16 The Varangian legend: testimony from the Old Norse sources Sverrir Jakobsson • I n the eleventh century there existed, within the great army of the Byzantine empire, a regiment composed mainly of soldiers from Scandinavia and the Nordic countries. This regiment was known as the Varangian Guard (tagma tōn Varangōn). The purpose of this paper is to assess the impact the existence of this regiment had on prevailing attitudes towards the Byzantine empire within the Old Norse linguistic and cultural community. The Varangian Guard is well known from Byzantine sources of the period. John Skylitzes’ chronicle Synopsis historiarum contains one of the earliest references to the term ‘Varangian’, connected with the events of the year 1034.1 From then on, Va- rangians appear in various sources.2 According to Michael Psellos’ Chronographia, the founding of the Varangian Guard took place during the reign of Basil II (976–1025), although Psellos calls these soldiers “Tauroscythians” rather than Varangians.3 This has often been connected with the evidence of Arabic and Armenian sources, according to which the nucleus of this regiment was formed by 6,000 mercenaries despatched by Prince Vladimir of Kiev in 989 to help the emperor Basil II quash a rebellion.4 From then on, Scandinavians formed the bulk of the guard, until expatriate Anglo-Sax- ons began to join in large numbers as a result of the Norman invasion of England in 1066. From the 1070s onwards, the Varangian Guard became predominantly English.5 Among notable Varangians serving the empire during the initial stage, when the force 1 John Skylitzes, Synopsis historiarum 394–95. The chronicle itself was written several decades lat- er. See also above, 53–87. 2 See Morrisson 1981, 131–34; Bibikov 1996, 203. 3 Psell. vol. 1, 9. 4 See Obolensky 1971, 255–56; Poppe 1976. The main source is the Arab Christian writer, Yahya of Antioch (Histoire vol. 2, 423–26). The number 6,000 is from Stephen of Taron, Histoire Univer- selle vol. 2, 164–65. However, Stephen uses the same number on other occasions to denote a large army: ibid., 156. See also Seibt 1992, 297–98. 5 See Ciggaar 1974. 345 346 Sverrir Jakobsson was predominantly Scandinavian (i.e. from 989 to the 1070s), was a certain Araltes, “son of the king of the Varangians [basileōs men Varangias ēn uios]”, who is mentioned in the Strategikon of Kekaumenos.6 This Araltes has commonly been identified with King Harald Hardrada of Norway (1046–1066). From sources such as these, it is pos- sible to gain some insight into contemporary Byzantine attitudes about the Norsemen, i.e. the view from the centre to the periphery. The view from the other side is more murky. Almost all our reliable knowledge about the Varangians stems from contemporary Greek sources. There is a distinct lack of Latin or Old Norse sources with the same validity. And our Slavonic sources, which have mostly been the focus of research into the history of the Varangians before 989, pose their own problems of interpretation.7 Yet there is no dearth of material relating to the Varangians in Old Norse sources from a later period. In this paper I shall focus on Old Norse sources from the thirteenth and fourteenth centuries, and on how they should be interpreted as representations of the contemporary Byzantine empire. These sources can be divided into two groups. The first consists of the Kings’ Sagas (Konun- gasögur), narratives dealing with the history of Scandinavian kings, in which there are sections about their relations with the Byzantine empire during the period between the First and the Fourth Crusades (1096–1204).8 The second consists of the Kings’ Sagas dealing with an earlier period (the tenth and eleventh centuries), as well as Sagas of the Icelanders (Íslendingasögur). The second group of narratives are set in the heyday of the Varangian Guard; but their problem as sources is that they were composed no sooner, and very often much later, than the Kings’ Sagas of the first group. The purpose of this paper is to delve into the Old Norse narratives containing information about the relationship of the Nordic peoples with the Byzantine empire during these two periods, and to extract from them such facts as are of use for the exploration of the image of the Byzantine empire in the north. The problems under discussion here are connected with periodisation, medieval ideas of sovereignty and the relationship between periphery and centre in an age before the advent of world systems. Crusader kings in Constantinople For a brief period in the early twentieth century, the Sagas of the Icelanders seemed to offer an exciting alternative view of the history of the Byzantine empire from the viewpoint of the Varangians themselves. The last manifestation of this optimism was Væringja saga by the Icelandic scholar Sigfús Blöndal, published posthumously in 1954. 6 Str, ed. and Russian tr. Litavrin, 298–99; ed. and tr. Roueché, 97.06. 7 See the overviews by Stender-Petersen 1953, 5–20 and Rahbeck Schmidt 1970. 8 On the Kings’ Saga genre see Ármann Jakobsson 2005.  The Varangian legend 347 However, by that time, serious doubts had been raised within saga studies in general about the value of these particular narratives, which mainly deal with events in Iceland from c. 930–1030, and in which events occurring abroad are mostly extraneous to the main plot. In his heavily edited English translation of Væring ja saga, The Varangians of Byzantium, Benedikt Benedikz offered Blöndal’s scholarship in a thoroughly revised form, with much more scepticism about the factual accuracy of the accounts.9 In Væring ja saga, Blöndal described the relationship between the Scandinavians and the Byzantines chronologically according to the occurrence of the events recount- ed, rather than following the age of the sources. This view of the sources has been ech- oed by subsequent scholars, often due to their unfamiliarity with the particular prob- lems relating to Old Norse sources.10 This view gives central importance to events and other historical titbits, with the sources of information becoming secondary to the dis- cussion. In order to shift our understanding of this relationship and the development of the Old Norse discourse about the Byzantine empire, the emperor and the imperial city of Constantinople, this order will be reversed, beginning with the oldest sources which deal with more recent periods. The earliest alphabetical texts in the Old Norse literary language are from the twelfth century. Before then, runic inscriptions were the dominant literary medium in Scandinavia, and Scandinavians even left traces of their presence in runic inscriptions in Constantinople and Athens.11 On the runestones of Norway, Sweden and Gotland, the names of ‘Greekland’ [Grikkland] and the Greeks appear more often than those of any other land or people.12 Similarly, the terms Grikkland and Grikkir/Girkkir occur in skaldic poetry, which is generally thought to have originated in the eleventh century, for instance at the court of King Harald Hardrada.13 The terminology is interesting in itself, as ‘Graecia’ and ‘Graecus’ were Latin terms which the Byzantines did not use themselves, and could even in some contexts be seen as pejorative.14 These terms were, however, generally used in Old Norse sources from that time onwards. However, neither skaldic poetry nor runic inscriptions contain longer narratives. These were introduced with the advent of an Old Norse adaptation of the Latin script on parchment. The earliest literary recordings of dealings between Scandinavians and Byzantines are thus necessarily no older than the twelfth century. Yet already at that time there are references to contemporary Byzantine events. The death of Alexios I 9 See VB. 10 See for instance the otherwise very useful overviews by Bibikov 1996 and Ciggaar 1996. 11 See Svärdström 1970; Larsson 1989; above, 187–214. 12 See Shepard 1984–1985, 230; Jansson 1984, 45–51. 13 See Jesch 2001, 99–102. 14 See Kaldellis 2007, 186, 296. 348 Sverrir Jakobsson Komnenos in 1118 is recorded in the Íslendingabók of Ari Thorgilsson, which was com- posed sometime between 1122 and 1133.15 In the twelfth century universal history Ver- aldar saga there are records of Byzantine emperors up to the Carolingian period; from then onwards, the western emperors are listed instead.16 The expeditions of Scandinavian kings to Constantinople do not receive much prominence in our twelfth-century narratives. The oldest surviving account of the cru- sade of King Sigurd the Jerusalem-Farer, the Historia de antiquitate regum Norwag- iensium by Theodoricus Monachus (composed c. 1180), does not mention his sojourn in Constantinople at all. The oldest surviving account in Old Norse, Ágrip af Nóreg- skonunga sögum (c. 1190) is relatively succinct: He went to Miklagarðr and received great honour from the reception of the emperor and great presents, and he left his ships there to commemorate his stay, and he took a great and valuable figurehead from one of his ships and placed it at the Church of St Peter.17 Nevertheless, the emphasis on the honour and gifts received from the emperor ini- tiates a theme echoed in later accounts of this very crusade, the bulkier Kings’ Sagas composed in the first half of the thirteenth century, for example Morkinskinna and Heimskringla.18 It seems that honour and gifts were the emperor’s to bestow and the Norwegian king’s to receive, which prompts some reflection about their relationship. King Harald had received honorary titles from successive emperors in the 1030s and 1040s and, according to the Kings’ Sagas, some decades later his great-grandson re- ceived dignity and presents from Alexios Komnenos. A similar description is given of the journey of the Danish king Eric (d. 1103) to the Holy Land some years earlier in the thirteenth-century Knýtlinga saga.19 This version draws upon the poem Eiríksdrápa, composed by the contemporary Icelandic lawspeak- er Markus Skeggjason (d. 1107), in which the various dignities bestowed upon Eric by foreign kings are enumerated at some length. Although Eric seems to have benefitted from being associated with the monarchs of France and Germany, the greatest digni- ty which he received was—according to the poem—from “the lord himself ” (harra sjölfum) in Miklagarðr. Gold, clothes and fourteen warships are listed among the gifts granted to Eric.20 In the narrative in Knýtlinga saga, the dignities that Eric and Sigurd 15 Landnámabók 25. 16 Veraldar saga 69–70. 17 Ágrip af Nóregskonunga sögum 48–49. 18 See Mork. vol. 2, 71–100 and Heimskringla 239–54. On the Morkinskinna version see Ármann Jakobsson 2013. The version of Heimskringla is discussed briefly by Fledelius 1996, 215. 19 Danakonunga sögur 232–39. 20 Norsk-islandske skjaldedigtning 419.  The Varangian legend 349 received from the emperor become the subject of direct comparison, the gold offered to Eric being contrasted with the games the emperor held for the Norwegian king in the Hippodrome in Constantinople, “and opinion is divided upon which choice was considered more noble”.21 The same attitude toward the emperor and his court is noticeable in Orkneyinga saga, an early thirteenth-century account of the pilgrimage of Earl Ragnvald in 1153– 1155. A man named Eindridi the Young (ungi), who had served for a long time as a mercenary in Constantinople, encourages the earl to travel to the Holy Land and not to be content simply to listen to stories from there. He argues that the earl will be “most respected, when you arrive there into the company of noblemen”.22 The pursuit of hon- our is thus made into the principal purpose of this pilgrimage. Following an adventur- ous journey, Ragnvald arrives in Constantinople to acclaim from “the emperor and the Varangians”. The emperor Manuel I Komnenos (1143–1180) gives Ragnvald and his companions “a great amount of money and offered them mercenaries’ payment, if they wished to remain there. They stayed a long time through the winter in altogether splendid revelry”.23 However, Ragnvald and his men choose to return, and in the end it is noted that they were considered to be much worthier men after this pilgrimage than they had been before. The account of the reception of the earl by the emperor is much briefer in Orkneyinga saga than in Morkinskinna, Heimskringla and Knýtlinga saga, and the gifts given by the emperor are referred to in the context of mercenary pay. Mention is also made of the Varangian Guard, perhaps as a suitable regiment for the service of any Scandinavian mercenaries. When seen in this light, the chrysobull sent by Alexios III Angelos (1195–1203) to the monarchs of Norway, Denmark and Sweden in 1196 seeking their military assis- tance is not very surprising.24 The Byzantine emperor could confidently expect these kings to be well disposed towards the empire, and at the very least putative allies in wars against its enemies.25 Is it possible to read more into this chrysobull and regard the relationship between the emperor and the Scandinavian kings as that between a liege lord and his vassals? It is certainly the case that this was a unilateral relationship, for the emphasis in the Old Norse is on dignities and gifts granted by the emperor, never on an exchange of gifts, as between monarchs of more equal stature. However, there is no 21 Danakonunga sögur 237. See also the version in Saxo Grammaticus, Gesta Danorum vol. 2, 74–83. 22 Orkneyinga saga 194. 23 Orkneyinga saga 236. 24 Sverris saga 192–94. 25 In the Historia de profectione Danorum in Hierosolymam, which describes the participation of a group of Danes and Norwegians in the Third Crusade, the emperor Isaac II Angelos is said to have hired some of these crusaders as mercenaries on their way home. See SMHD vol. 2, 490. This event occurred a few years before 1196. 350 Sverrir Jakobsson conclusive proof that the Byzantine emperor did regard the Nordic kings as anything more than junior partners within a larger community of sovereigns, in which the em- peror of Rome was bound to be pre-eminent. It is also evident that the Scandinavian kings accepted the emperor’s pre-eminence and considered it an honour to visit him and pay their respects. Some of the sources for the travels of these Scandinavian crusad- ers give a clear indication that the emperor was regarded as the foremost monarch in Christendom. This is implied by the turn of phrase in the Eiríksdrápa of Markus Skeg- gjason (noted above, 348), and more indirectly by the relative importance placed on the visit to Constantinople in all of the aforementioned accounts. The overwhelmingly positive relationship between the Byzantine emperor and various Scandinavian monarchs during the twelfth century is interesting in itself.26 But what is also noteworthy is that the image of the Byzantine empire in Old Norse sources was not subject to radical shifts after this period. The state of the empire following the debacle of the Fourth Crusade receives scant attention in Scandinavian sources. Very little attention was paid to the religious schism between the empire and Latin Christi- anity. In fact, the existence of such a schism seems to have been news to the Icelanders when they learnt of its putative resolution at the Council of Lyon in 1274!27 In Old Norse sources, little effort was made to make sense of the new political realities in the Balkans and Asia Minor in the thirteenth and fourteenth centuries. Instead, their his- toriographers’ attention turned towards a past more distant than the twelfth century, the period of the Varangian Guard. Emperors and kings Olaf Tryggvason (995–1000), the king of Norway associated with the Christianisation of Norway, Iceland and Greenland, was a character of great importance within Old Norse historiography and his reign is usually seen as marking a watershed between the pagan and Christian periods. In the late twelfth-century Ólafs saga Tryggvasonar, Olaf is also cast as the person who introduced Christianity to Rus. Even if hardly a reliable source for events two hundred years earlier, this account is nevertheless an early exam- ple of how the Christianisation of Scandinavia was connected with events in the east.28 In the narrative, Olaf is made out to be the chief missionary to Rus, actually travelling 26 The common scholarly opinion has been that the ‘special relationship’ between Scandinavians and the Byzantine empire ceased during the crusading period: Bagge 1990, 172. 27 See Sverrir Jakobsson 2008a, 175. 28 Jan Ragnar Hagland regards four Norwegian kings as having “strong, personal contacts” with the east, beginning with Olaf Tryggvason in 995: see Hagland 2005, 154.  The Varangian legend 351 to the Byzantine empire in order to bring missionaries from there.29 But although Ólafs saga Tryggvasonar links Olaf with the Byzantine empire through his mission work and depicts him as the noble servant of two great eastern monarchs, Prince Vladimir of Kiev and King Boleslaw of Poland, no attempt is made to connect him with the Va- rangian Guard. The main reason for this is probably that another Norwegian king, Harald Hardrada, was already renowned for his Varangian connections. As noted above (346), King Harald’s exploits in the Byzantine empire are men- tioned by contemporary authors such as Kekaumenos. Harald is also the subject of some discussion in the Gesta Hammaburgensis ecclesiae pontificum by Adam of Bremen, written sometime in the 1070s. Concerning Harald’s youth, Adam is quite laconic, sim- ply stating: “He was a powerful and triumphant man, who had formerly participated in several wars against barbarians in Greece and the regions of Scythia”.30 The oldest surviving history of the Norwegian kings, the Historia de antiquitate regum Norwag- iensium of Theodoricus Monachus, also mentions that Harald came to Norway from ‘Grecia’ and carried home with him great treasure. His exploits abroad are summarised thus: This Harald had performed many bold deeds in his youth, overthrowing many heathen cit- ies and carrying off great riches in Rus and in Ethiopia (which we call Bláland in our mother tongue). From there he travelled to Jerusalem and was everywhere greatly renowned and vic- torious. After he had travelled through Sicily and taken much wealth by force there, he came to Constantinople. And there he was arraigned before the emperor; but he inflicted enough shameful humiliation upon that same emperor and, making an unexpected escape, he slipped away.31 As noted by scholars, this narrative seems to be based partly on skaldic poems which were later used in more voluminous works in Old Norse, such as Morkinskinna and Heimskringla. The anecdote about a quarrel with the emperor corresponds to a degree with the tale told by Kekaumenos: Harald wished in the time of the emperor Monomachos to get royal permission to return to his own land, but it was not forthcoming. Indeed, the road out was obstructed. Yet he slipped away and took the throne in his own country in place of his brother Olaf.32 However, the details of the quarrel are different, as in the Historia de antiquitate regum Norwagiensium Harald seems to have been the subject of some accusation and appears to have somehow disgraced the emperor in making his getaway. Apart from Kekau- 29 Ólafs saga Tryggvasonar 40–42. On the historical value of this account see Jackson 2011, 121–24. 30 Adam of Bremen, Gesta 346. On the concept of Scythia see Janson 2011, 46–49. 31 Monumenta Historica Norvegicae 57. 32 Str, ed. and Russian tr. Litavrin, 300–01; ed. and tr. Roueché, 97.22–25. 352 Sverrir Jakobsson menos, the only source earlier than Theodoricus to mention this incident is the Gesta regum Anglorum by William of Malmesbury, who suggests that Harald had defiled a noble lady (which might explain the words “probrosa ignominia”).33 Harald’s Sicilian expedition is also mentioned by Kekaumenos and his wars in Rus might correspond to what Adam of Bremen calls the region of the Scythians, but neither eleventh-century source mentions Harald warring in Ethiopia. Nor was this exciting detail taken up by the more extensive narratives composed about Harald in the thirteenth century. Legends connected with Harald Hardrada were elaborated in the thirteenth-cen- tury Kings’ Sagas such as Morkinskinna and Heimskringla. They use the skaldic poems more extensively and provide greater detail on central events such as the invasion of Sicily, in which Harald is portrayed as a rival of the Byzantine general George Mani- akes.34 An interesting variation is provided by Saxo Grammaticus, who describes Har- ald fighting a dragon in a Byzantine dungeon.35 Nevertheless, the main outline of the plot is the one provided by Theodoricus, and it focuses on certain details.36 The first is the immense wealth gathered by Harald during his service with the Byzantine emperor. The second concerns the intrigues which made him leave Constantinople in a clandes- tine manner. By contrast, the narrative concerning Olaf Tryggvason focuses on religious mat- ters. By making this apostle to the north also responsible for bringing Christianity to Rus, Icelandic historiographers forged a clear link between eastern and western Chris- tianity, an idea which kept its appeal throughout the middle ages and figures promi- nently in works from the last quarter of the fourteenth century. These two kings were both active in the period when Nordic warriors were pre- dominant in the Varangian Guard. Their stories served as prototypes for accounts of less prominent persons who were said to have served the Byzantine emperor. These were mainly Icelanders, and the accounts of them were written down slightly later than the tales about King Olaf and Harald, mainly in the thirteenth and fourteenth centu- ries. The development of this ‘Varangian legend’ will be examined further in the fol- lowing section. 33 William of Malmesbury, Gesta regum Anglorum vol. 1, 261. 34 Mork. vol. 1, 84–118; Heimskringla 69–91. 35 Saxo Grammaticus, Gesta Danorum vol. 2, 10–13. See Fledelius 1996, 213–14. 36 The view of Harald and his relationship with the empire could, however, vary from source to source; see for example Bagge 1990, 179–90.  The Varangian legend 353 Legends of the Varangians The body of literature commonly known as the Sagas of the Icelanders (Íslendingasögur) had its heyday in the thirteenth and fourteenth centuries. The earliest known examples were composed in the second quarter of the thirteenth century, in the wake of the large compilations of Kings’ Sagas such as Morkinskinna and Heimskringla. It has even been suggested that the Sagas of the Icelanders were originally elaborations of shorter episodes dealing with Icelanders at the Norwegian court. This conjecture is supported by the fact that some of the earliest known sagas feature the exploits of Icelanders who were, at least at some point in their careers, retainers of the Norwegian king.37 By the second half of the thirteenth century the evolution of the sagas was well under way, with notable examples such as Laxdœla saga, Heiðarvíga saga and Brennu-Njáls saga being composed in this period. Most Sagas of the Icelanders are set in the decades before and after the introduc- tion of Christianity in Iceland, an event which has traditionally been dated to the year 999. The Christianisation then serves as a chronological and structural turning point in the sagas, creating a divide between the old pagan times and the new and improved customs introduced by the Christian faith. It is evident that as a history of particular events, the sagas are of limited value since the action takes place two or three hundred years before the time of their composition.38 There are, however, reasons for those studying Old Norse views of the Byzantine empire to be interested in this genre. It so happens that this chronological structure places the action of the sagas within the period when the Varangian Guard was at its peak, at least from the point of view of Scandinavians. It thus became a common narra- tive device to locate characters, who for some reason had to be removed from the thrust of the action in Iceland, at the court of the most glorious monarch in Christendom, the Byzantine emperor. There, the exploits of these characters were usually not listed in much detail, as it could be taken for granted that they had been exalted by serving such a noble master. If of little value as factual sources about the fate of particular individuals, what is the historical value of the Varangian episodes in the Sagas of the Icelanders? Are they nothing more than literary topoi? This is surely not the case, seeing that literary stere- otypes can shed light on a society’s thought processes. The reason why the Varangian motif was so popular in this particular genre is linked to the Byzantine empire’s posi- tion within the prevalent worldview of medieval Icelanders, and to a large degree of other Scandinavians as well. 37 See Bjarni Einarsson 1961. 38 For an overview of the genre see Vésteinn Ólason 2005. 354 Sverrir Jakobsson Since Iceland is central to the saga genre, Varangians usually appear in two contexts: either as men who have served in the guard but have returned to Iceland; or as protag- onists who have to leave Iceland and seek their fortune elsewhere, in this case in Con- stantinople. Different motifs are used according to the different contexts. One of the earliest instances of the first motif is in Hallfreðar saga, an early thir- teenth-century text which became part of a saga cycle connected with Olaf Trygg- vason. In this case, the eponymous hero is courting a woman who is betrothed to a wealthy and popular farmer called Gris Sæmingsson: “he had travelled abroad all the way to Miklagarðr and received much honour there”.39 Gris’ past is only referred to on one occasion, when Hallfred is about to duel with him, but is discouraged by the death of King Olaf. Gris proves surprisingly sympathetic to his plight and refuses to consider this an act of cowardice: “It is not so, I had less honour from the emperor and yet I considered it a great event when I lost my lord; the love towards a liege lord is fiery”.40 In their different ways, both Hallfred and Gris exemplify the ideal of service to a noble lord. It is not evident from the saga whether the wealth and social position of Gris are related to his former service with the emperor, but this is stated more clearly in two other thirteenth-century texts, Laxdœla saga and Hrafnkels saga Freysgoða. In the for- mer, young Bolli Bollason travels abroad and visits the courts of Norway and Den- mark. He then continues until he reaches Constantinople: He spent a brief time there until he acquired for himself a place in the Varangian Guard; we have not heard any tales of a Northman joining the service of the emperor before Bolli Bollason did. He spent very many winters in Miklagarðr and was considered the stoutest fellow in all hardship and always closest to the front ranks. Varangians had a high opinion of Bolli, while he was at Miklagarðr.41 All this is just a prelude to his return to Iceland, when great emphasis is placed on the glitz and glamour accompanying the return of a Varangian to his northern homeland: Bolli brought out with him much wealth, and many gems that dignitaries had given him. Bolli was such a richly-adorned fellow when he came back from this journey that he would wear no clothes but of scarlet or silk, and all his weapons were gilded: he was called Bolli the courteous. He made it known to his shipmates that he was going west to his own region, and he left his ship and goods in the hands of his crew. Bolli rode from the ship with eleven men, and all his fol- lowers were dressed in scarlet, and with gilded saddles, even though Bolli was peerless among them. He had on the silken clothes which the emperor had given him, he had around him a scarlet cape; and he had the sword Fótbítr [Foot- or Leg-Biter] girt on him, the hilt of which was ornamented with gold, and the grip woven with gold. He had a gilded helmet on his head, and a red shield on his flank, with a knight painted on it in gold. He had a lance in his hand, as is the 39 Vatnsdœla saga, Hallfreðar saga, Kormáks Saga 144. 40 Vatnsdœla saga, Hallfreðar saga, Kormáks Saga 192. 41 Laxdœla saga 214–15.  The Varangian legend 355 custom in foreign lands; and wherever they took quarters the women paid heed to nothing but gazing at Bolli and his ornaments, and those of his followers. 42 This kind of conspicuous wealth is reminiscent of reports of the great treasure of Har- ald Hardrada in the Kings’ Sagas; the wealth of Byzantium seems even greater in the context of medieval Iceland. But the value of his jewellery was more than just that of precious stones and metals in general. There was also symbolic value in the fact that most of these precious things were presents from a noble master. In that sense, Bolli is no different from Gris Sæmingsson, although his conspicuous showmanship is a far cry from the quiet dignity of the latter. Both gained in honour and wealth by associating with the noble lord in Constantinople. In Hrafnkels saga Freysgoða, a text from the last quarter of the thirteenth centu- ry, two secondary characters are returned Varangians. One is the brother of the main character’s chief antagonist, a person by the name of Eyvind Bjarnason. He was a sailor who went “abroad and ended in Miklagarðr, where he was much honoured by the king of the Greeks, and he stayed there for a while”.43 When he returns after seven years he wears coloured clothes and has a fine shield, and he had “educated himself a great deal and had become the bravest of men”.44 Another character in the saga, Thorkell Thjostarsson, had also been abroad for seven years “and gone to Miklagarðr, but I am now a retainer of the emperor”.45 Neither of them displays any conspicuous wealth and, although they evidently gained some social prestige from their stay in Constantinople, neither of them is elevated to the lofty heights of Bolli Bollason.46 Do the fates of these characters, portrayed in historical narratives composed much later than the time in which in the events took place, bear any relationship with those of actual Varangians? In the thirteenth century, memories of people returning from Miklagarðr were perhaps not so faint. In 1217 Sturla Sighvatsson, the eighteen-year-old son of a chieftain, gained some notoriety when he tried to take a sword from a local farmer and managed to wound him seriously in the process. In this, he was quietly en- couraged by his father.47 But why were this father and son prepared to disturb the peace in the region for the sake of a sword? The artefact in question was called Brynjubítr, ‘Mail-Biter’, and had been brought from Constantinople by a person known as Sigurd 42 Laxdœla saga 224–25. For a comparison, see below, 363–87. 43 Austfirðinga sögur 100. 44 Austfirðinga sögur 125. 45 Austfirðinga sögur 111. 46 According to Sigfús Blöndal, the evidence for the existence of these two men is of no historical value: SB, 310–11. He was, however, much more inclined to accept the existence of Gris Sæmings- son. But, as argued above, the main value of these narratives about individual Varangians is as evidence of the prevailing view of the Byzantine empire in thirteenth-century Iceland. 47 Sturlunga saga vol. 1, 261. 356 Sverrir Jakobsson the Greek (grikkr).48 This man had participated in dramatic events in the region some twenty years earlier, by which time he had already acquired his nickname. Sigurd’s most notable achievement in these battles had been to save a wealthy farmer by herding him into a church “and then he stood before the church and proclaimed that he would defend it, as long as he was able to stand”.49 While not of the highest rank in Iceland, Sigurd was evidently remembered as a valiant man and a defender of Christian values. Although the Varangian Guard is not mentioned in connection with Sigurd, he had clearly served in Constantinople in some way and had brought home a sword as proof, an artefact coveted by noble lords after his death. There are also several examples in the sagas of characters who end their careers in the Varangian Guard, having left their troubles behind in Iceland. The narratives con- cerning them are usually quite laconic, as events abroad seldom form the main plot in the sagas. In the early thirteenth-century Heiðarvíga saga, two men seek their fortune in Constantinople at different times following troubles in Iceland. One of them, Gestr, has slain a noble chieftain and is pursued by the chieftain’s son, Thorstein. Their jour- ney ends in Miklagarðr, where Gestr joins the Varangian Guard. Thorstein finds him there and wounds him during a wrestling match. The Varangians want to kill Thorstein for violating the rules of the contest, but Gestr intercedes and even pays for Thorstein’s journey home. In return, Thorstein promises to stop his pursuit, provided that Gestr will not return to the Nordic countries (Norðrlönd).50 Later in the saga, Bardi Gudmundarson is exiled following a series of killings. He visits the kings of Norway and Denmark, returns to Iceland and is married there, but then returns to Norway, where he divorces his wife. Finally, he travels to Rus (Garðaríki) and eventually joins the Varangian Guard: and all the Northmen thought highly of him, and held him in great affection. Every time the kingdom needed to be defended, he took part in the expedition and became known for his hardiness and had a large regiment of men around him. Bardi spent three winters there and received great honour from the king and all the Varangians.51 He is eventually killed against overwhelming odds, in an unspecified battle. Another important saga character connected with the Varangian Guard was Kol- skegg, brother of the famous Gunnar from Hlíðarenda, who is one of the central char- 48 For a comparison, see below, 363–87. 49 Sturlunga saga vol. 1, 208. The notion of church sanctity was heavily contested in Iceland in the 1190s, see Sverrir Jakobsson 2008b. 50 Borgfirðinga sögur 243–44. A similar pursuit occurs in the Grettis saga, in a form heavily influ- enced by Romance literature. See Guðmundur Andri Thorsson 1990. As this text may be of very late date (early fifteenth century) it will not be discussed here. 51 Borgfirðinga sögur 325.  The Varangian legend 357 acters in the Brennu-Njáls saga. Gunnar dies a heroic death after tragically refusing a settlement to go into exile for three years. Kolskegg, who has loyally supported his brother throughout his adventures, decides to honour the settlement and leaves Ice- land, eventually joining the court of the Danish king Sven Forkbeard: One night Kolskegg dreamed that a man approached him; a radiant man, who woke him up, saying: “Arise and come with me.” “What do you want of me?” asked Kolskegg, to which the man replied “I will give you a bride, and you shall be my knight.” Kolskegg believed he had agreed to this, whereupon he awoke. Kolskegg consulted a wise man about the dream, who interpreted it as meaning he would travel to southern lands and become the knight of God. Kolskegg was baptised in Denmark, but did not like it there, so travelled east to Rus, where he wintered. He then voyaged to Miklagarðr, where he entered service. The last that was heard of Kolskegg was that he had taken a wife in Miklagarðr; becoming the leader of a Varangian band and remaining there until his dying day. He is now out of this story.52 The story of Kolskegg has markedly Christian overtones. By serving in the Varangian Guard, Kolskegg has become a knight of the Lord. It is possible to identify a certain dichotomy based on these examples. Those who return from the empire gain great wealth and even greater glory from serving the noble emperor (most notably Bolli Bollason). They mirror Harald Hardrada, the prototype for examples of immense wealth from the east. Those who end their lives as Varangians achieve an advantage, either in reputation (such as Bardi Gudmundarson) or in be- coming a knight of God (such as Kolskegg). Here, the parallel is closer to Olaf Trygg- vason, who reportedly ended his life as a hermit “in Greece, the Holy Land and Syria”.53 A View from the periphery In the nineteenth and early twentieth centuries, those episodes in the sagas that men- tion the Byzantine empire received much scholarly attention. The main focus of schol- ars such as Gustav Storm and Sigfús Blöndal was to establish whether the Old Norse sources contained reliable information about the history of the Byzantine empire and to ascertain the facts relating to the Varangian Guard in the tenth and eleventh cen- turies.54 As the sagas’ credibility as sources about the distant past began to diminish, so did interest in these episodes. It can, however, be argued that the main value of the sagas’ evidence is as a source for the Old Norse world itself, especially its prevailing attitudes and mentalities. First, there is the question of the relationship of Nordic monarchs to the Byzan- tine emperor. It has often been noted that early medieval ideas of sovereignty revolved to a degree around “the legal axiom embodied in the Corpus Juris Civilis, namely that 52 Brennu-Njáls saga 197. 53 Ólafs saga Tryggvasonar 242. 54 See Storm 1884 and SB. 358 Sverrir Jakobsson theoretically, the emperor was lawful overlord and supreme monarch of Europe: every king and prince was inferior to him”.55 According to the testimony of the Kings’ Sa- gas, this was the prevailing view in the Old Norse world of the twelfth and thirteenth centuries. Nevertheless, neither the Byzantine empire nor the Nordic states can be characterised as ‘feudal’ at this time. The bond between the emperor and the Scandi- navian monarchs was not that of a lord and his vassals; it was more a symptom of the fragmented sovereignty characterising society in general. In the middle ages, ultimate sovereignty had an ultimate source—God himself—and the Byzantine emperor could be seen as one of his most distinguished representatives. Thus, service to the emperor was also service to a higher Lord, as exemplified by Kolskegg’s dream. From a Byzantine viewpoint, the relationship between the empire and other countries was not, and could not be, a relationship between equals. It was axiomatic to Byzantine political thinking that their emperor was the kosmokrator—the lord of the world; and, as seen in the De cerimoniis, by the tenth century they had developed the concept of a hierarchy of subordinate states, revolving in obedient harmony around the throne of the universal autocrat in Constantinople. Within his own lands a prince could be a fully sovereign ruler, but in relation to the empire he occupied a subordinate position in the hierarchical structure of the Commonwealth.56 The relationship between Nordic Varangians and the Byzantine empire also raises the issue of periphery and centre. Questions concerning peripheries and centres have been key to the study of development theory in the past few decades. Usually, howev- er, the focus has been on economic relations between areas. In his seminal study on world-systems, Immanuel Wallerstein defines a world-system as “an economic but not a political entity”, in contrast to political empire, which he regards as a “primitive means of economic domination”.57 According to Wallerstein, an economic system depends on a system of government which directs the flow of economic goods from the periphery to the centre. The medieval period, in Wallerstein’s view, was characterised by the absence of such a system. In the twelfth century there existed “a series of empires and small worlds”.58 Since then, Janet Abu-Lughod’s study of medieval world-systems has modi- 55 Ullmann 1949, 3. See also André Grabar’s theory about the ‘Family of Princes’: “for the pious Emperor of Byzantium, God is simultaneously Father and Brother, head of the army and com- rade in arms, He Who in time of war ensures victories and in time of peace just government. Above all, for the Emperor God is a friend; the basileus has the Master of the universe as a friend, and as a result—are not the goods of friends common?—the basileus who is loved by God be- comes a universal sovereign himself ” (Grabar 2007, 5). 56 See Obolensky 1970. 57 Wallerstein 1974, 15. 58 Wallerstein 1974, 17.  The Varangian legend 359 fied this simplistic picture of the medieval economy. In the view of Abu-Lughod, there were a number of such world-systems in the thirteenth and fourteenth centuries, but no single system exercised hegemonic power over the others.59 While the Scandinavian countries were on the periphery of the European economy during most of the middle ages, the same cannot be said of the Byzantine empire. Firstly, gold coins struck in the empire were long the preferred specie for international transactions. Until the second half of the thirteenth century, Venetian and Genoese merchants used gold coins from Constantinople or Egypt rather than striking their own.60 And secondly, Constantino- ple was Christendom’s largest and most prosperous city and the gateway to Central Asia. It is thus no wonder that Venetian merchants coveted and benefitted from con- trolling trade with Constantinople, extracted trading concessions from the emperor in the eleventh century, and then conquered the City in the Fourth Crusade.61 This revision of history leaves north-western Europe as a very marginal area in economic terms for much of the time. Even from Wallerstein’s Eurocentric perspective, Europe as a whole cannot be regarded as a hegemonic power during the middle ages. As already conceded by Wallerstein, north-western Europe did not simply have a sub- sistence economy, and its social relations grew out of the disintegration of the Roman empire. In Wallerstein’s words, “The myth of the Roman empire still provided a certain cultural and even legal coherence to the area. Christianity served as a set of parame- ters within which social action took place. Feudal Europe was a ‘civilisation’, but not a world-system.”62 This murky entity—‘civilisation’—amounts to the cultural and legal coherence provided by the myth of empire, and to the parameters set by the church that defined Christendom. There can be no doubt that both the Roman empire and the Christian church were of enormous importance for defining the identities of those who saw themselves as belonging to this world. And yet it seems facile to think of this entity as something other than a world-system. How can the expansion of Europe in the elev- enth, twelfth and thirteenth centuries be explained, if not in economic terms? What was different, however, was the relative importance of culture and the economy within this system. If Scandinavia was on the periphery, the nature of that peripheral status is open to debate. Was it mainly political, cultural or economic? The most important studies on centres and peripheries concentrate on their economic aspects, but that leaves the relationship between centres and peripheries within medieval Christianity largely un- 59 Abu-Lughod 1989, 32–38. 60 Abu-Lughod 1989, 15, 67. 61 Abu-Lughod 1989, 105, 119. 62 Wallerstein 1974, 17–18. 360 Sverrir Jakobsson accounted for. Even if it did not constitute an economic world-system, there existed a unity within the Catholic world of the middle ages, provided by the church and the legacy of the Roman empire. Rome, Jerusalem and Constantinople were the cultural and political centres of this entity. The distance of the north from the political, cultural and economic centres had to be compensated for. A journey to the centres of power could increase the cultural capital of the participants. It is a topos in narratives describing such journeys that the prestige of those who went on them increased. This was reflected in several ways. For instance, a person who had spent time with foreign dignitaries was supposed to have adopted good manners. He had adapted himself to the manners of noble men. It was also an advantage to be able to show tokens of the respect one had gained at the hands of foreign potentates, and gifts from a noble lord usually served as such tokens. The gilded exuberance of Bolli Bollason becomes very understandable from such a perspec- tive. Conclusion A ‘history of the Varangian Guard’, based mostly or entirely on Old Norse sources, will necessarily be the story of a legend. The legend of the Varangians which has been preserved in texts from the thirteenth and fourteenth centuries was to some degree a reflection of the past. However, that past was probably not the heyday of the Varangi- an Guard, but rather the experiences of crusaders in the twelfth century. Significantly, the thirteenth-century fragmentation of the Byzantine empire never became solidly anchored in the Old Norse works that form the textual basis of the present analysis. The Varangian legend revolved around a few major themes. One was the wealth and prestige to be had through service to the emperor. The prototypical Varangian in this sense was Harald Hardrada, with his vast treasure; but less exalted travellers, such as the relatively obscure Sigurd the Greek, who lived in the north of Iceland around 1200, also had the capacity to bring home tokens of their service, encapsulated in a sword that local magnates considered worth fighting for. However, the road from Con- stantinople to the north went both ways, and the fates of those destined to end their lives in the Byzantine empire also became part of the Varangian legend. Here the em- phasis was much less on material wealth and the tokens of honourable service, and far more on the glory that came posthumously from having served a true Christian lord and, ultimately, the Lord himself. For these travellers, being a Varangian was not just means to an end, but an end in itself.  The Varangian legend 361 Bibliography Primary Sources Adam of Bremen, Gesta Hammaburgensis eccle- Den norsk-islandske skjaldedigtning 800–1200. siae pontificum. Ed. W. Trillmich & R. Buch- Vol. B–1. Rettet tekst. Ed. Finnur Jónsson. Co- ner, Quellen des 9. und 11. Jahrhunderts zur penhagen 1910. Geschichte der Hamburgischen Kirche und des Ólafs saga Tryggvasonar af Oddr Snorrason munk. Reiches, 137–499. Darmstadt 1961. Ed. Finnur Jónsson. Copenhagen 1932. Ágrip af Nóregskonunga sögum, Fagurskinna— Orkneyinga saga. Ed. Finnbogi Guðmundsson. Nóregs konunga tal. Ed. Bjarni Einarsson. ÍF ÍF 34. Reykjavik 1965. 29. Reykjavik 1985. Saxo Grammaticus, Gesta Danorum. Danmark- Austfirðinga sögur. Ed. Jón Jóhannesson. ÍF 11. shistorien, 2 vols. Ed. K. Friis-Jensen, tr. P. Reykjavik 1950. Zeeberg. Copenhagen 2005. Borgfirðinga sögur. Ed. Sigurður Nordal & Guðni Stephen (Asolik) of Taron, Histoire universelle, Jónsson. ÍF 3. Reykjavik 1938. vol. 1, French tr. E. Dulaurier. Paris 1883; vol. Brennu-Njáls saga. Ed. Einar Ólafur Sveinsson. 2, French tr. F. Macler. Paris 1917. ÍF 12. Reykjavik 1954. Sturlunga saga, 2 vols. Ed. Jón Jóhannesson, Danakonunga sögur. Ed. Bjarni Guðnason. ÍF 35. Magnús Finnbogason & Kristján Eldjárn. Reykjavik 1982. Reykjavik 1946. Grettis saga. Ed. Guðni Jónsson. ÍF 7. Reykjavik Sverris saga. Ed. Þorleifur Hauksson. ÍF 30. Rey- 1936. kjavik 2007. Heimskringla, vol. 3. Ed. Bjarni Aðalbjarnarson. Vatnsdœla saga, Hallfreðar saga, Kormáks Saga. ÍF 28. Reykjavik 1951. Ed. Einar Ólafur Sveinsson. ÍF 8. Reykjavik John Skylitzes, Synopsis historiarum. Ed. H. 1939. Thurn. Berlin & New York 1973. Veraldar saga. Ed. Jakob Benediktsson. Copen- Landnámabók. Ed. Jakob Benediktsson. ÍF 1. hagen 1944. Reykjavik 1968. William of Malmesbury, Gesta regum Anglorum, Laxdœla saga. Ed. Einar Ólafur Sveinsson. ÍF 5. 2 vols. Ed. R. A. B. Mynors. Oxford 1998– Reykjavik 1934. 1999. Monumenta Historica Norvegicae. Latinske Yahya of Antioch, Histoire de Yahya-ibn-Sa‘id kildeskrifter til Norges historie i middelalderen. d’Antioche, 3 vols. Ed. and French tr. I. Ed. G. Storm. Kristiania 1880. Kratchkovsky, A. Vasiliev et al. Paris & Turn- hout 1924–1997. Secondary Literature Abu-Lughod, J. L. 1989. Before European He- T. Hägg (eds) 1990. Hellas og Norge. Kontakt, gemony. The World System AD 1250–1350, New komparasjon, kontrast, 169–92. Bergen. York & Oxford. Bibikov, M. 1996. “Byzantinoscandica”,  in BIII, Ármann Jakobsson 2005. “Royal Biography”, in 201–11.  McTurk (ed.) 2005, 388–402. Bjarni Einarsson 1961. Skáldasögur. Um uppruna —. 2013. “Image is Everything: The Mor- og eðli ástarskáldsagnanna fornu. Reykjavik. kinskinna Account of King Sigurðr of Nor- Ciggaar, K.  N.  1974. “L’emigration anglaise  à way’s Journey to the Holy Land” Parergon 30, Byzance après 1066. Un nouveau texte en latin 121–40. sur les Varangues à Constantinople” REB 32, Bagge, S. 1990. “Harald Hardråde i Bysants. To 301–42. fortellinger, to kulturer”, in Ø. Andersen & 362 Sverrir Jakobsson —. 1996. Western Travellers to Constantino- —. 1971. The Byzantine Commonwealth. ple. The West and Byzantium, 962–1204: Cul- Eastern Europe, 500–1453. London. tural and Political Relations. Leiden. Poppe, A. 1976. “The Political Background to the Fledelius, K. 1996. “Royal Scandinavian Travel- Baptism of the Rus’: Byzantine-Russian Rela- lers to Byzantium: the Vision of Byzantium in tions between 986–89” DOP 30, 195–244. Danish and Norwegian Historiography of the Rahbeck Schmidt, K. 1970. “The Varangian Early 13th Century, and in the Danish Histor- Problem. A Brief History of the Controversy”, ical Drama of the Early 19th Century”, in BIII, in Varangian Problems (Scando-Slavica sup- 212–18.  plementum 1), 7–20. Grabar, A. 2007. “God and the ‘Family of Princ- Seibt, W. 1992. “Der historische Hintergrund es’ Presided Over by the Byzantine Emperor”, und die Chronologie der Taufe der Rus’ in EOE, 1–7. (989)”, in A.-E. N. Tachiaos (ed.) 1992. The Guðmundur Andri Thorsson 1990. “Grettla” Legacy of Saints Cyril and Methodius to Kiev Skáldskaparmál 1, 100–17. and Moscow, 289–303. Thessalonica. Hagland, J. R. 2005. “Armenske biskopar i Nor- Shepard, J. 1984–1985. “Yngvarr’s Expedition den på 1000-talet”, in FBN, 153–63. to the East and a Russian-Inscribed Stone- Jackson, T. N. 2011. “Rus’ and Scandianavia: the Cross” Saga-Book 21, 222–92. Orthodox-Latin Division in Old Norse-Ice- Stender-Petersen, A. 1953. Varangica. Aarhus. landic Literature and in Reality”, in ECWVG, Storm, G. 1884. “Harald haardraade og 120–32. væringerne i de græske keiseres tjeneste” His- Janson, H. 2011. “Scythian Christianity”, in torisk tidskrift 2, 354–86. ECWVG, 33–57. Svärdström, E. 1970. “Runorna i Hagia Sophia” Jansson, S. B. F. 1984. Runinskrifter i Sverige. Fornvännen 65, 247–49. Stockholm. Sverrir Jakobsson 2008a. “The Schism that Nev- Jesch, J. 2001. Ships and Men in the Late Viking er Was: Old Norse Views on Byzantium and Age. The Vocabulary of Runic Inscriptions and the Rus” BSl 66, 173–88. Scaldic Verse. Woodbridge. —. 2008b. “The Peace of God in Iceland in Kaldellis, A. 2007. Hellenism in Byzantium. The the 12th and 13th Centuries”, in P. Krafl (ed.) Tranformations of Greek Identity and the Re- 2008. Sacri canones servandi sunt. Ius canoni- ception of the Classical Tradition. Cambridge. cum et status ecclesiae saeculis XIII–XV, 205– Larsson, M. G. 1989. “Nyfunna runor i Hagia So- 13. Prague. phia” Fornvännen 84, 12–14. Ullmann, W. 1949. “The Development of the McTurk, R. (ed.) 2005. A Companion to Old Medieval Idea of Sovereignty” The English Norse-Icelandic Literature and Culture. Ox- Historical Review 64, 1–33. ford. Vésteinn Ólason 2005. “Family Sagas”, in Mc- Morrisson, C. 1981. “Le role des varanges dans Turk (ed.) 2005, 101–18. la transmission de la monnaie byzantine en Wallerstein, I. 1974. The Modern World-System. Scandinave”, in PNB, 131–40. Capitalist Agriculture and the Origins of the Obolensky, D. 1970. “The Relations Between European World-Economy in the Sixteenth Cen- Byzantium and Russia (11th–15th Century)”, tury. New York. in XIIIth International Congress of Historical Sciences, 1–13. Moscow.