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


Settlement Landscapes in the North Atlantic: The Northern Isles in Context, ca. 800–1150 AD
Jane Harrison

Journal of the North Atlantic, Special Volume 4 (2013): 129–147

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


Site by Bennett Web & Design Co.
129 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 Introduction and Themes This article investigates a group of Viking-Late Norse coastal settlements in the Orkney Islands and, to a lesser extent, Shetland Islands (Scotland, UK), and looks at how the character and landscape setting of these settlements might help us understand the social organization of the Viking diaspora community. By the 10th century, this community was well established under the rule of the Orkney Earldom, a semiindependent magnate territory with origins in the 9th century and ruled by one or more peripatetic Earls. The Earls owed allegiance—sometimes little more than nominal—to Norway and/or Scotland at different times; after 1194 AD and the Northern Isles’ involvement in a failed coup in Norway, the Earls’ independence was increasingly constrained. During the Viking-Late Norse period, Orkney seems to have maintained particularly strong trade and exchange links with the west coast of Norway, as demonstrated by grave-goods and artifacts, such as bone combs and steatite vessels (e.g., Graham-Campbell and Batey 1998:25–36), and supported by descriptions in the Orkneyinga Saga of the sometimes difficult personal links maintained by the Orkney Earls with Norway (Palsson and Edwards 1978). The Viking-Late Norse economy was a broadbased and successful adaptation to the resources and strategic location of the Orkney Islands. Crops were grown (in particular, barley and oats) targeting the light sandy soils found along the coasts, and cattle and sheep kept, grazing on higher ground or less-fertile coastal areas, with evidence for at least some being sustained through winter on seaweed, stripped turf, and wetland meadow hay. The incomers had introduced flax and a comprehensive exploitation of marine resources: shellfish were gathered from rocky foreshores, and seabirds and their eggs were harvested from cliffs. Fish were not only caught close to the shore, but also larger cod and saithe from deeper sea fishing grounds, which required experienced crews and larger boats. Seaweed was used for fodder, fertilizer, and fuel. This much has been demonstrated by evidence from sites on Mainland, Sanday, and Westray in Orkney and from sites in Caithness (e.g., Barrett 2012; Dockrill et al. 2010; Griffiths and Harrison 2005, 2009, 2010, 2012; Griffiths et al. 2006, 2007, 2008, 2009; Hunter et al. 2007). The range of artifacts found at sites like Skaill in Sandwick (Bay of Skaill); Skaill in Deerness, and on the Bay of Birsay—such as Irish ring-pins and Norwegian steatite—also highlights the importance of trade and exchange (Buteux 1997; Griffiths and Harrison 2011; Morris 1989, 1996), while written sources document the continued involvement of Viking war-bands in mercenary and raiding activities into the 11th century (Owen 2005, Palsson and Edwards 1978). Burial records in Orkney, including the Scar boat burial, and the discovery of the Skaill silver hoard (Ashmore 2003, Graham-Campbell 1995, Owen and Dalland 1999) also suggest a military “chiefly” social structure. A society bound by complex and shifting ties of duty, local loyalty, and reciprocity, requiring leaders constantly “to mark and legitimise power and status” (Jørgensen 2000:76; cf. Scandinavia: Poulsen and Sindbæk 2011a, Skre 2011). The Orkney economy fully exploited the resources of the sea and land to sustain a rural population, but it was also hierarchical, supporting the enterprises of Settlement Landscapes in the North Atlantic: The Northern Isles in Context, ca. 800–1150 AD Jane Harrison* Abstract - During the Viking-Late Norse period (ca. 800–1150 AD), a complex network of cultural connections were forged between Scandinavia and areas of Viking settlement in the North Atlantic. This article focuses on how diaspora communities settling in the Northern Isles of Scotland adapted familiar ways of constructing their settlement landscape to new environments. Viking people in this period lived mostly on the coast and islands. Their dispersed settlements were often developed on natural mounds or mounds created by earlier clusters of buildings. Throughout the Viking-Late Norse age, many such mounds were built up in a tell-like layering of buildings, yards, and middens, visually dominating the surrounding landscape and coastal waters. This article argues that the people building settlement mounds were thus monumentalizing claims to local power. Mounds also physically represented social networks and adapted symbolism already associated with mounds in Scandinavia. Across the Sólundarhaf: Connections between Scotland and the Nordic World Selected Papers from the Inaugural St. Magnus Conference 2011 Journal of the North Atlantic *Department for Continuing Education - Archaeology, University of Oxford, 1 Wellington Square, Oxford OX1 2JA, UK; jane.harrison@conted.ox.ac.uk. 2013 Special Volume 4:129–147 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 130 minor magnates and Earls involved with networks of trade, exchange, and aggression. The known Viking-Late Norse settlement that went with this economy in the Northern and Western Isles and northeastern Scotland was located on the coasts, and comprised dispersed clusters of buildings and outhouses (Harrison 2007, Lamb 1980, OSMR 2013, RCAHMS 2013). The principal buildings were stone-built rectangular longhouses, a form familiar in Scandinavia and across the Viking North Atlantic (e.g., Hamerow 2002, Stummann Hansen 2003). Longhouse buildings tended to expand organically with the addition of annexes and smaller linked buildings. As will be discussed below, in the Northern and Western Isles, these longhouse settlements were built on either natural or archaeological mounds, sited to visually dominate bays and sea approaches. This attraction to mound locations has been previously documented in the Western Isles (Sharples 2005, 2012) and in Iceland (Vésteinsson 2010). In the Northern Isles, coastal settlements were also surrounded by a range of archaeological mounds from earlier periods. Mounds made by people were unavoidable in the landscape, and the Viking-Late Norse diaspora community apparently chose to create their own variation on the mound theme. But what was the significance of mounds? In this article, I will consider the social and economic implications of living on settlement mounds: how local power relationships embedded in socioeconomic networks might have been affirmed by the nature of the settlement landscape. In this examination, mounds are central. The remainder of the article will look more closely at the nature of settlement in Orkney, briefly at the Scandinavian background and at the detail of settlement mounds. It will be argued that considerable effort was invested in building such settlements because they translated an effective vocabulary of landscape symbolism from Scandinavia and adapted it to the social imperatives of a diaspora community. This symbolism was linked to the assertion of local power and the mound-located clusters of buildings played a key role in the social, political, and economic organization of the most valuable coastal areas. Settlement in Orkney As outlined above, Viking-Late Norse settlement in Orkney is found almost exclusively on the coastal margins of the islands, and all the excavated, dated sites of size and significance are located on either the windblown-sand bays of West and East Mainland (Mainland is the largest island in the archipelago), or of the islands of Sanday and Westray (including Papa Westray), or on the fertile coastal strip of the island of Rousay (Fig. 2; Graham-Campbell and Batey 1998:155–172, 180–200; Harrison 2007; OSMR 2013; RCAHMS 2013; Ritchie 1993). The same pattern is found in the Western Isles where evidence of Viking-Late Norse settlement is found exclusively on the sandy coastal machair (Parker Pearson 2012). Although the majority of Viking sites have been found on the coast of the Shetland Isles, settlement off the coastal margin has been discovered in excavations at Unst (e.g., Bond 2007). These coastal areas, such as along West mainland in Orkney, were extremely attractive to Viking settlers, providing sheltered locations well-situated to following a broad-based economy exploiting land and sea (Fig. 1). The sandy soils were easy to work, and the combination of shore, lowland, and higher ground provided an accessible range of environments. Apart from the relatively few excavated sites, settlement distribution has to an extent been inferred from relevant Norse place-names (e.g., Thomson 2008b:57–74), but marine erosion of coastal mounds in soft sand landscapes has also exposed long stretches of layered settlement (Dawson 2003, Lamb 1980, Steedman 1980). Some of these erosion exposures have produced Viking-Late Norse finds, for example around Ladykirk in Sanday, the Bay of Skaill in Westray, and Newark in Deerness (OSMR records 00144, 680, 682, 683, 02883). Viking-Late Norse settlement sites have been excavated at Pool on Sanday (Hunter et al. 2007); Tuquoy, Langskaill, and Quoygrew on Westray (Barrett 2012, Moore and Wilson 2003, Owen 1993); Westness and Swandro on Rousay (Kaland 1995, NABO 2012); Marwick Bay (Griffiths and Harrison 2009); Birsay Bay (Morris 1989, 1996; Ritchie 1977) including the mound known as Saevar Howe (Hedges 1983); the Bay of Skaill on West Mainland (Griffiths and Harrison 2005, 2009, 2010, 2012; Griffiths et al. 2006, 2007, 2008, 2009); and Skaill in Deerness, East Mainland (Buteux 1997, Gelling 1984). All of these sites are on the coast-edge, some have been partially eroded by the sea, and several are known to be multi-period, including Marwick Bay, Buckquoy in Birsay Bay, Pool on Sanday, Swandro on Rousay, and Skaill in Deerness. And all—except perhaps Westness, which is not fully published—are sites built on natural eminences and/or over earlier settlement, which were subsequently built up further during the Viking-Late Norse period by the superimposition of buildings, yards, activity areas, and middens. (Some of this build-up was achieved rapidly by infilling earlier buildings, sometimes over 0.5 m at once.) These settlement sites are “tell-like” (Lamb 1980), with deep stratigraphy and evidence for the incorporation of earlier structures and deposits in new building. In some cases, sequences begin with 131 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 pre-Viking elements, for example at Pool on Sanday and at Buckquoy in Birsay (Ritchie 1977). But the re-working also encompasses sequences of buildings and yards within the Viking-Late Norse period, as at the Bay of Skaill; Tuquoy and Quoygrew on Westray; Marwick Bay, West Mainland, and probably Saevar Howe on Birsay Bay. Building sequences beginning in the Viking period were also initiated in prominent locations or on existing natural mounds, for example on the Bay of Skaill, West Mainland. This pattern may also be true of some of the Sanday settlement mounds such as Beafield (OSMR 003290). The resultant mounds varied in size, but are usually over 20 m and up to 80 m in diameter and up to 4 m high. Similar economically well-placed coastal settlements on mounds or prominent locations—many also on sand—are found elsewhere: in the Western Isles (Crawford 1978; Sharples 2005, 2012); in NE Scotland, for example at Freswick in Caithness (Batey 1987); and in Shetland, for example at Jarlshof, Sandwick, Unst, and Old Scatness (Bigelow 1985; Bond 1998, 2007; Dockrill et al. 2010; Hamilton 1956). Interestingly, Northern Norway is notable for its so-called “farm mounds”, which have a similar coastal distribution and multi-period character, although so few have been extensively excavated that drawing conclusions about their origins and dating is problematic (Urbánczyk 1992:105–121). However, several have a Viking-Late Norse sequence and, unlike the excavated Orkney coast-side mounds, many continue into the post-medieval period (Bertelsen1979, 1984, 2011). Mounds on Sanday, in particular, are a slightly different case. While many—Pool and probably Crosskirk and Ladykirk (OSMR records 00144, 00710)—end with a Late Norse sequence, other mounds—like Beafield—are still occupied by modern farms. This difference may be partly because Sanday’s overwhelmingly sandy, flat, and narrow arms of land offered even less potential for settlement movement. It is interesting that mound locations were also targeted by Viking settlers in Iceland, although the imperatives shaping Figure 1. location of the Orkney Islands and some sites mentioned in the text. J. Harrison 2013 Journal of the North Atlantic Special Volume 4 132 settlement in not-yet-settled landscapes must have been different (Vésteinsson 2010). Despite this pattern, settlement mounds have not been considered as a related group across the North Atlantic and, while there is only space in this article to focus on Orkney, the author’s Ph.D. research ranges more widely. However, the settlement mounds were not the only mounds in the coastal landscape. The sandy bays and prime coastal strips in Orkney are also rich in other, often uncharacterized archaeological mounds. 946 mounds of any period or sort are recorded for the entire archipelago; around 75% of them are on Mainland, Westray/Papa Westray, Sanday, and Rousay (OSMR 2013, RCAHMS 2013). Unexcavated archaeological mounds are obviously difficult to classify with confidence, especially the larger ones, which may be multi-period or comprise a structural sequence ending with re-use for burial. But the vast majority of the smaller mounds (up to about 8 m diameter) are interpreted as burial mounds or enigmatic “burnt” mounds (633 OSMR records—114 are larger, chambered tomb mounds). Around a third of the total remains unclassified. However, the smaller burial/burnt mound-types are usually to be found on low, wet meadow lands in the center of bays, on higher ground encircling the bays, and even in more inland areas—a wider and slightly different distribution than the larger settlement mounds (Harrison 2007, Steedman 1980). Fifty-two of the mounds recorded in the OSMR are characterized as “settlement mounds”; the overwhelming majority of those remain unexcavated and any dating assayed on the basis of finds only, with four having been interpreted as Viking-Late Norse. However, these mounds are, like the excavated examples above, larger and coastal, often located on the arms of bays, and sited to dominate either the surrounding landscape or the sea approach. (One of the excavated sites is included in this group, but many were assigned other categories; one of the Bay of Skaill mounds is classified as a possible broch.) Thus, in Viking-Late Norse Orkney, mounds of various sizes and in a range of locations were a regularly encountered landscape element for the diaspora community. Furthermore, the known Viking mound settlements had a landscape relationship with surrounding mounds, even if only of topographical similarity. Windblown sand-dominated Marwick Bay (West Mainland) provides a good example of mound distribution in a sandy bay: smaller mounds such as the Know of Flaws occupy the lower central area, and at least two burnt mounds are found on the higher ground; larger mounds cluster on the fertile land in the center, such as “The Castle” (the Castle name appears to be only that, with no physical evidence or memory of any castle structure) between the farms of Langskaill and Netherskaill, and the known settlement mound on the southern arm of the bay (Fig. 2; Harrison 2007). The settlement site is being eroded by the sea, and radiocarbon dates from recent rescue excavation indicate that the mound was occupied first in at least the Iron Age, with upper levels of superimposed Viking-Late Norse structures, activity areas, and midden (Griffiths and Harrison 2010; Griffiths et al. 2009). This mound is particularly visible when approached by sea; it is also next to what appears from topographical and geophysical survey to be an early chapel site. A similar pattern to that in Marwick Bay is found for most of the excavated sites: Viking settlement built over earlier settlement, or on natural mounds, sharing a prime landscape unit with earlier smaller mounds (and indeed often with broch mounds, or—in the case of the Bay of Skaill—the mounds covering prehistoric settlement at Skara Brae) and having an association with early or important ecclesiastical sites. This pattern is true, for example, of the Birsay Bay sites, Skaill in Deerness, the Bay of Skaill in West Mainland, Tuquoy and Langskaill on Westray, and probably Westness in Rousay. Some of the eroding coastal mound sites with possible Viking-Late Norse dates observed during RCAHMS surveys, in more recent coastal surveys, and by the author may be included in this pattern: for example, the Cross Kirk site in Backaskail Bay; Sanday, which sits by a chapel site and on a bay with a known broch site; Langskaill in northwest Sanday; and Saviskaill in north Rousay (Harrison 2007, Lamb 1980, OSMR records 00710, 00386, 00481). These areas are not only predominantly on windblown sand but the distribution is almost identical to that of Skaill (ON skáli) and ~skaill names (Langskaill, Backaskaill/Backaskail/Backiskaill, and Saviskaill) (Fig. 3; Harrison 2007: Marwick 1952, 1970; Thomson 2008b:1–24). There are 41 OSMR “~skaill” archaeological records; allowing for duplications of records for different aspects of the same site, all but five (in St. Andrew’s and Orphir in Mainland, Eday, Gairsay, and Egilsay) are located within one of the coastal areas identified above, rich in mounds and Viking settlement archaeology (around a quarter are directly linked to known Viking-Late Norse archaeology). Of those five, the Langskaill records on the tiny strategic island of Gairsay relate in part to a supposed Viking-Late Norse stronghold, and the two Egilsay Skaill records include a possible Viking-Late Norse coastal mound. J. Storer Clouston (1932) identified 33 Skaill sites in his History of Orkney, including, interestingly, a long vanished Langskaill in Birsay Bay and another on the Bay of Skaill. All but four of the over 50 Skaill/~skaill 133 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 names surviving today share the same coastal/prime location distribution. This coincident distribution of coastal settlement mounds, significant Viking-Late Norse sites, and Skaill names warrants investigation. However, the meaning of the Skaill place name is disputed. It has not been as closely studied by place-name scholars in the specifically Orcadian context as have other Norse names like, for example, Houseby or ~staðir (e.g., Crawford 2006). Skaill has two contrasting meanings in Old Norse: one meaning used in Viking-Late Norse Orkney and the other, more widely at least, in Scandinavia, and in northern England (where more intensive studies have been made; e.g., Fellows-Jensen 2000). Skaill (ON skáli) can mean hut/shed, temporary dwelling place or hall, or sleeping-hall (Cleasby et al. 1957; Rygh 1898:74). In northern England, the meaning hut/shed seems to apply (Fellows-Jensen 2000). In Orkney, however, skáli seems to refer to a hall and even a quite specific sort of hall. In the Orkneyinga Saga, compiled in the late 12th and early 13th century by an Icelandic historian, skáli is consistently used with a high-status connection. The drinking and feasting halls of the Earls themselves and their leading henchmen such as Thorkel Fostri and Svein Asleifsson are referred to using skáli or lángskali. Thorkel entertained and assassinated Earl Einar Sigurdsson in a skáli (Palsson and Edwards 1978:Ch. 16); Earl Erlend and his followers drank for several days in another (ibid.:Ch. 94, see also Ch. 66). Raymond Lamb notes that Skaill names were linked to prime agricultural land and significant or head farms where the parish church is found (for example Skaill in Deerness, Bay of Skaill, and Rousay); yet their medieval rental assessments are consistently low (Lamb 1997, Marwick 1952, Thomson 2008b). Lamb (1997:15) and Thomson (1995, 2008b) have both suggested this may be because the skális, rather than rich farms with high tax assessments, held public functions within the local community; in other words, the name referred to a function—to a “social role”—linked with the leading figures in Viking-Late Norse Orcadian society. More straightforwardly, it would perhaps be surprising to find so many bays in the Orkney Islands called “temporary shed” Bay! Skaill names were linked to several of the excavated Viking-Late Norse longhouse settlements: Skaill in Deerness; Skaill in Sandwick (Bay of Skaill); Birsay Bay; Westness and Swandro on Rousay; and Langskaill/ Tuquoy on Westray. Considering skáli place names in conjunction with significant Viking-Late Norse mound sites might, therefore, contribute to Figure 2. Mounds on Marwick Bay, West Mainland, Orkney Islands. J. Harrison 2013 Journal of the North Atlantic Special Volume 4 134 an understanding of the social structures shaping the settlement landscape. As explored further below, the concurrence might imply a network of focal social points spread across the main islands. In the more detailed study of settlement mounds, the main case study will be the structures and features excavated between 2005–2011 by the Birsay-Skaill Landscape Archaeology Project at the Bay of Skaill, West Mainland (Fig. 4; Griffiths and Harrison 2005, 2009, 2010, 2011, 2012; Griffiths et al. 2006, 2007, 2008, 2009). The excavations in wind-blown sand-drift geology revealed a highstatus Viking-Late Norse longhouse over 26 m long with associated outbuildings, middens, and working areas on the so-called East Mound (≈65 m in diameter, ≈3 m high; Fig. 5), and the more fragmentary remains of another considerable longhouse, with an Figure 3. Orkney Islands sand-drift geology, Viking-Late Norse sites, and Skaill names. extensive and deep mid- Figure 4. Castle of Snusgar mound and East mound on the Bay of Skaill, West Mainland, Orkney Islands, looking northwest. 135 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 excavated settlement mounds, like the one recorded by the Project in Marwick Bay 5 km north, were constructed of “tell-like” deep stratigraphy formed through a complex process of infilling, rebuilding over, and incorporating elements of existing stone structures, activity areas, and middens. Viking- Late Norse building began on pre-existing natural mounds, probably also for the Snusgar mound over earlier archaeology, and between 2–3 m in depth of stratigraphy survives (Fig. 6; Griffiths and Harrison 2009, 2010, 2012 in particular). This deep superimposition of structures and middens has been revealed at all the sites in the coastal mound group excavated sufficiently extensively: the Birsay Bay sites; Skaill in Deerness, Pool on Sanday, Langskaill and Tuquoy on Westray, Quoygrew on Westray (Barrett 2012), and is also seen in coastal mound erosion exposures, as well as in Shetland sites including Jarlshof and Old Scatness (Dockrill et al. 2010, Hamilton 1956), and the western Isles site of Bornais (Sharples 2005, 2012). Viking-Late Norse settlement sometimes encompassed more than one mound, including at Skaill in Deerness, Birsay Bay and probably Quoygrew den sequence on the mound to the west, called the Castle of Snusgar (≈75 m in diameter, over 3 m high). A sequence of radio-carbon dates suggest that the Snusgar mound may have been first occupied in the Iron Age and that Viking-Late Norse activity began in the 10th century, while the structures on the East Mound may have originated in the late 10th century, but the main focus of activity was the 11th century (Griffiths and Harrison 2009, 2010). Both were abandoned in the twelfth century. The mounds are a bare half kilometer from an early church site—St. Peter’s Kirk—and are the find-spot of Scotland’s largest Viking silver hoard (later 10th century; Graham-Campbell 1995). Burial mounds have been recorded in the center of the bay and on higher ground. A Viking burial was identified in the top of an otherwise uncharacterized but possibly Viking-Late Norse settlement mound subject to marine erosion on the south arm of the bay (Morris 1985). This site lies next to the prehistoric settlement of Skara Brae. The main farm has a Skaill name like the bay itself; a broch is being eroded from the headland of the north arm of the bay (OSMR record 01256; Harrison 2007). The two Figure 5. The longhouse on the East mound, Bay of Skaill looking southeast. Photograph © Birsay-Skaill Landscape Archaeology Project. J. Harrison 2013 Journal of the North Atlantic Special Volume 4 136 in Westray, and on Marwick Bay. At Bornais, like at the Bay of Skaill, the main focus of settlement over the period may have shifted between mounds (Sharples 2012). Thus, these sites were probably more than single “farmsteads”, but rather a cluster of ancillary buildings around a central hall or longhouse or what have in other contexts been called magnate farms (see below). The combination of prime economic coastal location, longhouse-focused settlement, early churches, and often a range of traded/ exchanged items suggests that these may have been locally focal sites—“central places”—during the Viking-Late Norse period. These coastal landscapes will be now examined in more detail; crucially, what was it about mounds that motivated Viking-Late Norse people to construct and reconstruct settlement on natural mounds, or mounds including earlier structures? How did this interest in mounds fit with economic and social arrangements? To tackle these questions the following themes will be considered: was the focus on settlement-mound building in Orkney drawing on ideas about mounds imported from Scandinavia, or was the emphasis new; how might the mounds have reinforced local social organization; and what links with the local environment were revealed in the building of settlement mounds? It will be argued that the crafting of mounds was a purposeful process embedded in cultural ideas, creating a settlement landscape that underpinned the social organization of local communities. Not only is the location and landscape context of these settlements informative, but so is the manner in which the increase of mound width and height was achieved and the range of materials used in that process. Finally, the significance of mounded coastal landscapes in the development of Viking Age socio-economic structures in Orkney will be discussed. Mounds in Scandinavia The social and economic character of coastal Northern Norway in the Early Viking Age (ca. 800–1050 AD) provides perhaps the best analogy with the Northern Isles after ca. 850 AD. Both were rugged coastal economies with wide trading and exchange links, remaining relatively politically independent from developing polities to the south (and east) until the 12th century (Skre 2008). Both were governed by hierarchical magnate cultures characterized by personal power relationships negotiated in public assemblies (Poulsen and Sindbæk 2011a). Perhaps significantly, the “farm mounds” of coastal Figure 6. Structured midden on the Castle of Snusgar mound facing southeast. 137 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 largest Central Places, such as Uppåkra, continued as stable power centers into the early Viking period, but the tunanlegg declined in number (Larsson 2007). Inger Storli notes that those tunanlegg sites that did persist held particularly dominant landscape positions; it was increasingly important for competing elites to find mechanisms for validating power (Storli 2010:138–139). The hall also emerged as a central motif in the physical expression of dominance: “structures of power were related physically to the great halls” (Poulsen and Sindbæk 2011b:27). Halls may have provided a ritual space; certainly they hosted assemblies that reinforced authority (Herschend 1993, Munch et al. 2003:265–272). In Scandinavia, the narrative of authority was often expressed using burial mounds—interpreted as both territorial and social markers (Myhre 2000). (Viking burial mounds on the Isle of Man were described as the “land charters of illiterate men” [Fletcher and Reilly 1988:97]). They helped define farmstead boundaries before the Viking Age, but the increase in burial mound numbers in Norway after 800 AD may have signalled increasing competition for status as well as land rights (Jørgensen 2000:79). The mounds represented the important dead and made visible a farmer’s right to the surrounding land’s resources, but they were also physical signals of social position. The symbolism associated with burial mounds may have been complex and dynamic. Neil Price (2010) re-examined evidence from the Kaupang cemeteries in particular, to explore the impact of the process of interring the dead. He argued that elaborate mortuary ritual forged close associations between ancestry and the mounded burial landscape. Also highlighted were several burials in which later graves were positioned—clearly deliberately—directly over earlier inhumations. This positioning fashioned clear links with earlier important people and perhaps implied the handing down of position and status. Eva Thäte surveyed the frequent re-use of burial mounds in Viking Age Scandinavia; she also noted briefly that, in a variation of the idiom, Norse settlers in the Northern Isles used the houses of previous inhabitants for burials (Thäte 2007:120–127). This is a very interesting difference. We have seen how prime Viking-Norse settlement in Orkney was located close to prehistoric burial mounds. However, Orcadian Viking burials made into existing features have, so far, been found more often dug into midden or earlier settlement, than into older burial mounds, for example at Buckquoy, Saevar Howe, Oxtro Broch, Bay of Skaill, and Sties of Brough on Sanday (Harrison 2007; Hedges 1983; Morris 1989, 1996; Ritchie 1977). This pattern suggests that settlers, coming to a place already Northern Norway also provide the closest comparison to the settlement mounds of the Northern Isles (Bertelsen 1979, etc.). These farm mounds—layers of turf building material and midden—were under construction from ca. 900 BC, but appear to take a particularly significant part in landscape organization from the Viking Age (Berglund 1997). The predominantly coastal settlement in Norway was dominated by physically isolated single or clustered farmsteads economically linked with other farms in their area, such as Lurekalven in Nordhordland, western Norway, although villages such as the classic site of Vorbasse had developed in the more fertile areas of SW Norway and southern Scandinavia (Magnus 2002:11–13, 21). At a regional level, settlement was more complicated and varied than this; for example in Northern Norway, first tunanlegg assembly sites and then magnate farms developed as local central places (Berglund 1997, Grimm 2010, Storli 2006). Mounds of various kinds were prominent in the Scandinavian landscape, and this section focuses on this array of mounds in their settlement context. The nature of lordship before the medieval period meant that sustaining power over people and resources hinged on appropriate places—places that made rights visible and at which public assemblies could be held. Mounds played a particularly significant part in the creation of such places in Scandinavia from ca. 800 AD. (Although, as in the Northern Isles, elevated locations and settlement continuity were established ways of emphasizing power before the Viking period.) The language of mounds in the Viking Age landscape was complex, so the following elements need to be considered briefly: burial mounds, “Central Place” complexes and other assembly sites, and the north Norwegian “farm mounds”. From the Migration period in Scandinavia (ca. 350–550 AD), power was gradually consolidating into fewer hands, especially in the fertile southwest and the resources-rich Lofoten Islands. During this time, high-status courtyard or tunanlegg sites were first constructed, especially in southwest and northern Norway; these locations seemed to serve as local, probably elite-managed, assembly sites (Grimm 2010). Southern Sweden (Skåne and the Mälaren region) and southern Scandinavia saw the rise of major “Central Place” complexes with large highstatus halls in conspicuous locations, such as Gudme in Lundeberg and Tissø in Zealand (Magnus 2002). Then from the Early Viking Age, ca. 800 AD, with more dynamic hierarchies and fiercer competition for power, magnate farms with major hall buildings, such as Borg in Lofoten, became more conspicuous in the settlement landscape (Munch et al 2003). Some of the J. Harrison 2013 Journal of the North Atlantic Special Volume 4 138 large feasting halls (for more on Central Place theory, see Fabech and Ringtved 1999, Grimm 2010, Skre 2011). Such complexes were long-lived features in the Scandinavian landscape creating dramatic venues for socially vital assemblies for feasting, ritual, gift-giving, and the offering of tax-in-kind. Stable power centers in southern Scandinavia like Jelling, Gudme, and Uppåkra were exceptionally rich and enduring. Uppåkra, for example, spanned the whole 1st millennium AD (Larsson 2007, Randsborg 2008). Yet they used the same language of mounds as the locally significant “magnate farm” Central Places of Early Viking Age northern Norway, such as Borg in Lofoten (Munch et al. 2003). All these Central Places deployed some or all of a range of what might be termed mound-related characteristics in their construction. Firstly and obviously, large prestige halls were built in a prominent location. Secondly, central halls usually stayed in one place as structures around changed, but the hall itself was also rebuilt several times on that same spot. The largest longhouse among the farms at Vorbasse, unlike the other houses, “rested in more or less the same location through ten generations” (Magnus 2002:12). The large hall at Lejre, Sjæland “had been built and rebuilt on the same spot over centuries”, as had that at Borg in Lofoten; and at Uppåkra, the hall had been built and rebuilt in seven phases (Larsson 2007, Munch et al. 2003, Poulsen and Sindbæk 2011b). Thirdly, many mound halls shared the landscape with existing mounds, usually burial mounds, for example at Borg in Lofoten and other northern Central Places like Tjøtta in Nordland and Bjarkøy in Troms (Storli 2010). Then the halls were often purposefully built over older significant features; the high-status 8th–10th century AD hall at Huseby in Skiringssal was built over deliberately flattened barrows, and the first hall at Borg in Lofoten was built over two early Iron Age burials (Munch et al. 2003, Skre 2007). Finally, the construction of major buildings in visually dominant mound locations demanded extra construction work. The exposed positions were not necessarily the most topographically sensible locations for, or the easiest places on which to raise, substantial longhouses. When first hall at Borg in Lofoten was re-fashioned into a huge longhouse 82 m long, it seemed imperative to whoever commissioned the building not only to retain the elevated location, and to emphasize it by directly superimposing one hall on another, but to stay on precisely the same spot. This construction involved considerable preparation: a substantial rubble platform had to be laid to keep the heart of the longer building exactly over that of the preceding hall (Munch et al. 2003). At Huseby in Skiringssal, only the preparation of a substantial soil and stone occupied by others, considered the appropriation of non-burial-related landscape features more relevant to establishing authority. As most Pictish burials on Orkney seem to have been only stone-marked, or covered with small cairns, the ancestors under earlier prehistoric mounds belonged to a distant and non-contiguous period (Ashmore 2003, Graham- Campbell and Batey 1998). Perhaps for the incomers, using midden mounds, representing fertility and abundance, or recent structures for burials made a more relevant gesture. Thus, rather than superimposing graves to recall the significant dead, they evoked enduring power by layering settlement. In Orkney, therefore, the narrative of success was written with settlement mounds. Recent evidence from a burial mound site in Norway highlights the possible significance of investigating the composition and detail of mound stratigraphy. Terje Gansum and Terje Oestigaard (2004) studied the deep stratigraphy of two huge supposed burial mounds at Haugar in Tønsberg. The mounds were found to be huge cenotaphs, despite being claimed by13th-century historian Snorri Sturluson as the burial mounds of the sons of Harald Finehair (ca. 860–940 AD), the first king to claim sovereignty over “all” Norway. The mounds’ creation in the 9th century had involved the manufacture, transportation, and spreading of vast quantities of charcoal, along with turf and soils, apparently from different parts of Tønsberg. Considerable organizational power must have been exercised to bring together people from across the area to build these imposing status-symbols. The mound itself, without a burial, seems to have been enough. Mounds and the process of building mounds were thus more broadly associated with the signalling of local power. The mounds were also connected with formal assemblies—the gatherings of leaders and retainers for feasting, decision-making, and affirmation of authority. Such events were enshrined in early Norse law as regional law-giving and administrative thing assemblies, where mounds provided both focal points and physical references to landscape history (Sanmark and Semple 2010). But, before the process was formalized, similar gatherings must have been held, and at other suitable locations such as the tunanlegg sites and central longhouses. Interestingly, the tunanlegg also incorporated small constructed mounds into their layout, mounds often built over earlier features (Grimm 2010:ch. 4). There was an increasingly nuanced association between mounds and the affirmation of leadership in a public context, and this association was a persistent landscape metaphor. This metaphor continues to be compelling when considering Central Place complexes built around 139 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 building platform on the summit ensured the conspicuous hall could dominate the landscape around. Similarly, the visually dominant halls of Uppåkra in Scania were raised on preparatory house platforms (Larsson 2007). Many of these characteristics can be discovered in “lesser” Central Places, that is relatively minor magnate farms providing central functions for surrounding communities. For example, an otherwise unexceptional building in Alby in Sweden, interpreted as a small feasting venue, was raised on a considerable stone terrace to ensure visual dominance (Hjulström et al. 2008). The metaphor translated down the social scale as a way of demonstrating ever more local significance. Farm mounds are an interesting case. No convincing overall practical argument for the creation of these accumulations of turf building material and midden has been forthcoming (although Mook and Bertelsen [2007] suggest that living on mounds rich in gradually decaying material helped raise the temperature inside the farm [see also Urbánczyk 1992:105–121]). A significant number of these mounds originate in the Late Norse period (ca. 1100 AD), and it is perhaps in this period that farm mounds stood out as one of the few landscape features marking local (non-religious) authority (Berglund 1997). Imposing burial mounds were a thing of the past in Christian Norway, as were chiefly feasting halls. However, a powerful local farmer could still use a settlement mound to proclaim the time-depth of rights held over land, and the family’s dominance of the immediately surrounding area. Viking settlers therefore took a rich mound-related landscape vocabulary into the Orcadian context. They were familiar with landscapes in which visually dominant mounds were associated with the physical expression of rights and status. Earlier features were sometimes subsumed into mounds, or mounds re-used—in any case, earlier monuments were widely referenced by the form. Halls and mounds were a particularly powerful combination, with the main building often staying in a precisely fixed location despite being frequently reconstructed. The stability of location and the constructing of layers perhaps suggested the accumulation of a history of authority. Mounds and mound-related buildings seemed to take a focal role: whether in the process of construction, or in what they represented, or in what transpired in the buildings. The Orcadian Mounds What happened, then, as Viking leaders and settlers negotiated, or otherwise acquired, land rights and places to live and work in already well-settled Orkney? The next sections look at the settlement landscape in more detail to assess which elements of the mound “vocabulary” suggested above can be discerned in the Orcadian context. As we have seen, there were many constructed mounds in the pre-Viking settlement landscape of Orkney, especially in the sandy bays predominantly selected by the incomers. In the Western Isles, the machair was also rich in earlier archaeological mounds (Sharples 2012). Just as in Marwick Bay and at the Bay of Skaill (see above; Griffiths and Harrison 2012, Griffiths et al. 2009), the Viking- Norse settlement at Skaill in Deerness shares the bay with a broch mound, earlier settlement mounds, and burnt mounds (Buteux 1997), the landscape at Westness (Skaill) on Rousay is scattered with mounds and reminders of past settlement (Kaland 1995), and coastal mounds at Northskaill in northwest Sanday share the area with prehistoric mounds (Harrison 2007, Owen and Dalland 1999). Arriving in this already structured and well-exploited landscape, the first Viking settlers on Orkney needed to exercise all possible means to legitimize their claims to control land and to reinforce their growing cultural dominance. Shaping the landscape was one means, and might be done deliberately and also achieved incidentally as the result of actions integral to everyday life. The latter activity was routine and unexceptional, but a vital part of living in and exploiting environments. It is argued that landscapes thus became “infused with meaning” (Parker Pearson 1999:139). They came to bear the changing imprint of a creative and essentially material relationship between people and their surroundings, particularly potent in a pre-Modern context (Thomas 2007). Places—or spaces with association and history— were and are “produced by people” in many different ways (Giddens 1984; Thomas 1999:83–89, 2012). Ordinary subsistence routines—fishing, cooking, and cleaning out the byre—contributed to that process, as much as grander schemes like constructing a longhouse. The concept of habitus—the significance of habitual, everyday actions—suggests that routine activities wove social norms into the physical fabric (Bourdieu 1977). Nor was that engagement solely practical, even in the challenging environments Viking-Norse people encountered in Orkney or Northern Norway. Landscapes offer—afford— both practical and symbolic potential (Gibson 1977). So a burial mound, physically marking a boundary, also reminded passers-by of ancestral rights over land; a longhouse provided dignified shelter and reminded locals of the outcomes of the important meetings it hosted. Tim Ingold has developed these ideas with a more kinetic analysis of the constantly formative, reflexive, and material J. Harrison 2013 Journal of the North Atlantic Special Volume 4 140 relationship built over time between people and their places—observing the movements of everyday practice that become enmeshed in the landscape (e.g., Ingold 1993:52–58). These ideas of process and the role of routine tasks in creating visually informative, socially significant landscapes are of particular relevance in this analysis, as will be discussed further in the next section. How did the diaspora community “produce” their landscapes? In the already mound-rich coastal areas, many of the Viking-Norse longhouse settlements were built literally over and into the mounds of Pictish/Iron Age, and earlier, dwellings: on the Bay of Skaill, at Skaill in Deerness, and at sites on Birsay Bay such as at Pool on Sanday, at Langskaill on Westray, and on Marwick Bay. Building over and incorporating existing structures sent an unambiguous and unsubtle message: the past of the area was co-opted and the land takeover demonstrated beyond doubt. An immediate link was created with previous important people, and in referencing mounds the arrivistes appropriated landscape symbolism familiar to their own communities and indigenous groups. Those with social ambitions continued the theme, settled in a prime location on a visually dominant mound, whether natural or created by earlier settlement, and built a hall—perhaps a skáli. Then succeeding generations increased the mound, a process that wrote a narrative of dominance into the landscape. The exceptional aspect of this narrative in Orkney is the extent to which the focus was on augmenting settlement mounds, during the later 10th and 11th centuries in particular. This process was reliant on organic material and midden-type deposits and thus analogous to the methods used to create Norwegian farm mounds. The known Viking-Late Norse settlement was dominated by these mound forms, and therefore by a particular approach to building, until the marked 12th-century change in the settlement pattern (see below). (The question of where any obviously lower status settlement was located in the Northern Isles is interesting and problematic. Very few settlements less notable for trade/exchange goods, size, or location have been discovered, except perhaps Quoys in Deerness [Steedman 1980] and possibly Beachview in Birsay Bay [Morris 1996]. Even Quoygrew in Westray, characterized as wealthy but perhaps not high status, remains debatable, as the main building may still lie unexcavated within one of the mounds [Barrett 2012]. It may be that many of the less wealthy were housed within the longhouse clusters, and that any other less substantial buildings are either still beneath occupied buildings— now mostly found on then less-favored land, higher up around the bays—or that archaeologists have yet to find or recognize a different category of settlement less vulnerable to coastal erosion. However, it is also probably the case that the mound settlements, although dominant in their area, were not the most socially elevated sites of this period; those were perhaps the “chiefly citadels” on the Brough of Deerness [Barrett and Slater 2009] and Brough of Birsay and the Bu sites like Orphir on Mainland.) The settled mounds undoubtedly increased in size—both widened and were built higher—as Viking-Late Norse buildings, yards, and middens were superimposed. To achieve this in virtually treeless Orkney and Shetland, local building practices were adopted, using stone and some turf rather than wood, construction materials that were also familiar to settlers from western and in particular northern Norway. The Rousay flagstone formations of the Orkney Islands and the abundant beach cobbles provided useful and plentiful stone for building. But the settlement form was distinctive, as rectilinear longhouses took over from curvilinear cells (Buteux 1997, Ritchie 1977). Just as elements of local construction and economic practice were amalgamated with imported practices, the representational landscape language of mounds seems to have been adapted. The focus was on settlement and on the process of building mounds that incorporated stonebuilt hall-longhouses, their ancillary buildings, and middens. The remainder of the article explores how the translated idiom acquired substance. The landscape social context was not dissimilar to western Norway, although more generally fertile, but it was for the Vikings recently settled, not ancestral, land. The Mounding Process: Building a Settlement Mound This section will study the archaeological detail of settlement mounds to reveal the effort and deliberation involved in repeated rebuilding on such sites (Harrison 2007). These mounds were not usually in places so space-limited that direct superimposition was the only option, nor were stone and turf, the basic building materials for walls and floors, in short supply or difficult to access. These facts seem to suggest that building over was purposeful. This method also necessitated the use of the highly organic materials produced by agriculture and fishing and therefore relied on the farmers’ understanding of such products’ properties. The majority of the Viking-Norse settlement mounds in Orkney are built in pure windblown-shell sand-drift geology (including the Bay of Skaill, Marwick Bay, and Birsay Bay sites; Skaill in Deerness; the Sanday sites; and Tuquoy and Langskaill in Westray). The drifting and rapid piling of sand itself added 141 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 were only partially demolished so their remnants could be incorporated into the foundations of higherlaid surfaces. Other surfaces around and between structures were consolidated with organic middenmaterial or levelled through selective building up. Such efforts on a sand-rich mound would have been impossible without the supplementing and stabilizing of loose sand with a range of organic material. Various combinations of ash, byre cleanings (principally dung, hay, and turf), curated domestic midden (waste from bait and food preparation and cooking) and old turf were used—spread over and mixed with sandy surfaces. The following sequence illustrates these techniques. On the East Mound, Bay of Skaill, the earliest datable excavated structure, ancillary to the longhouse, was a small north–south building possibly in use in the late 10th century. A stretch of wall belonging to that structure was then incorporated into the walling of a new and differently aligned building. The original floor of the second building was a third of a meter higher, laid on old floors and a levelling infill of midden-rich sand. Then that building was filled with nearly a meter depth of more organic sand to prepare for a further realignment of the out-building. An original wall had to be re-faced to allow its upper courses to be re-used as foundations. Later again, when the building was shortened to the west, this wall was lowered and, along with the accumulated infill, capped with clay and flags to form a passage-way. This process was revealed on excavation as an archaeological “layer-cake” comprising structural stone and more or less sandy deposits augmented with decayed turf and peat, ash, byre cleanings, and domestic rubbish (Griffiths and Harrison 2010). A metal-working area was in use to the west of that out-building during the middle period of its history. As a result of all the alterations to the structure, the thick sand-interleaved sequence of metalworking floors was discovered at a level well above the original, capped eastern wall, but sealed below the bottom courses of the western wall of the final structure. These changes seemed to have happened (along with at least one further reconfiguration) over less than a century and within an area less than 8 m2. The building was abandoned sometime in the 12th century. Space available for expansion was not used. The archaeological evidence suggests it was as much work to superimpose structures as it would have been to prepare new foundations on virgin windblown sand. Long trenches excavated down the flanks of both the East and Snusgar mounds revealed how spreads of sand with midden material and rich midden together also provided bulk for the mounds to the mounds’ bulk. But flying in the face of the biblical injunction on wise places to build also demanded a particular approach to construction. Pure windblown sand is freely moving under pressure and in the wind so, before any walls were built or flags lain, pure sand had to be stabilized (see Crawford’s 1978 lament on the difficulties of containing pure sand when excavating). As structures, yards, drains, and middens were superimposed in sand mounds, surfaces had to be prepared with organic additions. Only by repeating this process as sand drifted, could stone wall foundations be reconfigured and re-aligned, or various surfaces re-used and built over. Two brief case studies will illustrate the techniques required by the process of concentrating on building over rather than out or anew. These are, however, methods that can be detected in reports on similar sites already published, recorded in the local Sites and Monuments Records, or visible in eroding sections across the Northern Isles (Harrison 2007). On the East mound and the Castle of Snusgar mound in the Bay of Skaill, careful preparation and particular construction approaches were required to re-build on one location, sometimes on pure sand, often to exploit existing features. Those techniques included: spreading clay layers or organic midden material to create firm surfaces on pure or mainly sand layers; digging wall foundations down into older, settled middens, or into areas already consolidated with organic material (yards for example); infilling older buildings, quite often in one, or a rapid series of events, to create a platform for a new building; and re-aligning and re-facing walls, or re-using stretches of partially buried walls to build annexes or differently configured structures (Figs. 5, 6). Floors accumulated rapidly within buildings, even without deliberate in-filling; flagged floors were usually repaired by adding a second or more layer of slabs. Sand blows would adhere to organicrich yards, middens and surfaces adding, more gradually, to the bulk of the mound and requiring the laying down of further often quite thick layers of stabilizing midden-enriched sand or turf. Storage middens themselves were carefully managed, and their accumulation also added bulk and structure (see below) to mounds. All these techniques created deep mound stratigraphy. The East Mound at the Bay of Skaill, Sandwick developed well over 2 m of extra height in what may have been little more than three generations (radiocarbon dates for the site are still undergoing final Bayesian analysis). On that East Mound, in the main longhouse and the smaller buildings linked to it in the south, some lengths of walls were strengthened, shored, or realigned to allow re-building; other redundant walls J. Harrison 2013 Journal of the North Atlantic Special Volume 4 142 Did this work with midden-type material also in any way strengthen the ideological construction of the settlement mounds? This much-moved and used “rubbish” resulted from daily agricultural, fishing, and domestic activities and therefore encapsulated the rhythms and actions of everyday life. Economic activity was essential to the building program as well as to sustenance; more midden material enabled more, and more effective, building, and thus economic success was reflected in the development of mound settlements. The targeted re-use of such material also demonstrated the fisher-farmers’ awareness of the full potential of agricultural and domestic by-products. In prehistoric archaeology, it has been argued that the accumulation and re-use of middens helped create a strong sense of place: “a visual reminder of past presence and activity” (Lamdin-Whymark 2008:57, Thomas 2012). Joshua Pollard (2001) has suggested that the residues from settlement activity deposited in Neolithic pits and on large midden accumulations had gained potency from their involvement in social, and indeed economic, life. This enhanced significance may have been particularly true of the remains of feasting (McOmish 1996). The use of midden material in building mounds may have had similar resonances in the Viking-Late Norse period—connecting control over land, and people’s use of that land, with the development of the settlements. There are certainly examples of the careful management of middens, as stores of material for other purposes as well as “terracing” or structural midden spreads. At the Bay of Skaill sites, several accumulations of ash were found carefully clay-capped to secure the material (one of these also had spade marks showing where material had been removed for redeposition). At Pool, one area of midden was even centrally located in the settlement and surrounded by a wall (Hunter et al. 1993, 2007). Thus organic-rich material was not only re-used immediately, but also stored for later use to stabilize and build-up surfaces, and terraced to help create the curves of the mounds. Settlement mounds were visual representations of past activity in the surrounding landscape, and their bulk demonstrated the inhabitants’ control over local economic resources. However, perhaps more specifically for mounds associated with halls, the plentiful midden used in their construction proclaimed not only ownership of plenty of stock and successful fishing, farming, and gathering, but also generous feasting and control over local labor. The conspicuous curves of growing sand mounds intimated success and social dominance. The regular, perhaps generational, remodelling of buildings may also have been significant, to be compared perhaps with superimposed graves beyond the buildings. The deep stratigraphy of the Snusgar mound in particular showed how different combinations of material were laid at different times and how layers were often contained using rough curves of stone, clearly not paths, which followed the mounds’ contours (Fig. 6; Griffiths and Harrison 2009). These same layering techniques and use of organic material are found at Pool on Sanday, as well as at Old Scatness in Shetland (Dockrill et al. 2010), Bornais in South Uist (Sharples 2005), Tuquoy and Langskaill in Westray, and Skaill in Deerness. Pool, for example, was a multi-period mound originating in the Neolithic, within which the layering of Viking structures over Iron Age buildings and the continued superimposition of buildings into the late Norse period is recorded (Hunter et al. 2007). Midden-rich material was spread to create working surfaces, as thick foundations for flagging and site levelling before re-structuring. The incorporation of idiosyncratic relict walling is recorded, often conditioned partly by what went before, and it is suggested that “the location itself was somehow more important than the design” (ibid.:112). Midden spreads were also carefully structured as at the Bay of Skaill. The excavators of Pool term such extensive midden deposits retained by low arced walling “terraced” midden. Again at Pool, as at most of the excavated sites, there seems to have been sufficient space to spread settlement. Thus, building sequences on coastal settlement mounds concentrated effort on referencing and incorporating older buildings. Structural alterations seem to have been aimed at achieving change while maintaining a very particular location in a visually dominant position. Midden accumulations were crucial to this layering-up of mounds. Some midden will have been used to fertilize infields (and small kitchen gardens like the one found on the Castle of Snusgar mound). But some must have been stored for subsequent use in building, and material otherwise found in middens was also used directly from the source (byre or hearth, etc.)—to consolidate surfaces for example, or as managed spreads to enhance and firm the slopes of the mound itself. As buildings were remodelled to incorporate elements of the preceding one, ash (especially seaweed and peat ash), turf, domestic midden and byre waste were all essential to construction. Different recipes were used according to location: ash from cooking hearths was used to create firm working surfaces, midden and byre cleanings were applied to outside surfaces and wall cores, while old turf was used to back stone wall facings and to level up areas. As Hunter et al. (2007:142) observed, all this necessitated a “complex cycle of deposition, removal and redeposition” . 143 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 coincidence of skáli—feasting-hall—place names with the areas dominated by significant Viking- Late Norse mound sites might provide further clues to the social organization of areas around Skaill in Deerness, the Bay of Skaill, the Skaills on Rousay, Langskaill on Westray, and the ~skaill sites on Sanday and in Marwick Bay. The halls and their cluster of buildings, perhaps encompassing more than one mound, may have acted as focal locations for their areas in two different but linked ways: as places where passing, powerful men stopped to exercise their authority and largesse, but also central places regulating the activities of the local population. The skáli sites are linked in the Orkneyinga Saga with the veizla system, the reciprocal arrangements through which peripatetic Earls and important men demanded formal hospitality, or veizla, from supporters, who were in return offered protection, gifts, and the chance of advancement (Lamb 1997:14–15, Thomson 2008b:16–17). There are references in the Orkneyinga Saga to important political players being á veizlu in Skaill locations during the negotiation of alliances and attempts to heal quarrels (e.g., Palsson and Edwards 1978:ch. 16 dealing with Earl Einar and Skaill in Deerness, chs 65 and 74 dealing with Earl Paul Håkonsson and Skaill, Rousay). This veizla system would operate through a network of linked sites across the cultural landscape of the Orkney Earldom. At Snusgar, the size of the hall—at over 26 m, one of the longest in the Orkney Isles—allowed for a considerable gathering to feast, and the outbuildings’ capacity provided for ample storage (while, interestingly, the barn-end of the longhouse allowed for very few animals to be quartered; the emphasis was on storage and working areas). There was a need in the settlement structure of Orkney to have focal social points—to have halls associated with formal hospitality and the activities and decisions of powerful groups. So, on the one hand, skális were places that important people passed through, stages for public feasting where tribute-in-kind could be collected and consumed. On the other hand, the local economy and society also needed a focus for local collaboration. Although the settlement pattern was dominated by scattered farmsteads, aspects of the broad-based economy demanded that groups co-operated—in deep-sea fishing, harvesting, building, maintaining the health of flocks and herds, and making decisions about land-use and common land. The skáli-type halls on mounds could have acted as local central places, gathering points where local decision-making meshed with the demands of the over-arching Earldom system. For the farmers, fighters, and fisherman of Bay of Skaill must surely have also had to co-operate to meet the demands of the veizla system. and re-use of burial mounds in Scandinavia. All the alterations made to main longhouse buildings seemed to have remained obvious. There was little if any attempt to disguise the process of accretion. At the Bay of Skaill sites, at Pool on Sanday, and at sites in Birsay Bay—and clearly visible in plans of Jarlshof—reconfigured stretches of walls are often built in a different way, of different widths and on noticeably different alignments. Both Hunter et al. (2007) and Niall Sharples (2005) remarked on the piecemeal and varied walls at Pool and Bornais, respectively. In the long East-mound longhouse, changes in coursing and alignment are clear; the contrasts seem deliberate. Perhaps the visibility of change helped remind those using the buildings of settlement longevity and enduring local dominance, despite inevitable personal leadership change. Thus, elements of the landscape vocabulary of mounds in Orkney reflected those discussed for Scandinavia. But the emphasis was much more on the settlement mound and its cluster of buildings. The main building stayed in a precisely fixed location while being frequently reconstructed, relying on the extensive and particular use of organic material to achieve that result. Buildings, including previous structures, and what happened in them and around them were being referenced. The next section looks at the socio-political context and argues that social position in the landscape was monumentalized in the building process. Mounds and the Socio-political Landscape What role might have been played in the development of Viking Age socio-political structures in Orkney, by the crafting of mounded coastal settlements? Here mound symbolism is examined in the context of the developing and, by the early 10th century, established Orkney Earldom. The Earldom was a hierarchical society and, at least until the 12th century, structured by personal contacts maintained by a peripatetic leadership through formalized hospitality. The first section discussed the coincident distribution of the prime coastal mounded landscapes so environmentally attractive to Viking-Norse settlers, with Skaill—ON skáli—or ~skaill place names. This place-name evidence can contribute to a reconstruction of how power relationships in the Earldom might have been organized. Place names are of course difficult to work with: their age is uncertain, their locations shift, they change in meaning, and they disappear. Two disappeared Langskaills have already been noted (and see Barrett [2012:27–28] regarding historical changes in the “Quoygrew” settlement). Against this cautionary background, I would only suggest that the J. Harrison 2013 Journal of the North Atlantic Special Volume 4 144 In the Orkneyinga Saga, the farmers are repeatedly referred to as a group who are both put upon and together wield a degree of power, such as in Thorkel’s negotiations with Earl Einar (Palsson and Edwards 1978:ch. 14). This degree of cooperation suggests groups who would have needed places to meet. Power centers on the Brough of Deerness and the Brough of Birsay may have been the pinnacle of the Earldom settlement pattern, but the mound settlements seemed to have articulated the wider social landscape of the coast. Conclusion Thus, skáli-places and mounds played an important role within the settlement organization of the Viking-Late Norse Orkney Islands. The mounds were social reference points in the landscape as well as geographic ones. They were visually dominant and played a “central place” role. They referenced earlier Viking and pre-Viking settlement, as well as other mound forms, creating an association with land rights and power; they also represented the linkage of daily activity across the landscape to veizla-related feasting and assembly. They thus reflected and reinforced the social organization of the Earldom during the period from perhaps the early 10th century into the 12th century. The emphasis was on mound-located settlement and the process of building mounds. In Norway, the language was more diverse and, although some high-status settlement mounds took similar roles in this period and in particular in Northern Norway, burial mounds remained potent landscape symbols in the negotiation of local power relationships into the 11th century. In Orkney, high-status Viking-Norse mound settlements were therefore part of the symbolism underpinning Earldom power structures. It was a settlement landscape appropriate to that culture. Economy, social organization, and long-rooted but evolving symbolism had constructed a landscape which reflected and reinforced social arrangements. In the 12th century, the settlement landscape began to change. Many of the mound settlements were abandoned into the 12th century or, like Tuquoy on Westray and Skaill in Deerness, seemed to change character. Most settlement moved off high-maintenance mounds and the sand, onto higher ground around the bays: this shift of settlement was less obvious only on predominantly flat and sandy Sanday, where there were few alternative locations (Davidson et al. 1983, 1986). Some of the few coastal sites that continued became lower-status specialized sites, for example, in fishing as at St. Boniface on Papa Westray (Lowe 1998) and perhaps Quoygrew on Westray. One or two were reconstructed as minor “castle sites”, as at Westness on Rousay and probably Skaill in Deerness—small rectangular or square buildings with a tower (Grieve 1999). The same 12th-century abandonment of the sand coastal sites is seen in the Western Isles (Parker Pearson 2012). The authority of the Earls declined after the 1194 battle of Florvåg, and their opportunities for mercenary and trading activity diminished. The administrative systems of Latinate Christian kingship began to permeate the Earldom; the economy became more specialized. Much of the settlement of authority moved to the proto-urban center of Kirkwall on Mainland. Thus, the veizla system no longer dominated the relationship between magnates and the local communities. The symbolism of the settlement mounds lost social relevance, and mound building no longer seemed worth the effort. It is at this time, when coastal settlement mounds were being abandoned in the Northern and Western Isles, that the farm mounds of northern Norway show a flourishing and new dominance of the landscape. Perhaps across the network of cultural connections linking the Northern Isles and Scandinavia, the symbolism of settlement mounds had found a renewed currency in the local power structures of later Viking northern Norway. Acknowledgments My thanks go to David Griffiths, the Project Director of the Birsay-Skaill Landscape Archaeology Project, for which I have been privileged to be Assistant Director and Excavation Supervisor, and to Carol and Bruce Hallenbeck, who have so generously provided financial support for my DPhil at the University of Oxford. Thanks also to all the experts, supervisors, students, and volunteers who have joined us on the Project. You have been a great pleasure to work with. I would like to thank in particular: Diane Alldritt and Ingrid Mainland; Kat Hamilton, Vix Hughes, and Andy Beverton; Sue Hanshaw and Fay Pendell; and, of course, Pauline, Freddie, and Michael Brass, the wonderful landowners. I am also very grateful for the encouragement and comments of James Barrett and Alex Sanmark in preparing the article and for the help, in particular on research trips, of Julie Gibson, the Orkney County Archaeologist, and Roderick Thorne on Sanday. All remaining mistakes are mine. Literature Cited Ashmore, P. 2003. Orkney Burials in the First Millennium AD. Pp. 35–50, In J. Downes and A. Ritchie (Eds). Sea Change: Orkney and Northern Europe in the Later Iron Age AD 300–800. The Pinkfoot Press, Balgavies, Angus, UK. Barrett, J.H. (Ed.) 2012. Being an Islander: Production and Identity at Quoygrew, Orkney AD 900–1600. McDonald Institute for Archaeological Research, University of Cambridge, UK. 145 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 Barrett, J.H., and A. Slater. 2009. New excavations at the Brough of Deerness: Power and religion in Viking Age Scotland. Journal of the North Atlantic 2:81–94. Batey, C.E. 1987. Freswick Links, Caithness. A reappraisal of the Late Norse site in its context. Parts i and ii. BAR British Series 179 (i) and (ii), Oxford, UK. Berglund, B. 1997. Changes in the power structure around AD 1100 on the northern Norwegian coast, Pp. 263–270, In G. de Boe and F. Verhaeghe (Eds.). Papers of the Medieval Europe Brugge 1997 Conference Volume 6. Instituut voor het Archeologisch Patrimonium,Belgium Bertelsen, R. 1979. Farm mounds in North Norway, A review of recent research. Norwegian Archaeological Review 12(1):48–56. Bertelsen, R. 1984. Farm mounds of the Harstad area. Acta Borealia 1:7–25. Bertelsen, R. 2011. Tilkomsten av Fiskevær, med særling blikk på kysten mellom Vestfjordenog Lopphavet. Pp. 78–88, In L.I. Hansen, R. Holt and S. Imsen (Eds.). Nordens Plass i Middelalderens Nye Europa: Rapport til det 27 Nordiske Historikermøte, Tromsø, 11–14 August 2011. Orkana, Stamsund, Norway. Bigelow, G.F. 1985. Sandwick, Unst, and Late Norse Shetland Economy. Pp. 95–127, In B. Smith (Ed.). Shetland Archaeology: New Work in Shetland in the 1970s. Shetland Times Ltd, Lerwick, UK. Bond, J. 1998. Ashes to the Earth: The making of a settlement mound. Pp. 81–96, In R.A. Nicholson and S.J. Dockrill (Eds.). Old Scatness Broch, Shetland: Retrospect and Prospect. North Atlantic Biocultural Organization, University of Bradford, Bradford, UK. Bond, J. 2007. Viking Unst Project: Excavations at Hamar and the upper house, Underhoull: Field season 2007 Interim Report no. 2. Shetland Amenity Trust and University of Bradford, Bradford, UK. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge University Press, Cambridge, UK. Buteux, S. 1997. Settlements at Skaill, Deerness, Orkney: Excavations by Peter Gelling of the prehistoric, Pictish, Viking and later periods, 1963–1981. BAR (British Series 260). Archaeopress, Oxford, UK. Childe, V.G. 1950. Skara Brae. Edinburgh, UK. Cleasby, R., Vigfusson, G. and W. Cragie. 1957. Icelandic- English Dictionary. Oxford, UK. Clouston, J.S. 1932. History of Orkney. W.R. Mackintosh, Kirkwall, UK. Crawford, B.E. 2006. Kongemarkt og jarlmarkt, stedsnavn som bevis? Betydningen av Houseby, Harry og –staðir navn på Orknøyenes West Mainland. Viking 69:195–214. Crawford, I. 1978. Archaeological prospect and practical technique in an environmental region: The Western Isles Machairs. World Archaeology 10(1):51–62. Davidson, D.A., R. Lamb, and I. Simpson. 1983. Farm mounds in North Orkney: A preliminary report. Norwegian Archaeological Review 16(1):39–59. Davidson, D.A., D.D. Harkness, and I.A. Simpson. 1986. The formation of farm mounds on the Island of Sanday, Orkney. Geoarchaeology 1:45–60. Dawson, T. (Ed.). 2003. Coastal Archaeology and Erosion in Scotland. Historic Scotland, UK. Dockrill, S., J. Bond, and V. Turner. 2010. Excavations at Old Scatness, Shetland. Vol. 1: The Pictish Village and Viking Settlement. Shetland Heritage Publications, Lerwick, UK. Fabech, C. and J. Ringtved (Eds.). 1999. Settlement and landscape: Proceedings of a conference in Arhus, Denmark, May 4–7, 1998, Moesgård. Jutland Archaeological Society, Aarhus, Denmark. Fellows-Jensen, G. 2000. The Vikings in the British Isles: The Place-Name Evidence. Acta Archaeologica 71:135–146. Fletcher, M., and P. Reilly. 1988. Viking settlers in the Isle of Man: Some simulation experiments. Pp. 95–117, In C.L.N. Ruggles and S.P.Q Rahtz (Eds.). Computer and Qualitative Methods in Archaeology 1987. BAR Intl Series 393, Oxford, UK. Gansum, T., and T. Oestigaard. 2004. The ritual stratigraphy of monuments that matter. European Journal of Archaeology 7(1):61–79. Gelling, P. 1984. The Norse buildings at Skaill, Deerness, Orkney, and their immediate predecessor. Pp. 12–39, In A. Fenton and H. Palsson (Eds). The Northern and Western Isles in the Viking World: Survival, Continuity, and Change. John Donald Publishers Ltd, Edinburgh, UK. Gibson, J.J. 1977. The theory of affordances. In R. Shaw and J. Bransford (Eds). Perceiving, Acting, and Knowing. Erlbaum, Hillsdale, NJ, USA. Giddens, A. 1984. The Constitution of Society: Outline of a Theory of Structuralism. Polity Press, Cambridge, UK. Graham-Campbell, J. 1995. The Viking-Age Gold and Silver of Scotland. National Museums of Scotland, Edinburgh, UK. Graham-Campbell, J., and C.E. Batey. 1998. Vikings in Scotland: An Archaeological Survey. Edinburgh University Press, Edinburgh, UK. Grieve, S.J. 1999. Norse castles in Orkney. Unpublished Ph.D. Dissertation. University of the Highlands and Islands, Inverness, Highland, UK. Griffiths, D., and J. Harrison. 2005. Birsay-Skaill landscape archaeology project, Orkney. Data structure report 2004–2005, BS04 (Birsay) and SG04 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., and J. Harrison. 2009. Birsay-Skaill landscape archaeology project, Orkney, phase X, data structure report 2008, SG08 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., and J. Harrison. 2010. Birsay-Skaill landscape archaeology project, Orkney, phase XII, data structure report 2010, SG10 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., and J. Harrison. 2011. Settlement Under the Sand: New Viking discoveries in Orkney. Current Archaeology 22(1):12–19. Griffiths, D., and J. Harrison. 2012. Birsay-Skaill landscape archaeology project, Orkney, phase XIII, data structure report 2011, SG11 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., J. Harrison, and S. Hanshaw. 2006. Birsay- Skaill landscape archaeology project, Orkney, phase IV. Data structure report 2005, SG05 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. J. Harrison 2013 Journal of the North Atlantic Special Volume 4 146 Griffiths, D., J. Harrison, and S. Hanshaw. 2007. Birsay- Skaill landscape archaeology project, Orkney, Phases V, VI and VII. Data structure report 2006, SG06 (Snusgar), survey and excavation phase VI. Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., J. Harrison, and S. Hanshaw. 2008. Birsay- Skaill landscape archaeology project, Orkney, phases VIII, and IX. Data structure report 2007, SG07 (Snusgar). Unpublished University of Oxford Report, Oxford, UK. Griffiths, D., J. Harrison, and M. Athansan. 2009. Birsay- Skaill landscape archaeology project, Orkney, phase XI, Data structure report 2009: Marwick Bay MW09. Unpublished University of Oxford Report, Oxford, UK. Grimm, O. 2010. Roman Period Court Sites in Southwestern Norway: A Social Organisation in an International Perspective. Museum of Archaeology, University of Stavanger, Stavanger, Norway. Hamerow, H. 2002. Early Medieval Settlements: The Archaeology of Rural Communities in Northwest Europe, 400–900, Oxford University Press, Oxford, UK. Hamilton, J.R.C. 1956. Excavations at Jarlshof, Shetland. Her Majesty’s Stationery Office (HMSO), Edinburgh, UK. Harrison, J. 2007. Working with sand: Coastal living in Viking and Late Norse Orkney. Unpublished M.Sc. Thesis. University of Oxford, Oxford, UK. Hedges, J.W. 1983. Trial excavations on Pictish and Viking settlements at Saevar Howe, Birsay, Orkney. Glasgow Archaeological Journal 10:73–124. Herschend, F. 1993. The origin of the hall in Southern Scandinavia. TOR 25:175–199. Hill, J.D. 1995. Ritual and rubbish in the Iron Age of Wessex: A study on the formation of a specific archaeological record. Tempus Reparatum, Oxford, UK. Hjulström, B.H., S. Isaksson, and C. Karlson. 2008. Prominent migration period and later building: Alby, Botkyrka, Södemanland, Sweden. Acta Archaeologica Supplementa 79:62–78. Hunter, J.R., J.M. Bond, and A.W. Smith. 1993. Some aspects of early Viking settlement in Orkney. Pp. 272–284, In C. Batey, J. Jesch, and C.D. Morris (Eds.). The Viking Age in Caithness, Orkney and the North Atlantic. Edinburgh University Press, Edinburgh, UK. Hunter, J.R., J.M. Bond, and A.N. Smith (Eds.). 2007. Investigations in Sanday, Orkney. Vol.1: Excavations at Pool Sanday, a Multi-Period Settlement from Neolithic to Late Norse Times. The Orcadian Ltd/Historic Scotland, Kirkwall, UK. Ingold, T. 1993. The Perception of the Environment: Essays on Livelihood, Dwelling, and Skill. Routledge, London, UK. Jørgensen, L. 2000. Political organization and social life. Pp. 72–85, In W.W. Fitzhugh and E.I. Ward (Eds). Vikings: The North Atlantic Saga. Smithsonian Institute Press, Washington, DC, USA. Kaland, S.H.H. 1995. The settlement at Westness, Rousay. Pp. 308–317, In C. Batey, J. Jesch and C.D. Morris (Eds.). The Viking Age in Caithness, Orkney and the North Atlantic. Edinburgh University Press, Edinburgh, UK. Lamb, R. 1980.The Archaeological Sites and Monuments of Sanday and North Ronaldsay. RCAHMS List No. 11, Edinburgh, UK. Lamb, R. 1997. Historical background to the Norse Settlement. Pp. 13–16. In S. Buteux (Ed.). Settlements at Skaill, Deerness, Orkney: Excavations by Peter Gelling of the Prehistoric, Pictish, Viking, and Later Periods, 1963–1981. BAR (British Series 260). Archaeopress, Oxford, UK. Lamdin-Whymark, H. 2008. The Residue of Ritualised Action: Neolithic Deposition Practices in the Middle Thames Valley. BAR Br Series 466. Archeopress, Oxford, UK. Larsson, L. 2007. The Iron Age ritual building at Uppåkra, southern Sweden. Antiquity 81:11–25. Lowe, C.E. 1998. Coastal Erosion the Archaeological Assessment of an Eroding Shoreline at St. Boniface Church, Papa Westray, Orkney. Historic Scotland/Sutton Publishing, Stroud, UK. Magnus, B. 2002. Dwellings and settlements: Structures and characteristics Pp. 5–32, In J. Jesch (Ed). The Scandinavians: From the Vendel Period to the Tenth Century. Boydell Press, Woodbridge, UK. Marwick, H. 1952. Orkney Farms- Names. Kirkwal,UK. Marwick, H. 1970. The Place Names of Birsay. Aberdeen University Press, Aberdeen, UK. McOmish, D. 1996. East Chisenbury: Ritual and rubbish at the British Bronze Age-Iron Age transition. Antiquity 70:68–76. Mook, R., and R. Bertelsen. 2007. The possible advantage of living in turf houses on settlement mounds. Acta Borealia 24. Moore, H., and G. Wilson. 2003. Langskaill DSR on 2002 assessment and 2003 excavation at Langskaill, Westray, Orkney. Unpublished Data Structure Report. Orkney Sites and Monuments Record. Morris, C.D. (Ed.). 1985. Skaill, Sandwick, Orkney; Preliminary investigations of a mound-site near Skara Brae. Glasgow Archaeological Journal 12:82–92. Morris, C.D. 1989.The Birsay Bay Project. Coastal sites beside the Brough Road, Birsay, Orkney: Excavations 1976–1982, Volume 1. University of Durham, Department of Archaeology, Monograph Series 1, Durham, UK. Morris, C.D. 1996. The Birsay Bay Project. Sites in Birsay Village and on the Brough of Birsay, Orkney: Excavations 1976–1982, Volume 2. University of Durham, Department of Archaeology, Monograph Series 2, Durham, UK. Munch, G.S., O.S. Johansen, and E. Roesdahl. 2003. Borg in Lofoten: A Chieftain’s Farm in North Norway. Tapir Academic Press, Trondheim, Norway. Myhre, B. 2000. The Early Viking Age in Norway. Acta Archaeologica 71:35–47. North Atlantic Biocultural Organisation (NABO). 2012. Rousay, Orkney: NABO field school. Available online at http://www.nabohome.org/fieldschools/rousay. html. Accessed February 2013. Orkney Sites and Monuments Record (OSMR). 2013. OSMR website. Available online at http://www. library.uhi.ac.uk/smr/orksmr.php.Accessed February 2013. 147 J. Harrison 2013 Journal of the North Atlantic Special Volume 4 Owen, O.A. 1993. Tuquoy, Westray, Orkney. Pp. 318–339, In C. Batey, J. Jesch, and C.D. Morris (Eds.). The Viking Age in Caithness, Orkney, and the North Atlantic. Edinburgh University Press, Edinburgh, UK. Owen, O. (Ed.). 2005. The World of Orkneyinga Saga: “The Broad-Cloth Viking Trip”. The Orcadian Ltd, Kirkwall, UK. Owen, O., and M. Dalland. 1999. Scar. A Viking Boat Burial on Sanday, Orkney. Historic Scotland and Tuckwell Press, East Linton, UK. Palsson, H., and P. Edwards. 1978. Orkneyinga Saga. Penguin Books, London, UK. Parker Pearson, M. 1999. The Archaeology of Death and Burial. Sutton Publishing Ltd, Stroud, UK. Parker Pearson, M. (Ed.) 2012. From Machair to Mountains: Archaeological Survey and Excavation in South Uist. Oxbow, Oxford, UK. Pollard, J. 2001. The aesthetics of depositional practice. World Archaeology 33:315–333. Poulsen, B., and S.M. Sindbæk. 2011a. Settlement and Lordship in Viking and Early Medieval Scandinavia. Turnhout, Brepols, Belgium. Poulsen, B., and S.M. Sindbæk. 2011b. Introduction. Pp. 1–28, In B. Poulsen and S.M. Sindbæk (Eds). Settlement and Lordship in Viking and Early Medieval Scandinavia. Turnhout, Brepols, Belgium. Price, N. 2008. Dying and the dead: Viking Age mortuary behaviour. Pp. 257–273, In D. Skre (Ed). The Viking World. Routledge, London, UK. Price, N. 2010. Passing into poetry: Viking-Age mortuary drama and the origins of Norse mythology. Medieval Archaeology 54:123–155. Randsborg, K. 2008. Kings’ jelling. Acta Archaeologica Supplementa 79:1–28. Ritchie, A. 1977. Excavations of Pictish and Viking-age farmsteads at Buckquoy, Orkney. Proceedings of the Society of Antiquaries of Scotland 108:174–227. Ritchie, A. 1993. Viking Orkney. Batsford/Historic Scotland, London, UK. Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS). 2013. Canmore. Available online at http://canmore.rcahms.gov.uk/. Accessed February 2013. Rygh, O. 1898. Norske Gaardnavne: Forord og Indledning. Christiania, Norway. Sanmark, A., and S. Semple. 2010. Topography of outdoor assembly in Europe with reference to recent field results from Sweden. In H. Lewis and S. Semple (Eds.). Perspectives in Landscape Archaeology: Papers Presented at Oxford 2003–5. Bar Intl. Series 2103, Archeopress, Oxford, UK. Sharples, N. 2005. A Norse Farmstead in the Outer Hebrides: Excavations at Mound Three, Bornais, South Uist. Oxbow Books, Oxford, UK. Sharples, N. (Ed.). 2012. An Iron Age Farmstead in the Outer Hebrides. Excavations at Mound 1, Bornais, South Uist. Oxbow Books, Oxford, UK. Skre, D. (Ed.). 2007. Kaupang in Skiringssal. Kaupang Excavation Project Publication Series, Vol. I. Aarhus University Press, Denmark. Skre, D. 2008. The Viking World. Routledge, London, UK. Skre, D. 2011. Centrality, landholding, and trade in Scandinavia ca. AD 700–900. Pp. 197–212, In B. Poulsen and S.M. Sindbæk (Eds.). Settlement and Lordship in Viking and Early Medieval Scandinavia. Brepols,Turnhout, Belgium. Steedman, K.A. 1980. The archaeology of the Deerness peninsula, Orkney. Unpublished research, Orkney Sites and Monuments Record. Storli, I. 2006. Hålogaland Førrikssamlingen. Politiske Prosesser i Perioden 200–900 e. K. Novus, Oslo, Norway. Storli, I. 2010. Court sites of arctic Norway: Remains of Thing sites and representations of political consolidation processes in the northern Germanic world during the first millennium AD? Norwegian Archaeology Review 43(2):128–144. Stummann Hansen, S. 2003. Scandinavian building customs of the Viking Age: the North Atlantic perspective. Pp. 241–254, In S. Lewis-Simpson (Ed.) Vínland Revisited: The Norse World at the Turn of the First Millennium. Historic Sites Associations of Newfoundland and Labrador, St. John’s, NL, Canada. Thäte, E.S. 2007. Monuments and Mind: Monument Reuse in Scandinavia. Acta Archaeologica, Lundensia, Sweden. Thomas, J. 1999. Time, Culture, and Identity: An Interpretive Archaeology. Routledge, London, UK. Thomas, J. (Ed.). 2007. Place and Memory: Excavations at the Pict’s Knowe, Holywood, and Holm Farm, Dumfries and Galloway, 1994–1998. Oxbow, Oxford, UK Thomas, J. 2012. Introduction. Pp. 1–12, In H. Anderson- Whymark and J. Thomas (Eds.). Regional Perspectives on Neolithic Pit Deposition: Beyond the Mundane. Oxbow Books, Oxford, UK. Thomson, W.P.L. 1995. Orkney farm names: A re-assessment of their chronology. Pp. 42–63, In B.E. Crawford (Ed.). Scandinavian Settlement in Northern Britain. Leicester University Press, Leicester, UK. Thomson, W. 2008b. Orkney: Land and People. The Orcadian Ltd/Kirkwall Press, Kirkwall, UK. Urbańczyk, P. 1992. Medieval Arctic Norway. Semper, Warszawa, Poland. Vésteinsson, O. 2010. On farm mounds. Archaeologica Islandica 8:13–39.