“Many Roman authors provide a manipulated version of events. Unfortunately, for this reason, most of the details of these battles have been lost to time.”

By Byron Waldron

DURING THE mid and late third century AD, the Roman Empire suffered from a nasty combination of civil wars, economic decline and foreign invasions. In the case of external enemies, the Sasanian Empire came into being in western Asia, and a slew of newly attested peoples and confederations emerged in North Africa and beyond the Rhine and Danube. These included the Franks, Alemanni and Goths.

Not since the Middle Republic had the Romans faced so many formidable opponents at the same time. Perhaps unsurprisingly, the empire suffered a string of defeats on a scale not seen since the Hannibalic and Cimbrian Wars.

However, for the most part, the events of the period have been transmitted to us through the accounts of Latin and Greek authors writing within the Roman Empire. For this reason, most of what we are told about Roman defeats in the third century is written from the perspective of the defeated.

Are Roman accounts of these defeats to be taken at face value?

Let us consider seven Roman defeats from the third century, to see how Roman authors engaged with some of the darkest moments in their history.

The Battle of Mesene, 232/3

According to the Roman historian Herodian (6.5-6), who was writing in the mid third century, in the 230s the emperor Severus Alexander launched an ambitious three-pronged invasion of the Sasanian Empire. However, the three operations failed to effectively support one another, and one of the armies was surrounded by the Sasanians in southern Mesopotamia, most likely in Mesene. This army was reportedly wiped out by Ardashir I, the energetic founder of the Sasanian regime.

Over the course of the campaign the Sasanians also suffered significant casualties, both because one of the Roman armies ravaged the Sasanian territory of Media, and because of the determined resistance of the Romans in the battle of Mesene, who likely fought hard as they knew they faced annihilation. As a result, both sides were unwilling to continue fighting, until Ardashir renewed the war three or four years later.

Except for two Byzantine histories (Cedrenus 450; Zonaras 12.15), Herodian’s account is the only surviving source to mention the defeat in Mesene or acknowledge that the overall campaign was indecisive. In the abbreviated account of the lost history of Aurelius Victor (Historiae Abbreviatae 24.2), a Roman historian writing in the fourth century, it is claimed that Alexander defeated Ardashir. Later authors followed Victor’s version of events (e.g. Festus, Breviarium 22; Historia Augusta, Alexander 55). However, Herodian lived during the events he describes and has provided by far the most detailed surviving account. Perhaps Romans were not inclined to favour or remember such a negative account of the campaign of Alexander, usually regarded as a good emperor.

The Battle of Misiche, 244

In 244 the emperor Gordian III invaded Sasanian-held Lower Mesopotamia, and according to an inscription produced by Ardashir’s successor, the illustrious Shapur I (Ka’ba-i Zardost 6-9 in Maricq 1958), the new Sasanian king confronted the Romans in a great pitched battle near Misiche, in which Gordian was slain. The Sasanian priest Abnon likewise acknowledges the battle in an inscription, and various Sasanian monuments depict Gordian’s corpse beneath Shapur’s steed.

In contrast, most Roman accounts do not mention a battle (e.g. Oracula Sibyllina 13.13-19; Aurelius Victor, Historiae Abbreviatae 27.8; Zos. 1.18.2-19.1). Rather, they variously claim that Gordian was murdered by either his own soldiers or his praetorian prefect Philip the Arab, who succeeded him as emperor.

This version of events does not accord well with Philip’s apparent deification of Gordian, or the fact that, while on the return march from Sasanian territory, the new emperor erected a very large cenotaph to Gordian’s memory on the Roman side of the border (Ammianus [23.5.7-8] was an eyewitness to the cenotaph).

To so honour the emperor whom either he or the soldiers had slain would likely have been self-destructive, especially considering how volatile the soldiery could be in this period. He could ill afford to upset his base of support.

A Greek source tradition, preserved through a small selection of Byzantine texts, partially backs up the Sasanian claim (e.g. Georgius Monachus 32; Zonaras 12.17; Scutariotes 36). According to this account, Gordian fell from his horse during a battle with the Sasanians and suffered a wound, which festered and took his life.

With the emperor having died on or near foreign soil, senatorial elites in the empire’s west were likely inclined to view Philip with suspicion. He was only the third person to not be a senator when he became emperor, and military rebellions were common in this period. But returning to the Greek account, the relative banality of the detail of a festering wound lends credence to its historicity (MacDonald 1981; cf. Potter 1990, 201-212.).

Operationally, Rome must have been on the backfoot. In the treaty that followed Gordian’s death, Philip agreed to pay a massive ransom of half a million gold coins to secure the safety of his army, as well as pay an annual subsidy to the Sasanians. He also appears to have vowed not to assist Armenia against the Sasanians, and he possibly agreed to abandon some fortresses in the Middle Euphrates and beyond Singara (Potter 1990, 221-225; cf. Edwell 2008, 173-181).

The Battle of Beroea, 250/1

In 250/1 a coalition of Goths and probably other peoples invaded Thrace, led by the wily king Cniva. Near Beroea (Augusta Traiana), they launched a surprise attack against units serving under the emperor Decius and destroyed an auxiliary force. The defeat was severe enough that it forced Decius to withdraw from Thrace and rebuild part of his army. This allowed Cniva to besiege and capture the provincial capital Philippopolis without imperial interference.

However, the details of the battle at Beroea are unclear, as we are told of this battle through a fragment of a lost account by the third century historian Dexippus (Codex Vind. Hist. gr. 73). The sixth century historian Jordanes (Getica 102), who read Dexippus, also mentions the battle, but his short account does not preserve the details alluded to in Dexippus’ fragment.

Most accounts of the Gothic war do not mention the battle, and the sixth century historian Zosimus (1.23.1) even claims that Decius had enjoyed nothing but success until his eventual defeat at Abritus (a reference to an initial success against the Goths outside Nicopolis). But there is no reason for Dexippus to have invented the defeat. The Romans suffered a defeat that temporarily paralysed Decius’ efforts against Cniva, with disastrous results, but the battle drifted out of Roman memory, probably because Roman authors paid it little attention. Perhaps this is unsurprising, as the defeat was overshadowed by the next battle in this list.

The Battle of Abritus, 251

Roman sources could not hide the magnitude of the disaster at Abritus. Cniva outplayed Decius, using a deceptive three-part formation to lure the Roman army into a swamp. Here, Decius and the Imperial army were butchered. Even the camp treasury was seized, as coin finds in north-eastern Europe show (Bursche & Myzgin 2020).

The third century historian Dexippus narrates that the battle was fought in Roman territory, in Moesia Inferior (Dexippus, Chronicle, fragment 17 in Martin 2006; Jordanes, Getica 103). Here, Decius was attempting to intercept the Goths, who were departing the empire with Roman captives in their possession, including people of high rank.

However, some later accounts instead claim that the battle took place in ‘barbarian’ territory (e.g. Aurelius Victor, Historiae Abbreviatae 29.4; Eutropius 9.4; Epitome de Caesaribus 29.2). It appears that Roman authors sought to present Decius’ death in a more heroic light by having the ambush take place in mysterious foreign lands rather than on home turf (Potter 2018). Zosimus’ account (1.23.1-2) attempts to reconcile the two versions, with Decius attempting to catch the Goths not before they cross the Danube, but before they cross the Don (in present-day Russia).

Zosimus also claims that Decius’ successor, Trebonianus Gallus, collaborated with the Goths, communicating with the enemy and persuading Decius to march into the trap (Zosimus 1.23.2-3; see also Zonaras 12.20). Roman collusion is not necessary to explain Cniva’s success. Accounts of his earlier victories at Beroea and Philippopolis already demonstrate his tactical and operational talents. But how did Roman authors even know about this supposed conspiracy? If it was well known that Gallus had helped the Goths to destroy the Imperial Roman army, how could Gallus have received the military support of the Lower Danubian frontier that allowed him to become the next emperor? Rather, Roman authors likely fabricated internal betrayal to take the sting out of Rome’s defeat, again driven by their suspicion of the successor, who was in Moesia at the time, and their disbelief that the Goths could win without assistance.

The Battle of Barbalissos, 252/3

The aforementioned inscription of Shapur (Ka’ba-i Zardost 10-19) records that the Sasanian king invaded Syria and annihilated a Roman army of 60,000 men near Barbalissos. Shapur then relates that, in the same campaign, he captured 37 cities in Syria and Cappadocia, which he proceeds to name, including Antioch, Rome’s most important city in the Syrian provinces.

Roman sources speak of the invasion of Syria, including the capture of Antioch and successful Roman resistance at Emesa. However, while several accounts yet again focus on treachery, namely that an Antiochene turncoat, Mareades, served under Shapur and his son Hormizd (Oracula Sibyllina 13.89-130; Historia Augusta, Thirty Usurpers 2; Malalas 12.26; Peter the Patrician, fragment 171 in Banchich 2015), not a single Roman account mentions the battle at Barbalissos.

While we need not believe that as many as 60,000 Romans fell in the battle, the wiping out of a Roman army explains how the Sasanians were able to capture numerous Roman cities, including multiple locations of administrative, strategic and religious importance, such as Hierapolis, Beroea, Apamea, Larissa, Seleucia, Alexandretta, Doliche and Germaniceia.

The Battle of Edessa, 260

As with Decius’ death at Abritus, Roman authors could not ignore Shapur’s capture of Valerian. Shapur’s inscription (Ka’ba-i Zardost 19-37) narrates that he defeated Valerian’s army of 70,000 men in a great battle near the cities of Carrhae and Edessa, which the Sasanians were besieging, and that the king took prisoner not only the emperor but also commanders, senators and officials as well as Valerian’s praetorian prefect. This allowed Shapur to capture 36 cities in Syria, Cilicia, Cappadocia, Pamphylia and Lycaonia, including Tarsus, Tyana, Iconium, Caesarea and Sebasteia.

Several Roman sources mention the campaign. Zonaras (12.23) narrates that Valerian engaged the Sasanians in battle when the latter were besieging Edessa, and that Shapur used superior numbers to encircle the Romans. Most of the Romans were killed, and Valerian and his retinue were captured. Eutropius (9.7), Festus (Breviarium 23), Agathias (4.23.7) and the Epitome de Caesaribus (32.5) also note that Valerian was defeated. A fragment from the lost history of Peter the Patrician (fragment 173) reports that Shapur, learning that the Roman army had been struck by plague and that Valerian was thus not confident, began to follow the Roman army from close behind. Another fragment of Peter (201) claims that Shapur tricked Valerian with deceptions.

The abbreviated account of the lost history of Aurelius Victor (Historiae Abbreviatae 32.5) reports that Valerian was surrounded by a ruse, and the Byzantine historian Scutariotes (38) relates that he was made a captive after being outwitted by the Sasanians. The Byzantine historian Symeon Logothete (80) states that Valerian was captured in the city of Caesarea alongside 40,000 men, and Cedrenus (454) reports the same but says it was 20,000 men.

We cannot trust any of the numbers given. Just as Shapur represents the Roman army as massive, Zonaras does the same for the Sasanian army. Shapur could well have used superior numbers of cavalry to target the Romans in the open plains between Edessa and Carrhae and envelop their lines. As we have seen, the Romans may have also been weakened by plague.

However, the likelihood that Shapur had won the battle in part through cunning operational manoeuvres and tactics, a mark of a good commander, appears to have encouraged certain authors to redirect emphasis away from the fact that Romans were bested in battle (Potter 2018).

The account of the Historia Augusta is lost, but a surviving passage (Valerian 1.1-2) asserts that Shapur did not wholly defeat the Romans but had captured the emperor through fraud. Zosimus (1.36) does not even mention a battle. Rather, Valerian’s army was struck by plague, and so the emperor, who is represented as weak and empty, tried to buy off Shapur. Peter also mentions the plague and sending of gifts in the aforementioned fragment 173, but in Zosimus’ account Shapur persuades Valerian to speak to him in person and promptly has him taken prisoner.

The Byzantine historian Syncellus (715-716) claims that, because the Roman army was suffering from famine in Edessa, they had become inclined towards mutiny. For this reason, Valerian, fearful of his own men, marched his army against Shapur as if to fight another battle (he does not narrate a previous battle), only to surrender himself to the Sasanian king. He agreed to betray his army to Shapur, but the Roman soldiers learned of this and escaped with difficulty, with some killed (see also Zonaras 12.23). This is unreliable as it duplicates the war fought between Aurelian and the rival emperor Tetricus, when Tetricus betrayed his army to Aurelian, having developed a distrust of his own soldiers (see e.g. Aurelius Victor, Historiae Abbreviatae 35.3-4; Eutropius 9.13).

The Battle of Placentia, 271

There is no denying the martial talents of the emperor Aurelian, but were it not for one source, we would never have known of the defeat that he suffered near Placentia. In 271 Italy was invaded by the Juthungi, a people renowned for their warriors, especially their cavalry (Dexippus, Chronicle, fragment 8). Likely leading a coalition of Germanic peoples, they ambushed and defeated Aurelian’s army.

Although the emperor prevailed in the subsequent battles, the initial defeat caused panic in Rome (Historia Augusta, Aurelian 18.3-6). This defeat, alongside corruption, likely helped to foster conspiracies and a revolt in the city (Aurelius Victor, Historiae Abbreviatae 35.6; Eutropius 2.14; Historia Augusta, Aurelian 21.5-6, 38.2, 39.8; Zosimus 1.49).

The only writer who narrates this defeat is the anonymous author of the Historia Augusta (Aurelian 21.1-4), which was likely produced at the start of the fifth century. In contrast, the eighth century Epitome de Caesaribus (35.2), while listing facts about Aurelian, briefly notes that the emperor won victories in Italy at Placentia, at Fanum Fortunae (beside the Metaurus), and lastly in the plains of Ticinum.

The Historia Augusta is a problematic and often unreliable text, but in this case the Epitome looks to be mistaken (Saunders 1992, 323-324 n. 62). A defeat at Placentia explains why Aurelian then needed to pursue the Juthungi to the Metaurus in central Italy, and it helps us to understand the tense situation in Rome, which culminated in the erection of the Aurelian Wall.

Despite its problems, the Historia Augusta appears to draw on the same source tradition as the Epitome, thus their agreement on the location. The Epitome’s author, concerned with producing only a brief notice on Aurelian’s Italian campaign, appears to have failed to differentiate between the campaign’s overall outcome and the outcomes of the three individual battles.

Final thoughts

From a geopolitical perspective, the Third Century Crisis arguably came to an end in 298, when Diocletian forced the Sasanian king, Shapur’s son Narseh, to agree to a humiliating peace treaty. Galerius, the Caesar (junior emperor) of Diocletian, had won a major victory against the Sasanians in Armenia.

However, even this final episode in the military history of the third century was preceded by a Roman defeat. In c. 296 Narseh invaded Roman Mesopotamia and bested Galerius in the area between Carrhae and Callinicum. Roman authors did not actually shy away from this defeat, as it was used to shed light on the relationship between Galerius and Diocletian (Eutropius 9.24; Festus, Breviarium 25; Ammianus 14.11.10; Jerome, Chronicle 227). But as the defeats of Aurelian and Galerius demonstrate, Rome’s third century enemies could win victories against even the most formidable of Rome’s warrior emperors.

In the third century they faced enemies who could go toe to toe with Roman armies and come out on top. Rather than explore what Shapur and Cniva did to overcome their armies, or even report the name of the Juthungian who initially bested Aurelian, many Roman authors provide a manipulated version of events. Sometimes they blame treachery; sometimes they ignore the defeat altogether. Unfortunately, for this reason, most of the details of these battles have been lost to time.

Byron Waldron is a Postdoctoral Researcher at the University of Cyprus. Here he conducts research for the project ‘Group Minds in Ancient Narrative’, which is funded by the European Research Council. Byron is the author of Dynastic Politics in the Age of Diocletian, AD 284-311 (Edinburgh University Press, 2022), and has written articles on Latin literature, Roman history and Persian history for edited volumes and journals, including the Bulletin of the Institute of Classical Studies, the Journal of Late Antiquity and Zeitschrift für Papyrologie und Epigraphik. He has also written documentaries for HistoryMarche, including popular series on Aurelian and the Third Samnite War.

Selected Bibliography

Banchich, T. M. 2015: The Lost History of Peter the Patrician: An Account of Rome’s Imperial Past from the Age of Justinian, Oxford.

Bursche, A. & K. Myzgin. 2020: The Gothic Invasions of the Mid-3rd c. A.D. and the Battle of Abritus: Coins and Archaeology in East-Central Barbaricum, JRA 33, 195-229.

de Blois, L. 2019: Image and Reality of Roman Imperial Power in the Third Century AD: The Impact of War, London.

Dodgeon, M. H. & S. N. C. Lieu. 1991: The Roman Eastern Frontier and the Persian Wars AD 226–363: A Documentary History, London & New York.

Edwell, P. M. 2008: Between Rome and Persia: The Middle Euphrates, Mesopotamia and Palmyra under Roman Control, Oxford & New York.

Leadbetter, W. 2009: Galerius and the Will of Diocletian, London & New York.

MacDonald, D. 1981: The Death of Gordian III: Another Tradition, Historia 30.4, 502–8.

Maricq, A. 1958: Classica et Orientalia 5: Res Gestae Divi Saporis, Syria 35.3/4, 295-360.

Martin, G. 2006: Dexipp von Athen: Edition, Übersetzung und begleitende Studien, Tübingen.

Potter, D. S. 1990: Prophecy and History in the Crisis of the Roman Empire: A Historical Commentary on the Thirteenth Sibylline Oracle, Oxford & New York.

Potter, D. S. 2018: Decius and Valerian, in D. W. P. Burgersdijk & A. J. Ross (eds), Imagining Emperors in the Later Roman Empire, Leiden, 18-38.

Potter, D. S. 2022: Decius and the Battle near Abritus, in R. Evans & S. Tougher (eds), Generalship in Ancient Greece, Rome and Byzantium, Edinburgh, 139-150.

Saunders, R. T. 1992: Aurelian’s Two Iuthungian Wars, Historia 41.3, 311-327.

Stover, J. A. & G. Woudhuysen. 2023: The Lost History of Sextus Aurelius Victor, Edinburgh.

Watson, A. 1999: Aurelian and the Third Century, London & New York.