Interpretations of the character of Norse settlement in the Northern and Western Isles embrace the full spectrum of possible relationships between the Norse and the native Picts – from wholesale genocide to peaceful assimilation. Modern genetic evidence is consistent with large folk migration to the Northern Isles, and smaller scale settlement in the Western Isles, but both genetics and place names lack chronological resolution. The Hebrides may have been repopulated by Celtic peoples during the Middle Ages and the high proportion of Scandinavian ancestry in Orkney and Shetland may relate to the long period of close political, economic, and social ties with Norway, maybe commencing before the Viking Age. On balance, the archaeological evidence implies large-scale migration, followed by Norse political, linguistic, and cultural domination, but with some coexistence of indigenous and immigrant identities, expressed differently in each area.
The Hebrides were linked with the Isle of Man as a single kingdom under the Lordship of the Isles, and shared a Norse inheritance, including traces of a ship levy and clinker-built vessels. It has been argued that the linguistic evidence suggests Gaelic was a lateroverlay on an almost entirely Old Norse-speaking population, although the place names are a mixture of Gaelic and Old Norse.
Much of the archaeological evidence was discovered so long ago that its value is limited. The picture has been dominated by graves, often chance discoveries as a result of erosion of sand dunes, such as Machrins on Colonsay and Ballinby on Islay. We do know the dead were dressed in full Scandinavian costume, and were well equipped. At Kiloran Bay (Colonsay), and Carn a’Bharraich (Oronsay), they were placed in rowing boats. The cemetery of a small community has been excavated more recently at Kneep on the Isle of Lewis. It includes men, women, and children, some buried with grave-goods, and some without. Isotope analysis of the teeth of a middle-aged woman buried in the late 10th century and dressed in traditional Norwegian folk costume reveals that she had been brought up in western Scotland. As a second-generation settler it is significant that, in death, her relatives were still keen to dress her in Scandinavian costume. As on the Isle of Man it appears that the Hebridean Norse settlers were negotiating their cultural identity through an emphasis on Scandinavian dress and custom. Unlike the Isle of Man, it seems that this strategy embraced their wives as well.
Norse settlements have been identified within grass-covered mounds on the sand dunes of the machair. Recent excavations on North and South Uist tell a similar story of takeover of native sites, with limited cultural continuity. At the Udal, on North Uist, characteristic Norse longhouses were built amongst the ruins of Pictish farms, on five settlement mounds, although the first structure built by the new occupants was a defensive enclosure, on the highest point of the site. New Norse styles of pottery, metalwork and combs appear, and the introduction of ceramic platters for baking barley cakes indicates a change in cuisine as well.
On South Uist settlement mounds are found throughout the machair and indicate that although farms again kept the same location from prehistoric times, new artefacts and new buildings appear during the Norse period. At Cille Phaedir a 10th-century timber hall represents a radical departure from local building tradition, accompanied by the characteristic consumption of barley cakes. At Bornish, a long-lived Pictish settlement was levelled during the creation of a 10th-century Norse farmstead, and the builders appear to have feasted before they laid the floor. By the late 11th century they were living in a classic Scandinavian-style bow-sided hall, 20 metres long, with substantial stone footings, robbed from an adjacent Pictish structure. The inhabitants of Bornish were more affluent – they were importing pottery from Wessex, and making antler combs. They were also maybe displaying mixed cultural messages. Niall Sharples has suggested that a Norwegian bone mount was deliberately buried because of its personal association with the owner, while houses were rebuilt overlapping earlier structures to demonstrate an ancestral link with the past.
The Life of Findan, a 9th-century continental source, is the most important contemporary document regarding the settlement of Orkney. It provides an apparently historical account of an Irish aristocrat’s escape from Norse slave traders on Orkney and his subsequent stay with a bishop, generally assumed to have been a Pict; the incident is dated c.840. The accepted origin myth, however, is to be found in the Orkneyinga Saga, written c.1192–1206. It claims that Orkney was settled by Earl Rognvald, fleeing from Norway in the 9th century, and it paints a vivid picture of life in Norse Orkney. Certainly from the mid-9th to the 12th centuries Orkney was the political focus of a semi-independent Norse state, whose ambit extended into Caithness. It did not become part of Scotland until 1468 when it was given to James III as part of the dowry of Margaret of Denmark.
A few place names and carved Pictish symbol stones and settlements indicate an indigenous pre-Norse Pictish population, but Scandinavian names obliterated all but a handful of the indigenous names. They extend to the smallest farmstead and every landscape feature and suggest little interaction.
Excavation of a number of sites around the Bay of Birsay indicates Norse takeover of an embryonic native power centre on the tidal promontory. According to the Orkneyinga Saga Earl Thorfinn ‘had his permanent residence at Birsay, where he built and dedicated to Christ a fine minster, the seat of the first Bishop of Orkney’. In the second half of the 9th century, Pictish buildings are overlain by rectangular structures ascribed to the Norse. A number of substantial Norse halls were constructed, with wall benches and box beds. Iron-working took place on the island, and silver was also melted down. The community cannot have been self-sufficient, however, and joints of beef and mutton were brought from farms in the bay, such as that at Buckquoy. Here successive generations of cellular buildings in distinctive Pictish figure-of-eight form were replaced by rectangular halls, although Pictish artefacts continued in use, indicating some continuity in population. At Pool, on Sanday, there is a clearly identifiable period of overlap between the two cultural groups. Pictish buildings were adopted and reused by inhabitants who had access to Norse material, particularly soapstone.
12. Brough of Birsay: aerial view
Altogether c.130 Scandinavian-style burials have been recognized on Orkney, a number consistent with significant population migration, and reflecting a more broadly-based settlement than in the Hebrides. All can be placed within 850–950, although most were discovered in the 19th century and records are poor.
The cemetery at Westness on Rousay has been excavated relatively recently. It consisted of 32 inhumations, but only eight graves – four male and four female – were accompanied. One grave was of a mother in her twenties, buried with her baby. She had been dressed wearing a pair of oval Norse brooches and a string of beads, but also wore a remarkable silver and gold penannular Celtic brooch. This wealthy woman may have been the head of the founding family. Two male warriors were each buried in small rowing boats. At least one had died in battle as the broken tips of four arrows were lodged in his back. The rest were buried without grave-goods in stone-lined cists, and these probably represent the graves of the pre-Norse population. Their position was marked on the surface by boulders, and none had been disturbed by the Norse. On the same promontory there was a boat house and a farmstead comprising a large hall furnished with low wall benches either side of the hearth, and two byres: one for cattle, and one for sheep.
Another boat burial was found in 1991, eroding out of a low cliff at Scar, on Sanday. The boat was a small rowing vessel which – from the presence of igneous rock particles in the caulking – must have been built in Norway, and brought to Orkney by a larger vessel. A woman in her seventies had been given pride of place in the centre. Her grave-goods included a maplewood box, an iron cooking spit or weaving batten, and a fine whalebone plaque, of north Norwegian type. She was also accompanied by a child, aged about 10, and a male in his thirties who had been squashed into one end of the boat; indeed his foot may have been broken in order to force him into the space. Nonetheless, he was no slave for he carried a comb, a sword, and a quiver of arrows, as well as 22 whalebone gaming pieces. Although the objects would place the burial in the late 9th century, radiocarbon dating suggests a date closer to the mid-10th century, indicating that many of the objects were heirlooms and that this was a group maintaining an old cultural and religious identity in the face of growing Christianization of the Norse colonists.
13. Westness, Rousay. Male boat grave with the prow and stern packed with stones to create a central chamber for the body
At St Magnus on Birsay, and at Deerness, there were private stone chapels as early as the 10th century, and by the 11th century considerable resources were invested in church building throughout Orkney. At Newark Bay a rectangular stone building underlay the later church. It was surrounded by c.250 burials which respected it apart from two burials inserted into the floor. Two 10th-century coins were found beneath the stone floor and 20 radiocarbon dates from the burials are consistent with mid-10th-century usage of the cemetery. It has been suggested that the occurrence of Papa place names on Orkney reflects the survival of Pictish Christian enclaves that were responsible for the early conversion of the Norse. It seems clear, however, that by 900 the Picts had been eclipsed – politically, linguistically, culturally, and socially. Their aristocracy had been displaced by Norse war leaders and only at the lower levels of society can survival of a native element in the population be perceived.
Shetland appears fleetingly in the Orkneyinga Saga, but it is generally assumed that it was colonized at the same time as Orkney, and enjoyed close connections with Norway. Hitherto there has been little archaeological investigation of Norse sites, and our knowledge has relied upon Jarlshof where, in 1934, Alexander Curle excavated the first Norse farmstead to be identified in the British Isles. Although it was described as a township, Jarlshof was never more than a farmstead, rebuilt many times, with the Norselonghouses erected adjacent to a Pictish broch. At Old Scatness Norse artefacts have also been found within an abandoned broch, suggesting that here and at Jarlshof they may have been targeting Pictish estate centres. The Norse finds coincide with economic changes, including the introduction of flax cultivation (possibly for linen for use in fishing nets) and more intensive exploitation of marine resources, although barley and oats were still grown.
Fieldwork by Stefan Stumann Hansen on the small island of Unst has sought to map fragments of a Norse landscape. The island is now largely depopulated, with a present population of c.600 inhabitants. However, the remains of Norse buildings have been recorded at 30 locations, and several have been excavated, including Hamar, Sandwick, and Soterberg. These sites are difficult to date. There is a small number of accompanied burials on Unst, such as Clibberswick, with 9th-century finds. The settlements may start in that period, but many are long-lived and continue in use until the 14th and 15th centuries.
Most comprise typical Norse longhouses with wall benches and central hearths; soapstone vessels, spindle whorls, and net sinkers are common finds. As Shetland was already treeless, the timbers for the main structural beams must have been imported. At Underhoull a bow-sided hall was constructed with footings from stone recycled from an Iron Age broch and other Pictish structures. A paved floor and drain at one end suggested it had been used as a cattle byre. The extravagant use of timber for the construction of Scandinavian-style longhouses can hardly be regarded as particularly functional. Pictish architecture tended to make use of stone rather than timber. The longhouse must have been regarded as a statement of cultural identity of almost symbolic importance to the settlers, allowing them to ‘feel at home’ and to express a sense of community in opposition to other native communities and identities.
Associated economic changes support the idea of a Norse takeover in Orkney and Shetland. James Barrett has shown that whereas in the pre-Norse period fishing was largely a littoral activity, there was now large-scale exploitation, and an increased dietary importance of marine protein. Such fundamental changes in subsistence activity are more likely to result from large-scale colonization than from the influence of a small immigrant elite. This is particularly likely in the case of fish consumption, which was probably relatively low-status. Unlike changes in language or costume there would have been little motivation for the local population to adopt new economic practices unless this was necessitated by demographic change. This intensification of fishing must therefore have been related to the distinct cultural practices of a large immigrant population, and if not genocide, must still imply considerable population replacement.
Nonetheless, although Norse cultural traditions have retained a political importance in both the Northern and Western Isles, it would be a mistake to see them as a straight Scandinavian import. The incoming populations already demonstrate cultural mixing, and elements of indigenous culture were also retained and adapted, at least in the lower social strata. To examine the colonization of virgin territories without indigenous inhabitants we must travel further across the North Atlantic.