Another son was killed while attempting to flee into Hungary, but the approach of autumn presented Sviatopolk with his most formidable rival in the form of Jaroslav advancing south towards the Dnieper with his Novgorod militia and a thousand Varangian mercenaries. Meanwhile, Sviatopolk had recruited his own mercenary auxiliaries from the Pechenegs, a Turkic people who represented a fearsome military presence, alternately as mercenaries for hire or as predatory raiders, in the steppe country extending into southern Russia from the north shores of the Black Sea. When Jaroslav’s forces reached Liubech on the west bank of the Dnieper some 90 miles north of Kiev, Sviatopolk’s host was ranged along the eastern bank. For three months the opposing armies faced each other across the river, until Jaroslav contrived to isolate Sviatopolk’s druzhina from their Pechenegs and trap them on a thinly frozen lake where the breaking ice ensured their destruction.
Jaroslav entered Kiev in triumph and Sviatopolk fled west into Poland to find refuge with his wife’s father, the Polish king Boleslaw. He returned to Russia in 1018, bringing with him a Polish army led by his father-in-law and reinforced with German, Hungarian and Pecheneg mercenary contingents which met and defeated Jaroslav’s force of Varangians and militias of Kiev and Novgorod in the contested border zone along the western Bug river. Now it was Sviatopolk who advanced on Kiev while Jaroslav took flight north to Novgorod. But the real victor was almost certainly Boleslaw who paid off his mercenary forces with the plundering of Kiev before making his own departure back to Poland, and taking possession of the Cherven border towns en route.
Meanwhile, Jaroslav was back in Novgorod and levying new taxation to pay for more Varangians to renew his offensive. Left vulnerable without his Polish allies, Sviatopolk was likewise engaged in the south and still trying to recruit fresh forces of Pecheneg mercenaries even while Jaroslav was advancing again on Kiev in the spring of 1019. The last battle between the two was fought beside the River Alta not far from Kiev where Jaroslav won a great victory and Sviatopolk again took flight, although he is said to have fallen ill and died before reaching Poland. Having regained the ascendancy, Jaroslav was still not yet secure because another of the numerous sons of Vladimir was about to mount his own challenge.
This was Mstislav whose power base lay at Tmutorokan where he had carved out his own impressive dominion in the south-east around the Sea of Azov. Perhaps tempted towards the Dnieper by Jaroslav’s evident preference for Novgorod, Mstislav moved north of the steppe sometime around 1024 and based himself at Chernigov within striking distance of Kiev. Once again, according to the Primary Chronicle, Jaroslav ‘sent across the sea for Varangians’ to face Mstislav’s ‘Kasogians and Khazars’ recruited from his subject peoples around the Sea of Azov and supplemented with Severians (Slav auxiliaries, presumably from Chernigov). These Severians would seem to have played the greater part in defeating Jaroslav’s northmen in an extraordinary battle fought at night in a thunderstorm at Listven, and Mstislav’s exclamation in his hour of victory is a remark of telling military significance: ‘Who does not rejoice at this? Here lies slain a Severian and here a Varangian, and yet the druzhina is unharmed!’
After the defeat at Listven, Jaroslav withdrew back to the north and yet, for whatever reason, Mstislav did not pursue his flight from the battlefield, but instead conceded Kiev (whose people had earlier refused him entry) to his brother while retaining Chernigov as his own power base. Relations remained cautious and Jaroslav continued to build up his forces at Novgorod until 1026 when the two met again near Kiev and formally agreed to a division of the lands of the Rus along the line of the Dnieper, those to the east for Mstislav and those to the west for Jaroslav. ‘Thus they began to live in peace and brotherhood’, according to the Primary Chronicle. ‘Conflict and tumult ceased and there was tranquillity in the land.’ Under that arrangement each brother would have been responsible for defence of his own dominions, although the two did join forces in common cause on at least one occasion, until Mstislav died of sickness sometime around the year 1036 and with no surviving son, leaving Jaroslav as samovlastets (‘sole ruler’) – and ‘Grand Prince’ in the full sense, although the Turkic style of khagan (or ‘khan’) was still in use until the twelfth century – over all the lands of the Rus.
All of which is intended to provide some outline of the Russian political and military context into which Harald Sigurdsson made entry – although assuredly not as commander of defence forces – on his arrival from Scandinavia in the summer of 1031. Dispensing with the account offered at this point in Snorri’s Harald’s saga, the greater detail found in Orkneyinga saga will provide the more useful starting-point for any attempt at a realistic reconstruction of this particular passage of his warrior’s way.
By way of example, one such detail of which Snorri makes no mention is the reference in Orkneyinga saga to Rognvald Brusason and his fellow survivors of Stiklestad making their way to the court of King Onund (presumably at Sigtuna south of Uppsala) on arrival in Sweden and thus it would be reasonable to assume that Harald also made his way to Onund’s court, probably by prior arrangement to join them there, when he followed them into Sweden. Having crossed the Baltic to Russia, probably aboard ships provided and provisioned by Onund, Orkneyinga saga describes their going directly to Novgorod where ‘King Jaroslav [or Jarisleif in the Norse name-form used in the sagas] welcomed them kindly out of respect for the holy king Olaf’.
It is at this point that the saga introduces the ‘Varangian dimension’ of Jaroslav’s welcome when it specifies that ‘all of the Norwegians’ joined Jarl Eilif to ‘take over the defences of Gardariki’. The interest of Orkneyinga saga in these events centres upon Rognvald Brusason, who was later to succeed his father as jarl of Orkney, and an account of his having remained in Russia while Harald went on to Byzantium. Lines from the skald Arnor Thordsson (usually called Arnor jarlaskald by reason of his service as court-poet to the Orkney jarls) are quoted in support of the saga’s claim for Rognvald’s fighting ten battles in Russia where he was held in the highest regard by Jaroslav and in whose service he ‘defended the country in the summers, but stayed in Novgorod over the winter’.
With the benefit of just those few more precise details, it already becomes possible to elaborate upon Snorri’s account of Harald’s entry into his career as a Varangian. There is, first of all, a useful geographical indication in the saga reference to Rognvald’s company going directly to Novgorod because it points to their ships’ having passed through the Gulf of Finland and continued along the waterway to Lake Ladoga where Staraja Ladoga (or Aldeigjuborg in the Norse) represented the most usual port of entry for Scandinavian arrivals in Russia.
Staraja Ladoga had long been a centre of great importance on Russia’s Baltic coast, and not least by reason of its proximity to the Finno-Ugrian peoples of the northern forests from whom the Rus obtained the furs they needed for trading along the east-way, initially by way of the Volga to Bulghar, and later down the Dnieper to the lucrative market of Byzantium. Archaeological evidence has identified Staraja Ladoga as the site of the earliest settled Scandinavian presence in Russia, so it must have been one of the very first fortified trading-posts known as goroda (from which term, of course, is derived the Norse name of Garðar) and, assuredly also by the eleventh century, one of the wealthiest. Indeed, lordship of Ladoga had been the bride-price asked of Jaroslav by Ingigerd when she arrived in Russia for their wedding and, once granted, she immediately conferred it upon Rognvald Ulfsson who had been her escort on the journey from Sweden. Formerly jarl of Gautland, Rognvald thus became ‘jarl’ (according to the saga, but perhaps more properly styled boyar in this Russian context) of Staraja Ladoga. Rognvald Ulfsson had died at some point during the decade before Harald reached Russia, however, and been succeeded in the lordship by his son Eilif who is already styled jarl in Orkneyinga saga, so when Rognvald Brusason’s company sailed on from Ladoga to follow the River Volkhov down to Novgorod (called by its Norse name of Holmgarð in the sagas), Eilif Rognva
ldsson probably travelled with them to provide an escort for distinguished visitors to Jaroslav’s court.
It is at this point that a cautious reading between the lines of Orkneyinga saga would indicate Harald and Rognvald Brusason having been differently assigned in Russian service. Rognvald, first of all, would already have been known to Jaroslav as a prominent member of King Olaf’s retinue in Novgorod just two years earlier and the clear inference from the saga is of his having been recruited to Jaroslav’s own druzhina, and not only ‘out of respect for the holy king’ when he himself was held in high regard on his own account and was to offer such sterling service through the ten ‘teeming arrow-storms’ recalled by Arnor’s verses. Interestingly, Arnor’s use of the term ‘arrow-storm’ might be read in this instance as a more specific reference than a stock skaldic kenning, because Jaroslav’s pressing military concern in the mid-1030s was the defence of Kiev and the middle Dnieper against the Pechenegs of the steppe whose characteristic and most feared warrior-type was the mounted archer.
To which should be added a reminder that Rognvald was an Orkneyman, which would have neatly excepted him (but not Harald) from ‘all the Norwegians’ who were assigned to Eilif ‘to take over the defences of Gardariki’. In fact, there is good reason to believe that Eilif would have been in need of Varangian reinforcements in the summer of 1031 because an entry in the Primary Chronicle under the previous year records that ‘Jaroslav attacked the Chuds and conquered them’. The Chuds, whose territory lay in what is now Estonia, were one of the numerous Finno-Ugrian tribes of importance to the Rus as the source, either by way of trade or tribute, of the greatly prized furs, so Jaroslav’s ‘conquest’ of the Chuds would have been a campaign to impose or to enforce the rendering of just such tribute.
Presumably by reason of their strange tongue and shamanic culture, these Finno-Ugrian denizens of the northern forests and sub-Arctic tundra were believed by the northmen to be possessed of sinister occult powers, as was evidenced by the weapon-proof reindeer coats Thore Hund acquired from the Lapps to armour himself and his housecarls at Stiklestad. Similar beliefs were apparently shared by the Rus and yet, while these peoples must have first appeared as shadowy hunter-gatherers magically materialising out of the dark forests, some of their kind nonetheless represented a formidable military presence and one which reflected the influence of warrior cultures from the steppes. Indeed, some Finno-Ugrian tribes boasted their own warrior elites, not least among these the Chuds who are known to have been recruited as mercenary auxiliaries by the Rus in Vladimir’s time.
There is every likelihood, then, that Jaroslav’s campaign of 1030 – whether of initial conquest or in retribution for refusal of tribute already imposed – would have been hard fought against fierce resistance. If only by reason of the proximity of Staraja Ladoga, Eilif’s forces would have formed a part of Jaroslav’s host, but the further implication of the saga evidence is of Eilif Rognvaldsson as Jaroslav’s voevoda in the north, effectively commander-in-chief of all the forces of the principalities of Novgorod and Staraja Ladoga and thus most prominently involved in the campaign against the Chuds. In which case, his Varangians, who invariably bore the heat and burden of such warfaring, may have sustained heavy losses and so Eilif would have welcomed a newly arrived phalanx of battle-hardened Norwegian housecarls to replenish his mercenary forces – and all the more so when Jaroslav would already have been mustering a great army for another war of conquest.
This campaign, entered in the Primary Chronicle under the year 1031, is of key importance here because it represents the one Russian military operation for which there appears to be quite specific evidence of Harald’s personal involvement. This evidence is found in lines attributed to Harald’s skald Thjodolf and quoted by Snorri which tell of Harald’s fighting beside ‘Rognvald’s son’ as they drove hard against the Læsir ‘to whom harsh terms were given’. Læsir is the Norse form of Liasi (or Lyakh in the Slavonic) by which is meant the Poles, and so Thjodolf’s reference is generally accepted as closely contemporary evidence for the involvement of Harald and Eilif in the invasion of Poland by Jaroslav and Mstislav with a large army which ‘ravaged the Polish countryside’, according to the Primary Chronicle, and successfully recaptured the Cherven towns seized by Boleslaw in 1018.
Thjodolf’s strophe also makes a more puzzling reference to ‘both chieftains’ (meaning Harald and Eilif) fighting against an enemy it calls the Austr-Vinðum, a people unknown (at least by that name) to any other source and whose identity poses a problem for translators of the saga. The most convincing attempt at translation of Austr-Vinðum is probably as ‘East Wends’, because the Wends (or Winida in Old High German, which would reasonably correspond to the Norse Vinðum) were a Slavic people settled in north Germany and familiar to the skalds and saga-makers as an enemy of more than one Scandinavian king of the time, so it is not entirely implausible to imagine that an easterly extension of their eleventh-century settlement could have fallen victim to the ravaging Rus in 1031.
What can be said of the Russian invasion is that it was surely prompted by the anarchy (also noticed in some detail by the Primary Chronicle) which had convulsed Poland after the death of Boleslaw in the previous year and so offered Jaroslav an ideal opportunity to reclaim the towns which had been a focal point of Russo-Polish contention since their seizure by Vladimir during his westward expansion of the 980s. Indeed, Jaroslav was to take further advantage of Poland’s disordered state when he relocated some numbers of Polish prisoners to his own new settlements along the Ros river which formed an extension of Russia’s lines of defence against the mounting tide of Pecheneg hostilities – and there is reason to believe that it was in this more southerly theatre of operations that Harald was to spend the later period of his Russian service.
While the date of Harald’s arrival in Russia can be securely placed in the summer of 1031 and the date of 1034 is generally accepted by historians for his arrival in Constantinople, the saga record supplies scarcely any indication, and still less detail, of his activities during the intervening three years. Indeed, were it not for the reference to Læsir made by the skald Thjodolf history would have no evidence for his involvement in the Polish campaign of 1031. The other skaldic strophe quoted by Snorri provides even less helpful detail in its fairly formulaic paean of praise for Harald’s military accomplishment in Russia and, indeed, would seem to add to the uncertainty with its couplet bearing on the duration of Harald’s stay in Russia: ‘You were the next, and the next after year, O warlike one, in Gardar.’
In fact, there is a question mark over authorship of this strophe when Snorri ascribes it to Bolverk Arnorsson and the same lines quoted in Fagrskinna are attributed to Valgard of Voll. Less is known of Valgard than of Bolverk, who is thought to have been a brother of the more famous Thjodolf, but both are reliably included in the list of Harald’s court-poets and so the historical authority of the lines in no way depends on the precise identity of their author. Their implication for the duration of Harald’s stay in Russia would seem to pose a problem, though, because, as Sigfús Blöndal observed in his study of the Varangians, the reference could be taken to mean that Harald spent no more than the two years after Stiklestad in Russia, although neither Snorri nor the Fagrskinna author reads it that way. Indeed, Snorri claims that Harald spent ‘several years in Gardariki and made expeditions east of the Baltic’, a statement convincingly interpreted by Blöndal as evidence for his being ‘employed on the arduous pólútasvarf’.3 By this term pólútasvarf (which will occur again at a crucial point in Harald’s Varangian service) is meant the winter round of tribute-gathering from subject peoples conducted by the druzhina, which would have been accompanied by Varangian mercenaries. These expeditions, conducted on horseback and by sled along frozen rivers in the depth of the northern winter, would have been an arduous duty indeed and probably a dangerous one too, when the least welcome could be expected from hosts confronted with the demands of heavily armed representatives of a distant
overlord.
As to the dating problem implied by ‘the next, and the next after year . . . in Gardar’, Blöndal’s suggestion that the reference may have applied only to the period of Harald’s stay with Eilif in Novgorod would seem to provide the most plausible explanation – and one with further bearing on Harald’s Russian service when it makes all the more apparent the liberties taken by Snorri Sturluson with the source material he had drawn from the Orkney Jarls’ saga. Assuming, of course, that the Jarls’ saga account has been accurately preserved in Orkneyinga saga (and there is no reason to think otherwise), then Snorri stands accused of gross exaggeration in his claim for Harald having shared command of Russian forces with Eilif Rognvaldsson. What Orkneyinga saga actually says is that all the Norwegians who had come to Novgorod with Rognvald Brusason joined Eilif’s forces, and presumably in the capacity of Varangian mercenaries. Thus their taking over the ‘defences of Gardariki’ only makes sense if ‘Gardariki’ is understood to mean Jaroslav’s northern dominions centred upon Novgorod, which would also correspond to Eilif’s sphere of command as Jaroslav’s voevoda but still falls a long way short of responsibility for defence of all the lands of the Rus.