While the distinctly Scandinavian atmosphere in Novgorod and Ladoga – where, for example, the Norse tongue was to be heard spoken – would not have been so very unfamiliar, he would have been increasingly aware of the advance of Slavic influence on the Rus, and not least in the orbit of Jaroslav’s court. Not so much farther afield, he would have encountered the Balt and Finno-Ugrian peoples, while further south – as, for example, on the Polish campaign – Harald was to become much more widely acquainted with the variety of cultures which had already exerted their influence on the Rus, and not least in the military sphere. There is, of course, no way of knowing the extent to which Harald himself was similarly influenced, in particular as regards his weaponry and war-gear, although it was usually characteristic of the mercenary employed abroad to bring his own style of arms and armour with him and afterwards to return home with those of the warriors with whom he had fought. Harald may very well have brought a sword to Russia, but probably little else unless he had been supplied with a helmet and mail-coat at the Swedish court. It is more than likely, of course, that he would have equipped himself with new war-gear among the Rus, where the more fashionable members of a druzhina displayed the influence of both Byzantium and the steppe in their adoption of lamellar armour (formed of upward-overlapping metal, horn or leather plates laced together with leather thongs), in the decorated metalwork of their helmets, or even in their use of the typically Turkic single-edged curved sabre.
Interestingly though, Varangian warriors would seem to have stayed loyal to the ring-mail, the double-edged straight-bladed sword and, most especially, the characteristic battle-axe they had known and used in the northland, so it can be fairly safely assumed that Harald and his troop would have been similarly armed when they left for Byzantine service. The most likely Russian influence would probably have been on their more basic clothing, where straight-legged Scandinavian trousers would long since have been replaced with the baggy Slav style adopted by the Rus, while fur cloaks which had proved their worth in sub-Arctic winters would soon realise a new value in the marketplaces of Grikaland.
As to Harald’s choice of route to Byzantium, the version of his saga in Morkinskinna supplies its own unfortunate illustration of the hazards awaiting a saga-maker who misunderstands the evidence he finds in skaldic poetry. Having read Thjodolf’s description of Harald’s march through ‘the land of the Langbards’ and Illugi Bryndælskald’s mention of his combat with Franks, the author of the Morkinskinna saga picked up other skaldic references to construct a route which took Harald from Russia, through Wendland in northern Germany, to France and on to Lombardy in the north of Italy before he reached Constantinople. Thjodolf’s apparent reference to Wends has already been considered here, but his reference to Langbardaland actually meant the Byzantine province of Longobardia in southern Italy and not Lombardy in the north. Similarly, by Frakkar, or ‘Franks’, Illugi was referring to the French Norman mercenaries who mutinied from Byzantine forces in Sicily to join a revolt raised against imperial lordship in the south of Italy. Thus neither skald was referring to the route of Harald’s journey to Constantinople in 1034, but both were actually celebrating his service with the Varangians of Byzantium in the southern Italian campaign of 1040.
Although no saga source specifically says as much (perhaps because the voyage to Byzantium was considered insufficiently remarkable for the skalds to notice in such detail), it is virtually certain that Harald’s way to Constantinople followed the same Dnieper route taken by the Rus merchant fleets which he would have seen being assembled when he looked down from the Snake Ramparts raised to protect the fortified marshalling yard at Vitichev some 28 miles downriver from Kiev. The most thorough description of this Dnieper route is that contained in De Administrando Imperio, a treatise written for the education of his son by the Byzantine emperor Constantine VII in the mid-tenth century, which tells how tribute-collection began in the November of each year and how the furs and slaves thus acquired were brought down the various rivers flowing into the Dnieper around Kiev. There the tribute became the cargo loaded aboard the monoxyla, similarly brought to Kiev where they were sold to the Rus by Slav boat-makers, which formed the merchant fleet setting out downriver in June and bound for the Black Sea.
The one item of closely contemporary evidence for Harald’s arrival at Constantinople is found in a strophe by the skald Bolverk which would fully correspond to his having followed the Dnieper route, and yet unmistakably indicates his ships having been of Scandinavian type, possibly including knorr if they were also shipping any quantity of goods for trade, but with warrior crews and rather more dignity than would have attended the barge-like monoxyla. Even so, their own craft would still need skill and care in the handling, especially when they reached the notorious stretch of rapids along the lower Dnieper, some of which required crews to climb overboard so as to manhandle their vessels between rocks and others where the cargo had to be unloaded and the boat carried along the bank by its crew while guards kept careful watch for a Pecheneg ambush.
Steppe raiders were still a hazard on the passage through the last of the rapids, fast flowing but also fordable and vulnerable to attack from archers on the overlooking cliffs. Beyond this point and out of danger from predators, the craft could be brought to shore and their crews rested on St Gregory’s Island before continuing out into the Black Sea where the course held close to the shoreline until it reached the Danube estuary and there turned southward across open sea to the Bosporus – and it would seem to be this passage which is described by Bolverk in his strophe celebrating Harald’s first sight of the Byzantine capital:
Bleak gales lashed prows
hard along the shoreline.
Iron-shielded, our ships
rode proud to harbour.
Of Miklagard, our famous prince
first saw the golden gables.
Many a sea-ship, fine arrayed,
swept toward the high-walled city.
Byzantine Empire, 1034–1041
At that time, the empire of the Greeks was ruled by Queen Zoe the Great with Michael Katalakos.’ Thus Snorri Sturluson begins his account of Harald in Byzantium, and with a sentence which supplies a key item of evidence for the date of his arrival in Constantinople while also introducing two personalities who are attributed particular significance for his subsequent career in imperial service.
The extraordinary Zoe – more properly, of course, styled Empress than ‘Queen’ – would seem to have figured no less prominently in Varangian tradition than in the formal history of Byzantium. As one of the three daughters of the dissolute emperor Constantine VIII, Zoe was porphyrogenita (‘born into the [imperial] purple’) and thus empress in her own right, as well as conferring the imperial title upon no fewer than three husbands and ruling jointly, albeit briefly, with her younger sister Theodora.4 In 1002, and while still an attractive young princess, Zoe had been promised in marriage to the Holy Roman Emperor Otto III, but he died before the ceremony could be solemnised and more than a quarter of a century was to pass before the imminent demise of her ailing father made it imperative that a husband be found for her so as to ensure the succession. Consequently, she was almost fifty at the time of her first marriage, in the year 1028, to the elderly Byzantine aristocrat Romanus Argyrus who relinquished his own wife to accept the union only under extreme duress and just in time to become the new emperor Romanus III following the death of his father-in-law on the very day after the wedding. Hardly surprisingly, and despite application of all available medical alchemy, the marriage bed proved barren and was soon deserted by Romanus for that of a mistress. Already humiliated, Zoe was further enraged by being deprived access to the imperial treasury and first vented at least some of her fury on her sister Theodora, whom she despatched to a nunnery in 1031 before she herself fell prey to the sinister ambitions of John the Orphanotrophus.
A eunuch who had risen from modest origins in Paphlagonia on the southern shore of the Black Sea to the prestigious offic
e of director of the capital’s principal orphanage, John presented his youngest brother Michael to the embittered empress, who found the handsome teenager so irresistibly attractive that she took him as her lover and the city buzzed with rumours of poisoning when her estranged husband soon afterwards succumbed to an inexplicable illness. Romanus had been found dead in his bath-house only hours before the Patriarch of Constantinople was summoned to the palace at dawn on Good Friday, 12 April in the year 1034, to join the newly widowed empress in marriage to a man almost forty years her junior and to consecrate him as the Emperor Michael IV.
By virtue of that precise record of the date of consecration of a new Byzantine emperor, Snorri’s statement that the empire was ‘ruled by Zoe the Great and Michael Katalakos’ at the time of Harald’s arrival in Constantinople firmly places it after the Easter of that year. Incidentally, Snorri’s use of the cognomen ‘Katalakos’ indicates Varangian tradition as his source of information, because Michael IV was formally known as ‘the Paphlagonian’ and Katalak represents the Norse form of his popular nickname Parapinakes, meaning ‘clipper of the coinage’ and deriving from his family’s trade of silversmith while also making allusion to his own alleged practice of devaluing the currency.
A more specific timing for Harald’s appearance in the Byzantine capital might be deduced from what is known of the seasonal organisation of traffic along the Dnieper route. The Rus trading fleets customarily set out from Kiev in June to allow their return from Constantinople before the rivers froze up again in the autumn and also because the water level reached the point best suited for negotiation of the rapids at that time of year. For that reason alone, Harald would quite certainly have chosen the month of June for his own departure. The duration of the voyage from Kiev to Constantinople has been estimated at some ten weeks and so – even allowing for his sea-ships having made a better time on the Black Sea crossing than the more cumbersome monoxyla – Harald would have been unlikely to have reached his destination before August.
In fact, that estimate of his time of arrival would correspond perfectly well to Snorri’s placing Harald’s first assignment as a Varangian mercenary in Byzantine service in ‘that same autumn’, but perhaps less convincing is the saga-maker’s claim for Harald’s presenting himself to Zoe immediately upon his arrival in Constantinople. While it is scarcely likely that a porphyrogenita empress would have made herself available for duty as a receptionist for mercenary recruits, it is perhaps just possible that she might have granted an audience to a warrior prince of distinguished Norse descent and held in high esteem by the Grand Prince of Kiev.
Such would seem to be the only plausible explanation of Snorri’s account – and yet the other saga sources suggest a quite different situation when they tell of Harald’s attempt to conceal his true identity in Constantinople on account of the Byzantine policy of discouraging high-born recruits in their mercenary forces. For that reason, he is said by Morkinskinna and Flateyjarbók to have entered imperial service under the pseudonym Nordbrikt, which aroused the suspicions of an Icelander in command of a detachment of the Varangian Guard to the point where he tried, albeit unsuccessfully, to discover Harald’s real name from Halldor Snorrason. That reference alone would point to Halldor as the original source of the story and might be thought to endorse its authenticity, were it not for a closely contemporary document – known as the Book of Advice to an Emperor and dated to the 1070s – which has no hesitation in confirming Harald’s true identity and distinguished kinship having been known to the Byzantines, while also confirming Michael IV as the reigning emperor who received Harald ‘with a proper courtesy’ upon his arrival in Constantinople.
This account of Araltes (calling him, of course, by his Greek name-form) is considered the most authoritative record of his Byzantine service, and not least because its anonymous author mentions having served alongside Harald in the Bulgarian campaign of 1041, yet it appears to make no reference to his military activities prior to the invasion of Sicily which was launched in 1038. What is known of the earlier years of his Byzantine career, during which he appears to have served as a Varangian mercenary before his promotion to the Varangian Guard proper, derives from the skalds and the saga-makers, whose evidence requires some measure of corroboration by the more formal historical record.
As before in Russia, Snorri is equally impatient in Byzantium to promote Harald to the highest possible level of command, even to that of the ‘acknowledged leader of all the Varangians’, a statement which closer examination places under some measure of doubt. Whether or not Harald really was welcomed to Constantinople by the emperor or empress, there appears to have been nothing at all unusual about his entry into their service as a Varangian mercenary or in command of the warriors he had brought with him from Russia, who would have formed a unit typical of foreign contingents in the Byzantine armed forces.
The use of foreign fighting-men, under various forms of arrangement, had always been the practice of the Byzantine military and can be recognised as a legacy from the old Roman Empire, of which the new imperium founded upon the ancient Greek city of Byzantium by Constantine the Great in the early fourth century considered itself to be the true successor. This eastern empire, which has long been called ‘Byzantine’ by historians, was known to the Rus and to the northmen as that of ‘the Greeks’ (and reasonably so when Greek had replaced Latin as its official language in the early seventh century), yet the Byzantines themselves, almost all of whom were of Greek and Slav ethnicity, still considered themselves ‘Romans’ (in the Greek form of Rhomaioi) in Harald’s time. Indeed, the Advice tells of ‘Araltes, the son of an emperor of Varangia . . . [having] determined to go and pay homage to the most blessed Emperor Michael the Paphlagonian, and to see for himself the ways of the Romans’.
Just as there had been few, if any, Italians in the ‘Roman Army’ of the western empire through most of its last four hundred years when the greater part of its military comprised fighting-men recruited from the free peoples east of the Rhine, foreign contingents were to make up more than half of the strength of Byzantine forces by the later eleventh century. Even before then, the ‘thematic’ structure of imperial military organisation, which had first appeared in Asia Minor during the seventh century, had already entered into serious decline. Since the ninth century, the eastern empire had been organised into themata (or ‘themes’) under military governors usually known as strategoi, each of whom raised his own army corps on the basis of land held in return for military service. Much like the later western European feudal system, this land-holding was passed from father to son and with it the obligation to provide a soldier, either a family member or a proxy, equipped with arms and armour, as well as his own mount when cavalry represented the principal component of Byzantine land forces.
By the eleventh century, these themes had become the increasingly independent provinces of a land-owning aristocracy, where smaller holdings were absorbed into larger estates supplying a diminishing flow of manpower to the thematic armies. In consequence, the strategoi became ever more dependent upon the hire of mercenaries and thus drew in more and more foreign fighting-men to maintain the strength of the forces of the empire: Normans and Italians, Germans and Magyars, Pechenegs and Khazars, Arabs and Slavs, and – most importantly here, of course – Scandinavians who had come from the northlands by way of Russia, thus being known first to the Byzantines as Tauro-Skuthai (‘Northern-Scythians’) and afterwards as Rhos (the Greek form of Rus) before they acquired their more celebrated name of Varangoi or ‘Varangians’. Such mercenaries are on record as early as the reign of Michael III in the 840s and further notices of Rhos serving with imperial forces punctuate the chronicles of the following century, especially after the Russo-Byzantine trade agreements of 907 and 911 which actually include provision for a military levy. A reference to two Rus ships with the imperial fleet sent to Italy in 968 represents an early, if not the earliest, item of evidence for the deployment of Varangians in Byzantine naval op
erations, and is of particular interest here because Snorri makes specific mention of Harald’s first assignment in imperial service having been with the ‘fleet in the Greek sea’ in the autumn of 1034.
Before investigating that particular passage of his warrior’s way in further detail, however, it should be emphasised that all these Varangians – by whatever name they might be called in the Greek sources – represented effectively auxiliary contingents within a larger imperial force and so, while each unit was probably led by one of its own kind, all would have served under the supreme command of a Byzantine general. Indeed, Snorri himself indicates as much when he tells of Harald having ‘kept his men together as a separate company’ before adding that ‘the commander-in-chief of the fleet was a man called Gyrgir’. The same Gyrgir is mentioned in Fagrskinna, where it is explained that this was the Norse name-form of Georgios Maniakes, the outstanding Byzantine general of his time and a figure to whom the sagas attribute a key significance in the course of Harald’s Varangian career.
It should be explained that the Varangian mercenaries mentioned thus far were distinct from the ‘Varangian Guard’ proper, known to the Byzantines as the ‘Varangians of the City’ and including the emperor’s personal bodyguard, which formed its own regiment of the imperial forces based at Constantinople. The longest-standing division of these forces was the Tagmata, which comprised four elite regiments of horse, each under its own commander who was usually styled domestikos. Not unlike the Household Cavalry of the British army in more recent centuries, the Tagmata’s first duty was to serve the emperor as his lifeguard, whether on campaign or in the capital itself where it also performed a range of ceremonial duties, while an associated infantry regiment known as the Numeri provided a garrison for defence of the city. Naturally enough, the professional soldiery making up these regiments was largely drawn from the Byzantine Greek aristocracy, yet it was just that characteristic of the Tagmata which brought it under suspicion, and not at all unreasonably in the intrigue-ridden climate of Constantinople where its regiments were openly associated with rival political factions. Consequently, the loyalty of the Tagmata was never considered entirely reliable during times of internal political crisis, and a source of more trustworthy lifeguards was offered by the emperor’s own foreign mercenary troops, whose loyalty depended upon nothing more complex than the generosity of their paymaster. Some divisions of these forces were assigned duties in the capital much like those of the Tagmata and thus formed a part of the emperor’s lifeguard which became known as the Hetairia, of which the best-known component was the Varangian Guard established as a regiment in its own right by Basil II, elder brother of Zoe’s father, Constantine VIII.