A. Sanmark 96 Journal of the North Atlantic Special Volume 5 96 Introduction The locations and characteristics of the Norse assembly (thing) sites in Shetland have hitherto remained largely unexplored, which is hardly surprising considering the scarcity of source materials. Shetlandic assemblies are not mentioned in the Orkneyinga Saga or any other Norse source, and the bulk of the evidence consists of a few references and place-names in medieval Shetland documents. According to these sources, there was a lawthing site at Tingwall and a number of so-called thing parishes, i.e., parishes with names containing the ON element þing, which most likely were administrative districts with their own thing site (Fig. 1). In this article, an integrated approach has been taken where all the documentary sources have been reviewed together with archaeological evidence and topographical information in order to identify the most likely locations of the thing sites. All sites have been visited and recorded with a GPS and photography, and also mapped in ArcGIS using additional data from Canmore, the on-line database of The Royal Commission on the Ancient and Historical Monuments of Scotland (http://canmore.rcahms.gov.uk). On the basis of this data, the characteristics of the proposed thing sites are examined in the light of comparative evidence from other areas of Norse settlement. For reasons of space, this article will not address the validity of the interpretation of the “stone ring” on Papa Stour and the mound at Setter farm in Delting as thing sites (Crawford 1984:48–50 [HU16SE 24], O’Grady 2008:210–11 [HU36SE 1]).1 It has previously been pointed out that thing sites and thing organizations are known from most areas of Norse settlement, although the nature of these systems varied slightly between regions (Sanmark 2010a:182). Shetland provides an interesting case study as it fits into the overall pattern, although with some unusual traits. The island group formed part of the Norwegian colonies, skattlands, which before 1266 also included Orkney, the Faros, and the Western Isles. Norwegian power gradually increased over time, although it is not clear exactly when Norway took control over Shetland after the first period of Norse settlement in the 9th and 10th centuries. What is known, however, is that after 1195 and the Shetlanders’ participation in the uprising against the King of Norway, Shetland was removed from the Orkney Earldom and instead placed directly under the Norwegian kingdom. This state of affairs continued until 1469 when the islands became part of Scotland (Smith 2011:104). At the end of this paper, the question of whether royal presence is necessary or not for the creation of assembly sites will be examined. The Shetland thing sites—those known and unknown today—are unlikely to have been established at the same time or for the same reasons. As evidenced in Scandinavia, some assemblies were short-lived and/or temporary, while others formed stable meeting points over several hundred years (Sanmark 2009; Sanmark and Semple 2008, 2010; Semple and Sanmark 2013). There is no evidence of when the first thing sites were established in Shetland. Evidence from the Scottish Mainland, however, suggests that this may well have happened rather early in the settlement process. In this area, there is evidence of at least three or four thing sites (O’Grady 2008:195–201, 211–17) despite the fact that Norse rule lasted for a much shorter period than in Shetland, and the maximum period of use Patterns of Assembly: Norse Thing Sites in Shetland Alexandra Sanmark* Abstract - The assembly (thing) sites in Shetland have hitherto not been systematically examined, and their locations are more or less unknown. The aim of this article is therefore to identify the locations of the assemblies in the so-called thing parishes and analyze their characteristics, using comparative evidence from other areas of Norse settlement. As part of this process, it is proposed that Rauðarþing, one of two “lost” parishes, was located on the island of Yell, rather than on the Shetland Mainland as previously argued. Close examination of the proposed thing locations has revealed a number of striking features, most of which have parallels in Scandinavia. This finding demonstrates that great care went into the selection of thing sites, although with some consideration for local conditions. On the basis of the strong site characteristics, a new potential thing site has been identified in the area of Benston in Nesting on the Shetland Mainland. Finally, it is argued that the first thing sites were established by early Norse settlers in the time before the Norwegian kings had established firm rule in Shetland. Debating the Thing in the North I: The Assembly Project Journal of the North Atlantic *Centre for Nordic Studies, University of the Highlands and Islands, Kirkwall, Orkney KW15 1QX; alexandra.sanmark@ uhi.ac.uk. 2013 Special Volume 5:96–110 A. Sanmark 97 Journal of the North Atlantic Special Volume 5 97 Figure 1. The Shetland thing parishes and thing sites. © Crown Copyright/database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. A. Sanmark 98 Journal of the North Atlantic Special Volume 5 98 for the sites is therefore between the 9th and the 11th centuries. Altogether, this means that we can visualize a rather organic phase in Shetland, from the 9th century onwards, with assemblies set up by powerful individuals/families as a way of taking or claiming control over a particular area. The Shetland lawthing is referred to in documents from 1307 and 1577 (Ballantyne and Smith 1999:2,183–224), and its existence is further supported by place-name evidence and late traditions. Orkney too had its own lawthing, first recorded in 1496 (Ballantyne and Smith 1999:30 [1510], 32 [1516]; Clouston 1914:74, no. xxxii). A lawthing was a representative assembly where royal law was introduced and enforced and it constituted the highest instance in the medieval Norwegian court system. Below the lawthings were regional and local things, the nature of which varied slightly over time and between areas (Sanmark 2006:47–48, with references). The skattlands deviated slightly from Norway proper, where there was one main lawthing per legal district, i.e., one in the Gulathing, the Frostathing, the Borgarthing, and the Eidsivathing. It is argued that the lawthing was first introduced into the Gulathing in the tenth century by King Hákon the Good (Helle 2001:30–34). Lawthings were generally successors of althings, general assemblies for all free landholders, which are known from several areas in Scandinavia, such as Gotland (gutnalþing) and Gamla Uppsala, while the Icelanders established their own althing around 930 (Friðriksson 1994:105, Helle 2001:30–34, Hollander 1964:315, Myrberg 2008, Peel 2009:xxvii). In the Faros, this transition is implied in medieval documents. The “Sheep letter” (seyðabrævið ) from 1298 states that “the althing” had in 1273 received a decree from the Norwegian king, Magnus the Lawmender, to use his new law, while a letter from 1350 shows that a lawthing was now in existence (Poulsen and Zachariasen 1971, Smith 2009:38–39, Sølvará 2002:42). The reign of Magnus was a turning point in Norwegian judicial history, as his law for the first time placed the legal system fully in the hands of the king (Helle 2001:155–156, Sanmark 2006). As evidenced in the Faros, Magnus was clearly concerned with implementing his changes, and it has been argued that this was pursued also in Shetland, both by Magnus and his son Hákon Magnusson (Smith 2009:42–44, 2011). The Shetland lawthing may therefore have been introduced in the late 13th century, replacing an older althing, as seems to have been the case in the Faros (cf. Smith 2009:38–39). It is conceivable, however, that this change took place earlier in Shetland as the Norwegian kings monitored these islands more closely after the events of 1195. The Shetland Thing Sites The location of the thing site at Tingwall is the most easily pinpointed (Fig. 2). Tingwall is recorded as the Shetland lawthing in a document from 1307, and a number of judicial actions were announced and witnessed here in 1506, 1510, 1525, 1532, 1577, and 1578 (Ballantyne and Smith 1999:2–3,25–28,30– 31,34–35,183–224,239). According to the 1577 document, meetings were held at Ting Holm, now a small promontory in Tingwall Loch (Ballantyne and Smith 1999:196). In 1701, John Brand reported that on the Ting Holm there were “three or four great Stones” on “which the Judge, Clerk, and other Officers of the Court did sit”, and a similar statement was made a century later (Brand 1883:183, Edmonston 1809:130).2 The other thing participants are said to have been sitting in the open field, below Tingwall Church, overlooking the Holm (Brand 1883:183, Bruce 1908:18, Smith 2009:39–41). The use of Tingwall as a thing site is further strengthened by the place name Grista, 750 m northwest of Law Ting Holm. This name is derived from Griðastaðir, from ON griðr = sanctuary, Figure 2. Law Ting Holm, Tingwall. © Crown Copyright/ database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. A. Sanmark 99 Journal of the North Atlantic Special Volume 5 99 asylum and ON staðr = farm, abode (Cleasby 1874: 215, 586; Jakobsen 1936:100; Stewart 1987:235). The connection between thing and griðr is clearly shown in Snorri’s Edda, where the famous episode of the killing of Baldr takes place at a thing site. The Æsir were, however, unable to take their revenge as the thing was “a place of such sanctuary” (“svo mikill griðastaðir”) (Faulkes 1987:49, Pálsson 2003:71, Riisøy 2013 [this volume]). According to early Norwegian law, griðr concerned three places: in church, at a thing, and at a feast. In these places specific regulations applied, which meant that stricter punishments were enforced than if the same crime had been committed elsewhere (Larson 1935:281, The Law of the Frostathing IV:58).3 Tingwall traditions state that criminals would receive sanctuary at Grista if they managed to run through the crowd assembled around Law Ting Holm (Stewart 1987:255).4 A similar combination of names occurs by Thingwall on the Wirral in Cheshire, England. On the boundary between Thingwall and the neighboring farm Storeton, the field-name le Gremotehalland was recorded in 1330, where the first component could be either OE *grið- (ge)mōt or ON griða-mót (Dodgson 1972:256, Paton 2011:20). As mentioned above, Shetland documents provide clues to the local thing organization via the medieval parish names (Stewart 1987:300). Seven such parishes occur in the source material: • Sandsting, recorded in 1500 (Stewart 1987:300). • Aithsting, recorded in 1507 (Stewart 1987:300, Jakobsen 1936:125). • Nesting, recorded in 1490 (Stewart 1987:300, Jakobsen 1936:125). • Lunnasting, recorded in 1490 (Stewart 1987:300, Jakobsen 1936:125). • Delting, recorded in 1490 (Stewart 1987:300, Jakobsen 1936:125) • Rauðarþing, recorded in the 14th century (Gunnes and Kjellberg 1979:no 151). • Þvæitaþing, recorded in the 14th century (Gunnes and Kjellberg 1979:nos 130, 150) The locations of these parishes, apart from Rauðarþing and Þvæitaþing, are known, as they are still in existence, although their medieval boundaries have not been ascertained (Fig. 1). The two “lost” parishes were, according to two 14th-century documents, “in the diocese of Orkney”, which consisted of both Orkney and Shetland. One of the documents explicitly states that Þvæitaþing was in Shetland, and it is therefore assumed that Rauðarþing was too (Ballantyne and Smith 1999:3; Gunnes and Kjellberg 1979:nos 130,150,151; Smith 2009:41–42), a suggestion which is further supported by the evidence presented below. The names of the other Shetland parishes do not contain the ON element þing and cannot therefore in themselves provide any information on the assembly organization. Although various different types of thing districts existed in the Viking homelands, parishes were not one of these, and Shetland is the only known example across the Viking world of a merged parish and thing organization. Traces also remain of three other types of Norse administrative units: herað, fjórðungar (quarters), and áttungar (eighths). Herað units are implied by the “Herra” place names, all of which refer to districts or groups of farms (Stewart 1987:130),5 while a document of 1490 mentions “Vogafiordwngh’” and “Mawedes otting ”, interpreted as “the Waas quarter” and “the eighth at Mavis Grind” (Ballantyne and Smith 1999:22, Indrebø 1936:253, Smith 2009:43). These units are all known from the Scandinavian countries, although their form and function varied strongly between regions. The situation described in the earliest Gulathing law may be the most accurate description of the situation in Shetland. According to this law, the fylki (province) was divided into quarters, eighths, and heraðs. According to the early Gulathing law, only the quarters had things belonging to them, while herað things are not mentioned until in the later Law of Magnus the Lawmender (Bauge 1980–1982:379–380, Fladby 1980–1982:276–278, Helle 2001:76–81). The evidence thus suggests that Shetland may have been viewed as a fylki and divided accordingly (Indrebø 1936:253), with quarter things. The Norwegian eighths were connected to the naval levy fleet (ON leiðung), and their presence in Shetland may indicate that it was attempted to introduce this system in Shetland too. Seeing the continuously changing administrative patterns in Scandinavia, these three different types of districts were not necessarily contemporary or uniformly enforced. The thing parishes may represent a yet later phase of reorganization, carried out by Magnus the Lawmender and Hákon Magnusson (Smith 2009:42–44, 2011).6 I will return to this question below. As the first element in the parish names are placenames, these are often used as starting points in the search for the thing sites (Smith 2009:41, Stewart 1987:300). This idea is derived from the one site that can be located with certainty, that of Tingwall in Tingwall parish, and also from Scandinavian naming traditions.7 On this basis, approximate locations of the thing sites for each parish can be suggested, in the area around the recorded place-names (Fig. 1). For Sandsting, Delting, Lunnasting, and Aithsting, this is rather straight forward, but less so for Nesting, Þvæitaþing, and Rauðarþing. Sandsting is named after the farm of Sand, recorded in 1355 (Ballantyne and Smith 1999:5, A. Sanmark 100 Journal of the North Atlantic Special Volume 5 100 Jakobsen 1936:125, Stewart 1987:300). The exact location of the thing is difficult to pinpoint, but as large parts of this area are rather low-lying and boggy, the assembly meetings must have been held at least as high up as Innersand or modern Foraness (recorded in 1869–1899; Fig. 3, Stewart 1987:216). For Delting, there is more detailed information, albeit of late date. The parish name is derived from Dale, recorded in 1507 (Stewart 1987:72). According to local tradition, the thing site was “on a patch of rising ground at the side of the burn of Sandgarth, and immediately above the waddel [ford] … at the burn mouth” (Greig 1892:82). This description fits in well with the geography around Dale farm, and the assembly was most likely situated on Gullaness, a small peninsula sticking out into the tidal zone of the voe (Fig. 4). The situation for Lunnasting is slightly different as the parish name is not directly derived from Lunna farm (recorded in 1507), but from hlunnr-eið (Stewart 1987:80, 300). On this basis, a good approximation of the thing-site location can be put forward.8 ON eið denotes an isthmus or portage, and ON hlunnr refers to the wood rollers used for pulling boats across land (Stewart 1987:80, Waugh 2010:545). Just below the 17th-century Lun- Figure 3. The area around Sand, Sandsting. © Crown Copyright/database right 2011. An Ordnance Survey/ EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. Figure 4. Gullaness by Dale in Delting. © Crown Copyright/ database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. na House is a narrow isthmus, recorded as a portage site in modern times (McCullough 2000:183–185, Ritchie 1997:94–95). This area must be the location of the thing site, which in turn has given name to the parish (Fig. 5). Finally, Aithsting, named after Æiði farm (derived from ON eið) and recorded already in 1299 (Ballantyne and Smith 1999:1; Jakobsen 1936:125; Stewart 1987:80, 300). Aith is today a rather built-up area, and remains of the thing site are difficult to trace. One possible location for the assembly, easily accessible by the water’s edge, is Gotda Taing, a small peninsula shown on the 1878 OS map (http://maps.nls.uk/os/6inch/; Fig. 6).9 For Nesting, the task of locating the thing site is slightly more complex, as the parish name does not refer to a known place-name. There are, however, two late documents that mention a tax-paying district called Neipnating in Nesting (Ballantyne and Smith 1999:App. 1 [ca. 1510], Smith 2009:41).10 Neipnating is derived from ON *Gnípnaþing, i.e., the thing at the steep hills/peak (Smith 2009:41, footnote 43; Zoëga 1910:168). Neap farm is situated just above a rather rocky and steep-sided peninsula where, according to a late 19th-century source, “The spot where the court was held at Neep is still to be seen, A. Sanmark 101 Journal of the North Atlantic Special Volume 5 101 Figure 6. The area around Aith, Aithsting. © Crown Copyright/ database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. Figure 7. The area around Neap, Nesting. © Crown Copyright/ database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. with the bench before it, and was called Neepsnating” (Edmondston 1809:132). This “stone bench” may well be in the northern part of the peninsula where a number of stones are seemingly exposed, forming a semi-circle (Fig. 7; Smith 2009:41).11 The nearby Outrastone Bay is an interesting placename, as a possible interpretation is ýtrast í þing, i.e., “outermost in the (area of the) thing”, which would make sense geographically (P. Gammeltoft, Department of Scandinavian Research, University of Copenhagen, pers. comm.). It is not clear, however, if the name specifically refers to a thing site at this location, or whether it refers to a thing district. More importantly, it is not clear how Neipnating relates to Nesting. Brian Smith has suggested that as Neap lies on a promontory (ness), this is what has given its name to the parish and that Neap therefore was the assembly site for Nesting (Smith 2009:41). Another, perhaps more likely suggestion, is that Gnípnaþing was an older assembly district, perhaps abolished by reorganization, but which survived as a tax-paying unit. This issue will be revisited below. The task of locating the thing sites of Þvæitaþing and Rauðarþing is equally complex. It has been argued that Þvæitaþing was located on the west side of the Shetland Mainland where all place-names containing twatt (ON *þveit = clearing) appear (Ja- Figure 5. The area around Lunna, Lunnasting. © Crown Copyright/database right 2011. An Ordnance Survey/ EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. A. Sanmark 102 Journal of the North Atlantic Special Volume 5 102 name for Mid Yell Voe and must have given name to the parish (Fig. 9; Ballantyne and Smith 1999:2; Cant 1975:15, 1987:87, 86).12 This parish name has not survived, but courts were clearly held here at least in the 16th century.13 In Mid Yell, there is a place called Gardiestaing, although there is no visible “taing” in the area.14 Taking the evidence about the court in Giærd/Gerde into account, this name is most likely derived from Gardiesting.15 After close examination of the suggested locations of all the Shetland thing sites, a number of striking features and characteristics have been found, which will now be discussed in turn. 1. Isthmus and small island location There are several examples of thing sites located at isthmuses (Fig. 1), and the names of two of these, Aith in Aithsting and the assembly in Lunnasting, contain ON eið. Aith has been seen as slightly problematic as an isthmus since this name refers to a 4-km wide area, where the ground rises significantly. It has, however, been convincingly argued that “when eið forms all or part of the place-name which describes the location … the isthmus in question is part of the Norse economic landscape” (Waugh 2010:545). kobsen 1936:110–111,126; Sandnes and Stemshaug 1976:324–325; Smith 2009:42; Stewart 1987:300, 302–303). Jakobsen suggested that Þvæitaþing included both Walls and Aithsting, as today’s Twatt is found in modern Aithsting (Jakobsen 1936:126). Seeing that Aithsting formed its own district, this seems unlikely. Bearing this in mind and also that all but one of the twatt names are found within an area along a 3-km stretch close to Walls, this seems to be where the thing site was most likely located (Fig. 8). The location of Rauðarþing has so far proven more elusive. The one suggested location concerns the northern part of Northmavine, which may previously have been called *Rø, as its northernmost area is now named North Roe (Jakobsen 1936:126, Smith 2009:42). Detailed study of the evidence, however, suggests that Rauðarþing instead encompassed parts of the island of Yell. A document from 1538 refers to a “court” held “in Giærd in Rødefiord parish in Jelle [Yell]”, while a 1586 document mentions a court held in “Gerde in Redeførd parish” (Ballantyne and Smith 1994:38–39, 1999:40). Ræyðar fyrðe (now Reafirth), recorded in 1307, seems to be the old Figure 8. The suggested location of Þvæitaþing and the general area of its local assembly. © Crown Copyright/database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. Figure 9. Gardiestaing in Rauðarþing. © Crown Copyright/ database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. A. Sanmark 103 Journal of the North Atlantic Special Volume 5 103 Eið is thus seen to refer to isthmuses that were used as portages of goods or people and probably also stretches of land which served as portages, but which cannot be defined as isthmuses. This description fits Aith well, and it may be of significance that the inlet at the other end of the suggested portage route was called Effirth (ON Eiðfjörðr) (Heggstad et al. 1975:86, Waugh 2010:545–546; cf. McCullough 2000:201–204, Nymoen 1995:34–35, Stewart 1987:89).16 An isthmus and portage location may be the case for Gardiestaing too. This site is located by the sea, at Yell’s narrowest point. Today the walk across land between the two inlets is less than 1 km. There is no eið place-name recorded here, but this has been identified as a possible portage for smaller boats (McCullough 2000:187–190). Finally, Sand in Sandsting was also located on a rather narrow strip of land, where the surrounding waters linked into Effirth. The only thing site seemingly located on a small island was Law Ting Holm. This island setting is suggested by the ON name (holmr = a small island), but was presumably true until Tingwall Loch was drained in the 1850s (Smith 2009:41). Today there is a causeway, 40 m long and almost 2 m wide, consisting of a double row of boulders, linking the Holm to the shore (Fig. 10). A causeway consisting of boulders appears on a painting by Sir Henry Dryden, dated ca. 1855. The use and function of a causeway for the thing meetings was moreover described in 1701 by John Brand: “All the Country concerned to be there, stood at some distance from the Holm on the side of the Loch, and when any of their Causes was to be Judged or Determined, or the judge found it necessary that any person should compear before him, he was called upon by the Officer, and went in by these steping stones, who when heard, returned the same way he came” (Brand 1883:183–184, my emphasis). In 2011 The Assembly Project (TAP) excavated parts of the causeway and demonstrated that it is lying on top of pottery from the 18th or 19th century and is therefore of rather recent date (Coolen and Mehler 2011:26–28). This chronology means that John Brand’s description predates the current structure, although the reference to “steping stones” fits the boulder construction rather well. There are several possible scenarios that would reconcile the recent dating of the existing causeway and Brand’s earlier Figure 10. View of the causeway leading from Law Ting Holm towards Tingwall church. Photograph © Natascha Mehler, University of Vienna. A. Sanmark 104 Journal of the North Atlantic Special Volume 5 104 frequent practice in Scandinavia, and when excavated, they tend to be poor in Norse finds and features (Coolen and Mehler 2011; Sanmark 2009; Sanmark and Semple 2008, 2010). Another possible example of monument reuse is the isthmus of the Lunnasting assembly, where there are archaeological remains from multiple periods (Fig. 5). The most striking feature is the probable broch mound, known as Chapel Knowe, with possible remains of a medieval church on top (MacKie 2002:116, Ritchie 1997:94–95 [HU46NE 4]). Sites with large mounds were frequently used for assemblies in Scandinavia and also in Norse Scotland (O’Grady 2008:195–201, Semple and Sanmark 2013). To the southwest of Chapel Knowe are a few moundlike features, which could be Viking-Age burials, another Scandinavian thing-site feature (Fig. 5; Sanmark 2009; Sanmark and Semple 2008, 2010). As the other Shetland thing sites cannot be pinpointed, possible monument reuse cannot be determined. It is important to note, however, that the archaeological remains around these sites lack major visible features. A site that is strikingly different from Law Ting Holm and Lunnasting is Gullaness in Delting. Here a geophysical survey carried out by TAP did not identify any archaeological features, and the only recorded site in close vicinity is a chapel by Dale farm, 250 m from the peninsula (HU46NW 2). 4. Sheltered location Several of the Shetland thing sites are located in relatively sheltered areas, often at the bottom of inlets, as seen at Aith, Dale, Gardie, and, to some extent, Sand (Fig. 1). This finding is not surprising, considering the frequency of strong winds, and is rather similar to the location of the two suggested thing sites in Greenland (Sanmark 2010a), but constitutes an interesting example of adaptation to local conditions. A site that breaks this pattern is Neap. Outdoor meetings would have been problematic at times, as the suggested site is at the tip of the peninsula and therefore very exposed to wind. This open exposure is admittedly the case with many places in Shetland, but in comparison to the other suggested thing sites, this location is much more windswept. One suggestion is that the potential Gnípnaþing meetings may have been held in more sheltered places around the modern farm of Neap. 5. Proximity to churches and chapels One of the most striking features of the Shetland thing sites is the proximity to churches and chapels. Possible chapel remains have been identified at Gardiestaing, Sand, Lunna, Dale, and Aith. As these description. The first one is that one or more earlier causeways have existed, of which all traces have now been removed, at least in the trench excavated by TAP. A second possibility is that temporary causeways have existed, made of wooden planks and/or stones that were laid out for the meetings. A third possibility, at least for the earliest phase of the assembly, is that participants waded to reach the island. Thing sites in Scandinavia are frequently located by fording places, and the entering of a thing site may have involved the ritual crossing of “holy waters”, as suggested by the eddaic poem Grimnir's Sayings (Grímnismál; Sanmark 2009:231–232). One suggestion is therefore that such a crossing was required for those taking part in the thing meetings during the (early) Norse period. As the site remained in use until the 16th century, and belief systems and traditions changed, the appeal of this manner of accessing the island may have faded, and a causeway may have been constructed. This possibility would explain the mention of “these steping stones” from 1701. It may of course also be a combination of these suggestions, and the causeway may have been rebuilt on several occasions. One possible point in time for such a (re)construction is the 18th or 19th century, when interest in old assembly sites was reawakened (S. Semple, Durham University, Durham, UK, pers. comm.). 2. Proximity to fresh water Since the location of so few Shetland thing sites can be pinpointed, the exact relationship between them and freshwater streams cannot be quantified. However, as streams are present around all the sites discussed above and as this is the one feature found by all assemblies studied in Scandinavia and Iceland too, it is worth taking into consideration. Streams were naturally needed for practical purposes, but may well have had the same function as the “holy water” discussed above (cf. Sanmark 2009, 2010b). 3. Monument reuse The clearest case of monument reuse is Tingwall. The excavations carried out in by TAP revealed that Law Ting Holm was the site of an Iron Age settlement, which may have roots further back in time. The settlement deposit was levelled at some point, perhaps in order to use the Holm for assembly meetings. No clear evidence of Norse activity was found, although the final stages of analysis are still ongoing. The seeming lack of Norse evidence should not necessarily be interpreted to mean that Ting Holm was not used as a thing site. The evidence linking the assembly meetings to the Holm is too strong to be discounted on the basis of negative evidence. Moreover, the reuse of older sites as assemblies was A. Sanmark 105 Journal of the North Atlantic Special Volume 5 105 isthmuses. One was presumably control, as it is difficult to arrive in secrecy to these narrow stretches of land (Fig. 11). Isthmuses may also have been seen as liminal, being located between the land and the sea, and forming integral parts of sailing routes (cf. Nymoen 1995:36). As suggested for Tingwall, the crossing of water to access the sites most likely served as a ritual boundary, as seems to have been the case in Scandinavia and Mainland Scotland. Assemblies by fords are very common in Scandinavia (Sanmark 2009:231–232), and there are also several examples of assemblies held on islands, such as Selaön and Adelsön in Södermanland and Enhälja in Uppland, Sweden (Sanmark 2009:225–227, Vikstrand 2001:247–248). Island locations fit in well with the idea of assemblies as sacred or liminal sites where special regulations applied, as many island cult sites are known, perhaps most notably from the Helgö (“Holy island”) place-names (Sanmark 2009:231–233, Vikstrand 2001:238–252). A second trait of the Shetland thing sites is the reuse of older monuments. As monument reuse was commonly occurring in Scandinavia for thing sites, burials, and settlement (Thäte 2007, Herschend 2009, Zachrisson 1994), this trait should hardly surprise us. There were naturally various reasons for monument reuse, but an important one seems to have been to connect with the local landscape, where natural and human-made features were seen to be “inhabited” by supernatural beings and ancestors (Sanmark 2010b:176). Another significant reason seems to have been to state land ownership via the addition of new burials to older burial grounds (Zachrisson 1994). When the Norse settled in a new area, these concepts seem to have been brought and translated to their new environment, as, e.g., at Helgafell in Iceland, where Thorolf Mostur-Beard and Thorstein Cod-Biter were seen to dwell after death, feasting with their ancestors (Sanmark 2010b:176, with references). By reusing Shetland monuments and perhaps in this way integrating them with their own past, the settlers ultimately made the local landscape their own. They stated their claim to the area, drawing on the past and the ancestors to legitimize their presence and power, in both a religious and political sense. A third contributing factor may have been practicality (cf. Semple and Sanmark 2013). Reused mounds, as seen across Scotland, would have served as useful site markers and perhaps platforms for speakers, while at the same time carrying religious and ancestral connotations. Another theme brought out by this article is that a variety of Norse administrative organizations existed in Shetland, from the herað units to quarters, eighths, parishes, and the elusive Gnípnaþing, buildings have not been properly dated, although some may be of medieval origin, the relationship between them and the thing sites cannot be verified (HU46NW 2, HU59SW 1, HU34NW 3, HU35NW 16, HU46NE 4; Ritchie 1997:94–95). Tingwall is an especially interesting case, as this site is located just below Tingwall Church (Fig. 2), which was presumably built between 1788 and 1790, close to or on top of the earlier St. Magnus Church, probably dating from the late 12th century. This site was the seat of the Arch Deacon and the most important church in Shetland (Cant 1975:21). The lawthing of 1307 was held here, most likely a temporary measure due to bad weather, as there is no tradition of thing meetings regularly taking place inside buildings until the 16th/17th century (Sanmark 2009:230). Discussion This article has pointed to a number of salient traits of the Shetland thing sites that can be used to draw conclusions regarding their wider role and function. The most common trait is the isthmus location, which occurs in Orkney and frequently in Scandinavia (Waugh 2010). In Norway, King Olaf Tryggvason is reported to have summoned an assembly for four districts at “Dragseith” (Hollander 1964:199, Westerdahl 2006b:41–42). Other examples include Tingvoll by Tiltereidet (Møre og Romsdal) and Eide by Eidsfjorden, which is one of the suggested sites for the Gulathing lawthing, or for the local thing in the skipreiða of Gulen (Helle 2001:52–53, Ringstad 2006:51–53, Westerdahl 2006b:41–42). A related example from Sweden concerns the thing site at Högsby, Handbörd herað, Småland. The OSw hundred name Andhbyrdh(e) denotes that the assembly is opposite the place where boats had to be carried across the water. This place name fits in with the local geography as Högsby is located where an esker (a major land route) crosses a river (Andersson 1965:27, Brink 2004:213). The reasons for assemblies being located by eið-names are many. One is, of course, communication, as these sites were accessible from several directions, and people from different parts of the administrative districts could therefore reach them more easily (cf. Westerdahl 2006b:42). In Norway, portages formed meeting places and crossroads from prehistoric to medieval times (Nymoen 1997:19), while Inuit trading and assembly sites in Greenland were often located on the narrow isthmuses where the Inuit carried their kayaks across land (Petersen 2006). There must, however, have been other reasons for the preference of holding assemblies on A. Sanmark 106 Journal of the North Atlantic Special Volume 5 106 which cannot be classified. As mentioned above, the thing parishes may be the latest addition to the administrative system, as part of a royal 13th-century reorganization (Smith 2009:42–44, 2011). This idea is to some extent supported by evidence presented in this paper, as the proposed thing sites are all rather centrally located within the parishes, which suggests they were chosen to fit in with this system. The reorganization is further supported by the link between thing sites and churches/chapels, as research from Scandinavia suggests that over the course of the Middle Ages, thing sites were increasingly moving to the vicinity of parish churches (Sanmark 2009). The Shetland sites therefore seem to be linked to administrative units of rather late date. This view does not mean, however, that all the sites were newly created at this point in time (contra Smith 2009). It has been illustrated for Sweden that thing sites established before the 11th century frequently reused older features, while for later sites, the overriding concern seems to have been to fit in with the administrative organization. As a result of this process, some old sites were kept in use, while new sites were also created (Sanmark 2009; Sanmark and Semple 2008, 2010). Two examples of the latter are Aspa in Södermanland and Arkel’s thing site in Uppland, which were “empty sites” with no pre-existing features when they came into use, although they were embellished according to the standards of the time. Examples of thing sites established on sites with a very long biography are Kjula ås in Södermanland and Anundshög in Västmanland, both of which were used for burial from the early Roman Iron Age (Sanmark and Semple 2008, 2010, 2011). Using this model, it can be argued that the two Shetland sites, Tingwall and Lunnasting, which clearly reuse older monuments, are older thing sites that were integrated into the parish system. This argument is strengthened by an examination of the thing sites on the Scottish Mainland. Two of the three sites that can be rather exactly identified, Dingwall in Ross-shire and Thingsva in Caithness, are found on top, or in the immediate vicinity, of earlier archaeology (O’Grady 2008:195–201).17 As stated above, these sites must have been created and used between the 9th and the 11th century at the most, showing the keen interest in monument reuse by the Norse during this time. The age of the other Shetland sites are unknown, but it is possible that some, or all, of these were newly created in the 13th century. The most likely such case is Gullaness in Delting, which seems to have been created on empty land in the same way as Aspa and Arkel’s thing site in Sweden. Finally, seeing the fluid administrative arrangements in Shetland, it seems clear that other sites than Figure 11. Dingieshowe, thing site located on a narrow isthmus, Orkney, Scotland. Photograph © David Griffiths, University of Oxford. A. Sanmark 107 Journal of the North Atlantic Special Volume 5 107 are further indications that this area could have been used as a thing site. The place-name Vassa, derived from ON Vatnseið18, refers to a stretch of low-lying land, just over 100 m wide (Stewart 1987:80), and a second portage can be envisaged between the Vadill of Garth and Trowie Loch.19 There are also two broch mounds, one in the loch itself and one on the shore (MacKie 2002:115 [HU45SE 18, HU45SE 14]). Finally, an issue that may further strengthen the case is that a court for Nesting, Lunnasting, Whalsay, and the Out Skerries was held in Brough in 1604 (Donaldson 1958:131–133) In conclusion, this article suggests that thing sites were established in Shetland from the early phases of settlement. The resemblances between thing-site locations in Scandinavia and Shetland show that similar concerns were present in the selection of suitable sites, although with some consideration for local conditions. Above all, however, the existence of thing sites in Mainland Scotland, in the areas ruled by the Norse, but which were not included among the Norwegian skattlands, reinforces the idea that thing sites formed part of the culture brought by the Norse settlers to their new homelands, and were not necessarily enforced by a centralized kingdom and its machineries. Acknowledgments I would like to express my gratitude to a number of people who have read and commented on earlier drafts of this paper: Oscar Aldred, Barbara Crawford, Brian Smith, two anonymous reviewers, and my colleagues from The Assembly Project, Natascha Mehler and Sarah Semple. Brian Smith has also helped with obtaining material from the Shetland Archives, while Peder Gammeltoft has generously shared his knowledge of place-names. I am moreover grateful to the Aithsting History Group, Jenny Murray, Brydon Leslie, Julian Arculus, Andrew Jennings, and Fredrik Sundman for helping me on my search for the Shetland thing sites. Literature Cited Andersson. T. 1965. Svenska häradsnamn. Uppsala University, Uppsala, Sweden. 397 pp. Ballantyne. J.H., and B. Smith. 1994. (Eds). Shetland Documents 1580–1611. Shetland Islands Council and Shetland Times, Lerwick, UK. 329 pp. Ballantyne. J.H., and B. Smith. 1999. (Eds). Shetland Documents 1195–1579. Shetland Islands Council and Shetland Times, Lerwick, UK. 359 pp. Bauge, S. 1980–1982. Fjerding. Cols. 379–380, In J. Granlund (Ed.). Kulturhistoriskt lexikon för nordisk medeltid. Från vikingatid till reformationstid. Vol. 4. Rosenkilde og Bagger, Köpenhamn, Denmark. Brand, J. 1883. A Brief Description of Orkney, Zetland, Pightland-Firth, and Caithness. William Brown, Edinthose discussed here must have existed. Using the site characteristics outlined above, it is possible to suggest at least one more possible thing site, situated in the area around the Loch of Benston in the southern part of Nesting. This site is located on a major ness, which equally could have given name to the thing district (Fig. 12; cf. Smith 2009:41). One of the most important pieces of evidence is the place-name Freester, from ON Friðrsetr/Friðarsetr. ON friðr translates as peace, personal security, or inviolability, and ON setr as seat, residence, or dairy lands (Cleasby 1874:173, 526; Heggstad et al. 1975:127; Stewart 1987:236; Zoëga 1910:150, 375). This nomenclature could be a parallel to Grista, derived from griðr, by Tingwall. The two terms, friðr and griðr, seem to have been interchangeable in relation to assemblies as The Law of Uppland (Sweden) refers to Disaþings friþer, i.e., the peace that lasted for the duration of the thing meetings and markets at Gamla Uppsala (Holmbäck and Wessén 1933:205, Schlyter 1877:119, Sundqvist 2002:198–199), and the term frið-staðr was at times be used for asylum (Cleasby 1874:173, 215; Zoëga 1910:172). There Figure 12. The area around the Loch of Benston, Nesting. © Crown Copyright/database right 2011. An Ordnance Survey/EDINA supplied service. All rights reserved 2010. Map created by Alex Sanmark and Brian Buchanan. A. Sanmark 108 Journal of the North Atlantic Special Volume 5 108 Gunnes, E., and H. Kjellberg. (Eds.). 1979. Regesta Norvegica IV. Kjeldeskriftfondet, Oslo, Norway. 515 pp. Heggstad. L., F. Hødnebø, and E. Simensen. 1975. Norrøn ordbok. Det Norske samlaget, Oslo, Norway. 518 pp. Helle, K. 2001. Gulatinget og Gulatingslova. Skald, Leikanger, Norway. 240 pp. Herschend, F. 2009. The Early Iron Age in South Scandinavia: Social Order in Settlement and Landscape. Department of Archaeology and Ancient History, Uppsala University, Uppsala, Sweden. 499 pp. Hollander, L.M. (Ed. and Tr.). 1964. Heimskringla: History of the Kings of Norway. University of Texas Press, Austin, TX, USA. 854 pp. Holmbäck, Å., and E. Wessén. (Eds) .1933. Östgötalagen och Upplandslagen. Svenska Landskapslagar. Tolkade och förklarade för nutidens svenskar. Första Serien. Hugo Gebers förlag, Stockholm, Sweden. 461 pp. Indrebø, G. 1936. Fjordung: Granskingar i eldre norsk organisasjons-soge. Bergens museums Årbok. Historisk- antikvarisk rekke 1935:1. Bergen, Norway. 287 pp. Iversen, F., N. Mehler, A. Sanmark, and S. Semple. In press. Negotiating the North, Society for Medieval Archaeology Monographs. Maney, Leeds, UK. Jakobsen, J. 1936. The Place-Names of Shetland. David Nutt, London, UK. 273 pp. Larson, L.M. 1935. (Ed. and Tr.). The Earliest Norwegian Laws. Being the Gulathing Law and the Frostathing Law. Morningside Heights, New York, NY, USA. 451 pp. MacKie, E.W. 2002. The Roundhouses, Brochs, and Wheelhouses of Atlantic Scotland c. 700 BC–AD 500: Orkney and Shetland Isles Pt. 1: Architecture and Material Culture. British Archaeological Reports British Series, Archaeopress, Oxford, UK. 388 pp. McCullough, D.A. 2000. Investigating portages in the Norse maritime landscape of Scotland and the Isles. Unpublished Ph.D. Dissertation. University of Glasgow, Glasglow, UK. 379 pp. Myrberg, N. 2008. Room for all? The Gotlander’s al and the all-thing in Roma. Viking and Medieval Scandinavia 4:133–58. National Library of Scotland. 1843–1882. Ordnance survey maps - six-inch 1st edition, Scotland. Available online at http://maps.nls.uk/os/6inch/. Accessed 27 September 2011. Nymoen, P. 1995. Sjøveien over land - Om eid og båtdrag i Midt-Norge. Spor 1:34–36. Nymoen, P. 1997. “Der er en eiendommelighet ved den Norske Kyst” - om ferdsel over land og vann. Kysten 1:16–20. O’Grady, O. 2008. The setting and practice of open-air judicial assemblies in medieval Scotland: A multidisciplinary study. Unpublished Ph.D. Dissertation. University of Glasgow, Glasglow, UK. 649 pp. Available online at http://theses.gla.ac.uk/506/. Accessed 2 October 2011. Paton, D. 2011. Things aren’t always what they seem … A critique of recent approaches to early medieval assembly sites in Britain with a case-study focussing of Thingwall, Wirral. Unpublished B.A. Thesis. Department of Archaeology, University of Chester, Chester, UK. Pálsson, H. (Ed.). 2003. Snorra-Edda. Mál og menning: Reykjavík, Iceland. 308 pp. burgh, UK. 247 pp. Brink, S. 2004. Legal assembly sites in early Scandinavia. Pp. 205–16, In A. Pantos and S. Semple (Eds.). Assembly Places and Practices in Medieval Europe. Four Courts Press, Dublin, Ireland. 251 pp. Bruce, Mr. 1908. Description of Ye country of Zetland. Printed for private circulation. No prov. Available from the Orkney Library and Archive. 99 pp. Cant, R.G. 1975. The medieval churches and chapels of Shetland. Shetland Archaeological and Historical Society, Lerwick, UK. 50 pp. Cleasby, R. 1874. An Icelandic-English Dictionary. Enlarged and completed by Gudbrand Vigfusson. Clarendon Press, Oxford, UK. 901 pp. Available online at http://www.archive.org/details/icelandicenglish- 00cleauoft. Accessed 3 October 2011. Clouston, J.S. (Ed.) 1914. Records of the Earldom of Orkney 1299–1614. Scottish History Society, Edinburgh, UK. 515 pp. Coolen, J., and N. Mehler. 2011. Archaeological excavations at the Law Ting Holm, Tingwall, Shetland, 2011. Data Structure Report/Interim Report. The Assembly Project Field Report No. 4. Available from http:// www.khm.uio.no/english/research/projects/assemblyproject/. 30 pp. Crawford, B.E. 1984. Papa Stour: Survival, continuity, and change in one Shetland island. Pp. 40-58, 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. 347 pp. Dodgson, J. McN. 1972. The Place-Names of Cheshire. Part Four: The Place-Names of Broxton Hundred and Wirral Hundred, Cambridge. Cambridge University Press, Cambridge, UK. 340 pp. Donaldson, G. (Ed.) 1958. The Court Book of Shetland 1602–1604. Scottish Record Society, Edinburgh, UK. 176 pp. Edmondston, A. 1809. A View of the Ancient and Present State of the Zetland Islands. Volume I. Longman, Hurst, Rees, and Orme. John Ballantyne and Co., Edinburgh, UK. 364 pp. Faulkes, A. 1987. (Ed. and Tr.) Snorri Sturluson: Edda. Everyman, London, UK. 260 pp. Fellows-Jensen, G. 1996. Tingwall: The significance of the name. Pp. 16–29, In D. Waugh (Ed.). Shetland’s Northern Links. Scottish Society for Northern Studies, Edinburgh, UK. 254 pp. Fladby, R. 1980–1982. Attung. Pp.. 276–278, In Granlund, J. (Ed.). Kulturhistoriskt lexikon för nordisk medeltid. Från vikingatid till reformationstid. Vol. 1. Rosenkilde og Bagger, Köpenhamn, Denmark. Friðriksson, A. 1994. Sagas and Popular Antiquarianism in Icelandic Archaeology. Avebury, Aldershot, UK. 212 pp. Graham, J.J. 1993. The Shetland Dictionary. Shetland Times, Lerwick, UK. 124 pp. Available online at http:// www.shetlanddialect.org.uk/john-j-grahams-shetlanddictionary- intro Accessed 27 September 2011. Greig P.W. 1892. Annals of a Shetland Parish: Delting. C and A Sandison, Lerwick, UK. 100 pp. A. Sanmark 109 Journal of the North Atlantic Special Volume 5 109 nådigste befallning utgifven af d. C.J. Schlyter. Vol. 13, Glossarium ad Corpus iuris Sueo-Gotorum antiqui. Ordbok till Samlingen af Sweriges gamla lagar. Berlingska boktryckeriet, Lund, Sweden. 818 pp. Semple, S., and A. Sanmark. 2013. Assembly in North West Europe: Collective concerns for early societies? European Journal of Archaeology 16(3):518–542. Smith, B. 2009. On the nature of tings: Shetland’s law courts from the middle ages until 1611. New Shetlander 250:37–45. Smith, B. 2011. Hákon Magnusson’s root-and-branch reform of public institutions in Shetland, c. 1300. Pp. 103–112, In S. Imsen (Ed.) Taxes, Tributes, and Tributary Lands in the Making of the Scandinavian Realm in the Middle Ages. ROSTRA books 7. Tapir Akademisk Forlag Trondheim, Norway. 357 pp. Stewart, J. 1987. Shetland Place-names. Shetland Library and Museum, Lerwick. UK. 353 pp. Sundqvist, O. 2002 Freyr’s Offspring. Rulers and Religion in Ancient Svea Society. Uppsala, Sweden. 420 pp. Sølvará, H.A. 2002. Løgtingið 150 - Hátíðarrit 1. Løgtingið og høvundarnir, Tórshavn, Faroe Islands. 400 pp. Thäte, E.S. 2007. Monuments and Minds. Monument Reuse in Scandinavia in the Second Half of the First Millennium AD. (Acta Archaeologica Lundensia Series in 4o No. 27). Wallin and Dalholm, Lund, Sweden. 336 pp. Vikstrand, P. 2001. Gudarnas platser. Förkristna sakrala ortnamn i Mälarlandskapen. Gustav Adolfs akademien, Uppsala, Sweden. 482 pp. Waugh, D. 2010. On eið-names in Orkney and other North Atlantic islands. Pp. 545–554, In J. Sheehan and Ó. Corráin (Eds.). The Viking Age: Ireland and the West. Papers from the Proceedings of the Fifteenth Viking Congress, Cork, 18–27 August 2005. Four Courts Press, Dublin, Ireland. 569 pp. Westerdahl. C. 2006a. (Ed.). The Significance of Portages: Proceeding of the First International Conference on the Significance of Portages, 29 Sept–2nd Oct 2004, in Lyngdal, Vest-Agder, Norway, arranged by the County Municipality of Vest-Agder, Kristiansand International Conference on the Significance of Portages. Archaeopress, Oxford, UK. 273 pp. Westerdahl. C. 2006b. On the significance of portages. A survey of a new research theme. Pp. 15–52, In C. Westerdahl (Ed.). The Significance of Portages: Proceeding of the First International Conference on the Significance of Portages, 29 Sept–2nd Oct 2004, in Lyngdal, Vest-Agder, Norway, arranged by the County Municipality of Vest-Agder, Kristiansand International Conference on the Significance of Portages. Archaeopress, Oxford, UK. 273 pp. Zachrisson, T. 1994. The odal and its manifestation in the landscape. Current Swedish Archaeology 2(1994):219–238. Zoëga, G.T. 1910. A Concise Dictionary of Old Icelandic. Clarendon Press, Oxford, UK. 551 pp. Peel. C. (Ed. and Tr.). 2009. Guta lag: The Law of the Gotlanders. Viking Socety for Northern Research, University College of London, London, UK. 235 pp. Petersen, R. 2006. The Greenland portages as meetingsites. Pp. 179–186, In C. Westerdahl (Ed.). The Significance of Portages: Proceeding of the First International Conference on the Significance of Portages, 29 Sept–2nd Oct 2004, in Lyngdal, Vest-Agder, Norway, arranged by the County Municipality of Vest-Agder, Kristiansand International Conference on the Significance of Portages. Archaeopress, Oxford, UK. 273 pp. Poulsen, J.H.W., and U. Zachariasen. (Eds.) 1971. Seyðabrævið. Rættarbót Hákunar hertuga Magnussonar fyri Føroyar. Tórshavn: Føroya fróðskaparfelag, Faroes. 65 pp. Riisøy, A.I. 2013. Sacred legal places in the Edda. In A. Sanmark (Ed.). The Assembly Project: Debating the Thing in the North 1. Journal of the North Atlantic Special Volume 5:28–42. Ringstad, B. 2006. Tingvoll. Pp. 25–80, In M. Stige and T. Spurkland (Eds.). Tingvoll kyrkje. Gåta Gunnar gjorde. Senter for middelalderstudier, skrifter nr. 21, Trondheim, Norway. Ritchie, A. 1997. Shetland. Exploring Scotland’s Heritage series. The Stationary Office, Edinburgh, UK. 148 pp. Ross-Shire Journal. 2011. Car park was court of law for Vikings! Available online at http://www.ross-shirejournal. co.uk/News/Car-park-was-court-of-law-for- Vikings-7305257.htm. Accessed 21 November 2011. Sandnes, J., and O. Stemshaug (Eds.). 1976. Norsk stadnamnleksikon. Det norske samlaget, Oslo, Norway. 359 pp. Sanmark, A. 2006. The communal nature of the judicial system in early medieval Norway. Collegium Medievale 19:31–64. Sanmark, A. 2009. Thing organisation and state formation. A case study of thing sites in Viking and medieval Södermanland, Sweden. Medieval Archaeology 53:205–241 Sanmark, A. 2010a. The case of the Greenlandic assembly sites. Journal of the North Atlantic, Special Volume 2:182–196. Sanmark, A. 2010b. Living on: Ancestors and the soul. Pp. 162–184. In M. Carver, A. Sanmark, and S. Semple (Eds.). Signals of Belief in Early England: Anglo- Saxon Paganism Revisited. Oxbow, Oxford, UK. Sanmark, A., and S. Semple. 2008. Places of assembly: Recent results from Sweden and England. Fornvännen 103:245–260. Sanmark, A., and S. Semple. 2010. The topography of outdoor assembly in Europe with reference to recent field results from Sweden. Pp. 107–119, In H. Lewis and S. Semple (Eds.). Perspectives in Landscape Archaeology. BAR International Series 2103. Archaeopress, Oxford, UK. 119 pp. Sanmark, A., and S. Semple. 2011. Tingsplatsen som arkeologiskt problem, Etapp 3: Anundshög [Excavation Report from Anundshög, Sweden]. The Assembly Project, Report 3. Available from http://www.khm.uio. no/english/research/projects/assembly-project/. Schlyter, C.J. 1877. Corpus iuris sueo-gotorum antiqui. Samling af Sweriges gamla lagar, på kongl. maj:ts A. Sanmark 110 Journal of the North Atlantic Special Volume 5 110 of use. This may be the case, but if the suggestion put forward in this article is accepted, the thing site clearly remained in use. 14Taing is defined as “flat land projecting into the sea” (Graham 1993:91). 15Smith (2009:43) has dismissed the idea of Gardiesting on the basis that Gardiestaing may well have been an appropriate name for this area in the past, prior to known mercantile activities in the 1750s. 16The many reasons for portages have been summarized in Nymoen (1995) and Westerdahl (2006b). For the Northern Isles, the most important may well have been to avoid wild waters/tides and not having to wait for the right wind. The necessity and frequent usage of portages by the Norse and in other cultures is demonstrated in The Significance of Portages (Westerdahl 2006a). 17The Dingwall site is located on top of an as yet undated mound, and a class I Pictish symbol stone has been found during the augmentation of the church building located in the immediate vicinity of this mound. These pieces of evidence suggest that this was a significant site in the early Middle Ages (O’Grady 2008:195–198, Ross-Shire Journal 2011). The Thingsva assembly site is located on top of/ next to a broch mound (O’Grady 2008:198–201). The suggested thing sites in the Western Isles (i.e., Eilean Thinngartsaigh in Harris, Tiongal in Lewis, Hinnisdale [Tinwhil], Trotternish in the Isle of Skye) need further investigation before any conclusions can be drawn (Fellows-Jensen 1996:23–24, O’Grady 2008:201–203). 18ON vatn = lake (Stewart 1987:80). 19This area is a tidal zone, and Trowie Loch was presumably also bigger in the past as a large area of drainage ditches is present between this site and Benston Loch. Endnotes 1The Papa Stour “stone ring” will be discussed in Iversen, Mehler, Sanmark and Semple, forthcoming. 2Arthur Edmonston (1809:130) stated that “the site of the bench and surrounding seats can still be traced.” 3For a discussion of the regulations, see Sanmark (2009:232). 4Several versions of this tradition exists, see, e.g., Smith (2009:41). 5Herra on Yell, which splits into the Utherra and the Inherra. The former includes the farms on the west side of Whalfirth Voe, while the latter referred to the east side of the voe and west of the hill of Camb. Herra on Fetlar contained eight tuns north of Loch of Papil (Northdale, Baelans, Southdale, Crossbister, Tafts, Newerhouse, Clothin, Velzie), and was divided into the Upper and Lower Herra. Herra in Lunnasting contains all the crofts around Vidlin Voe to the east and south. The Harray, Tingwall parish, Litillogartht, and Hamyrisland said to be “within the Harray” in 1525, and Herrislee Hill is close by (Stewart 1987:130). 6This reorganization is suggested to have involved rents, taxation, and the judicial system (Smith 2009:42–44, 2011). 7For Sweden see, e.g., Torstuna in Torstuna härad, Frösåker in Frösåkers härad, and Frötuna in Frötuna och Länna skeppslag. Frötuna later became the name of the parish when the church was erected by Frötuna village (Vikstrand 2001:151, 169, 187). 8Jakobsen (1936:125, 143), however, interpreted the name to be based on ON lundr “grove”. 9This has been seen as derived from *götu-tangi, “a tongue of land, across the neck of which a path runs” (Jakobsen 1936:45; P. Gammeltoft, Department of Scandinavian Research, University of Copenhagen, Denmark, pers. comm.). 10The ca. 1510 document is very corrupt and lists a number of places, including Nepnatang, under Lunnasting although they are located in Nesting. The form Neipnating comes from a 1628 document (National Archives of Scotland, E41/7, [1628]), which on the whole is a lot more trustworthy (B. Smith, Shetland Museum and Archives, Lerwick, Shetland, pers. comm.). 11After the publication of Smith (2009), a person came forward relating a tradition that this was the site of the Neipnating (B. Smith, pers. comm.). The exposed stones are unlikely to be an archaeological feature, but rather exposed bedrock, although that does not disqualify them from having been used for thing meetings. 12The name has been seen to derive from ON reyðr (rorqual, finner-whale) and fjörðr (fjord) (P. Gammeltoft, pers. comm.; Stewart 1987:87). Indeed the voe on the other side of Mid Yell is now called Whalefirth. The hills stretching up to 4 km south of Gardie are named Hill of Reafirth, highlighting the earlier importance of this name. 13Smith (2009:41) has argued Þvæitaþing and Rauðarþing formed part of his suggested royal reorganization and that these two names did not “catch on” and fell out