BMCR 2023.09.24

The Roman frontier with Persia in north-eastern Mesopotamia: fortresses and roads around Singara

, The Roman frontier with Persia in north-eastern Mesopotamia: fortresses and roads around Singara. Archaeopress Roman archaeology, 96. Oxford: Archaeopress, 2023. Pp. 148. ISBN 9781803273426.



This book delivers precisely what its title promises. It offers an in-depth analysis of the region concerned from the second century to the seventh, concentrating on the period up to 363. As Comfort notes in his Introduction (ch.1), there has been increasing interest in this region in recent years, even if access on the ground has been difficult.[1] To some extent this has been compensated for by satellite photography and access to earlier Corona images (p.6).

The structure of the book is straightforward. Its core comprises two extensive chapters, one devoted to forts and fortresses (ch.3) and one to roads (ch.4). This is preceded by a short introductory chapter (ch.2), ‘Geographical and historical background’, and followed by a brief assessment (ch.5) of ‘The frontier in north-eastern Mesopotamia’. There is a one-page conclusion (ch.6). The book is impressive for its wealth of images, often satellite shots, and its grounding in earlier research; indeed, the volume is dedicated to the memory of those pioneers of aerial archaeology, Antoine Poidebard and Sir Aurel Stein, as well as to David Kennedy. There are many similarities between the present work and that of Kennedy and Riley, Rome’s Desert Frontier from the Air (London, 1990).

North-eastern Mesopotamia, including Singara, Thannuris, Circesium, Castra Maurorum (for which various identifications are assessed) and many smaller places are extensively analysed, their locations probed and the nature of their surviving fortifications (if any) considered. As Comfort notes, it is often hard to be sure of the dating of building work; likewise, whether a structure is Hellenistic, Roman or Sasanian (pp.17-19). Good use of earlier surveys is made and maps and plans from the work of Antoine Poidebard (La trace de Rome dans le désert de Syrie, Paris, 1934) and Louis Dillemann (Haute Mésopotamie orientale et pays adjacents, Paris, 1962) are deployed to good effect alongside the author’s own maps. Usually it is easy to find the places described on one of the three maps by the author, though sometimes one needs to have recourse to those of his predecessors (included in the book).

The volume is more a work of reference than a book that will be read cover to cover. Since it is not possible to cover the wealth of detail it contains, I shall instead try to signal some particularly interesting points about fortifications in the border region that emerge from his discussions. Comfort displays, for instance, a salutary willingness to identify remains as Sasanian structures, making good use of the recent work by Eberhard Sauer and others on the Gorgan wall (Persia’s imperial power in late antiquity: the great wall of Gorgan and frontier landscapes of Sasanian Iran), which have expanded our knowledge of their techniques. Thus, contrary to David Oates (Studies in the ancient history of Northern Iraq, Cambridge, 1968), he argues that a barracks found at Ain Sinu, to the east of Singara, represents a Persian construction (pp.21-4), not a forward Roman base. Following the surveys of T. Wilkinson and D. Tucker (Settlement development in the North Jazira, Iraq, Baghdad, 1995), he also locates other Persian forts around Singara, including Qohbol/Ghobal to the north-east (pp.32-5, cf. 120 on its importance) and ‘Site 54’ further to the north-east, closer to the Tigris (pp.87-8). Comfort, while accepting (rightly) the existence of a clear frontier, highlights (e.g. p.121) the absence of linear fortifications here (like the Gorgan wall) and offers some explanations as to why neither the Romans nor the Persians sought to erect any. He is no doubt right that their effectiveness would have been limited, and that they might have unduly strained the states’ resources and not have allowed for any adjustments in the frontier itself. But one might add also that such work would have been felt to be particularly provocative because from at least 422 there was a treaty (between Rome and Persia) that specifically banned either side from erecting fortifications close to the frontier. In ch.4 there are extensive and helpful discussions of the Tabula Peutingeriana’s itineraries on the eastern frontier, in which Singara appears twice (pp.101-2). The author’s painstaking efforts to reconstruct the roads of the region and to make sense of unclear place names in the ancient sources are to be commended. His own map 3 (p.91) clearly indicates where he believes the Roman roads to have lain.

The relevance of studies such as this to wider work is well illustrated by a new article by Michael Whitby, ‘The location of Mindouos and Roman fortification activity on the eastern frontier in the

years 527-9’, in Byzantinoslavica 81 (2023), in which the author argues that the fortress that Belisarius tried to build in 527 lay not in the vicinity of Dara and Nisibis, but rather well to the east, in the vicinity of Bezabde. It is an intriguing hypothesis, which could indicate a more aggressive attitude on the part of Justinian than is usually supposed at the outset of his reign, for which Peter Heather (among others) has argued. Through investigations such as Comfort’s, it may one day be possible to substantiate such theories – but for the moment, we may note that he makes no pronouncements on the attempted construction of a Roman fort in the 520s in this region.

In conclusion, this is a valuable and well executed book.[2] Already with his similar article, ‘Fortresses of the Tur Abdin and the confrontation between Rome and Persia’, Anatolian Studies 67 (2017), 181-229, he had covered the region immediately to the north of that covered here. We may hope that the author continues his investigations in the coming years.



[1] The author completed a Ph.D. thesis on the roads in Euphratesia, Osrhoene and Mesopotamia from 363 to 602 (University of Exeter, 2009,

[2] In some sections the copyediting has been a little slipshod (typos seem particularly abundant, for instance, on pp.18-19, in the footnotes), while Resaina is spelt thus in the text, but Ressaina on the maps. On pp.30-1 reference is made to a map by Poidebard as fig.3, which is nowhere to be found in the vicinity but is surely to be identified with fig.68 at p.93.