Eagle Hill Masthead

Journal of the North Altantic
    JONA Home
    Aim and Scope
    Board of Editors
    Editorial Workflow
    Publication Charges

Other Eagle Hill Journals
    Northeastern Naturalist
    Southeastern Naturalist
    Caribbean Naturalist
    Neotropical Naturalist
    Urban Naturalist
    Prairie Naturalist
    Eastern Paleontologist

Eagle Hill Institute Home

About Journal of the North Atlantic


Parishes and Communities in Norse Greenland
Orri Vésteinsson

Journal of the North Atlantic, Special Volume 2 (2009–10): 138–150

Full-text pdf (Accessible only to subscribers.To subscribe click here.)

JONA Vol. 1 abstract page


Site by Bennett Web & Design Co.
138 Journal of the North Atlantic Special Volume 2 Introduction A central preoccupation of archaeologists and historians is finding ways to assess the quality of life of people in the past. This task is never easy because it is an inherently subjective issue—one person’s good life is another person’s nightmare— but scholars nevertheless normally feel justified in assuming that certain basic aspects of life can be quantified to give at least a relative sense of differences in life quality. Most commonly, these aspects include the basic needs of humans for food, shelter, and security—the lower strata of Maslow’s (1943) hierarchy of needs—although esteem (status) is also commonly considered. It is, however, fair to say that the availability, quantity, and quality of food is the single-most studied factor. It is not only the most absolute necessity, but there is also relatively plentiful evidence about it which is amenable to study and reconstruction. The logic of studying food availability is simple: as life must be good for those who have plenty of food, so it must be bad for those who have little. This measure—the differential access to food within a society—forms the basis for theories about social structure as well as numerous explanations for historical change; if food availability increases or it becomes more scarce, this might be the root cause for, or the effect of, other changes observed. The problem is that it can be difficult to know how sensitive people, either individuals or society as a whole, were to food availability. If people were always on the brink of starvation, then it is obvious that any change in food supply will have made a big difference to them—a difference of life and death—but if they had a comfortable margin, then such changes take on a different meaning, suggestive more of issues like adaptation, cultural preference, or status differentiation. In Norse Greenland, the grand narrative has always been about failure: the question about why the Norse settlements came to an end in the late middle ages. The available evidence has tended to be interpreted in the light of the end result, and it is fair to say that the question of food procurement has been central in this debate in recent decades. It has been reasonably postulated that because of its ecological marginality, Norse Greenlandic society would have had difficulties in feeding itself and that any changes in conditions—deteriorating climatic conditions in particular—could have pushed it over the edge (e.g., Barlow et al. 1997). This line of reasoning has proven fruitful in the sense that intensive research into it has now largely disproven its basic premise; it now appears that obtaining enough calories was not a problem for the Norse Greenlanders (Dugmore et al. 2009). The emerging picture is of an economy based on intensive, and successful, agricultural practices (Adderley and Simpson 2006, Arneborg 2005, Buckland et al. 2009, Schweger 1998) and of climate change increasing as much as decreasing the availability of wild species, such as the harp seal, for hunting (Ogilvie et al. 2009). In short, very little now seems to suggest that the Norse Greenlanders left because their subsistence system failed them. The more sophisticated understanding we now have of Norse Greenlandic subsistence suggests that more complex reasons than simple want of food affected the development and eventual demise of the settlements. While the subsistence system is thereby not absolved from all responsibility, this newer perspective suggests that it might be fruitful to look also at other needs that might have been wanting in Norse Greenland. Parishes and Communities in Norse Greenland Orri Vésteinsson* Abstract - The isolation of the two Norse Greenlandic settlements, from each other as well as from the rest of the world, is a well rehearsed topic. Ideas about communications within the two settlements are, on the other hand, not much developed. One way of looking at intra-settlement communication is through the parish system. The parishes arguably reflect the community structure, but they also provided the framework for much of the social interaction in the everyday lives of ordinary Greenlanders. The parish system can be reconstructed by analysing the different types of churches and their spatial and chronological distribution in relation to the location of farm sites. Based on fieldwork in southern Greenland as well as comparisons with Icelandic data, a reconstruction of parishes in Eystribyggð is proposed. This analysis reveals significant differences between the structure of Greenlandic and Icelandic parishes, the former being more centralized but also much larger with correspondingly less pastoral care available to each household. These differences highlight the particular nature of Norse Greenlandic society and may help to explain why that society came to an end in the late middle ages. Special Volume 2:138–150 Norse Greenland: Selected Papers from the Hvalsey Conference 2008 Journal of the North Atlantic *Department of Archaeology, University of Iceland, Sæmundargötu 2, 101 Reykjavík, Iceland; orri@hi.is. 2010 2010 O. Vésteinsson 139 The Study of Parish Systems In this paper, I propose to use the parish system as a window on what sort of communities the Norse Greenlanders lived in. Community relates to slightly higher levels in the hierarchy of needs: those of security, belonging, and esteem—factors that are well known to influence people’s behavior and choices of abode. Communities of sedentary (if highly mobile) peoples like the Norse Greenlanders have structural characteristics which can be reconstructed through settlement patterns. Importantly, these elements also allow systematic comparison with other places to get a sense of how different the Norse Greenlandic communities were from, e.g., their neighbors’ in Iceland. Parishes—the basic units of church organization, consisting of a church, its priest, and the households to which he provided pastoral care and which paid him and/or the church owner for those services— are vitally important for this sort of reconstruction. Rural parishes began to take shape in Northern Europe in the 10th century, and as a rule, they seem to reflect pre-existing community structures. More importantly, however, they provided the framework for people’s everyday lives and for much of the social interaction that took place. Parishes reinforced community feeling not only by the coming together for mass, but by communal feasting associated with church festivals as well as possibly by communal projects in maintaining and furnishing the church. It is arguable that the communal aspect of parishes is even more important in regions of dispersed settlement, like Northern Norway, Iceland, and Greenland, where people lived in separate farmsteads and had few occasions to meet each other outside church gatherings. The shape of the parish, in terms of which farms belonged to it, decided which people met each other on a regular basis. In such communities, the role of the priest also became that much more important as he was the only individual who travelled regularly between the farms in the community. He must have been an important intermediary for all sorts of interactions and would have been well placed to influence matters within his parish. Given that evidence for the building blocks of parishes—the churches and the farms—does exist in the archaeological and historical records for Norse Greenland, an attempt to reconstruct the parishes can be made. This reconstruction is problematic in many ways, however, and will need to rely on some hypothesizing. Here, the Icelandic parish system will be used as a model, both to aid in the reconstruction of the Norse Greenlandic one—on the reasonable but unprovable assumption that they are likely to have been similar—as well as to bring out the possible differences between the two societies. In Iceland, there is by now good evidence that in the early 11th century, within decades or even years of the decision to convert to Christianity, small churches or chapels were built in great numbers all over the country (Vésteinsson 2000:45–57, 2005). Reasonable estimates suggest that as many as 1700 of these structures were built, one for every second or third farm in the country. These simple buildings were, as a rule, associated with cemeteries, but their small size (most are less than 20 m2) argues against them having had a principal function as a place of gathering for groups larger than a household or two. Churches big enough for larger gatherings had been built before the end of the 11th century, but it seems that it was not until the second half of the 12th century that enough priests had been ordained to serve all these larger churches, which can as a group be associated with the ca. 330 churches served by one or more resident priests, i.e., the parish churches. Before the end of the 12th century, a hierarchical system had emerged whereby churches can be grouped in three categories: 1. Churches or chapels without a resident priest and whose owners did not receive any tithes or dues and were not exempt from paying tithes and dues to other churches. There were ca. 1000 of these chapels. 2. Churches without a resident priest which received tithes and possibly other dues from at least the farm they stood on and sometimes a few others. There were ca. 350 of this intermediate type of church. 3. Churches with a resident priest which received tithes and dues from a number of farms, including farms with chapels. There were some 330 of these parish churches. In the early 13th century, the intermediate type of church sometimes received tithe and dues from a number of farms, no doubt on account of a limited supply of priests, but this had become rare by the 14th century, when such annexes as a rule had no other revenue than the tithe and dues payable by the household to which they belonged. The revenue of the parish churches will have increased correspondingly, and the later middle ages saw a very gradual but nevertheless inexorable attrition in the numbers of chapels and of the status of annexes, making the system more and more centralized. The system was nevertheless very resilient, and according to a survey of several northern parishes in 1486–87, some 62% or 78 out of 125 chapels and annexes in 29 parishes were still in function (Diplomatarium islandicum V:352–57). Burial ceased at some of these chapels as early as the 12th century, but at others it continued throughout the middle ages and even beyond. The area served by a priest, the ministry, could therefore consist of several tithe areas, most 140 Journal of the North Atlantic Special Volume 2 commonly a large one with 5+ farms in addition to one or two single-household ones. The number of annexes and chapels a priest might have to service varied considerably, but the larger ministries with many annexes tended to have more priests, and there are examples of parish churches with as many as three priests and three deacons, some of whom would have been kept busy the year round servicing the annexes and chapels. In Iceland, there were on average 1.3 priests for every parish church and some 25% of the churches had a permanent staff of two priests or more. The size of the ministries varied considerably, from a single farm (usually an estate with several households) to as many as 40 farms, but figures between 5 and 15 were typical. There was a relationship between settlement density and sizes of ministries in that the greater the density of settlement, the smaller the ministries tended to be, with a correspondingly larger number of priests per capita. In regions of more dispersed settlement, ministries tended to be large, with many annexes and chapels and fewer priests per inhabitant. This tendency was particularly pronounced in the Eastern and Western fjords, the landscapes most similar to that of Greenland. Based on this summary of the main contours of the Icelandic parish system (see further in Vésteinsson 1998; 2000:45–57, 94–112, 240–45; 2005), we can now turn to Norse Greenland to consider the system there. In order to reconstruct the Norse Greenlandic parish system, three types of evidence need to be considered: First, the evidence for the equivalents of the Icelandic parish churches; second, the evidence for possible annexes and chapels; and third, the evidence for the numbers of farms that could have made up the Norse Greenlandic parishes. The reconstruction presented below is for Eystribyggð (the Eastern settlement), but reference is also made to Vestribyggð (the Western settlement) as the evidence allows. The Parish Churches of Norse Greenland Identifying the parish churches in Norse Greenland is relatively straightforward, although by no means unproblematic (Vebæk 1966 gives a useful summary of the research history). A number of large church ruins have long been known in Norse Greenland, and most of those can with reasonable certainty be identified with churches named in medieval documents. There are two main documents: a list of Greenlandic churches in Flateyjarbók, a manuscript written in Iceland in 1387–94, and Ívar Bárðarson’s description of Greenland, presumably written sometime after his return from Greenland in 1368, but with a more problematic transmission and surviving only in 17th-century manuscripts. There are signifi- cant correspondences between these documents, but also differences in detail. The impression they give of the number of churches is, however, very similar. That impression is consistent with information collected in the 17th century in Iceland (Halldórsson 1978:37–39, 235–36). These sources (see Table 1) suggest that there were 10–14 parish churches in Eystribyggð (including the cathedral and two monastic churches; Fig. 1) and 1–4 in Vestribyggð. The main difference between Flateyjarbók and Ívar’s description is that Flateyjarbók does not mention the two monastic churches, whereas Ívar does not mention churches at Herjólfsnes, undir Höfða, Table 1. Medieval sources for Norse Greenlandic churches, printed in Halldórsson (1978:37–39, 79, 133–37, 234–36). Source Ruin group Flateyjarbók Ívar Bárðarson Gronlandia number Modern place-name Herjólfsnes Herjólfsfjörður Ø111 Ikigaat Vatnsdalur í Ketilsfirði Áróskirkja í Ketilsfirði 2 kirkjur í Ketilsfirði Tasiussaq, Tasermiut Vík í Ketilsfirði Pétursvíkurkirkja Taserssuaq, Tasermiut Stórt klaustur inn frá Vatnsdalsbyggð Ø105 Tasermiutsiaat Systraklaustur langt inni í Hrafnsfirði Ø149? Narsarsuaq, Unartoq Vogar í Siglufirði Vogakirkja Siglufjörður Ø162? Narsaq, Unartoq undir Höfða í Austfirði Austkarsfjörður Ø66 Igaliku Kujalleq Garðar í Einarsfirði Nikuláskirkja í Einarsfirði Ófundinnfjörður Ø47 Igaliku Einarsfjörður Harðsteinaberg Brattahlíð í Eiríksfirði Hlíðarkirkja } 3 kirkjur í Eiríksfirði { Ø29a Qassiarsuk undir Sólarfjöllum undir Sólarfjöllum Ø23 Sillisit Ísafjörður Dýrneskirkja Ø18 Narsap Ilua Hvalseyjarfjörður Hvalseyjarfjarðarkirkja Ø83 Qaqortukulooq Garðanes í Miðfjörðum Ø1 Nunataaq Sandnes í Lýsufirði Steinsneskirkja Lýsufjörður V51 Kilaarsarfik Andafjörður Hóp í Agnafirði Agnafjörður V23a? Ánavík í Ragnafirði Ragnafjörður V7 Ujarassuit 2010 O. Vésteinsson 141 Harðsteinaberg, or Garðanes. Ívar’s omission of Herjólfsnes may be due to the incomplete transmission of the text; he mentions the place several times, and there is good archaeological evidence that the church was in operation into the 15th century. It is also strange that he omits undir Höfða, which from its substantial ruin and its location must be regarded as one of the main churches of Eystribyggð, the parish center for the whole of Vatnahverfi. It is tempting, on the other hand, to see the omission of Harðsteinaberg and Garðanes as indications of change. The information available to the compiler of Flateyjarbók was probably not up to date, and quite likely many decades old, and it is therefore possible that these churches were no longer in use when Ívar was in Greenland. His comment that the church undir Sólarfjöllum owned all of Miðfirðir may suggest that the parish of Garðanes had been merged with that of undir Sólarfjöllum. Radiocarbon determinations on human bones from Ø1 (Garðanes) suggest that the cemetery there may have been used into the 15th century, but the continued use of the cemetery does not preclude the possibility that the church was demoted to an annex. According to a 17th-century copy of a medieval manuscript called Gripla, there were 12 churches in Eystribyggð and 4 in Vestribyggð (Halldórsson 1978:37). Another manuscript, referred to as Grönlandiae vetus chorographia, also surviving only in 17th-century copies (the fullest in Arngrímur Jónsson’s Gronlandia), lists 10 churches in Eystribyggð by location (Halldórsson 1978:38, 235– 36). The difference may be that the list does not refer explicitly to the churches in Hvalseyjarfjörður or Dýrnes, although both place-names are mentioned. The omission of the two monastic churches from the Flateyjarbók and Gronlandia lists may be a matter of definitions, i.e., that they were not included because they were monasteries. However, from Ívar’s account, it is clear that they also served as parish centers, and Icelandic monastic churches were also parish churches, so this seems an unlikely explanation. Another possibility might be that both foundations were late, and that they post-dated the Flateyjarbók list. In that case, the parishes between Ketilsfjörður (Tasermiut) and Vatnahverfiwill originally have been extremely large, and the Flateyjarbók list must then predate 1308, when the monasteries are first referred to (Diplomatarium norvegicum X:15). Figure 1. Eystribyggð in Greenland with sites and place names mentioned in the text. 142 Journal of the North Atlantic Special Volume 2 The majority of these churches have been associated with particular ruins (Fig. 2), and on the whole there is consensus about those identifications. Although the names of the two churches in Ketilsfjörður have got mixed up in either or both sources, Ívar’s description of their locations is convincing, and one must have been in Tasiussaq and the other on or near Lake Taserssuaq (both in Ketilsfjörður or Tasermiut fjord). The identification of a church ruin on the lake with Pétursvík at Ø140 (Vebæk 1966:206–208) does not seem to have met with widespread approval but there can be little doubt that there must have been a church in that general area. The identifications of Vogar with Ø162 (Narsaq) and the convent with Ø149 (Narsarsuaq) are generally agreed upon, but pose some problems (Arneborg 2004:255). For one thing, Ívar’s text seems to imply that their locations are the other way around, i.e., that the convent is east or south of Vogar, not northwest as Ø149 is. He also says that the convent is “langt inni í þeim firði” (far up that fjord), which is hardly an apt description of the location of Narssarsuaq Ø149. Furthermore, he locates the convent in Hrafnsfjörður, which he says is the next fjord west of Ketilsfjörður. That fjord is Sarqa and, further in, Søndre Sermilik, normally associated with Álftafjörður of the medieval fjordlists, while a few lines below he mentions Hrafnsfjörður in a context which makes it clear that it is Alluitsup Kangerlua, which fits other evidence better. There is clearly some confusion here, but while there can be no doubt that Siglufjörður is the same as Unartoq and that Vogar therefore can be either Ø149 or Ø162, the convent should be looked for in either Søndre Sermilik or Alluitsup Kangerlua. To this author, it seems to make most sense that Ø149 is Vogar; that the much smaller ruin at Ø162 is an annex or chapel, and that the convent remains unidentified somewhere in the inner parts of Alluitsup Kangerlua. In order not to court controversy, however, the traditional identifications will be used in the reconstruction below. The location of Harðsteinaberg is the only serious remaining problem in parish church identifications in Eystribyggð, and the one that has proven most intractable. It should be somewhere between Garðar (Igaliko Ø47) and Brattahlíð (Qassiarsuk Ø29) if the Flateyjarbók list is in true geographical order. In this area, a possible church has been tentatively identified at Qinngua Ø39 (Guldager et al. 2002:42–45), but an excavation in 2001 failed to support this interpretation (Jette Arneborg, Danish National Museum, Copenhagen, Denmark, pers. comm.). It is not inconceivable that either Ø33 or Ø35 are Harðsteinaberg, but wherever it was, it seems likely that, like Garðanes, it was demoted to annex church status sometime before the mid-14th century, subsumed for its part by Brattahlíð parish. Figure 2. Plans of Norse Greenlandic church ruins. From Keller 1989:193. 2010 O. Vésteinsson 143 Arneborg et al. 2002), and on the whole, radiocarbon datings of the human bones point towards the 11th and 12th centuries, whereas dating on human bones from the cemeteries of the parish churches concentrate on the 13 and 14th centuries (J. Arneborg, pers. comm.). Interestingly, the bones from Sandnes (V51) in Vestribyggð have dates stretching back into the 11th and 12th centuries, but no chapels have so far been identified in Vestribyggð. These findings do suggest that the chapels lost their burial rights by the 13th century, but more research is needed to ascertain whether they also ceased to function at such an early date. More research is also needed to determine if burial really only commenced at the parish churches in the 13th century or whether this result is due to sample bias, the later graves being in better condition and more likely to be excavated and dated than the earlier, generally more disturbed ones. So far, eight chapels have been securely identified (including Þjóðhildarkirkja and Ø162 associated with Vogar), and interestingly, all but Ø162 are found in the areas of densest settlement in Eystribyggð, in Igaliku Kangerlua (Einarsfjörður) and Tungdliarfik Tunulliarfik (Eiríksfjörður). A ninth site may be indicated by a fragment of a church bell found near Qorlortorsup on the eastern side of Vatnahverfi(Arneborg 2004:253), and the find of a gravestone in the so-called Middle settlement may relate to a chapel or church there (Arneborg 1994). It is reasonable to assume that more of these sites will come to light in the course of further research, but intensive surveys of the Brattahlíð region, Vatnahverfi, and Hvalseyjarfjörður in recent years (Algreen Møller and Koch Madsen 2006a, 2006b; Guldager et al. 2002; Vésteinsson 2008), have not added many more sites, so it seems that the current figures reflect the order of magnitude. If the chapels were contemporary with the parish churches, this distribution suggests that their purpose was not to provide religious services to the greatest number of people, in which case they would be expected to be found at greater distances from the parish churches, but rather that they represent the wealth and status of farmers on a slightly lower tier than those who controlled the parish churches. The proximity of some of the chapels to parish churches can then be seen in terms of assertion of non-dependence on a powerful neighbor. Such patterns are well known in Iceland (e.g., Friðriksson et al. 2004, Vésteinsson 2006:100). Finally, it is worth pointing out that while there is no reason to doubt that Þjóðhildarkirkja was the first church in Brattahlíð, it later seems to have been contemporary with the earlier of the two stone churches for several decades if not more. The stone churches are not accurately dated, but the later one is believed to have been built around In Vestribyggð, no church ruin has been found that can be identified with Hóp, but its location has plausibly been associated with V23a (Roussell 1941:98). Annexes and Chapels in Norse Greenland Since 1932, a number of small churches have come to light in Eystribyggð which clearly do not appear on any of the medieval church lists and are furthermore substantially different from the parish church ruins, both in being smaller and being mostly inside circular enclosures. Initially, it was felt that these small churches (which will hereafter be referred to as chapels) predated the parish churches known from the written sources, not least because some of them were located close to parish church sites, suggesting relocation (e.g., Ø48 neighboring to Garðar Ø47 and Ø64 neighboring to undir Höfða Ø66), and the case seemed to be clinched by the excavation of the so-called Þjóðhildarkirkja in Qassiarsuk (Ø29a - Brattahlíð), in the 1960s (Krogh 1982:33–52). This very small chapel is in the same home-field as two phases of a very substantial stone church, and there is consensus that it must have been their precursor (e.g., Arneborg 2004:251–52). There has, however, also been a growing realization that it is in no way unthinkable that the chapels were contemporary with the larger parish churches, and that they represent a lower tier in an hierarchical organization (Krogh 1976:308). That would certainly be the inference from the Icelandic evidence and is the interpretation favored here. It is supported by evidence from Grænlendinga þáttr (Halldórsson 1978:109–10), where it tells of a church at a farm called Langanes (which remains unidentified) at which no priest resided, but where the bishop nevertheless celebrated mass in front of a large crowd and where burial was allowed. The phrase used, “eigi er heimilisprestr” could be lifted directly from an Icelandic charter (e.g., Diplomatarium islandicum I:466), and it is clear that the Icelandic audience would have been familiar with the arrangement where a church with burial rights could be the scene of a major gathering without having its own priest. Of course, it may be that the presumably Icelandic author of this text simply assumed that the church organization of Greenland was the same as in Iceland. As there does not seem to be any strong narrative reason for including this rather technical information, it is, however, just as likely that the author, who clearly was familiar with Greenland, simply knew that Langanes church was an annex and used this in the reconstruction of events. Recent research into these chapels led by Jette Arneborg (2002) has shown that burial took place in their churchyards (Algreen Møller et al. 2007, 144 Journal of the North Atlantic Special Volume 2 minor farms is more problematic. It was, however, necessary, as a number of sites which have a dwelling- type ruin and a home-field are in other respects so insignificant that they can hardly represent more than temporary or intermittent habitation. Such sites can be regarded as short-term experiments or as cottages, dependent on a larger farm, which had either intermittent or simply small-scale occupation. In the Vatnahverfireports (Algreen Møller and Koch Madsen 2006a, 2006b), the authors consistently distinguish between “mindre/lille gårdsanlæg” and larger farm sites, and the same criteria (not described but easily deduced) were applied to the other reports. The results are presented in Table 2. Although these survey areas are more or less contiguous and all are in the heart of Eystribyggð, there are significant differences between them, especially in the lower proportion of permanent farms in Vatnahverfi. This difference becomes even more marked when farms with two farm mounds are considered. There are 13 such sites in the Brattahlíð region and 2 in Hvalseyjarfjörður—an addition of 25% in the number of potential farm units in both areas—while there are only 2 in Vatnahverfi—potentially upping the number by only 4.4%. It seems, therefore, that the areas north of Igalikup Kangerlua were not only more densely settled, but also that the settlement there was more stable. It is important to keep this in mind because it suggests that there might be even more variation in this regard in other parts of Eystribyggð. In order to use these proportions to estimate the number of permanent farms in other parts of Eystribyggð, it is necessary first to subtract the number of new sites (invariably in the “other sites” category) that were discovered in the recent surveys. Equally intensive surveys of other areas can be expected to produce a similar number of new discoveries, so any calculations based on the overall number of sites must reckon without them. With 9 new sites in the Brattahlíð region, 1 in Hvalseyjarfjörður, and 14 in Vatnahverfi, that gives an expected proportion of permanent farms as 53%, minor farms as 16%, and other sites as 31%. The number of known Norse sites in Eystribyggð is close to 500, which should, using these proportions, break down to 265 permanent farms, 80 minor farms, and 155 other sites. 1300, making a 12th-century date reasonable for the earlier one. Radiocarbon dates show that burial continued in the cemetery around Þjóðhildarkirkja down to the 13th century (Lynnerup 1998:148), suggesting that it continued to have a function long after the stone church was built. Considering their different locations in the Brattahlíð home-field, it may be that the two churches belonged to different households, but a more plausible scenario may be similar to the one reported for Eyvindarmúli in Southern Iceland in the 16th century, where a rich landowner maintained a separate chapel out in the home-field for her private devotion, while the parish-church next to the dwelling received the congregation (Vésteinsson 2000:50–51). Like at Þjóðhildarkirkja, burial was also practiced around the chapel in Eyvindarmúli, suggesting that such private chapels also could serve as exclusive cemeteries of high-status families or households. The Number of Farms in Norse Greenland Systematic survey of the Norse ruins in Greenland has been underway since the 1880s, and although much work remains to be done, the general distribution seems well established and distribution maps have been widely published (the most detailed, and still useful, is at the back of Krogh 1982). Recent work continues to add new sites, typically minor sites of one or two ruins rather than farms, but more significantly it is providing a much more detailed and clear picture of sites long ago identified (Algreen Møller and Koch Madsen 2006a, 2006b; Guldager et al. 2002; Vésteinsson 2008). Although it has been obvious since the 19th century that not all Norse sites represent farms—some are shielings or economic buildings in out-fields—and that some farms were more permanent, wealthier, and had more households than others (see e.g., Roussell’s 1941 classification system), the distribution maps do not reflect such differences, and no complete inventory of Norse sites in Greenland is available in print. This knowledge gap is a major obstacle to parish reconstruction, which depends on the numbers and locations of farms being at least roughly known. In order to get a handle on this overall inventory, the reports of the most recent field surveys were studied (Algreen Møller and Koch Madsen 2006a, 2006b; Guldager et al. 2002; Vésteinsson 2008), and the 176 sites described in them grouped into farms, minor farms and other sites. Distinguishing between farms and non-farms was generally unproblematic. The former have, as a rule, more buildings, at least one of which can be characterized as a dwelling, and a grassy area definable as a home-field. Distinguishing between farms and Table 2. Classification of sites in three recently surveyed regions of Eystribyggð. Region Brattahlíð Hvalseyjarfjörður VatnahverfiTotal # % # % # % # % Permanent farms 45 56 6 43 29 35 80 45 Minor farms 6 8 2 14 16 20 24 14 Other sites 29 36 6 43 37 45 72 41 Total 80 100 14 100 82 100 176 100 2010 O. Vésteinsson 145 Considering the bias of the presently available data just mentioned, the figure of 265 should be considered as a maximum. The medieval manuscript Grönlandiae vetus chorographia mentioned above records the information that there were 190 farms in Eystribyggð (and 90 in Vestribyggð) (Halldórsson 1978:39), a figure which is reassuringly close to the calculated one. This tally presumably reports the number of properties, equivalent to the Icelandic concept of lögbýli. These properties could have multiple households, and it is therefore to be expected that this figure is lower than that of archaeologically defined farm sites. Here, the figure of 190 will be treated as a minimum and 265 as a maximum for the number of farms in Eystribyggð. With 14 parish churches, that gives 13.5 and 19 farms per parish, respectively, and with 12 parish churches, it yields 16 and 22 farms per parish, respectively. In Iceland, the national average was 11 farms per parish (Vésteinsson 2000:241), and although there were equally large parishes in Iceland, the Greenlandic ones can only be described as very large. Although these figures give cause enough for reflection, the case can be pushed a bit harder to produce a sense of the parish system by trying to map the figures onto the landscape. Reconstructing Parishes in Eystribyggð One way of going about reconstructing parish boundaries in Norse Greenland would be to draw von Thiessen polygons onto the landscape. Given that this landscape is riven by fjords and high mountains, this would, however, produce some very strange results, and here a more contextual, and therefore inherently more debatable, method will be employed. This approach tries to estimate what was the shortest possible route (in most cases by sea) from any given farm to a church, placing each farm in the parish of the church it is closest to. This is by no means unproblematic because alternative routes (e.g., overland vs. by sea) often exist and it is in most cases difficult to assess without detailed knowledge of local conditions, which would be more realistic. Where a farm or farms could be placed in either one of two parishes on the basis of distance, it was as a rule attributed to the parish with fewer farms on the reasonable, but not necessarily correct, assumption that there was a concern to divide the farms equally between parishes. Finally, indications about parish boundaries given by Ívar Bárðarson were taken into consideration. In one case, he clearly states that eight farms in Langey paid tithe to the church in Hvalseyjarfjörður, but in all other cases, he simply describes the area “owned” by a particular church. In some cases, this clearly refers to landownership, but in others, especially in the southern part of Eystribyggð, it seems to indicate the extent of the parishes as well (see Keller 1991:135–36). The results of this exercise can be seen in Table 3 and Figure 3. The map shows Vogar parish based on Ø162, but if Vogar is Ø149 and the nunnery was further north, then that would make their parishes more evenly sized in terms of the number of farms. The two parishes in Ketilsfjörður, where the locations of the churches remain unknown, are shown as one parish, and following the hypothesis that Garðanes and Harðsteinaberg parishes were merged with undir Sólarfjöllum and Brattahlíð, respectively, an attempt to split these parishes is not made either. The number of farms is given as a range for each parish, with the lower number corresponding to a sum of 190 and the higher to a sum of 265. The degree of uncertainty varies with the confidence that can be placed in the identification of actual farm sites within each parish. The greatest uncertainty is about the parish of Dýrnes, which Ívar claims is the largest in Greenland. This assertion really only makes sense if the farms in the so-called Middle settlement (Albrethsen and Arneborg 2004, Vebæk 1956), where no church has been found (although a gravestone find is reported in Arneborg [1994]), belonged to Dýrnes. That is a very long way indeed to go to church (more than 160 km by sea) and if true, has profound implications about the nature of pastoral care in Norse Greenland. It hardly needs stressing that these numbers and boundaries can only be considered as very rough approximations. Accuracy cannot be aimed for, and other reasonable premises would produce different results. It is maintained, however, that this analysis gives a clear enough sense of the shape and contours of the parish system to support important observations about the nature of Norse Greenlandic society. Table 3. Reconstructed numbers of farms in 14 parishes in Eystribyggð. Suggested number Lower Higher of farms estimate estimate Herjólfsnes 17 15 20 Árós 10 8 12 Pétursvík 7 5 7 klaustur 12 10 15 Vogar 14 12 15 Nunnuklaustur 31 30 35 undir Höfða 30 25 35 Garðar 9 8 10 Harðsteinaberg 6 5 8 Brattahlíð 17 15 20 Garðanes 6 5 7 undir Sólarfjöllum 8 5 10 Hvalseyjarfjörður 30 25 32 Dýrnes 34 22 40 Total 231 190 266 146 Journal of the North Atlantic Special Volume 2 all the farms in Eystribyggð (and presumably all in Vestribyggð) belonged to very large parishes, while in Iceland less than a fifth did. Keeping these differences in mind, the implications of the second structural aspect—the apparently much smaller number of chapels and annexes in Norse Greenland than in Iceland—become even more profound. It is possible that chapels and annexes are drastically under-identified in Norse Greenland, but as we have seen, there is not much hope that they are. Instead, it seems that they are mainly conspicuous by their absence from the Norse Greenlandic landscape. Figure 5 compares Eystribyggð with Vestfirðir, the northwestern region of Iceland most similar to Greenland in terms of geography and settlement patterns. It is a landscape cut through with deep fjords, and settlement there is largely confined to narrow Discussion Two structural aspects of the Icelandic and Norse Greenlandic parish systems can be compared. One is parish size, and Figure 4 presents an attempt to draw out the difference. Given the uncertainty about the Norse Greenlandic parish sizes, the numbers for Greenland can be distributed differently across the size categories and so they cannot be taken literally, but Figure 4 allows three main conclusions: first, that the systems are similar in general outline; second, that very small parishes (1–5 farms) did not exist in Norse Greenland; and third, that very large parishes (25+ farms) were much more common in Greenland, making up more than a third of the whole, while in Iceland very large parishes are less than a tenth. To put it differently: more than half of Figure 3. Reconstructed parish boundaries in Eystribyggð. 2010 O. Vésteinsson 147 Figure 4. Parish sizes in Iceland and Eystribyggð compared. Icelandic figures from Vésteinsson 2000:241 (Table 9), based on actual figures for ca. 66% of Icelandic parishes. Figure 5. Comparison of the proportions of farms with parish churches, chapels, or annexes, or with neither, in Eystribyggð in Greenland and Vestfirðir in Iceland. in Eystribyggð, but this was offset by a much greater number of annexes and chapels. It is reasonable to equate the latter group of churches with a middle class—small landowners and well-off tenants—who maintained a degree of independence in relation to the parish centers, which were either ecclesiastical benefices or bases of secular power. That this class of people is so poorly attested in Norse Greenland suggests that Norse Greenlandic society was a more decidedly two-tier society, with a small upper class and a more homogenous lower class. The larger parishes also relate to this pattern, suggesting that the Norse Greenlandic communities were more centralized, with relatively fewer centers, but that each center not only had a larger resource base but also controlled a greater proportion of the available revenue than their counterparts in Iceland. In the North Atlantic, there seems to be a correlation between greater settlement dispersal and larger parish sizes and fewer churches and chapels. This relationship is certainly the case within Iceland, and the data from Norse Greenland suggest that the communities there exhibit the same trend in extreme. Vestribyggð was even more exaggerated in this respect than Eystribyggð, with ca. three 30- farm parishes and no chapels or annexes. That this correlation exists may suggest that parish structure coastal strips. In the Icelandic context, Vestfirðir is a region of large parishes and relatively few chapels and annexes, but the difference with Norse Greenland is nevertheless striking. In Vestfirðir, there were 35 parish churches, and 133 chapels or annexes have been identified (the actual total is likely to be higher), while farms without church or chapel were ca. 360 in the 14th and 15th centuries to which these figures apply. The average parish in Vestfirðir was, in fact, slightly larger than 148 Journal of the North Atlantic Special Volume 2 was to a large degree a function of the economic base, that the poorer the land and hence greater the dispersal of settlement, the larger the ecclesiastical units had to be. That might imply that the more dispersed settlement units were smaller in terms of population and/or less productive than units in more densely settled areas because otherwise they would have been able to support at least the same relative number of churches. This line of argument might hold in a purely agrarian economy, but in the North Atlantic, it is arguable that units in areas of extreme settlement dispersal were often highly productive; they could occupy isolated pockets of good land for animal husbandry and have access to extensive pastures, but more likely they had access to plentiful wildlife. The vulnerability of such units might not relate so much to the economic base, but rather to problems of manning the operations. It is possible that in some cases such units were only viable with a relatively large crew (e.g., enough to man a boat or to round up herds dispersed over large areas), and that they had to rely on recruiting workers (perhaps seasonally) from somewhere else, individuals who would have had limited incentive or opportunity to settle down in the region. Even if such units could be operated by single family households, they might not have been considered such desirable places to live in by family folk either. Eating seal until it comes out of the ears is only an exciting prospect for a hungry person. Places which have nothing to offer except calorific abundance are inherently vulnerable to competition from places that offer more excitement, especially in good times. This is the vulnerability of areas of dispersed settlement, and the risk of periodic abandonment of the units will have had to be factored in to decisions about building infrastructure like churches. Such churches had to spread the risk of abandonment by having larger parishes, and they could not tolerate any siphoning off of revenue to lesser churches. It follows that while the risks of building a church in areas of dispersed settlement could be high, the profits could also be exponentially greater than in areas of more predictable, and hence more safely dividable, revenues. This in turn suggests that church owners in such areas of dispersed settlement would be likely to become actively involved in maintaining the population level of their parishes. If farms became abandoned, the church owners would have been well placed, and it would have been in their interest, to buy them in order to encourage speedy reoccupation. In this light, it may not seem so farfetched to take Ívar’s description literally to mean that many of the parish churches of Eystribyggð actually owned all the land in their parishes. In this sort of system, not only the tithe and other church dues went to the church-lord but also rent, presumably both of land and livestock, and quite likely he or she might also be responsible for collecting taxes, tolls, and fines for the royal and ecclesiastical treasuries. The high degree of economic centralization implied by this is manifested in the building works at these centers, the stone churches and feasting halls. The emphasis on providing a setting for social interaction and religious services is apparent and it suggests that the church-lords tried to out-do each other in order to hold on to and attract enough people to man the farms that kept the system afloat. However, while they could put on magnificent feasts and religious celebrations for their parishioners, the church-lords could not really make up for the social drawbacks of the isolation of the farms, the root-cause for the shape of the system. To get a sense of how very dispersed settlement in Norse Greenland was, it can be noted that Eystribyggð was some 15,000 km2 and had 190–260 farms, while Vestfirðir in Iceland, a region of high settlement dispersal in the Icelandic context, was 10.000 km2 with some 530 farms. In Hvalseyjarfjörður, the mean distance to the nearest neighbor was 4.1 km (i.e., when all the farm sites were occupied), which is a significantly higher figure than the area of greatest settlement dispersal in Iceland (Hornstrandir, with 2.5 km mean distance to the nearest neighbor [Vésteinsson 2006:93 (Table 5.1)]). Settlement was denser in parts of the Brattahlíð region and Vatnahverfi, but there were also other areas where distances between farms were much greater, suggesting that this is a fairly representative figure. This sparse settlement pattern means, in effect, that social interaction between farms in Norse Greenland would have been rare and in most cases would have required special effort, such as going to a feast or a mass, or the gathering for joint expeditions. These occasions would have been separated by long periods of isolation, and for some members of the households—young children, the old and infirm, as well as women—such interactions would be even fewer than for others, especially young men. It is quite possible to imagine a society where this very low level of social interaction is the norm and where no one thinks it is anything to complain about. Once other standards intrude, however, things can change rapidly. In the earliest Icelandic church legislation, The Old Christian Law Section, originally composed in the 1120s, there is a palpable sense of worry about how to maintain sufficient levels of pastoral care in a country of dispersed settlement. This angst is most apparent in the regulations about baptism, requiring the priest always to be ready to baptize at short notice and grown-ups to be able to perform a short-order baptism in case a priest 2010 O. Vésteinsson 149 could not be reached (Grágás 1852:4–5), but it can also be inferred from the law’s mention of visitation of the sick, hearing last confession, and administering extreme unction (Grágás 1852:12, 21). Birth and death were the two most significant moments in a medieval Christian’s life, and the system of pastoral care developed from the 10th century onwards intended a priest to be at hand at these moments to make sure that the soul stayed on the path to salvation. This system, modelled on European village societies, also intended a priest to be a constant presence, guiding his flock through every step of their lives, achieved by regular church attendance. Applying these principles in the North Atlantic was clearly problematic. Short-cuts were developed, and the Icelandic system of itinerant priests who serviced the chapels and annexes away from the parish church was clearly, in part at least, a response intended to provide a more effective cure of souls. How the Norse Greenlandic church adapted to the more extreme conditions of that country is difficult to know. It is reasonable to assume that the parish churches as a rule had more than one priest, and it is conceivable that some of these would have been itinerant, if not so much to service chapels and annexes—because there were not many of those—but simply to visit the households of their far-flung ministries and provide pastoral care. That Greenlandic conditions were known to call for special solutions is evident by the account in Eiríks saga rauða of how the Greenlanders buried their dead at their farms in unconsecrated ground, placing a wooden stick vertically on the corpse’s breast which was then taken out when a priest came by and holy water poured into the hole (Íslenzk fornrit IV:217). Such stories do not have to be taken literally (although see Magnússon 1972), but they do attest to a perception of Greenland as a place of great distances and isolation, a perception the Greenlanders themselves cannot have been completely unaware of. The knowledge that one inhabits a frontier where all modern amenities are not available does not have to have negative effects on a society as long as there is a sense of progress: a realistic prospect that at least some of those amenities will be provided in the not-so-distant future. While the parish system was developing in Greenland in the course of the 12th and 13th centuries, with steady improvements in the cure of souls and build-up of the church centers, the Norse Greenlanders may well have been content with their lot. However, once the system became consolidated and ceased to expand or improve, the realization that this is as good as its gets will have started to have an impact on the Norse Greenlandic psyche. Once the limits of growth had been reached, the negative aspects of life in Greenland will have become all the more glaring. Conclusion The isolation of the Norse Greenlanders was twofold. They were not only isolated from the rest of the world, but they also lived in an isolation from each other which was of an order of magnitude greater than in any other whole society in medieval Christendom. The settlement patterns established in the late 10th and early 11th centuries shaped the parish system that took form in the 12th century and continued to develop throughout the 13th century as evidenced by major building-works at the parish centers, mostly completed or halted by the first half of the 14th century. Compared to Iceland, this parish system was highly centralized, suggesting that Norse Greenlandic society was more abruptly divided in two tiers, an upper class and a lower class, and that the Norse Greenlandic upper-class, the church-lords, were much more completely in control of the surplus production. As long as this system continued to grow, and competition between the church-centers is evident, isolation and long distances would have been perceived as obstacles to be overcome, but once growth slowed down and stability was reached, the quality of life in Norse Greenland would have started to be considered as poor, irrespective of how much food could be put on the tables. Had Jesus been a Greenlander, he might have said: “Man does not live by seal alone,” and he would have been right. Literature Cited Adderley, W.P., and I.A. Simpson. 2006. Soils and palaeoclimate based evidence for irrigation requirements in Norse Greenland. Journal of Archaeological Science 33:1666–79. Albrethsen, S.E., and J. Arneborg 2004. Norse ruins of the southern Paamiut and Ivittuut region. SILA – The Greenland Research Center at the Danish National Museum and Danish Polar Center, Copenhagen, Denmark. 95 pp. Algreen Møller, N., and C. Koch Madsen. 2006a. Nordboerne i Vatnahverfi. Rapport om recogniscering og opmåling af nordboruiner i Vatnahverfisommeren 2005, (SILA Feltrapport 24), København, Denmark. 36 pp. Algreen Møller, N., and C. Koch Madsen. 2006b. Gård og Sæter, Hus og Fold - Vatnahverfi2006. Rapport om besigtigelser og opmålinger i Vatnahverfi, sommeren 2006, (SILA Feltrapport 25), København, Denmark. 39 pp. Algreen Møller, N., L. Lund Johansen, C. Koch Madsen, L. Felding, P. Baltzer Heide, and K. Smiarowski. 2007. Vatnahverfi2007. Udgravninger i norrøne kirkegårde og møddinger, (SILA: Feltrapport 26), København, Denmark. Arneborg, J. 1994. Den norrøne bosætning i Mellembygden. Appendiks af Karsten Secher. Grønlandsk kulturog Samfundsforskning 94:7–21. 150 Journal of the North Atlantic Special Volume 2 Keller, C. 1989. The Eastern Settlement reconsidered. Some analyses of Norse medieval Greenland. Ph.D. Dissertation. Universitetet i Oslo, Oslo, Norway. Keller, C. 1991. Vikings in the West Atlantic: A Model of Norse Greenlandic Medieval Society. Acta Archaeologica 61:126–41. Krogh, K.J. 1976. Om Grønlands middelalderlige kirkebygninger. Pp. 294–310, In G. Kolbeinsson (Ed.). Minjar og menntir. Afmælisrit helgað Kristjáni Eldjárn 6. desember 1976. Menningarsjóður, Reykjavík, Iceland. 576 pp. Krogh, K.J. 1982. Qallunaatsiaaqarfik Grønland. Erik den Rødes Grønland, 2. udg, Nationalmuseet, København, Denmark. 132 pp. Lynnerup, N. 1998. The Greenland Norse. A biologicalanthropological study. Meddelelser om Grønland. Man and Society 24, Commission for Scientific Research in Greenland, Copenhagen, Denmark. 149 pp. Magnússon, Þ. 1972. Staursetning. Árbók hins íslenzka fornleifafélags 1971:108–12. Maslow, A.H. 1943. A theory of human motivation. Psychological Review 50(4):370–96. Ogilvie, A.E.J., J.M. Woollett, K. Smiarowski, J. Arneborg, S. Troelstra, A. Kuijpers, A. Pálsdóttir, and T.H. McGovern. 2009. Seals and sea ice in medieval Greenland. Journal of the North Atlantic 2:60–80. Roussell, A. 1941. Farms and Churches in the Mediaeval Norse Settlements of Greenland. Meddelelser om Grønland 89.1. Reitzel, København, Denmark. 354 pp. Schweger, C.E. 1998. Geoarchaeology of the GUS site: a preliminary framework. Pp. 8–14, In J. Arneborg, and H.C. Gulløv (Eds.). Man, Culture, and Environment in Ancient Greenland (Danish Polar Center Publication 4), Danish National Museum and Danish Polar Center, Copenhagen, Denmark. 212 pp. Vebæk, C.L. 1956. Mellembygden. Tidsskriftet Grønland 1956(3):92–98. Vebæk, C.L. 1966. Kirke-topografien i Nordboernes Østerbygd. Tidsskriftet Grønland 1966(6):198–212. Vésteinsson, O. 1998. Íslenska sóknaskipulagið og samband heimila á miðöldum. Pp. 146–66, In G.J. Guðmundsson and E. Björnsson (Eds.). Íslenska söguþingið 28.–31. maí 1997. Ráðstefnurit I. Sagnfræðistofnun Háskóla Íslands, Sagnfræðingafélag Íslands, Reykjavík, Iceland. 391 pp. Vésteinsson, O. 2000. The Christianization of Iceland. Priests, Power, and Social Change 1000–1300, Oxford University Press, Oxford, UK. 318 pp. Vésteinsson, O. 2005. The formative phase of the Icelandic Church c. 990–1240 AD. Pp. 71–81, In H. Þorláksson (Ed.). Church Centres. Church Centres in Iceland from the 11th to the 13th Century and their Parallels in Other Countries, Snorrastofa, Reykholt, Iceland. 218 pp. Vésteinsson, O. 2006. Communities of dispersed settlements. Social organization at the ground level in tenth- to thirteenth-century Iceland. Pp. 87–113, In W. Davies, G. Halsall, and A. Reynolds (Eds.). People and Space in the Middle Ages, 300–1300. Studies in the Early Middle Ages 15. Brepols, Turnhout, Belgium. 366 pp. Vésteinsson, O. 2008. Archaeological Investigations in Hvalseyjarfjörður, Eystribyggð 2005. Fornleifastofnun Íslands, Reykjavík, Iceland. 58 pp. Arneborg, J. 2002. Kirke, kristendom, and Storbønder i Grønland - et nyt Forskningsprojekt. Pp. 8–26, In L. Bisgaard and R.S. Christensen (Eds.). Kristningen af Norden – et 1000 års-jubilæum. (Mindre Skrifter udgivet af Center for Middelalderstudier Syddansk Universitet, Odense vol. 21), Odense, Denmark. Arneborg, J. 2004. Det europæiske landnam. Nordboerne i Grønland, 985–1450 e.v.t. Pp. 219–78, In H.C. Gulløv (Ed.). Grønlands Forhistorie, Gyldendal, København, Denmark. 434 pp. Arneborg, J. 2005. Greenland irrigation systems on a west Nordic background. An overview of the evidence of irrigation systems in Norse Greenland c. 980–1450 AD. Pp. 137–45, In J. Klápste (Ed.). Water Management in Medieval Rural Economy (Památky Archeologické - Supplementum 17. Ruralia V), Institute of Archaeology, Academy of Sciences of the Czech Republic, Prague, Czechoslovakia. Arneborg, J., J. Toubro Hansen, and C. Poulsen. 2002. Kirkearkæologiske undersøgelser, (SILA: Feltrapport 5), København, Denmark. 40 pp. Barlow L., T. Amorosi, P.C. Buckland, A.J. Dugmore, J.H. Ingimundarson, T.H. McGovern, A.E.J. Ogilvie, and P. Skidmore. 1997. Interdisciplinary investigations of the end of the Norse Western Settlement in Greenland. The Holocene 7(4):489–99. Buckland, P.C., K.J. Edwards, E. Panagiotakopulu, and J.E. Schofield. 2009. Palaeoecological and historical evidence for manuring and irrigation at Garðar (Igaliku), Norse Eastern Settlement, Greenland. The Holocene 19(1):105–16. Diplomatarium islandicum eða íslenzkt fornbréfasafn sem hefir inni að halda bréf og gjörninga, dóma og máldaga og aðrar skrár, er snerta Ísland eða íslenzka menn. 1853–1972. 16 vols. Hið íslenzka bókmenntafélag, Kaupmannahöfn/Reykjavík, Iceland. Diplomatarium norvegicum. Oldbreve til kundskab om Norges indre og ydre forhold, sprog, slägter, säder, lovgivning og rettergang i middelalderen. 1847–1976. 22 vols. Kommissionen for Diplomatarium Norvegicum, Oslo, Norway. Dugmore, A.J., C. Keller, T.H. McGovern, A.F. Casely, and K. Smiarowski. 2009. Norse Greenland settlement and limits to adaptation. Pp. 96–113, In W.N. Adger, I. Lorenzoni, and K. O’Brien (Eds.). Adapting to Climate Change: Thresholds, Values, Governance. Cambridge University Press, Cambridge, UK. Friðriksson, A., O. Vésteinsson, and T.H. McGovern 2004. Recent investigations at Hofstaðir, northern Iceland. Pp. 191–202, In R.A. Housley and G. Coles (Eds.). Atlantic Connections and Adaptations: Economies, Environments, and Subsistence in Lands Bordering the North Atlantic, Oxbow Books, Oxford, UK. 304 pp. Grágás. 1852. Islændernes Lovbog i Fristatens Tid, Udgivet efter det Kongelige Biblioteks Haandskrift og oversat av Vilhjálmur Finsen, Første del. Nordiske Litteratur-Samfund, København, Denmark . 250 pp. Guldager, O., S. Stumann Hansen, and S. Gleie. 2002. Medieval Farmsteads in Greenland. The Brattahlid region 1999–2000. Danish Polar Center, Copenhagen, Denmark. 142 pp. Halldórsson, Ó. 1978, Grænland í miðaldaritum, Sögufélag, Reykjavík, Iceland. 453 pp. Íslenzk fornrit IV. 1935. Eyrbyggjasaga. Grænlendinga sögur. E.Ó. Sveinsson, and M. Þórðarson (Eds.). Hið íslenzka fornritafélag, Reykjavík, Iceland. 326 pp.