So too, the skaldic reference to Harald fighting beside Eilif – ‘in phalanx tight with Rognvald’s son’ being the literal translation – need mean nothing more than his having been in action with Eilif’s forces, even though his personal qualities and royal kinship would have very probably have afforded him a status beyond that of a rank-and-file warrior, even placing him in some capacity of command, if only over the new Norwegian Varangian recruits to whom he would already have been a familiar comrade-in–arms. While Rognvald Brusason was almost certainly admitted to Jaroslav’s druzhina, there cannot be said to be any real evidence for Harald’s serving in Russia in any other capacity than that of a Varangian mercenary, albeit one recognised for his remarkable qualities even by the Grand Prince Jaroslav himself. A couplet attributed to the skald Thjodolf, although preserved only in Flateyjarbók and not quoted by Snorri, would seem to allude to just this point in Harald’s career when it tells how ‘Jarisleif saw the way in which the king [Harald] was developing; the fame grew of the holy king’s [Olaf’s] brother’.
The relationship between Harald and Jaroslav, founded on genuine mutual respect and sustained over more than a decade, was certainly formed before Harald left for Byzantium. His military qualities might even have come to Jaroslav’s notice during the Polish campaign, but the skald’s reference to his growing fame would seem to suggest a later date for Jaroslav’s recognition of the true potential of this unusually ambitious seventeen-year-old Varangian. At some point between his service with Eilif and his departure for Byzantium, Harald had evidently gravitated southward from Novgorod to Kiev which served as the focal point of assembly for Russian trading fleets bound down the Dnieper route to the Black Sea and the beckoning marketplaces of Grikaland (literally the ‘land of the Greeks’), as the empire of the Byzantines was known to the northmen.
Kiev’s key location along the east-way to Byzantium was the reason for its displacement of Novgorod as the new capital centre of the Rus towards the end of the ninth century. The earlier prominence of Novgorod derived from its proximity to the northern source of furs and its access to the Volga route along which they could be traded for silver from the east. But the progress of the Rus in that direction was so constrained by the might of the Bulgars on the middle Volga and the power of the Khazars (a highly sophisticated people of Turkic origin) along its lower reaches that they were rarely able to venture further south than the great marketplace of Bulghar. So it was that their trade with the Arabs was largely conducted through powerful ‘middle men’ in an arrangement demanding great outlay of effort in return for suspiciously chiselled profits, whereas the Dnieper route to Constantinople promised direct dealing with the famously wealthy Greeks in a market ever eager for the most exotic and luxurious of merchandise.
The Rus were still northmen in character and culture throughout the ninth century, and so it was only to be expected that their first contact with Byzantium – launched from Novgorod in 860 – was as viking raiders, but the mighty walls of Constantinople (or Miklagarð as it was called in the Norse) and the fire-breathing warships of the imperial fleet had driven off the raid of 860 as they were to do again on occasions through the following two centuries. Trading, rather than raiding, was clearly going to be the safer and more profitable approach to the Greeks, as would be confirmed by the generous trade treaties (of which the texts, terms and names of signatories still survive) made by the Byzantines with the Rus in 907 and 911.
Hence the new importance of Kiev, standing high above the point where the northern riverways flow into the broad stream of the Dnieper and already established not only as a Slav settlement but apparently also as a tribute-collection point for officials of the Khazar khans. The Primary Chronicle tells of the Rus seizure of Kiev during the infancy of Rurik’s son Igor – the first reliably historical prince of the Rus and great-grandfather of Jaroslav – but only in the form of folk-tales which would be of little value were it not for the apparent authenticity of its claim that the Slavs, who themselves had only been on the Dnieper since the seventh century, were happy to welcome the Rus as their overlords instead of the displaced Khazars. Indeed, the Slavs were to have a key role in the annual supply of the vessels – called monoxyla and of typically Slav design as a single hollowed-out tree trunk, much like a giant dug-out canoe before it was built up and widened with planking – which formed the merchant fleets for the voyage from Kiev down the Dnieper to the Black Sea and Constantinople. It is unlikely, however, that Harald would have gone directly from Novgorod to Byzantium by way of Kiev in 1034 and much more probable that he had been earlier drawn south by the demand for mercenary forces in the region of the middle Dnieper.
There is evidence in the saga record for the custom of the Rus in Jaroslav’s time having been to hire their Varangians on a contractual basis, providing their maintenance over a twelve month period and, upon its completion, rewarding their services either in coin or in kind, usually in the form of furs which had been rendered as tribute or taken as plunder. If such had been the case with Harald and the other Norwegians recruited by Eilif in the summer of 1031, a twelve months’ contract would not only have covered their service on the Polish campaign and the winter round of pólútasvarf, but possibly also on the expedition of 1032 from Novgorod to the ‘Iron Gates’, by which is meant the domain of the Ob-Ugrian tribes in the far north-eastern region of the Pechora river under the Urals. The most remote of all the Finno-Ugrian peoples, these were considered so terrifying that Russian tradition believed them to have been locked behind iron (or copper) gates until the Day of Judgement and, indeed, a similar expedition to the Ob river beyond the Urals disappeared entirely without trace in 1079.
Whether or not Harald and his Varangians were engaged on so daunting an enterprise before their contract expired, they evidently survived the experience to be rewarded with their payment due from the proceeds of tribute collected through the winter. By the later summer of 1032, then, they would have been free to enter mercenary service elsewhere and the most promising opportunity is indicated by the entry under that year in the Primary Chronicle which notices that ‘Jaroslav began to found towns along the Ros’. Jaroslav’s intended policy of re-settlement of his Polish captives along the Ros river had already been mentioned in the chronicle entry under the previous year and this new entry confirms its implementation.
Something more needs to be said about this, however, because the foundation of townships along the Ros was just one component of a wider policy for defence of the middle Dnieper which had come under increasing pressure from incursions by Pecheneg raiders. These Pechenegs were just one of a long sequence of Turkic-speaking nomad warrior tribes, which had begun with the Huns and was to culminate in the Mongol invasion, who swept westward across the vast swathe of grassland known as the steppe which extended over some five thousand miles of Eurasia from Manchuria in the east to the Hungarian plain.
This steppe provided the natural domain of a people who were constantly on the move, settling only in tent encampments, and whose livelihood depended upon their livestock, cattle and sheep bred for meat, fleece and hide, and most importantly upon the hardy ponies which served as well for following flocks and herds across great distances as for lightning raids upon the more sedentary peoples whose lands were overrun by the nomad warriors. ‘We cannot fight them,’ replied the Magyars of Hungary to a Byzantine suggestion that they should rise against the Pechenegs, ‘because their land is vast, their people are numerous, and they are the devil’s brats!’ Such, then, was the new presence faced by the Rus in the last decades of the ninth century when, having only recently broken the power of the Khazars to gain control over the trade route to Byzantium, the Pechenegs irrupted on to the steppe north of the Black Sea in their drive towards the fertile plains of Hungary.
Ferocious raiders who fought with lance and spear, sabre and hand-axe, their most characteristic weapon was the composite bow formed of a light wood (or, better still, bamboo) core strengthened with horn and bound with
sinew, painstakingly bonded with glues and skilfully shaped into a curved weapon, much smaller than the northern self-bow (such as the traditional English longbow) and yet of equal power, more efficient and perfectly designed to fill every requirement of the mounted archer. Whether as friend, at least of a sort, or more often as foe, the influence of the Pecheneg steppe warrior had a dramatic impact on the character of the Rus, and not exclusively in terms of military practice, although it was probably first apparent in the urgent adoption of cavalry warfare by a people whose ancestors had always fought on foot.
While the first Scandinavian venturers into Russia, who would have crossed the Baltic after the spring thaw, found their ships perfectly suited to a system of rivers up to half a mile wide and linked by overland portages across which clinker-built craft were easily carried or rolled on logs by their crews, they would soon have discovered that the same routes, when hard-frozen through the winter, could serve equally well as thoroughfares between impenetrable forest for warriors travelling on horseback or by horse-drawn sled. Horse travel is one thing, of course, and cavalry warfare quite another, so the Rus on the Dnieper must have lost no time in adapting to the new presence of fighting-men who spent virtually their entire lives, whether at work or at war, in the saddle. The influence of the steppe warrior culture on the Rus becomes most vividly apparent in the second half of the tenth century and in the person of Jaroslav’s grandfather, the warlike Sviatoslav of Kiev who inflicted a number of defeats on the Pechenegs and yet is known – most graphically from a first-hand Byzantine account – to have adopted the style and appearance of a steppe khan. Ironically enough, when he had included a Pecheneg contingent (as allies or as mercenaries) in the host he led on his last campaign to the lower Danube, Sviatoslav was making his desolate homeward retreat, following his surrender to the Byzantines, when he was slain by Pechenegs as he passed up the same Dnieper rapids where Rus convoys heavy-laden with goods for Constantinople faced the choice of paying off a Pecheneg ambush or hiring a Varangian escort to fight it off.
By contrast to the devastating sweep of later Mongol hordes, the effectiveness of the Pecheneg incursions lay in repeated raiding until the social order of the afflicted territory collapsed under the unrelenting strain. Such was the character of their campaign against the middle Dnieper where it made agricultural settlement almost impossibly difficult, and to oppose it Sviatoslav’s son and eventual successor Vladimir devised a system of earthworks some 3 or 4 metres in height, their interiors reinforced by logs laid parallel to the rampart which was fronted by ditches up to three times as wide.
During the last twenty-five years of his reign, Vladimir erected some 300 miles of these defences, known as the Snake Ramparts, which were raised just too high for a steppe pony to clear at its full speed and so intended to slow down the nomad horsemen, denying them the advantage of a surprise attack and obstructing their line of retreat while Rus warriors came in pursuit. Fortified strongpoints were added to this defensive network, similarly constructed as earthwork with timber reinforcement and built at points along the line of ramparts, some of which were large enough to accommodate a cavalry squadron, which itself attests the new accomplishment of the Rus in mounted warfare. Within this defended region of the middle Dnieper, Vladimir established a number of fortified towns as well as unfortified settlements, all of them peopled with thousands of settlers brought in from subject and conquered tribes such as the Slovenes and the Chuds.
Impressively reassuring as they must have been to the Rus and their settlers, the Snake Ramparts would seem to have offered as much a provocation as a deterrent to the Pechenegs, who considered the Dnieper valley their own summer grazings, so their incursions continued and with a renewed vigour during the wars of succession which followed Vladimir’s death. In consequence, and despite his preference for Novgorod and the north, Jaroslav’s attention was drawn down to his southern frontier along the Dnieper where he followed his father’s example in extending the Snake Ramparts and establishing new townships, such as those along the Ros where he settled prisoners from his Polish conquest. Relocation of prisoners of war, and presumably in some numbers, was a military operation requiring larger forces than the druzhina, who would probably have considered it a duty beneath their dignity anyway. Neither was it a short-term operation, because the newly settled communities would probably need some measure of supervision and no less a measure of armed protection should Pechenegs make an appearance. The solution, as always for Jaroslav, would have been to recruit Varangians and so Harald and his company, not only veterans of the Polish campaign but also experienced in dealing with subject peoples, would have been the ideal choice. To which can be added just one key fragment of evidence and it is supplied by Adam of Bremen, probably Harald’s most hostile historian but who can still offer occasional items of information preserved in no other source, such as his reference to Harald having ‘fought many battles with the Saracens by sea [in Byzantine service] and the Scythians by land’.
The Scythians were one of the very earliest steppe warrior peoples, although of Iranian rather than Turkic origin, and first recorded north of the Black Sea in the seventh century BC, flourishing thereafter until they were displaced by the Sarmatians three hundred years later. Although there had been no Scythians around for fourteen centuries by the time Adam was writing, the name was still retained in literary currency as a generic term for ‘barbarians’. Byzantine writings, for example, are known to refer to Varangians as ‘Tauro-Scythians’, meaning ‘northern barbarians’, but in its eleventh-century usage the term ‘Scythian’ almost invariably meant Pechenegs. Even though it is not entirely beyond possibility that Harald could have encountered Pechenegs during the earlier years of his Byzantine service, he was very much more likely to have done so in Russia on the Dnieper and there, while he would certainly have run the risk of ambush on the way to Byzantium, his greatest likelihood of meeting them in ‘many battles’ was while engaged on the Ros and the Snake Ramparts.
For all Jaroslav’s efforts, the Pecheneg menace remained undiminished through the mid-1030s, coming to its point of crisis in 1036 when the death of Mstislav left something of a hiatus on the middle Dnieper. Jaroslav was in the north and more immediately engaged with the installation of his eldest surviving son, Vladimir, as prince of Novgorod, when a Pecheneg host seized the opportunity to besiege Kiev. In response, ‘Jaroslav gathered a large army of Varangians and Slovenes’ (according to the Primary Chronicle) and came south to lift the siege. In a ferocious battle fought into the evening on the fields outside the city he inflicted a crushing defeat which effectively marked the end of the Pecheneg ascendancy because within twenty years they had been driven from the Russian steppe by the next wave of Turkic warrior nomads, a people known to the Rus as Polovtsy and to the Byzantines as Cumans, although they called themselves the Kipchaks.
This triumph over the Pechenegs would have served as a fitting climax to the ten ‘arrow-storms’ in which Rognvald Brusason is said to have fought so valiantly for Jaroslav, had not the Orkneyman returned with the young prince Magnus who had been invited back to Norway as his father’s successor the year before. Neither could the saga-makers include the victory at Kiev among Harald Sigurdsson’s Russian battle-honours because by 1036 he had already been some two years in Byzantine service.
It was love for a woman which prompted Harald’s departure for Constantinople, at least according to the ‘Separate’ version of his saga in Flateyjarbók which tells how he was refused the hand in marriage of Jaroslav’s daughter Elizaveta until he had won greater wealth and glory. More recent historical opinion casts doubt upon the dating of the Flateyjarbók story, because Elizaveta – who is known in the sagas by her Norse name-form of Ellisif and called ‘the bracelet-goddess in Gardar’ in Harald’s own poetry – was scarcely ten years old in 1034. Nonetheless, the couple were eventually to be married, although not until Harald had returned from Constantinople a full decade later, so there may indeed have been some sort of prio
r arrangement because Jaroslav evidently made a habit of marrying his daughters to foreign magnates and Harald was, by that time, preparing his return to claim kingship in Norway.
Whatever promises might have been made to him in 1034, the more immediate motive behind Harald’s further venture along the east-way was assuredly his great desire (a greed freely admitted in his sagas) for fame and riches which, as he would have heard from others returned from Varangian service in the east, were generously available to a warrior such as he in the land of the Greeks. The promise of such wealth would likewise have been the lure for his comrades-in-arms, because Snorri’s Harald’s saga tells of his arrival in Constantinople ‘with a large following’ of fighting-men, which would have comprised his Varangian troop in Russia, probably still including those of Olaf’s housecarls who had come to Russia with Rognvald and been recruited alongside Harald into Eilif’s forces.
For these men, Harald Sigurdsson would surely have also offered a natural leader and not only by reason of his kinship to Olaf and descent from Harald Fair-hair, because he would by now have been a truly formidable fighting-man in his own right. Harald is well known to have been exceptionally tall (even allowing for exaggeration in the saga estimate of his height at ‘five ells’ or seven feet six inches). Although he was scarcely nineteen years of age when he left Russia, it is salutary to remember that Cnut had been much the same age when he fought his way to kingship of England in 1015 and, in strictly military terms, Harald was the more widely experienced of the two. He had been just fifteen when he fought at Stiklestad and there is nowhere any indication of his being discouraged by what he had seen there of the realities of warfare, although there can be little doubt that the death of Olaf – which Harald himself may well have witnessed at first-hand – assuredly cut a long, deep scar in his psyche. Both mentally and physically then, he was ideally equipped by nature for the profession of arms and, while his first experience of battle would have corresponded to the expectations instilled by the heroic culture in which he had been raised, a more expansive education – and not only in the way of the warrior – awaited him in Russia.