Of key significance in Snorri’s account is the passage telling of ambitious young soldiers leaving the main army to join Harald’s troop when they learned of the greater booty being shared by the men under his command, because the most usual source of contention between Maniakes and his mercenaries in Sicily is known to have been the sharing of plunder. When disputes of this sort are set beside Maniakes’ code of iron discipline, it is not difficult to imagine the sort of problems which would have been presented by a contingent of northmen under their own young officer, brimming over with self-confidence and mercenary avarice.
So too, the Normans – who had proudly aristocratic commanders of their own – are known to have clashed with the heavy-handed Georgios over battle-booty in the Sicilian campaign, but one thing of which Maniakes cannot be accused is failure in pursuit of victory. Within two years of campaigning in Sicily, he had reclaimed virtually all the island from a Saracen enemy possessed of superior numbers. The Byzantine army certainly included first-class fighting-men in its Norman and Varangian mercenaries – and likewise in the force of sturdy Armenians led by Katalokon Cecaumenus (who has been suggested as one possible candidate for authorship of the account of Harald in the Advice) – but others, such as the reluctant Longobard recruits from Apulia, were of lesser quality and so the elite units must have borne the heat and burden of the day in the fiercely fought battles which led to the defeat of the Saracen commander Abdallah at Traina in 1040.
The immediate aftermath of Traina heralded Maniakes’ dramatic fall from imperial favour – and entirely as a consequence of his own fearsome temper. He was never going to be well-disposed towards the naval commander assigned to the Sicilian invasion, because the admiral Stephen had been a caulker in the shipyards of Constantinople until his brother-in-law the emperor Michael appointed him to the rank of patricius and to command of the fleet which brought Maniakes’ army to the shores of Sicily. Having accomplished that duty, presumably to the best of his modest abilities, Stephen was later held personally responsible for allowing Abdallah to escape by ship to Tunis in the wake of the defeat at Traina. So explosive was Maniakes’ rage at this oversight that he actually took a whip to the unfortunate admiral, provoking Stephen’s bitter complaint to the emperor accompanied by allegations of treason on the part of the general, charges sufficiently serious to prompt the recall of Maniakes to the capital where he was to spend most of the next two years in prison.
Before his loss of command, Maniakes was to assert his arrogance of power over his subordinates with disastrous consequences, first causing affront by seizing a fine warhorse which had been chosen as his own prize by Arduin, commander of the Longobard contingent, and then by denying the Normans the full share of booty which they believed to be their due. As a result, the Norman mercenaries defected to join the latest rebellion brewing up among imperial subjects in the south of Italy, thus depriving the Byzantine forces of their best mercenary cavalry just as Maniakes was replaced in command by the scarcely comparable admiral Stephen. In the event, Stephen was dead within the year and replaced in his turn by an obscure eunuch known only as Basil who soon managed to lose almost all that had been won in Sicily, leaving Messina as the one remaining imperial possession on the island.
It does seem very likely that the defection of Norman mercenaries who felt themselves to have been short-changed by the overbearing Maniakes might have formed the subject of a story brought back to Iceland – quite possibly by Halldor – and thus found its way into Snorri’s saga in the form of his reference to ‘Latin-men’. While there is good reason to believe the Varangians having similarly resented Maniakes, there is no question of Harald and his troop having followed the Normans, either in their defection or in their alliance with the Italian rebellion, because it is perfectly clear from the evidence of the skalds alone that when Harald and his troop were despatched to Italy they were fighting against the same Normans who had earlier been their comrades-in-arms in Sicily.
There had been trouble already in Byzantine Italy, where rebels had seized the town of Bari in 1038, and there was still greater trouble ahead through the three decades which it took the Normans to break the last imperial hold on the Italian mainland in 1071. In 1040, however, the empire was still prepared to put down any insurgency in its Italian provinces. Indeed, Bari was retaken in that same year just before a new rebellion broke out in Mottola where it claimed the lives of the catepan and other imperial officials before the rebel leader made his peace and submitted to the emperor. A newly appointed catepan arrived towards the end of the year with the support of a force of Varangians, some of whom must have been assigned to him from the army in Sicily because they included the troop commanded by Harald, who ‘led the march in the land of the Longobards’ – according to the skald Thjodolf – when a separatist revolt in Apulia had Norman cavalry as its cutting-edge.
Norman mercenaries had first emerged in Italy in 1015, when a band of young pilgrims had been recruited to fight for the Longobards against imperial forces. When word of this new source of demand for fighting-men got back to Normandy, other young warriors looking for action and profit made their way south until a steady tide of Normans was flowing down into Italy. Eventually Norman mercenaries were also to be found in Byzantine forces – being available, as was ever the way of the professional, to fight for whichever paymaster might be recruiting – engaged against the Saracens in Sicily and Pechenegs in the Balkans.
Already in the later 1030s, the sons of Tancred de Hauteville, who himself had been just another knight in the service of the dukes of Normandy, can be recognised as the first representatives of a formidable Italo-Norman dynasty. William (called ‘Iron-arm’) and his brother Drogo led the Norman contingent with Maniakes in Sicily in 1038, while their more famous brother – the afore-mentioned Robert called Guiscard, or ‘the crafty’ – was to be appointed by the Pope to new dukedoms of Apulia and Calabria in 1059, as also of Sicily which he first invaded with his younger brother Roger in 1061 and which was finally wrested from the Saracens in 1072.
Only recently mercenaries in imperial service, these Normans now presented the most serious opposition facing Byzantine forces in Italy. It has been suggested that it was their highly effective development of the close-formation cavalry charge which gave the Normans the edge over the Byzantines, ‘who hated the solid lines of horsemen with levelled lances’,8 and this may well have brought them the victory over superior numbers of imperial troops in the two major battles fought in southern Italy in the spring of 1041. Varangians served with the Byzantine forces in both of these conflicts and are said by the Greek annalists to have suffered heavy losses first at Olivento in April and again in the second battle fought early in the following month at Montemaggiore, where a great part of the catepan’s army was drowned in the full flood of the Ofanto river.
The skald Illugi tells of Harald’s going early ‘to disturb the peace of the Frakkar’, by whom can only be meant the Normans, and so it would seem likely that he and his troop would have been engaged in the fighting on at least one of these occasions. Nothing more is known of Harald’s part in this Italian campaign, other than that he clearly came out of it alive and with full honour on the evidence of his subsequent promotion to the rank of manglavites. His was not the only contingent of Varangians serving in Italy, because others are mentioned by the annals when Harald is known to have been already on campaign with the emperor against the Bulgars, so it is not impossible that he and his troop may have been withdrawn to the capital before the more serious of the defeats inflicted on Byzantine forces in that spring.
Snorri himself has nothing to say of Harald in Bulgaria, because the solitary reference to that campaign found in his saga is a single phrase in a strophe from the skald Thjodolf quoted at the foot of its opening paragraph. Thjodolf’s verse is principally concerned with the battle of Stiklestad, although set down many years after the event, and yet refers to Harald as Bolgara brennir or ‘burner of the Bulgars’, thus supplying the on
ly fragment of skaldic testimony to corroborate the eminently authoritative evidence of the Advice for his part in the emperor Michael’s final suppression of the Bulgar rebellion in 1041.
An outline of the background might be helpful at this point, because there has been little reference to Bulgaro–Byzantine relations here since the Bulgarian tsar Samuel destroyed the emperor Basil’s army at Trajan’s Gate in the year before the foundation of the Varangian Guard. Basil had sworn to take his revenge on the Bulgars for that devastating defeat and, although it took him a full twenty-five years to do so, that vengeance was terrible indeed. Having rebuilt the Byzantine military into a war-machine capable of outfighting the Bulgars in their own rough Balkan terrain, Basil had already reclaimed most of the eastern extent of the peninsula for the empire by 1004. Ten years later, he defeated a Bulgar host in the narrow pass of Cimbalongus north of Serrae and earned himself enduring infamy after the battle when he put out the eyes of 15,000 prisoners, leaving one out of every hundred with the sight of a single eye so as to be able to lead their comrades home. When the tsar Samuel, already a sick man by that time, beheld the return of so many grievously mutilated warriors, he is said to have suffered an apoplexy and died a few days later, yet his people fought on for four more years until finally surrendering to Basil the Bulgar-slayer in 1018.
The name ‘Bulgar’ derives from the Old Turkic bulgha, meaning ‘to mix’, and the earliest ancestors of the people who were known by that name in the tenth and eleventh centuries would have been a Turkic steppe tribe akin to the Huns and Pechenegs – as also, indeed, to the later Mongols. That original stock of nomadic warrior-herdsmen had long since been diffused by its passage through the various gene pools of the Balkans, and most influentially those of the Slav pastoralists, by the time of the creation of the ‘First Bulgarian Empire’ under its tsar Samuel, who was himself of largely Armenian descent. There were, of course, other branches of the same original stock – principally the Volga Bulgars, who adopted the Islamic faith from their Arab trading contacts – but the Bulgar subjects of the Byzantine empire were those settled around Macedonia.
As frighteningly vindictive as he had been after his victory over the Bulgars in battle, Basil was to show a remarkable generosity towards them once they had become imperial subjects, and most particularly in allowing them to pay their taxes in kind rather than in cash, but the same generosity was soon to be discontinued during the reign of Michael IV, especially when the emperor’s health had entered into its terminal decline and responsibility for imperial finances had passed to John the Orphanotrophus. Now the great power at the Byzantine court, he was soon to demand taxation rendered in hard currency, thus provoking a tide of hostility among the Bulgars, which only awaited the emergence of some sort of leader before it erupted into full-scale revolt.
Just such a figurehead appeared in 1040 when one Peter Deljan declared himself tsar in Belgrade. This ‘Tsar Peter’ is sometimes identified as an escaped slave and sometimes as a grandson of the great Samuel, yet he could well have been both because he was possibly a son of Samuel’s son and short-lived successor, Gabriel Radomir. His claim would seem to have had some measure of legitimacy, because he was joined in early autumn by an ally in the person of Alusian – presumably Peter’s cousin, because he was not only a grandson of Samuel but also a son and younger brother of the last two Bulgarian tsars – who had escaped from house arrest in Constantinople (where he had been held by order of the Orphanotrophus but on an unknown charge) to take his place beside Peter Deljan at the head of the rebellion.
With its joint lords of revolt in place, the Bulgar rising was unleashed against an imperial authority already under pressure from Saracens in Sicily and Normans in southern Italy, so the rebels would seem to have had a head-start for their surge through Macedonia and into northern Greece, where they inflicted a heavy defeat on the garrison at Thebes. The emperor Michael was in his palace at Thessalonika when the rebellion broke out and lost no time in hurrying back to Constantinople to organise his military response – which may well have included the recall of Harald and his troop from Italy and, if so, then in all likelihood on the recommendation of Michael’s brother, the general Constantine.
Constantine remained in Thessalonika and held the city when the Bulgars arrived to lay it siege. He was supported by forces apparently drawn from a regiment of the Tagmata, but identified only as the ‘Tagma of the Great-hearted’, who made a magnificent sortie in the last week of October to win an impressive victory in throwing back the besieging Bulgars, who suffered casualties in the thousands and the rest of their forces put to flight. Nonetheless, the rebels were evidently able to recover and to fight on, because they had driven westward to storm Dyrrachium on the Adriatic coast before the end of the year. In the spring of 1041, however, their prospects were beginning to darken when an impressive imperial army – assuredly now including Harald and his Varangians – was in Thessalonika with the emperor who declared his intention of leading his forces in person.
This was, indeed, an extraordinary announcement because, although still only in his twenties, Michael the Paphlagonian was a chronic invalid obviously nearing the end of his life. He had long suffered from epilepsy, but what Psellus calls his ‘internal trouble’ had caused him to become bloated to the point of semi-paralysis, his legs hideously swollen and afflicted by gangrene so that every movement must have been a torment. Yet Michael was determined to lead his army on a carefully planned operation, which would have been confident of victory even if the rebel leaders had not fallen out with each other and thus ensured its absolute certainty.
When he was accused of treachery by Deljan, Alusian had struck back with a cook’s knife (at least according to Psellus, who describes the whole grisly exchange enacted at a banquet) to put out both of his cousin’s eyes and then to slice off his nose. Now in full command of the rebels, Alusian led them against the Byzantine army but with so little success that he had to flee into hiding when their onset was thrown back. Presumably recognising the inevitability of total defeat, Alusian entered into his own secret negotiations with the emperor, offering surrender in exchange for restoration of such honours as were due to him and to his family. When these conditions were accepted, Alusian proceeded with a pre-agreed charade of advancing his horsemen once more towards the enemy before suddenly abandoning them in a theatrical rush to throw himself upon the emperor’s mercy. While he was granted a full pardon and returned to Constantinople, the Bulgar host persisted in its dogged resistance, now with the mutilated Deljan at its head and refusing to surrender until it was finally crushed by the emperor’s army at the battle of Prilep.
When the skald Thjodolf makes mention of Harald having fought ‘eighteen fierce battles’ before his return to Norway, it is reasonable to assume that these would have been the major engagements in which he saw action as a Varangian mercenary in the service of Michael IV. So too, when Thjodolf’s description of his king as ‘burner of the Bulgars’ is set beside the reference made by the author of the Advice to ‘the revolt of Delianos in Bulgaria [where] Araltes went on campaign with the emperor, with his own troops, and demonstrated deeds against the foe worthy of his birth and nobility’, Harald was quite certainly involved in this later phase of the Bulgarian war and (if Thjodolf’s words might be taken literally) possibly deployed on a firebrand-wielding intimidation of the rebel heartland.
What can also be said with certainty is that this campaign was to be his last in imperial service – leading on to the probability of the battle at Prilep having been the last that he fought as a Varangian mercenary. Once again though, Harald clearly served the emperor with sufficient distinction to merit promotion, and this time to the rank of spatharokandidatos when he accompanied Michael’s triumphal return to the capital.
Constantinople, 1041–1042
Probably the most extraordinary aspect of the Varangian Guard is the fact of the personal protection of the Basileus (as the Byzantine emperor was formally styled in Greek)
, who was ‘held to be the sole legitimate sovereign of the Christian world’ and represented as ‘the earthly counterpart and vice-regent of the Christ Pantokrator’,9 having been entrusted to fighting-men out of the remote northland whose two best-known characteristics were their mighty battle-axes and notorious appetite for alcohol. Customarily referred to in the Greek annals as ‘the axe-bearing Guard’ – and on one occasion by a contemporary observer as the ‘Emperor’s wine-bags’ – at least there is some reflection of ceremonial dignity encoded in the Greek titles bestowed on the various ranks of the Varangians of the City.
The title of manglavites, for example, had its origin in manglavion, the name given to a short whip borne by officers of the Hetairia preceding the emperor in procession and used by them to clear the way for the progress of the imperial party. Although thought to be largely honorary in Harald’s time, the rank of manglavites still carried with it the privilege of wearing a gold-hilted sword, such as that described in Laxdæla saga when Bolli Bollason returns home from his service in the palace guard with his sword ‘now inlaid with gold at the top and shank and with gold bands wound around the hilt’. This reference is of particular interest because it confirms an officer of the Varangian Guard having used gold decoration on his own sword (in some cases a valued heirloom and given its own name, as in the case of Bolli’s ‘Leg-Biter’), while Michael Psellus describes a quite different weapon, ‘a single-edged sword of heavy iron’ known as the rhomphaia, which was ‘slung from the right shoulder’ of every palace guardsman. This must have been a dress sword and thus quite distinct from the double-edged blade of the traditional northern type which accompanied the spear (of a heavier type than that found in the Byzantine armoury) and, of course, the famous two-handed axe to make up the more typical complement of Varangian battle-weaponry.