by John Marsden
The eleventh-century Byzantine chronicler Michael Psellus remembers Basil as having been ‘well aware of the disloyalty of the Romans’ and understandably so because he was just eighteen when he became emperor in 976 and the first dozen years of his reign were plagued by contention with two rival claimants to his imperial crown. That distraction encouraged the ambitious Bulgarian tsar Samuel to extend his dominions at the expense of the Empire and when Basil led an army into Thessaly against him it suffered devastating defeat in a Bulgar ambush at the pass called Trajan’s Gate. While Basil would eventually take the ferocious revenge which was to earn him the soubriquet of Bulgaroctonos (‘Bulgar-slayer’), he was more immediately concerned with the most dangerous of his rivals, Bardas Phocas, whose forces were already converging on Constantinople in 988. With his own military still crippled by the losses inflicted at Trajan’s Gate, Basil turned for assistance to Vladimir of Kiev, who supplied him with the force of 6,000 axe-wielding Scandinavian fighting-men which threw back Bardas and his army in the early weeks of 989 and went on to achieve their total destruction at Abydos on the coast of the Dardanelles some three months later.
By way of return for his generosity, Vladimir requested for himself a Byzantine bride – in the person of Basil’s sister, the porphyrogenita Anna – and even agreed to facilitate the union by accepting conversion of himself and his people to the Orthodox faith. Although Basil was in no hurry to arrange the marriage, which was not solemnised until the summer of 989 and only then after Vladimir had increased the pressure by occupying the Byzantine outpost of Cherson on the Black Sea, he was so greatly impressed by the performance of these Varangoi that he formed them into an elite regiment of the Hetairia and the one which was also to provide his own personal bodyguard. So it was that Basil created the celebrated Varangian Guard which was to serve his successor emperors for more than two hundred years until the fall of Constantinople to the Fourth Crusade in 1204.
It is often assumed – sometimes, indeed, actually asserted – that it was this Varangian Guard of which Harald was later to take supreme command. An unsuspecting reading of Snorri’s saga might give no less an impression, but it is one emphatically denied by the Book of Advice to an Emperor when it supplies precise detail of the ranks which Harald held as an illustration of how it was both unnecessary and undesirable to promote foreigners, however able they might be, to positions of the highest rank. By way of reward for his service in the Sicilian and Italian campaigns, Harald was first appointed to the rank of manglavites (which reliably confirms his entry into the ‘Varangians of the City’) and later promoted to that of spatharokandidatos on his return with the emperor from the suppression of the Bulgarian revolt in 1041. While there are references in Byzantine sources to commanders of the Varangian Guard with the title of Akoluthos (or ‘Acolyte’ in the sense of ‘follower’, presumably of the emperor himself), the officer in command of the manglavites held the rank of protospatharios, but neither of the two ranks held by Harald approached such seniority. Indeed, their comparatively modest status is unmistakably acknowledged by the author of the Advice when he observes with approval that Araltes ‘did not grumble about the titles of manglavites and spatharokandidatos with which he had been honoured’.
In fact, no less an authority than Sigfús Blöndal is of the opinion that ‘if he held no higher rank in imperial service, then he is unlikely to have ever held independent command of an army larger than the small force needed to reduce a small fortress or minor township’.5 Just such a ‘small force’ need not have been so very much larger than the ‘five hundred men’ said by the Advice to have accompanied Harald on his arrival in Constantinople, although there is no reason to doubt Snorri’s claim for new recruits having been attracted to his troop by the growing reputation of its commander, and especially by his well-documented accomplishment in garnering profit by plunder. A mercenary warrior, by definition, fought for financial gain, and there can be little doubt that it was the prospect of great wealth which had brought Harald to the land of the Greeks in the first place – a factor which must never be overlooked in any realistic reconstruction of his military career in Byzantine service.
Having described Harald’s arrival in Constantinople, Snorri goes straight on to tell how in the ‘same autumn he sailed with some galleys together with the fleet into the Greek sea’. This would seem to have been a customary form of introduction into Varangian service and especially in the early years of the reign of Michael IV when Arab pirates or corsairs (kussurum in the Norse of the sagas) represented a persistent menace to ship and to shore throughout the eastern orbit of the Mediterranean.
Byzantine annals for the 1030s record a great fleet from Arab-held Sicily and north Africa raiding the islands of the Aegean, the Greek mainland and the shores of Thrace, even taking the town of Myra in Asia Minor before it was overtaken by a naval force, commanded by the strategos of the Cibarrote, which destroyed most of its ships and either butchered or enslaved their crews. The Cibarrote, on the southern coast of Asia Minor, was one of the maritime themes responsible for the upkeep of the ‘Fleet of the Themes’, which bore much of the responsibility for defence of the Aegean and its surrounding waters against Arab piracy. A similar policing duty was borne by one of the two divisions of the imperial fleet, which was also regularly assigned to naval support of military expeditions (the other division being responsible for transport of the imperial family and defence of the capital).
Foreign mercenaries were often recruited to man the ships of these fleets, many of them being Varangians chosen to begin their service to the Empire in its navy on account of the seafaring and sea-fighting expertise for which the northmen were traditionally renowned. While the capital ships of the fleet (larger vessels known as dromoi) were manned by a complement more than two hundred strong and equipped with siphons to unleash the devastating incendiary flare known as ‘Greek Fire’, the Varangians were more usually deployed to crew the lighter, faster and more manoeuvrable craft called ousiai and considered especially effective in pursuit of the similarly speedy lateen-sailed Arab dhows. So when Snorri writes of Harald having ‘sailed widely around the Greek islands and inflicted heavy damage there upon the corsairs’, it would seem virtually certain that he and his company were serving aboard these ousiai, each of which carried a crew of about 110 including 50 or more soldiers. Snorri’s reference would also indicate his theatre of operations as the Aegean Sea, although there is a further possibility, which may be implied by the Advice and is certainly suggested by a reference in one of Harald’s own verses to sailing ‘on the sea of Sicily . . . with a warrior crew’. Unless this refers to his later involvement with Maniakes’ invasion in 1038, it could be taken as evidence for Harald’s earlier naval service having included an assignment to the auxiliary fleet stationed off the Sicilian coast by Michael IV to safeguard the seaway between Sicily and North Africa.
For all the efforts of his skalds to portray Harald at sea as a free-spirited viking, he and his men would have been subject to the same command structure as other foreign mercenaries engaged on naval service, taking their orders from officers of the Hetairia who reported in their turn to the strategos in command of the Fleet of the Themes or to the admiral directing the operations of an imperial fleet. Nonetheless, the claim made by the skald Bolverk for Harald ‘having begun a fight where he wanted to have one’ during these early years in Byzantine service might not be entirely without foundation if a superior officer had been sufficiently impressed by Harald’s undoubted abilities and rapidly acquired experience to encourage his initiative by allowing him more than usual licence in pursuit of the enemy.
There is also every reason to expect that Harald would have had his own keen financial interest in seeking out and overhauling a well-laden Arab pirate craft, because his saga in Flateyjarbók has a precise note of the profit thereby accrued. Just 100 marks was due to the emperor’s coffers for every corsair taken, and once that had been rendered any surplus plunder was kept by Harald and h
is men. Varangians in naval service were paid less generously than those in the Hetairia, so the entitlement to a share of captured booty must have represented a lucrative source of performance-related pay and one of which Harald would certainly have taken full advantage. Although Snorri has nothing to say about this remunerative aspect of Harald’s sea-service, he soon has occasion to tell of the ‘hoard of money and gold and treasure of every kind’ acquired on the subsequent land campaign in ‘the parts of Affrika which the Varangians call Serkland’.
Snorri’s source for these references would appear to have been surviving verses by the skald Thjodolf, one of which is actually quoted in the saga, but doubt is cast upon his interpretation of that source material by the absence of any record of a Byzantine expedition to Africa during the reign of Michael IV. When the poetry of the skald represents the most closely contemporary evidence, the disparity between saga account and historical record must be attributed, once again, to a misinterpretation on the part of the saga-maker, which can only derive from the use of the Norse word Serkir (often translated as ‘Saracen’) as a generic term for Arabs and Arabic-speaking peoples, and likewise to the use of Serkland for any region of the Islamic conquest wherein they might have been encountered.
Snorri had already mentioned the Byzantine general Georgios Maniakes, even implying his involvement with Harald in the action against Arab corsairs in the Greek islands, and yet, while Maniakes was effectively commander-in-chief of Byzantine forces at that time, he was unlikely to have been in immediate command of naval operations. So when Snorri places an anecdote (which will be considered in more detail later) concerning Harald and Maniakes immediately before the chapter bearing on Serkland, it would suggest that Harald and his troop had been brought ashore when the campaign against the corsairs was extended inland to strike at their support bases in Asia Minor, where there was conflict with the Arabs during the early years of Michael’s reign and where Georgios Maniakes and the emperor’s brother Constantine, strategos of Antioch, were in command of the Byzantine forces.
Having driven the Saracens out of Asia Minor by the end of 1035, the Byzantine campaign pressed on through Armenia and Syria towards the Euphrates and it may well have been that theatre of operations to which Thjodolf referred in his strophe telling of Harald’s capture of ‘twice-forty towns in Serkland’ before he ‘bore his war-making to smooth Sicily’. Although the total of ‘twice-forty towns’ is realistically read only as an approximation – the figure ‘eighty’ being thought to represent a traditional indication of ‘very many’ – there is no reason to doubt the placing of these conquests prior to Harald’s known involvement in the Sicilian invasion some two years later. Other lines from Thjodolf, preserved in Flateyjarbók and probably known to Snorri, although not quoted in the saga, refer to the ‘king of Africa’ finding it difficult to guard his people against Harald and these may also allude to the campaign in Serkland or, perhaps more probably, to the Caliph of Tunis whose son commanded the Arab forces encountered in the invasion of Sicily. Another half-strophe of skaldic verse, also preserved in Flateyjarbók and in Morkinskinna, clearly does recall the fighting in Serkland and is worthy of quotation here because it is attributed to Harald himself, although composed in Norway many years after his return from Byzantium.
One other time there was, when I
reddened blades from my homeland;
the sword singing in the Arab town
– and yet that was long ago.
It has thus far been possible to follow the chronology of Snorri Sturluson’s Harald’s saga in Heimskringla quite closely – as, indeed, was my original intention – but his sequence of events through the next five or six years of Harald’s Byzantine career diverges so far from that suggested by more reliable historical record as to require some rearrangement here. Snorri may well have already entangled anecdotes refering to the Sicilian invasion of 1038 with others, similarly derived from Varangian tradition, relating to Harald’s earlier activities located by the skalds in the regions they know only as ‘Serkland’.
At this point in the saga, however, its narrative enters upon a further series of anecdotal accounts concerning four sieges (only one of which is specifically placed in Sicily) preceded by two episodes refering to Harald’s contentious relationship with Georgios Maniakes, which might be assigned either to the Sicilian invasion or the earlier campaigning in Asia Minor. The historicity of almost all these stories is dubious at best, especially when none has the support of any skaldic verses, while some of them can be safely dismissed as apocryphal. There is another chapter, however, which quite clearly derives from court-poetry – two strophes quoted in the saga and both attributed to the sightless skald Stúf Kattason – describing Harald in the Holy Land and thus representing reliable evidence for a genuinely historical Varangian assignment, even though its true character is largely disguised by the aggrandising enthusiasm of the skald and the saga-maker.
The initial difficulty with the saga account is Snorri’s placing of Harald’s journey to the Holy Land after his return from the Sicilian campaign (which had taken him on into southern Italy in its later phase) and thus in the year 1041, when the more reliable evidence of the Advice places him in Bulgaria with a contingent of the Varangian Guard accompanying the emperor in suppression of the rebellion. The historical record of the turbulent events crammed into these few months would allow only just enough time for Harald’s withdrawal from the Sicilio-Italian theatre, his promotion to the rank of manglavites, and his participation in the Bulgarian campaign (all reliably recorded by the author of the Advice) and would thus entirely exclude the many weeks required for an expedition as far as Jerusalem.
The first of the strophes accredited to Stúf tells of the ‘weapon-bold warrior’ having journeyed from Grikaland ‘Jerusalem-wards’ and of Palestine being rendered ‘unburned in submission to his hand’, while the second tells of his enforcing justice on both sides of the Jordan, having ‘made an end of men’s treacheries, inflicting sure trouble for proven crime’. Although evidently derived from these verses, Snorri’s account is clearly intended to portray Harald in the guise of a pilgrim making generous offerings at the holy places and enshrined relics, its underlying motive probably being to ensure that its hero was not to be outdone by other northern magnates such as Cnut of Denmark and Jarl Thorfinn of Orkney, both of whom made pilgrimages to Jerusalem in the eleventh century. While the skaldic verses can be taken as authoritative evidence for Harald’s journey to the Holy Land, it is their reference to his punishment of criminal elements which might be said to supply the vital clue to what really brought him there.
There is, first of all, no question of his having ‘conquered’ Palestine for the empire, when relations between Constantinople and the Caliphs of Egypt had been remarkably harmonious since 1027, in which year it had been agreed that the emperor Constantine VIII should be allowed to rebuild the church of the Holy Sepulchre. There had been no further progress in that direction through the last year of the ailing Constantine’s reign, nor through that of his immediate successor, Romanus Argyrus, until the succession of Moustansir-Billah to the Caliphate in 1035. Perhaps because he was the son of a Byzantine mother, this new caliph was generously tolerant of other faiths and strongly opposed to religious persecution, having already freed many thousands of Christian captives before he signed a thirty-year peace treaty with Michael IV in the following year. Included in this treaty was a renewal of the agreement to permit restoration of the Holy Sepulchre and so architects, masons and other Byzantine craftsmen were immediately despatched to Jerusalem, escorted on their journey by a troop of Varangians entrusted with their protection from attack by desert bandits. While there is every reason for confidence in the ability of formidable Varangian fighting-men to deal with such predators, their deployment on this escort duty is of particular significance because the choice of such elite soldiery would indicate the working party having been accompanied by very high-ranking pilgrims, probably even members of
the imperial family, taking advantage of a rare opportunity to visit the holy places.
The most likely imperial personages to undertake such a pilgrimage were the empress Zoe’s sisters Eudocia and Theodora, both of whom were nuns at that time, and their eminence alone would have ensured that all the travellers would have been welcomed with open gates at every point along the journey, a reception naturally interpreted by the skald as the willing ‘submission’ to Harald as he progressed along the same route. All of which corresponds so well to much of the detail encoded in Stúf’s verses as to propose the Varangian escort having been none other than Harald and his troop, a possibility – even if no more than that – which would provide the most plausible basis for Snorri’s saga account, and might also have some further bearing on Harald’s involvement in events in Constantinople some eight years later.
