Allowing for Snorri having misplaced the date of Harald’s expedition to Jerusalem, his account may still be correct in its claim for Harald having afterwards returned to the capital, particularly if he had made the journey in the capacity of military escort for porphyrogenita pilgrims. All of which might even point to the likelihood of Harald and his troop having already been admitted to the Varangian Guard proper by 1036, quite possibly as a result of their qualities demonstrated on active service in Asia Minor having attracted the attention of the emperor’s brother Constantine. In which case, Harald would surely have established himself as a formidable presence in the Varangian ranks by the following year when he and his men were assigned to Georgios Maniakes’ forces for the invasion of Sicily – and it is that campaign which offers the most likely setting for the saga accounts of personal contention between two of the most extraordinary soldiers of their time.
Crete had been lost to the Empire two centuries before, following the appearance in the Mediterranean of a fleet carrying 10,000 Muslim warriors who had been first expelled from Andalusia and some years later from Alexandria which they had captured in 818. Once more on the loose in the Middle Sea, they seized Crete, compelling the conversion of its inhabitants to Islam before enslaving them and turning their island into a pirate base for the corsairs who became the terror of the eastern Mediterranean. Within just a couple of years, Saracens had claimed Sicily too, although in this case at the invitation of a Byzantine renegade, a former admiral who rose in revolt after being dismissed from imperial service and sought the support of a North African emir. Thus in 827 a hundred Arab ships were brought to Sicily, the ex-admiral was slain and the island transformed into another, still more vexatious, stronghold of corsairs and forward base for Saracen incursions across the Strait of Messina even to the Dalmatian coast.
The Empire had eventually managed to reclaim Crete in 961, but Sicily remained in Saracen hands. Basil II had been planning its recapture and may well have achieved it had he not died in the year before his fleet was due to sail, so the island remained a threat to the security of Byzantine southern Italy, a plague upon ship-borne trade so important to the imperial economy, and an affront to an empire which still considered itself ‘Roman’. Through the decade after Basil’s death, however, Saracen Sicily became the battleground for a struggle between its two Arab chieftains, one of whom sought imperial assistance against his brother and rival. In response, the catepan (the local equivalent to the title of strategos elsewhere in the empire) of Italy was sent to Sicily in 1037, but meanwhile the other brother had appealed for help from the Caliph of Tunis who similarly despatched forces under the able command of his son Abdallah-ibn-Muizz, who attracted increasing support and so well outfought the catepan as to force his withdrawal (although in apparently good order) to the Italian mainland. The emperor Michael resolved to try again, and this time with his finest available forces – including 300 Norman mercenaries from Salerno as well as an elite company of Varangians from Constantinople – under the command of his finest general, who was, of course, the afore-mentioned Maniakes.
Said to have been Turkish-born and evidently of lowly origins, Georgios Maniakes was possessed of the physique of a giant and naturally endowed with ‘all the attributes of a man born to command . . . a voice of thunder, hands strong enough to shake gates of brass and the scowl upon his face terrible to behold’, as recalled by the chronicler Michael Psellus, who knew him personally. For all these qualities, Georgios had climbed only slowly to the peak of command, having started out on his military career as a menial servant – ‘among the baggage-men’, according to Psellus – but steadily rising through the ranks until he ‘attained the highest position open to a soldier’. First emerging as a commander – in the post of strategos of the lesser theme of the Euphrates Cities – during the reign of Romanus Argyrus, he rose to real prominence with his capture of the great city of Edessa on the mid-Euphrates in 1032, and so when the emperor Michael appointed him strategos of Longobardia he would have had every reason to believe Maniakes capable of achieving no less a triumph in Sicily.
This, then, was the Gyrgir who is said by Snorri to have come up against Harald while both were marching overland (although neither the campaign nor its location is specified). The Varangians were the first troops to arrive in the place where it was planned to set up camp for the night and so Harald was able to choose the higher ground, rather than the lower lying marshy land, on which to have his men pitch their tents. When Gyrgir arrived, he immediately ordered their tents taken down so as to make his own camp on the better ground, but Harald refused to comply, insisting on the Varangians’ independence of any command other than that of the emperor and empress to whom they had sworn allegiance (a privilege, incidentally, of which there is no known record). An argument between the two was approaching the point of drawn weapons when it was agreed to resolve the matter by the drawing of lots, in which Harald outwitted his superior officer by sleight of hand. Anecdotes based around the same ‘triumph by trickery’ formula are found so often in Snorri’s storytelling – and, indeed, throughout the whole saga literature – that this tale would be considered of dubious authenticity, even if a strikingly similar battle of wills were not to occur in a later passage of this same saga where Harald scores over his nephew Magnus in a dispute over precedence in ship-berthing.
All that is known of the mighty Maniakes, and especially of his merciless mode of military discipline, would put the story itself almost entirely beyond the bounds of credibility, and yet it might still have some genuine basis if it reflects the inevitability of contention between two men possessed of such powerful personalities formed by such dramatically contrasting backgrounds. In the event of such a confrontation, and when Maniakes is said by Psellus to have been so tall that ‘men who saw him had to look up as if at the summit of a mountain’, it is scarcely likely that any concession would have been made to a young prince, however physically impressive, still in his early twenties and recently arrived out of the north. Indeed, there is much to be said for the trenchant analysis made by Benedikt Benedikz when he suggests that Harald ‘as a somewhat undisciplined junior, was rather frequently carpeted by the Chief, and that the smart repartee and spectacular action contained in [the “Separate”] Harald’s saga and Heimskringla alike are much-expanded self-justifications, originally told by Harald and blown up by his flatterers’.6
Curiously enough, and especially in view of the prominence accorded him in Snorri’s saga, there is no mention of Georgios Maniakes in any surviving skaldic verse, yet the skalds do supply their own further confirmation of Harald’s involvement in the invasion of Sicily. Indeed, Valgard of Voll would seem to have been in Harald’s company during the later years of his Byzantine service and makes specific reference to the Sicilian campaign when he writes of Harald taking ‘a great force south of the broad lands . . . eventually Sicily was depopulated’. So too does Bolverk who would seem to be describing Harald’s courageous part in seizing a beach-head for the landing of Maniakes’ forces when he tells how ‘ships ran to shore and the lord [Harald] fought nobly, winning sand beneath him for a great army in the south of Sicily’.
Just such would have been a fully plausible deployment of Varangians, in view of the renowned Scandinavian accomplishment in seaborne warfare. It is also probable that a troop such as that under Harald’s command might have been deployed in the capture of smaller towns or coastal forts (of the type called riba-t, many of which were raised in Arab-held Sicily), although the northmen are not known to have used siege engines on their own account and the earlier record of viking sieges invariably attributes their success to guile. Snorri does not put a name to any of the fortified towns which he claims to have fallen to Harald’s Varangians, and neither does he indicate even the approximate whereabouts of three of them, specifically locating only one of the sieges in Sicily, and that the most glaring example of what has been called an ‘itinerant folk-tale’ (meaning a story long current in variou
s different traditions before its attachment to whichever current hero).
In fact, three of his four siege stories might be recognised as tales of that kind, the first of them telling of a Sicilian town which falls to Harald’s forces by means of a firestorm contrived by capturing birds who have flown out of the town’s buildings, fixing kindling, wax and sulphur to their tails and setting it alight, so as to return the poor creatures to their roosts as flying firebombs. Not only utterly implausible, the story is also notoriously unoriginal when it was also told of Harald’s contemporary, the Norman Robert Guiscard in Sicily, of the Russian princess Olga taking revenge upon a Slav town a hundred years earlier, and of the Danish viking Guthrum at the siege of Cirencester in the century before that.
Indeed, similar stories have been attached to Alexander the Great and Genghis Khan, but Snorri’s version probably originated in Armenia where the legend was associated with an emir of Baghdad before finding its way into the currency of Varangian tradition, which has been described by the modern authority Omeljan Pritsak as ‘a kind of “hatchery” for old Icelandic storytelling’. Pritsak’s essay goes on to explain how ‘Constantinople was the meeting place for peoples of different cultures, and the higher milieu of the imperial capital stimulated the soldiers to tell stories in which truth and fantasy could be easily combined’.7
A great many tales from those same wellsprings were brought back to Iceland by returning Varangians and thus passed down to the saga-makers, by which time some number of them had become attached to the heroic legendry accumulating around Harald. Such was assuredly the source of Snorri’s tale of the incendiary birds, as it must likewise have been for the story of Harald’s faking his own illness and death so as to enable his Varangians to gain entry into another besieged town with his ‘funeral’ cortege before casting aside the charade and seizing the victory. It is thought likely that Snorri’s version of this tale originated in southern Italy, where it was associated with Robert Guiscard’s capture of a monastery, before being picked up by Varangian tradition, yet a similar episode is attached to a more ancient hero by Saxo Grammaticus and other examples are found elsewhere in medieval sources, although usually those informed by Norman tradition.
Snorri’s third variation on this same theme tells of Harald having a tunnel dug beneath the town under siege, while maintaining an attack on its walls so as to distract the attention of the defending garrison. This ruse is perhaps more plausible than the first two, but the story is scarcely original when it had been widely current since classical times – and if Harald had been engaged in such a tactic, it would surely have involved Byzantine military engineers and been under the direction of a superior officer. Even so, it is perhaps worth noticing that Snorri once again mentions the ‘vast hoard of booty’ won by Harald’s troops after they had burst out of the tunnel to seize the town.
The fourth anecdote making up this quartet of siege stories in the saga is the one taken most seriously by historians and by reason of its involving Harald’s companion-in-arms, Halldor Snorrason, who was one of Snorri’s own ancestors and thus quite certainly the original source for a tale included in Heimskringla but in no other version of Harald’s saga. The story itself is simple enough, telling how Harald’s Varangian troop laid siege to the ‘largest and strongest, wealthiest and most populous’ of the towns they captured and one well defended against assault by a moat surrounding its walls. Harald’s ruse in this instance was to send some of his men, apparently unarmed and unconcerned, to engage in games within view of the town walls but beyond the range of the defenders’ weapons. This activity continued for some days while the townsfolk mocked them with taunts, eventually becoming so confident as to leave their gates open and thus providing the opportunity for Harald’s men to draw the weapons hidden in their cloaks and make a charge into the town. Harald let them lead the assault and yet he himself took longer than expected to follow up with the rest of his force, thus exposing the vanguard – including Halldor and his fellow-Icelander, Ulf Ospaksson, both described by Snorri as ‘outstanding warriors and very dear to Harald’s heart’ – to bear the brunt of fierce fighting. Some Varangians had been killed and many wounded before Harald reached the gates and even as he did so his own standard-bearer was slain. At which point he called upon Halldor to take up the standard, a command which the Icelander refused, accusing Harald of timidity in holding back from the fray. While Snorri offers swift assurance that ‘these words were spoken in anger rather than truth’, other evidence preserved in the Icelandic sources indicates Halldor’s loyalty to his future king having been tempered with a critical cutting-edge and, indeed, other versions of Harald’s saga include the same acrimonious exchange with Halldor, but which they otherwise append to the tale of the ‘fake funeral’ ruse. ‘Let the trolls carry the standard for you, you coward’ is perhaps the most convincingly idiomatic of these alternative renderings, all of which might be taken to suggest Halldor’s response to Harald having been a favourite feature of the storytelling for which he became renowned in Iceland in the years after his return home from the Norwegian court.
Snorri’s version of the story tells how Halldor himself was wounded in the fighting, suffering a deep gash to his face which left him marked with ‘an ugly scar for the rest of his life’ and it seems very likely that his scar might have prompted a variety of tales explaining how it came to be inflicted, each one including his sharp retort to the king as its punch-line. None of which lends especial credibility to Snorri’s story of the ‘Varangian games’, which may even have been contrived for inclusion in the saga as a convenient opportunity to introduce the genuinely historical characters of Halldor and Ulf, both of them reliably identified as sons of prominent Icelandic families, into the narrative of his Harald’s saga.
There is nowhere any indication of when these two remarkable men first joined Harald’s retinue, although it is not impossible that they might have been among those Icelanders who fought in Olaf’s army at Stiklestad and afterwards followed Harald to Russia and Byzantium. It is just as likely, though, that they were already employed in Byzantine mercenary service (possibly in company with their fellow-Icelander, Bolli Bollason, whose Varangian career is well-known from Laxdæla saga) when they joined Harald’s troop, perhaps attracted by its reputation for profitable plundering or accepting personal invitations which might well have been offered to fighting-men of the outstanding quality described by Snorri. However and whenever they were first recruited into Harald’s company, Halldor and Ulf evidently rose swiftly to become his principal lieutenants by the time of the Sicilian campaign, afterwards accompanying the rise and fall of his career in the Varangians of the City before returning with him to Scandinavia.
While Ulf spent the rest of his life in Norway where he was made a lenderman, and in Harald’s service where he held the premier military post of king’s marshal (or stallari), Halldor would seem to have settled uneasily into life at court and chose to go home to Iceland sometime around 1051, bearing with him all the tales of Varangian adventure which were to establish his reputation as a storyteller of outstanding authority. As the favourite son of a famous chieftain of Helgafell, Halldor makes appearances in a number of Icelandic sources, but the most impressive testimonial to his renown as a teller of tales is found in Morkinskinna, where the Tale of the Story-wise Icelander tells of a young man who came to King Harald in Norway where his storytelling kept the court enthralled through all twelve days of Yule. Harald himself was particularly impressed by the performance and so asked the young Icelander where he had learned the story he told. ‘It was my custom in Iceland to journey each summer to the Althing and it was there that I learned the story, piece by piece each year, from the telling of Halldor Snorrason.’ ‘In that case,’ said Harald, ‘it is little wonder that your knowledge is so excellent, and good fortune will attend you now.’
It should be said, of course, that Halldor’s reminiscences are unlikely to have been preserved intact and uncorrupted through almost two h
undred years of oral transmission, and even then that they represented only a small proportion of the Varangian lore – assuredly including many even taller tales – which found its way into Icelandic tradition and thus provided Snorri with his reservoir of source material. Consequently, Halldor is not necessarily to be blamed for those occasions when the saga’s chronology becomes unhelpfully confused or when its aggrandising enthusiasm bursts the bounds of historical credibility. Nonetheless, such passages from the saga are still of interest when they might reflect something of the feelings of Varangians about their commanders, as appears to be the case in Snorri’s chapter which contrasts Harald with Georgios Maniakes in the context of a campaign only loosely identified and yet bearing unmistakable resemblance to aspects of the Sicilian invasion.
Snorri claims that Harald tried always to keep his own men out of the heat of battle when the forces were in action together and yet drove them fiercely against the enemy whenever they were engaged as a separate unit. Thus victories were won when Harald was in sole command and the troops acclaimed him as a better commander than Maniakes, who countered that the Varangians were not giving him their full support and responded to the criticism by ordering Harald to take his men off on their own while he himself remained in command of the rest of the army. When Harald did so, Snorri tells of his taking not only his Varangians, but also a contingent of ‘Latin-men’, by which must be meant Norman mercenaries, and this reference alone would associate the story with the Sicilian campaign where Byzantine forces did include a Norman contingent, although Snorri’s narrative places it earlier, presumably in Asia Minor, and probably conflates tales told of more than one theatre of operations.