Oxford University Press's
Academic Insights for the Thinking World

The Battle of Hadrianople

Peter Heather, a leading authority on the late Roman Empire and on the barbarians, a teacher at Worchester College, University of Oxford and author of The Fall of the Roman Empire: A New History of Rome and the Barbarians is our guest on the OUPblog this week. Heather’s book proposes that centuries of imperialism turned the neighbors Rome called barbarians into an enemy capable of dismantling the Roman empire. In the article below Heather looks at The Battle of Hadrianople. Be sure to come back tomorrow when Heather answers a few questions for OUP.


On August 9th 376, Valens, Emperor of the Roman east, gave battle to a mixed force dominated by Goths on a plain some 8 hours march north of the city of Hadrianople: modern Edirne in European Turkey. The result was catastrophe. By nightfall the emperor and two-thirds of his army, the pick of the eastern army, lay dead on the battlefield. Valens’ body was never recovered. What had gone wrong? It used to be taught that Hadrianople represented the dawn of the Middle Ages: the roman-empire.jpgfirst triumph of cavalry over the Roman legion. But while there certainly was a substantial Gothic cavalry army present at the battle, most Goths fought on foot. And infantry were to remain the staple of warfare right through the collapse of the Roman west over the century which followed. Besides, more or less the same group of the Goths was to fight multiple engagements against Roman forces in the years which followed – 395, 397, 402 (twice) and so on – and the result was always some kind of draw, with a minor tactical advantage swaying first one way and then the other. What all this really shows is that something really odd went on that August day in the heart of the Balkans. Hadrianople was a fluke, not the start of a new order in warfare at all. But what kind of fluke; why did Valens lose so badly?

Contemporary accounts of ancient battles are never what you’d like them to be. History was a branch of rhetoric for Greeks and Romans, and the audience expected individual heroics, not strategic and tactical analysis. That said, our main account of Hadrianople, written by the historian Ammianus Marcellinus, is about as good as it gets. Ammianus wasn’t present on the battlefield, but he was a former soldier, with many military contacts and he’d clearly been talking to someone who survived. And from his account, several key ingredients of the Roman Armageddon emerge without too much interrogation.

First, Valens was misled by faulty intelligence. The war plan for 378 envisaged him waiting for his western colleague and nephew, the Emperor Gratian, before taking on the Goths. But Gratian was delayed and Valens’ army was left sitting in the Balkans from early June to early August: 2 whole months for morale to fritter away. In this context, Valens received a report that only 10,000 Goths were massing near Hadrianople, not the full Gothic force. The emperor was confident he had enough men to fight against this many enemies and rushed forward to battle. But as it turned out, the Gothic army was larger than reported, and Valens found this out only after the fighting had begun.

Second, Valens really did rush into battle. Having decided, after much discussion, that he would fight, his army plunged forward on the morning of August 9th from its base around the city of Hadrianople. They left at dawn and the Gothic wagon circle (‘as if turned by a lathe’ in Ammianus’ words) came into view at about 2pm, after an eight hour march. Battle would begin before the Roman force could eat or rest. And in the Balkans, the average midday temperature in August reaches 30 degrees and above. The Romans were hot, tired, and hungry and the Goths made it all worse, lighting huge fires upwind of their opponents, so that billowing smoke and heat poured down on them.

Third, and not surprisingly given both the two month wait and the morning’s dash from Hadrianople, discipline and order was not what it should have been. Two Roman regiments opened the fighting without explicit orders, when Valens – ever cautious – was still contemplating in fact whether to discuss terms for a truce. The Roman army was by no means fully ready for battle at this point, some of it still moving up the road from Hadrianople.

The outlines of what followed, likewise, are clear enough. The Roman heavy infantry – pick of the army – rushed forward in the centre, fighting its way towards the Gothic wagon circle. At first all seemed well. The Romans pushed their opponents back to the wagon circle and looked to be about to storm even that when disaster struck. The Roman left wing was also pushing forward – perhaps to encircle the wagons – when the other Goths Valens didn’t know about suddenly ‘dashed out as a thunderbolt does near high mountains’ and routed it. This left the heavy infantry of the centre unprotected, open to a massive flanking attack, against which they could not manoeuvre:
Their companies were so crowded together that hardly anyone could pull out his sword or draw back his arm…arrows, whirling death from every side, always found their mark with fatal effect since they could not be seen beforehand nor guarded against… and in the press of ranks no room for retreat could be gained anywhere.
The Roman main body was trapped and could not make either the superiority of its equipment or its training count. And thus it was that two-thirds of Valens’ army went down in the massacre which followed.

In sum: really not a bad account. We have a much better sense of what went on at Hadrianople than at just about any of the other battles which punctuate the period of west Roman collapse in the late fourth and fifth centuries. But even so, there are some crucial gaps. First, what size was Valens army? We know two-thirds of it fell, but how many men was this? Ammianus doesn’t say. From another source – a list of the Roman army’s order of battle from 395 – you can work out that 16 elite Roman regiments were never reconstituted after the battle, so bad were their losses. This has led some scholars to estimate Roman dead at 20,000 and Valens’ army at 30,000 in total. But we don’t know, in fact, how large a late Roman regiment really was, perhaps only 500 men, and how much of a unit has to die before you decide not to reconstitute it? My own view, would be that Valens led more like 15,000 men to the Balkans, so that about 10,000 died. Second, how many Goths were present? We know Valens thought he was facing 10,000, but how many extra Goths turned out to be there. Again Ammianus doesn’t say.

Not knowing how big the 2 opposing forces were, you simply cannot be certain of the significance of the different episodes Ammianus reports. The harder you press it, in fact, the more the account slips through your fingers. Just how many Roman troops were still on the road when battle commenced? Would the Roman assault on the wagon circle have failed anyway because there were more Goths there than Valens thought, or because the hot, hungry, and tired Roman infantry were bound to run out of steam? And just how big a charge was it that destroyed the Roman left wing and trapped the centre in a desperate killing zone? There were 2 major Gothic groups – called Tervingi and Greuthungi – loose in the Balkans in the summer of 378. And when he thought he was facing only the 10,000 Goths, Valens was periodically negotiating with just one of them: the Tervingi. The charge, however, was commanded by the leaders of the Greuthungi. One possible scenario, therefore, is that Valens’ army threw itself on the Tervingi and their wagon circle only to be ambushed, from the Goths’ right, by the whole of the Greuthungi. It’s a satisfyingly dramatic vision, but is it correct? We have no way to be sure, but the great thing with Ammianus is that, If you’re the kind of person who loathes and detests the fact that we will never know for certain, then first millennium history is not for you! If, on the other hand, you are blown away by the fact that, thanks to Ammianus, , we can – 1629 years on from the battle – get as far as asking these kind of detailed questions, then attempting to solve the complex puzzles thrown up by the deep past will bring you huge satisfaction.

Recent Comments

  1. […] of The Fall of the Roman Empire: A New History of Roman and the Barbarians taught us about the Battle of Hadrianople. Today he has kindly answered a few questions for […]

  2. […] University Press Blog has posted an article on the battle by historian Peter Heather. The article is excellent in its own right, but it whets […]

  3. Peter Donnelly

    Good article. The only thing I would take issue with is the notion that Valens was “ambushed” by the Greuthungi. This was a hot August day, and Ammianus himself describes how the battle raised a choking cloud of dust. The returning cavalry must have been raised a similar cloud that would have been visible long before they reached the battlefield.

    I’ve examined Ammianus’s account in some detail at http://mysite.verizon.net/res1bup4/adrianople.htm .

Comments are closed.