Edward Watts's Riot in Alexandria takes as its object a brief but destructive late fifth-century ascetic and episcopal campaign against the material culture and intellectual and political power of non-Christian believers and philosophers in the region of Alexandria. Watts discusses these episodes as, in part, the result of very traditional tensions between the followers of rival teachers of philosophy and rhetoric in the ancient and late ancient world, and in part as the result of an evolving set of discursive possibilities for the fashioning of episcopal selves among a succession of Alexandrian bishops over the course of more than a century and a half, from the reign of Alexander (313–328) and Athanasius (328–373) to that of Peter Mongus (477–490).

The sole source for the episodes upon which Watts focuses his study is a hagiographic text of the late fifth or early sixth century, Zacharias Scholasticus's Life of Severus. That well-known text describes the destruction and looting of an Isaic temple in the town of Menouthis that followed a fight between a Christian student—one Paralius, whose brothers were members of the nearby Enaton monastic community—and a group of pagan students whom Paralius had offended by mocking local pagan intellectuals and the traditional gods. Immediately following the 485 c.e. raid, Watts argues, the significance and meaning of these events were opportunistically narrativized by one of their primary authors, the embattled Alexandrian archbishop Peter Mongus. In Watts's rendering, these events are easily legible as decisive episodes in the history of a much longer process of pagan decline in Alexandria.

Putting aside for the moment a set of very worrisome problems that underlie the book's organizing claims, there is much here to admire. Primary among the book's merits is Watts's attention to the crucial role of small communities—and in particular the narratives that sustained the identities of these communities—in shaping apparent episodes of religious violence in late antiquity. This is not to say that Watts questions the role of specifically “religious violence” in the events he claims to analyze; indeed, Watts seems to take a sort of innate pagan–Christian animosity very much for granted as the actual basis of the “riot.” In addition to this, however, Watts is unusually cognizant of the manner in which specific historical narratives shaped the ways in which it was possible for individuals and communities to imagine violence, and the ways in which violence fitted into ongoing processes of memory and memorialization. Watts sadly does not seem to avail himself of the plentiful critical treatments of narrative, memory, and memorialization now extant, for memory and memorialization are in many ways what his project is actually about.

Now, as Watts notes, Alan Cameron has recently argued that even among the notoriously perilous hagiographic texts of late antiquity, Zacharias Scholasticus's Life of Severus, with its extreme Christian triumphalist tone and troubling intentional distortions, is even less trustworthy as a resource for historians than most saints' lives. Indeed, Cameron points to the very section of Zacharias's text upon which Watts depends as the evidentiary basis for his narrative of what actually happened in Alexandria and its suburbs in the autumn of 485 as the locus of particularly problematic apparent untruths and occlusions.

Cameron's article appeared in 2007, and in an apparent attempt to salvage his project—begun in 2006, according to the book's acknowledgements—Watts has made one interpretive move in particular and set forward a series of assertions in order to rehabilitate Zacharias's text. The interpretive move is his intriguing suggestion that the section of the text in which one encounters the story of the Christian–pagan student brawl and the raid on Menouthis is in fact a Life of Paralius composed sometime prior to Zacharias's Life of Severus, and rather untidily incorporated into that text. This putative text, which is otherwise unknown, was, Watts insists, part of a campaign by a presumptive Alexandrian Christian student-philoponoi faction to “claim . . . ownership of the riot” (266). The assertions in service of which Watts conjures this spectral text are set forth in an appendix that tries desperately to refute Cameron's arguments concerning Zacharias's text, at least insofar as these have relevance for Watts's own project.

The degree to which Watts succeeds in countering Cameron will rest very much with the individual proclivities of the reader. In fact, however, this question of the specific merits or shortcomings of Zacharias's Life of Severus as a resource for understanding what actually happened “on the ground” in Alexandria in the closing decades of the fifth century is but a busy distraction from the much larger question of the use of hagiographic texts at all to make the sorts of factual claims about inter-communal violence in late antiquity that Watts seeks to make in this book.

The question of the proper use of hagiographies by modern historians is by now an old question, of course. Indeed, few would now argue that historians can afford to ignore such texts; it is a commonplace (and has been for decades) that saints' lives can illuminate in deeply beneficial ways the world of late antiquity as it was imagined by their authors. But the last decade and a half (at least) have seen a decisive and disturbing drift in the acceptable uses of these texts away from their employment as resources for our understanding of what one might call “the late ancient imaginary” to an increasingly troubling dependence on such texts as repositories of “factual information.” This is a trend for which Watts does not bear any particular responsibility, of course. But his participation in this process provides yet another example of an otherwise admirable monograph whose defining claims tragically rest on an evidentiary basis marred by ill-considered handling of hagiographic material. There is much we may simply never know about late ancient history, and this is of course regrettable. But to pretend that we can know more than we actually can know about that world by misusing or naively reading late ancient hagiographic texts is more regrettable still.