This is one of several translated excerpts from Byzantine sources produced and mounted with historical introduction and commentary by Paul Stephenson.




Cinnamus, 10-13.

When he heard of these affairs(1), the emperor rushed to the Danube with his entire army, taking also an allied force(2) of Ligurian horsemen, whom our men call Lombards, and Turks. He pitched camp beside the river's banks and prepared for battle. However, Stephen's body was feeble, and he was being tended somewhere in the heart of his land. Yet he did not wish to be heedless, and sent troops as quickly as he was able, commanding that they prevent the emperor from crossing. The Hungarians carried out the orders. To neutralise this resistance the emperor connived as follows. He detached his allies, then ordered them along the river to a place called Tempon, where a hill which rose in the Hungarians' land ran as far as the river, where they should cross. He stayed with the remaining force of Romans opposite the fort of Chramon, giving a false impression that he would shortly cross from there. After this the Romans crossed without trouble.(3) The Hungarians could not withstand their first assault, and hastily turned in flight. The pursuit lasted as far as the river, where the Hungarians dashed in throngs onto a bridge across the flood which collapsed.(4) Many, swept away by the river, disgorged their spirits, but many fell into the Romans' hands, including Akous and Keled(5), who were very distinguished men among the Hungarians. After this success the emperor took the fort of Chramon without resistance, and immediately crossed back into Roman lands. He strengthened the city of Branicevo with a garrison of troops under the command of Kourtikios, and returned to Constantinople. Shortly afterwards the Hungarians besieged and took Branicevo, killing some of the Romans within, and capturing others. There were some who sought salvation in flight. The emperor was infuriated by this and convicted Kourtikios on a charge of treason, lashing him many times on the back although (they say) he did not desert the walls until the whole enemy army charged into the city and put buildings to the torch.

At this time the Serbs, a Dalmatian race, conspired to revolt and subjected to themselves the fortress of Ras. For this reason the emperor avenged himself similarly on Kritoplos, who had been charged with the defence of that fortress. He led him through the marketplace wearing women's clothes and mounted on an ass. Returning to Branicevo, the second time, he set about rebuilding it in haste. Since this endeavour took some time, the army began to suffer severely from the winter weather and dearth of provisions. Learning of this, the Hungarian king resolved to cross the Danube as soon as possible and launch a surprise assault on them. However, in the land of the Hungarians was a certain woman, Latin by birth, who was exceptionally wealthy and distinguished. She sent word to the emperor of what was to take place. Because he was unable to engage them in battle with an equivalent army, since as related his troops were depleted by disease and dearth of provisions, he reinforced the city where possible and departed. Furthermore, in order to avoid the Hungarian army, he passed through the impenetrable and craggy region known as the Evil Stairs. There the Hungarian army swooped without warning upon the regiments guarding the rear, but inflicted no injury upon the Romans. Having collected sections of the awnings which hung in the imperial tent, which had been discarded through lack of pack animals, they departed, and the army of Romans escaped unscathed.

Choniates, 17-18.

During the summer season the Hungarians crossed the Danube and sacked Branicevo. They demolished the walls and transported the stones to Semlin. They also plundered Sofia, breaking and tearing up earlier peace treaties. The secret reason for this conflict was the fact that Almos, the brother of Stephen who ruled the Hungarians, had fled to the emperor and had been welcomed cordially. The professed and specious reason was the [Hungarians'] accusation that citizens of Branicevo had assaulted and robbed Hungarians who had travelled to the locality to trade, committing heinous crimes against them. Since this evil had erupted unexpectedly while the emperor was dwelling at Philippopolis, he considered the predicament scrupulously and resolved to expel the Hungarians from there. He spent some time preparing his household troops to defend against the enemy, then sailed equipped and swiftly-propelled ships along the Danube from the Black Sea, falling upon the foe by both land and water. He then crossed the river in the general's trireme, ferrying the army to the far shore where the cavalry dispersed the Hungarian host with their couched lances. Demonstrating his remarkable endurance he remained in enemy lands and captured Frangochorion, the richest land of the Hungarians which lies between the rivers Sava and Danube with plains suited to driving horses. He also took Semlin and attacked Chramon, from which he wrested great spoils. After further struggles with this race, he offered them peace, and compelled those other Barbarian peoples at the western frontier of the Roman Empire whom he had so often defeated in battle to become allies.


It is clear from the translations presented above that the accounts composed by Cinnamus and Choniates diverge dramatically when describing John II's encounters with Hungarian forces between 1127 and 1129.(6) Cinnamus seems to offer a full and plausible account. However, his description is replete with allusions to John's weaknesses as a commander. Moreover, he divides the campaign into two episodes. The first episode begins with the author implying that John was panicked by a surprise Hungarian attack and marched straight to the Danube. Next, Cinnamus reports that the emperor was not met by the Hungarian king, who was recovering from an illness many miles from the Danube. This fact is not mentioned by Choniates, and had little effect on the outcome. Therefore, I would suggest that Cinnamus intention in referring to Stephen's absence and weakness is to diminish John's victory. Third, Cinnamus implies that John was not personally involved in the initial, and decisive victory achieved by the Lombard and Turk mercenaries. In fact the only martial endeavour directed by John in Cinnamus's account is the capture of the fort at Chramon which fell "without resistance", after which, Cinnamus states, the emperor immediately recrossed the river into his own lands. He delayed only to strengthen Branicevo with a garrison under a certain Kourtikios.

Cinnamus records that a second episode was initiated "a short time after"by the Hungarians launching a renewed attack on Branicevo. He reports that, rather than meeting the Hungarians in battle John's first response was to humiliate the commander, Kourtikios, who had fought bravely to prevent the fortress being captured. At this point Cinnamus records that John dealt with the commander of Ras in a similarly humiliating fashion, but provides no further information on John's dealings with the Serbs.(7) The cumulative effect of Cinnamus's innuendo is to imply that John acted as a vengeful and spiteful coward. Moreover, in an aside he suggests that John was criticised for his actions at the time.(8) Furthermore, in devoting precious time to acts of vengeance John is shown to have neglected his men and precipitated his own final humiliation. The Hungarians took heart from the fact that John's troops were stuck at the Danube, demoralised by the cold and suffering through lack of provisions. They set out to renew their assault. John was only saved through luck, as word of the imminent attack reached him, allowing him time to retreat. If this were not shameful enough, John's informant was a both a woman and a Latin. The emperor fled the region through the Evil Stairs to avoid meeting the Hungarian army. However, the Hungarians followed and fell upon his rearguard. Although they inflicted no harm on the troops the Hungarians disgraced John by stealing the hangings from the imperial tent, which the emperor had left behind in his ill-prepared haste.

Choniates directly contradicts Cinnamus's account. He informs us that John took time considering and preparing his forces to meet the Hungarians, then summoned his fleet to prepare the ground for a land assault. Both facts are fully corroborated by the Hungarian Chronicle, which further recounts how the Hungarians' ships had been bombarded with Greek Fire.(9) Choniates also informs us that the capture of Chramon was part of a campaign which involved several engagements after the fall of Semlin. Writing some years after Cinnamus, Choniates is adamant that John did not retire, but stayed on in enemy territory and showed remarkable endurance.(10) His is a very definite statement, and his references to the stamina and endurance of the emperor bring to mind the twenty-two verse encomia of Theodore Prodromus which constantly stress elements of John's andreia.(11) There is no evidence that John intended for his conquest of Frangochorion to be permanent or to involve occupation or annexation of any lands beyond the Danube.(12) His long-term plans did not involve a protracted dispute with Stephen II, who seemed keen to make political capital by raiding Byzantine lands. For this reason the whole episode, as portrayed by Choniates, smacks of a carefully orchestrated show of strength. As such, it was a considerable success for John, who forced the Hungarians to accept peace on his terms.

Cinnamus's failure to record John's series of victories and the treaty he concluded beyond the Danube is remarkable when one considers how thorough are his accounts of all subsequent campaigns in this area. He generally provides considerable detail which is missing from Choniates's account. Moreover, his encomiasts would have made sure that John's victories in Frangochorion would have been well known in Byzantium. It is therefore possible that Cinnamus was quite deliberately drawing attention to the emperor's departure mid-campaign, and implying that this took place before the series of victories commemorated in contemporary panegyrics and later picked-up on by Choniates. Therefore, we must seek an explanation for this deliberate obfuscation.

For further commentary, see P. Stephenson, 'John Cinnamus, John II Comnenus and the Hungarian Campaign of 1127-1129', Byzantion lxvi (1996), 177-87.


1. Stephen II had just sacked Belgrade and used the stones to construct the walls of Semlin on the opposite bank of the Danube.

2. The Greek is symmakhikon (or symmakhoi), auxiliaries or allies, but also a self-contained ethnic mercenary unit. See J. Shepard, The use of Franks in eleventh-century Byzantium, Anglo-Norman Studies XV (1992), 280-1.

3. This sentence is a little obscure. Cinnamus has just informed us that John drew Hungarian attention from his ploy (dokesin) with the pretence that he was about to cross at Chramon. In the next sentence he informs us that indeed John did just that -- hardly a sophisticated feint. If the Romans had fallen upon the Hungarian army we must imagine that some time had passed between John despatching his mercenaries (symmakhikon) and the battle; two incidents related in consecutive sentences. We must also further conjecture that the mercenaries crossed the river and took up a position beyond the Hungarians, facilitating the Roman crossing. This seems to require a little too much of the reader [...] Cinnamus surely means to imply that the emperor had little personal involvement in the battle. Furthermore, we should beware of over-interpretation since, as we will see below, Choniates tells an entirely different story.

4. The editor of the Bonn edition clearly believes that the Hungarians were driven towards a bridge over the river Danube and perished in the fast flowing waters beneath when it collapsed (in the Latin translation Meineke even supplies the name Ister). This would indeed be the natural interpretation since no other river is mentioned by name, and Cinnamus refers only to the potamos. However, there was no bridge across the Danube at this point, so the river into which many Hungarians fell must have been a tributary. Therefore, it is possible that the Romans crossed the Danube and drove the fleeing Hungarians towards a fast-flowing bridged river which ran into the Danube. As we will see below, this is confirmed in the Hungarian Chronicle,which states that the fateful river was the Karas (Karasso/Carasul) which flows into the Danube opposite the site of Branicevo. See Chronici Hungarici Compositio Saeculi XIV, ed. A. Domanovszky, Scriptores Rerum Hungaricarum, ed. E. SzentpŽtery, Budapest 1937, I, 441. Brand does not record this in his translation, but seems to prefer this explanation, referring to a river, not the river (i.e. the Danube).

5. G. Moravcsik, Byzatinoturcica,Berlin 1958, II, 59-60, 158.

6. A point noted by Chalandon, Les Comnene,II, 58.

7. Choniates seems to allude to unrest among the Serbs at this time. However, this may refer to an episode which took place several years before the Hungarian Campaign, which Cinnamus fails to record. Clearly the insertion here is to juxtapose this incident with the Kourtikios story, thereby highlighting a trait in John's treatment of his commanders which the author clearly intends his audience to condemn. Choniates, 16, records that John fought a successful campaign against the Serbs shortly after his victory against the Cumans in 1122. This is corroborated by a twelfth-century Latin source. Letopis' Popa Dukljanina, ed. F. Sisic, Belgrade-Zagreb 1928, 368-9: "After the death of King Vladimir, Queen Jaquinta's son George claimed the realm. In the second year of his reign he wanted secretly to arrest Branislav's sons. However, he achieved little because they were made aware of his plans and fled to Dyrrachium and their uncle Goyslav. Only Grubessa was captured and put in gaol in Scutari. At that time the Dux Calo-John Cumanus mustered a mighty army and marched with Goyslav and his nephews against King George. The king assembled his people and prepared them for battle. As the battle raged, a division of King George's army fell. Many were slaughtered and more still captured. The king and some of his entourage escaped and fled to Oblik. After this the Dux and his men attacked and captured the city of Skodra. There they released Grubessa from custody, and on imperial orders the people instituted him as king. The Dux relinquished command of the army to him and returned to Dyrrachium". The fact that John is called Cumanus seems to confirm that his victory against the Cumans was a very recent event. See also Chalandon, Les Comnene,II, 69-73.

8. The Greek phasin, translated here as "they say", may well refer to a psogos,a censorious tract or pamphlet. These were certainly written by critics of the Comnenian emperors. For example, see Alexiad, XIII, i, ed. Leib, III, 89.

9. See the alternative version of events in the Hungarian Chronicle, 439-42.

10. So adamant in fact that one must consider that he was aware of Cinnamus's implied criticism and sought to correct it. This raises the thorny question of whether Choniates read Cinnamus and suggests that the answer should be yes.

11.Theodoros Prodromos. Historische Gedichte, ed. W. Hšrandner, Vienna 1974, nos. 1-21, 24. P. Magdalino, The Empire of Manuel I Komnenos, Cambridge 1993, 419-20, stresses the importance John's encomiasts attributed to the andreia of the emperor. Indeed, John was the emperor who revived the ceremony of the imperial triumph in 1133. On that occasion four encomia were delivered each celebrating the ideology of warfare and the qualities of the emperor as leader and blood-letter.

12. F. Makk, The Arpads and the Comneni, Budapest 1989, 25

Copyright: Paul Stephenson, April 2000

Revised January 2012