H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 64 Introduction The last decade has seen an increasing interest in assemblies in northwestern Europe. Recent publications include studies of assemblies in Ireland (e.g., FitzPatrick 2004, Warner 2004), Scotland (e.g., Driscoll 2004, O’Grady 2008, Sanmark 2013 [this volume]), England (e.g., Baker et al. 2011, Pantos 2004, Semple 2004), and the Nordic countries (e.g., Brink 2004, Friðriksson et al. 2005, Sanmark 2009, Vésteinsson et al. 2004). While there is a long history of discussion of assemblies by historians, placename specialists, and others, the more recent contributions also include archaeological approaches. Several of these studies have been published in two compilations (Barnwell and Mostert 2003, Pantos and Semple 2004), drawing attention to important common features with regard to localization and physical characteristics of assembly sites (Pantos and Semple 2004:18–19). However, variation among assembly sites has also been emphasized (Sanmark and Semple 2008:256). When Nordic assembly sites are discussed, the sites in question are with few exceptions Swedish and Icelandic. This geographical focus is natural given that most of the recent contributions from the region are based on material from Sweden and Iceland (but see Jørgensen et al. 2010, Sanmark 2010), although it may not represent the total picture of Scandinavian assembly sites. In light of the influence of Danes and Norwegians on the British Isles and the general Northern Atlantic region in the Viking and Middle Ages, improved knowledge of assemblies—and not least, knowledge of the social organizations the assemblies were part of—within these countries should be of interest even outside the national context. My PhD project “Borders, Assemblies, and Power: The Administrative Landscapes of Western Norway AD 800–1400” aims to identify assembly (thing) sites in central western Norway, and to study the development of administrative landscapes in that region during the transition from pre-state societies to kingdom/state. Relatively little is known about Norwegian assembly sites (cf. Ødegaard 2013 [this volume]), and in addition to identification of sites, an important aim of the project is to attempt to understand their place and roles in the administrative structure of the medieval law province. An improved knowledge of the territorial units of the administrative organization, particularly with regard to the local-level units, is prerequisite for this, thus necessitating extensive use of written sources to establish an overview of the administrative structure of the region. While I am still in the process of investigating the geographical units involved in rule and administration, the results so far seem to indicate that some areas deviate from the expected pattern. One of these areas is Hardanger, where the medieval local-level units at present are poorly understood, and it is this area that is the focus here. With the aim of improving the knowledge of early local administrative units, this article presents a retrogressive method combining variously dated written sources and relevant archaeological material (i.e., remains of large boat houses) as a possible way to trace older administrative districts and landscapes. Hardanger Contextualized: Main Features of the Gulathing Law Province Hardanger belonged to the law province of Gulathing in the medieval period, held to have been established in the early 10th century A.D. (Helle 2001:27). The Gulathing Law, known in its written form from the mid-13th century although parts Tracing Medieval Administrative Systems: Hardanger, Western Norway Halldis Hobæk* Abstract - As part of an ongoing Ph.D. project focusing on medieval assembly sites in western Norway, this paper discusses administrative units in the Hardanger district. An improved understanding of the medieval administrative organization, particularly local-level units, is of vital importance for the study of assembly sites. Using Hardanger as a case study, this paper discusses how reconstruction of older administrative units can be attempted by a detailed retrogressive study combining written sources and archaeological material. Although preliminary, the investigations indicate that in some parts of western Norway, among them Hardanger, the administrative structure differed from the general model described for the region. This finding implies a larger degree of heterogeneity in social and political organization within the medieval Norwegian kingdom than hitherto assumed. Debating the Thing in the North I: The Assembly Project Journal of the North Atlantic *Museum of Cultural History, University of Oslo, Pb. 6762, St. Olavs Plass, 0130 Oslo, Norway; Halldis.Hobak@khm.uio.no. 2013 Special Volume 5:64–75 H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 65 of the content are held to be considerably older, mentions four territorial units relating to the secular organization of the region: fylki, fjórðungr, áttungr, and skipreiða (see Eithun et al. 1994, Robberstad 1969).1 The original core area of the Gulathing law province comprised three medieval fylkir: Horðafylki, Sygnafylki, and Firðafylki (Fig. 1). Fjórðungr and áttungr mean “one fourth” and “one eighth”, respectively, and represent subdivisions of the fylki within the Gulathing law province (Helle 2001:76–80 with references). The fjórðungr and áttungr are less well understood than the fylki. The áttungr does not seem to be mentioned in any other source than the law text,2 and it has been questioned if this unit ever existed as more than a normative attempt to build a systematically structured judicial system. However, there is general agreement that Sunnhordland, Nordhordland, Voss, and Hardanger constituted the fjórðungar of Horðafylki in the Middle Ages (Helle 2001:77). The skipreiður were territorial units that originally related to the naval defence system called the leiðangr, where each skipreiða was obliged to build and supply a war ship (Bjørkvik 1970:546). The leiðangr is generally held to have been organized by King Haakon the Good in the 10th century (e.g., Ersland 2000:42–53), and according to Snorri Sturluson’s rendering from the 13th century, the king introduced the skipreiða system simultaneously (Hødnebø and Magerøy 1979:93). This description was widely accepted by historians in the 19th and early 20th centuries, while more recent studies have argued that the introduction and organization of the leiðangr varied between regions. However, there is agreement that skipreiður were first introduced in the Gulathing law province, probably by King Haakon in the 10th century, from where the system was expanded to the entire kingdom by ca. 1200 (Andersen 1977:264–265, Ersland 2000:48–52). An early dating of the skipreiða system in western Norway is also to some degree supported by archaeological studies (Myhre 1985, 1997). In addition to defensive purposes, the skipreiður soon gained important fiscal and administrative functions as taxation units, where the contribution of a ship in times of war was extended to a regular tax by the end of the 12th century (Andersen 1977:268– 269). As part of this development, the term leiðangr came to mean both the naval defence and matters related to it (e.g., leiðangsskip “ship belonging to the leiðangr”) as well as an annual tax. The skipreiður also functioned as the lowest level in the judicial system (Imsen 1974), and the expected administrative structure within the Gulathing law province can therefore be sketched out as in Figure 2. As the lowest level of administrative unit, the skipreiður are of vital importance for assessing the level and scope of assembly sites. However, the spatial aspect of these units is an under-researched topic, and the only study dealing specifically with the subject contains only coarse maps and overviews (Bull 1920). Attempting to improve the knowledge of the medieval skipreiður in the research area is therefore a priority for my project. Because the judicial and fiscal functions of the skipreiður were retained well into early modern times, various written sources mention skipreiða units within the Gulathing law province. A cadaster from 1647 (Fladby and Marthinsen 1995, Fladby and Winge 1995) is the oldest source providing systematic data from the entire research area, and contains sufficient information to map the skipreiður in relative detail. Added to this is information gleaned from the corpus of Norwegian medieval documents, where about half of the skipreiður in the research area are mentioned, and tax and tithe rolls from the 1520s and 1560s.3 Comparison between the early modern sources and the various medieval references provides a basis for tracing the medieval skipreiða organization of the Gulathing law province. This approach, with retrogressive use of various written sources, entails a wide array of challenges regarding both sources and inferences based on them, but enables an outline of the medieval local administrative organization as well as assessment of stability and changes in the skipreiða structure that cannot be attained otherwise. However, the tracing of skipreiður has proved difficult in Voss and Hardanger, the two eastern fjórðungar of Horðafylki (Fig. 1). Both these areas have relatively high numbers of preserved medieval sources (Imsen 1990:42–43), especially compared to the coastal areas of the fylki where the skipreiða units are well known. Despite this, the term skipreiða does not seem to occur at all in Voss, and it is first known from Hardanger in sources from the late 15th century (see below). The problems with identifying the expected lower levels of the administrative system in these areas have necessitated an in-depth investigation. The situation in Hardanger is discussed in detail in the following section. Case Study: Tracing the Local Administrative Units of Hardanger Hardanger is a mountainous region, with settlement mainly along the branching fjord that provides connection between the settled areas and access to the neighboring regions (Figs. 1, 3). The starting point for most works on the administrative H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 66 Figure 1. The three medieval fylkir in the research area, with the fjorðungr division of Horðafylki indicated. The inserted map shows the research area’s location in Norway. Map underlay © Kartverket. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 67 structures in Hardanger has been the will of King Magnus the Lawmender from 1277 (DN 2 no. 3; Lange and Unger 1858:3–4). This is the oldest source enumerating the skipreiður in Norway, and states that Horðafylki comprised 32 skipreiður. Twentytwo skipreiður in Nordhordland and Sunnhordland are known from historical sources (medieval letters mention nine of them, all 22 figure frequently in various sources from ca. 1518 and onwards). By arithmetical logic it has been assumed that the remaining ten were situated in Voss and Hardanger, five in each fjórðungr.4 The absence of skipreiður in medieval sources and the unclear picture emerging Figure 2. Overview of the medieval administrative structure within the Gulathing law province, based on the Gulathing Law and recent historical studies. Figure 3. Ten settlement areas in Hardanger are frequently mentioned in medieval and later sources. All ten were medieval parishes. Map underlay © Kartverket. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 68 from later sources has been explained as being because the skipreiða-organization was not properly established. That might actually be the case, but as a fjórðungr is considered too large a district for local assemblies (Helle 2001:78–79), Hardanger must be assumed to have had some kind of administrative organization on a smaller scale. To clarify the administrative structure of Hardanger, the main written sources available5 have been examined to see how areas are designated (skipreiða, parish, tinget—see below), which administrative units are mentioned when, and which areas individual farms were located in. Emerging from these sources are ten smaller communities occurring in medieval as well as early modern records (Fig. 3). All ten areas were parishes, and in each case, use of the term parish is contemporary with use of the areaname without any designation at all. Whether parish was attached to the names or not does not seem to have been important, and this is probably an indication that the extent of the parishes coincided with established communities with well-known borders. The earliest reference to the noun skipreiða found within Hardanger is as late as 1462 (DN 8 no. 375; Unger and Huitfeldt 1874:401–402). The document in which it is mentioned concerns the sale of a farm in “Hardhen skypredhæ”, “Hardhen” being the farm Håra in Røldal. Røldal is a small mountain community located inland between western and eastern Norway, and on these grounds the idea of Røldal as a medieval skipreiða has been rejected (Olafsen 1920:127). Asgaut Steinnes has argued similarly, adding that the entire leiðangr-contribution from Røldal in the 17th century was only two pund of butter (approx. 10 kg). It thus seems out of the question that Røldal could have had the economic capacity to be responsible for a warship of its own, and Steinnes (1933:148) suggested that the designation skipreiða might simply be a mistake (but cf. below). The next area designated skipreiða is Kinsarvik in 1472. Then, from the 1550s to ca. 1630, three communities (Kinsarvik, Ullensvang, and Øystese) are called skipreiður in documents, while only Jondal is designated skipreiða in tax and tithe rolls from 1563 and 1567.6 In a cadaster from 1646, all four areas are called skipreiður. Only a year later, six skipreiður (Granvin and Ulvik in addition to the former) are listed.7 The changes do not seem to be the result of omissions in either the 1563–1567 rolls or the 1646 cadaster, as the six communities not designated skipreiða are mentioned several times as parishes and without designation. The four areas called skipreiður are also mentioned as parishes and without designation. It is tempting to assume that similarly to the situation found for the designation parish, the ambiguous use of skipreiða ca. 1470 to 1647 reflects that the communities and their borders in this period were so well known that the administrative designations did not matter. However, a closer scrutiny of the sources contradicts this assumption: the 1472 reference is explicit that Kinsarvik skipreiða encompassed Odda parish, and the border between Jondal and Øystese skipreiður seems less than stable in the 16th century.8 The inconsistent use of administrative designations therefore seems to at least in part reflect factual changes in administrative structure during the period, but what could be the cause for this instability? Thing districts? A registry of an extraordinary tax paid from Hardanger in 1521 represents another puzzle, using a designation not known from any other source: eight of the areas are called -tinget, i.e., the thing (NRJ vol. III ; Huitfeldt-Kaas 1905:133–156). While noted by several authors, to my knowledge no attempt has been made to clarify what this might imply (e.g., Olafsen 1920:127, Steinnes 1933:148). The -tinget entries are used as headings, followed by lists of tax payers located at farms, and clearly designate geographical areas. The term -tinget is difficult to interpret in this context as anything other than a geographical area with judicial as well as administrative functions. Reconstruction of the -tinget districts by reference to the individual farms named as belonging to them shows that all but one correspond to parishes. The exception is Øystese-tinget, which encompassed two parishes (Figs. 3, 4). In a study of administrative systems in southwestern Norway, Steinnes (1974:52–55) found that the same areas were alternately called skipreiða, tinghá, and tingstad in the 15th and 16th centuries (the latter two meaning “thing district” and “thing site”, respectively). For Trøndelag in central Norway, Audun Dybdahl (1997:215) has demonstrated that tinglag (“thing district”) replaced the term skipreiða from the 15th century. Covering regions south and north of the research area, these studies open up the possibility of interpreting the eight areas called -tinget in Hardanger as skipreiður. Could they be the only recorded traces of the medieval skipreiður that later were merged into larger units? A study by Steinnes in the early 1930s offers valuable insights for assessing whether these eight areas could have functioned as skipreiður. Steinnes (1933:150–160) was able to show that the average leiðangr paid by each of the better-known skipreiður H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 69 in the Viking Period are uncertain, and while the sources for calculating population size in the medieval period are better, the estimates are still disputed (Øye 2002:252). Nevertheless, an attempt to assess the number of medieval skipreiður that the population of Hardanger could reasonably be expected to have supported suggests five as a maximum (Liland 1951:42). This figure should be seen as a qualified guess rather than an accurate estimate, but it corresponds with the calculations based on agrarian production capacity. Five ships thus seem to be a reasonable contribution from the area. Following this argument, it does not seem likely that the eight tinget-areas were medieval skipreiður—they were, in terms of economy and population, probably too small to have had this function. A more probable explanation for these eight areas as judicial-administrative units is that they constituted districts for heraðsþing, local assemblies. The existence of such a local level has been presupposed in Sunnhordland and Nordhordland corresponds closely to one fifth of the total leiðangr paid by the Hardanger region. Thus, the total contribution from Hardanger equalled the contribution from five units responsible for one ship each in the outer parts of Horðafylki. King Magnus the Lawmender’s law from 1274 stated that leiðangr should be calculated according to the economic capacity of the farms within the contributing area (Landsloven III 6:1; Taranger 1979:32–33; see also Dybdahl 2001:11–12). This implies that the economic capacity of the eight tinget-districts could be estimated to have equalled that of five average skipreiður in Horðafylki, and if they as potential medieval skipreiður should have shouldered construction and maintenance of eight war ships, the burden seems unreasonably heavy. Prior to 1274, the Gulathing Law stated that the contribution to the leiðangr was to be calculated according to population size (G 296–298; Eithun et al. 1994:160–162). Estimates of population size Figure 4. The eight areas with names ending in -tinget in the tax roll from 1521. Øystese-tinget comprised two parishes. Røldal is not mentioned in the tax roll, but there are indications that this area could also have functioned as a judicialadministrative unit in the early 16th century. Map underlay © Kartverket. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 70 Grimm 2004, 2006; Myhre 1985, 1997, with references), and eight such boat houses are known from Hardanger (Grimm 2006:414, Myhre 1997:170). Bjørn Myhre (1987:38–40, 1997:176–178, 182) has found that the distribution of medieval boat houses in western Norway corresponds well with the expected pattern of the leiðangr organization. Based on the remains of large boat houses, Oliver Grimm (2004:53–54, 2006:126–128, 218, 293) suggested that there were seven skipreiður in Hardanger (see also Stylegar and Grimm 2003:95–97). These studies are most valuable, as they bring archaeological material firmly into the discussion of the medieval skipreiða organization. The problem, however, is how to tell the boat houses that belonged to the naval defence from the ones that did not. The question is addressed by Myhre (1997:176), who uses the location of boat houses at or close to skipreiða- centers11 as a criterion for their identification as leiðangr boat houses. While resulting in sound identifications in areas where the skipreiða organization is better known, this approach is, as Grimm (2004:54, 2006:126–128, 218) pointed out, less certain in Hardanger because the written sources are late and might not reflect the medieval situation. Within Hardanger, the criterion of spatial proximity to skipreiða-centers in practice translates to proximity to central areas in the 17th-century skipreiður/parish churches. The seven suggested skipreiður would, whether assessing medieval population size or agrarian production capacity, be very unevenly sized. Seven also seems a high number of ships to be contributed from Hardanger, in view of the arguments by Liland and Steinnes cited above. The best-known of the Hardanger boat houses is Skiparsto in Kinsarvik (Fig. 4), 30 m long and 16 m wide (Fett 1954a:34). It is dated to the medieval period12, and generally held to belong to the leiðangr (e.g., Christie 1986:68, Ersland 2000:53, Grimm 2006:414, Myhre 1997:176). Another of the large boat houses is situated in the western part of Hardanger, in Skipadalen at Valland, Kvam (Fig. 4). This boat house has been associated with the leiðangr (Olafsen 1920:127) but is not identified as a skipreiða boat house in later works (Grimm 2006:414, Myhre 1997:178), apparently because of its peripheral location in Vikøy as well as its distance from parish churches. Yet the boat house in Skipadalen is not only of the same type as Skiparsto with solid stone-built walls, but—measuring 32 by 15.5 m—it is also comparable in size (Fett 1954b:12). A limited archaeological excavation has dated the boat house to the medieval period (Fett 1954b:12, Grimm 2006:414, Myhre 1997:172). (Helle 2001:78), but the geographical units associated with it have not hitherto, to my knowledge, been identified in the research area. This interpretation would also explain the reference to Røldal as a skipreiða in 1462 (DN 8 no. 375; Unger and Huitfeldt 1874:401–402); the document was issued in Stavanger law province, and if Røldal functioned as the district for a heraðsþing, it is understandable that the writer used the familiar designation of local-level judicial-administrative units within his own law province (i.e., skipreiða). A letter from 1501 mentions “rettom tingstad i Røgædal”, “the correct assembly site in Røldal” (DN 2 no. 1016; Lange and Unger 1851:747–748), and this strengthens the argument that Røldal might indeed have functioned as a local judicial unit.9 Steinnes (1933:149, 156) found a correspondence between leiðangr contributions and the number of skipreiður within the regions he studied, and concluded that his calculations showed that there actually were five skipreiður in Hardanger. This logic seems a more substantial argument than the arithmetical approach mentioned above, but it leaves the absence of the designation skipreiða in the rich medieval source material of the region unexplained. Steinnes himself pointed out what he saw as indications that the skipreiða-organization was still being shaped in the 16th and 17th centuries, among them their very different economic capacities, with Øystese skipreiða of the early 17th century paying three times as much leiðangr as Granvin did. Furthermore, the borders of the skipreiður in the outer regions of Horðafylki in several cases cross the borders of the presumably younger parishes, while the borders of skipreiður in Hardanger always coincide with parish borders (Steinnes 1933:148). In summary, the designation skipreiða is not found in Hardanger until the late medieval period, and when it appears there are indications that the geographical units associated with it were not an ancient, stable form of organization in the area. Yet 12th-century accounts explicitly describe Hardanger’s contribution to the leiðangr,10 so how were Hardanger’s obligations towards the naval defence organized? The Skipreiða Organization in Light of Archaeological Material Archaeological material may have the potential to shed light on this question. The remains of large boat houses intended to shelter warships have been used to investigate naval organization in the Iron Age as well as the medieval period (e.g., H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 71 As noted by Edvard Bull (1917:21) the boat house in Skipadalen is located at the border between Vikøy and Øystese parishes, the two communities that constituted Øystese-tinget in 1521. While peripheral in each of the parishes, the location is very central for the joint tinget-district (Fig. 4). Rather than a reason to exclude this boat house from the leiðangr system, might its location instead provide a clue to the spatial organization of the leiðangr? Could the two communities constituting a judicial unit also have been a unit with military obligations? However, it might be worth considering if a possible district for the Skipadalen boat house could have been even larger. Farms in Vikøy are listed under Øystese-tinget in 1521, and Vikøy clearly was part of Øystese skipreiða in 1647. Yet the Vikøy farms listed under Øystese-tinget in 1521 belonged to Jondal skipreiða in 156313, a notable deviation from Vikøy’s strong connections with Øystese in all other regards. If the communities constituting Jondal-tinget and Øystese-tinget in 1521 held their responsibilities for naval defence jointly in the medieval period, the Skipadalen boat house would still have a central location for this (so far hypothetical) unit (Fig. 4). Could a division of such a large, medieval skipreiða explain Vikøy’s affiliation with Jondal in 1563? This line of reasoning leads over to questions relating to the assumed judicial functions of the skipreiður. The eight tinget-districts are only mentioned in the early part of the 16th century, but if they go back to the medieval period as judicial-administrative units, they would presumably represent a level below the skipreiður. If so, there would be one level more in the judicial structure of Hardanger than expected (cf. Fig. 2). As Hardanger in the medieval period was smaller both in population size and geographical extent than the coastal fjórðungar of Hordafylki, it is not immediately evident why an extra judicial level would be needed. Is it possible that while the tinget-districts functioned as the lowest judicial level, the skipreiður in Hardanger did not gain judicial functions, but had military functions only? If so, this could explain the situation in the written sources, with the apparent instability of the skipreiður reflecting changes in functions and re-structuring as they were transformed into judicial- administrative units. Conclusions For the questions raised above to become anything more than speculations, further research on the skipreiða organization of Hardanger is needed. There are, however, already indications that it differed from the western fjórðungar of Horðafylki, and archaeological material (i.e., the boat houses) might hold the key to clarifying some of these issues. The possibility of medieval districts for heraðsþing, local assemblies, also hints at important deviations from the general administrative model described for western Norway. This possibility of an alternative structure has consequences for the work on identifying assembly sites in the region. If, instead of the five hitherto expected skipreiður as the lowest judicial-administrative level, there were eight14 smaller units in Hardanger, the implication is that there is a minimum not of five, but eight medieval assembly sites awaiting identification. Methodologically, the detailed study of various written sources and the combination of these seem a promising way forward to reconstruct older administrative landscapes. By comparing the glimpses of the chronologically different situations gleaned from the written sources, and in particular when considering them in light of the archaeological material, it is possible to ask new questions of the data. Attempts at answering these questions will hopefully result in improved knowledge of the geographical aspects of medieval rule and administration in western Norway. The Norwegian kingdom is held to have been established in and expanded from western Norway in the Viking Age (Krag 2008:646–647), and Hardanger was thereby located within one of the core areas of royal power as well as the core area of the Gulathing law province. This geographic centrality lends particular interest to the preliminary findings presented above. The significant deviations regarding judicial-administrative functions and spatial organization of the skipreiður found within Hardanger seem to suggest that the organization of pivotal structures and institutions in the medieval kingdom varied not just between its main regions; the organization and development of administrative systems within the law province may also have been more heterogeneous than previously assumed. This variation in social and political organization in medieval Norway may also be of relevance for studies of administrative structures and systems of governance in other parts of the North Atlantic region. Acknowledgments I would like to express my gratitude to Frode Iversen, Ingvild Øye, and Alexandra Sanmark for help and advice in the preparation of this article, as well as to guest editor John Baker and two anonymous reviewers for their very useful comments. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 72 FitzPatrick, E. 2004. Royal inauguration mounds in medieval Ireland: Antique landscape and tradition. Pp. 44–72, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Fladby, R., and L. Marthinsen (Eds.) 1995 (2nd Edition). Skattematrikkelen 1647. Vol. XII. Sogn og Fjordane Fylke. Norsk Lokalhistorisk Institutt, Oslo, Norway. 452 pp. Fladby, R., and H. Winge (Eds.). 1995 (2nd Edition). Skattematrikkelen 1647. Vol. XI. Hordaland Fylke. Norsk Lokalhistorisk Institutt, Oslo, Norway. 457 pp. Friðriksson, A., H.M. Roberts, G. Guðmundsson, G.A. Gísladóttir, M.Á. Sigurgeirsson, and B. Damiata 2005. Þingvellir og þinghald að fornu. Framvinduskýrsla 2005. FS 307—02143. Fornleifastofnun Íslands, Reykjavík, Iceland. 34 pp. Grimm, O. 2004. Schiffsbezirke. Pp. 52–55, In J. Hoops (Ed.) 2004. Reallexikon der Germanischen Altertumskunde, vol. XXVII. Walter de Gruyter, Berlin, Germany. Grimm, O. 2006. Großbootshaus—Zentrum und Herrschaft. Walter de Gruyter, Berlin, Germany. 448 pp. Gundersen, D. (Ed.). 1996. Sverres saga. Norges nasjonallitteratur, bind 1. Gyldendal Norsk Forlag, Oslo, Norway. 240 pp. Helle, K. 2001. Gulatinget og Gulatingslova. Skald, Leikanger, Norway. 240 pp. Huitfeldt-Kaas, H.J. (Ed.) 1905. Norske Regnskaber og Jordebøger fra det 16de Aarhundrede. Vol III. Det Norske historiske Kildeskriftfond, Christiania, Norway. 830 pp. Hødnebø, F., and H. Magerøy (Eds.). 1979. Snorres kongesagaer 1. Gyldendal, Oslo, Norway. 344 pp. Imsen, S. 1974. Skipreide. Pp. 298, In R. Fladby, S. Imsen, and H. Winge (Eds.). 1974. Norsk historisk leksikon. Næringliv, rettsvesen, administrasjon, mynt, mål og vekt, militære forhold, byggeskikk m.m. 1500–1850. J.W.Cappelens Forlag, Oslo, Norway. 368 pp. Imsen, S. 1990. Norsk bondekommunalisme fra Magnus Lagabøte til Kristian Kvart. Del 1. Middelalderen. Tapir Forlag, Trondheim, Norway. 226 pp. Jansen, J. (Ed.). 1938. Norske lensrekneskapsbøker 1548–1567. Vol. III. Skatten av Bergenhus len 1563. Riksarkivet, Oslo, Norway. 199 pp. Jansen, J. (Ed.). 1941. Norske lensrekneskapsbøker 1548–1567. Vol. IV. Rekneskap for Bergenhus len 1566–1567. Riksarkivet, Oslo, Norway. 212 pp. Johannessen, O.-J. (Ed.) 2000. Jordebok for Bergen Bispedømme ca. 1600. Med Jordebok for hospitalet i Bergen ca. 1617–21. Norske kyrkjelege jordebøker etter reformasjonen vol. 3. Kjeldeskriftfondet /Riksarkivet, Oslo, Norway. 299 pp. Jørgensen, A.N., L. Jørgensen, and L.G. Thomsen. 2010. Assembly sites for cults, markets, jurisdiction and social relations. Historic-ethnological analogy between North Scandinavian church towns, Old Norse assembly sites and pit house sites of late Iron Age and Viking Period. Pp. 95–109, In L. Boye, P. Ethelberg, L. Heidemann Lutz, S. Kleingärtner, P. Kruse, L. Matthes and A.B. Sørensen (Eds.). Arkæologi i Slesvig/ Archäologie in Schleswig: Sonderband “Det 61. Internationale Sachsensymposion 2010” Haderslev, Danmark. Wachholtz, Neumünster, Germany. 332 pp. Literature Cited Andersen, P.S. 1977. Samlingen av Norge og kristningen av landet. 800–1130. Universitetsforlaget, Oslo, Norway. 382 pp. Baker, J., S. Brookes, and A. Reynolds. 2011. The law of the land: Finding early medieval assembly sites. British Archaeology 120:46–49. Barnwell, P.S., and M. Mostert (Eds.). 2003. Political Assemblies in the Earlier Middle Ages. Brepols, Turnhout, Belgium. 213 pp. Bjørkvik, H. 1970. Skipreide. Pp. 546–551, In F. Hødnebø (Ed.) Kulturhistorisk leksikon for nordisk middelalder fra vikingtid til reformasjonstid. Vol. XV. Gyldendal Norsk Forlag, Oslo, Norway. 724 pp. Brink, S. 2004. Legal assembly sites in early Scandinavia. Pp. 205–216, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Bull, E. 1917. Skipstomter og naustetomter fra vikingetid eller middelalder. Foreningen til Norske Fortidsmindesmærkers Bevaring. Aarsberetning for 1917:17–23. Bull, E. 1920. Leding. Militær- og finansforfatning i Norge i ældre tid. Steenske Forlag, Kristiania, Norway. 175 pp. Christie, H. 1986. Leidangsmateriell på kirkeloftet. Hikuin 12:67–72. Dalen, K., and A. Dalen 1960. Røldal bygdebok. Røldal kommune, Røldal, Norway. 563 pp. Driscoll, S.T. 2004. The archaeological context of assembly in early medieval Scotland: Scone and its comparanda. Pp. 73–94, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Dybdahl, A. 1997. Skipreider og fylker i Trøndelag. Historisk Tidsskrift 76(2):211–248. Dybdahl, A. 2001. Leidangen i Frostatingslagen og Trondheim len. Skattelisten 1590/91 i en historisk kontekst. Med noen kompletteringer. Tapir Akademisk Forlag, Trondheim, Norway. 167 pp. Døssland, A. 1990. Med lengt mot havet. Fylkeshistorie for Møre og Romsdal I. 1671–1835. Det Norske Samlaget, Oslo, Norway. 434 pp. Eithun, B., M. Rindal, and T. Ulset (Eds.). 1994. Den Eldre Gulatingslova. Norrøne tekster nr. 6. Riksarkivet, Oslo, Norway. 208 pp. Ersland, G.A. 2000. Kongshird og leidangsbonde. Pp. 13–154, In G.A. Ersland and T.H. Holm. 2000. Norsk Forsvarshistorie, bind 1. Krigsmakt og kongemakt 900–1814. Eide forlag, Bergen, Norway. 334 pp. Fett, P. 1954a. Førhistoriske minne i Hardanger 3. Ullensvang prestegjeld. Historisk Museum, University of Bergen, Bergen, Norway. 39 pp. Fett, P. 1954b. Førhistoriske minne i Hardanger 5. Kvam prestegjeld. Historisk Museum, University of Bergen, Bergen, Norway. 23 pp. Fett, P. 1954c. Skiparstod. Prøvegraving og restaurering 1954. Unpublished report, doc. no. 038371 in Topografisk Arkiv, Museum of Cultural History, University of Bergen, Norway. 2 pp. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 73 Pantos, A., and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Robberstad, K. (Ed.). 1969 (3rd Edition). Gulatingslovi. Det Norske Samlaget, Oslo, Norway. 433 pp. Rygh, O. 1910. Norske Gaardnavne. Vol. XI. Gaardnavne i Søndre Bergenhus amt. W.C. Fabritius og Sønner A/S, Kristiania, Norway. 643 pp. Sanmark, A. 2009. Administrative organization and state formation: A case study of assembly sites in Södermanland, Sweden. Medieval Archaelogy 53:205–241. Sanmark, A. 2010. The case of the Greenlandic assembly sites. Journal of the North Atlantic Special Volume 2:178–192. Sanmark, A. 2013. Patterns of assembly: Norse thing sites in Shetland. Journal of the North Atlantic Special Volume 5:96–110. Sanmark, A., and S. Semple 2008. Places of assembly: New discoveries in Sweden and England. Fornvännen 103:245–259. Semple, S. 2004. Locations of assembly in early Anglo- Saxon England. Pp. 135–154, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Sogner, S.B. 1961. Herred. Pp. 492–494, In F. Hødnebø (Ed.). Kulturhistorisk leksikon for nordisk middelalder fra vikingtid til reformasjonstid. Vol. VI. Gyldendal Norsk Forlag, Oslo, Norway. 710 pp. Steinnes, A. 1933. Gamal skatteskipnad i Noreg II. Re-published by Kildebokforlaget 1982, Oslo, Norway. 228 pp. Steinnes, A. 1974. Styrings- og rettsskipnad i Sørvest-Noreg i mellomalderen. Historisk-topografisk oversyn. Det Norske Samlaget, Oslo, Norway. 129 pp. Stylegar, F.-A., and O. Grimm 2003. Place-names as evidence for ancient maritime culture in Norway. Norsk Sjøfartsmuseum Årbok 2002:79–115. Taranger, A. (Ed.). 1979 (5th Edition). Magnus Lagabøters Landslov. Universitetsforlaget, Oslo, Norway. 206 pp. Tryti, A.E. 1987. Kirkeorganisasjonen i Bergen Bispedømme i første halvdel av 1300-tallet. Master’s Thesis. University of Bergen, Bergen, Norway. 557 pp. Unger, C.R., and H.J. Huitfeldt (Eds.). 1874. Diplomatarium Norvegicum. Vol. 8. P.T. Mallings Boghandel, Christiania, Norway. 924 pp. Unger, C.R., and H.J. Huitfeldt (Eds.). 1878. Diplomatarium Norvegicum, Vol. 9. P.T. Mallings Boghandel, Christiania, Norway. 899 pp. Unger, C.R., and H.J. Huitfeldt-Kaas (Eds.). 1888. Diplomatarium Norvegicum.Vol.12. P.T. Mallings Boghandels Forlag, Christiania, Norway. 918 pp. Vésteinsson, O., Á. Einarsson, and M.Á. Sigurgeirsson. 2004. A new assembly site in Skuldaþingsey, NE-Iceland. Pp. 171–179, In G. Guðmundsson (Ed.). Current Issues in Nordic Archaeology: Proceedings of the 21st Conference of Nordic Archaeologists 6–9 September 2001. Akureyri, Iceland. 214 pp. Kolltveit, O. 1977. Granvin, Ulvik og Eidfjord i gamal og ny tid. Bygdesoga I. Granvin, Ulvik og Eidfjord Bygdeboknemnd, Odda, Norway. 528 pp. Krag, C. 2008. The creation of Norway. Pp. 645–651, In S. Brink and N. Price (Eds.). 2008. The Viking World. Routledge, London, UK. 717 pp. Lange, C.C.A., and C.R. Unger (Eds.). 1851. Diplomatarium Norvegicum. Vol 2. P.T. Mallings Forlagshandel, Christiania, Norway. 948 pp. Lange, C.C.A., and C.R. Unger (Eds.). 1858. Diplomatarium Norvegicum. Vol. 4. P.T. Mallings Forlagshandel, Christiania, Norway. 908 pp. Liland, S. 1951. Skipreidene i Hordaland fylke. Frå Fjon til Fusa. Årbok for Nord- og Midhordland sogelag 1951:38–71. Magerøy, H. (Ed.) 1976. Diplomatarium Norvegicum. Vol. 21. Kommisjonen for Diplomatarium Norvegicum/ Norsk historisk kjeldeskrift-institutt, Bergen, Norway. 1025 pp. Munch, P.A. (Ed.). 1843. Registrum prædiorum et redituum ad ecclesias dioecesis Bergensis sæculo P.C. decimo quarto pertinentium, Vulgo Dictum “Bergens Kalvskind” (Björgynjar Kálfskinn). Guldberg and Dwzonkowksi, Christiania, Norway. 132 pp. Munch, P.A. 1849. Historisk-geografisk Beskrivelse over Kongeriget Norge (Noregsveldi) i Middelalderen. Wilhelm Grans Forlag, Christiania, Norway. 256 pp. Myhre, B. 1985. Boathouses as indicators of political organization. Norwegian Archaeological Review 18:36–58. Myhre, B. 1987. Naust, skip og leidang. Pp. 28–47, In I. Øye (Ed.). 1987. Kystliv. Onsdagskvelder i Bryggens Museum III. Bryggens Museum, Bergen, Norway. 118 pp. Myhre, B. 1997. Boathouses and naval organization. Pp. 169–183, In A.N. Jørgensen and B.L. Clausen (Eds.). 1997. Military Aspects of Scandinavian Society in a European Perspective, AD 1–1300: Papers from an International Research Seminar at the Danish National Museum, Copenhagen, 2–4 May 1996. Publications from The National Museum, Studies in Archaeology and History Vol. 2. The Danish National Museum, Copenhagen, Denmark. 264 pp. O’Grady, O.J.T. 2008. The setting and practice of open-air judicial assemblies in medieval Scotland: A multidisciplinary study. Vol. I–II. Ph.D. Thesis. University of Glasgow, Glasgow, UK. 649 pp. Available online at http://theses.gla.ac.uk/506/. Accessed 2 Februrary 2012 Olafsen, O. (Ed.). 1918. Jordebog for Hardanger Fogderi af 1646. Hardanger Historielag, Norheimsund, Norway. 109 pp Olafsen, O. 1920. Skibredeinddelingen i Hardanger. Pp. 126–129, In E. Bull (Ed.). Leding. Militær- og finansforfatning i Norge i ældre tid. Steenske Forlag, Kristiania, Norway. 175 pp. Olafsen, O. 1921. Kvam i fortid og nutid. Kvam kommune, Nordheimsund, Norway. 796 pp. Pantos, A. 2004. The location and form of Anglo-Saxon assembly places: Some “moot points”. Pp. 155–180, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 74 Warner, R. 2004. Notes on the inception and early development of the royal mound in Ireland. Pp. 27–43, In A. Pantos and S. Semple (Eds.). 2004. Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Ødegaard, M. 2013. State formation, administrative areas, and thing sites in the Borgarthing Law Area, Southeast Norway. Journal of the North Atlantic Special Volume 5:42–64. Øye, I. 2002. Landbruk under press 800–1350. Pp. 215– 414, In B. Myhre and I. Øye 2002. Norges landbrukshistorie I. 4000 f.Kr.-1350 e.Kr. Jorda blir levevei. Det Norske Samlaget, Oslo, Norway. 495 pp. Endnotes 1The Gulathing Law also used the term herað, a territorial unit that has been extensively discussed by Norwegian historians. In Eastern Norway, herað was a judicial unit corresponding to the Danish herred and Swedish härad. In western Norway, however, herað probably referred in a general way to a settled area (bygd), not units in an administrative structure. The Gulathing Law distinguished between heraðsréttr, i.e., rural law, and kaupangrsréttr, i.e., town law (G 120; Eithun et al. 1994:99). The few place-names containing herað in western Norway are also consistent with the meaning “settled area” (Helle 2001:78-79). For a short overview of the discussion of the herað, see Sogner (1961). 2From Voss, numerous medieval documents name eight áttungar in the area. However, these are subdivisions of the largest parish in Voss fjórðungr, and presumably too small to correspond to the “fylkis-áttungr” described in the Gulathing Law (Helle 2001:79–81). Romsdalen further north was also organized in áttungar in the middle Ages (Døssland 1990:34–36); however, this region did not belong to Gulathing, but instead to the neighboring Frostathing law province. 3Medieval Norwegian documents are compiled in Diplomatarium Norvegicum (normally referred to as DN, for volumes used in this article see Lange and Unger 1851,1858; Magerøy 1976; Unger and Huitfeldt 1874, 1878; Unger and Huitfeldt-Kaas 1888), and the corpus contains about 19,000 documents predating 1570. The 1520s tax rolls are published in Norske Regnskaber og Jordebøger fra det 16de Aarhundrede (normally referred to as NRJ; in this article, volume 3 published by H.J. Huitfeldt-Kaas in 1905 is used); and the 1560s tax and tithe rolls in Norske Lensrekneskapsbøker 1548–1567 (normally referred to as NLR; in this article, volumes III and IV published by J. Jansen in 1938 and 1941, respectively, are used). 4For examples of this rather arithmetical approach see Bull 1920:123–126, Munch 1849:117–118, and Robberstad 1969:319–320. In his comprehensive study of the Gulathing law province Knut Helle argued similarly, although more cautiously, and estimated five skipreiður in Hardanger and Voss respectively (Helle 2001:77). 5In addition to the already mentioned sources, a 1646 cadaster from Hardanger (Olafsen 1918) as well as two ecclesiastical cadasters from ca. 1360 (Munch 1843) and 1590 (Johannessen 2000) are utilized. The discussion in this article summarizes the findings of a more in-depth study being undertaken by the author (H. Hobæk, unpubl. data). 6References to Kinsarvik skipreiða—1472 in Magerøy 1976:436 (DN 21 no. 569) and 1566 in Unger and Huitfeldt 1878:821–822 (DN 9 no. 794), Ullensvang skipreiða—1558 in Magerøy 1976:125 (DN 21 no. 161) and 1627 in Unger and Huitfeldt-Kaas 1888:72–73 (DN 12 no. 94); and Øystese skipreiða—1564 in Unger and Huitfeldt 1878:819 (DN 9 no. 791). References to Jondal skipreiða—1563 in Jansen 1938:82 (NLR vol. III), 1567 in Jansen 1941:183 (NLR vol. IV). 7Kinsarvik, Ullensvang, Øystese, and Jondal skipreiður 1646: Olafsen 1918. Kinsarvik, Ullensvang, Øystese, Jondal, Granvin, and Ulvik skipreiður 1647: Fladby and Winge 1995:157–241. 8A skipreiða named Kinsarvik must be assumed to have encompassed the community bearing this name, and a document concerning the property rights to the farm Freim (gnr. 28; Rygh 1910:446) from 1472 states that the farm was located in “oddæ Sokn ok j Kinserwiks skipreede” (DN 21 nr. 569; Magerøy 1976:436). To constitute a coherent geographical unit, as skipreiður are understood to have been, a skipreiða encompassing Kinsarvik and Odda must also have comprised Ullensvang (Fig. 3). As Kinsarvik and Ullensvang are mentioned simultaneously as separate skipreiður from the mid-17th century and onwards, the skipreiða structure in this area appears to have changed. On Jondal and Øystese skipreiður, see note 13. 9The letter from 1501 relates testimonies on the course of events in a murder case in Røldal, which the ombudsman of Ullensvang parish in Hardanger investigated on behalf of the overlord of Bergenhus fief (DN 2 no. 1016; Lange and Unger 1851:747–747). This is the earliest source relating anything about Røldal’s affiliation within secular administration, and implies that the community from this time onwards was a judicial unit of its own and belonging to Hardanger. Røldal’s connection to Odda and Ullensvang in judicial-administrative matters is further evidenced in 17th and 18th century sources (Dalen and Dalen 1960:162). While Hardanger in general belonged to the diocese of Bergen, Røldal and Eidfjord parishes belonged to the diocese of Stavanger in the medieval period (Tryti 1987:114). Both parishes are in all extant sources said to have belonged to Horðafylki (Olafsen 1920:126, Rygh 1910:438) and thus to have diverging ecclesiastical and secular affiliation. While Eidfjord probably originally belonged to Hardanger also within the ecclesial administration (Kolltveit 1977:190), it is a possibility that Røldal’s ecclesial affiliation reflects an older situation with concurring affiliation to Stavanger diocese/Stavanger law province. With the scarce sources available for Røldal, no definite conclusions are possible at this point, but comparison with other districts located at borders of administrative units in Horðafylki may shed light on this question (H. Hobæk, unpubl. data). H. Hobæk 2013 Journal of the North Atlantic Special Volume 5 75 10King Sverre’s saga mentions that the king was attacked by an army of farmers, i.e., a leiðangr army, in 1180, where inhabitants of Hardanger and Sunnhordland fought under the command of Jon Kutissa, royal representative (lendrmaðr) of Sverre’s opponent King Magnus. Hardanger continued to defy Sverre and, according to the saga the region, was sacked in 1189 because the inhabitants had sent men and leiðangr to the king’s opponents (Gundersen 1996:66–67, 190). 11Bjørn Myhre explains the term “skipreide center” as an administrative center, mentioned in written sources, within the respective skipreiður. Often this center would give its name to the skipreiða (Myhre 1997:176–178, see also 1985:50 with references). 12The medieval period is AD 1030 to 1537 in a Norwegian context. A limited archaeological excavation was conducted in the Skiparsto boat house in Kinsarvik in 1954 (Fett 1954c), but no carbon datings have been done for the site, and the boat house is roughly dated to the Middle Ages (Grimm 2006:350, 414; Myhre 1997:172, 176). The same applies for the boat house in Skipadalen, Kvam—a limited excavation in 1918 yielded medieval pottery sherds, indicating a similar date for the building (Fett 1954b:12 with references). More precise datings are not available, and the boat house is roughly dated as medieval (cf. Grimm 2006:348, 414; Myhre 1997:172, 174). However, boat houses of this type are generally dated to the high/late Middle Ages (i.e., 1150–1537 in a Norwegian context; see Myhre 1997:173). Both of the names Skipadalen and Skiparsto contain the element “ship”, combined with elements meaning “the valley” and “landing place”, respectively. Names containing “ship” are known in connection with several large boat house remains, and a survey of maritime place-names mentions four additional skip-names from Hardanger (Stylegar and Grimm 2003:93–94, 102). However, little research has been done on maritime place-names in Norway, and while the skip-names confirm the maritime functions of the sites, questions regarding their dating and wider significance are at present unanswered (Stylegar and Grimm 2003:79–81). 1322 farms in Vikøy parish are mentioned in the 1521 tax roll, all of them listed under Øystese-tinget (NRJ vol. III; Huitfeldt-Kaas 1905:138–140). One of these (Ytre Birkeland gnr. 6; Rygh 1910:493) is not mentioned in 1563. The 1563 tax roll lists 22 farms from Vikøy parish under Jondal skipreiða, 21 of them identical to the farms mentioned in 1521 (NLR vol. III; Jansen 1938:82-83). The last farm listed here is not mentioned in the 1521 tax roll (Skåleim gnr. 30; Rygh 1910:496). In total, 23 farms in Vikøy parish are listed in the 1521 and 1563 tax rolls, all of them included in Øystese skipreiða in 1647 (Fladby and Winge 1995:224–230). One of these farms is also said to be located in Øystese skipreiða in 1564 (Sandven gnr. 11; Rygh 1910:494, mentioned in Unger and Huitfeldt 1878:819 [DN 9 no. 791]). The tax roll from 1563 also contains an entry of farmers who paid an additional tax as land owners, listed by parishes. However, of the 12 identifiable farms mentioned under Øystese parish (NLR vol. III; Jansen 1938:86), only one is actually located within this parish (Nedre Vik gnr. 43; cf. Olafsen 1921:808). Eight of the farms belonged to Vikøy parish and are also listed under Jondal skipreiða earlier in the same roll (NLR vol. III; Jansen 1938:82–83, 86), and three belonged to Jondal parish (gnr. 41 Augastad, 42 Drage, and 45 Tørvik; Rygh 1910:512–513). The Jondal farms do not occur elsewhere in the 1563 roll, but are listed under Jondal-tinget in 1521 (NRJ vol. III; Huitfeldt-Kaas 1905:134–135) and under Jondal skipreiða in 1647 (Fladby and Winge 1995:163). The 1563 tax roll thus illustrates the ambiguous use of administrative designations frequently encountered within Hardanger, but also clearly shows close connections between Øystese, Vikøy, and Jondal at this time. 14In the early part of the 16th century, there were nine judicial units, as Røldal seems to have constituted a judicial district of its own and at this time belonged to Hardanger. However, it is not clear whether this was also the case in the medieval period (see note 9).