Late Antiquity, roughly spanning the period from the ascension of Diocletian in 284 to the decline of the Merovingian Empire in 751, has always appealed to me because it sits between familiar worlds. It is close enough to classical Rome to feel recognizable, yet distant enough that the political, military, and cultural landscape often looks very different from the one suggested by older schoolbook narratives of “decline and fall.” Although the period is still sometimes lazily grouped under the label of “the Dark Ages,” the work of scholars such as the great Peter Brown has shown how misleading that framing can be: Late Antiquity was not simply an age of collapse, but one of transformation, adaptation, and cultural reinvention.

The eastern frontier of the Roman Empire is especially fascinating in this regard: a zone of cities, fortresses, roads, armies, allies, and rival imperial ambitions, where small episodes can reveal much larger problems of power, evidence, and interpretation.

Historical reconstruction is not simply copying events from a textbook. The further back one goes, the less often “what happened” is a clean, agreed-upon sequence. Historians may have only a battle name, a few commanders, scattered dates, brief or biased accounts, and modern debates hinging on small details. Modern books often turn this material into smooth campaign narratives, but the evidence underneath is usually fragmentary, contradictory, and shaped by authors’ agendas, audiences, and knowledge. This is best explained in a case study: The twin battles (?) of Thannuris and Mindouos.

The events associated with Thannuris and Mindouos on the eastern frontier of the Roman Empire in the late 520s are a case in point. They belong to the early Iberian War between Rome and the Sasanian Empire. The war began over influence in the Caucasus, especially in Iberia (roughly today's Georgia), but quickly drew in the fortified frontier of northern Mesopotamia, where both empires tested each other's defenses. In the contested zone between Roman Dara and Sasanian Nisibis, Roman fortification efforts led to clashes which difficult to reconstruct in detail.

The specific difficulty is how to connect the surviving accounts of 528. Procopius, in History of the Wars, describes a Roman attempt to build a fortress at Mindouos, a Sasanian response, Roman defeat, the capture of Coutzes (a frontier dux from Phoenice Libanensis), and the destruction of the works. Pseudo-Zachariah places the disaster opposite Thannuris and describes a Sasanian trap with trenches and obstacles. Malalas confirms a major defeat and names commanders. Procopius' The Buildings complicates things further by hinting that Thannuris may not have been a single location.

Later historians have tried to reconcile these accounts, but no consensus has emerged. Are Thannuris and Mindouos two names for a single confused event, or two separate actions? Did Procopius misattribute locations, or have later interpreters merged distinct episodes? Or are modern scholars overcomplicating what sixth-century sources saw as one event?

This is what I call the "Thannuris-Mindouos problem".

Note: I refer to various sources, both primary and secondary. All works I used for research and analysis are listed in the bibliography section at the end. To improve readability for the purpose of this essay, I refrain from using footnotes or inline citations.

To understand why these accounts diverge, it helps first to look at the frontier environment in which the events unfolded.

A frontier built on pressure points

To see why this problem matters, consider the sixth-century eastern frontier: not a simple line, but a complex landscape of cities, fortresses, roads, and contested zones. Roman and Sasanian strategies relied on controlling movement and fortified positions, not just on battles.

Two names are especially important: Dara and Nisibis. Nisibis had long been one of the great frontier cities of the region and was in Sasanian hands. Dara, fortified under Anastasius in the early sixth century, became one of the major Roman counterweights to Nisibis. The existence of Dara itself had been controversial because it placed a powerful Roman fortress close to the Sasanian frontier. It was not just a defensive site; it was a statement of strategic intent. Any additional Roman fortification in the same broad zone could be read by the Sasanians not as a minor local improvement, but as part of a larger attempt to alter the balance of the frontier.

Roman-Sasanian Frontier in the 5th and 6th century (Cplakidas / CC BY-SA 3.0)

This helps explain why a projected Roman fort at or near Mindouos could have mattered so much. If the Romans were building a new fortified point between Dara and Sasanian territory, or in a zone that threatened movement toward Nisibis, the Sasanians had good reason to respond. The work may have protected Roman construction elsewhere, supported operations against Arab raiders, strengthened the approaches to Dara, or extended Roman control over a sensitive route. Even if the exact location is disputed, the strategic logic is clear enough: on this frontier, building could be an act of war.

In 527-528, several Roman operations aimed to strengthen this frontier, including at Thannuris and, more famously, Mindouos. These overlapping efforts, and the way sources preserved or simplified events, have made later reconstruction difficult and prone to forced coherence.

The primary sources: different windows onto the same frontier

The main sources do not give us one identical narrative. They preserve different pieces of the same frontier crisis, and each must be used for what it is strongest at.

Procopius is the most famous of the relevant authors and provides the clearest account of the Mindouos fortification framework. He was close to Belisarius and may have had access to informed testimony from his circle, but we cannot say for sure if he witnessed the disaster at Mindouos or Thannuris. In his account in De Bellis (On the Wars), Justinian orders Belisarius to build a fortress at Mindouos, near the frontier. The Sasanians gather to obstruct the work; reinforcements arrive under Coutzes and Bouzes; a battle follows; the Romans are defeated; and the Sasanians destroy the unfinished fortification.

That sequence is coherent on its own, but it lacks the most detailed tactical explanation of the disaster. Pseudo-Zachariah Rhetor supplies that detail. His account, associated with Thannuris, focuses less on a named fortification project and more on the battlefield: a prepared Sasanian trap, an aggressive Roman advance, disorder among the leading elements, cavalry escape, and heavy infantry losses. It is the strongest surviving explanation of how the Roman defeat may have unfolded.

John Malalas is less tactically detailed, but he confirms that a major defeat occurred in 528 and preserves additional information on the command. Procopius’ On Buildings (De Aedificiis) adds a different kind of complication by noting two Thannourioi, one large and one small, suggesting that “Thannuris” may not have denoted a single, unambiguous location on the map.

Each source, therefore, preserves something valuable: Procopius’ On the Wars is strongest for the fortification project and the name Mindouos; Pseudo-Zachariah is strongest for the battlefield trap and the action opposite Thannuris; Malalas is useful for scale, commanders, and the sense of a major defeat; and Buildings matters because it complicates the local topography. None of them alone can settle the whole problem.

One battle, two battles, or overlapping episodes?

The question, then, is how these pieces relate to one another.

The simplest answer is a one-event model. On this reading, the Roman attempt to build at Mindouos and the battle opposite Thannuris are two perspectives on the same failed operation: construction provoked a Sasanian response, the Sasanians prepared the ground, the Romans attacked, the army was defeated, and the unfinished work was destroyed. This has the advantage of economy and fits the way ancient authors often preserve different aspects of the same event.

Bellisarius and his wife Antonina (AI rendering with DALL-E 3, based on the mosaic in the Basilica of San Vitale in Ravenna)

The danger is that this model can become too smooth. Pseudo-Zachariah describes a prepared wilderness trap opposite Thannuris, not a construction-site clash, and Malalas does not tie the defeat to Mindouos or fort-building. If everything is forced into Procopius’ framework, the distinctiveness of the Thannuris tradition may be flattened.

At the other end is the hard two-battles model, most clearly associated with Hughes. Here, Thannuris and Mindouos are treated as separate defeats: one battlefield disaster preserved by Pseudo-Zachariah and Malalas, and another failed fortification episode preserved by Procopius. This takes the differences between the sources seriously, but the evidence for two fully developed set-piece battles is not equally strong.

A third possibility may be the most cautious: one major battle, plus one or more related fortification episodes. Under this model, the battlefield disaster of 528 belongs most securely to the Thannuris tradition, while the Mindouos material preserves a real Roman building project whose exact relationship to the battle remains uncertain. It is less elegant than a single clean narrative, but it avoids both forcing all evidence onto one battlefield and inventing two large battles where the sources may justify only one.

The geography problem: where was Mindouos?

The second major issue is topography. Procopius says that Mindouos lay “on the left as one goes toward Nisibis” (ἐν ἀριστερᾷ ἐς Νίσιβιν ἰόντι) (Proc. Wars 1.13.2, trans. Greatrex 2022, 176). Much turns on how this phrase is understood and, above all, from where the imagined movement begins. Is Dara the assumed point of departure? Is the route one of the known roads in the Dara-Nisibis region? Is the phrase precise geographical guidance, or a looser directional marker?

Roman Dara, Ambar, and Ammodios near the c.528 frontier and Persian Nisibis (from Lillington-Martin, “Forts on Frontiers Facing ‘Βάρβαροι.’”)

The near-Dara model

One model places Mindouos close to Dara, around the area of Kasriahmethayro, south-east of the city, and near quarries. This is the model most strongly associated with Lillington-Martin and later supported by Greatrex. It keeps Mindouos within the immediate Dara-Nisibis frontier system, gives practical meaning to the presence of workmen and quarrying, and offers something concrete: visible remains near a plausible frontier setting.

Lillington-Martin’s case is especially important because it is topographical and archaeological, not merely literary. He identifies a small site near the quarries around Kasriahmethayro, approximately 6.3 kilometers south-east of Dara, with views across the Syrian plain and toward the frontier zone. The site has often been described as a fortlet or watchtower rather than a large city or major fortress. That makes sense if it guarded workers, quarries, roads, or a forward observation point; it becomes harder if Mindouos is imagined as a major strategic counterweight in its completed form.

Locations of evidence for structures near the C6th Roman-Persian frontier (from Lillington-Martin, “Forts on Frontiers Facing ‘Βάρβαροι.’”)

One answer is that the visible remains need not represent the whole intended project. If the Romans were still building, a small fortlet, watchtower, guard post, or first stage of a larger unfinished fortification could leave less impressive remains than the political and military crisis it provoked.

Possible site of Mindouos and 6 km SE of Dara (37° 8'09.56" N, 41° 0'18.25" E) (from Lillington-Martin, “Forts on Frontiers Facing ‘Βάρβαροι.’”) 

The near-Dara model also gains force from Procopius’ wording if Dara is understood as the implied starting point for the road to Nisibis. In that case, a location south-east of Dara and near the frontier makes good sense. Greatrex’s later commentary strongly supports this reading, and Lillington-Martin’s topographical case gives it a physical anchor.

The Tur Abdin alternative

Whitby’s later critique pushes in a different direction. If Mindouos was intended as a major strategic counterweight, a small fortlet near Dara may seem underwhelming, especially since Dara itself was already the dominant Roman fortress in the region. Whitby, therefore, looks farther east and south-east, toward the southern edge of the Tur Abdin and the broader route system leading to Bezabde, Cizre, Nisibis, and the Tigris.

This alternative broadens the strategic geography. Mindouos would not be a mere appendage of Dara, but part of a more ambitious attempt to control routes across or around the Tur Abdin and threaten communications between Nisibis and the east. The drawback is that this model is harder to anchor in a single archaeological candidate. It solves some strategic problems by moving the site into a more important corridor, but creates new identification problems.

The map, therefore, does not settle the problem. It multiplies the questions.

Thannuris, Tanurin, Tell Tuneinir, and El-Thuneir

Thannuris enters the problem above all through Pseudo-Zachariah’s battle narrative. His account points toward Thannuris, often associated with Tanurin or Tell Tuneinir; the name also appears in relation to the modern El-Thuneir / al-Thunayr area. This southern or desert-facing identification fits the tactical setting well: a wilderness opposite Thannuris, prepared trenches, pits, stakes, and openings, cavalry able to escape, and infantry trapped in broken ground.

The Tell Tuneinir or El-Thuneir/Tanurin zone gives the account room to work. It provides a known place, a desert-facing landscape, and enough space for a prepared battlefield trap rather than a tight construction-site clash.

Satellite image of the archaeological site of Tell Tuneinir (Sat image by Google Earth) 

But the problem is not solved by simply saying “the battle was at Thannuris.” Procopius’ Buildings refers to two Thannourioi, one larger and one smaller. That opens the possibility that “Thannuris” may have functioned as more than one local identifier. If so, two questions become unavoidable: which Thannuris did Pseudo-Zachariah mean, and could Mindouos have been connected with the smaller Thannuris?

One speculative reconciliation is possible, though it should be clearly labeled as my own hypothesis rather than as an established position in the secondary literature. If Lillington-Martin’s near-Dara site were not only Procopius’ Mindouos, but also the smaller Thannuris mentioned in Buildings, then Pseudo-Zachariah’s Thannuris and Procopius’ Mindouos might not be as far apart as they first appear. The battle could have taken place in front of, or opposite, a smaller Thannuris/Mindouos site near the Roman construction zone.

The weakness of this model is obvious. No ancient source states that Mindouos and the smaller Thannuris were the same place, and I am not aware of a modern scholar who makes that exact identification. It also requires a double identification: Mindouos would have to be equated with the smaller Thannuris, and Pseudo-Zachariah’s Thannuris would have to refer to that same smaller site rather than to the better-known Tanurin/Tell Tuneinir/El-Thuneir area.

The alternative is to keep the Thannuris battle and the Mindouos fortification apart. In that case, the Thannuris/Tanurin/Tell Tuneinir/El-Thuneir zone remains the stronger candidate for the battle described by Pseudo-Zachariah, while Mindouos remains a separate and unresolved fortification problem. This avoids the double equation, but it leaves us with a more fragmented campaign.

The battle itself: a Sasanian trap and a Roman disaster

Amid the geographical uncertainty, the tactical shape of the Thannuris battle is somewhat clearer, especially in Pseudo-Zachariah. The Sasanians used prepared ground - trenches, pits, stakes, and openings - to channel the Roman attack. The Romans advanced aggressively, perhaps too aggressively. The leading elements became disordered. Some mounted men escaped, but the infantry was trapped and suffered heavily.

This is a valuable battle narrative because it is not merely a list of casualties. It explains the mechanism of defeat. The Sasanians did not simply overpower the Romans in open combat. They used field engineering and deception. The battlefield itself became a weapon. Obstacles broke momentum, separated arms, and turned Roman aggression into vulnerability.

That tactical picture also fits broader patterns of frontier warfare. The Sasanians were not passive defenders. They were capable of preparing ground, exploiting Roman command problems, and using combined defensive and offensive action. A trench system does not need to imply a static siege. It can function as a trap: a way to disorder an attacking enemy before a counterstroke. If the Romans expected to push through or clear a blocking position, openings in the obstacle system might have encouraged precisely the kind of forward rush that the Sasanians wanted.

The Roman side appears to have suffered from command fragmentation. Belisarius was present, but he may not yet have had the kind of authority later associated with his name. Other commanders, including Coutzes and Bouzes, were involved. Malalas preserves further names and gives the episode a wider command structure. Such an army could be formidable, but it could also be difficult to control, especially if several commanders acted with more courage than coordination.

The fate of Coutzes, captured in Procopius' account and treated differently elsewhere, is one of those small but telling discrepancies that remind us how unstable the dossier can be. It should not dominate the interpretation. The more important issue is not exactly what happened to one commander, but why the sources attach the Roman disaster to different frameworks: Mindouos and construction in Procopius; Thannuris and a prepared battlefield in Pseudo-Zachariah; a broader campaign defeat in Malalas.

The infantry's fate is especially significant. Pseudo-Zachariah emphasizes that mounted survivors escaped, while the infantry suffered the worst losses in the trench system. That gives the battle a particular structure: not a simple rout of an entire army at once, but a collapse in which mobility determined survival. Cavalry could flee. Foot soldiers caught in broken ground, among trenches and obstacles, could not. This also explains why the battle was remembered as a disaster, even if not every Roman contingent was destroyed.

Belisarius before fame

The episode is also important because of Belisarius. Later history remembers him as Justinian’s great general, the victor of Dara in 530, the conqueror of the Vandals in 533-534, and the commander who entered Rome in the Gothic War. But in 528, he was still early in his career. The defeat at Thannuris-Mindouos occurred before his reputation had fully formed.

This creates a problem for interpretation. When a later-famous commander appears in an early defeat, sources and historians may be tempted either to shield him from responsibility or to become suspicious precisely because later reputation can distort the narrative. Bury, for example, stresses that Belisarius held only a subordinate position and was not responsible for the disaster. That may be true, but Procopius also had reasons to present Belisarius carefully.

Court of Emperor Justinian I. The figure next to the right shoulder of Justinian is usually interpreted as Belisarius, the one next to the left shoulder is often interpreted as the eunuch Narses, who later played an important role in the Gothic War. (from the mosaic in the Basilica of San Vitale, Ravenna, Emilia-Romagna, Italy)

This does not mean Belisarius should be blamed for the defeat. The safest view is that he was involved in a major Roman operation that went badly wrong, but that command was shared or fragmented. The disaster resulted from a combination of Sasanian preparation, Roman aggression, difficult terrain, and weak unity of command. That explanation does not require Belisarius to be either villain or victim; it places him within the operational reality of the frontier.

Modern interpretations: a spectrum of solutions

Modern interpretations fall along a spectrum. At one end are synthetic accounts, such as Bury's, which effectively combine the evidence into a single campaign narrative: a Roman fortification effort, a Sasanian response, a Roman defeat, and the destruction of the works. This approach has the virtue of clarity, but it can smooth over the differences between Procopius, Pseudo-Zachariah, and Malalas.

Other reconstructions are more explicitly harmonizing. Syvänne, for example, offers an integrated military reconstruction linking the near-Dara identification of Mindouos to the battlefield disaster, emphasizing troop composition, command confusion, Sasanian preparation, and Roman tactical collapse. This is useful because it asks concrete military questions, but it also shows the danger of making the reconstruction more detailed than the evidence can firmly support.

Lillington-Martin and Greatrex give the near-Dara model its strongest topographical form. Lillington-Martin provides the physical candidate near Kasriahmethayro and the quarries south-east of Dara; Greatrex's later commentary supports that reading and treats it as the most persuasive solution. Whitby, by contrast, challenges whether such a small site can bear the strategic weight assigned to Mindouos and looks instead to the southern or south-eastern Tur Abdin and the wider route system between Nisibis, Bezabde, Cizre, and the Tigris.

Hughes represents the clearest hard two-battles approach, separating Thannuris and Mindouos into distinct defeats and arguing that Procopius conflated them. That solution takes the differences between the sources seriously, but it may also impose too neat a division on the evidence, which does not yield two equally detailed battle narratives.

The modern debate, then, is not a simple contest between one right and one wrong answer. It is a set of choices about what to privilege: narrative economy, tactical detail, topographical evidence, strategic plausibility, or source separation. Each solution explains some things well and leaves other problems behind.

What can be said with confidence?

So, after all this, what can we responsibly say?

First, the events of 528 were part of a wider Roman-Sasanian crisis on the eastern frontier, connected to Roman attempts to strengthen or extend their fortified positions along the routes between Dara, Nisibis, the Tur Abdin region, and Arab allies, and to the contested routes of northern Mesopotamia. 

Second, the sources agree on the broad failure even when they differ on the details: the Romans suffered a serious defeat, commanders were killed or captured, Belisarius escaped, and the Roman effort failed. The battle tradition belongs most securely to Thannuris, while Procopius preserves essential information about the Mindouos fortification project.

Third, the locations remain disputed. The near-Dara identification around Kasriahmethayro has strong practical and topographical arguments, while Whitby’s south-eastern Tur Abdin alternative has strategic force but less secure archaeological anchoring. The two Thannourioi mentioned in Buildings add another layer of uncertainty.

The relationship between Thannuris and Mindouos therefore remains unresolved: one event-complex, adjacent or overlapping locations, one battle plus a related fortification episode, or two distinct defeats. That is not a disappointing conclusion. It is an honest one.

Why this matters

The Thannuris-Mindouos problem is not famous. It is not one of the great iconic battles of antiquity. Yet precisely because it is obscure, it is useful. Famous battles often come wrapped in familiar narratives; lesser-known frontier actions expose the machinery of reconstruction more clearly. We can see the seams.

The case also reminds us that ancient battles were not always remembered for the same reasons we want to reconstruct them. We may want exact locations, force ratios, command hierarchies, and battlefield layouts. The sources may care more about imperial policy, moral judgment, local suffering, the fame of commanders, or the theological meaning of disaster.

There is, therefore, no such thing as simply “following the sources” if the sources do not line up. One must decide how to follow them: whether to privilege the famous historian or the less polished chronicle, the source with the place-name or the source with the tactical mechanism, the archaeological candidate or the broader strategic logic. Every answer has consequences.

Conclusion: the value of uncertainty

The Thannuris-Mindouos problem may never be resolved to everyone’s satisfaction, but that is not a failure of the exercise. It is exactly what makes the case revealing. The evidence does not assemble itself into a single obvious story. It has to be weighed, compared, and sometimes held in tension.

For the general reader, this can feel frustrating. We want to know where the battle was, what the armies did, and who was responsible. But uncertainty is part of understanding the past. The Roman-Sasanian frontier was complicated, the sources are partial and human, and the narratives we build from them remain provisional.

From a wargamer’s point of view, this uncertainty is not a problem to be hidden away in the notes. It is also an opportunity. A confused command structure, an unclear sequence of events, and competing versions of the battlefield allow different interpretations to be explored rather than simply described. One reconstruction might emphasize reckless Roman aggression, another divided command, another the effectiveness of the Sasanian trap, and another the possibility that later accounts compressed more than one episode into a single disaster. The point is not to pretend that one version is beyond doubt, but to make the uncertainty visible and historically plausible.

What matters is that the question forces us to see the process behind the narrative. And that process is where much of the real historical work begins.

Bibliography

Primary Sources

Procopius. The Wars of Justinian. Translated by H. B. Dewing. Revised and modernized by Anthony Kaldellis. Indianapolis: Hackett Publishing, 2014.

Procopius of Caesarea. The Persian Wars: A Historical Commentary. By Geoffrey Greatrex. Cambridge: Cambridge University Press, 2022.

Procopius. Buildings. Translated by H. B. Dewing. Loeb Classical Library 343. Cambridge, MA: Harvard University Press, 1940.

John Malalas. The Chronicle of John Malalas. Translated by Elizabeth Jeffreys, Michael Jeffreys, and Roger Scott. Melbourne: Australian Association for Byzantine Studies, 1986.

Pseudo-Zachariah Rhetor. The Chronicle of Pseudo-Zachariah Rhetor: Church and War in Late Antiquity. Translated by Geoffrey Greatrex, Robert R. Phenix, and Cornelia B. Horn. Liverpool: Liverpool University Press, 2011.

Secondary Sources

Bury, J. B. History of the Later Roman Empire from the Death of Theodosius I to the Death of Justinian. 2 vols. London: Macmillan, 1923.

Greatrex, Geoffrey. Rome and Persia at War, 502-532. ARCA Classical and Medieval Texts, Papers and Monographs 37. Leeds: Francis Cairns, 1998.

Greatrex, Geoffrey, and Samuel N. C. Lieu. The Roman Eastern Frontier and the Persian Wars, Part II, AD 363-630: A Narrative Sourcebook. London: Routledge, 2002.

Hughes, Ian. Belisarius: The Last Roman General. Yardley, PA: Westholme, 2009.

Syvänne, Ilkka. Military History of Late Rome, 518-565. Barnsley: Pen & Sword Military, 2024.

Whitby, Michael. “The Location of Mindouos and Roman Fortification Activity on the Eastern Frontier in the Years 527–529.” Byzantinoslavica 81, nos. 1–2 (2023): 7–20. https://doi.org/10.58377/byzslav.2023.1.

Not directly related to the research above, but a great introduction to the eponymous "World of Late Antiquity."

Brown, Peter. The World of Late Antiquity: AD 150-750. New York: W. W. Norton, 1989.