The Territorialization of Power in the Icelandic Commonwealth

kapittel 5. the territorialization of power in the icelandic commonwealth kapittel 5 The Territorialization of Power in the ­Icelandic Commonwealth Sverrir Jakobsson Introduction The Icelandic historian Gunnar Karlsson (b. 1939) has described Icelandic society before 1262 as being ‘unique in existing without any central power for centuries after Christianity had brought to the country the art of writ- ing on parchment in the Latin alphabet’.1 This pithy statement depicts in a nutshell both the particular and general about the Icelandic Common- wealth, which existed until the advent of state power in 1262–1264, when all parts of the country accepted the overlordship of the king of Norway. It can be assumed that before the development of executive power in the country, Icelandic society was similar to many other communities that also lacked such power structures. Consequently, research into medieval Icelan- dic society during this period may yield conclusions that are applicable and significant to the study of other such societies. On the other hand, one factor not applicable to all pre-state societies is that sources about medi- eval Icelandic society are abundant, meaning that we are able to explore various elements of this culture that usually elude us in studies of other similar societies. Thus, in this concise observation, two important histori- cal phenomena are brought into focus: the introduction of literacy and the development of state power. In many countries these two developed side by side, but not in Iceland. Therefore Icelandic medieval sources are useful for examining the development of state power and the evolution of social discourses concerning power. In this study, I intend to analyse these two themes, looking at their rela- 1 Gunnar Karlsson, Iceland’s 1100 Years: The History of a Marginal Society (London, 2000), p. 1. 101  sverrir jakobsson tionship with space.2 The interaction between power and space is both com- plex and interesting. In modern times, what predominantly ­characterises state power is that it is territorial, and this has been the case for a long time. In fact, territorial power in Europe is generally considered to have originated in the Middle Ages.3 On the other hand, the Old Norse concept of ríki (usually translated as ‘state’) often occurs in medieval Icelandic sources but not in a way that connects it particularly with territorial authority.4 This offers an indication of the different types of power relations that existed in Iceland compared to other societies that contained some elements of state power. In twelfth- and thirteenth-century Norway there existed a territo- rial royal power that was supported by a system tax payments and fines.5 On the other hand, the power of the goðar, the chieftains of the Icelandic Commonwealth, was not territorially defined in this manner, and there are few signs of hierarchies that were related to territorial entities. According to traditional Icelandic law, in the form that was codified in the thirteenth century, there were two entities that might be defined as being territorially demarcated: the quarters (each comprising one fourth of the country) and the communes (ON hreppr). It is noteworthy that neither of these entities were used as basis for power hierarchies. According to the law, chieftains could only have clients within a particular quarter. However, this principle seems to have been disregarded, as in the commonly cited example of the chieftain Jón Loftsson (1124–1197) having parliamentary followers in the Western quarter.6 The communes were assuredly geographically defined 2 This study is based on an earlier version in Icelandic, ‘Rými og vald í í­slensku samfélagi á seinni hluta þjóðveldisaldar’, Heimtur. Afmælisrit til heiðurs Gunnari Karlssyni sjötugum, eds. Guðmundur Jónsson, Helgi Skúli Kjartansson and Vésteinn Ólason (Reykjavík, 2009), pp. 324–39. I want to thank Vala S. Valdimarsdóttir for her assistance in adapting this material into English. 3 See for example Saskia Sassen, Territory, Authority, Rights. From Medieval to Global Assemblages (Princeton & Oxford, 2006), pp. 41–53. 4 Johan Fritzner, Ordbog over det gamle norske sprog, 3 volumes (Kristiania, 1883– 1896), III, pp. 110–11. 5 See for example Sverre Bagge, «Borgerkrig og statsutvikling i Norge i middel­ alderen», Historisk tidsskrift, 65 (1986), 145–97 (especially pp. 165–82). 6 See for example Helgi Skúli Kjartansson, Fjöldi goðorða samkvæmt Grágás. Lecture in a conference, held by The Sigurðar Nordal Institute July 24–26 1988 (Reykjavík, 1989), pp. 15–16; Gunnar Karlsson, Goðamenning. Staða og áhrif goðorðsmanna í þjóðveldi Íslendinga (Reykjavík, 2004), p. 195. 102 kapittel 5. the territorialization of power in the icelandic commonwealth entities, yet power within them seems to have been decentralized, since there were a total of five administrative officers and there is no indication that the power of each officer was geographically defined.7 This changed in the twelfth and thirteenth centuries, but there is no ­general agreement as to the origin of this development.8 What mainly ­characterizes the change is that the power of chieftains–formerly dependent on their personal relationships with their clients–became regional within a territorial unit that modern historians have termed as ‘state/ríki’. These changes were not reflected in the traditional law, which might indicate that they were of a fairly late provenance. In this essay, I intend to illustrate the connection between these political and social changes and the develop- ment of a new discourse of power, which became ever more prominent in the twelfth century. At that time, literacy was introduced into Iceland with profound social consequences. Following the country’s conversion to ­Christianity from the eleventh century onwards, new standards of morality took over, along with a new system of knowledge. A new discourse about space also developed, whereby space became defined with reference to events from the history of Christianity and the saints.9 This affected attitudes to local territorial power structures, since the discourse about foreign hierar- chies had implications for similar hierarchies at home. The discourse about the state of the world had to be translated and provided with meaning in an Icelandic context, but it also influenced the way people understood their own society. A few factors that are related to the discourse about space and power will be examined in particular in this study. The first one concerns spatial discourse in a society in which hierarchies were not territorially defined. What was the nature of regional power in such a society and what were its geographical limitations? Secondly, I will try to analyse the introduction of literacy into Iceland; who benefitted from it and how was it related to both new and pre-existing systems of discourse? Who were the ones that 7 See for example Grágás. Lagasafn íslenska þjóðveldisins, ed. Gunnar Karlsson, Kristján Sveinsson and Mörður Árnason (Reykjavík, 1992), p. 180. 8 See for example Jón Viðar Sigurðsson, Frá goðorðum til ríkja. Þróun goðavalds á 12. og 13. öld (Reykjavík, 1989), pp. 55–60; Gunnar Karlsson, Goðamenning, pp. 280, 286–287, 292–94. 9 Cf. Sverrir Jakobsson, ‘Heaven is a Place on Earth. Church and Sacred Space in 13th century Iceland’, Scandinavian Studies, 82 (2010), 1–20. 103  sverrir jakobsson claimed the power to discuss space and define its meaning? Thirdly, I will discuss the development of regional power in the early thirteenth century when the supporters of chiefs came to be defined in a new way: by regional markers. I will attempt to establish how this new discourse about space and territorial units was related to actual changes in power relations. Finally, I will analyse the regional units of government that had come into existence in the thirteenth century, examining whether one can trace the origin of ­territorial governance in Iceland to this period. In doing so, I hope to offer some general insights into the power structures inherent in the state- formation process, in contrast to other forms of authority. Non-territorial power relations in the Icelandic Commonwealth One of the most important questions in the political and social history of the Icelandic Commonwealth is: what was the nature of the relationships between chieftains and farmers, and what type of power did the chieftains wield?10 An important issue here is the degree to which the power of chieftains was territorial, that is, the role played by the chieftains in their local communi- ties. It has been generally accepted that chieftains were in charge of various affairs in the vicinity of their habitations.11 On the other hand, there are few examples of any specific duties chieftains had towards their neighbours, apart from protecting the community against miscreants that robbed the region’s inhabitants and otherwise behaved disgracefully.12 On the other hand, there are countless examples of chieftains’ duties concern towards their individual followers, farmers who were personally bound to chieftains. Recently, it has been argued that the social group who attended the Thing (the regional assem- bly led by the chieftains) had no duty or inclination to protect all farmers from evildoers, as is evidenced by the fact that the chieftains were comparatively few and unevenly distributed around the country.13 10 Gunnar Karlsson, ‘Goðar og bændur’, Saga, X (1972), 5–57. 11 See Gunnar Karlsson, ‘Goðar og bændur’, pp. 9–18. 12 Gunnar Karlsson, ‘Goðar og bændur’, p. 10. 13 Gunnar Karlsson, Goðamenning, pp. 156–57. 104 kapittel 5. the territorialization of power in the icelandic commonwealth In many ways, recent research by scholars such as Gunnar Karlsson seems to undermine traditional ideas about the local authority of chieftains. It seems evident that the main pillar in the chieftains’ hierarchy of power was their personal relationship with their followers, rather than their status within the region. However, the question is, if this was the case, then what was territorial leadership all about? Here we must look at the concept of the region (ON herað, pl. heruð), which is elusive, as these entities had no basis within traditional law. 14 However, the concept is crucial when one has to value the nature of local government, assuming that it was a possible concern of chieftains. ­According to the law, quarters and communes were separate and fairly well defined entities, but this was not true in the case of the regions. The concept of the herað seems to have had various meanings in the Old Norse language; for example a populated area as opposed to an uninhabited area, or as a rural area as opposed to a village. In translations from Latin it frequently has the same meaning as regio. In Norwegian documents it is used for various entities such as states or administrative districts. In sources, the herað of Oslo is mentioned but also the herað of Telemark, as well as a region in Denmark named ‘Fyn’, but these entities varied greatly in size.15 In Iceland, the meaning of the concept is not well defined. In the law collection Grágás it often has the same meaning as a rural commune (ON hreppr), and this also seems to be the case in the saga of Guðmundur dýri in the Sturlunga collection where the ‘herað of Reykjardalr’ is referred to.16 The events in that saga take place in the last decades of the twelfth century. In Þórðar saga kakala and Þorgils saga, both of which take place in the middle of the thirteenth century, there is, however, mention of heruð in Borgarfjörðr, Skagafjörðr and Eyjafjörðr, which are much larger entities.17 This seems to indicate a development in time, as herað still defined a commune at the end of the twelfth century, but Borgarfjörðr is first called herað as late as in 1247, in connection with Snorri Sturluson’s inheritance. The concept of the ‘regional state’ (Icelandic heraðsríki) is in many ways appropriate for geographically compact areas, governed by a chieftain or a 14 The origin of the word is unknown but it is possibly related control over an army, cf. Ásgeir Blöndal Magnússon, Íslensk orðsifjabók (Reykjavík, 1989), p. 321. 15 See Fritzner, Ordbog over det gamle norske sprog, I, pp. 792–93. 16 See Grágás, p. 534; Sturlunga saga, I, p. 160. 17 See Sturlunga saga, II, pp. 84, 192, 193, 207. 105  sverrir jakobsson lord.18 Such regional states existed in Iceland in the fifth and sixth decades of the thirteenth century, and naturally, the origin of this phenomenon has been researched extensively.19 On the other hand, the study of the concept of the herað itself is difficult due to its manifold meanings. In the law codes that were written down in the last decades of the Commonwealth era one cannot detect a shift in its meaning; in these documents it is synonymous with the hreppr/rural commune when used to define a territorial entity. However, the study of the emergence of this concept is central to this ­current analysis this for two reasons: on the one hand, it represents the first occasion in Iceland where a chieftain–or a ‘foreman’ (ON. formaður) as it is termed in Þorgils saga skarða–enjoys territorial power. Secondly, the so-called ‘regional governance’ (ON heraðsstjórn) is the only example of chieftains involving themselves in issues concerning a territorially defined entity. Therefore, the concept is indisputably connected with the origin of regional power in Iceland. The main difficulties in reaching a scholarly definition of ‘regional ­governance’ are, on the one hand, the question of chieftainships and the ­traditional inclination of scholars to assume that they were regional in nature. On the other hand, the problem is the concept of a regional par- liament, which has influenced scholarly discourse about regions as a phe- nomenon. In Grágás and Íslendingabók we see that in each quarter there were three or four spring parliamentary sessions, and three chieftains were supposed to attend each one.20 It is tempting to see such spring sessions as geographically demarcated entities that were the precursor to the admin- istrative districts (Icelandic sýslur) that were formed in the fifteenth and sixteenth centuries, a sort of regional assembly.21 If this is true, then it is tempting to regard the regions that had formed shortly before the end of 18 Heraðsríki is an idiom coined by modern historians; in ON. the much more vague term ríki would have been used to signify such an area, along with many other things. 19 See for instance Jón Viðar Sigurðsson, Frá goðorðum til ríkja; Gunnar Karlsson, Goðamenning, pp. 271–315. 20 See Íslenzk fornrit I. Íslendingabók, Landnámabók, ed. Jakob Benediktsson ­(Reykjavík, 1968), p. 12; Grágás, p. 413. 21 The concept of the heraðsþing only appears once in Grágás and seems rather to refer to a parliament for the quarter rather than some sort of a spring parliament district, see Grágás, p. 249. 106 kapittel 5. the territorialization of power in the icelandic commonwealth the Icelandic ­Commonwealth era as entities that somehow equalled an older ‘parliamentary district’ (Icelandic þingumdæmi).22 However, such temptations should be resisted. Even though these spring parliamentary sessions were equally spread around the country, there is no indication in contemporary sources that any territorially distinct entities developed around them. A good example of this is the two spring assemblies held in the region of Breiðafjörðr, Þórsnesþing and Þorskafjarðarþing. Our sources about them are fragmentary but they do not indicate that these assemblies served terri- torially demarcated districts. In Gísla saga Súrssonar it is stated that Thorkell the Wealthy from Dýrafjörðr had business across Breiðafjörðr in Þórsnesþing along with the sons of Súr. They accepted an invitation to the home of Þorsteinn the Codbiter, the son of Þórólfr Mostrbeard, and in return they invited the sons of Þorsteinn–Þorgrímr and Börkr the Stout–to come to the spring assembly Hválseyri in Dýrafjörður. Then Þorgrímr settled in Sæból in Haukadalur. According to the saga, ‘Þorgrímr had a chieftainship and the brothers got a lot of support from him.’23 It does not seem to be vital that Þorgrímr’s chieftainship was in Þórsnes at the site of Þórsnesþing. In the saga we are also told that Börkr the Stout, Þorgrímr’s brother, who lived at Helgafell in Þórsnes, went with many of his men across the bay to Þorskafjarðarþing ‘to meet his friends’ there.24 These events were supposed to have happened in the tenth century but the narrative stems from the thirteenth century. In a narrative dealing with more recent events, Þorgils saga ok Hafliða, we are told about Ólafsgildi, a feast held in Reykhólar in 1119, during which the participants discuss the purchase of grain at Þórsnesþing on the other side of the fjord.25 Similar tangles often occur in accounts of parliamentary spring sessions, but these are only problematic if we assume that the parliamentary districts were territorially demarcated. If that assumption is the case, then there is nothing curious about narratives of chieftains from the Western Quarter attending spring assemblies in more than one place. 22 See for example Gunnar Karlsson, Goðamenning, p. 65. 23 Íslenzk fornrit VI. Vestfirðinga sögur, eds. Björn K. Þórólfsson and Guðni Jónsson (Reykjavík, 1943), p. 19. 24 Íslenzk fornrit VI, p. 89. 25 Sturlunga saga, I, p. 27. 107  sverrir jakobsson The next question to consider is what the concept of the heraðsstjórn referred to if it did not indicate a territorially demarcated region controlled by a chieftain. There are least three possibilities. First of all there is an indi- cation that the concept of the herað signified a territorial unit that existed before any lord had claimed overlordship of the area. In this case, regional governance was probably the joint responsibility of several men, one or more chieftains and farmers in the neighbourhood. There are some examples of this in the Sagas of Icelanders (ON Íslendingasögur).26 Another use of the concept of the heraðsstjórn is the leadership of ­chieftains within a certain area, who were therefore in charge of a proto- state in its initial stages of development. This would, for example, apply to the ­authority of Þorgils Oddason (c. 1080–1151) and his son Einar ­Þorgilsson (c. 1120–1185) in the Breiðafjörðr region. According to Þorgils saga ok Hafliða, Þorgils was among the greatest chieftains in Iceland, yet he did not have absolute power in his home region. Thus, one chieftain could be superior to others due to a network of friends and influence at the Althing but his powers were territorially defined just as those of late medieval kings. For example, Þorgils Oddason was regarded as bold when in 1120 he defined the land he possessed as ‘the entire region in which I possess land’.27 In the sagas, it is stated that he owned land in Saurbær and Hvammsdalur, and he possessed some land by Laxá where his protégé Ólafr Hildisson was slain. Yet, when he claims that territorial units other than the farmland belong to his sphere of authority, it is seen as an example of his hubris. A few decades later the authority of Þorgils and his sons in the region had increased and they were in charge of several manors. In 1148 Oddi Þorgils- son, the older son of Þorgils, had settled at the farm Skarð in Skarðsströnd, but his brother Einar Þorgilsson lived in Sælingsdalstunga.28 The rule of Þor- gils and his sons over such manors was short-lived since Skarð soon reverted to its hereditary owner. Before 1148 the law-speaker Snorri Húnbogason (c. 1100–1170) used to live there, and later on the sons of Snorri—Þorgils and Narfi—lived at their hereditary manor.29 Böðvar Barkarson, Þorgils’s 26 See for ex. Íslenzk fornrit VI, p. 141; Íslenzk fornrit VII. Grettis saga Ásmundarso- nar, Bandamanna saga, ed. Guðni Jónsson (Reykjavík, 1936), p. 226. 27 Sturlunga saga, I. p. 32. 28 Sturlunga saga, I, p. 66 29 Sturlunga saga, I, pp. 64, 81. 108 kapittel 5. the territorialization of power in the icelandic commonwealth son-in-law, resided at Sælingsdalstunga after he had sold Hvammur to Sturla Þórðarson in 1150.30 In the spring of 1151 both Þorgils and his son Oddi died, but Einar became ‘a chief since he was supported by many; relatives and friends that his father Þorgils had gathered around him. He also did not lack zeal and bold- ness’.31 Einar inherited his father’s network of supporters, but his future was not glorious since he proved not to be his father’s equal. The hierarchy that was built up by Þorgils in the Dalir region was partly based on his ­personal qualities as a leader; his son could gain an access to his ­supporters but ­nevertheless found it hard to maintain his position if he did not ­demonstrate these same qualities. During Þorgils’ lifetime other leaders in the district seem to have been forced to yield to him; Snorri Húnbogason gave his manor to the son of Þorgils, and it is also stated that Þorgils, in his later years, contrived with his son in killing Skeggi Gamlason, a prominent farmer, and did so with impunity.32 However, when Einar Þorgilsson had become the leader of this group and made the mistake of shielding outlaws in Gilsfjörðr, ‘rumour had it that his regional governance proved to be different from that of Þorgils.’33 Sturla Þórðarson (1115–1183) was one of those who saw this as an opportunity to seize the initiative. He was considered to be ‘a great man, of a good family, and likely to become a lord’ and it became his task to undermine the hierarchy built up by Þorgils and his sons, the system of regional governance.34 When Sturla had become the most powerful chief in these parts, he also tried to define his powers in a territorial manner. Other chiefs had to unite in resistance, but Sturla drove away those farmers in the neighbour- hood who had allowed his competitors to stay overnight.35 In this manner, Sturla strove for regional authority in a manner similar to the way that Þorgils Oddason had done before. However, his hierarchy never became territorially defined since he was always competing with less powerful chieftains. 30 Sturlunga saga, I, p. 68. 31 Ibid. 32 Sturlunga saga, I. pp. 66-67. 33 Sturlunga saga, I. p. 72. 34 Sturlunga saga, I, p. 66. 35 Sturlunga saga, I. p. 96. 109  sverrir jakobsson The third way in which one can define regional governance in spatial terms is to use the concept as defined by law, in connection with the com- munes that were indeed territorially demarcated entities. Then we would have to assume that regional governance could be the provenance of one man and that there might exist proto-states at the level of the commune, even though the governance of communes was divided up between five persons according to the law. One such communal leader would have been Torfi Valbrandsson, who in Landnámabók is depicted as gathering a team in order to kill a band of outlaws.36 Yet such leadership of communes would have been ephemeral in nature, or connected with special projects. One can also contemplate the possibility that the communes were defined as separate entities from the start, as part of the formation of teams of ­settlers led by a special leader.37 This would mean that each inhabited commune was organized during Iceland’s initial settlement period, and that from then on one chieftain (and his descendants) was the leader of each region. Thus, the main difference between the authority of chieftains in the tenth and eleventh centuries on the one hand, and that of the regional lords of the thirteenth century on the other, would be that the older entities were smaller. However, this hypothesis does not fit with the picture of the ­distribution of power that the law clearly draws up. In the Sagas of ­Icelanders there are also quite a few examples of more than one chieftain fighting in order to gain i­nfluence within a territorially distinct region. The ideas of Jón Viðar Sigurðsson (b. 1958) about chieftains being regional leaders of this kind fit with this theory, but his premise is that testimony of the law about ­administration of the Commonwealth is more or less invalid.38 He also takes for granted that authority and hierarchies must always have been regional; that they could not have been personal at any point. From what has been argued before, it can be concluded that authority of chieftains used to be personal in nature before the thirteenth century, based on their relations with their followers: the thingmen. In this context, one can extrapolate the existence of domains or other geographically ­demarcated 36 Íslenzk fornrit, I, p. 75. 37 See for example Jón Viðar Sigurðsson, Chieftains and Power in the Icelandic ­Commonwealth (The Viking Collection. Studies in Northern Civilization, 12), transl. Jean Lundskær-Nielsen (Odense, 1999), pp. 196–97. 38 See for example Gunnar Karlsson, Goðamenning, pp. 40, 155–57. 110 kapittel 5. the territorialization of power in the icelandic commonwealth areas, since the geographical entities within the system–quarters and dis- tricts–had little or no connection with the hierarchies of chieftains. The dominant tendency among modern scholars to assume that there existed parliamentary districts that formed a core of the later regional government does not hold up, since the parliaments were never territorially demarcated. This so-called regional governance was the only duty of a chieftain, which might be territorially defined, but the evidence is far from unambiguous as to whether this really was the obligation of chieftains. Regional governance may have been a joint duty held by the most powerful men in a given area, both chieftains and farmers; it may have been a provisional arrangement for specific tasks, or an obligation that one expected from chiefs d­ ominating a big area who had started forming a proto-state. Until the thirteenth c­ entury, such experiments generally did not last for many generations, since competition, distribution of power and lack of regional authority was built into the law and hierarchies. The Church, literacy and a new discourse about space Literacy began to gain ground in Iceland in the first half of the twelfth cen- tury, and the oldest written sources about the history of Iceland derive from this period. From the nature of the earliest Icelandic texts one can deduce that they mainly belonged to the church, since homilies and translations of saints’ lives are prominent. There was also some production of scientific texts, for example about the position of stars, but such texts were useful for exact calculations regarding the church year. These texts are the earliest sources about the dominant worldview of some Icelanders; their ideas about geography and world history. There is no indication that a large group was using the written word at this time. However, literate texts soon became very influential, since ­knowledge written on parchment was more likely to last than education conveyed through speech. It became possible to refer to a piece of knowledge in the same form, and thus standardize the knowledge that was communicated. This is highly important for the development of the discourse about the most important elements of worldview, such as space. It is disputable whether the church of Iceland was powerful or not in the twelfth century. Until 1190, the division between religious and secular power was vague, but at that point the consecration of chieftains was deemed 111  sverrir jakobsson illegal by the Archbishop of Nidaros. Before this separation of secular and clerical powers, the church could only protect its interests through close cooperation with chieftains, landowners and the most powerful members of the secular class. Yet the advent of literacy bestowed upon clergymen a supreme authority over a new kind of discourse; an authority that can hardly have been very obvious at first. The church of Iceland did not gain an advantage in learning over the secular classes until the written word had become dominant within educated discourse. From the start, the hierarchy of the church marked a shift in the tradi- tional social and cultural relationships that existed in Iceland. To begin with, ecclesiastical authority was territorial, as its main units were territorially demarcated bishoprics and parishes, parallel to quarters and districts within the secular hierarchy. This was different from the authority of chieftains, which was based on personal relations and had no reference to territorially defined unities. The main difference between these two systems was that within the church there was nothing similar to the personal hierarchy of the chieftains. Therefore, the discourses of clergymen about space must have been different to those that existed in secular hierarchies. Yet it is likely that what characterizes written texts from this period is the prevailing discourse that took place in the church, since literacy only became predominant within other hierarchies after the fall of the Commonwealth. In geographical descriptions from the twelfth and thirteenth centuries, there are descriptions of territorial hierarchies in other countries. In the world description of Hauksbók there is a description of ‘where countries lie in the world’, some countries (kingdoms) within each continent are mentioned, and within the countries, there is mention of stórheruð (plur. macroregions).39 This discourse is reflected in other medieval Icelandic texts that also contain descriptions of countries. In the manuscript AM 194, 8vo (from 1387), it is explained that the states have borders and the borders and/or outer limits of Norway are described.40 Yet such descriptions are 39 Hauksbók udgiven efter de arnamagnæanske håndskrifter number 371, 544 and 675, 4to samt forskellige papirshåndskrifter, eds. Eiríkur Jónsson and Finnur Jónsson (Copenhagen, 1892–1896), pp. 153–56. The concept stórherað is used for regions within larger states, such as India or Parthia. 40 Alfræði íslenzk: Islands encyklopædisk litteratur I. Cod.mbr. AM 194, 8vo (Samfund til udgivelse af gammel nordisk littteratur, 37), ed. Kristian Kålund (Copenhagen, 1908), p. 11. 112 kapittel 5. the territorialization of power in the icelandic commonwealth not standardized, and even though units such as storheruð are mentioned in various texts, herað did not have a standardized meaning in g­ eographical literature. In the extensive geographical description of the Old Norse trans- lation of the bible, Stjórn, words such as partur (part) or hluti (portion) are more frequently used than heruð (regions). In this text, there is also the usage of the Latin concept of provincia.41 World descriptions of clergymen give readers insight into the discourse about regional hierarchies in other countries and there standardized concepts were not used, at least not in the thirteenth century. In the section Skáldskaparmál in the Prose Edda there is the following definition of territorial governance: Because a king rules over many countries, he has tributary kings [ON skattkonungar] and earls, who govern with him, and judge according to the law and protect the land against war in distant countries, and their judgments and punishments shall be equal to that of the king. But in one country there are many regions [ON heruð] and it is the way of kings to make governors rule in these regions—in the Danish tongue these men are called hersar or lendir menn but counts in Saxony and barons in England. And they shall be judges and defenders of the region that is under their authority.42 The important thing about this paragraph is that in the thirteenth century ríki and herað evidently meant the same thing. If we accept the com- monly held belief that Snorri Sturluson (1179–1241) is the author of this text, then this is one of the oldest examples of the two being considered identical. Snorri Sturluson’s herað in Borgarfjörðr is one of the oldest examples of a regional district in Iceland, as mentioned before. It is thus hardly a coincidence that Snorri seems to have been interested in systems of regional power in other countries. In the Prose Edda there is a systematic description of the concept of the herað and it is made clear that a herað submits to the authority of one person. 41 Stjórn. Gammelnorsk bibelhistorie fra verdens skabelse til det babyloniske fangen- skab, ed. Carl Rikard Unger (Christiania, 1862), pp. 67–100. 42 Edda Snorra Sturlusonar, ed. Finnur Jónsson (Copenhagen, 1931), p. 161. 113  sverrir jakobsson Since Snorri Sturluson was a courtier of the king of Norway it is tempting to assume that his attempt to coordinate the use of concepts about regional power systems takes into account the situation in Norway and the discourse of power at the royal court. Yet the matter is far from simple. In Heimskringla one can read that landed men [ON lendir menn] were powerful in Norway in the eleventh century ‘since in each district [fylki] it seemed as if lendir menn ruled over farmers’.43 Heruð, on the other hand, were smaller entities and most likely some sort of entities similar to those in Denmark.44 Historians actually believe that there was some sort of fusion of various phenomena since districts [ON fylki] prevailed in Trøndelag and in the western part of Norway, but regions [ON heruð] were the dominant form in Vik and the eastern region. However, in the thirteenth century the concept fylki was in most places used as a synonym for the Latin word provincia.45 It is not clear in Norwegian law whether there is coherence between fylki and lendir menn, i.e. whether each fylki was ruled by one lendr maðr, as seems to be the assumption in Heimskringla.46 In the present context, however, the most important thing is how the concept is used in Icelandic texts such as Heimskringla and the Prose Edda. In these texts it is evident that fylki or herað are seen as territorial entities, under the authority of a lendr maðr. The idea of territorial governance in Norway influenced the Icelanders’ ideas about such administration in Iceland. Snorri Sturluson must have been very interested in the situation of lendir menn, as he became one himself just before he was sent home to Iceland in 1220.47 The question is whether he saw himself as lendr maðr ruling a territorially demarcated region, probably Borgarfjörðr. 43 Íslenzk fornrit XXVII. Heimskringla II, ed. Bjarni Aðalbjarnarson (Copenhagen, 1944), p. 58. 44 Sølvi Bauge, ‘Fylke’, Kulturhistorisk leksikon for nordisk middelalder fra vikingetid til reformationstid V. Frälsebrev-Gästgiveri (Copenhagen, 1960), 39–40, cf. Poul Rasmussen, ‘Herred. Danmark’, Kulturhistorisk leksikon for nordisk middelalder fra vikingetid til reformationstid VI. Gästning-Hovedgård (Copenhagen, 1961), 488–91, Sølvi Bauge Sogner, ‘Herred. Noreg’, op. cit, 492–94. 45 Sølvi Bauge, ‘Fylke’, 39-40. 46 See Arne Bøe, ‘Lendmann’, Kulturhistorisk leksikon for nordisk m ­ iddelalder fra vikingetid til reformationstid X. Kyrkorätt-Ludus de Sancto Canute duce ­(Copenhagen, 1965), 498–505. 47 Sturlunga saga, I, p. 278. 114 kapittel 5. the territorialization of power in the icelandic commonwealth A new discourse of territorialized power seems to have been brewing during this period. From Sturlunga saga we know that those who opposed Snorri’s business were ‘Southerners … and mainly those related to Ormr Jónsson’.48 There is no other mention of men from the quarters in this ­manner in Sturlunga, except in relation to ecclesiastical matters. Indeed, such solidarity amongst men from the Southern quarter does not occur again until in 1238, although in 1231 there is talk of Áverjar (The River People) as a specific group, supporting the family from Oddi.49 The southern chieftain Gissur Þorvaldsson (1209–1268) has a big crowd with him at the Althing in 1229 and 1232 but on neither occasion is it stated that his group belongs to a certain quarter of the country. Yet in 1229 the Austfirðingar (men from the Eastfjords) are grouped together at the Althing, and it is stated that this group had as its leaders the brothers Ormr and Þórarinn, the sons of Jón Sigmundsson. The followers of the chieftain Kolbeinn the Younger (1208–1245) and his kinsman Kolbeinn Sighvatsson (d. 1238) who settled at Snorri Sturlusons’s farm in the South in 1234, are called ‘Northeners’. These men came from Skagafjörðr and Eyjafjörðr. When Sturla Sighvatsson (1199–1238) appropriated for himself Snorri’s possessions in Borgarfjörðr it is also stated that he appropriated for himself ‘the whole region [herað]’.50 In the battles at Bær in 1237 and at Örlygsstaðir in 1238 it is made clear that the fighting groups are named after certain regions. The followers of Sturla Sighvatsson and Sighvatr Sturluson (1170–1238) at Bær are known as ‘the men from the Westfjords’ and some of their followers are called Borgfirðingar (men from Borgarfjörðr) during the battle of Örlygsstaðir. By then the supporters of Gissur Þorvaldsson have become known as Sunnlendingar (Southerners). The followers of Kolbeinn the Younger are called Skagfirðingar (men from Skagafjörðr) in 1231 and again at the battle of Örlygsstaðir.51 Thus, within a relatively short period, regional appellations proliferated in order to make sense of battles between secular groups, whereas until then they hardly ever occur unless in the context of ecclesiastical hierarchies. 48 Ibid. 49 Sturlunga saga, I, p. 345. The River in this case would be Rangá, in the heartland of the region where the family from Oddi held hegemony. 50 Sturlunga saga, I, pp. 333–34, 376–77, 392. 51 Sturlunga saga, I, pp. 405, 417–39. 115  sverrir jakobsson The discourse about territory and power seems to have changed greatly during a short period after 1220. It is tempting to link this difference to a change in political strategies. Chieftains like Snorri Sturluson envisioned a different situation where they could turn their personal leadership as chieftains into territorial power. This was the basis for the regional districts that had formed throughout the country in the fifth and sixth decade of the thirteenth century. Even though regional appellations had begun to apply to secular political rivals in 1238 it is evident that the ensuing civil war between the family of the Sturlungs and their rivals stimulated this development. Political rulers formed permanent groupings and led raids in the region of their rivals. For instance, in the 1240s Kolbeinn the Younger attacked farmers living in the Westfjords and in return his rival Þórður kakali Sighvatsson (1210–1256) conducted raids in the Northern Quarter. In the following decades, there was no turning back from this praxis. In these circumstances, there formed a new rationale for the rule of the Norwegian king in Iceland. While hierarchies were not regional, the king could only have power over individuals, the chieftains that submitted to him. When the regional proto-states [ON heraðsríki] had been formed, however, a few leaders could submit themselves to the king on behalf of their regions. With this in mind the structure of the events in Iceland from 1262–1264, when men from various parts of the country submitted to the king of Norway, becomes easy to explain. In the Northern Quarter, three farmers from Skagafjörðr and three from Eyjafjörðr took an oath, but other parts of the quarter seem not to be defined as entities. In the Westfjords five chieftains, each with three farmers, all took an oath. The most remarkable example is that of the Borgfirðingar who had no chief, but nevertheless three farmers took an oath on their behalf.52 Borgarfjörðr had become a territorial unit, regardless of lordship in that area. If one tries to define the development from personal to territorial power within a temporal framework, there seems to be a turning point with the forming of territorial ecclesiastical hierarchies in the eleventh and the twelfth centuries. The discourse about the constitution of the church was ­different from that of secular power, since the hierarchy of the church was ­territorial, modelled on international institutions. As the discourse of the church became more independent in the twelfth and thirteenth centuries, 52 See Sturlunga saga, II, pp. 281–82. 116 kapittel 5. the territorialization of power in the icelandic commonwealth the more it influenced society. In this respect the separation of secular and ­ecclesiastical power in 1190 must have been a turning point. After this, regional dominions [heraðríki] formed more rapidly. The clearest example of this is the state of the Sturlungar in Breiðafjörðr, as their main rivals for power in the region became servants of the church and thus no longer suited for secular power. Amongst these were, for example, the descendants of Einar Þorgilsson and family at Skarð. As a result of this, a power vacuum appeared that ensured the Sturlungar the leadership amongst chieftains in these parts.53 This fusion of power did not entail a shift in the legal basis of power, which remained the same throughout the Commonwealth period. Yet territorialization became imminent partly through ecclesiastical and foreign influences passed on through the medium of literacy. One can view the descriptions of regional governance in the Prose Edda and Heimskringla as an attempt to add new dimensions to the discourse of power and place regional hierarchies on the agenda. It is hardly a coincidence that both these works are connected to the first Icelander who came to Iceland as a lendr maðr–the subtle politician Snorri Sturluson. With this in mind, it is indeed ironic that in 1262 Snorri’s district of Borgarförðr had become an anomaly among the regional proto-states, as it did not have a chief. Conclusion The unique position of Icelandic society regarding the development of state power has often been commented upon, but in this study, I have tried to draw out further ramifications of this uniqueness. It is clear that the absence of state power, as demonstrated by the law codes, sagas and contemporary sources, was indeed very different from the social situation in most Euro- pean states in the Middle Ages. What mainly characterizes power relations in the Icelandic Commonwealth is that hierarchies were built on personal relationships, not territorialized. The rule of chieftains extended to their followers, the thingmen, but this rule was not geographically demarcated. 53 See Sverrir Jakobsson, ‘The Process of State-formation in Medieval Iceland. The Process of State-Formation in Medieval Iceland’, Viator. Journal of Medieval and Renaissance Studies, 40:2 (Autumn 2009), 151–70. 117  sverrir jakobsson There existed entities, demarcated in space, such as communes and quarters, but they were not related to hierarchies. In the thirteenth century, a new system, built on territorialized space replaced the old system, mainly based on personal associations. Many events led to this development; one important factor is that the Icelanders started discussing spaces and hierarchies in a different way. I have argued that a dif- ferent discourse about regional hierarchies was mainly connected with two factors developing side by side. One was the initiation of an ecclesiastical power structure that was different from secular hierarchies, resting on a ter- ritorial basis. The other was the initiation of literacy, which was important since with it the nature of discourse changed, giving knowledge a more per- manent shape a more standardised method of dissemination. In geographi- cal texts and written sources of various kinds, one can find a standardised discourse about regional hierarchies around the world, where concepts such as the herað could gain a new provocative meaning. An example of this is the description in the Prose Edda of hierarchies in kingdoms, which could easily be compared to the circumstances in medieval Iceland, clarifying the ambitions of chieftains to form regional hierarchies. In the Prose Edda there are descriptions of regional hierarchies abroad, which stemmed from the period during which Snorri Sturluson attempted to create such a structure around himself in Iceland. Later, other chiefs followed his example with more success. Around the year 1220, a development occurred that proved to be a ­turning point, namely that lords throughout the country began to define their powers as territorial. This shift in thinking began to gain momentum after 1238, and the followers of such lords also eventually came to regard themselves differently; the groups that supported certain chiefs gradually became identified with reference to space. In the fifth decade of the ­thirteenth century, the followers of Kolbeinn the Younger were called ‘Northeners’, those following Gissur Þorvaldsson were known as ‘Southerners’ and those of Þórðr kakali were called ‘men from the Westfjords’. The forming of regional groupings, such as these, was a novelty at the time, and there are few traces of them in contemporary sources until around the time of the battle at Örlygsstaðir. The systematic change that this led to occurred before the Icelanders became subject to the king of Norway, but it was this same transformation that gave territorial authority a more permanent form. 118