2014 Viking Settlers ofJournal of the the North North Atlantic Atlantic: An Isotopic Approach Special Volume 7 2014 Journal of the S.V.North Arge Atlantic Special Volume 7:1–17 Viking Faroes: Settlement, Paleoeconomy, and Chronology Símun V. Arge* Abstract - The paper presents a synopsis of the current evidence for the settlement chronology and Viking Age to Early Me- dieval paleoeconomy of the Faroe Islands. Special emphasis is placed on the recent interdisciplinary research carried out in the village of Sandur, on the island of Sandoy, as part of the Heart of the Atlantic project. A particularly important outcome of this recent work has been the wide application of scientific dating methods to the early settlement remains. Recent AMS radiocarbon dates push the earliest settlement of the islands further back in time than traditionally thought, results that are of great importance because the Faroes were the first stepping-stone for the Viking diaspora west across the North Atlantic. Introduction the oldest phases at the site, those dating to the Vi- The Faroe Islands are a group of some 18 islands king period, rather than the later Medieval deposits. located in the North Atlantic almost midway be- The excavation method employed did not follow tween Norway, Iceland, and Scotland. The islands, stratigraphic sequence, and consequently Medieval separated by narrow fjords and sounds, together artifacts were mixed with those from the Viking pe- have an area of ~1400 km2. When the first Viking riod during recovery. settlers arrived, they encountered a landscape Research on settlement-development histories in characterized by grasses, sedges, and ericaceous the Faroes has revealed a settlement pattern charac- shrubs. Woodlands—small groups of juniper and terized by a long continuity. In general, the locations birch—seem to have been of minor importance. In of settlements in existence today appear to be the other words, the landscape was rather similar to same places settled during the Viking landnám (first what we see today (Bradshaw et al. 2010, Lawson settlement) period. Therefore settlements at the et al. 2005). The rugged topography of the islands so-called heimrust—the proper settlement area in restricted the settlements mainly to the coastal strips the village—in some instances have been traced ar- along the sounds and the fjords. chaeologically from the present back into the Viking Whether these settlers came directly from the Period, and the Viking farm in Kvívík is an excellent east—from a Norwegian homeland—or from the example of one such site (Arge 2005). south—via northern Scotland and Ireland, as indi- Where conditions have been favorable, this cated by archaeological and recent genetic evidence continuity of settlement has led to the creation of (Als et al. 2006)—they brought with them a Norse farm-mounds, a phenomenon also known from other or Hiberno-Norse culture, which was subsequently regions of the Norse and Viking world (Arge 2005). adapted to local conditions in the North Atlantic. In the decades since the excavations in Kvívík were carried out, other Viking settlement remains have been mapped and investigated around the Faroe Settlement and Architecture Islands, both within the infields and in the outfields The first proper archaeological excavation in as well (Arge et al. 2005). I turn now to a brief sum- the Faroes took place in 1941 at the site of “Niðri á mary of some of these. Toft”, located in the village of Kvívík on the island of Streymoy (Fig. 1). Through the excavation of this Overview of Recent Archaeological Research site, which was carried out over several seasons in the 1940s and 1950s, archaeological remains from Toftanes the islands’ early history were brought to light for It was not until the excavation of the site of “Á the very first time. This site has since been regarded Toftanesi” in the village of Leirvík, during the period as the classic example of a Faroese Viking farm of 1982–1987, that a Viking Age farm was investigat- (Dahl 1951). However, more recent re-analysis of ed which provided a clearer picture of the layout and the excavation has modified this interpretation to function of a Viking Age settlement compared to the some extent (Matras 2005). site at “Niðri á Toft” in Kvívík (Fig. 2; Hansen 2013). Methodologically, I believe most of the attention The farm consisted of four buildings. The dwell- in the excavations at “Niðri á Toft” was focused on ing, a longhouse, measured some 20 m in length, * Søvn Landsins - Føroya Fornminnissavn, Faroe Islands National Heritage. Archives, Library and Museums, Kúrdalsvegur 15, PO Box 1155, FO-110 Tórshavn; svarge@savn.fo. 1 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Figure 1. The Viking farm at “Niðri á Toft” in Kvívík. Photograph © S.V. Arge. Figure 2. The Viking farm at Toftanes, Leirvík, seen from north. Photograph © S.S. Hansen. 2 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge with an internal width of 5 m. The curved walls from the British Isles (Hansen 2013). The site’s ar- were 1 m thick and were made of outer and inner tifact assemblage also includes imported goods and dry-stone walling infilled with turf to provide a jewellery originating from the Irish Sea region, most more windproof structure. The middle of the west- notably two ringed pins of Hiberno-Norse type and ern half of the building had a fireplace almost 5 m a jet bracelet (Hansen 1993, 2013). The settlement long, while the eastern end may have contained a at Toftanes, which is the best-known example of a byre. Added to the southern wall of the longhouse Viking Period farm in the Faroes, has been dated to was a small structure with a floor space of ~12 m². the 9th–10th centuries (Vickers et al. 2005, Edwards The wall of the western gable was probably wooden. and Schofield 2013)). Contrary to the following On the northern side of the longhouse was another case, Toftanes represents an example of a permanent building, measuring 13 m long and 4 m wide. Its settlement located within the infields. walls were constructed of only a single course of dry-stone walling without turf, and its function has Argisbrekka preliminary been interpreted as an outhouse, perhaps From early on, philologists paid special attention for use in craft and/or food-production activities. A to the existence of place-names containing the Celt- small building, 5 m long and 3 m wide, was located ic name element “ærgi” or “argi”, which may have close to the northern side of the longhouse. The side entered the Norse language during the 9th century walls of this structure were constructed in the same (Matras 1933, 1956). It was assumed that this place- manner as those of the dwelling; while the west ga- name referred to something like shielings, small ble wall was probably wooden, the eastern end was dwellings located in the highlands summer grazing missing, due to erosion by a stream. The floor was pastures. Archaeological surveys of the 18 localities paved with flat stones and covered with thick layers retaining such place names in the Faroes—all but of ashes and charcoal, and there was a small stone- one situated in the outfields—found small ruins at built ember pit in the eastern end. This structure has several of these localities. Archaeological excava- been interpreted as a firehouse (Hansen 1991, 2013). tion at one of these sites, Ergidalur, in the outfield As is the case with other farmsteads known from of the village of Hovi on the island of Suðuroy, this period, the buildings at Toftanes were primarily revealed that it was a Viking-period summer or sea- constructed of wood, though with thick insulating sonal settlement (Dahl 1970:362). outer walls of stone and earth. Even though building Prompted by the planned damming of Lake timber was scarce in the Faroes—or was at least Eiðisvatn in the northernmost part of the island of not a widely available local natural resource—the Eysturoy, extensive archaeological excavations Norwegian/Scandinavian tradition of wooden were conducted at the site Argisbrekka from 1983 stave-building was employed, albeit with modifi- to 1987 (Mahler 1991, 1998, 2007). The site is lo- cations to suit local natural conditions—it was part cated in the outfield of the present village of Eiði, of a widespread tradition throughout the so-called at an altitude of 130 m above sea level (Fig. 3). Viking World (Hansen 1999, Stoklund 1984, Thor- The archaeological features can be grouped into steinsson 1982). roughly two settlement areas, with the remains The artifacts recovered at Toftanes are of par- of 8 and 13 buildings unearthed in the two areas, ticular interest. These include schist querns and a respectively. Within these two areas, there were large number of steatite objects, such as fragments two to three places containing lesser constructions, of bowls and saucepans, spindle whorls and line- or each consisting of a residential house and one or net-sinkers for fishing, and whetstones. As steatite is two outhouses, interpreted as working and storage not locally available, these objects must have been structures. All were constructed with walls of turf, imported, presumably from Norway (Hansen 2013). sand, clay, and pebbles and all were relatively Only a few artifacts made of local stone (e.g., basalt small, measuring around 7 to 8 m long and 3 to 4 and tufa) were found. Also recovered were very m wide or smaller. Stratigraphic evidence indicates large numbers of well-preserved wooden objects, that two of the shielings were in operation simulta- such as bowls, spoons, and barrel staves. A large neously, during the area’s last active period in the amount of the wooden objects consisted of cords Viking Age (Mahler 2007). Functionally, the Faro- of twined juniper branches, several meters long, ese ærgir are reminiscent of the Norwegian fulltime which no doubt were used as handles for barrels and shielings. as ropes for roof stones (Larsen 1991). While most Stratigraphic and radiocarbon evidence indicate of the stone artifacts may have been imported from a commencement of activities in the eastern area Norway, the quernstone material seems to originate of the site sometime during the 9th century, with a 3 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Figure 3. The Viking shieling at Argisbrekka, Eysturoy. Map showing the primary investigation area. The shaded house structures show the latest structural remains in each of the two settlement areas. From Mahler (2007). 4 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge cessation of all activity in this area sometime dur- enactment for the Faroes that dealt primarily with ing the middle of the 11th century (Mahler 2007). the legislation of sheep farming. It therefore seems Interestingly, with the exception of the presence of safe to conclude that a combination of a greater em- local ceramics, the Argisbrekka assemblage does phasis on fishing and extensive sheep farming led to not differ significantly from that recovered at the the disappearance of the ærgir as part of an outdated contemporary primary (non-shieling) farm site at agricultural practice in the Faroes (Mahler 2007). Toftanes. The local ceramics are a unique category of artifact in Faroese archaeology, documented from “Við Kirkjugarð” in Sandur, Sandoy the late 10th into the 19th centuries (Arge 1991, 1997; The village of Sandur is one of the largest and Mahler 2008). wealthiest agricultural communities in the Faroes, We must conclude that in the Viking period the and has likely been a prominent community since ærgi and the permanent farm were closely linked el- the first farming settlers arrived in the Viking period ements of the domestic economy, traces of which are (Fig. 4; Arge 2001). Until the first half of the 20th widely distributed throughout the Faroes. The oper- century, the local church site was rather isolated ational method was, however, adjusted to local con- from the rest of the village (Krogh 1983), a situation ditions, as was the case with regard to sheep-milking which is uncommon in the Faroes. However, ever methods (Thorsteinsson 1977). Shielings are not since the discovery of an 11th-century coin hoard mentioned in the Sheep Letter of 1298, a special (the only one of its kind yet found in the islands) in Figure 4. Sandur. The church site is central in the photo. The investigated site Junkarinsfløttur is located in the eroding cliff edge along the green fields between the church site and the beach. Photograph © S.V. Arge. 5 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge the churchyard in 1863, there had long been specula- orated with an interlaced motif of Irish origin, and tion that the site would yield further archaeological some small silver fragments. Additionally, one of the remains (Steen Jensen 2004). graves contained a clipped Cufic (Arabic) coin, the When the first archaeological excavation in San- only such coin found in the Faroes thus far; this find dur took place in 1969/1970, efforts were focused has been identified as a late 9th-century imitation of within the actual church (Krogh 1975, 1983). The an Abbasid-style dirhem, and it may be suggestive of results were quite interesting and rather unexpected, burial in Sandur during the mid-10th century (Black- with the remains of five successive churches discov- burn 2005, Graham-Campbell 2005). Compared to ered below the present one, which was built in 1839. Viking-period burials in both Scotland and Iceland, The oldest, a small single-aisled stave church, was it is remarkable that 3 of the 7 graves contained sil- dated to the 11th century. Shortly after this initial ver and/or silver-ornamented artifacts, including the excavation, plans to extend the churchyard led to silver coin fragment. These burials clearly indicate the archaeological excavation of test trenches in the presence of a high-status settlement at the site. the ~3000-m2 extension area, revealing extensive In contrast to sites such as Toftanes, the archaeo- Viking-period remains. Since then, Føroya Fornmin- logical remains uncovered in the Sandur church- nissavn (the Faroe National Museum) has periodi- yard—the site of “Við Kirkjugarð”—reflect a variety cally undertaken archaeological excavations in the of features and structures which, together with the churchyard to allow for the release of additional waste and midden deposits (including ash, slag, burial plots. In 1972, a small-scale excavation in the and large deposits of fire-cracked stones), appear old churchyard revealed the eastern gable and well- to document activities more characteristic of an in- preserved stone-paved floor of a structure. These dustrial zone than domestic space (Arge 2001). For remains were interpreted as part of a boat-shaped example, the best preserved building, a 5 m x 5 m longhouse, potentially the dwelling of a farm. It structure excavated in 2008–2009, contained thick is very likely the coin hoard was placed below the layers of burnt material, including large pieces of paved floor of this building by the end of the 11th charcoal and significant amounts of charred barley, century, based on the description of the hoard’s dis- and was clearly associated with some kind of barley covery, indicating that the building was in use during processing. AMS radiocarbon assays of 2 charred the 11th century at least (Krogh 1975, 1983). barley samples produced very early dates (Arge and In 1989, 11 graves were uncovered in the Friel 2009; Arge et al., forthcoming), which I shall churchyard, 7 of which were excavated (Arge and return to later in this paper. Hartmann 1992). Prior to this discovery, the only known Viking-period burial site in the Faroes was “Á Sondum” in Sandur, Sandoy at “Yviri í Trøð” in the village of Tjørnuvík. Ex- On the northern shore of the bay of Sandsvágur cavated in 1956, the site has been dated to the 10th lies the farm site of “Á Sondum”, an excellent exam- century, based on both archaeologically and radio- ple of a settlement-period farm mound. Referenced carbon evidence (Arge 2001:11, Dahl and Rasmus- in written records in 1412, “Á Sondum” is the only sen 1956). The graveyard at Sandur appears to be farm site in the Faroes mentioned by name in a me- well-regulated, with a series of burials placed end- dieval document (Thorsteinsson 1979). The site was to-end in a number of more-or-less parallel rows. first archaeologically assessed in 1994 with a small- All of the graves are aligned almost east–west, and scale excavation of the eroding cliff face, conducted all of the excavated skeletons lay with their skulls as part of a Nordic amateur archaeologist camp and pointing west. Though the state of preservation of overseen by myself and a Danish colleague (Jen- the osseous material was very poor, dental enamel sen 1995). In 2006–2007, the Heart of the Atlantic was found preserved in all of the excavated graves. project provided an opportunity to re-investigate the The artifacts recovered from these graves can be site. These excavations revealed a 3-m-deep multi- classified as personal belongings, and include, for period sequence of midden and structural remains, example, silver and bronze finger rings, bone beads, from 19th-century midden deposits at the top of the glass and amber objects, and iron knives. Two of the eroding section to a large, truncated, Norse recti- individuals had each been buried with an iron knife, linear structure—a Viking longhouse—towards the the handle of which was entwined with thin silver base (Church et al. 2013). The longhouse featured a threads; one of these individuals, a young man, also 2-m-wide external wall, a large central hearth, floor had a pouch, possibly woven, containing a leather surfaces, and internal features, including a barrel pit. purse with 7 plain lead weights. Other notable finds Four radiocarbon dates on carbonized barley grains from this grave included a bronze strap-end orna- from the central hearth and an external midden mented with an animal head, a bronze fragment dec- dated the longhouse occupation to the initial Viking 6 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Figure 5. “Á Sondum,” Sandur. Remains of the southern wall of a Viking longhouse at the bottom of the eroding cliff. Photograph © S.V. Arge. 7 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge colonization of the 9th century (Fig. 5; Church et al. of stratified midden deposits preserved in the sandy 2013). matrix (Arge et al. 2010). Similar (though far less well-preserved) midden deposits were recovered at the site of “Á Sondum” across the bay, and these Environmental Archaeology and Viking Faroese two sites must be characterized as Viking- to Me- Paleoeconomy dieval-period settlement mounds (Arge 2001, Arge “Undir Junkarinsfløtti” in Sandur, Sandoy et al. 2010, Church et al. 2005). The reports on the In 1989, a phosphate survey of a long-cultivated archaeofauna from these two sites are the first sub- field north of the church in Sandur suggested the stantial published zooarchaeology from the Faroes presence of a large settlement area (Arge 2001). In (Brewington 2006, 2010, 2011). The excavated de- the summer of 2000, erosion of the cliff face just posits at “Undir Junkarinsfløtti” have been divided below these fields exposed deep cultural deposits. into five phases dating from the 9th century to the This discovery spurred the launch of an interdisci- Late- or Post-Medieval period. The exceptional plinary, international collaboration at the site—the preservation conditions at “Undir Junkarinsfløtti” Heart of the Atlantic project—beginning in 2003 have produced a massive archaeofaunal assemblage; as part of the Leverhulme Trust-funded Landscapes to date, nearly 60,000 bone and shell fragments have Circum-Landnám project. This work has been a been analyzed. The preliminary analysis of materi- North Atlantic Biocultural Organization (NABO) al presents a diverse range of economic practices cooperative effort, involving the Faroe National Mu- employed by the Norse settlers at a key time and seum as lead institution, together with institutions geographical position in their expansion across the such as the City University of New York, University North Atlantic. Characteristic of the archaeofauna is of Bradford, Durham University, and University of the large proportions of wild resources, particularly Stirling (Arge et al. 2010). seabirds, fish, and molluscs (Fig. 6). There is a rel- Excavations at the site of “Undir Junkarinsfløtti” atively small percentage of domestic mammals in were focused not only on the multiple phases of the total number of specimens identifiable to species structural remains represented but also on the ~3 m level in all three phases. Thus, the economic strategy Figure 6. “Undir Junkarinsfløtti,” Sandur. Major Archaeofauna. The phase UJF 0 represents the earliest, ephemeral deposits; UJF 1 is dated to 9th–12th centuries cal AD; UJF 2 to 11th–12th centuries cal AD, and UJF 3 to 11th–13th centuries cal AD. From Brewington (2011). 8 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge at the site appears to have relied heavily upon the have had a stable diet from the landnám up into the exploitation of a broad spectrum of the local wild Middle Ages, in contrast to Greenland where the resources to supplement a mixed agricultural base of diet shifted from a relatively terrestrial to a predomi- animal husbandry and cereal cultivation. nantly marine base (Arneborg et al. 2008). Domestic mammals recovered included sheep, Paleobotanical remains were also retrieved from cows, and pigs, with single bones of goat and dog at “Undir Junkarinsfløtti” and at “Á Sondum,” pro- (Fig. 7). Significant numbers of pig bones were viding evidence of plant use at the sites. A major recovered throughout the sequence, indicating sus- consideration for the Norse in the Faroes would tained pig keeping up to and beyond the 13th century, have been the procurement of wood; the islands a situation unique compared to Iceland and Green- never sustained extensive woodland, and heather land (Brewington 2011). Pig-herding is not known and juniper were the only wood resource available in post-Reformation traditional Faroese agriculture at settlement (Church et al. 2005). However, frag- (Arge et al. 2009). Seabirds also comprised a large ments of various coniferous species would have proportion of the archaeofauna. The settlers at the arrived as driftwood and could have been collected site of “Undir Junkarinsfløtti” remained dependent from the shore. Wood charcoal was very rare in upon seabirds, especially puffins, far longer and to the archaeological assemblage in Sandur, and con- a much greater degree than any of the other Viking sisted of locally derived roundwood, coniferous Age settlers of the North Atlantic islands (Brewing- driftwood, and imported oak (Church et al. 2005). ton and McGovern 2008). A wide range of marine Rather, peat and turf were the main fuel sources in resources, particularly Atlantic cod, were also re- the treeless Faroese landscape. A hulled six-row covered, further emphasizing the key role of wild barley monoculture appears to have been in place, resources in the Faroese paleoeconomy. with small-scale yet intensive cultivation undertak- Recent isotopic analyses of Faroese human and en. Cereal cultivation seems to have played a lesser animal bones indicate that 20–40% of the dietary role in the economy than in other areas of the east- protein was of marine origin, a level somewhat ern North Atlantic, however, and some of the barley higher than that for contemporary Icelanders but less may have been imported (Church et al. 2005). than that of the Norse Greenlanders (Arneborg et al. “Undir Junkarinsfløtti” represents a key site for 2008). Both the Icelanders and the Faroese seem to investigating the early Faroese paleoeconomy. To Figure 7. “Undir Junkarinsfløtti,” Sandur. Major Domesticates. From Brewington 2011. 9 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge understand this archaeological evidence, we must supported the traditionally accepted archaeological assume that this site is part of an extensive settle- chronology for the Viking-period settlement of the ment area, which includes the previously mentioned islands. One quite obvious reason for these differing church site, “Við Kirkjugarð”. The archaeological outcomes between the two camps has been the ab- record from the area leaves us with an impression of sence of research aimed at the sampling of paleobo- a high-status Faroese community with strong links tanic remains from archaeological contexts. Howev- to the outside world. er, recent years have seen a renewed emphasis on the dating of human-induced landscape and vegetational change (e.g., Hannon 1999). Most recently, the topic Settlement Chronology has been addressed as part of the aforementioned A recurring issue in the debate regarding the Landscapes Circum-Landám project, leading to a earliest settlement (landnám) of the Faroes has been more complete picture of the vegetational history of the related questions of when did the first settlers the islands (Lawson et al. 2005). arrive and where did they come from. This was It has gradually become more common to incor- also the subject of my master’s thesis in medieval porate a range of archaeological science methods archaeology in 1986 (Arge 1990). At that time, I had into research projects in the Faroes. The Heart of to conclude that, based on the available evidence, it the Atlantic project in Sandur exemplifies this new was not possible to trace the oldest settlement of the approach, with 36 AMS radiocarbon dates having Faroes any further back than about the year 1000, been obtained from the deeply-stratified (~3 m) or in broad terms to the Late Viking/Early Medie- sites of “Undir Junkarinsfløtti” and “Á Sondum” val period (Arge 1991:114, 1991:fig. 5). However, (Church et al. 2005). At “Undir Junkarinsfløtti,” referring to the lack of sufficient analyses of the 24 radiocarbon assays were taken in an effort to archaeological data at the time, I suggested that establish an absolute chronology for the site. Five one or two of the sites treated in the thesis—“Niðri phases identified in the midden deposits: the earli- á Toft” in Kvívík, for example—might be dated est was dated to the 9th–12th centuries A.D. and the somewhat earlier. Even so, evidence for earlier (pre- latest, representing the amended soil and topsoil, to 1000) occupation was lacking, despite the presence the late- and post-Medieval periods. Additionally, of (limited) radiocarbon data (Arge 1991:table 1). previously sampled material from the site at Tof- Until recently, radiocarbon dating in the Faroes has tanes has now been re-dated under the Landscapes been used to support traditional typological dating. Circum-Landám project, and the results from these Despite the fact that a few of these radiocarbon dates three sites show contemporary early-Viking period turned out to be a bit “too old”, most were consistent activities (Church et al. 2005, Edwards and Borth- with the established archaeological chronology. wick 2010b, Vickers et al. 2005). This approach has begun to change in recent One early archaeological investigation that has years, however, with several interesting studies been particularly intriguing for quite some time was making more prominent use of radiocarbon dating carried out in 1952 at the site of “Undir Keisarafløt- in their analyses. There have historically been two ti,” in the village of Norðragøta, Eysturoy. Structural different “camps” of researchers using radiocarbon remains had been exposed on the beach at the site dating to investigate the timing of the earliest settle- by severe storms during the 1940s (Arge 1990:80) ment of the Faroes: the archaeological one, using ra- (Fig. 8). The small-scale excavation in 1952 re- diocarbon to supplement traditional archaeological vealed stone-lined drains and the remains of a stone dating methods, and the paleobotanical one, using wall that had been erected on a turf layer. This turf radiocarbon to date the onset of human impacts on layer was situated some 60 cm below the surface at vegetation. Interestingly enough, these two camps high tide, and geological analysis suggested that the have been very divided over the years, largely fail- land had sunk about 2 m since the formation of the ing to communicate effectively with each other. For turf layer. The finds recovered from the site includ- instance, radiocarbon dates from the paleobotanical ed twined juniper ropes. Conventional radiocarbon investigations of so-called “ancient” or “Celtic” analysis of a sample of the ropes taken a number of fields indicated pre-Viking settlement activity in the years ago produced relatively early dates (A.D. 670– 6th–8th centuries (Jóhansen 1971, 1979). These dates 860 [K-6120], see Fig. 9;); though these pre-date are controversial, however, contrasting sharply with the traditionally-accepted settlement of the Faroes, the archaeological evidence (Edwards and Borth- they do fall in line with other early dates recently wick 2010a). Radiocarbon dates from archaeologi- obtained from other sites in the islands. The most cal investigations, on the other hand, have usually recently acquired of these early dates come from 10 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge the previously-mentioned excavations in the church early Viking settlement starting sometime between yard in Sandur, “Við Kirkjugarð” (Fig. 10; Arge and the 7th and 9th centuries (Ballin Smith 2007:294, Friel 2009). Two radiocarbon dates obtained from a 2013:229–230). semi-circular black/burnt pit feature within the 5 x I will end this paper by referring to yet another 5 m structure suggest that the material within the pit set of dates which resulted from the 2006–2007 ex- was deposited at some point between the end of the cavations at the site of “Á Sondum”. As mentioned 7th century and the end of the 9th century (see Fig. 9; earlier, the excavations investigated the remains of Arge et al., forthcoming). a Viking longhouse, one of the earliest-dated Vi- The employment of more intensive radiocarbon king longhouses yet investigated in the Faroes. This dating strategies in modern excavations in the Far- longhouse and associated external midden deposits oes has gradually pushed the dates for the initial were found to truncate an extensive wind-blown Viking settlement back in time, as clearly illustrat- sand deposit, which contained patches of burnt peat ed in Table 1. These recent data, together with the ash of anthropogenic origin. Carbonized barley results of past palynological and paleobotanical grains from two of these ash deposits provided four studies, force us to reconsider the traditional landám radiocarbon dates. These dates strongly suggest chronology for the Faroes. Furthermore, these dates two separate episodes of settlement at the site—one might be considered in light of some radiocarbon within the 4th–6th centuries and a second one within dates obtained in recent investigations at the ceme- the 6th–8th centuries AD. The majority of this set- tery at Norwick, Unst, as well as at the site of Old tlement was likely subsequently destroyed by the Scatness, Dunrossness, in Shetland, likely indicating construction of the later Viking longhouse (Fig. 11; Figure 8. “Undir Keisarafløtti,” Norðragøta. Investigation of Viking remains on the beach 1952. Photograph © J. Rasmussen. 11 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Church et al. 2013). These early dates pose new traditional fuel-procurement method as well as the questions regarding the early history of the Faroes. processing of barley clearly indicate human activi- Neither structures nor artifacts were associated ty at the site, including some form of settlement— with the early dated deposits. Nevertheless, the not necessarily permanent. A few pieces of charred Figure 9. Radiocarbon dates from recent excavations in the Faroes. 12 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Figure 10. The church site in Sandur, “Við Kirkjugarð.” In the foreground, the excavated early Viking structural remains in the extended churchyard; in the background, across the bay, the early settlement site of “Á Sondum”. Photograph © R. Friel. 13 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge barley do not tell us anything about the ethnicity Brewington at CUNY; as well as colleagues within the of those who left these traces, however, nor where Landscapes Circum-Landnám Project, especially Kevin these people came from; these are questions for fur- J. Edwards at Aberdeen University, Andy Dugmore at ther research. Despite a clear lack of existing cor- Edinburgh University, and Ian Simpson at Stirling Uni- versity, as well as Helgi D. Michelsen and Hákun An- roborating archaeological evidence in the Faroes, dreasen at the museum in Tórshavn. Thanks also to Mike we must be open to re-evaluating the archaeolog- Church for helping produce Figure 10. Further, I would ical record. Also necessary will be the retrieval of like to thank the funding bodies—the Faroese Research additional samples from other archaeological sites Foundation, the Anadarko Faroes Company, BP Amoco in the islands. Exploration Ltd. (Faroes), and US National Science The Faroes were the first stepping-stone for the Foundation—for much appreciated financial support Viking diaspora across the North Atlantic, a process over the years. that culminated in the first European discovery of continental North America in the 11th century A.D. Literature Cited The main settlement event in the Faroes has hitherto been believed to be the arrival of the Vikings in the Als, Th.D, T.H. Jørgensen, A.D. Børglum, P.A. Petersen, O. Morse, and A.G. Wang. 2006. Highly discrepant 9th century, as represented by extensive archaeo- proportions of female and male Scandinavian and Brit- logical evidence. Our new data radically alter the ish Isles ancestry within the isolated population of the timeline for this colonization! Faroe Islands. European Journal of Human Genetics 14:497–504. Arge, S.V. 1990. Landnamet på Færøerne. En diskussion og Acknowledgments vurdering af teorierne om, hvornår det fandt sted, med For fruitful collaboration and inspiring discussions, I særlig vægt på teorier baseret på arkæologisk materiale. would like to thank all my colleagues in the Heart of the Master’s Thesis in Medieval Archaeology, Aarhus Uni- Atlantic Project, especially Steve Dockrill, Julie Bond, versity, 1986. Published by Føroya Fornminnissavn and Robert Friel of the University at Bradford; Mike Church Middelalder-arkæologisk Nyhedsbrev, Tórshavn, Faroe at Durham University; and Tom McGovern and Seth Islands, and Højbjerg, Denmark. 163 pp. Figure 11. Investigation of the eroding cliff section at “Á Sondum”. Sandur, 2007. Photograph © S.V. Arge. 14 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Arge, S.V. 1991. The landnám in the Faroes. Arctic Anthro- Blackburn, M. 2005. Coinage and contacts in the North pology 28(2):101–120. Atlantic during the seventh to mid-tenth centuries. Arge, S.V. 1997. Í Uppistovubeitinum: Site and settle- Pp. 141–151, In A. Mortensen and S.V. Arge (Eds.). ment. Fróðskaparrit 45:27–44. Viking and Norse in the North Atlantic. Select Papers Arge, S.V. 2001. Forn búseting heima á Sandi. FRØÐI from the Proceedings of the Fourteenth Viking Con- 2:4–13. gress, Tórshavn, 19–30 July 2001. Annales Societatis Arge, S.V. 2005. Cultural landscapes and cultural en- Scientiarum Færoensis Supplementum XLIV. Føroya vironmental issues in the Faroes. Pp. 22–38, In A. Fróðskaparfelag, Tórshavn, Faroe Islands. 445 pp. Mortensen and S.V. Arge (Eds.). Viking and Norse in Bradshaw, R.H.W., G.E. Hannon, M. Rundgren and T. the North Atlantic. Select Papers from the Proceedings Giesecke. 2010. Which trees grew on the Faroe Islands of the Fourteenth Viking Congress, Tórshavn, 19–30 before people arrived? Pp. 36–49, In S.-A. Bengtsson, July 2001. Annales Societatis Scientiarum Færoensis P. Buckland, P.H. Enckell and A.M. Fosaa (Eds.). Supplementum XLIV. Føroya Fróðskaparfelag, Tór- Dorete—her book—being a tribute to Dorete Bloch shavn, Faroe Islands. 445 pp. Arge, S.V., and R. Friel. 2009. Fornfrøðiliga rannsóknin and to Faroese nature. Annales Societatis Scientiarum Við Kirkjugarð heima á Sandi. Kvørnin 12:60–62. Færoensis. Supplementum 52. Faroe University Press, Arge, S.V., and N. Hartmann. 1992. The bural site of við Tórshavn, Faroe Islands. 307 pp. Kirkjugarð in the village of Sandur, Sandoy. Fróðskap- Brewington, S. 2006. Interim report on archaeofauna arrit 38–39:5–21. from Undir Junkarinsfløtti, Sandoy, Faroe Islands. Arge, S.V., G. Sveinbjarnardóttir, K.J. Edwards, and P.C. NORSEC Zooarchaeology Laboratory Report No. 31. Buckland. 2005. Viking and medieval settlement in the CUNY. 14 December 2006. 25 pp. Faroes: People, place, and environment. Human Ecol- Brewington, S. 2010. Third interim report on analysis ogy 33(5):597–620. of archaeofauna from Undir Junkarinsfløtti, Sandoy, Arge, S.V., M. Church, and S. Brewington. 2009. Pigs in Faroe Islands. NORSEC Zooarchaeology Laboratory the Faroe Islands: An ancient facet of the island’s pal- Report No. 46. CUNY. 25 February 2010. 22 pp. eoeconomy. Journal of the North Atlantic 2:1–14. Brewington, S. 2011. Fourth interim report on analysis Arge, S.V., R. Edvardsson, and R. Friel. 2010. The of archaeofauna from Undir Junkarinsfløtti, Sandoy, Heart of the Atlantic Project: Excavations at Undir Faroe Islands. NORSEC Zooarchaeology Laboratories Junkarinsfløtti and á Klettum: Field Seasons 2004– Report No. 55. A product of the North Atlantic Bio- 2006. Interim Report Number 2 (Data Structure cultural Organization (NABO) Research Cooperative. Report). Compiled by R. Edvardsson and R. Friel. CUNY. 27 March 2011. 17 pp. Bradford Archaeological Sciences Research, Brewington, S., and T.H. McGovern. 2008. Plentiful University of Bradford, UK. 202 pp. puffins: Zooarchaeological evidence for early sea- Arge, S.V., R. Friel, J.E. Cussans, and A.R.R. Mustchin. bird exploitation in the Faroe Islands. Pp. 23–30, In Forthcoming. The Heart of the Atlantic Project: Exca- C. Paulsen and H.D. Michelsen (Eds.). Símunarbók. vations at Við Kirkjugarð. Field Seasons 2008–2009. Heiðursrit til Símun V. Arge á 60 ára degnum 5. Sep- Interim Report Number 1 (Data Structure Report). Compiled by R. Friel. Føroya Fornminnissavn, FO- tember 2008. Faroe University Press, Tórshavn, Faroe 100 Tórshavn, Føroyar. Division of Archaeological, Islands. 252 pp. Geographical and Environmental Sciences, University Church, M.J., S.V. Arge, S. Brewington, T.H. McGovern, of Bradford, UK. J.M. Woollett, S. Perdikaris, I.T. Lawson, G.T. Cook, Arneborg, J., N. Lynnerup, and J. Heinemeier. 2008. Far- C. Amundsen, R. Harrison, Y. Krivogorskaya, and oese Isotopes—a “half-baked” study. Pp. 15–22, In E.Dunbar. 2005. Puffins, pigs, cod, and barley: Pal- C. Paulsen and H.D. Michelsen (Eds.). Símunarbók. aeoeconomy at Undir Junkarinsfløtti, Sandoy, Faroe Heiðursrit til Símun V. Arge á 60 ára degnum 5. Sep- Islands. Environmental Archaeology 10:179–197. tember 2008. Faroe University Press, Tórshavn, Faroe Church, M.J., S.V. Arge, P.L. Ascough, J.M. Bond, G.T. Islands. 252 pp. Cook, S.J. Dockrill, A.J. Dugmore, K.J. Edwards, T.H. Ballin Smith, B. 2007. Norwick: Shetland’s first Viking McGovern, and I.A. Simpson. 2013. The Vikings were settlement? Pp.287–297, In B. Ballin Smith, S. Taylor, not the first colonizers of the Faroe Islands. Quater- and G. Williams (Eds.). West over Sea. Studies in nary Science Review 77(2013):228–232. Scandinavian Sea-Borne Expansion and Settlement Dahl, S. 1951. Fornar toftir í Kvívík. Varðin 29:65–96. Before 1300. A Festschrift in Honour of Dr Barbara Dahl, S. 1970. Um ærgistaðir og ærgitoftir. Fróðskaparrit Crawford. Koninklijke Brill NV, Leiden, The Nether- 18:361–368. lands. 581 pp. Dahl, S. and J. Rasmussen. 1956. Víkingaaldargrøv í Ballin Smith, B. 2013. Norwick: Shetland’s earliest Vi- Tjørnuvík. Fróðskaparrit 5:153–167. Tórshavn, Faroe king settlement? Pp. 217–233, In V.E. Turner, J.M. Islands. Bond and A.-C. Larsen (Eds.). Viking Unst. Excava- Edwards, K.J. 2005. “On the windy edge of nothing”: A tion and Survey in Northern Shetland 2006–2010. historical human ecology of the Faroe Islands. Human Shetland Heritage Publications, Lerwick, Shetland, Ecology 33(5):585–596. UK. 266 pp. 15 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Edwards, K.J., and D.B. Borthwick. 2010a. The pollen Jóhansen, J. 1979. Cereal cultivation in Mykines, Faroe content of so-called “ancient” field systems in Suðu- Islands A.D. 600. Danmarks Geologiske Undersø- roy, Faroe Islands, and the question of cereal cultiva- gelser Årbog 1978:93–103. tion. Pp. 96–116, In S.-A. Bengtsson, P. Buckland, Krogh, K.J. 1975. Seks kirkjur heima á Sandi. Mondul P.H. Enckell and A.M. Fosaa (Eds.). Dorete—her 2:21–54. book—being a tribute to Dorete Bloch and to Faroese Krogh, K.J. 1983. Gård og Kirke. Samhørighed mellem nature. Annales Societatis Scientiarum Færoensis. gård og kirke belyst gennem arkæologiske undersø- Supplementum 52. Faroe University Press, Tórshavn, gelser på Færøerne og i Grønland. Hikuin 9:231–244. Faroe Islands. 307 pp. Larsen, A.-C. 1991. Norsemen’s use of juniper in Viking Edwards, K.J., and D.B. Borthwick. 2010b. Peaceful wars Age Faroe Islands. Acta Archaeologica 61–1990:54– and scientific invaders: Irishmen, Vikings, and palyno- 59. logical evidence for the earliest settlement of the Faroe Lawson, I.T., M.J. Church, T.H. McGovern, S.V. Arge, Islands. Pp. 66–79, In J. Sheehan and D.Ó. Corráin J. Woollett, K.J. Edwards, F.J. Gathorne-Hardy, A.J. (Eds.). The Viking Age: Ireland and the West. Papers Dugmore, G. Cook, K.-A. Mairs, A.M. Thompson, from the Proceedings of the Fifteenth Viking Congress, and G. Sveinbjarnardóttir. 2005. Historical ecol- Cork, 18–27 August 2005. Four Courts Press, Dublin, ogy on Sandoy, Faroe Islands: Paleoenvironmental Ireland. 569 pp. and archaeological perspectives. Human Ecology Edwards, K.J., and E. Schofield. 2013. Radiocarbon dating 33(5):651–684. at Toftanes and the wider Faroese context. Pp. 177–185, Mahler, D.L. 1991. Argisbrekka: New evidence of shiel- In S.S. Hansen. Toftanes. A Viking Age Farmstead in ings in the Faroe Islands. Paper read at “The Norse the Faroe Islands. Acta Archaeological 84(1). of the North Atlantic Conference” April 17–21 1988. Graham-Campbell, J. 2005. The Viking Age gold and sil- Bowdoin College, Maine, USA. Acta Archaeologica ver of the North Atlantic region. Pp. 125–140, In A. 61–1990:60–72. Mortensen and S.V. Arge (Eds.). Viking and Norse in Mahler, D.L. 1998. The stratigraphical cultural landscape. the North Atlantic. Select Papers from the Proceedings Pp. 51–62, In Outland Use in Preindustrial Europe. H. of the Fourteenth Viking Congress, Tórshavn, 19–30 Andersson, L. Ersgård, and E. Svensson (Eds.). Lund July 2001. Annales Societatis Scientiarum Færoensis Studies in Medieval Archaeology 20. Institute of Ar- Supplementum XLIV. Føroya Fróðskaparfelag, Tór- chaeology, Lund University, Lund, Sweden. shavn, Faroe Islands. 445 pp. Mahler, D.L. 2007. Sæteren ved Argisbrekka. Økono- Hannon, G.E. 1999. The use of plant macrofossils and pol- miske forandringer på Færøerne i vikingetid og tidlig len in the palaeoecological reconstruction of vegetation. middelalder. Economic development during the Vi- Doctoral thesis, Swedish University of Agricultural king Age and Early Middle Ages on the Faroe Islands. Sciences, Alnarp 1999. Acta Universitatis Agriculturae Annales Societatis Scientiarum Færoensis Supple- Sueciae Silvestria 106. Alnarp, Sweden. mentum 47. Faroe University Press, Tórshavn, Faroe Hansen, S.S. 1991. Toftanes: A Faroese Viking farmstead Islands. 525 pp. from the 9–10th centuries AD. Acta Archaeologica Mahler, D.L. 2008. Towards a typology of Viking Age 61–1990:44–53. and Medieval Faroese ceramics. Pp. 131–139, In C. Hansen, S.S. 1993. Viking-Age Faroe Islands and their Paulsen and H.D. Michelsen (Eds.). Símunarbók. southern links in the light of recent finds at Toftanes, Heiðursrit til Símun V. Arge á 60 ára degnum 5. Sep- Leirvík. Pp. 473–486, In C.E. Batey, J. Jesch, and C.D. tember 2008. Faroe University Press, Tórshavn, Faroe Morris (Eds.). The Viking Age in Caithness, Orkney Islands. 252 pp. and the North Atlantic. Selected Papers from the Pro- Matras, C. 1933. Stednavne paa de færøske Norðuroyar. ceedings of the Eleventh Viking Congress, Thurso and Aarbøger for Nordisk Oldkyndighed og Historie 1932 Kirkwall, 22 August–1 September 1989. Edinburgh, 3.Rk. Gyldendahl, Copenhagen, Denmark. 322 pp. Scotland, UK. 554 pp. Matras, C. 1956. Gammelfærøsk ærgi. Namn och Bygd Hansen, S.S. 1999. I Jan Pedersens fodspor på Oma. Nyt 44(1–4):51–67. om Vikingetidens Gårdsanlæg. Pp. 253–272, In I.T. Fu- Matras, A.K. 2005. The Viking settlement “Niðri á Toft”, glestvedt, T. Gansum, and A. Opedal (Eds.). Et hus med Kvívík, Faroe Islands: A reanalysis. Pp. 99–108, In A. mange rum. Vennebok til Bjørn Myhre på 60-årsdagen. Mortensen and S.V. Arge (Eds.). Viking and Norse in Bind A. AmS-Rapport 11A. Arkeologisk museum, the North Atlantic. Select Papers from the Proceedings Stavanger, Norway. of the Fourteenth Viking Congress, Tórshavn, 19–30 Hansen, S.S. 2013. Toftanes: A Viking Age farmstead in July 2001. Annales Societatis Scientiarum Færoensis the Faroes Islands. Acta Archaeologica 84(1). Supplementum XLIV. Føroya Fróðskaparfelag, Tór- Jensen, N.M. 1995. Udgravningsberetning for am- shavn, Faroe Islands. 445 pp. atørarkæologisk undersøgelse á Sondum, Sandoy, Steen Jensen, J. 2004. Møntskatten fra Sand, Færøerne. SNR:6194, 1995. Føroya Fornminnissavn (Unpub- Nordisk Numismatisk Årsskrift 1997–99:65–93. lished report), Tórshavn, Faroe Islands. Stoklund, B. 1984. Building traditions in the Northern Jóhansen, J. 1971. A paleobotanical study indicating a World. Pp. 96–223, In A. Fenton and H. Pálsson pre-Viking settlement in Tjørnuvík, Faroe Islands. (Eds.). The Northern and Western Isles in the Viking Fróðskaparrit 19:147–157. World. Edinburgh, Scotland, UK. 721 pp. 16 2014 Journal of the North Atlantic Special Volume 7 S.V. Arge Thorsteinsson, A. 1977. Heimildir um seyðamjólking í Føroyum. Fróðskaparrit 25:84–94. Thorsetinsson, A. 1979. Ruddstaðir í Brekkum—ein mu- turgøla frá 1412. Mondul 1:14–21. Thorsteinsson, A. 1982. Færøske huskonstruktioner fra vikingetid til 1800-årene. Pp. 149–161, In B. Myhre, B. Stoklund, and P. Gjærder (Eds.). Vestnordisk byggeskikk gjennom to tusen år. Tradisjon og foran- dring fra romertid til det 19. århundrede. West Nordic Building Customs from the Roman Period to the 19th Century. AmS-skrifter 7. Arkeologisk museum, Stavanger, Norway. 287 pp. Vickers, K., J. Bending, P. Buckland, K.J. Edwards, S.S. Hansen, and G. Cook. 2005. Toftanes: The paleoeco- logy of a Faroese landnám farm. Human Ecology 33(5):685–710. 17