52 Journal of the North Atlantic Special Volume 2 *Balliol College, Broad Street, Oxford OX1 3BJ, UK; lesley.abrams@balliol.ox.ac.uk. Introduction During the Viking Age, everyone in the Norse world had one experience in common. In Norway, Denmark, and Sweden, and in the settlements in Russia, eastern and northern Britain, Ireland, Normandy, and the North Atlantic, they all began to abandon their traditional religion for the Christianity which was practiced by the superpowers of their day. Christianity is of course a religion based on a more or less unchanging book, with well-articulated doctrines and strictly prescribed ritual practices. But this fi xed center has been accompanied by a remarkable fl exibility when it comes to applying religion to real life. In every conversion context, Christianity has had to convince people that their own religious practice was, fi rst, inadequate, and then, unacceptable. Yet in the process, the Church has had to face the primary fact of any missionfi eld: that what could change and what had to stay the same varied from place to place. Wherever it faced the challenge of conversion, Christianity worked its way into traditional habits of understanding life and the world, and as a result, Christian societies of very different character were created, bringing different measures of continuity and change depending on local circumstance. This process is obscure—only fi tfully illuminated across the Scandinavian world— and for Greenland even more than other settlements of the Viking Age the question of how the transition from one religion to another worked in practice can only have hypothetical answers. The exercise of speculation may, however, help us to think more critically about the character of Greenland’s society in its fi rst generations. Historiography of Conversion For much of the middle ages, the model of historical writing was Bede’s Historia ecclesiastica gentis anglorum, the Ecclesiastical History of the English People (Colgrave and Mynors 1969). Bede’s work was completed in northern England in AD 731, but it had tremendous staying-power. His history of the English conversion fundamentally conditioned the way that subsequent Christian writers chose to represent the process of religious change. It is easy to forget how infl uential these constructed Christian templates have been on our thinking. Although there is as yet no defi nite proof that Bede’s Historia was known to the earliest Icelandic historians, his chronological works were used and referred to, and Benedikt Benedikz (1976:340) supposed that the Historia “was absorbed into the bloodstream of the writing of the golden age of Icelandic historiography” (Phelpstead 2006:54–57, Turville-Petre 1972:104–107). In order to consider the religious condition of the early Greenlanders, we need to remember this and ask just how infl uential ecclesiastical convention might have been in shaping the settlement’s record of its past. Ari the Learned’s history of Iceland, Íslendingabók (Benediktsson 1968, Grønlie 2006), was probably inspired by Bede’s Historia, and if in their turn, educated Greenlanders had written the history of their settlement and its conversion to Christianity, it would in all likelihood have been in Bedan terms. They did not do so, as far as we know, and, while we can regret the lack of a surviving history of this sort, there may be an unexpected advantage as a result. In Bede, pagan people were converted Early Religious Practice in the Greenland Settlement Lesley Abrams* Abstract - While the beginnings of Christianity in Greenland are very poorly recorded, the settlement has played a prominent role in the discussion of paganism, the conversion, and early Christianity in the Viking world, thanks to the sagas in which Greenlanders feature. In particular, the range of religious practice that is refl ected in the literary representations of the past is very striking; the rituals of the seeress, Thorbjorg, the Christian practice of Eric’s wife, Thjodhild, and Gudrid’s pilgrimage to Rome and profession as a nun offer contrasting perceptions of lived religion in the late Viking Age. While the absence of other relevant sources relating to Greenland is clearly a disadvantage, it leaves us free to question entrenched assumptions about the early religious life of the community. While, as elsewhere, the conversion to Christianity in Greenland would have had a practical impact, ranging from the creation of political and economic alliances to changes in social custom (including burial and memorialization), I argue in this paper that Greenland might have been somewhat different from other Scandinavian communities overseas. Discussing and drawing on the written and material record, I propose that we might gain from resisting the narrative of Christian convention, which requires sudden, dramatic, and emphatic change, in favor of a different understanding of religious practice. If we entertain the possibility that some societies may have had more room for religious diversity than Christian sources would allow, it could be argued that the early community of Greenland, instead of conforming to Christian stereotype, experienced an extended period of diversity; a mixed society encompassing traditional religious practice and a largely domestic Christianity could have continued for some time until Christianity gained a suffi cient degree of institutionalization to impose a more conventional Christian way of life. 2009 Special Volume 2:52–65 Norse Greenland: Selected Papers from the Hvalsey Conference 2008 Journal of the North Atlantic 2009 L. Abrams 53 in a moment—by winning on the battlefi eld, for example, or being rounded up and taken down to the river for baptism on the order of a king. People “became” Christian because kings said they should and because missionaries were on hand to effect the ritual which would transfer them from one religion to another. Baptism equalled conversion. However, it is quite clear that in real medieval life, outside the conventions of narrative, conversion unwound on a much slower time scale. Surviving texts from betterdocumented regions, not to mention analogy from anthropological study of more modern populations, suggest that, no matter how the Church wanted it to be seen, conversion was not something effected in a moment, no matter how important some moments were. Furthermore, traditional, pre-conversion, religion and Christianity were not two different but equal entities which could simply be exchanged, replacing one with the other. Writing and the conventions of written recording arrived with Christianity, and all written accounts refl ect that connection, consciously or not. Their conventions require sudden dramatic change and the substitution of one religion with another, but this does not capture the complexity of the situation. Any story set in the ninth, tenth, or eleventh centuries written along those lines would presumably be inaccurate. If, on the other hand, Greenlanders had composed a saga account of the origins of Christianity, we might have had a Leifs saga or a Þjóðhildar saga, attributing the conversion to its central character. But this too would be problematic. Saga-writers and (re-writers) had various aims and concerns, family validation and lineage construction high amongst them. The genre could have been used retrospectively to attribute (or misattribute) to particular individuals the momentous change that had occurred in the settlement’s religious character when it became Christian. The concern to give religious weight to forbears of important contemporaries certainly had an impact on what literature survives. It has been suggested, for example, that Eiríks saga was rewritten to magnify the exploits of Thorfinn Karlsefni, the ancestor of Hauk Erlendsson, who produced a version of the saga some time before his death in AD 1334 (Wahlgren 1993:704; see also Vohra 2008). In Grænlendinga saga, Thorstein Eiriksson’s prophecy of the future of Thorfinn’s wife, Gudrid—she will make a pilgrimage to Rome, she will build a church, she will become a nun— establishes her as the overwhelmingly appropriate ancestor of churchmen and churchwomen, a holy forbear, as Ólafur Halldórsson (1978:392–394, 452; 2001:42) has noted. How accurately it reflects her real religious life is another matter. The role attributed to Eirik’s wife Thjodhild as the founder of Greenland’s first church (Fig. 1) similarly derives from a story in Eiríks saga, which throws the spot- Figure 1. The replica in Qassiarsuk of the small church excavated at the same site in the 1960s and commonly associated with the church of Thjodhild. See also Figure 4. Photograph © Jette Arneborg. 54 Journal of the North Atlantic Special Volume 2 light on Gudrid’s association with Eirik’s family. In addition to distortion motivated by family concerns, the sagas may misrepresent more generally by suggesting that the first people to adopt the new religion were those of the highest status. So this vernacular convention may be misleading as well. It is of course no surprise that literary conventions make for bad history, if history means an exact representation of past events. If we want to think about religious life in the early stages of the settlement, the absence of both ecclesiastical chronicles and sagas recounting the interaction between pagans and Christians and the attitudes of Greenlanders on both sides of the religious divide is a real drawback, whatever the fl aws of the genres. But, to look on the bright side, their absence gives us the opportunity to construct an independent picture of the early stages of Christianity in Greenland. A narrative of the progress of Christianity in the fi rst generations is out of the question. But another look at the small amount of written and archaeological evidence, seen in the light of comparative anthropological material, suggests that some entrenched assumptions about religious life in the early generations of the Greenland settlement can be questioned. Converting to Christianity Exactly what were people doing when they converted to Christianity? What happened to their lives? These questions are usually addressed by looking at sources steeped in ecclesiastical convention. In classic conversion narratives, conversion was an instrument of royal policy. If it was not forced upon people by oppressive secular power, it was usually a choice—conventionally, Christianity was the deserved winner of a contest over who would be in charge of managing society’s relations with the supernatural. The choice involved material considerations (would the new religion bring greater wealth?), political considerations (would it increase the king’s power?), social considerations (would it strengthen family ties?), or legal considerations (would it keep better order?). These are clearly matters of authority, not faith, matters of practical living, not believing. The touchpoint was the performance of ritual, which acted as a binding force, but also clarifi ed who did not belong. While faith was extremely important, it was faith in the new power that really mattered. Ruth Mazo Karras (1997:110) has noted that the sagas portray Christ as a more effective overlord and a better patron, and offer his superiority in this department as a good reason to convert. These are material, not philosophical or theological considerations. Nevertheless, I suspect that there was far more discussion, debate, and disagreement than the conventional conversion narratives allow for. I fi nd it hard to imagine that medieval people were simply passive consumers of the new religion. Reconstructions of the conversion period in sagas certainly imagine otherwise, as in Njáls saga, where the poetess Steinunn “lectured [the missionary Thangbrand] for a long time and tried to convert him to paganism. [He] listened to her in silence but when she had fi nished he spoke at length, turning all her own arguments against her” (ch. 102; Magnusson and Pálsson 1960:219–222; Sveinsson 1954:265–267). The saga author may have been speculating about the past as much as we are. But even if a society could be offi cially “converted” in one decisive moment, as in Iceland in AD 999, we must question how far this could have gone. New beliefs must be taught, new practices accepted. The process of convincing individuals and making the new religion a fundamental part of their lives must have taken time. The transformative message would fi nd different responses depending on whether the audience was forwardlooking and eager to take on the new, or reactionary and resistant to change. Greenland’s early settlers could plausibly have been of either sort—while pioneers are progressive, “ex-pat” societies are often more conservative than those they have left behind. Anthropology can offer better recorded contexts for comparison. Students of Africa, for example, have claimed to be able to chart the movement of Christianity into the continent with some precision; they can identify the names and biographies of the missionaries who were sent from the various colonial powers and can assemble details about the training of these men and women, their methods, where they were sent, and what they did—all this is available through surviving records, and it has been used to write the history of Africa’s conversion to Christianity. But, as Adrian Hastings (1994:437–438) has pointed out, the reality on the ground, “the black advance” of Christianity, was in fact “far more low-key, often entirely unplanned or haphazard.” Africans had far more agency in their conversion than the missionary-centered narrative reveals. Although the written sources accurately refl ect the missionary life, they avoid (or are ignorant of) other truths. African converts, in the words of David Maxwell (1999:3–4), “appropriated the symbols, rituals, and ideas of Christianity and made them their own,” creating an indigenous Christianity after an extended dialogue between missionary Christianity and African culture. Perhaps the model of an extended dialogue would suit the fi rst stages of Christianity in the early medieval world as well. Maybe conversions were more low-key, haphazard, and unplanned in the early middle ages than the sources suggest. This might explain a disturbing statement from Hamburg-Bremen, probably added 2009 L. Abrams 55 to the history of its Church around AD 1300, about Goutlande, Swetide, [and] Grenelande: The peoples claim that they are in part Christian, even though they are without faith and without confession and without baptism. In part, they even worship Jupiter and Mars, although they are likewise Christian (Schmeidler 1917:286, Tschan and Reuter 2002:228). Given that the text goes on to say that Icelanders’ noses freeze and come off if they wipe them, we are dealing here with something other than factual reporting. But the tradition represented in this statement—of a less than embedded Christianity and a society that did not conform to contemporary convention—may in its own suggestive way be as reliable as the other sources on which we rely. Christianity in Greenland We are very much in the dark when it comes to hard facts about the religious situation in Greenland in the early settlement. Several of the fi rst settlers were fi rmly associated with paganism in later vernacular tradition, especially in the two surviving sagas which focus on the voyages to Vinland, Eiríks saga rauða and Grænlendinga saga (Magnusson and Pálsson 1965, Sveinsson and Þórðarson 1935). Eirik is a committed pagan; Thorbjorg the seeress conducts rituals at a feast at Herjolfsness (Eiríks saga ch. 4); and one of the Vinland travellers, the hunter Thorhall, composer of a poem for “my patron, Thor,” seeks supernatural assistance when the food supply runs low (Eiríks saga ch. 8). The text on a rune-stick from Narsaq (Fig. 2) may suggest that Greenland’s residents were familiar with mythological stories—no more so, of course, than later Christian writers in Iceland, but the suggested dating of the rune-stick to the period AD 985x1025 may make it an early witness, and therefore potentially from a pagan context (Stoklund 1993: 47–50).1 Unfortunately, there is very little physical evidence to attest to active paganism in the settlement, beyond a small fragment of steatite found in 1932 in the barn and byre complex at Qassiarsuk (Ruin 19, Ø29) decorated with a Thor’s hammer (Fig. 3; Arneborg n.d.:31,36; Krogh 1967:23). If the settlers were not yet Christian, the lack of any discovered remains of so-called pagan burials is puzzling—almost all the excavated burials have been in churchyards (Lynnerup 1998:11–33, 51–52)—but there could be reasons other than the abandonment of paganism to account for the lack of burials with grave-goods. Furnished inhumations may have been associated with particular circumstances or social groups, as they were not the only form of burial in Iceland (or Scandinavia) in the late tenth century (Eldjárn and Friðriksson 2000:550–551). Pre-Christian graves could perhaps have been moved to churchyards at a later date. Furthermore, problems of preservation and discovery in the less-settled and more extreme Figure 2. The rune-stick from Narsaq. Photograph © National Museum of Denmark. 56 Journal of the North Atlantic Special Volume 2 landscape of Greenland may conspire against the identifi cation of such graves. If we turn to evidence of the conversion itself, it is in short supply. The kinds of external forces that usually brought missionary initiatives and political and social pressure for conversion can not be identifi ed for certain. Greenland, like Iceland, had no king in whose interests the new religion might have acted. When Adam of Bremen, in the history of his Church which he wrote in the 1070s, the Gesta Hammaburgensis ecclesiae pontifi cum, declared that Christianity had “recently” reached Greenland (Schmeidler 1917:274, Tschan and Reuter 2002:218), it is diffi cult to know what he meant. Adam also said that Greenlanders (along with Icelanders and legates from Orkney) had travelled to Bremen and begged Archbishop Adalbert (AD 1043–1072) to send preachers, and that the archbishop had obliged (Schmeidler 1917:167, Tschan and Reuter 2002:134). The see of Hamburg-Bremen had been founded in the ninth century to evangelise the pagan peoples of the North, and it cited papal authority for its claim to jurisdiction over the whole of the Scandinavian world (Abrams 1995a:esp. 229–238). It seems to have sent a mission to Iceland in the 980s, and Isleif Gizurarson, the fi rst Icelandic bishop, was consecrated in Hamburg-Bremen in AD 1056, returning to his country, according to Adam, with letters from the archbishop addressed to the people of both Iceland and Greenland (Schmeidler 1917:273, Tschan and Reuter 2002:218). Perhaps, as Ólafur Halldórsson (1981:204) has said, the assumption was that Isleif, in the absence of other offi cial bishops in the region, would take responsibility for Greenland as well., However, Adam’s statement is not equivalent to real information about Figure 3. Steatite loom-weight from Qassiarsuk decorated with a Thor’s hammer. Photograph © National Museum of Denmark. 2009 L. Abrams 57 the religious situation in Greenland at the time. Isleif was a missionary bishop (Vésteinsson 2000:19–24), and the authority of the German see over Iceland’s developing Church could have been quite tentative. The latter’s activity was likely to have extended to Greenland only if the two formed a single ecclesiastical unit at the time. The Historia Norvegiae, written in the second half of the twelfth century, possibly in eastern Norway, says that Greenland was “discovered, settled, and confi rmed in the Catholic faith by Icelanders” (a telensibus reperta et inhabita ac fi de catholica roborata) (Ekrem and Mortensen 2003:54–55, for the date and location:11–24), but this statement of a more general Icelandic connection is not necessarily evidence of a concerted program of religious mission. Hamburg-Bremen was itself clearly active in some parts of the North Atlantic in the second half of the eleventh century; Adam names three Hamburg-Bremen appointees to Orkney in Archbishop Adalbert’s time (Abrams 1995b:28–30). However, after Adalbert’s death, bishops for Orkney were appointed in York (Crawford 1983:105–108). There is plenty of evidence that Hamburg-Bremen experienced increasing diffi culty maintaining its northern monopoly, especially against the English. Michael Gelting (2004) has argued that Hamburg- Bremen had lost control over the Danish Church as early as AD 1059. Increasing stress is refl ected in Adam’s pages as the archbishops enlisted papal support against the competition. Adam quotes a letter from Pope Alexander II (AD 1061–1073) to the Danish bishops reminding them of their obligations to Hamburg-Bremen (Schmeidler 1917:221–222, Tschan and Reuter 2002:181–182). Alexander also wrote to the Norwegian king, Harald harðráði (AD 1046–1066), reprimanding him for preferring English and French bishops to those of the German see (Schmeidler 1917:155–157; Tschan and Reuter 2002:125; Abrams 1995a:229–236; Sawyer et al. 1987:81–83, 92–94). However, while the papacy had initially promoted Hamburg-Bremen’s rights, it changed its tune in the last quarter of the eleventh century, when it needed allies against the German emperor. Thereafter, it supported Scandinavian independence at the expense of the archdiocese’s claims (Abrams 1995a:237–238). Popes wrote to the kings of Denmark, Norway, and Sweden, suggesting ways in which direct links with Rome could be strengthened, bypassing Hamburg-Bremen. Ólafur Halldórsson (1978:xi, 443; 2001:44) has cited a letter from Pope Leo IX to the archbishop of Hamburg-Bremen dated AD 1053 as evidence that the people of Greenland fell under the jurisdiction of the German see by the mid-eleventh century. However, Leo’s is not the earliest Hamburg-Bremen text to mention Greenland. Several documents enumerating the northern places controlled by the diocese lay claim to ecclesiastical authority in Greenland from the early ninth century—some centuries before it was even colonized. Hamburg-Bremen’s notorious forgeries were apparently created to support the see against increasing competition (Curschmann 1909:122–129, Schmeidler 1918:195–203, Tschan and Reuter 2002:23). Documents in the names of popes Gregory IV (AD 827–844), Nicholas I (AD 858–867), Anastasius III (AD 911–913), John X (AD 914–928), Leo IX (AD 1048–1054), Victor II (AD 1054–1057), and Innocent II (AD 1130–1143) claim Greenland among Hamburg-Bremen’s fl ock. Papal recognition of Greenland is probably most likely to have occurred after it acquired its own bishop, but, as we shall see, this does not appear to have happened before the twelfth century; direct contact between Greenland and the pope through the payment of tithe is not attested before the thirteenth century (Arneborg 2000:315). Although the privilege of Leo IX may have been authentic (Curschmann 1909:3, Schmeidler 1918:251–252), it seems that Adam had no knowledge of the document in the 1070s; he discusses the protracted negotiations between Leo and Adalbert relating to the creation of a patriarchate (Schmeidler 1917:175, Tschan and Reuter 2002:140–141), but does not cite the pope’s privilege. Unfortunately the document was lost in 1943 when the Hamburg archives were bombed. Even if the reference to Greenland was indeed original and not a later interpolation, Leo IX’s statement of authority over the settlement may have been more aspirational and symbolic than real. The challenge offered by foreign missionaries seems to have begun to leave its mark on Hamburg- Bremen’s documents in the second half of the eleventh century (Curschmann 1909:123–129), while a visit to Rome by two of its archbishops in the 1120s to plead their cause has aroused particular suspicion (Christensen 1976:31–35). The politics that encouraged this manipulation of the archive may lie behind Greenland’s development of a connection with the Church in Scandinavia, bypassing Iceland. The early twelfth century was a time of significant change in episcopal organization, when Hamburg- Bremen’s authority was being further undermined by the creation of new sees. As Jette Arneborg (1991:145) has pointed out, the establishment of a diocese in Greenland belongs to the period of episcopal foundations emanating from the new archbishopric of Lund, itself created in AD 1103–1104. New sees were created for Oslo, Bergen, Nidaros, and, indeed, for Iceland; Jon Ogmundarson, first bishop of Hólar, was consecrated by Asser, archbishop of Lund, in AD 1106. Although it did not 58 Journal of the North Atlantic Special Volume 2 give in easily, by the mid-twelfth century Hamburg- Bremen had lost its status as the primary player in northern mission. Christianity in the Literary and Material Record Other external forces are identifi ed as instruments of Greenland’s conversion in saga tradition. While Eirik’s son Leif was singled out as the agent who brought Christianity to the settlements, fi rmly placing responsibility for this crucial development with the founding family, the missionary activities of the Norwegian king Olaf Tryggvason formed the background context. Olaf’s Christianity probably had its roots in England. Sverre Bagge (2006) has argued that eleventh-century traditions of Olaf’s Christianization of Norway were picked up and highlighted by later writers such as Oddr Snorrason (ca. AD 1190) and Snorri Sturluson in his Heimskringla (ca. AD 1230). Various twelfth- and thirteenth-century sources credit the king with conversions throughout the North Atlantic zone (of Iceland in Ari’s Íslendingabók, ch. 7, for example; of Shetland, Orkney, the Faroes, and Iceland in Historia Norvegiae; Ekrem and Mortensen:94–95). We are told, says Oddr in his Saga Olafs Tryggvasonar (ch. 52; Andersson 2003:101–102, Jónsson 1932:154–155), that Olaf converted fi ve countries, but the list that Oddr gives includes a sixth—Greenland. In Heimskringla, the king was made responsible for the conversion of Orkney and Iceland (Saga Olafs Tryggvasonar, chs. 47, 73, 84, and 95) before he baptized Leif and commissioned him to preach in Greenland (chs. 86 and 96; Halldórsson 1981:205–207; Hollander 1964:218, 228). Eiríks saga, which stresses Leif’s role, could have been particularly influenced by Gunnlaug Leifsson’s Latin saga about the Norwegian king, composed in Iceland in the early thirteenth century but now largely lost, where Olaf’s missionary activities seem to have had particular importance (Wahlgren 1993:704). Although there is no great narrative of conversion in Eiríks saga and Grænlendinga saga, they both show interest in the religious life of the early settlement, but with varying emphases and different detail. Both sagas specify that Greenland was a “heathen country,” settled before the official conversion of Iceland to Christianity (Eiríks saga ch. 5, Grænlendinga saga ch. 2). Everything revolves around individuals in Eirik’s family circle. In Grænlendinga saga, Thorvald Eiriksson’s death in Vinland provides an opportunity to stress his Christianity; back in Greenland, we are told, Christianity was in its infancy, and Eirik had died a pagan (chs. 5–6). The saga ends with the proclamation of the Christian credentials of Thorstein’s wife Gudrid. Eiríks saga in its current form opens with the story of Aud the Deep-Minded, whose entourage included Gudrid’s Christian grandfather. Gudrid’s own Christianity is highlighted by its opposition to the paganism of Thorbjorg the seeress. Thjodhild is converted by her son Leif. The few further incidents or comments included in Eiríks saga which give color to the conversion story are all associated with Eirik’s family. Thjodhild refuses to sleep with him after she becomes a Christian. She builds a church at some distance from the farm (Figs. 1 and 4). Her son Thorstein complains about Christians being improperly buried in unconsecrated ground rather than in proper churchyards (Eiríks saga chs. 5–6). These details portray an alternative process of conversion, through peer pressure and social levers, rather than by oppressive external intervention, and they may be true or they may be invention. Stories nonetheless refl ect a community’s conception of its past, and they also explain what they fi nd in the landscape. Carbon-14 dates for skeletons found around the turf church at Qassiarsuk (Fig. 4), well known for its attribution to Eirik’s wife, Thjodhild, place Christian burials there at an early period (Arneborg 2001:127, 130; Arneborg et al. 1999:161, 163). Arneborg (n.d.:30, 33) has pointed out that the 143 graves in the churchyard were not much disturbed, concluding that surface markers must have existed. In one mass grave, the remains of 13 disarticulated individuals were found, and a number of other graves contained disarticulated skeletons (Krogh 1967:31–37, Lynnerup 1998:53–54). Shipwrecks, other deaths far from home, or the reburial of pagan remains could account for this feature of the cemetery. Arneborg (2001:129) has argued that, because several of the 9 dated skeletons were so early, the church belonged to the landnám phase and catered to the fi rst settlers, some of whom were already Christian. She noted that a domestic building next to the turf church, excavated in 1974, was probably contemporary with it (Arneborg n.d.:23). Arneborg has also provided potentially eleventh-century dates for 2 samples of human bone from the churchyard at Kilaarsarfi k in the Western Settlement associated with the Viking-Age farm of Sandnes (AD 1021–1151 and 1030–1116, at one sigma; Arneborg 2001:127–128, 130).2 Although the cemeteries at Qassiarsuk and Kilaarsarfi k provide important physical evidence of early Christian communities, the exact application of C14 results remains problematic. It seems overly precise, for example, to describe an ox bone with a calibrated date-range of AD 960–1040 (accurate to only one sigma) which was found in a communal grave in the cemetery of Qassiarsuk’s turf church as representing settlers and their cattle at “precisely” the landnám date of AD 985 (Arneborg et al. 1999:163). Three samples from that cemetery with 2009 L. Abrams 59 more which resemble its later church (Guldager et al. 2002:45–47, 55–57, 66–67, 87–89, 116–118). Arneborg’s ongoing research will represent a major step forward in our understanding of the chronology and context of Greenland’s small church sites. They belong to a type which can be identifi ed elsewhere in the North Atlantic world, and there is some disagreement about whether they represent “early” use, being replaced by a later rectangular style of building (Fig. 6), or whether they continued as a type throughout the middle ages. There is also the issue of whether these types of church were necessarily proprietary, and therefore replaced by new churches when centralized ecclesiastical organization developed, or whether they too could have been incorporated in the diocesan structure (Stummann Hansen and Sheehan 2006, Vésteinsson 2005:75–79, Wood 2006:92–108). While small churches may have gone out of use for burial, they could have continued to be used for prayer.3 Ólafur Halldórsson (2001:44), because he dated the settlement of Greenland to ca. AD 1000 rather than the more traditional date of ca. AD 985, has suggested that Greenland was never pagan, assuming that all Icelanders after the “offi cial” date of AD calibrated ranges of AD 894–996, 909–1017, and 995–1043 (similarly accurate to only one sigma) nevertheless suggest Christian burial at an early date, though whether they allow us to distinguish with certainty between the late tenth century and the early or late eleventh —a crucial distinction—is unclear. Niels Lynnerup (1998:48) has noted that a peak in the calibration curve of these samples makes for “rather large uncertainties in [the] datings when … translated to years AD.” More churches have been identifi ed with the early period of settlement by other means. A number of structures with circular banks, such as a small building surrounded by burials inside an enclosure at Inoqquassat (Ø 64) (Vebæk 1991:9), very like the small church at Qassiarsuk, were recorded close to farms by C.L. Vebæk (1991:7–19) and Knud Krogh (1967:90–1, 1982:121–3). Although Vebæk conducted small exploratory investigations, none of these sites has been properly excavated. One recent survey of the core area of the Eastern Settlement, centered on modern-day Qassiarsuk, has identifi ed at least 2 potential parallels to its turf church (at Ruin groups 522 and 529 [Ø 33 and 35], Qorlortoq and Qorlortup Itinnera; Fig. 5), as well as several Figure 4. Plan of the small church and surrounding cemetery excavated in Ø29a, Qassiarsuk by Knud Krogh in 1962–65. N 60 Journal of the North Atlantic Special Volume 2 Figure 6. Plan of the church in Ø1, Nunataaq, inside a rectangular churchyard. From Guldager et al. (2001: 89), reproduced with kind permission from the authors. Figure 5. Plan of the small church in Ø33, Qorlortoq. From Guldager et al. (2001:47), reproduced with kind permission from the authors. 2009 L. Abrams 61 be seen as historical fact. Equally, the recognition of the force of paganism in Greenland in this tradition might derive from a sense of narrative balance, or from reality, preserving a memory of religious diversity among the early settlers. Women’s religious leadership and households that encompass different religious practices are both conventional tropes and credible circumstances. Laws of a seventh-century king of Kent, Wihtred, name the penalty for a man who sacrifi ces to devils without his wife’s knowledge (ch. 12; Attenborough 1963:26–27). While Thjodhild is a classic representative of independent-minded Viking-Age womanhood, she would nevertheless not be out of place in the most conventional of ecclesiastical sources. Women loom large in missionary narratives like Bede’s, where they bring conversion to men who are slower to appreciate its value. Bede included in his Historia a letter from Pope Boniface V to Æthelburh, the Christian queen of Northumbria (ii.11; Colgrave and Mynors:172–175). Æthelburh’s “illustrious husband,” King Edwin, was “still serving abominable idols and hesitat[ing] to hear and obey the words of the preachers.” But Æthelburh was urged to “soften his hard heart as soon as possible.” The pope makes much of the Scriptural text “the unbelieving husband shall be saved by the believing wife” (I Corinthians 7:14). The court of Edwin and Æthelburh encompassed two religious traditions at one time, as did that of Clovis and Chlothild in Merovingian Francia in Gregory of Tours’ History of the Franks (II. 28–31; Thorpe 1974:141–5), and many more. It is diffi cult to say whether there are any hard facts in this convention; nor is it clear whether it could have fi ltered down through historical writing to infl uence the composition of vernacular sagas in medieval Iceland. Religious Diversity If the sagas’ interest in the practical problems of religious diversity preserved a memory of Greenland’s past, Iceland’s example should, of course, cast doubt on any suggestion that such a situation could have been viable for long. Thorgeir the Lawspeaker’s eloquent set-piece in Íslendingabók is based on the logic that pagans and Christians could not continue to live together in one society (ch. 7; Grønlie 2006:9, Hermannsson 1930:53). But Ari’s goal was the writing of Iceland’s past as Christian history. Furthermore, there were reasons for Greenland to have been different from Iceland. As we have seen, the coercive external forces that were instrumental in conversions elsewhere may not have been present; and, internally, Greenland may have had a more horizontal society, lacking the concentration of prestige and status in a few families which could 999 were Christian. If this itself seems doubtful (see Vésteinsson 2000 for the complexity of the situation in Iceland in the eleventh century), it is nonetheless perfectly credible that some of the settlers who went to Greenland could have been Christian. Íslendingabók claims that a missionary presence contributed to the presence of committed Christians in Iceland before AD 999 (ch. 7). Thanks to prolonged contact with the Christian world, increasing Scandinavian settlement within it, and the rise of Christianity in the homelands, individual Icelanders could also have converted abroad—that Leif Eiriksson became a Christian at the court of Olaf Tryggvason is probably more plausible than his subsequent missionary career. Some early settlers may have come from a Hiberno-Scandinavian milieu. Christian Keller (1991:134) has raised the possibility that the characteristic round churchyard of the turf churches refl ects an infl uence from Celtic Britain. The later identifi cation of some of those who emigrated to or visited Greenland as Christian, such as the unnamed poet from the Hebrides whose verses are cited in Grænlendinga saga (ch. 2) or Thorbjorn Vifi lsson (one generation away from a freed slave from the British Isles; Eiríks saga chs. 1 and 4), could, whether based on real people or fi ctional stereotypes, realistically refl ect a combination of religious infl uences in the new community right from the start. It is diffi cult to know how to treat the literary traditions of conversion. The sagas show a particular interest in religious oppositions. Gudrid’s dramatic dilemma in Eiríks saga (ch. 4)—whether to help the community by performing the pagan rituals or remain true to her Christianity—is one; Thjodhild’s refusal of conjugal rights is another. Pre- and postconversion attitudes also confront one another in the complaints of Thorsteinn Eiriksson’s ghost: “it is a bad custom, as has been done in Greenland since Christianity came here, to bury people in unconsecrated ground with scarcely any funeral rites. I want to be taken to church, along with the other people who have died here” (Eiríks saga ch. 6). The statement in Grænlendinga saga (ch. 5) that Eirik died before Greenland was converted implies an event other than the baptism of his wife and children, which he so manifestly survived. While refl ecting the Christian insistence on the progression from one religion to another, it might simply be a literary conceit: Eirik stood for the old religion and had to die before the new one could be thought to fl ourish. As ever with sagas, what you take from the text depends on your views on the issue of historicity. The accepted pattern of saga-narrative, necessitated by Iceland’s own history, was that pagan origins must be shown to give way to a Christian society—a literary explanation of the course of events. Alternatively, Eirik’s paganism and his family’s conversion can 62 Journal of the North Atlantic Special Volume 2 for ensuring social uniformity, there may have been no impetus for change. We might wonder whether a domestic religion might have been more tolerant than any public one, to which everyone had to conform. Though its progress in Greenland is obscure, a primarily domestic religion may have allowed more fl exible religious conditions to exist before the development of a more conventional bishop-led culture. As ecclesiastical organization encroached on family custom, a more public Christianity— the religion of society, not the household—won out over the private context. This process occurred throughout Christendom, at a pace determined by the nature of local secular power. Perhaps archaeology can give us an idea of what percentage of Greenland’s early farms had these private churches, and what proportion of the population could have been baptized and buried there. Development of an Institutional Church Although there were churches in early Greenland, without a bishop there could be no Church. Conditions had to be right to welcome a bishop, and resources had to be found for a substantial farm dedicated to his and his community’s livelihood. Greenland acquired its fi rst bishop some time in the twelfth century, probably the 1120s (the date is problematic; see Arneborg 1991:144). Einars þáttr Sokkasonar, composed perhaps around AD 1200 and surviving in Flateyjarbók, a manuscript of the late fourteenth century, makes the initiative emanate from Brattahlid and from a family which “stood head and shoulders above other men” (Halldórsson 1981:103–116, Jones 1964:191–203). The story offers no explanation of why Greenlanders did not turn to the Icelandic Church for a bishop. Instead, Einar went to Norway and requested one from King Sigurd Jórsalafari (AD 1103–1130), after which Arnald was consecrated by Archbishop Asser in Lund. However, the Icelandic annals, mostly written in the late thirteenth and early fourteenth centuries, record as bishop before Arnald the Icelander Eirik upsi Gnupsson, who went out to look for Vinland in AD 1121 or 1123 (Arneborg 1991:143–144; Storm 1888:19, 59, 112, 252, 320, 473). This bishop remains a mystery. While he makes an appearance on the infamous Vinland map (Seaver 2004:4–8), he does not fi gure in any other context. If he existed, he may have had more of a missionary than an institutional brief. Bishop Arnald himself is a shadowy figure, and some doubt whether he ever left home to take up his post; although the Icelandic annals record a succession of appointments, Arneborg (1991:145) has suggested that Helgi, who arrived in Greenland in AD 1212, might have been the first resident bishop have swung the balance in the Christian direction in the way that the Haukdælir promoted and monopolized episcopal power in Iceland (Vésteinsson 2000:19–37, 144–161). Thomas McGovern (1992) has suggested that Greenland was unique in other ways. He proposed that its society was characterized more by interdependence than independence and could have successfully functioned only with an unusual degree of social and economic co-ordination. It is possible that, thanks to its distinctive circumstances, relations between social groups might not have been articulated in the same way in Greenland as elsewhere in Christendom, and this would have had consequences for religious life. Perhaps the small farm-churches hold the key to understanding early practice. Whatever rituals existed in pagan society seem to have included the dead in the world of the living, and some responsibility for their aftercare devolved on those who continued to live in the world. Conversion on its own seems not to have fundamentally altered this ownership of the dead. The archaeology of Greenland’s farm-churches confi rms the claim in Eiríks saga that household religion, and most specifi cally burial by family members, continued even after the introduction of Christianity. The appearance of concerns about burial in Eiriks saga might be an indication of thirteenth-century pressures as much as those of the initial settlement-period; in the absence of suffi cient excavated data, it is impossible to say when burial in the larger churches became expected practice. In western European society, burial was traditionally the responsibility of the family until the institutional Church succeeded in monopolizing funerary practice, breaking up old patterns of authority and allegiance and replacing them with new ones (Blair 2005:58–65, 228–245, 463–471). Greenland’s small churches may have been the cult sites of a domestic religion, private establishments for the farmer and his household, as opposed to the larger cult buildings later established by the institutional Church. We might envisage a privatized religion, serviced by priests who lived at the farm (or were shared by several communities, or travelled occasionally from Iceland). This kind of religious practice could have allowed for greater diversity in the community than was “normal” elsewhere. A small number of Christian farmers with private churches in an otherwise traditional and conservative society may have existed for some time without disturbing the social balance, just as a number of pagan farmers could have carried on their domestic religion without opposition. As long as Christianity belonged primarily to the society of slaves and women— the politically incompetent—or to a small number of dispersed farmers (as in Iceland before AD 999?), it might not have threatened others, and, without mechanisms 2009 L. Abrams 63 perhaps as a way of retaining an identity associated with where they came from; conservative societies resist change and cling to traditions that others have abandoned. McGovern (1992:223–224) has argued that throughout its existence, the Greenland settlement was “different, unusual, and extreme by the standards of contemporary Scandinavia.” It would be interesting to allow at least the possibility that until the arrival of a bishop, the religious life of the settlement was different as well, preserving the conditions that allowed an “extended dialogue” between two religious traditions to continue for some time, until Christianity fi nally carried the day. Once the bishop arrived, Greenland became more like everywhere else in the Christian world. This suggestion may be as inaccurate and tendentious as Bede’s picture of instantaneous transformation. However, without more evidence, we can only entertain possibilities about what the past might have been like. Literature Cited Abrams, L. 1995a. The Anglo-Saxons and the Christianization of Scandinavia. Anglo-Saxon England 24:213–49. Abrams, L. 1995b. Eleventh-century missions and the early stages of ecclesiastical organisation in Scandinavia. Anglo-Norman Studies 17:21–40. Andersson, T.M. 2003. The Saga of Olaf Tryggvason. Oddr Snorrason. Cornell University Press, Ithaca, NY, USA and London, UK. Arneborg, J. 1991. The Roman Church in Norse Greenland. Pp. 142–150, In F. Bigelow (Ed.). The Norse of the North Atlantic. Acta Archaeologica 61. Arneborg, J. 2000. Greenland and Europe. Pp. 304–317, In W.W. Fitzhugh and E.I. Ward (Eds). North Atlantic Saga. Smithsonian Institution Press, Washington, DC, USA and London, UK. Arneborg, J. 2001. The Norse settlement in Greenland: The initial period in written sources and archaeology. Pp. 12–33, in A. Wawn and Þ. Sigurðardóttir (Eds). Approaches to Vínland. Sigurður Nordal Institute, Reykjavík, Iceland. Arneborg, J. No date. Saga Trails. Brattahlið, Garðar, Hvalsey Fjord’s Church and Herjolfsnes: Four chieftain’s farmsteads in the Norse settlements of Greenland—A visitor’s guidebook. National Museum of Denmark, Copenhagen, Denmark. Arneborg, J., et al. 1999. Change of diet of the Greenland Vikings determined from stable carbon isotope analysis and C14 dating of their bones. Radiocarbon 41(2):157–168. Attenborough, F.L. 1963. The Laws of the Earliest English Kings. Russell and Russell, New York, NY, USA. Bagge, S. 2006. The making of a missionary king: The medieval accounts of Olaf Tryggvason and the Conversion of Norway. Journal of English and Germanic Philology 105:473–513. Benediktsson, J. 1968. Íslendingabók. Landnámabók. Íslensk fornrit 1, Reykjavík, Iceland. Benedikz, B. 1976. Bede in the Uttermost North. Pp. (Storm 1888). Absenteeism was not unprecedented: a number of Hamburg-Bremen’s tenth- and eleventh- century appointments to Denmark and Sweden seem never to have visited their sees (Sawyer 1987:78–79, 93). Scholars disagree about whether Greenlanders themselves gained in power or lost it once the institutional Church arrived in Greenland, that is, whether the Church was an exploitative colonial force or an opportunity for domestic success. Arneborg (1991) and Keller (1991) have argued (contra McGovern, i.e., 1992:220–4) that the Church never really dominated the chieftains, who maintained their independence in cultural, political, and economic terms. But it is interesting that Greenland never produced its own bishops. On the Icelandic model, Einar or a son of the family should have become the first bishop, harnessing familylands to institutional power. Instead, Greenland’s first official bishop seems to have been a Norwegian, appointed by the archbishopric in Lund, and subsequent bishops continued the association with Norway. A lack of homegrown enthusiasm for the Church, a lack of wealth, an inadequate educational infrastructure, or an insufficiently hierarchical society without the right kind of elite could explain this reliance on outsiders. Although, as Keller (1991:128–137) has pointed out, Norwegian kings could hardly have taken responsibility for the operation of the Church in Greenland before AD 1261, Norwegian royal power or trade relationships could in some way have played a role in putting Greenland’s bishops in post before the settlement officially came under Norwegian rule. Under the bishops’ influence, Greenland may have become more Norwegian. Once Christian scrutiny and pressure to conform were more immediate, Greenland would certainly have become more conventionally Christian, with Christian practice penetrating more fully into social norms. Any leeway in religious identity would have been unlikely to survive the direct surveillance and institutional pressure of a resident bishop. Conclusion We know very little about the earliest Christianity in Greenland. You could say, however, that if we had more abundant written sources, we might be no better informed. History and anthropology both show how the record of conversion can mask its reality. Archaeology’s information on early churches and burials is therefore crucial, and more data could radically alter our picture. Although Christian convention required conversion to be rapid and complete, we should consider whether some of the settlers in Greenland might have in fact stuck to their pagan ways for longer than has been assumed, 64 Journal of the North Atlantic Special Volume 2 Krogh, K.J. 1982. Erik den Rødes Grønland. Nationalsmuseets Forlag, Copenhagen, Denmark. Lynnerup, N. 1998. The Greenland Norse. A biologicalanthropological study. Meddelelser om Grønland, Man and Society 24. Magnusson, M., and H. Pálsson. 1960. Njal’s Saga. Penguin Books, London, UK. Magnusson, M., and H. Pálsson. 1965. The Vinland Sagas. The Norse Discovery of America. Penguin Books, London, UK. Maxwell, D. 1999. Christians and Chiefs in Zimbabwe. A Social History of the Hwesa People c. 1870s–1990s. Edinburgh University Press, Edinburgh, UK. McGovern, T.M. 1992. Bones, buildings, and boundaries: Palaeoeconomic approaches to Norse Greenland. Pp. 193–106, In C.D. Morris and D.J. Rackham (Eds). Norse and Later Settlement and Subsistence in the North Atlantic. University of Glasgow, Department of Archaeology, Glasglow, Scotland, UK. Phelpstead, C. 2006. Pilgrims, missionaries, and martyrs: The holy in Bede, Orkneyinga saga, and Knýtlinga saga. Pp. 53–81, In L.B. Mortensen (Ed.). The Making of Christian Myths in the Periphery of Latin Christendom (c. 1000–1300). Museum Tusculanum Press, Copenhagen, Denmark. Sawyer, B., et al. 1987. The Christianization of Scandinavia. Viktoria Bokförlag, Alingsås, Sweden. Schmeidler, B. 1917. Adam von Bremen, Hamburgische Kirchengeschichte. Hannsche Buchhandlung, Hannover and Leipzig, Germany. Schmeidler, B. 1918. Hamburg-Bremen und Nordost- Europa. Dieterich’sche Verlagsbuchhandlung m.b.H., Leipzig, Germany. Seaver, K.A. 2004. Maps, Myths, and Men. The Story of the Vinland Map. Stanford University Press, Stanford, USA. Stoklund, M. 1993. Objects with runic inscriptions from Ø 17a in Narsaq—A Norse Landnáma farm. Pp. 47–52, In C.L. Vebæk et al. (Eds.). Meddelelser om Grønland, Man and Society 18. Storm, G. 1888. Islandske Annaler. Grøndal and Søns Bogtrykkeri, Christiania, Norway. Stummann Hansen, S., and J. Sheehan. 2006. Bønhústoftin and the early Christianity of the Faroe Islands, and beyond. Archaeologica Islandica 5:27–54. Sveinsson, E.Ó. 1954. Brennu-Njáls saga. Íslensk fornrit 12, Reykjavík, Iceland. Sveinsson, E.Ó., and M. Þórðarson. 1935. Eyrbyggja saga. Íslensk fornrit 4, Reykjavík, Iceland. Thorpe, L. 1974. Gregory of Tours. History of the Franks. Penguin Books, London, UK. Tschan, F.J., and T. Reuter. 2002. History of the Archbishops of Hamburg-Bremen. Revised Edition. Columbia University Press, New York, NY, USA. Turville-Petre, G. 1972. Legends of England in Icelandic manuscripts. Pp. 104–21, In P. Clemoes (Ed.). The Anglo-Saxons. Studies in Some Aspects of their History Presented to Bruce Dickins. Bowes and Bowes, London, UK. Vebæk, C.L. 1991. The Church Topography of the Eastern Settlement and the Excavation of the Benedictine Convent in Uunartoq Fjord. Meddelelser om Grønland, Man and Society 14. 334–43, In G. Bonner (Ed.). Famulus Christi. Essays in Commemoration of the Thirteenth Centenary of the Birth of the Venerable Bede. SPCK, London, UK. Blair, J. 2005. The Church in Anglo-Saxon Society. Oxford University Press, Oxford, UK. Christensen, A.E. 1976. Archbishop Asser, the emperor and the pope. The fi rst archbishop of Lund and his struggle for the independence of the Nordic Church. Scandinavian Journal of History 1:25–42 Colgrave, B., and R.A.B. Mynors, 1969. Bede’s Ecclesiastical History of the English People. Clarendon Press, Oxford, UK. Crawford, B.E. 1983. Birsay and the early earls and bishops of Orkney. Pp. 97–118, In W.P.L. Thomson (Ed.). Birsay. A Cultural and Political and Ecclesiastical Power. Orkney Heritage 2. Kirkwall, Orkney, UK. Curschmann, F. 1909. Die Älteren Papsturkunden des Erzbistums Hamburg. Leopold Voss, Hamburg and Leipzig, Germany. Ekrem, I., and L.B. Mortensen. 2003. Historia Norwegie. Museum Tusculanum Press, Copenhagen, Denmark. Eldjárn, K., and A. Friðriksson. 2000. Kuml og Haugfé. Úr heiðnum sið á Íslandi. 2nd Edition. Mál og menning, Reykjavík, Iceland. Gelting, M.H. 2004. Elusive bishops: Remembering, forgetting, and remaking the history of the early Danish Church. Pp. 169–200, In S. Gilsdorf (Ed.). The Bishop: Power and Piety in the First Millennium. LIT, Münster, Germany. Grønlie, S. 2006. Íslendingabók. Kristni Saga. The Book of the Icelanders. The Story of the Conversion. Viking Society for Northern Research, London, UK. Guldager, O., S. Stummann Hansen, and S. Gleie. 2002. Medieval Farmsteads in Greenland. The Brattahlid Region 1999–2000. Danish Polar Centre, Copenhagen, Denmark. Halldórsson, Ó. 1978. Grænland í miðaldaritum. Sögufélag, Reykjavík, Iceland. Halldórsson, Ó. 1981. The conversion of Greenland in written sources. Pp. 203–216, In H. Bekker-Nielsen, et al. (Eds). Proceedings of the 8th Viking Congress. Odense University Press, Odense, Denmark. Halldórsson, Ó. 2001. The Vinland sagas. Pp. 39–51, In A. Wawn and Þ. Sigurðardóttir (Eds). Approaches to Vínland. Sigurður Nordal Institute, Reykjavík, Iceland. Hastings, A. 1994. The Church in Africa 1450–1950. Clarendon Press, Oxford, UK. Hollander, L.M. 1964. Heimskringla. History of the Kings of Norway. University of Texas Press, Austin, TX, USA. Jones, G. 1964. The Norse Atlantic Saga. Oxford University Press, London, UK. Jónsson, F. 1932. Saga Óláfs Tryggvasonar. G.E.C. Gads Forlag, Copenhagen, Denmark. Karras, R.M. 1997. God and man in medieval Scandinavia. Writing—and gendering—the conversion. Pp. 100–114, In J. Muldoon (Ed.). Varieties of Religious Conversion in the Middle Ages. University Press of Florida. Gainesville, FL, USA. Keller, C. 1991. Vikings in the West Atlantic: A model of Norse Greenlandic Medieval Society. Pp. 126–41, In F. Bigelow (Ed.). The Norse of the North Atlantic. Acta Archaeologica 61. Krogh, K.J. 1967. Viking Greenland. National Museum, Copenhagen, Denmark. 2009 L. Abrams 65 Vésteinsson, O. 2000. The Christianization of Iceland. Priests, Power, and Social Change 1000–1300. Oxford University Press, Oxford, UK. Vésteinsson, O. 2005. The formative stage of the Icelandic Church c. 990–1240 AD. Pp. 71–81, In H. Þorlaksson (Ed.). Church Centres: Church Centres in Iceland from the 11th to the 13th Century and their Parallels in Other Countries. Snorrastofa, Reykholt, Iceland. Vohra, P. 2008. The Eiríksynnir in Vínland: Family exploration or family myth? Viking and Medieval Scandinavia 4:249–267. Wahlgren, E. 1993. Vinland sagas. Pp. 704–5, In P. Pulsiano (Ed.). Medieval Scandinavia. Garland, NY, USA and London, UK. Wood, S. 2006. The Proprietary Church in the Medieval West. Oxford University Press, Oxford, UK. Endnotes 1I should like to thank Judith Jesch and Rie Oldenburg for information on the Narsaq rune-stick. 2I am grateful to Jette Arneborg for information on the dates from Qassiarsuk and Kilaarsarfi k, and to Marc Pollard of the Oxford Research Laboratory for Archaeology and the History of Art for discussion of C14 methodology. 3I am grateful to Steffen Stummann Hansen and Christian Keller for discussion of these sites; I owe this last suggestion to Christian Keller.