Nonetheless, a ceremonial association with sword-bearing is clearly indicated by the spatharo- prefix found in the titles applied to ranks such as that of spatharokandidatos, to which Harald was promoted on his return from Bulgaria, and likewise that of the superior protospatharios, both of these having their origin in an earlier Hetairia lifeguard called the basilikoi anthropoi (literally the ‘emperor’s men’) which pre-dated the formation of the Varangian Guard. The Laxdæla saga account of Bolli Bollason’s return from Constantinople also supplies some evidence for the dress uniform of the palace guard when it describes the eleven men of his retinue ‘dressed in clothes of scarlet’ and Bolli himself wearing ‘a cloak of red scarlet given him by the emperor of Byzantium’. This has been taken to indicate ‘red scarlet’ as the uniform colour of the manglavites, while that worn by officers of the Hetairia holding the honorary rank of spatharokandidatos had traditionally been white (with the distinguishing badge of a golden torque) and may still have been so in the eleventh century. Again, it must be stressed that these were items of ceremonial garb as distinct from war-gear, which would have been essentially a helmet, shield and coat of ring-mail for a Varangian on campaign or even garrison duty in the capital.
The conical helmet might well have been of ‘spangenhelm’ construction, although some would have reflected Slavic or Turkic influence and, in the case of the more affluent warrior, may even have been decorated with precious metals. The longer a man remained in mercenary service, of course, the more likely he was to adopt items of foreign arms and armour, but Scandinavian military taste was invariably conservative in character and the mail-coat would seem to have been at least as characteristic of Varangian war-gear as was the axe. So it was in Harald’s case, although his mail-coat is said by Snorri to have been unusually long and so would have been similar in style to those worn by the mailed Norman knights with whom he served in Sicily. The shield most often used by Varangians of Harald’s time would have been of the traditional Scandinavian circular design, and thus much larger than the disc-shaped shields carried by Byzantine cavalry, although the long, tapered ‘kite’ shields favoured by the Normans were finding their way into Russian and Byzantine armouries through the eleventh and twelfth centuries when fragments of saga evidence also suggest shields of that type brought back to Iceland, presumably by returning Varangians.
One further item of military equipment particularly associated with Harald is his banner, known as Landeyðuna (‘Land-ravager’) and noticed on more than one occasion in Snorri’s saga where it is said to have been his most highly prized possession. The name alone bears testimony to the banner’s long active service before it was raised on his last campaign of 1066 and if this was the same banner which featured in Halldor’s anecdote connected with the siege story then it must have come into Harald’s possession while he was in Byzantine service, and perhaps in emulation of those carried by the standard-bearers of the Tagmata regiments. These regimental standards were highly honoured, in much the same way as the eagles carried by the old Roman legions, and yet smaller units within a regiment (companies or cavalry squadrons known as banda) also had their own banners, known as bandophorai, one of which might have found its way into Harald’s hands – but the story of Land-ravager is a long one indeed, and worthy of its own appointed place later in these pages.
What can be said of Land-ravager at this point – and especially if Thjodolf’s phrase ‘burner of the Bulgars’ can be taken at face value – is that it would have fluttered proudly over Harald’s troop on the march into Constantinople with the imperial forces returning victorious from the Bulgarian campaign in 1041. Michael Psellus – reporting, as usual, at first-hand – describes the entry of the emperor’s army into the city as ‘a brilliant occasion’, with the whole populace thronging to greet their emperor as he ‘returned in glory to his palace, bringing with him a host of captives, among whom were the most notable men of the Bulgars, including their leader, the pretender [Peter Deljan] himself, minus his nose and deprived of his eyes’.
The whole performance was a typically Byzantine blend of splendour and savagery, reaching its climax when the Bulgar prisoners were force-marched through the stadium of the Hippodrome in which the high and low usually gathered to watch horse-races, and yet for Psellus the really tragic figure was that of the emperor Michael himself: ‘I saw him on this occasion . . . swaying in the saddle of his horse. The fingers gripping his bridle were as if those of a giant, each of them as thick and as large as a man’s arm, the result of his internal trouble, while his features preserved not a trace of their former likeness.’ This was to be the last public appearance of an unlikely and yet remarkable Basileus whose personal tragedy has been justly recognised by one of the outstanding modern English historians of Byzantium: ‘Few emperors had risen from more lowly origins, or by more questionable methods; none suffered a more agonising end. He possessed wisdom, vision and courage . . . and in the reigns that followed, there would be many who regretted his loss’.10
One who certainly would have had his own reason to do so was the newly promoted spatharokandidatos of the Varangian Guard, because the passing of Michael the Paphlagonian – who had himself carried to a monastery of his own foundation on 10 December 1041 and there took the monk’s habit and tonsure before he died on the same evening – was to mark the decisive point of downturn in Harald’s fortunes in imperial service.
The fast-failing health of the emperor had been apparent in court circles for more than a year – and not least to his brother, John the Orphanotrophus, who was thus in need of a successor who could secure the avaricious interest of himself and his kin. His one remaining brother with manhood left intact was Constantine, but he was now as widely disliked as John himself. Stephen, the lowly shipyard worker promoted to admiral of the imperial fleet by reason of his marriage to Michael’s sister Maria, had been dead for over a year but had left a son – also named Michael – and it was he who offered the family its last potential candidate for the succession if he were to be accepted as an adopted son by the ailing emperor and an empress now in her sixties.
All of which had been neatly accomplished by the scheming Orphanotrophus before the emperor went away to war and so his favoured successor was already on hand when the death of Michael IV was announced. All that was needed was his endorsement by the porphyrogenita Zoe, now utterly alone and as gullible as ever, who could hardly refuse the proclamation of her adopted son as the emperor Michael V – although he was to be more popularly known as Michael Calaphates (‘the Caulker’) in satirical reference to his father’s erstwhile employment with the tar-brush. In the event, John’s manipulation proved disastrous for his family and, indeed, most immediately for himself because the uncle preferred by the new emperor was Constantine whom he raised to the eminence of nobilissimus and appointed to command of the forces as ‘Grand Domestikos’, while the Orphanotrophus was shortly to find himself aboard a galley carrying him into the exile from which he was never to return.
Not the least astonishing move in this reshuffling of imperial favourites was the restoration of the man only recently disgraced for taking a whip to the emperor’s late father, because Georgios Maniakes was released from imprisonment and offered the ships and men he needed to reclaim Italy from the new Norman ascendancy (where just four towns remained in imperial control) and to make yet another attempt to win back Sicily from Saracen control. The return of the formidable Maniakes was never going to bode well for Harald, especially when the young emperor apparently shared a similar ill-disposition towards Varangians, replacing them as his bodyguard with ‘young Scythians’ (almost certainly Pechenegs) said by Psellus to have been eunuchs ‘who knew his [Michael’s] temper and were well suited to the services he required of them’. The Varangian regiment itself was not disbanded and those of its members who had formerly served as the imperial lifeguard were simply reassigned to garrison duty in the capital. While these ominous developments cannot be placed into precise sequence, they mu
st have been implemented within the first few months of 1042 and so might seem to form a prelude to the more dramatic events which were to bring the short reign of Michael V to its gruesome conclusion in the third week of April.
One of the first of these is of particular importance here by reason of its involving the imprisonment of three officers of the Varangian Guard – namely, Harald Sigurdsson, Halldor Snorrason and Ulf Ospaksson. It should be said, first of all, that there is no reference to Harald’s arrest and confinement in the most closely contemporary evidence of skaldic verse or Byzantine records, so all that is known of this episode derives from later northern sources and principally from accounts included in the different versions of Harald’s saga, of which that set down by Snorri Sturluson must have its own claim to be the best informed. The original sources informing all these saga accounts can only have been stories brought back to the northlands by returning Varangians, as Snorri himself effectively admits in the saga, and so his family connection with Halldor Snorrason must lend at least some measure of first-hand authority to his version of events.
Nonetheless, Snorri’s account is grossly distorted by misinterpretation and confusion, and most unhelpfully in his sequence of events, as in his opening passage which links Harald’s imprisonment (from which he had evidently escaped by 21 April) with his decision to leave Constantinople, being eager to return to his homeland again ‘having heard’ that his nephew Magnus had become king of both Norway and Denmark (which indeed he did, but not until after the death of Hardacnut early in June of the same year). Snorri goes on to tell how Harald’s intention to depart so enraged the empress Zoe that she accused him of defrauding the imperial treasury of plunder won on expeditions under his command – even though the true reason for her anger was of a more personal nature.
Harald is said to have asked for the hand in marriage of ‘Maria, the beautiful maiden daughter of Queen Zoe’s brother’ and to have been refused, although ‘some Varangians who had served in Miklagarð and returned to Iceland’ claimed the empress herself wanted to marry Harald and that this had been the true objection to his request to depart the capital, ‘although a quite different story was given to the public’. Having thus confirmed Varangian tradition as his source of information for this passage of his narrative, Snorri immediately demonstrates how such source material can so easily and often fall short as reliable historical record. The only Maria known in the court circle of the time was the emperor’s sister and the admiral Stephen’s widow who would scarcely correspond to Snorri’s description, so if this ‘Maria’ existed at all she could not possibly have been a niece of the empress when Zoe had no brother and neither of her two sisters had ever married.
It is, of course, not at all unlikely that a young officer might have enjoyed some sort of liaison with a Byzantine noblewoman and yet no Varangian, however well born, would have been considered a proper choice of husband for such a lady when Greek society took so dim a view of the regiment’s character. Neither would an empress in her sixties have considered – or been allowed to consider – a marriage so far beneath her imperial dignity, even though another anecdote, which is found in Morkinskinna where it tells of Harald’s utterly ungracious response to Zoe’s request for a lock of his hair, would suggest that the two names were linked by salacious gossip (unfounded or otherwise) apparently current in mercenary barracks at the time and assuredly providing the source of both saga references.
The allegation of Harald’s defrauding the emperor (also attributed to Zoe in Fagrskinna) is quite another matter, however, and especially in view of so much evidence preserved elsewhere in the sagas, not least in that set down by Snorri himself which makes frequent mention of the wealth accumulated by Harald in Byzantine service. When first on campaign in Serkland, Harald was already gathering a ‘hoard of money and gold and treasure of every kind’ which he sent ‘in the care of trusted men to Holmgarð’ for Jaroslav’s safe-keeping, while two of Snorri’s four siege stories conclude with similar notes of the ‘immense booty’ collected when the towns had fallen to Harald’s guile. Presumably most of this plunder also found its way to Novgorod because, when Harald eventually reached Russia on his return from Constantinople, he is said to have spent the winter gathering together ‘all the gold he had taken from Miklagarð, together with other valuable treasure of all kinds’ into a hoard ‘greater than had ever been seen in the north in one man’s possession’.
At which point, Snorri comes up with the dubious explanation that much of this wealth had been acquired by means of Harald’s having thrice taken part in pólútasvarf, a term which he interprets as ‘palace plunder’ and a Byzantine custom whereby Varangians were allowed to ransack all the imperial palaces on the death of an emperor and to take as much of his treasure as they could carry away. There has been evidence found for a similar custom of ‘palace plunder’ known in Rome after the passing of a pope or his bishops (until abolished in 904) and likewise in Baghdad on the death of the caliph, but no such practice was known in Constantinople and Blöndal dismisses any such idea as ‘inconceivable’ in a Byzantine context (although he admits it is not impossible that guards on duty at the time of an emperor’s passing might have been allowed to take certain precious objects as personal mementoes).
Having said that, Blöndal does subject the whole question to a meticulous examination leading to his suggestion that the key may well lie in Snorri’s use of the term pólútasvarf. Russian in origin, of course, it was applied there to a tax-collecting expedition accompanied by a military escort, not infrequently comprised of Varangian mercenaries who were remunerated with a share of the tribute collected. Mention has already been made here of the probability that the young Harald would have been engaged on just such duties with Eilif Rognvaldsson in Novgorod, but he spent no more than two winters in the north of Russia and so would have had no time to amass any very great wealth before he left for Byzantium. Neither, indeed, would Snorri’s explanation suggest as much, because he specifically applies pólútasvarf to Harald’s Byzantine service, and thus leads Blöndal to his proposal of the term ‘having become Varangian slang for tax-gathering expeditions in Imperial service’.11 If so, then the inference must be of Harald’s extortion of unlawful revenue for himself and his men while engaged on tax-collection in imperial service, although the substance of the allegations brought against him in Constantinople bears only upon his holding back more than the legitimate share of battle-booty.
While the saga leaves no doubt of Harald’s Varangian career having been ‘mercenary’ in every sense of the term, it is still impossible to know whether or not he was justly accused, especially when there is the further possibility that Georgios Maniakes might have had some part to play in putting Harald and his lieutenants in prison. Released from his own imprisonment and restored to imperial favour, he was in a position to take revenge on those who had been his enemies just two years before, including Harald and his troop who may even have taken Stephen’s side in the acrimony which was to deprive Maniakes of command. If the general did have any knowledge, or even suspicion, of wrong-doing on Harald’s part, he would have been well placed to set retribution in process, a possibility further supported by stories in the Morkinskinna and Flateyjarbók versions of the saga which tell of at least one earlier occasion when Gyrgir had complained to the Emperor of Harald wanting to keep all the booty that he took for himself.
On this occasion, however, Snorri himself makes no attempt to implicate Maniakes (who, by this time, would already have left for Italy and Sicily) and lays the blame for Harald’s arraignment squarely upon the empress Zoe while thrusting the responsibility for his arrest and imprisonment upon Constantine Monomachus, ‘who at that time ruled as king of the Greeks’. At this point it is clear that Snorri’s chronology has gone completely awry and so much so as to make a nonsense of the conclusion to the next passage of his narrative, which tells of Harald having seen a vision of St Olaf promising to come to his aid, just before he, Halldor and Ulf were th
rust through a door on the street and into their cell inside a tall, roofless tower. As fully expected, the saint was as good as his promise and made yet another appearance in Constantinople, this time to ‘a lady of high birth’ he had once healed of some ill whose aid he now sought in the rescue of his brother. Following Olaf’s supernatural guidance, the noblewoman came to the prison on the following night, bringing with her two servants who set ladders against the wall and scaled the tower so as to lower down a rope and haul the three prisoners up out of their cell to freedom.
The more expansive variation on the same theme in Morkinskinna also has a vision of Olaf appearing to Harald as he approaches his prison, but in this version to advise him as to the whereabouts of a hidden knife. With this weapon and the assistance of his two loyal lieutenants, Harald is able to slay the fearsome serpent lurking in the dark cell before the three make their escape. Curiously, it is this story which is taken up by Saxo Grammaticus in his history, having learned it from the Danish king Valdemar who claimed to have the very same knife in his own possession. Snorri must have known of this story which would seem to have formed part of Halldor’s repertoire – one of the tales told of Halldor refers to a mocking rhyme claiming he had only sat upon the wurm and left the knife-work to Harald and Ulf – yet he would seem to have given it so little credence that it was omitted from his own saga. In fact, it might be thought rather more credible than any supernatural intervention by Olaf, because a snake or similar reptile would have been a very likely resident in a Byzantine dungeon and one easily re-cast by a saga-maker into some more alarming form of monster, especially in view of a number of similar ‘prisoner in snake-tower’ tales of south-east European origin current during the twelfth century.
They were evidently of no interest whatsoever to Snorri, however, because his account of the rescue by way of the roofless tower moves swiftly on with Harald’s urgent quest for revenge. Having been brought to freedom, he went directly to his Varangians who took up their weapons and followed him to find the emperor asleep in his bedchamber, from which he was seized to suffer the putting out of his eyes. At this point in the narrative, Snorri summons up the support of his most authoritative source of evidence in the form of two passages of skaldic verse by Harald’s court-poets, the first attributed to Thorarin Skeggjason, the second to the oft-quoted Thjodolf, and both informed by Harald himself for their description of his blinding of a Byzantine emperor – although one whom neither skald identifies by name.