text, and Visa Vel Lecta, eight thematic chapters on aspects of government, geography, warfare, and culture in Ammianus’ fourth century, with special focus on his description of that world. The combination of recapitulation and specialist studies, accompanied by unusually expansive endnotes and a highly detailed index, furnished the reader with a wide context for Ammianus’ history, as well as an examination of its peculiar qualities. In this it was compared by more than one reviewer to Syme’s magnum opus on Tacitus,² and, before the completion of the Budé edition with commentary in 1996 (and with the commentary begun by De Jonge still ongoing under Den Boeft et al.), it rendered many of the services expected of a commentary. By any standards, this was and continues to be a major contribution. At its first appearance, the book went a very long way toward bringing Momigliano’s “lonely historian” out of the cold for the more general scholarly audience (especially in the English-speaking world) and of redeeming him from the charge of derivativeness popular since the nineteenth century and of linguistic barbarity leveled at him since Gibbon. Not lost in this vast general project, though, were a number of specific arguments that elicited particular discussion, especially on the dating and composition of the History and on Ammianus’ own origin and identity.

On dating, Matthews was credited with at last doing away with the old theory that Ammianus published Books 26–31 of the extant work separately from the portion ending at Book 25, in which his hero Julian dies. He also securely set the terminus ante quem for composition of the whole at ad 391 or the year after (the destruction of the Alexandrian Serapeum in ad 392, of which Ammianus appears ignorant, being the key determinant). Furthermore, he seriously questioned the assumption that Ammianus came from a curial background and the extent of Tacitus’ influence on his historiography.

On the matter of Ammianus’ own origins in Antioch, a supposition long upheld by identifying the historian with the Marcellinus addressed in Libanius’ letter 1063, Matthews was...