Anglo-Danish Empire - A Companion to the Reign of King Cnut the Great

["Chapter 19 Cnut\u2019s Reign in England and Denmark 429 The first of these, named \u201cWrytsleof\u201d and titled dux, attested one of Cnut\u2019s char- ters from 1026 (S 962).47 The form Wrytsleof is presumably an Anglo-Saxon vari- ant of the Slavonic name Vartislav. Apart from the name, the dux in question remains unknown to us, just as the reason for his stay in England, or where this was, or for how long it lasted. The date of Wrytsleof\u2019s attestation gives some clue, however, for it followed Cnut\u2019s expedition to Denmark in 1023 and his likely military encounter with the Slavs, particularly the Obodrites. As I have already mentioned the possibility of the Slavs supporting the rebellious Earl Thorkell, one cannot exclude the possibility of Wrytsleof being a hostage to guarantee the loyalty of the Obodritian dynasty towards the Danish ruler. On the other hand, accounts of the turbulence that affected the Obodritian dynasty in the early 1020s could suggest that \u201cVartislav\u201d was a refugee in England applying for Cnut\u2019s protection. King Cnut could have found such a situation advantageous, for it would have given him a reason for intervening in Polabian affairs when necessary.48 A second Slav in Cnut\u2019s court may be inferred from John of Worcester, who mentions a marriage between an anonymous daughter of Sveinn Forkbeard and \u201cWyrtgeorn,\u201d called \u201crex Winidorum\u201d (king of the Wends [i.e., Slavs]).49 Although John\u2019s account lacks confirmation in other sources, scholarly opinion supports its reliability.50 The incident may refer to \u201cSantslaua,\u201d the sister of Cnut referenced in the Liber Vitae of the Winchester New Minster. On the other hand, the identification of the Slavonic king is more problematic. Perhaps John was referring to the same person who, as Wrytsleof, attested Cnut\u2019s charter. These remain for the most part speculations; taken together, however, they sug- gest that the territories of the Polabian Slavs, and of the Obodrites in particular, remained important in Cnut\u2019s plans to preserve Danish influence in this strate- gic region, following the policies of his predecessors. Conclusion: Interests, not Friends In conclusion, one must observe that Cnut\u2019s kinship with the Piasts did not de- termine or even influence his Baltic and Western Slavonic policy, which was guided solely by his interests as king of Denmark. Cnut strictly followed his 47 Keynes, \u201cCnut\u2019s Earls,\u201d 64\u201365. 48 Morawiec, Knut Wielki, 213. 49 Chronicle of John of Worcester, ed. Darlington and McGurk, 511. 50 Bolton, Empire of Cnut, 215\u201316. See also Gazzoli in this volume, pp. 413\u201314.","430 Jakub Morawiec predecessors in this respect. One of his policy\u2019s key elements \u2013 to maintain his influence among the Polabian Slavs, particularly the Obodrites \u2013 was dictated by the need both to secure the borders between the Danes, Saxons, and Slavs, and to withstand the ambitions of the Empire and other political partners, in- cluding Poland. This policy of Cnut\u2019s explains the presence of Slavonic nobles such as Wrytsleof (Vartislav) at his court, who might already have proved use- ful to him in these areas. Despite Cnut\u2019s blood ties with the rulers of Poland, it is hard to say anything specific about their mutual relations. Either they simply passed unnoticed in the record, or, as seems more likely, there was no particu- lar reason to record them. This matter also concerns Polish relations with the Empire, in which one can see clearly the opposing interests of both sides. It is therefore not surprising that Mieszko II received no help from his cousin, King Cnut of England and Denmark, during his conflict with Emperor Conrad, even during the hardest times of failure, exile, and humiliation. For all his kinship with the kings of Poland, Cnut was his own man.","Barbara E. Crawford Chapter 20 St. Clement of Rome: Patron Saint of Cnut and the Dynasty of Denmark This chapter firms up the notion that St. Clement had a significant association with the Anglo-Danish dynasty of Cnut, by way of showing the relevance of St. Clement to the political, religious, and cultural circumstances of eleventh- century northern Europe,. I shall first focus on the development of St. Clement\u2019s cult in the principality of Rus\u2019, taking note of the authoritative standing of this prestigious papal martyr in Kyiv (Kiev) and of his relevance for the newly Chris- tianized dynasty of Rus\u2019, at the time it was looking for a protective patron.1 As we shall see, that example may have served as a model for the rulers of the Nor- wegian kingdom, some of whom knew Kyiv well. How, when, and why the cult of Clement then spread to Denmark and England are questions which will be considered towards the end of this chapter. It is well known that St. Clement of Rome was a popular saint with the Danes.2 The spread of the cult of St. Clement in Denmark and Norway can be explained in part as a reflection of the need of polities and communities (which were new to Christianity) for a saintly figure who they hoped might help them realize their spir- itual and political aspirations. In attempting to trace the growth of the cult in these countries and in England, primarily from the evidence of church dedications, my studies have examined the way in which several groups \u2013 namely political leaders, the ecclesiastical hierarchy, the mercantile\/trading classes, and fishing communi- ties \u2013 looked to Clement as an apostolic patron with useful protective functions and impeccable credentials.3 The variable locations of churches dedicated to Clem- ent suggest that this saint appealed to different populations for different reasons, so that, while the spread of the cult received an impetus from royal promotion, his 1 This aspect was explored at a conference commemorating Clement\u2019s martyrdom in Cherson (Sebastopol) in Crawford, \u201cCult of St. Clement in North Europe.\u201d 2 The study of the cult of Clement in Scandinavia was initiated by Erik Cinthio\u2019s seminal \u201cThe Churches of St. Clemens\u201d (1968). His researches were broadened by my own project on the cult in Scandinavia and England, in Crawford, Churches dedicated to St. Clement in Medieval England. 3 Crawford, Churches dedicated to St. Clement in Medieval England, Section 1.3.2. https:\/\/doi.org\/10.1515\/9781501513336-021","432 Barbara E. Crawford appeal resonated more widely than could be explained had his cult simply been promulgated in royal circles (see Map 20.1).4 Map 20.1: Clement dedications in Northern Europe. Nonetheless, Clement\u2019s appeal to the royal dynasties of northern Europe is an impor- tant aspect of our understanding of the spread of his popularity, and Erik Cinthio has alerted historians and archaeologists to the special character of churches dedicated to Clement in Scandinavia and their possible link with Cnut the Great (1016\u20131035), 4 Cinthio\u2019s study of the Clement churches stresses their significance in relation to centers of po- litical power, in \u201cThe Churches of St. Clemens,\u201d 112, 113. See my summary of this aspect of his study, in Crawford, Churches dedicated to St. Clement in Medieval England, Introduction, 1.2.","Chapter 20 Patron Saint of Cnut 433 king of England, Denmark, and Norway.5 To see how these dedications became coter- minous with Cnut\u2019s empire, let us begin with the cult in the principality of Rus\u2019. The Cult of St. Clement in Kyiv First, we need to summarize the circumstances of the martyrdom of Pope Clem- ent I (88\u201399), a successor to St. Peter, in order to understand why he might have appealed to the newly Christianized dynasties in Scandinavia.6 Clement was be- lieved to have earned his martyr\u2019s crown as a result of his death by drowning in the vicinity of Cherson in the Crimea, where he had been exiled during the Tra- janic persecutions (100\u2013117 CE) because of his success in converting some pow- erful members of the Roman administration. According to the later legends, he was forced to work in the marble quarries. However, he was so effective at con- verting his fellow quarry workers that he was condemned to death by drowning with an anchor round his neck.7 The anchor became the symbol of his martyr- dom in localities where the saint was popular (see Figure 20.1). Figure 20.1: Martyrdom of St. Clement by Bernardino Fungai (ca. 1500). \u00a9 York Museums Trust (York Art Gallery). 5 Cinthio, \u201cThe Churches of St. Clemens.\u201d 6 Clement was the fourth bishop of Rome, although sometimes regarded as the direct succes- sor to Peter in the apostolic succession, as it was claimed he had been consecrated pope by Peter himself. See Crawford, Churches dedicated to St. Clement in Medieval England, 29. 7 Traditions about his supposed martyrdom arose in the late fourth and fifth centuries and are recorded in the apocryphal \u201cClementine literature,\u201d as well as the sixth-century Latin passio. For the sources, see Crawford, Churches dedicated to St. Clement in Medieval England, 30.","434 Barbara E. Crawford The development of these traditions meant that Clement was considered to have protective powers for those whose lives exposed them to danger of drowning. Such protective powers were particularly relevant for rulers like the Norwegian and Danish kings, whose authority rested on domination of the northern waterways. Cnut was a \u201cthalassocrat,\u201d the historical \u201csea-king,\u201d whose northern empire con- sisted of three joint kingdoms separated by wide stretches of turbulent sea: he had a fleet of sixty ships at his disposal, and a naval force of between two and three thousand. Clement, appropriately, could have been regarded as a suitable protector for him and his retinue and their ships when voyaging around the North Sea. The martyred pope\u2019s success in mission provided a further useful model for aspirational missionary kings, particularly \u00d3l\u00e1fr Tryggvason (r. ca. 995\u20131000) and \u00d3l\u00e1fr (later St. \u00d3l\u00e1fr) Haraldsson of Norway (ca. 1016\u20131030).8 Cnut\u2019s aspirations to be ranked among the foremost leaders of Christian Europe also follow his program for Christianizing the Danes and establishing a Christian kingdom by building churches and appointing bishops. Clement\u2019s reputation as a successful missionary has been seen to be reflected in the Homily written for the Feast of St. Clement by Abbot \u00c6lfric of Eynsham (ca. 1000). This text has been interpreted by Joyce Hill as an attempt to strengthen Clement\u2019s reputation as an \u201ceffective converter to the faith.\u201d9 It affords good contemporary evidence that Clement was regarded as a model for those missionary kings who aimed to convert the pagans of their own day. How the cult of St. Clement reached Europe and in what circumstances it was adopted are questions which have been discussed many times. The bring- ing of Clement\u2019s relics to Rome in 867\u2013868 by Constantine (later Cyril) and Methodius must have been an important causal factor in the growth and devel- opment of the cult in western Europe.10 As far as Scandinavia is concerned, however, the eastern route to the Black Sea and including Kyiv is likely to have been the route by which knowledge of the cult was transferred north. Because of the supposed location of Clement\u2019s martyrdom in the Black Sea (or in the Sea of Azov), his cult became well established in the principality of Rus\u2019. In 988, Prince Vladimir of Kyiv (r. 980\u20131015) was converted and received relics from the priests of Cherson, including the head of Clement. These helped in the mass baptism of the population of Kyiv, and they gave status to the young Christian- ized principality.11 8 Haki Antonsson, \u201cEarly Cult of Saints,\u201d 21. 9 Hill, \u201c\u00c6lfric\u2019s Homily,\u201d 105. 10 Hofmann, Die Legende von Sankt Clemens, 144\u201349. 11 Crawford, Churches dedicated to St. Clement in Medieval England, 35, citing Cinthio, \u201cThe Churches of St. Clemens,\u201d 112; Hofmann, Die Legende von Sankt Clemens, 173; Shchapov, \u201cThe Assimilation,\u201d 59.","Chapter 20 Patron Saint of Cnut 435 Map 20.2: Route to Kyiv from Scandinavia via the Baltic. There is a strong likelihood that the Norwegians \u00d3l\u00e1fr Tryggvason and \u00d3l\u00e1fr Haraldsson became very aware of the cult of this papal martyr when they visited Kyiv in the later tenth and early eleventh centuries, respectively. Kyiv was doubtless the source of their interest in the cult of Clement that each of them then took back north with him. The \u00d3l\u00e1frs are two examples of the widely traveled Viking adventurers who sailed down the rivers of Russia to the Byzan- tine empire and stopped off at the Kyivan court, or so their sagas tell us.12 The first \u00d3l\u00e1fr was brought up at the court of Prince Vladimir from the age of nine, 12 There were other instances of close ties between the Norwegian royal dynasty and the prin- ces of Kyivan Rus\u2019 in the eleventh century. See Haki Antonsson, \u201cThe Cult of St. Olaf,\u201d 150).","436 Barbara E. Crawford and he stayed there for nine years. According to Snorri Sturluson\u2019s account in Heimskringla, \u00d3l\u00e1fr \u201chaf\u00f0i \u00fear it mesta yfirl\u00e1t\u201d (had the best treatment there) from the king and \u201ck\u00e6rleik af dr\u00f3ttningu\u201d (affection from the queen), while his role as chieftain of a well-equipped military force is also mentioned.13 The saga of \u00d3l\u00e1fr Tryggvason by Oddr Snorrason, written in a (now-lost) Latin original in the late twelfth century and translated into Icelandic in the thirteenth, says that \u00d3l\u00e1fr requested baptism in Greece, where he was then \u201cpr\u00edmsigna\u00f0r\u201d (prime- signed).14 The composite \u00d3l\u00e1fs saga Tryggvasonar en mesta of ca. 1300 says of \u00d3l\u00e1fr that on returning to Norway from the British Isles in 997, \u201cl\u00e9t hann ok reisa Clemens kirk[iu]\u201d (he had a church built for Clemens) at the royal resi- dence in Trondheim, the main power base of the kingdom at this time (in Heim- skringla his residence is named \u201cSkipakr\u00f3kr\u201d).15 We do not know whether \u00d3l\u00e1fr took some relics of Clement back north with him from Kyiv, but he can hardly have failed to see the importance of the apostolic saint\u2019s cult in the capital of Kyivan Rus\u2019. Two generations later, the connection between the Norwegian kings and the court at Kyiv was continued when \u00d3l\u00e1fr Haraldsson sought refuge there after he was driven from Norway in 1028. This \u00d3l\u00e1fr is also credited with having founded a church in Trondheim which was dedicated to St. Clement.16 However, since this is said to have taken place immediately on his arrival in Norway to claim the kingdom there, from Normandy in 1016, perhaps he only refounded \u00d3l\u00e1fr Tryggvason\u2019s church. \u00d3l\u00e1fr Haraldsson could have been influ- enced also by the veneration accorded to Clement in Rouen, where some ac- counts claim that he was baptized.17 The association of both these kings with the church dedicated to Clement in Trondheim, although it is found only in later saga accounts, suggests that each had a personal commitment to the saint as a powerful advocate of his own claim to power in Norway. The source of their belief in the effectiveness of Clement could have been the experience of the status of his cult: in the Kyivan 13 Heimskringla I, ed. Bjarni A\u00f0albjarnarson, 251. English translation from Heimskringla, trans. Finlay and Faulkes, I, 155. 14 \u00d3l\u00e1fs Saga eptir Odd Munk, ed. \u00d3lafur Halld\u00f3rsson, 164 (chap. 13). 15 \u00d3l\u00e1fs saga Tryggvasonar en mesta, ed. \u00d3lafur Halld\u00f3rsson, I, 369. 16 Heimskringla II: \u00d3l\u00e1fs Saga Helga, ed. Bjarni A\u00f0albjarnarson, 70 (chap. 53). Heimskringla: \u00d3l\u00e1fr Haraldsson, trans. Finlay and Faulkes, II, 43. 17 For evidence of the importance of St. Clement\u2019s status at the ducal court, and for \u00d3l\u00e1fr\u2019s close connections with the ducal power centre, and for his baptism in Rouen, see Crawford, \u201cChurches to St. Clement in Norway,\u201d 113\u201314, and Crawford, Churches dedicated to St. Clement in Medieval England, 18, 193\u201395, 196.","Chapter 20 Patron Saint of Cnut 437 principality, in the case of \u00d3l\u00e1fr Tryggvason; and in Normandy and England, in the case of \u00d3l\u00e1fr Haraldsson.18 The close association of the Norwegian dynasty with the Clement church in Trondheim is absolutely confirmed by its choice as the burial place of both \u00d3l\u00e1fr Haraldsson (d. 1030) and his son Magn\u00fas (d. 1047).19 In this respect it is impor- tant to note that Prince Vladimir and his wife were buried in 1015 in the Tithe Church in Kyiv, which Thietmar of Merseburg says was the church of Christ\u2019s martyr, Pope Clement, where Clement\u2019s head was enshrined. This head relic was an exceedingly important symbol of the Kyivan state\u2019s independence from both Byzantium and Rome, and it was used later (in 1147) for the consecration of the metropolitan of Kyiv. On this occasion, with the head representing the required authority, Prince Izjaslav wanted the Kyivan bishops, rather than the patriarch of Constantinople, to officiate.20 Whether there were any relics in the church of St. Clement in Trondheim is unknown, but the parallel use of this saint\u2019s church as a royal burial place is most probably modeled on the Kyivan example. Map 20.3: Map of Kyiv (Kiev) in the Tenth and Eleventh Centuries. 18 See Garipzanov, \u201cThe Journey of St. Clement\u2019s Cult,\u201d 372: \u201cit is possible that the dissemination of this cult in early Christian Scandinavia was influenced by different regions simultaneously.\u201d 19 Heimskringla II: \u00d3l\u00e1fs Saga Helga, ed. Bjarni A\u00f0albjarnarson, 270 (chap. 244). Heimskringla III, ed. Bjarni A\u00f0albjarnarson, 107 (chap. 30). 20 Crawford, Churches dedicated to St. Clement in Medieval England, 35.","438 Barbara E. Crawford Furthermore, the sarcophagi of Vladimir and his wife were said by Thietmar to be \u201cpalam stantibus\u201d (displayed openly) in the Tithe Church. It has been suggested that this situation can be compared with the 1031 translation of \u00d3l\u00e1fr Haraldsson in Trondheim after his death at the battle of Stiklestad in 1030: \u201cVar \u00fe\u00e1 l\u00edkami ko- nungs borinn inn \u00ed Clemenskirkju ok veittr umb\u00fana\u00f0r yfir h\u00e1altari. Var kistan sveip\u00f0 pelli ok tjaldat allt gu\u00f0vefjum\u201d (Then the king\u2019s body was carried in into Clemenskirkja and set up before the high altar. The coffin was wrapped in precious cloth and all hung with velvet).21 The bodies of both rulers, Vladimir and St. \u00d3l\u00e1fr, were apparently made accessible and visible to visitors.22 The similar patterns of display for both revered, deceased rulers may suggest a connection, indicating some knowledge in Norway of the ritual observed in Kyiv. The church of St. Clement in this city was the burial place of Prince Vladimir and his wife, and the cult of St. Clement was a powerful adjunct to the prestige and status of their princely house. It is worth noting that Cnut\u2019s half-English son Sveinn (for Swe- gen) was in authority in Norway at the time, and that both he and his mother, \u00c6lfgifu (or Alfiva), were present at the translation ceremony of St. \u00d3l\u00e1fr\u2019s re- mains to St. Clement\u2019s Church in Trondheim.23 The St. Clement Churches in Trondheim and Oslo The location of the churches dedicated to St. Clement in northern Europe is usually an important indicator of their significance. The location of such a church in a power centre, or close to the ruler\u2019s residence, is a strong indication that it was founded by the ruler, who used the cult of the papal martyr as an adjunct to his own power.24 If the church is at the heart of the commercial quarter, we can assume that it was the church of sailors or merchant-traders. In many cases such locations are close to harbours or rivers, which were the main highways of the period.25 The former location of St. Clement\u2019s, Trondheim, is therefore significant. Al- though the church no longer exists, the site is known to have been close to the royal estate which was established at the trading center founded on the banks of 21 Heimskringla II: \u00d3l\u00e1fs Saga Helga, ed. Bjarni A\u00f0albjarnarson, 270 (chap. 244). Heimskringla: \u00d3l\u00e1fr Haraldsson, trans. Finlay and Faulkes, II, 270. 22 Garipzanov, \u201cThe Journey of St. Clements Cult,\u201d 373. 23 Haki Antonsson, \u201cThe Cult of St. Olaf,\u201d 143. 24 Cinthio called these \u201croyal estate churches,\u201d in \u201cThe Churches of St. Clemens,\u201d 111. 25 This is a feature of the later development of the cult, according to the subjective choice of a protector-saint. See Cinthio, who follows German \u201cpatrozinien-research,\u201d in \u201cThe Churches of St. Clemens,\u201d 111.","Chapter 20 Patron Saint of Cnut 439 Map 20.4: Map of Trondheim, Norway, showing the location of St. Clement\u2019s Church (KLEMENS KIRKE) close by the royal residence and warf (Kongs Gaard og Brygge) at the time of Olafr Haraldsson (1016\u201328). Based on Blom (1956), 228. the River Nid. This river leads out into the seaways of the inner sailing route along the Norwegian coast.26 St. Clement\u2019s was the very first church in \u00d3l\u00e1fr Tryggvason\u2019s newly founded trading centre, and the saint\u2019s protective powers for those sailing on the northern European seas certainly became an important basis of the cult. Pro- tecting the royal dynasty, however, was probably the prime purpose behind the transmission of the cult and the founding of this church. Its location near the royal residence clearly indicates the importance of the cult to the ruling dynasty. Another important early Clement church in Norway was in the medieval town of Oslo. This trading center in the south-eastern part of the country is be- lieved to have been founded by \u00d3l\u00e1fr Haraldsson\u2019s younger half-brother, King Haraldr har\u00f0r\u00e1\u00f0i (\u201charsh ruler\u201d) Sigur\u00f0arson (r. 1046\u20131066), in ca. 1050. The surviving remains of the twelfth-century stone church are still very impressive, and also impressive is the archaeological evidence for a burial ground around St. Clements, which has been dated to the early eleventh century.27 These earliest burials are on a different alignment from those of the twelfth century and were perhaps associated with a timber church which predated the stone one. As finds, the burials and the traces of the timber church are very significant for the earliest urban settlement in the locality, and they are 26 Lunde, Trondheims Fortid, fig. 136. Recent excavations have uncovered the remains of four churches on the site, the earliest of which dates back to the late tenth or early eleventh cen- tury. See: www.niku.no\/projeskter\/Klemenskirken\/. 27 Recent calibration of the skeletal material has given a more precise estimation of the burial dates to 995\u20131000\/1028. See Nordeide and Gulliksen, \u201cFirst Generation Christians,\u201d 22\u201323.","440 Barbara E. Crawford evidence for the earliest Christianization of the population in the district round Oslof- jord, the fjord which connects southern Norway with Denmark. However, this Clement church is 150 m away from the royal residence at Kongsg\u00e5rden in Oslo, so its associa- tion with the political authority is not as clearly suggested by physical proximity as is the case in Trondheim. One suggestion is that the church might have been founded by Map 20.5: Site of St. Clements in Gamlebyen, Oslo, ca. 1300.","Chapter 20 Patron Saint of Cnut 441 King Haraldr Bluetooth of the neighboring Danish kingdom to the south.28 He had brought Viken under his control in the 970s and 980s and is also said to have imposed Christianity. This suggestion is not completely implausible, since Haraldr\u2019s was the first generation of Danish rulers to be converted and to impose the new religion in the Dan- ish kingdom. On the other hand, the revised dates established by carbon-14 analysis of skeletal material do not indicate that this was a church founded by Haraldr Bluetooth. Nor is it likely in any case that Haraldr Bluetooth would have dedicated his foundation church to St. Clement. As we will see, there is no evidence that any church dedicated to Clement was in existence in Denmark until the 1030s. The early Danish kings are unlikely to have been influenced by the cult in Kyiv, where they never ventured. It is more likely that the St. Clement\u2019s church in Oslo was another early foundation by one of the Norwegian kings, albeit by one unknown, since the archaeological evidence in- dicates no more than an approximate date of its foundation in the very late tenth or early eleventh century. St. Clement\u2019s Cult in Denmark The Danish kingdom consisted of a cluster of islands on the eastern side of the North Sea and the peninsula of Sk\u00e5ne, controlling access to the Baltic. As the map shows, the concentration of churches dedicated to St. Clement in Denmark is quite dense (at least twenty-six churches, from Slesvig east to the province of Sk\u00e5ne in modern-day Sweden), whereas there are only five such churches, and one chapel, in Norway, and only one possible medieval example in Sweden. This list does not include the island of Gotland, which was an independent commonwealth until taken by the Danes in the twelfth century, which is proba- bly when the Clement church in Visby was founded as a merchants\u2019 church close to the harbour. The disparity between Denmark and rest of Scandinavia in numbers of St. Clement churches requires an explanation. It shows that the cult spread to Denmark (from Norway or England or both) and took firm root there in the early Danish urban settlements where Clement became a popular saint. Urban development was a marked feature of Denmark in the eleventh and twelfth centuries, whereas there was little in Norway apart from in Trond- heim and Oslo, at least before Bergen was founded at the end of the eleventh century. Sweden came late in the history of Christianization in the north, and conversion proceeded slowly. By the time churches were being built there, the 28 Schia, Oslo innerst i Viken. Crawford, \u201cChurches to St. Clement in Norway,\u201d 109.","442 Barbara E. Crawford cult of Clement had passed the peak of its popularity and St. Nicholas of Myra was taking over the saintly role of the seamen\u2019s protector. Map 20.6: Early urban centres in Denmark with Clement churches. Churches Dedicated to St. Clement in Denmark The main problem in associating the churches dedicated to St. Clement with ei- ther a political or a mercantile initiative is that we can rarely establish the date at which these churches were founded. For our study of the cult of Clement in northern Europe, Denmark is very important, for the number of dedications is striking, as noted above. There is, fortunately, one church in Denmark for","Chapter 20 Patron Saint of Cnut 443 which the date of foundation can be established by evidence, and that is St. J\u00f8rgensbjerg church in Roskilde (on Sj\u00e6lland or Zealand), originally dedicated to St. Clement. Its date indicates an association with the period of Cnut\u2019s rule. At this point, we should provide some political background to the relation- ship of the kingdoms of Denmark, Norway, and England in the early eleventh century, as the spread of the cult of Clement can perhaps be seen to be linked with the political situation of the time. The most notable feature of the period was the success of Cnut in conquering England in 1015\u20131016, followed some years later by his conquest of the kingdom of Norway in 1028. This made him the most powerful ruler of a North Sea \u201cEmpire,\u201d and the fostering of the cult of Clement in his homeland was probably an adjunct to Cnut\u2019s \u201cimperial\u201d author- ity. Denmark was a newly converted Christian kingdom, and the cults of the established saints of the Christian church may have been introduced by the clerical advisers whom the king brought with him from England, which he ruled as a model Christian king after his conquest in 1016. We certainly know that some other saints were introduced from England.29 It is likely that St. Clement was regarded as a desirable protector for the newly converted Danish dynasty. There are three aspects to his cult which help to provide an explanation for his popularity with Cnut. The first was the inesti- mable value of papal authority for a newly Christianized dynasty seeking to en- hance its status in the wider world of Christian kings and princes. The transfer of Clement\u2019s relics to Rome in the late ninth century brought his physical pres- ence to western Europe, with the result that his cult was widely publicized. He was the only pontiff, with Peter, who could be claimed as apostolic, with the added enhancement of martyrdom. Second, we know that Cnut had imperial pretensions, particularly after his participation in the coronation of Emperor Conrad in Rome in 1027, when he also visited Pope John XIX. As a result, the northern king achieved a new, elevated status in the world of Christian rulers, stimulating his desire to adopt a presti- gious saint as protector of his dynasty. Unfortunately, we do not know whether Cnut was given, or whether he acquired, any relics of Clement during this visit, although he does write in his letter to the English people that he was honored \u201cdonis pretiosis\u201d (with precious gifts) from the pope and the emperor.30 It is fairly certain that the church of San Clemente, where the relics were enshrined, would have been one of the sacred places he visited. Once he had returned north, the 29 Crawford, Churches dedicated to St. Clement in Medieval England, 15, 25, 57. See also Haug in this volume, pp. 383\u201392. 30 Gesetze der Angelsachsen, ed. Liebermann, I, 276 (Cn 1027, 5). EHD 1, no. 53.","444 Barbara E. Crawford adoption of this papal martyr as an especially prestigious patron would fit well with the enhanced conception of his authority which certain factors indicate he promoted.31 The third aspect of the cult is the element of competition that may have played a part in the adoption of Clement\u2019s cult by the Danish dynasty as a result of the enmity between Kings Cnut and \u00d3l\u00e1fr Haraldsson of Norway. As we have seen, the cult had already been established, or reestablished, by this \u00d3l\u00e1fr after \u00d3l\u00e1fr Trygg- vason\u2019s initial foundation of the Clement church in Trondheim. It would be very difficult to reject the saga evidence for the founding of the church in Trondheim by these kings, just as there is no doubting the founding of churches dedicated to Clement in many more places in Denmark, one of which, at Roskilde, dates from the reign of Cnut. Yet the two dynasties were rivals, and \u00d3l\u00e1fr Haraldsson was de- feated and killed at the battle of Stiklestad by Danish forces and disaffected Norwe- gians. Cnut presented himself successfully as a triumphant alternative to the Norwegian dynasty, and his son Sveinn (along with \u00c6lfgifu, Sveinn\u2019s English mother) was given power in Norway. So how did it happen that Clement, the favor- ite saint of the Norwegian kings (the two \u00d3l\u00e1frs), was adopted as a saint by the rival Danes? The answer may lie in \u201ccontesting patronage,\u201d the phenomenon of competing factions each striving to control a cult for its own purpose.32 It has been shown that the Anglo-Danish dynasty actually took up and promoted the cult of the martyred \u00d3l\u00e1fr Haraldsson, despite the fact that he was killed in battle against Cnut\u2019s Danish forces in Norway in 1030.33 The same phenomenon may be at work in the adoption of the cult of the Norwegian kings\u2019 protective saint by their ene- mies and conquerors. This phenomenon could explain the apparently contradictory situation in which the dominant political power does not suppress the cult of a saint which had been promoted by its victim, but actually takes over as the promoter of that cult. Danish success in the contest with \u00d3l\u00e1fr Haraldsson may have been taken as an indication that Clement had transferred his support from the Norwegian to the Danish dynasty, so that he could justifiably be regarded as the protector saint of the Danes. The date of the founding of the church in Roskilde, 1030\u20131035, 31 Bolton, Empire of Cnut the Great, 296\u2013303. 32 This explanation of the circumstances of the adoption of Clement\u2019s cult in Denmark is ex- plored in Crawford, \u201cThe Cult of Clement in Denmark,\u201d 274, and Churches to St. Clement in Medieval England, 25\u201328. 33 Townend, in \u201cLike Father, like Son?,\u201d 476, also demonstrates how \u201cone of Cnut\u2019s strategies in the establishment of his rule in England was to patronize the cults of English saints.\u201d In \u201cKn\u00fatr and the Cult of St. \u00d3l\u00e1fr,\u201d 264, he shows how competing factions each try to control a cult \u201cfor their own political purposes.\u201d","Chapter 20 Patron Saint of Cnut 445 corresponds so closely with the date of Olaf\u2019s defeat at Stiklestad, and with the Danish victory over the Norwegians, that it is tempting to propose a causal link between the victory in battle and the foundation of St. Clement\u2019s church in Ros- kilde. This place was the seat of power of Danish kings, according to Adam of Bremen, who calls it \u201csedes regum Danorum\u201d (seat of the kings of the Danes).34 Map 20.7: Map of Roskilde Showing Site of St. Clement\u2019s Church. 34 Gesta Hammburgensis, ed. Schmeidler, 50 (I.l). See also North in this volume, p. 282.","446 Barbara E. Crawford Nonetheless, the location of the church in Roskilde at a distance from the royal estate (in contrast to the Clement church in Trondheim, which was close to the royal estate in that city) suggests that this church may have been founded primarily for the retinues of the Anglo-Danish king, and moreover, close to the harbor where the royal fleet would be berthed. There were many groups of pro- fessional warriors, minters, craftsmen, and ecclesiastical personnel, who were accustomed to accompanying their king to and fro across the North Sea in the period of combined rule of his three kingdoms, and who would need their own place of worship. Perhaps St. Clement\u2019s in Roskilde was founded with their inter- ests in mind. Certainly, the style of architecture and the stone construction of this church point to English influence.35 Whatever the purpose of the founding of this church, there are many reasons to see it as a royal foundation reflecting the political and military contacts across the North Sea, which were such an impor- tant part of Cnut\u2019s maintenance of power at this time. Figure 20.2: View of St. Clement\u2019s from Roskilde Harbour, and North Doorway. Maritime Links: Churches in England The geographical situation of all these politically important Clement churches in Nor- way and Denmark both points to the maritime lifestyles of the powerful and indicates how important the sea routes were in the maintenance of power. All the urban churches in Scandinavia are located by harbours, with the exception of Lund, in Sk\u00e5ne, which, although an important power center in this period, is found in an in- land location. Clement\u2019s proximity to sea-routes raises the important issue of his 35 Crawford, \u201cThe Cult of Clement in Denmark,\u201d 249.","Chapter 20 Patron Saint of Cnut 447 Map 20.8: Urban churches dedicated to St. Clement in England.","448 Barbara E. Crawford association with seafarers, for whom his protective powers became the most enduring aspect of his cult in northern Europe. This may be one reason why his cult became so popular in Denmark, whose political and commercial life revolved around sea voy- ages. If we turn to England, it is notable that most of the urban churches dedicated to Clement in that country are also in locations close to sea routes and waterways. There is no doubt that Clement\u2019s cult was already well established in England by the eighth century, as his feast is included in the Martyrology of Bede (ca. 673\u2013735), with details about his martyrdom and the discovery of his body in a stone coffin, with his anchor nearby, when the sea receded.36 The date of his feast is listed in late eighth- and ninth-century calendars, which mention miraculous legends associated with his burial place, and the miracle of the fountain of spring water which Clement caused to gush forth from a rock for the benefit of the workers in the Cherson marble quarries. The Old English Martyrology, of a late ninth- century date, gives an account of his martyrdom under the date of his feast day (November 23), as well as details of the miracle of the child who was trapped at the saint\u2019s tomb when the sea flooded around the island, but who was restored alive to his mother the next year when the waters receded.37 A verse Menologium of ca. 1000 gives an explicit reference to Clement\u2019s death by drowning, implying that he provided protection from drowning for those who prayed to him.38 It is noticeable that the preponderance of the Clement dedications in England, no fewer than thirty-four, were located north of the River Thames, and in close asso- ciation with the area settled by the Danish immigrants in the ninth and tenth centu- ries. Unfortunately, it is very difficult indeed to pin down the date of the foundation of most of these churches. In some cases, notwithstanding, there is an explicit link with the Danes, such as the name \u201cSt. Clement Danes\u201d (\u201cecclesia Dacorum\u201d) in Lon- don, a term which can be traced back to the twelfth century. There is also a remark- able association with the major boroughs (or burhs, Anglo-Saxon urban foundations) in the Danelaw. Every one of the major Danelaw boroughs has a St. Clement church, which is often in close association with waterways, either harbours, rivers, or estuar- ies, and sometimes located close by a major bridge or river crossing. This association tells us that these churches were founded by, and for, those who voyaged overseas. Since these churches are all on the east-coast rivers, or on the east coast, it also tells us that those journeys would have been across the North Sea, above all to Denmark and Norway. On the one hand, it is not possible to define whether these churches date to the period of Danish settlement of the first Viking Age in the ninth and tenth 36 Crawford, Churches dedicated to St. Clement in Medieval England, 38\u201339. 37 Old English Martyrology, ed. and trans. Rauer, 219. 38 Lapidge, \u201cThe Saintly Life,\u201d 249.","Chapter 20 Patron Saint of Cnut 449 centuries, or to the period of Cnut\u2019s conquest of England and the creation of the Anglo-Danish North-Sea empire in the early eleventh century. Only excavation will prove the date of the foundation of these churches; so far there have been no excava- tions conducted at any of the English sites of Clement churches. On the other hand, the Scandinavian association with St. Clement is strong. It is likely not only that some of his churches at least date to the period before Cnut\u2019s conquest, but also that his reign increased the popularity of this saint whose special property was the pro- tection of seafarers from the danger of drowning. One striking feature of the distribution of churches dedicated to Clement in the British Isles is that there are very few in Scotland, and none in Wales or Ire- land, apart from one in Dublin. Other examples exist in the south and west of England, but of the overall number of his churches in England, which stands at fifty-two, as many as thirty-four were founded or dedicated north of the River Thames and predominantly in the southern Danelaw region. The absence of churches in the Celtic countries certainly suggests that Clement\u2019s cult was not deeply embedded in these communities (although his name appears in some of the early Irish martyrologies).39 This disparity in the location of churches in the different parts of Britain is reflected in the Scandinavian distribution (already referred to), where the churches are predominantly located in Denmark (or for- mer Danish territory), but sparse in Norway and almost absent from medieval Sweden. The link with the Danes on both sides of the North Sea is striking. In general terms, one can be sure that the cult existed in England before the Vi- king Age and was taken up with enthusiasm by the Danes once they had been converted. One can likewise be sure that this was the period when St. Clement\u2019s cult spread among the urban communities of eastern England, because of the close contact with Denmark across the North Sea. As already discussed, the at- traction of the cult of St. Clement to Cnut\u2019s dynasty suggests that there was a political element in the choice of this papal martyr as a saint worth cultivating \u2013 and worth taking over from the defeated Norwegians. In the eleventh century, when connections across the North Sea between the three kingdoms of England, Denmark, and Norway were constant, and the knowl- edge of Clement\u2019s significance was transferred with the political regimes, it seems that this saint\u2019s effectiveness extended to protection for those who voyaged across the seas. Eventually this became the most significant aspect of the cult for the popu- lations who lived in coastal communities, whose circumstances exposed them to the danger of drowning. Exactly when and where this aspect of the cult developed is an issue which needs further exploration, for it does not appear to be a feature of 39 Crawford, Churches dedicated to St. Clement in Medieval England, 43\u201344.","450 Barbara E. Crawford Clement\u2019s cult in the principality of Kyiv. Nonetheless, it is true of Scandinavia. Some of the Clement churches in Denmark were closely associated with royal estates and their significance was particularly political.40 Cinthio argued that if they were also near the harbours, this was because of the lifestyle of the politically important. Once the Anglo-Danish world of Cnut\u2019s dynasty collapsed and England and Den- mark went in separate political directions, the churches which had possibly been founded in strategic situations for the sea voyages of the rulers and their retinues became more closely associated with the merchant communities and seafarers in general. Clement\u2019s popularity remained high among the coastal urban communities and spread into rural areas in both England and Denmark. St. Clement Danes and Other Churches in England In England, as already noted, it is very difficult to know when the Clement churches were founded, and particularly so to know whether they date from before the Nor- man Conquest or after it.41 However, with regard to the famous church of St. Clement Danes on the Strand in London, we have some evidence that it did indeed exist in the time of Anglo-Danish rule. Located west of the City of London on the main route to Westminster, the strategic position of this church may be an indication of its mili- tary significance. The early association with the Danish community is quite clear, as we have seen, from the name \u201cecclesia Dacorum\u201d (church of the Danes) recorded in the twelfth century. There is also the interesting story that it became the burial place of Harold Harefoot, second son of Cnut by \u00c6lfgifu, who took power in England on his father\u2019s death in 1035. Harold reigned for only five years; after his death and burial at Westminster, according to John of Worcester, his half-brother Harthacnut had his body disinterred and thrown into the River Thames, from where it was recov- ered and taken for burial to the Danes\u2019 cemetery.42 This record of the burial of Harold Harefoot, in what is presumed to be the burial ground of St. Clement\u2019s on the Strand, is rather important for our understanding of the saint\u2019s association with the Anglo- Danish dynasty. 40 Cinthio (\u201cThe Churches of St. Clemens\u201d) examines the location of churches dedicated to St. Clement in Scandinavia and the political significance of their location close to royal power centres in many instances. 41 Clement was well known to the Normans, and a chapel in Rouen, close to the ducal Don- jon, on the River Seine can be dated certainly no later than 1025 and possibly as early as 1006. See Crawford, Churches dedicated to St. Clement in Medieval England, 194\u201395. 42 EHD 1, 317. The Danes\u2019 burial ground was said by Ralph of Diceto in the late twelfth century to be \u201capud Sanctum Clementem\u201d (by St. Clement\u2019s). See Works of Ralph de Diceto, ed. Stubbs, I, 185 (Abbreviationes chronicorum, s.a. 1040).","Chapter 20 Patron Saint of Cnut 451 The possible military or naval connection needs to be explored further when seeking to understand the significance of Clement\u2019s association with the Danish dynasty. This raises the issue of \u201cgarrison churches\u201d which, it has been postulated, were connected with the churches dedicated to Scandinavian saints in London, and particularly with churches dedicated to St. Olave or \u00d3l\u00e1fr Har- aldsson.43 Of course, these churches must postdate 1030, when \u00d3l\u00e1fr died at the battle of Stiklasta\u00f0ir (Stiklestad), after which his fame as a martyr spread quickly, resulting in the remarkable number of six churches dedicated to him in London. These churches are linked to strategic positions near the walls and gates, as well as the waterfront. In addition, Pamela Nightingale suggested that St. Clement Danes and St. Bride\u2019s, on the road to Westminster and near the River Thames, may also have been associated with encampments of troops close to their ships, in the years after Cnut\u2019s conquest when London was a city under military occupation. There is no doubt that London was used as an im- portant naval base during the reign of \u00c6thelred II, and also by the Danish kings who continued to use London as the chief base for their hired fleet of Scandinavian mercenaries, the li\u00f0smenn (lithsmen) paid for by the heavy gelds which were exacted.44 It was a mobile, naval force, originating in the forty ships\u2019 crews which Cnut retained in 1017 as a force for conquering the rest of England and securing his authority over his North Sea empire. Such a force would need its own base, and a band of li\u00f0smenn was based in London; \u201c\u00fea li\u00f0smen on Lunden\u201d (the men of the fleet in London) are said to have supported Harold Harefoot to be Cnut\u2019s successor in 1035.45 The association of these troops with St. Clement\u2019s is evident from the report that Harald\u2019s body, having been recovered from the Thames in 1040, was taken to the Danes and buried in their cemetery. These would have been Danes who had been his supporters in the succession crisis of 1035, and who can therefore be seen to have had St. Clement\u2019s church as their religious centre. This is circumstantial evidence for St. Clement Danes having functioned as a \u201cgarrison church\u201d for the period of the Danish rule, until the naval force was dismissed by Edward the Confessor in 1049\u20131050 and the heregeld (army-tax) abolished in 1051. For more than thirty years this naval force would have needed a riverine base in London for 43 Crawford, in Churches dedicated to St. Clement in Medieval England, 60\u201361, discusses Nightin- gale\u2019s theories of garrison churches in London in the period of Danish rule, in Nightingale, \u201cOrigin of the Court of Husting,\u201d 578. Recently the garrison theory has been reconsidered in connection with the churches and chapels dedicated to St. Clement and St. Olave in South Conesford, Nor- wich. See Shelley, \u201cSouth Conesford, Norwich.\u201d See also Reynolds in this volume, pp. 53\u201355, 61. 44 Crawford, Churches dedicated to St. Clement in Medieval England, 64\u201365. 45 ASC (E), ed. Irvine, 76 (s.a. 1036 [for 1035]).","452 Barbara E. Crawford the security of its fleet, along with its own church. The location of St. Clement\u2019s church on the Strand provided them with an anchoring place and a supply base, as well as access to the interior of England via the Thames, especially to Oxford, which was an important meeting-place for the regulation of the con- quered country. There was also a St. Clement\u2019s Church in Oxford, one located by the River Cherwell, which flows into the Thames. The important assemblies which were held in Oxford in 1015, 1018, and 1035 would have been attended by the li\u00f0smenn, particularly in 1035 when the succession dispute was settled, and when Harold Harefoot was elected as \u201cregent\u201d by all the councilors and the li\u00f0smenn of London. It is not beyond the bounds of possibility that these li\u00f0s- menn also had their own \u201cgarrison church\u201d as part of their naval base on the river in Oxford; if so, that church might be identified with St. Clement\u2019s.46 There are many other St. Clement churches in England, located in impor- tant maritime or riverine locations suitably placed for naval establishments, such as the ports on the south coast, or in Kent, like Rochester and Sandwich. The latter port (the battle station for the Anglo-Saxon fleet) was where Edward the Confessor attended mass in St. Clement\u2019s Church when reviewing the fleet in 1042.47 London, Oxford, and Sandwich are the most likely locations for \u201cgar- rison churches\u201d and the founding of a St. Clement church in all three ports points to the possible role of these churches in the military strategies of both the Anglo-Saxon and the Anglo-Danish regimes.48 There are other important strategic urban locations in East Anglia, like Norwich, where St. Olave and St. Clement churches coexist on the River Wensum. There were two St. Clement churches in Norwich, as also in Lincoln. 46 However, the earliest historical evidence for the existence of the St. Clement\u2019s church in Oxford dates from the early twelfth century. Archaeological material recovered on an island in the river in the late nineteenth century, during dredging works on the River Cherwell, includes metal objects with cavalry connections which date from the late tenth or early eleventh cen- tury and hint at a military establishment in this location or possibly at a late pagan equestrian burial. See Blair and Crawford, \u201cA Late-Viking Burial at Oxford?,\u201d 135\u201342, and Crawford, Churches dedicated to St. Clement in Medieval England, 88\u201389. 47 Crawford, Churches dedicated to St. Clement in Medieval England, 72: Gatch, \u201cMiracles in Architectural Settings,\u201d 229. The recent study of Sandwich by Clarke, Sweetinburgh, and Jones (Sandwich, 25\u201329) discusses the significance of St. Clement\u2019s Church in the early development of the port (late tenth or early eleventh century). 48 The possibility that \u00c6thelred had been the founder of some of the St. Clement churches, as part of his naval strategy, is suggested in Crawford, Churches dedicated to St. Clement in Medie- val England, 74, 205.","Chapter 20 Patron Saint of Cnut 453 Cnut\u2019s Church of St. Clement\u2019s in Roskilde and \u201cthe Machinery of Control\u201d It has already been noted that St. J\u00f8rgensbjerg church in Roskilde, originally dedi- cated to St. Clement, was located by the harbor in a way that suggests its function as the church founded for the use of seafarers, or for members of the royal Anglo- Danish retinue. These would be constantly voyaging between Denmark and Eng- land from the period of Cnut\u2019s invasion and conquest of England through the reigns of his sons, until the demise of the dynasty in 1042. There were also soldiers, crafts- men, and clergy, as we have seen, in whose interests St. Clement\u2019s in Roskilde may have been founded: the style of architecture and the stone construction of St. Clem- ent\u2019s points to English influence, even if the church was not constructed by an Anglo-Saxon master builder and his masons.49 Its stone construction indicates \u201cthat it was prestigious, and had extremely wealthy patrons.\u201d50 One could extrapolate from this assessment to make similar comments about the dedication to Clement. Whatever the purpose of the founding of this church, there are many reasons to see it as a royal foundation reflecting the political and military contacts across the North Sea which were such an important part of Cnut\u2019s maintenance of his power at this time. Cnut\u2019s ships\u2019 crews formed a mobile fleet which could be deployed any- where in his maritime empire, and they would have needed bases in both of his kingdoms. The church of St. Clement\u2019s could have been founded in the Danish half of his empire with their interests in mind. Clement\u2019s powers of protection for those engaged in frequently traversing the North Sea would be a very relevant consider- ation in the choice of patron saint for this church, located at the dynasty\u2019s Danish power base. Fortunately, we are able to pinpoint its foundation date closely, as a hoard of coins was discovered during excavations in the 1950s, located in the founda- tions of the tower. It was probably deposited as a foundation offering and can be assigned to the years 1030\u20131035.51 This date links the church\u2019s foundation with the reign of Cnut the Great, king of Denmark, England, and Norway in those years. We can less confidently assign the term \u201cgarrison church\u201d to this St. Clement\u2019s foundation, considering that it was constructed close to the dy- nastic power base in Cnut\u2019s home country. But we can probably regard it as one element of Cnut\u2019s \u201cmachinery of control\u201d which, it has been suggested, was put 49 Crawford, \u201cThe Cult of Clement in Denmark,\u201d 249, citing Andersen and Nielsen, \u201cEn Stor- mannsg\u00e5rd\u201d; Roesdahl, Viking Age Denmark; and Ulriksen, \u201cSct J\u00f8rgensbjerg Kirke.\u201d 50 Bolton, Empire of Cnut the Great, 171. 51 For details and references, see Crawford, \u201cThe Cult of Clement in Denmark,\u201d 248.","454 Barbara E. Crawford into operation by the Anglo-Danish king in western and central Denmark.52 This would be the date of the incorporation of Clement as a saintly protector of the dynasty and its naval storm-troopers following the defeat of \u00d3l\u00e1fr Haralds- son and the conquest of Norway in 1030. As has already been elaborated, it seems logical to see the adoption of Clement at Roskilde at this date as a result of contesting the patronage of this prestigious papal martyr with the defeated Norwegian royal dynasty which had imported the cult from Kyiv, and which had already built churches dedicated to him in their own urban power centres. Such a politically ambitious move may have been encouraged, if not initiated, by the English churchmen who were active in Cnut\u2019s power structures. With re- gard to Roskilde, Gerbrand, consecrated bishop of Roskilde before 1022, was one of these powerful ecclesiastics who operated on both sides of the North Sea.53 The English churchmen would have been familiar with Clement\u2019s role as protector of seamen, as his cult was long-established in England. In conclusion, the evidence for Roskilde\u2019s importance in Cnut\u2019s political con- trol of his own kingdom is undoubted, and the role of Clement as protector at the harbor church in this royal power center is surely significant. We can still see the situation as comparable in some way with that of St. Clement\u2019s in the Strand in London.54 The latter church was, of course, in a country governed by an alien dy- nasty which needed to maintain its position with military and naval forces. The use of a somewhat anachronistic term like \u201cgarrison chapel\u201d to define the role of these churches is unnecessary, and in the case of Roskilde probably inappropriate, but it does focus attention on their possible function in Cnut\u2019s maritime empire and in his \u201cmachinery of control.\u201d Disparate though the bits of evidence are, it cannot be denied that the evidence for the burial of Cnut\u2019s son Harold Harefoot in St. Clement\u2019s in the Strand provides support for the hypothesis that this Clement church was significant among the Anglo-Danish dynasty\u2019s ritual locations. It sug- gests that St. Clement may have had a role as the skytshelgen, the \u201cprotective saint\u201d of members of the dynasty and its entourages, military and administrative. 52 Bolton, Empire of Cnut the Great, chap. 7. 53 Bolton, Empire of Cnut the Great, 178. 54 Crawford, Churches dedicated to St. Clement in Medieval England, 67.","Chapter 20 Patron Saint of Cnut 455 General Popularity of the Cult in the Twelfth Century From this high point in the eleventh century, when the cult of St. Clement was pa- tronized by the powerful elites in Norway, England, and Denmark, we move for- ward into an era when the saint became popular with the wider population, for his churches are found in rural areas in eastern and central England and throughout Denmark. In many cases it is apparent that his popularity spread out from a partic- ular urban power center with a Clement church into the surrounding countryside. The continuing close link with waterways, fishing communities, and locations prone to flooding gives a strong impression that it was Clement\u2019s protective powers where the danger of drowning was a concern that made him so popular. The sym- bol of his death by drowning \u2013 the anchor \u2013 which was the means by which he gained a martyr\u2019s crown, is a well-established attribute in northern Europe. The anchor\u2019s firm association with safety for mariners in danger of drowning tells us clearly that Clement was regarded as the patron saint of seafarers.55 Figure 20.3: Altar panel from Skj\u00e6rv\u00f8y, Troms, Norway (after 1500), showing St. Clement with his papal tiara, holding papal cross and anchor (Oslo Universitets Oldsaksamling). 55 Crawford, Churches dedicated to St. Clement in Medieval England, 52\u201353.","456 Barbara E. Crawford Another aspect of Clement\u2019s cult associates his name with wells or springs. This reflects the story in the Acta about the miracle when he caused a fountain or spring to pour out of a rock to quench the thirst of his converts in the Crimea.56 Other associations developed in England do not seem to have a direct connection with any of the legends; these are with blacksmiths, wiremongers, and iron- founders, which have lasted up to recent times. It is possible that the association with ironworking may reflect the blacksmith\u2019s craft and the production of the metal anchor which was Clement\u2019s attribute. The company of ironfounders had Clement as their patron. Blacksmiths, in particular, held celebrations on St. Clement\u2019s Day, when they \u201cfired\u201d their anvils with gunpowder, and it may be this practice which led to his being considered a protector from thunderstorms in Germany. In this way, the cult of St. Clement was many-sided, and we have to see him as a saint who met the need of different sectors of medieval society at different times. But he particularly supplied the aspiring newly-Christianized rul- ers of Norway and Denmark with valuable strength and support. Cnut and his sons were undoubtedly aware of this aspect of the cult. Conclusion: Patron of an Empire Can it be concluded that Clement was also the skytshelgen (protecting saint) of Cnut and his dynasty? In a general sense, probably yes, in that this apostolic martyr provided powerful support for dynasties engaged in power struggles, whose especial concern was for those involved in sea travel. Erik Cinthio\u2019s study of the cult of Cnut shows that some of the Danish churches can be shown to date to the reign of Cnut, and that their specific location in the towns some- times indicates an association with royal estate. Accepting Cinthio\u2019s conclusion about the \u201cpolitical significance\u201d of the cult does not mean, however, that we should downplay Clement\u2019s effectiveness as a protector for all members of these societies who had maritime lifestyles exposing them to danger in their watery worlds.57 From being associated with the elites of the newly converted Scandinavian world in Norway and Denmark, Clement\u2019s cult spread to a wider section of the population in Denmark and in England, but did not spread so far in Norway. The 56 Crawford, Churches dedicated to St. Clement in Medieval England, 54\u201355. 57 Cinthio was seemingly unaware of the significance of the position of Clement churches in locations prone to flooding. Hofmann, Die Legende von Sankt Clemens, 171, mentions this geo- graphically important aspect of Clement\u2019s role as protector.","Chapter 20 Patron Saint of Cnut 457 saint\u2019s long-standing association with coastal communities and with seafarers meant that his churches became indelibly linked with seamen and fishermen, and their geographical location at harbours gave continuity to this association. This aspect is the one which is familiar today, while the earlier political signifi- cance of Clement\u2019s cult is little known. He was rivaled in his role as the seafarer\u2019s saint by Nicholas, whose cult rose to prominence in the twelfth century after his relics were brought to Bari in Southern Italy from Myra in 1079. Indeed, Clement was overtaken by Nicholas, who became very popular indeed \u2013 and whose cult was patronized by the kings and aristocratic members of Scandinavian society.58 Knowledge of Clement\u2019s cult has suffered from the poorly documented situ- ation regarding almost all his churches, and the lack of historical sources from the late tenth and early eleventh century. The far-reaching contacts of the Vik- ings and Varangians with the center of Clement\u2019s cult in Kyiv and the links with Cherson are only dimly recorded in the sagas, and the sagas have suffered from too much skeptical criticism in the past century. It is time that we recog- nized the value of this information and listened to what the saga-writers are telling us.59 The maritime empire of Cnut\u2019s dynasty did not last long and the important political role of Clement\u2019s cult in that Anglo-Danish world was fleet- ing. The evidence that we have suggests that his cult was an important adjunct to the power of that dynasty, difficult though this is to prove. Further research \u2013 in particular, more archaeological research \u2013 will help to firm up our knowl- edge of the significant role of the cult of St. Clement in northern Europe in the early eleventh century. 58 Crawford, Northern Earldoms, 212\u201313. A reference in the lai of Eliduc, dating from the mid- to late twelfth century, shows the equivalence of Clement and Nicholas at that time as protec- tors in stormy seas. They are both invoked during Eliduc\u2019s sea voyage with his lady Guilliadun from Totnes to France, when the ship\u2019s mast breaks and splits and the sail is completely torn. However, they reach harbor safely. See The Lais of Marie de France, trans. and ed. Burgess and Busby, 121. 59 See Bolton in the Epilogue.","","Timothy Bolton Epilogue Cnut and the Potential Uses and Abuses of the Late Narrative Sources from Northern Scandinavia That element of research on Cnut the Great which still seems to raise the great- est number of eyebrows is the use of late North Scandinavian narrative sources for parts of the history of Scandinavia during the reign of Cnut. These sources are: the synoptic Historia de antiquitate regum Norwagiensium of Theodoricus monachus (ca. 1177\u20131178 and certainly before 1188) and its Old Norse vernacu- lar sister text \u00c1grip af N\u00f3regskonungas\u01ebgum (probably soon after 1180); the later Norwegian and Icelandic sagas and saga-compilations: Fagrskinna, Mor- kinskinna (both ca. 1220), the so-called Oldest saga of St. \u00d3l\u00e1fr (ca. 1200), and its part-descendant, the Legendary Saga of St. \u00d3l\u00e1fr (early thirteenth century), as well as the surviving fragments of Styrmir K\u00e1rason\u2019s L\u00edfssaga (probably from the early decades of the thirteenth century) and lastly Heimskringla (ca. 1230) and Kn\u00fdtlinga saga (probably 1250s). Other more contemporary sources that shed light on these do exist for Cnut\u2019s reign, including his letter of 1027 and fragments of skaldic verse, but these are brief and few. The study of Cnut both as an English and a Scandinavian ruler, together with the totality of his domin- ions, forces us to consider as many of the Scandinavian sources of evidence as possible. An examination of how historians have engaged with, or avoided, late Scandinavian narratives in the last one and half centuries, reveals much, not only about the changing fortunes of these texts in that time, but also about some of the problems that followed the rejection of these sources or attempts to continue to work with them. It is the uses and abuses of this material by mod- ern historians that I shall attempt to set out here, followed by some observa- tions about how and where we might appropriately use this material in future.1 In keeping with the theme of this book, I shall restrict myself to studies relevant to Cnut. 1 Much of this discussion can be found in practical examples in Bolton, Empire of Cnut the Great, see esp. 251\u201359 and 275\u201387, and explained more fully in Bolton, Cnut the Great, 22\u201326. The present paper as presented at the conference formed the basis of what was written in the latter book. https:\/\/doi.org\/10.1515\/9781501513336-022","460 Timothy Bolton The Source-Critical Approach: its Arrival and Effects on the Uses of these Late Scandinavian Narratives In a short span of years following 1910, there was a sea-change in Scandinavian historiography and its critical approach to late narrative sources. This was initi- ated by two Swedes \u2013 the brothers and medieval historians Lauritz and Curt Weibull \u2013 closely followed by the Norwegian Halvdan Koht. Such events are well known in Scandinavian circles, but perhaps need a little rehearsing for an English readership.2 The earliest indications of the new source-critical ap- proach by one of these authors may be found in two articles by Lauritz Weibull printed side-by-side in Historisk tidskrift f\u00f6r Sk\u00e5neland (Historical Journal for Sk\u00e5ne) in 1910.3 The same author then set out his new approach in more sys- tematic detail in 1911 with his Kritiska unders\u00f6kningar i Nordens historia omkring \u00e5r 1000 (Critical Studies in Nordic History around the year 1000). This was closely followed in 1913\u20131914 by Halvdan Koht\u2019s \u201cSagaernes opfatning av vor gamle historie\u201d (The Sagas\u2019 Perception of Our Old History), which he initially gave as an address to the Norske Historiske Forening (Norwegian Historical As- sociation) on November 24, 1913 and then published in the following year. In 1915 Lauritz Weibull\u2019s brother Curt joined the charge, with his Saxo: Kritiska unders\u00f6kningar i Danmarks historia fr\u00e5n Sven Estridsens d\u00f6d till Knut VI (Critical Studies on the History of Denmark from the Death of Sven Estrithsson to Knut IV). Only one of these lengthy studies even mentioned Cnut the Great, but they were to have a great influence on his historiography in the years to come.4 Their important legacy in our subject was to bring vigorous source criticism into this region of medieval history. Primarily, they exposed the central prob- lem in using late-twelfth- and thirteenth-century narratives for the study of the 2 An excellent survey can be found in Lindqvist, \u201cEarly Political Organisation,\u201d 161\u201363. 3 These are \u201cDr\u00e5pen i Roskilde i Knut den stores och Sven Estridsens tid\u201d (The Murders in Roskilde in the Time of Cnut the Great and Sven Estrithsson) and \u201cKnut den stores sk\u00e5nska krig\u201d (Cnut the Great\u2019s Scanian War), published together in Historisk tidskrift f\u00f6r Sk\u00e5neland, 4 (1910\u20131913). I owe this reference to S\u00f8ren Balle, \u201cUlf Jarl og drabet i Roskilde\u201d (Earl \u00dalfr and the Murder in Roskilde), 35\u201336. 4 Lauritz Weibull\u2019s study stops just short of Cnut\u2019s reign, whereas Curt Weibull\u2019s is on Saxo Grammaticus, an author of the twelfth century. Much of Halvdan Koht\u2019s concern is focused on St. \u00d3l\u00e1fr as a central figure of Norwegian history, and in \u201cSagaernes Opfatning,\u201d 391, he notes that ruler was as important to the Norwegians as \u201cKarl den Store var for de tyske og de franske\u201d (Charlemagne was for the Germans and the French). Thus, Cnut, the great enemy of St. \u00d3l\u00e1fr, is mentioned briefly at 392.","Epilogue 461 late Viking Age. In Cnut\u2019s case, such sources were separated by a century or so from the events they describe, and were thus suspect in their record of events and especially so in their interpretations of those events.5 This consideration naturally leads to another: the question of where writers in the late twelfth and thirteenth centuries could have turned for the basis of their narratives, and to what extent they used pure invention to fill the gaps in such narratives.6 Notwithstanding these developments, earlier doubts about the veracity of such sources had appeared throughout the last decades of the nineteenth cen- tury and in the years leading up to 1910\u20131911. The Danish historian Kristian Er- slev, while never mentioning the eleventh century, let alone Cnut, published writings through the 1880s which contained the beginnings of a critical ap- proach to medieval sources.7 These works led to his \u201cErik Plovpennings Strid med Abel. Studier over \u00e6gte og u\u00e6gte Kilder til Danmarks Historie\u201d (Erik Plov- penning\u2019s Fight with Abel: Studies of Genuine and Illegitimate Sources of Dan- ish History) in 1890, an essay which contained a definitive statement about the value of primary sources over later materials.8 Only a few years before, in 1877, the English historian E. A. Freeman had recorded doubts about the late Scandi- navian sources.9 Of Cnut\u2019s northern wars, Freeman stated that \u201cthe Norwegian sagas and the rhetorical Latin of the Danish historian help us to abundance of detail, if only we could accept them as authentic.\u201d Attached to these words is a footnote to which he relegated a statement on a work he read through the 1844 translation of Samuel Laing, namely Heimskringla: \u201cI use it freely, though with caution, for Northern affairs.\u201d10 His relief is palpable, when, incidentally quite at odds with my own approach, he concludes: \u201cHappily, to unravel the difficul- ties and contradictions of their various statements is no part of the business of an English historian.\u201d A similarly dawning awareness of source problems appeared in King Cnut\u2019s first biography, which was written by Laurence M. Larson, a Nor- wegian-born \u00e9migr\u00e9 to America. Larson notes that Kn\u00fdtlinga saga\u2019s description of 5 For an excellent survey of the problems associated with such texts see Sawyer and Sawyer, \u201cAdam and Eve of Scandinavian History.\u201d 6 Koht himself asks openly, in \u201cSagaernes Opfatning,\u201d 391: \u201chvorledes er denne historien- opfatning opstaat? Er den fri speculation, eller har den r\u00f8tter i almene historiske vilkaar?\u201d (How did this perception of history come about? Is it free speculation, or does it have roots in general historical conditions?). 7 See his \u201cStudier til Dronning Margrethes Historie\u201d (\u201cStudies for the History of Queen Mar- grethe\u201d), for the earliest version of these ideas. 8 Erslev, \u201cErik Plovpennings Strid med Abel.\u201d 9 Freeman, History of the Norman Conquest, I, 452. 10 Freeman, History of the Norman Conquest, I, 452\u201354, as well as his notes ZZ, BBB, EEE, GGG, MMM and QQQ at the end of the volume, for examples of such use.","462 Timothy Bolton Cnut\u2019s outward appearance and abilities was \u201c[i]dealistic . . . [due to being] com- posed two centuries or more after his time\u201d and that the speeches in these ac- counts \u201care doubtless the historian\u2019s own.\u201d11 Probably due to these doubts, Larson was quick to cite skaldic verse, which he called \u201cfragments of contempo- rary verse\u201d and \u201ccourt poetry of the scalds,\u201d as supporting evidence in preference to saga-material \u2013 this was decades before Finnur J\u00f3nsson collected and edited the skaldic corpus.12 This source-critical approach was very necessary to the scholarship, from which it cleared out much dead wood. Larson\u2019s biography is a case in point. De- spite the statements of doubt noted above, his use of such material is naive and clearly of the period before the revelations of the Weibull brothers and Koht.13 Larson uses Heimskringla for the marriage of Sveinn Forkbeard without any dis- cussion of the problems of this source;14 Kn\u00fdtlinga saga he uses for the supposed late arrival of E\u00edrikr H\u00e1konarson and his forces in 1015;15 Heimskringla, again, for the erroneous embassy between Cnut and St. \u00d3l\u00e1fr in the mid-1020s, and for the pact between Magn\u00fas \u201cthe Good\u201d and Harthacnut;16 Fagrskinna, for Cnut\u2019s meet- ing with Emperor Conrad II outside of Rome;17 and he was completely taken in by the literary exaggerations of J\u00f3msv\u00edkinga saga.18 Indeed, the new source- critical approach to these texts came at probably the worst moment for Larson, and did great damage to his biography within months of its release. Since his preface to the biography is dated 1911, his book was almost certainly in prepara- tion at the same time as Lauritz Weibull\u2019s own study, and most probably in press while the latter\u2019s conclusions were making themselves felt in Scandinavian scholarship. In 1912, perhaps as a consequence of this, Larson withdrew almost entirely from publications on medieval history and confined his research to 11 Larson, Canute the Great, 323 and ix\u2013x. 12 Larson, Canute the Great, 206 and 122. In fact, the only editions of such material available were those fossilized within the individual editions of the larger saga narratives and the mate- rial in Vigfusson and Powell, Corpus Poeticum Boreale. In this early extensive use of this mate- rial he mirrors Lauritz Weibull\u2019s \u201cKnut den stores sk\u00e5nska krig\u201d of 1910. 13 As an aside intended only to amuse, Larson is also badly served by the passage of time in his insistence on translating Scandinavian bynames into modern English. Due to this, the Nor- wegian magnate Einarr \u00deambarskelfir becomes \u201cEinar Thongshaker,\u201d shifting the meaning of his byname from a martial one involving trembling bowstrings (or paunches) to something far more in the lingerie line. 14 Larson, Canute the Great, 31. 15 Larson, Canute the Great, 72. 16 Larson, Canute the Great, 204\u20136 and 98\u201399 and 336. 17 Larson, Canute the Great, 226. 18 See Larson, Canute the Great, 156\u201357 for one example among many.","Epilogue 463 works on American-Scandinavian history, the British Empire, and World War I, with the exception of his two modern English translations of Old Norse sources: The King\u2019s Mirror in 1917 and The Earliest Norwegian Laws: Being the Gulathing Law and the Frostathing Law in 1935. Reactions to the Emergence of this Source- Critical Approach From our perspective, a little over a century later, the emergence of these critical approaches to later medieval Scandinavian sources seems an inevitable part of the similar European trends sweeping medieval studies in the nineteenth and early twentieth centuries. For Scandinavian medieval studies in the English- speaking world and Scandinavia, however, the consequences of this source- critical approach were anything but straightforward or uniform. Historians in both these regions began to withdraw from such sources. In Scandinavia many scholars withdrew from the fields associated with these sources almost entirely, shifting their attention to the mid-twelfth century or later, periods for which pri- mary sources are more abundant. Historical commentary on the eleventh cen- tury, especially on Cnut\u2019s reign, either confined itself in the main to English or German primary or near-primary sources where these covered events relevant to Scandinavia, or took refuge in other disciplines, such as numismatics, archaeol- ogy, and art history, which boasted more abundant and trustworthy material. The three Scandinavian nations which were principally engaged in studies of Cnut seem to have reacted according to the availability of source material for their own histories. Denmark, though not served well by late Scandinavian narra- tives, was better served than its northerly neighbours by the more contemporary sources from England and Germany. Denmark thus had the least to lose from an outright rejection of the late Scandinavian narratives wherever these could not be corroborated by more reliable sources. The source-critical approach had pro- found effects in Denmark, in which only a small amount of historical scholarship engaging with Cnut, in any form other than historical survey, exists for much of the twentieth century.19 One exception is S\u00f8ren Balle\u2019s \u201cUlf Jarl og drabet i Ros- kilde\u201d (Earl \u00dalfr and the Murder in Roskilde), published in 1983. This essay 19 This is not to diminish, however, the excellent archaeological and numismatic scholarship produced in Denmark in this period, or to ignore the historical research on individual aspects of eleventh-century Danish history and military organization.","464 Timothy Bolton returned to the subject matter of the 1910 article on the same subject by Lauritz Weibull, and in so doing had to engage with saga accounts of this apparent polit- ical murder on Cnut\u2019s orders. It includes sections on the \u201cFort\u00e6llingerne om dra- bet p\u00e5 Ulf\u201d (Narratives about the Murder of \u00dalfr) and \u201cDateringen af drabet\u201d (Dating of the Murder), each of which rehearses much saga material, assessing this against itself and other sources.20 Lauritz Weibull\u2019s doubt about these sources is examined and for the most part agreed with; what is absent from Balle\u2019s essay, however, is anything like a conclusion regarding their worth. We are offered the saga narratives one after the other and warned of their pitfalls, but without much attempt to assess their worth beyond where they find support in other more reli- able English or numismatic evidence. The reader might be forgiven for suspect- ing that these narratives are there because the subject forces their inclusion, and that they are repeated at length for the sake of completeness rather than as an integral part of the building blocks of the analysis. Likewise, Niels Lund had to engage with some saga material in his contribution to The Reign of Cnut volume, which was published in 1994. Here, towards the end of his paper, his comments turn to Cnut\u2019s actions elsewhere in Scandinavia, which are of course intrinsically linked to those in Denmark. As he does so, his words \u201c[a]ccording to later Old Norse sources\u201d signal a shift away from the more reliable sources used previ- ously, to the late Scandinavian narratives, which he classifies as \u201csuggestions.\u201d21 Norway, though it had some occasional notices in the English and German historical sources, was the principal subject of the late Scandinavian narrative sources and so had the most to lose from any rejection of these. Thus the new source-critical approach made its presence strongly felt in historical studies throughout the half-century following the years 1910\u20131915. Saga sources were allowed back in, but only in cases where they had support from other more con- temporary sources.22 Nonetheless, a new approach emerged in 1977 with the publication of Per Sveaas Andersen\u2019s seminal study, Samlingen av Norge og Kristningen av Landet 800\u20131130 (The Unification of Norway and the Christiani- zation of the country in 800\u20131130). In this work the author sought to take a long and wide view of history across three and a half centuries; doubtless he was 20 Balle, \u201cUlf Jarl og drabet i Roskilde,\u201d 31\u201339. 21 Lund, \u201cCnut\u2019s Danish Kingdom,\u201d 38\u201339. 22 For an example of this approach, see the collection of essays in Harald Hardr\u00e5de, edited by Arno Berg in 1966. Much of the book is occupied by numismatic and archaeological studies, with historical contributions on Harald and Byzantium, Harald\u2019s queen, and Norwegian rela- tions with Denmark and England. In these, where late Scandinavian narrative material might have been used in studies before Koht\u2019s work, we find in general that these are cited only when they have the support of a non-Scandinavian primary source.","Epilogue 465 influenced by the longue dur\u00e9e approach of the Annales school, which examined grand themes of the centralization of society and, within that, the roles of the major power structures of society. This approach to Norway\u2019s history necessitated the use of late Scandinavian narratives. While there is no direct comment on methodology in Sveaas Andersen\u2019s work, it is clear that he sought to address the inherent problems of his sources by drawing attention to advances in literary studies on this material and by employing only the barest of details from such sources in an effort to eliminate as many later accretions as possible. In his intro- duction, Sveaas Andersen rehearses and endorses the same doubts raised by the Weibulls and Koht, but follows this up with a lengthy quotation from Anne Holts- mark on the sources that were employed by the thirteenth-century saga au- thors.23 When his study arrives at the reign of Cnut\u2019s contemporaries, we are reminded again of the textual relationships which created the sagas we have today. Citing the work of the literary scholars Sigur\u00f0ur Nordal (whose ground- breaking work on the interrelation of the narrative sources appeared in 1914, dur- ing the same period as the Weibulls\u2019 and Koht\u2019s studies), O. A. Johnsen (1916), Johan Schreiner (1926), J\u00f3n Helgason (1941), and Anne Holtsmark (1967), Sveaas Andersen does expend some effort to define an \u201celdre sagatradisjon\u201d (older saga tradition), whose positive and negative qualities he attempts to weigh.24 His bare narrative is mostly established by these sources, backed up by any relevant skaldic verse, numismatic, legal, or runic material or by relevant non-Scandinavian sour- ces where available. Thus, allowing for some brief speculation on the potential in- terpretations of the latter sources, he creates a sparse thumbnail sketch of a historical narrative.25 More often than not, somewhat worryingly, he introduces such details without footnotes, reducing our ability to approach the supporting sources critically as readers and giving the impression of a consensus-agreed base narrative. This position has prevailed in Norwegian scholarship that touches on Cnut, more or less until the present day. The same focus on larger social themes rather than detail can be found in Claus Krag\u2019s survey Norges historie fram til 1319 (Norwegian History up to 1319), published in 2000. This book produces a thumb- nail sketch of the period relevant to Cnut with barely a reference to individual sources beyond phrases such as \u201cI skaldekvadene h\u00f8rer vi at\u201d (We hear from the 23 Sveaas Andersen, Samlingen av Norge, 25, citing Holtsmark, \u201cSankt Olavs liv og mirakler.\u201d 24 Sveaas Andersen, Samlingen av Norge, 109\u201311 and 115, citing Sigur\u00f0ur Nordal, Om Olaf den helliges saga, O. A. Johnsen, Olavsagaens genesis, Johan Schreiner, Tradisjon og saga, J\u00f3n Hel- gason, Den store saga om Olav den hellige, and Holtsmark\u2019s study, which is cited above. His suggested reading adds the further literary study of Bjarni A\u00f0albjarnarson, Om de norske kongers sagaer. 25 Sveaas Andersen, Samlingen av Norge, 115\u201339.","466 Timothy Bolton skaldic verses that).26 This sketch is augmented by some brief general comments later in the book on the problems commonly associated with later Scandinavian sources, and there are assessments individual to each chapter alongside lists of sources and published collections of sources.27 The situation is much the same in J\u00f3n Vi\u00f0ar Sigur\u00f0sson\u2019s Det Norr\u00f8ne Samfunnet, Vikingen, Kongen, Erkebiskopen og Bonden (\u2018Old Norse\u2019 Society: Viking, King, Archbishop and Farmer), published in 2008, albeit this work refers to the sources more accurately. This Norwegian approach has its merits. While it comes with wider margins of potential error in its findings, due to the dearth of source material, it does seem likely that true \u201ckings\u2019 sagas\u201d (that is, those sagas not framed in such liter- ary style as J\u00f3msv\u00edkinga saga) relate, more or less correctly, the major political events of the eleventh century and perhaps some from the last decades of the tenth.28 These were events that occurred less than a century or two before they were put to parchment. While such accounts of events most probably suffer from erosion of detail (especially of the detail which played no further part in future events), and from a telescoping of the narrative, the general audience was proba- bly prevented by mutual scrutiny from making wanton inventions; and at least some of that audience must have continued to recite and enjoy contemporary skaldic verse about many of these events. What such accounts do badly, how- ever, is interpret why events happened, for the interpretations within these ac- counts frequently reveal concerns of the writer\u2019s own time rather than the events as they happened.29 With the exception of statements in skaldic verse, if these are introduced in the prose as the works of named poets or from named poems, direct speech in saga prose is never to be trusted. In addition, from the eleventh century onwards, genealogy does seem to have had an important social role in Scandinavia. As long as it relates to contemporary links (and not to a far distant and legendary past) and as long as it relates to geographically close relation- ships, genealogy may perhaps be trusted. The check on wanton invention that the medieval audience might give such material does not apply if such relation- ships are set either in a distant past or outside of Scandinavia (or perhaps even 26 Krag, Norges historie, 65. 27 Krag, Norges historie, 218\u201327 and 279\u201380. 28 It bears stating here that I would not advocate the use of such material further back than the last decade or so of the tenth century. Before this the deviation of such material from the few other sources we have, and the potential level of invention, is sufficiently high to prevent such use. 29 Here again, for fuller discussion of this serious concern with such sources, I direct the reader to Sawyer and Sawyer, \u201cAdam and Eve of Scandinavian History.\u201d","Epilogue 467 outside of northern Scandinavia). Examples of this last potential error are proba- bly the isolated and clearly erroneous statements of the so-called \u201cSupplement\u201d to J\u00f3msv\u00edkinga saga, that Eadric Streona was Queen Emma\u2019s brother and had fos- tered Edmund Ironside.30 In these cases, the geographical borders and distances ensured that the text\u2019s audience knew less and that the writer was freer to embel- lish for literary effect, here turning eleventh-century Anglo-Saxon politics into the form of drama that one finds in thirteenth-century Icelandic family sagas. That said, this approach cannot be proved so valid that the above \u201crules\u201d may be applied without significant caution and careful qualification. Early medieval Sweden has few notices in the contemporary or near- contemporary English or German historical sources, nor is it comprehensively served by the late Scandinavian narrative sources, which give but little coverage. The one exception is the Battle of Helge\u00e5, in which Cnut appears to have de- feated a number of his most powerful Scandinavian enemies, cementing his su- premacy in the region. Most probably the site of the battle was in Sk\u00e5ne, now the southernmost part of modern Sweden, then a region in medieval Denmark. In the twentieth century, debates about this battle were few and far between and driven by the research of a single Swedish scholar, Ove Moberg, from 1941 until 2008, when Bo Gr\u00e4slund wrote a single article in response to research that Mo- berg had carried out a generation earlier in 1986.31 Moberg, while partly studying under Lauritz Weibull himself, received guidance in his later studies from a skaldic scholar, E. A. Kock, as well as from the literary scholars Eilert Ekwall and J\u00f3n Helgason. This literary aspect of Moberg\u2019s training most probably lies behind his unconventional historical approach.32 English-language scholarship on Cnut in this period has been in much the same boat as the Danish studies, for it can afford to set the late Scandinavian narratives aside. Where these do have a part to play is in surveys of Cnut\u2019s whole life or dominions, and here some interesting solutions emerge. One need look no further than Sir Frank M. Stenton\u2019s seminal Anglo-Saxon England, which 30 Edited in Encomium, ed. Campbell, 92\u201393. 31 Moberg, Olav Haraldsson, Knut den Store och Sverige (Olaf Haraldsson, Cnut the Great and Sweden), 148\u201378 (chap. 4); see also Campbell\u2019s critical assessment of Moberg\u2019s work as a post- script to his Encomium Emmae. Moberg returned to this subject in \u201cKnut den stores motst\u00e5n- dare i slaget vid Helge\u00e5\u201d (Cnut the Great\u2019s Opponents in the Battle of Helge\u00e5), and again in \u201cSlaget vid Helge\u00e5 och dess f\u00f6ljder\u201d (The Battle of Helge\u00e5 and its consequences). He then re- argued it in English in his \u201cThe Battle of Helge\u00e5.\u201d Bo Gr\u00e4slund\u2019s response to Moberg appeared in 1986 as \u201cKnut den store och sveariket. Slaget vid Helge\u00e5 i ny belysing\u201d (Cnut the Great and Sweden: the Battle of Helge\u00e5 in a new light). 32 For these relationships, see the introduction to his Olav Haraldsson.","468 Timothy Bolton was published in 1943 with a second edition in 1947 and a third in 1971. When Stenton had to comment on Cnut\u2019s activities in Scandinavia, he cited the Anglo- Saxon Chronicle\u2019s report that Cnut went to Norway and seized control there in 1028, otherwise merely noting that \u201c[l]ater Old Norse authorities amplify this out- line.\u201d33 Further comment on these \u201cauthorities\u201d is given in his sources section and restricted to their \u201cbearing on English history.\u201d34 He notes their literary merit and the role they play in bringing out the importance of certain Scandinavians. However, he also says that they are hampered by \u201ctheir innumerable mistakes on points of fact [that] show the weakness of a tradition which is uncontrolled by written record.\u201d Nonetheless, he goes on to follow (or perhaps reinvent) the Nor- wegian approach, deploying the same sources over two densely packed, event- filled paragraphs in a representation denuded of any potentially worrying detail. The late Peter Sawyer follows suit, offering the most recent English-language sur- vey of the field in his \u201cCnut\u2019s Scandinavian Empire,\u201d published in The Reign of Cnut volume in 1994. His is a thorough survey, but what he does not cite in evi- dence is revealing. In his discussion of northern Scandinavia, Sawyer notes skaldic verse often, as well as the late-twelfth-century \u00c1grip af N\u00f3regskonungas\u00f6- gum, but not the later sources.35 While there is no explicit comment in this essay, other publications by Sawyer make clear his acceptance of the problems of using later Scandinavian sources for the study of the eleventh century.36 However, when his narrative in \u201cCnut\u2019s Scandinavian Empire\u201d must enter territory for which there is no source other than late sagas, this is done without comment on the nature of these sources and the citations in the footnotes are to Sveaas Ander- sen\u2019s survey (which, as noted above, is largely based on the sagas and bereft of footnotes of its own).37 Sawyer\u2019s choice of source here was doubtless informed by his years teaching and living in Norway. I do not include these examples to point fingers at individual historians, but instead to show the inherently prob- lematic nature of the current way in which such late Scandinavian sources are handled. That is, the source-critical approaches of the early twentieth century dictate that such texts must be set aside as untrustworthy, but then if one is to say anything about northern Scandinavia in the period, one is forced to use these same sources in some fashion, often in a convoluted or less-than-fully- acknowledged form. 33 Stenton, Anglo-Saxon England, 404\u20135. 34 Stenton, Anglo-Saxon England, 700\u2013701. 35 Sawyer, \u201cCnut\u2019s Scandinavian Empire,\u201d 20\u201322. 36 See Sawyer and Sawyer, \u201cAdam and Eve of Scandinavian History.\u201d 37 See Sawyer and Sawyer, \u201cAdam and Eve of Scandinavian History,\u201d 21\u201322.","Epilogue 469 An alternative approach in English-language scholarship is to be found in the appendices attached to Alistair Campbell\u2019s edition of the Encomium Emmae Reginae in 1949. Here, however, we have what appears to be a mixed approach, perhaps even a confused approach, one suggesting that Campbell was torn be- tween literary and historical camps. These disciplines had over the previous four decades drifted far apart from each other on the subject of these late Scan- dinavian narratives. Campbell came from a literary and linguistic background and was well versed in Scandinavian languages. He was clearly aware of the saga material and the criticisms directed at it. However, despite a lengthy dis- cussion of this material in his edition of the Encomium, he offers the reader no open statement of his position on such sources.38 We might discern some hints that he does not intend to cast out all such material, in his introductory address to \u201cthose who wish to use Scandinavian sources for the history of the eleventh century,\u201d and in his direction of these users to \u201cBjarni A\u00f0albjarnarson\u2019s excel- lent work Om de norske kongers sagaer [\u201cOn the Sagas of the Norwegian Kings\u201d] (Oslo, 1937) and to the enduring value of Sigur\u00f0ur Nordal\u2019s Om Olaf den helliges saga [\u201cOn the Saga of Saint \u00d3l\u00e1fr\u201d] (Copenhagen, 1914).\u201d Another hint may be taken from from Campbell\u2019s later rehearsal of Nordal\u2019s views on the interrelation and relative antiquity of the various saga materials.39 At first sight, the reader might even be forgiven for thinking that Campbell is in sup- port of the cautious use of such materials. The thought appears to be confirmed when we turn to the lengthy biographies of Cnut\u2019s Scandinavian followers in Campbell\u2019s Appendix III, which weigh up the various sources for them one against each other, resting in some substantial part on saga material.40 Nonethe- less, despite rehearsing such material and assessing it at length, Campbell goes on to state, in a few throwaway but fundamentally important lines, that Heim- skringla and Fagrskinna are \u201c[h]istorically worthless,\u201d while he observes that the so-called \u201cSupplement\u201d to J\u00f3msv\u00edkinga saga, which he includes in the biography of Thorkell the Tall and edits in full in his monograph, is of historical value only 38 Indeed, he fails to note or include in his bibliography the work of either the Weibulls or Koht. I assume his direction to that of Bjarni A\u00f0albjarnarson (see below) is meant to implicitly contain this recommendation. 39 Encomium, ed. Campbell, v and 80\u201381. 40 Encomium, ed. Campbell, 66\u201391. Much of what he discusses there had already been dis- cussed in a more naive fashion by A. S. Napier and W. H. Stevenson in their The Crawford Col- lection of Early Charters and Documents, 139\u201349. Campbell must owe some debt to this earlier publication; he did know of it, for he cites it at 82 and 86\u201387 of his Appendix III.","470 Timothy Bolton where its facts are confirmed by other earlier sources.41 The reader could be for- given for being confused, or at least could be left questioning why nearly thirty pages of a forty-four-page appendix have been given over to a thorough rehearsal of material that is then announced to add nothing other than late confirmation to our other more reliable sources.42 Recently another English version of the robustly source-critical approach has appeared in a 2016 article by Ann Williams on Thorkell the Tall, who, as we have seen in earlier chapters in this volume, was a powerful associate and sometime enemy of Cnut.43 The main thrust of Williams\u2019s paper makes the quite reasonable assertion to \u201cpay more attention to contemporary sources and to mistrust anything which cannot be corroborated,\u201d but she goes on to ques- tion the connection between the Thorkell who is described in our contemporary and near-contemporary English sources (augmented by a brief appearance in the skaldic poem Li\u00f0smannaflokkr) with the figure by this name in the late Scandinavian narrative sources. These late Scandinavian narratives identify him as a member of a dynasty which ruled over Sk\u00e5ne at the southern end of what is now Sweden. Urging caution, she notes the lack of prominence of Thor- kell in the narratives of the earliest saga sources, in which he is only ever a sup- porting actor in the drama there, using this to suggest that, in Scandinavia, \u201cwhile Thorkell\u2019s name survived, little was known of his actual deeds\u201d; and that, \u201c[a]s Thorkell\u2019s historical career faded into the mists of time, his legendary life began to develop,\u201d by which she means that subsequent Scandinavian writ- ers merely invented these details in their accounts.44 She continues in this vein, and turns her attention to the appearance of potential family members with simi- lar groups of names in both sets of sources, as first noted by Freeman in 1877 and since supported by Campbell and then tentatively by Simon Keynes, among others.45 These similarities, however, she sweeps aside with the line: \u201c[i]t is on 41 Encomium, ed. Campbell, 81, 90\u201391; with edition in Appendix IV, 92\u201393. 42 Campbell appears to have taken a similar line with what to do with skaldic verse. His paper, Skaldic Verse and Anglo-Saxon History (1970) is full of suggestions of material to use, but then undermines the potential use of these. See his section on the verse composed for Cnut (13\u201315), where he takes the differences of detail between such witnesses and the tradi- tional English historical record to indicate that the poets were ill-informed and vague. Com- pare to this the conclusions reached in Poole, \u201cSkaldic Verse and Anglo-Saxon History,\u201d Bolton, Cnut the Great, and the account of such verse set out in Townend, \u201cContextualising the Kn\u00fatsdr\u00e1pur.\u201d 43 Williams, \u201cThorkell the Tall.\u201d 44 Williams, \u201cThorkell the Tall,\u201d 151\u201352. 45 See Freeman, History of the Norman Conquest, I, 655\u201356; Encomium, ed. Campbell, 84\u201385, where he concludes \u201c[i]t would be a remarkable coincidence\u201d if these names did not correctly","Epilogue 471 coincidences like these that historians pounce like leopards, lifting juicy titbits out of their contexts and stringing them together into a plausible story \u2013 what does not fit . . . is ignored or explained away as the inevitable confusion of oral tradition.\u201d A further late Scandinavian source, the so-called \u201cSupplement\u201d to J\u00f3msv\u00edkinga saga edited by Campbell, is written off by Williams with a flourish: \u201cthe less said the better.\u201d46 It is a difficult source, but as Campbell noted, it con- tains a surprising amount of accurate material which we must account for in some way.47 A view entirely focused on primary sources is an excellent thing, but, as noted above, unachievable for the scholar who wishes to look at much of Cnut\u2019s activities in Scandinavia as well as those in England. As Cnut had an equal pres- ence in both of those regions during his lifetime, I believe we have a responsibil- ity to take in as many sources as possible and to see what can be done with them. To ignore our Scandinavian sources risks confining our attention to only the English part of his realms and weighting our view of him in that direction. A crucial part of Williams\u2019s argument rests on the assumption that medieval Scandinavian authors invented details of the life of Thorkell the Tall to fill a void created by an absence of sources between the contemporary skaldic verse and the subsequent thirteenth-century sagas. To borrow a term from modern forensic science, this would be a \u201cbreak in the chain of evidence.\u201d When seen in the light of a century of literary studies of the texts involved,48 this assumption does not bear much scrutiny. The more closely we look at this perceived void, the more it shrinks. The complex meter and linguistic intricacy of skaldic verse, which are features ensuring that it has been, in the most part, transmitted without substan- tial tampering, are well known and I will not rehearse them here.49 Equally well identify this kin relationship, concluding that this was \u201creasonably certain\u201d; and Keynes, \u201cCnut\u2019s Earls,\u201d 62, n. 97, and 66. 46 Williams, \u201cThorkell the Tall,\u201d 151\u201352. 47 Campbell, Encomium, 89. See also Bolton, Empire of Cnut the Great, 211\u201312, esp. n. 28. 48 The most important of these studies are listed above in n. 24. 49 For comment on this and a discussion of the various editions of such material see Bolton, Cnut the Great, 18\u201321. Ghosh, in King\u2019s Sagas and Norwegian History, has recently raised some doubt about the use of skaldic poems as historical sources. He questions the oft-noted stability of this kind of verse by pointing towards certain conclusions concerning a number of literary and textual studies of various skaldic verses. These studies show verses with variants which were most probably inserted by later poets repeating the material, and perhaps also by scribes copying parts of the works into the extant saga material (see 46\u201348 and the references therein). The conclusion of Ghosh is that if we can observe changes occurring in such verses, then all skaldic verses must be suspect. Although this subject deserves fuller attention else- where, it is enough now not only to agree with Ghosh that there are problems, but also to wish","472 Timothy Bolton discussed by others are the forms of introduction used by later saga-authors, which signal the difference between, on the one hand, the more reliable verses, which are being quoted from poems that were already in existence and often named and ascribed to known poets, and on the other, the lausav\u00edsur (loose- verses), which come without such indications that they existed before the saga- writer put pen to parchment.50 As has long been noted, much of the narrative of late Scandinavian sources is directly based on such verse.51 This is true of the sagas produced in Norway and Iceland in the heyday of such writing, the thir- teenth century, as well as of the so-called synoptic histories composed in Norway in the second half of the twelfth. Two of these synoptic histories are of interest here: one is Historia de antiquitate regum Norwagiensium, which is a Latin ac- count of the history of Norway from the mid-ninth century until 1130, written by \u201cTheodoricus Monachus\u201d;52 the other is a closely related sister text, the vernacu- lar \u00c1grip af N\u00f3regskonungas\u00f6gum. The \u00c1grip, though fragmentary at its begin- ning and end, covers the period from the late ninth to the early twelfth century.53 The monk Theodoricus\u2019s opening address to Eysteinn, archbishop of Nidaros that his study had discussed them in more detail. He does not note Russell Poole\u2019s use of just such variants to deduce probable English loan words behind some of the poems for Cnut, loan- words which argue powerfully for authenticity; see Poole\u2019s \u201cSkaldic Verse and Anglo-Saxon History,\u201d 284\u201386, and also in this volume, pp. 260\u201369. Moreover, all the studies cited by Ghosh as underpinning the existence of such variants concern the textual traditions of ex- tremely early verses from the ninth century or from more literary skald\u2019s sagas. Since his foun- dations thus lie outside the main corpus of material usually employed for historical purposes, we might wish that he had repeated such studies for eleventh-century \u201chistorical\u201d verses. In- deed, one of the authors cited by Ghosh (Poole, in \u201cVariants and Variability in Egill\u2019s H\u01ebfu\u00f0- lausn,\u201d 101) notes that there is greater and lesser flexibility in different verses, fixed in part by the fact that the flexible poem \u201ctells no real story.\u201d Certainly, we must accept Ghosh\u2019s criticisms and be mindful of them when using such verse as historical evidence in the future. Also note, however, that changes to individual words and in some cases entire lines do not in themselves necessarily give reason to reject this source. Narrative rather than expression is harder to shift in the complex meanings of such verses, and it is narrative that has been princi- pally of interest to the majority of historians. In addition, where such expressions of power, or echoes of such, have been employed by historians such as myself, this is usually within the con- text of other examples from multiple poems which are evidently unconnected; that argues for, rather than against, veracity. 50 Note, however, that Williams\u2019s endorsement of such verse, in her \u201cThorkell the Tall,\u201d 144, is lukewarm and rests on Eric Christiansen\u2019s general survey volume, The Norsemen in the Vi- king Age, rather than on the numerous detailed studies available. 51 See Ghosh, King\u2019s Sagas and Norwegian History, 70\u201394, and references there. 52 Text edited by Storm, Monumenta historica Norvegi\u00e6, 1\u201368. 53 Text edited most recently by Driscoll, \u00c1grip, 2\u201380.","Epilogue 473 (Trondheim), fixes his work within Norwegian writing of the period between his consecration in 1161 and death in 1188.54 He knew of events of September 1176, while the lack of any definitive statement about Eysteinn being away from his see suggests a date for this work in 1177\u20131178.55 Theodoricus makes frequent mention of Icelanders as separate from him and his audience, refers to \u00d3l\u00e1fr Tryggvason as \u201cour king,\u201d and shows a knowledge of Nidaros and its immediate region, all of which strongly indicates that he was Norwegian and lived in that urban site.56 Consensus places \u00c1grip after Theodoricus\u2019s work, as one that was composed in ca. 1190 and most probably before Oddr Snorrason began work on his \u00d3l\u00e1fs saga Tryggvasonar in ca. 1200, with which it appears to share similar passages.57 Again, the author reveals himself to be Norwegian, but this time through morphological \u201cNorwegianisms,\u201d through nicknames not otherwise found in Icelandic writing, and through the fact that the action and local knowledge is firmly based in Ni- daros (often referred to as kaupangr, \u201cthe town,\u201d implying familiarity).58 Both of these short synoptic histories were clearly produced by a historical school there. Theodoricus\u2019s Historia makes its debt to skaldic verse explicit in its prologue, stat- ing that it sets down these few details concerning the ancient history of the Norwe- gian kings, \u201cet prout sagaciter perquirere potuimus ab eis, penes quos horum memoria praecipue vigere creditor, quos nos Islendinga vocamus, qui haec in suis antiquis carminibus percelebrata recolunt\u201d (as I have been able to learn by assidu- ous inquiry from the people among whom in particular the remembrance of these matters is believed to thrive \u2013 Icelanders, who preserve them as much celebrated themes in their ancient poems).59 \u00c1grip acts out this statement, enclosing seven complete stanzas and two half-stanzas within the body of its narrative, as well as including references to the late-tenth-century poet Eyvindr sk\u00e1ldaspillir and his poem H\u00e1leygjatal.60 Nor do these authors\u2019 internal references to their sources stop there. Theo- doricus\u2019s prologue, when listing its sources, goes on to state that \u201c[v]eritatis vero sinceritas in hac nostra narratione ad illos omnimodo referenda est, quo- rum relatione haec annotavimus, quia nos non visa sed audita conscripsimus\u201d 54 Theodoricus: Historia, ed. McDougall and McDougall, xi\u2013xiii. 55 Theodoricus: Historia, ed. McDougall and McDougall. 56 Theodoricus: Historia, ed. McDougall and McDougall, ix\u2013xi. 57 \u00c1grip, ed. Driscoll, xii, and references there. 58 \u00c1grip, ed. Driscoll, x\u2013xi, and references there. 59 Theodoricus\u2019s Historia, ed. Storm, in Monumenta historica Norvegi\u00e6, 3 (Prologue). Transla- tion based on Theodoricus: Historia, ed. McDougall and McDougall, 1. 60 \u00c1grip, ed. Driscoll, 12 (chap. 6) and 26 (chap. 15). The latter of these refers the reader to the poem, naming both it and its poet, as support for the details in the narrative.","474 Timothy Bolton (the degree of pure truth in my narrative must be placed entirely at the door of those by whose report I have written these things down, because I have re- corded things not seen but heard).61 The interpretation of this phrase \u201caudita non visa\u201d (here \u201cnot seen but heard\u201d) has caused its share of trouble. A superfi- cial interpretation might take these things to be the spoken accounts of eyewit- nesses, rather than written sources. Jens Hanssen, however, pointed out as long ago as 1945 that this common Latin phrase does not preclude the use of written sources, and that it should be read as an opposition between chroni- cling the events of one\u2019s own time against those of the past compiled from other sources, either written or oral; David and Ian McDougall returned to this argument in 1998.62 The McDougalls observe that this interpretation was in keeping with the author\u2019s citation of at least one now-lost written text of great relevance for our purpose here: a catalogus of Norwegian kings, cited in refer- ence to Cnut the Great and to those who ruled Norway in his stead.63 It is also of relevance that Theodoricus declines to describe the accounts of the miracles of St. \u00d3l\u00e1fr and Bishop Gr\u00edmkell\u2019s exhumation of his body, with the words \u201cquia haec omnia a nonnullis memoriae tradita sunt, nos notis immorari super- fluum duximus\u201d (because all these things have been recorded by several, I re- gard it as unnecessary to dwell on matters which are already known).64 The McDougalls note that the Latin idiom chosen by Theodoricus, a variant of me- moriae tradere, normally means \u201cto record in writing.\u201d65 Thus it seems clear that the writers of the last decades of the twelfth century did not work exclu- sively from skaldic verse and oral accounts, but also had access to some written material. I have suggested before that at least the material relating to the details of St. \u00d3l\u00e1fr\u2019s martyrdom and beatification may have been hagiographic in con- tent, but this is where we hit a wall, for the composition of the Passio Olavi in the mid-twelfth century appears to have swept away almost all earlier mate- rial.66 A few extant scraps suggest that such material did exist then, and per- haps was of some age already in the mid-twelfth century. A single leaf from an 61 Theodoricus\u2019s Historia, ed., Storm, in Monumenta historica Norvegi\u00e6, 4 (Prologue). Transla- tion based on Theodoricus: Historia, ed. McDougall and McDougall, 2. 62 Hanssen, \u201cObservations on Theodoricus,\u201d and Theodoricus: Historia, ed. McDougall and McDougall, xiv and 56\u201357, n. 11. 63 Theodoricus\u2019s Historia, ed., Storm, in Monumenta historica Norvegi\u00e6, 44 (chap. 20). See also Theodoricus: Historia, ed. McDougall and McDougall, xiv and 92\u201393, n. 214. 64 Theodoricus\u2019s Historia, ed., Storm, in Monumenta historica Norvegi\u00e6, 44 (chap. 20). Trans- lation based on Theodoricus: Historia, ed. McDougall and McDougall, 33; see also xiv and 92, n. 213. 65 Theodoricus: Historia, ed. McDougall and McDougall, 92. 66 Bolton, Empire of Cnut the Great, 254\u201355.","Epilogue 475 Old Norse hagiographical collection, one which dates to ca. 1155\u00d71165 and in- cludes material on St. \u00d3l\u00e1fr\u2019s miracles, survives now in Copenhagen, A.M. MS. 325 v \u03b1 4to. Furthermore, knowledge of organized votive masses for St. \u00d3l\u00e1fr are witnessed in Exeter by the Leofric Collectar (British Library, Harley MS. 2961) in the 1050s, and in Sherborne by the Red Book of Darley (Cambridge, Corpus Christi College, MS. 422) in or after the early 1060s, while John of Worcester knew of some form of hagiographic tradition surrounding him in the 1130s.67 These last three references to hagiographic materials for St. \u00d3l\u00e1fr pre- date the foundation of Norwegian monasteries, which are the most likely candi- dates for a storehouse of historical writing. In Norway the earliest monasteries were Lyse (1146), Hoved\u00f8ya (1147), Munkeby (1150\u20131180), Halsn\u00f8y (1163), and T\u00f8nsberg in the second half of the twelfth century. It seems inconceivable that liturgical and hagiographic materials should be developed in England for a Norwegian saint such as \u00d3l\u00e1fr. The skaldic poem Gl\u00e6lognskvi\u00f0a, by \u00de\u00f3rarinn Loftunga, makes it clear that some part of his worship was promulgated from within the Norwegian royal court,68 while such bishops as existed in Norway at that time were missionaries looking to the king as their immediate patron. In this context, we should probably locate the origin of these materials and their now-lost written records in this court. In this way, we can only perceive this potential \u201cchain of evidence\u201d imper- fectly. On the other hand, it is clear that there was not a two-century-wide gap in the sources between the composition of terse skaldic verse in the eleventh century and the flowing prose of the fully-fledged Icelandic sagas in the thir- teenth, which had to be filled by fertile imaginations. There are solid stepping- stones in the middle to late twelfth century, and suggestions of other written sources even earlier. Even if we ignore the flimsier of these suggestions, we come to within decades of living memory of the events of the 1030s, and within the period in which someone who had met and discussed such events with an eyewitness might have lived. It remains true that a person aged twenty years in 1030, who might have lived to be eighty in 1090, could comfortably have met and discussed events of the past with someone still alive in 1150. 67 Chronicle of John of Worcester, ed. Darlington and McGurk, 510 (s.a. 1027) and 543 (s.a. 1046). 68 Townend, \u201cKn\u00fatr and the Cult of St. \u00d3l\u00e1fr,\u201d and Bolton, Empire of Cnut the Great, 271\u201375.","476 Timothy Bolton A Potential Alternative Approach to These Sources This shift of perception of the \u201cchain of evidence\u201d allows us not only to see our sources in a slightly different light, but also to entertain and test the notion that there may be some veracity in them that we might be able to detect. I shall now turn to describing how I have approached these late Scandinavian narra- tives, and how occasionally I have attempted to employ them in my own work. I don\u2019t think I have ever advocated, nor do I here, treating such a shift in per- ceptions as the license to take any report in these sources at face value. When dealing with such material, I think we must presume all such sources guilty until proven to be reasonably innocent. My approach differs greatly from the modern Norwegian model, in that I would prefer to focus on hard-fought-for de- tails, rather than make general thumbnail sketches in a longue dur\u00e9e tradition. The scholar who attempts to use these must strive to gather as many Scandina- vian sources as can be found; I hope to show below how greatly sources such as lists of poets and their patrons and amendments to medieval law codes can add to our trust in parts of the late Scandinavian narratives. Just as impor- tantly, the same scholar must try to understand such sources within their indi- vidual and generic contexts, in order to identify their potential weaknesses and strengths. The rewards may be numerically small, but each point in which it is reasonable to root the given details more plausibly in an eleventh-century real- ity than in some later literary invention, shines new light into a very dark pe- riod of history, and every secure foothold established there potentially changes our understanding of the entire period. Let me offer two short examples, which are already laid out in my 2009 and 2017 books on Cnut and will suffice to show how this methodology works, per- haps inspiring some confidence in its results. Firstly, since Ann Williams has called the links between Thorkell the Tall and the dynasty of the earls of Sk\u00e5ne into question in the paper discussed above, we might first turn our attention to these. The assertion of Williams that is relevant here is that the slightness of Thorkell\u2019s part in the saga narratives gives such reason to doubt the historicity of his part there that we may sweep it away as potentially a later invention.69 One sizeable problem with Williams\u2019s assessment lies in her choice of saga material to illustrate her arguments. This choice reveals either her lack of familiarity with these sources, or her debt to Campbell\u2019s appendices \u2013 documents nearly seventy years old when she turned 69 I have discussed this briefly in my Cnut the Great, 23 and 59\u201362.","Epilogue 477 to the subject. The main Scandinavian source that she names is the literary and late J\u00f3msv\u00edkinga saga. Indeed, while she does not directly charge any modern scholar with the errors she outlines, Williams\u2019s choice of this saga suggests that she was thinking of Campbell (who is cited elsewhere) or perhaps of Keynes (who utilized much of the material in Campbell\u2019s appendix for the Scandina- vian figures in his essay on \u201cCnut\u2019s Earls\u201d) as she wrote.70 Williams is quite right to point out the problems of J\u00f3msv\u00edkinga saga that should eliminate it from use as a historical witness; I think that no one would now claim that this saga is free of erroneous accretions, and elsewhere I have set it aside. However, both she and Campbell miss a number of sources which identify Thorkell (or to play devil\u2019s advocate, at least someone of his name who was understood by writers in the thirteenth century to be Thorkell the Tall) as a member of a dy- nasty who ruled Sk\u00e5ne in the late tenth and eleventh centuries. A list of these sources, even after noting their potential links to each other, argues against such a presumption of later invention. Both Williams and Campbell also acknowledge the so-called \u201cSupplement\u201d to J\u00f3msv\u00edkinga saga, albeit Williams does so very briefly. To this source we might add Heimskringla, which includes some brief comments on this ruling dy- nasty, including Thorkell\u2019s part in it.71 It must be admitted, however, that Heim- skringla might be taken to depend on the traditions set out in J\u00f3msv\u00edkinga saga and so cannot be held independent of that work. There again, we should also note that there is information in Heimskringla, not found in any other extant source, about the career of the skaldic poet \u00de\u00f3r\u00f0r Sigvaldask\u00e1ld, which claims that \u00de\u00f3r\u00f0r served the Sk\u00e5ne dynasty and Thorkell in particular.72 There \u00de\u00f3r\u00f0r is 70 Indeed, one suspects that certain phrases in Campbell\u2019s discussion of Thorkell and his son Haraldr might be the genesis of Williams\u2019s thesis in her \u201cThorkell the Tall.\u201d This is despite the fact that Campbell (Encomium Emmae, 84) supports the connection between the Thorkell in the English sources and his namesake in the late Scandinavian ones. Note Campbell\u2019s accom- panying assertion, with which he resolves which Scandinavian nobles fought at Helge\u00e5 and held jarldoms in Scandinavia: \u201cNow in this Snorri is falling into a practice in which he is very apt to indulge. When he has to find a person for some purpose, he seizes upon one who had some reality, however shadowy, rather than invent one.\u201d Campbell directs us to his war-time \u201cThe Opponents of Haraldr H\u00e1rfagri\u201d for other examples of the same, but does not mention that these are all related to a single battle of the late ninth century, a period so early that it is more likely to contain inventions and legendary accretions than eleventh-century events. 71 \u00d3l\u00e1fs saga Tryggvasonar, ed. Bjarni A\u00f0albjarnarson, 272\u201374 (chap. 34\u201335). \u00d3l\u00e1fs saga helga, ed. Bjarni A\u00f0albjarnarson, 54 (chap. 43). 72 \u00d3l\u00e1fs saga helga, ed. Bjarni A\u00f0albjarnarson, 54 (chap. 43). Williams appears not to know of this, given her statement that \u201cno skald seems to have felt moved to compose eulogies for Thorkell,\u201d in \u201cThorkell the Tall,\u201d 149.","478 Timothy Bolton identified as the permanent court-poet of Earl Sigvaldi, while it is stated that he was later in the train of Thorkell the Tall, Sigvaldi\u2019s brother, and \u201ceptir fall jarls \u00fe\u00e1 var \u00de\u00f3r\u00f0r kaupma\u00f0r\u201d (after the fall of the earl, \u00de\u00f3r\u00f0r became a merchant).73 \u00de\u00f3r\u00f0r was the father of perhaps the most celebrated and prolific skald of the whole period, Sigvatr \u00de\u00f3r\u00f0arson, who served both St. \u00d3l\u00e1fr and Cnut. Conse- quently, he is an improbable (but not impossible) candidate to choose if one wished to fabricate parts of his life a century after his death. Finally, another often overlooked source, the Sk\u00e1ldatal, also indirectly links Thorkell to this dy- nasty.74 This list of Scandinavian rulers and the poets who served them was compiled in the early thirteenth century, most probably as part of the research materials which stood behind the composition of Heimskringla.75 The so-called Kringla manuscript (of ca. 1260), which included Sk\u00e1ldatal, was destroyed in the Copenhagen fire of 1728 (with the exception of a single leaf not relevant to our purposes here). However, early modern manuscript witnesses that descend from Kringla put a patron, Haraldr \u00de\u00f3rkelsson, who is presumably the son of Thorkell the Tall, and his poet \u00dej\u00f3\u00f0\u00f3lfr Arn\u00f3rsson immediately after Sigvaldi at the end of the list of the earls of Sk\u00e5ne.76 Heimskringla claims that this Haraldr was the son of Thorkell, who was taken under Cnut\u2019s wing, and that he was the earl named in the skaldic poem Gl\u00e6lognski\u00f0a as part of the Danish government sent by Cnut to Norway in the last years of the 1020s.77 This does not amount to an enormity of extra material, and its component parts are all loosely related to each other, but does it reduce the efficacy of Williams\u2019s suggestion that a later writer invented Thorkell\u2019s link to the Sk\u00e5ne dynasty. Crucially, the link between them is not founded simply on the note in J\u00f3msv\u00edkinga saga that Thorkell and Sigvaldi were brothers, on which Williams focuses as a potential later invention. That such an invention could have been made becomes increasingly far-fetched the more we 73 \u00d3l\u00e1fs saga helga, ed. Bjarni A\u00f0albjarnarson, 54 (chap. 43). 74 The edition of Sk\u00e1ldatal by Finnur J\u00f3nsson, in his Snorra Edda (on which see 259, 268, and 284) is now superseded. A facsimile of the Uppsala manuscript (De la Gardie 11) has been pub- lished in Snorre Sturlassons Edda, ed. Grape, Kallstenius, and Thorell, and a modern edition and English translation may be found in Snorri Sturluson: The Uppsala Edda, ed. Heimir P\u00e1ls- son and trans. Faulkes, 100\u2013117, esp. 112, 114. 75 On this see Gu\u00f0r\u00fan Nordal, \u201cSk\u00e1ldatal and its Manuscript Context\u201d and Tools of Literacy. 76 The relevant parts of the manuscripts are Reykjav\u00edk, A.M. MS. 761a 4to, fols. 16v\u201317r (paper transcript of ca. 1700) and Uppsala, Universitetsbibliothek, MS. R. 685, fol. 25v (early eigh- teenth-century paper transcript of the Swedish antiquary Peter Salan); I have given their read- ings in Bolton, Empire of Cnut the Great, 206\u20137. See Gu\u00f0r\u00fan Nordal, \u201cSk\u00e1ldatal and its Manuscript Context,\u201d for discussion of this source. On Haraldr \u00de\u00f3rkelsson see Keynes, \u201cCnut\u2019s Earls,\u201d 66. 77 \u00d3l\u00e1fs saga helga, ed. Bjarni A\u00f0albjarnarson, 399 (chap. 239)."]