Journal of the North Atlantic J. Coolen 2016 Special Volume 8 93 Introduction The term gallows refers to a structure used for execution by hanging and is often used as a synonym for gibbet, although the latter generally signifies a gallows-type structure from which bodies were hanged on public display after the execution. While gallows could be temporary structures that were erected at the scene of crime or at frequented places such as town squares, many judicial districts had dedicated places of execution that served both as hanging and gibbeting sites in the medieval and early modern period. Throughout medieval and early modern Europe, gibbets and gallows defined and reinforced the boundaries of judicial authority and moral behavior and thereby contributed to the shaping of collective identities (Coolen 2013). The rituals associated with public executions and the subsequent display of the malefactor’s body have been described as rites de passage, which purified society of criminal elements and confirmed social and religious values (Evans 1996:99–108). Even centuries after being abolished, the tentative places of execution often remain embedded in the collective understanding of the landscape. Throughout Shetland there are a remarkable number of place-names which suggest the former presence of gallows (Fig. 1). One of the most striking sites is located near the main assembly area at the Law Ting Holm in Tingwall and would have been clearly visible to the attendants of the assemblies held there (Coolen and Mehler 2014:5f.). The close spatial and functional connection of the ancient court site and Gallow Hill has long been recognized (Turnbull 1845:60). However, the historical and administrative contexts of the other alleged gallows sites are more obscure. It has been suggested that the gallows sites are part of Shetland’s Norse heritage (Smith 2006), but this general statement will be challenged in this paper. Historical or archaeological evidence for the alleged places of execution in Shetland is sparse. Nevertheless, we can try to get a better understanding of these places by looking at their topography and the spatial relationship with known judicial sites and boundaries. In this paper, I present the evidence of the known or alleged places of execution in Shetland and investigate their topographical setting, especially with respect to presumed assembly sites and parish boundaries. The main research questions I address are: What is the evidence for places of execution in Shetland? How can these sites possibly be dated? and What is the relationship between the tentative places of execution and (post-) medieval court sites and districts? The paper includes the results of a survey carried out in connection with the fieldwork of The Assembly Project in Shetland in 2011 (Coolen 2012). Shetland’s Alleged Gallows Sites In 2006, Brian Smith delivered a paper on Shetland’s gallows sites. While this paper still awaits formal publication, the original text has been published on a weblog (Smith 2006). In that paper, Smith dealt with the evidence of former places of execution from place-names and local traditions and tried to interpret these sites within a historical and landscape context. He originally counted 13 sites and later added a fourteenth site on the isle of Whalsay.1 However, he now considers some of these sites problematic (B. Smith, Shetland Museum and Archives, Lerwick, UK, 30 May 2011 pers. comm.). Nonetheless, all sites are included in the present paper, even though the dubious character of some of the sites is acknowledged (Table 1).2 Smith (2006) noted several common features of Shetland’s gallows sites: they are all highly visible sites, often with steep access; most are marked by prehistoric structures (primarily cairns); some of Gallows, Cairns, and Things: A Study of Tentative Gallows Sites in Shetland Joris Coolen* Abstract - In this paper, I summarize the evidence of former gallows sites in Shetland from place-names, oral traditions, historical records, and archaeological remains. I make an attempt to date the alleged places of execution by comparison of their spatial distribution with known or presumed historical assembly or court sites and districts. I argue that the Gallow Hills are associated with the post-medieval judicial organization of Shetland rather than the Norse division and may therefore be later in date than has been suggested before. Furthermore, I show that some of the oral traditions associated with the gallows sites reflect notions of liminality and hidden world s that have parallels in other parts of northwestern Europe. Debating the Thing in the North: The Assembly Project II Journal of the North Atlantic *Institute of Prehistoric and Historical Archaeology, University of Vienna, Franz-Klein-Gasse 1, 1190 Vienna, Austria; joris. coolen@univie.ac.at. 2016 Special Volume 8:93–114 Journal of the North Atlantic J. Coolen 2016 Special Volume 8 94 the sites lie close to important medieval settlements; and finally, some are associated with place-names that refer to old judicial districts, namely “ting” and “herra”. Based on these observations, Smith argued that the gallows sites “had their origin early in the period when Shetland was a Scandinavian country”. With one possible exception (which shall be discussed later), there are no known visible remains of gibbets or gallows in Shetland. In some cases, however, the morbid history of the site is recalled in oral traditions. While the local folk tales are highly informative about the perception of places, it must be taken into account that they also have their own cultural biography and may well have developed later or may have significantly changed over time. In short, the evidence for the alleged places of execution in Shetland is scarce and needs to be critically assessed for each site individually. By comparing the nature of the sites and their position in the wider landscape within Shetland as well as in other regions, we might find similarities that could give us some more information about the history of these intriguing places. Geographical information systems (GIS) provide valuable tools to analyze the topography of the sites and their relationship with both the physical environment and the historic landscape. Historical Evidence The earliest reference to a gallows in Shetland dates from 1574. Five men, who had been sentenced to death for ransacking a shipwreck, were pardoned after they had been kept standing at the foot of the gallows with the noose attached for 2 hours (Balfour 1859:11). The trial was held at Scalloway Banks—the later site of the castle—and presided over by Lord Robert Stewart. The latter had acquired the earldom and sheriffship of Shetland and Orkney after having received a feu of the islands from his halfsister, Queen Mary of Scotland in 1564 (Anderson 1982, 1996:179). The case against the ship-wreckers shows that the lawthing, Shetland’s traditional law court, had been moved from its original site in Tingwall to Scalloway soon after Robert came into power. It was later held in Scalloway Castle, commissioned by Robert’s son and successor, Patrick Stewart. From the reign of the Stewarts onwards, the preserved court books give a good view of judicial practice in Shetland (Table 2). The trial of Christopher Jhonsoun in 1602 is the earliest recorded case leading to an execution (Donaldson 1954:18f.). It is also the only known case in which capital punishment was imposed at a local court meeting, namely at Wethersta in Delting. It is stated in the court book that the convict shall be hanged at the gallows hill, but it is not clear whether this refers to Figure 1. Location of place-names in Shetland, which refer to gallows. Sites: 1 Gallow Hill, Huesbreck; 2 Golgo, Sandwick; 3 Knowe of Wilga, Cunningsburgh; 4 Gallow Hill, Scalloway; 5 Gallow Hill, Tingwall; 6 Gallow Hill, Walls; 7 Gallow Hill, Brae; 8 Yamna Field (Gulga?), Gluss; 9 Watch Hill (Gallow Hill?), Eshaness; 10 Gallows Knowe, Mid Yell; 11 Gallow Hill, Fetlar; 12 Gallow Hill, Unst; 13 Muckle Heog (Hanger Heog), Unst; 14 Setter Hill (Gallow Hill?), Whalsay; and 15 Wilgi Geos, North Roe. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 95 Table 2. Overview of historically recorded death penalties and o ther references to execution sites in Shetland. Year Convict(s) Crime Sentence Place of execution References 1574 Gilbert McReich Looting Initially sentenced to Gallow Hill, Scalloway? Balfour 1859, 11 David Leslie death, later pardoned James Leslie Normand Leslie 1602 Christopher Jhonsoun Thievery Hanged Gallow Hill, Scalloway/ Donaldson 1954:18f. Delting? 1615 Christopher Esplein Thievery Hanged Gallow Hill, Scalloway Barclay 1967:116, Donaldson 1991:21 1615 Bothuel Erasmussone Iver Manssone Thievery (?) Hanged Gallow Hill, Scalloway Barclay 1967:117 1616 Katherine Jones-dochter Witchcraft, Strangled and burnt Gallow Hill, Scalloway Donaldson 1991:43 Jonka Dyneis contact with (females), hanged (male) Barbaray Scord the devil, Robert Boundsone thievery 1618 William Sutherland Murder Beheaded Outside Scalloway castle Donaldson 1991:71 1618 John Thomson Bestiality Strangled and burnt Gallow Hill, Scalloway Donaldson 1991:72f. 1625 Robert Ingsetter Thievery Hanged Gallow Hill, Scalloway Donaldson 1991:124 1625 Marioun Thomas-dochter Thievery Thrown down the cliffs “Luckymenis Ness” Donaldson 1991:124 1628 Mans Christophersson Thievery Hanged Gallow Hill, Scalloway Donaldson 1991:146 1628 William Cogill Thievery Hanged Gallow Hill, Scalloway Donaldson 1991:154 1644 Marion Peebles Margaret Guthram-daughter Witchcraft Strangled and burnt Gallow Hill, Scalloway Hibbert 1822:575, 593–602 ca. 1675 “Luggie” Witchcraft Burnt at the stake Scalloway Brand 1701:167, Sinclair 1685:237f. ca. 1680 Barbara Tulloch Witchcraft Burnt at the stake Gallow Hill, Scalloway? Sinclair 1685:231, Ellen King (= Helen Stewart?) Smith 2006, Turnbull 1845:60 1685 John Johnson Thievery Corporal punishment, Gallow Hill, Scalloway Smith 2006 banished the Gallow Hill in Brae, only few kilometers north of Wethersta, or the historically attested gallows site in Scalloway. Later on, capital offenses were exclusively dealt with at the court in Scalloway. The only gallows site that is explicitly mentioned in historical documents is the Gallow Hill in Scalloway, sometimes specified as the hill above Berry or Houll. According to the court books, several executions took place here in the 17th century. The convicts were either hanged or strangled and burnt at the stake. Other places of execution that appear in the court books are the garden of Scalloway castle Table 1. Overview of the discussed sites and the various strains of evidence for gallows. The italicized names are not marked on Ordnance Survey maps; the conventional names of these sites, or where th ey were mapped, are given in brackets. Historical Local Visible Prehistoric Place-name Township, parish, island Coordinates sources tradition remains monuments Gallow Hill Huesbreck, Dunrossness, Mainland HU 3929 1443 - X - Cairn Golgo Sandwick, Dunrossness, Mainland HU 4281 2348 - X - Standing stone (vanished); prehistoric burials? Knowe of Wilga Cunningsburgh, Cunningsburgh, Mainland HU 4236 2664 - X - Cairn Gallow Hill Scalloway, Tingwall, Mainland HU 3991 3963 X X - - Gallow Hill Tingwall, Tingwall, Mainland HU 4105 4279 - - ? Cairns Gallow Hill Walls, Walls, Mainland HU 2556 5104 - ? - Cairns Gallow Hill Brae, Delting, Mainland HU 3774 6810 ? - - - Gulga (Yamna Field?) Gluss, Northmavine, Mainland HU 3379 7660? - - - Cairn Gallow Hill (Watch Hill?) Eshaness, Northmavine, Mainland HU 2590 7809? - - - Cairn Gallows Knowe Halsagarth, Mid Yell, Yell HU 4878 9122? - X - - Gallow Hill Fetlar HU 5919 9070 - X ? - Gallow Hill Belmont, Unst HP 5752 0070 - - - Cairns Hanger Heog (Muckle Heog) Baltasound, Unst HP 6315 1081 - ? - Cairns Gallow Hill (Setter Hill?) Marrister, Whalsay HU 5480 6388? - - - - Wilgi Geos North Roe, Northmavine, Mainland HU 3444 9163 - - - - Journal of the North Atlantic J. Coolen 2016 Special Volume 8 96 (1618, beheading) and the cliffs at Luckymenis Ness (1625, falling), but these seem to have a more ad hoc character (Donaldson 1991:71, 124). If the other Gallow Hills in Shetland were indeed used as execution sites, the historical records provide a terminus ante quem. The question therefore is: how far back these execution sites could possibly date? It is unlikely that the alleged execution sites predate the Norse colonization at the turn of the 9th century, as there is no evidence that anybody had the authority required to inflict capital punishment and lastingly designate places of execution throughout Shetland before that time. The Viking settlers probably established a system of local thing councils, which were all subject to 1 general assembly, like in other parts of medieval Scandinavia (Sanmark 2013, Smith 2009). Throughout the Scandinavian countries, the althing acted as the main legislative body, initially following unwritten local laws, but had limited executive power. Outlawry was the ultimate sentence that could be imposed by the althing. From the 11th century onwards, efforts were made to record and eventually generalize the laws of the different assemblies in the Norwegian kingdom. The introduction and implementation of royal, centrally enforced legislation became the main task of the assemblies, and they became lawthings instead of althings. Shetland’s althing adopted the Gulathing version of Magnus Hákonarsson’s law code at the end of the 13th century, which officially remained valid until the 17th century. Although the Norse laws did prescribe capital punishment in rare cases, the thing courts still had no central executive power. In Iceland, for example, the althing seems not to have executed death penalties before 1564 (Bell 2010:31). Most crimes were sentenced with fines or outlawry, though blood vengeance also seems to have been common practice well into the 13th century. Death penalties, especially by hanging, were rare in medieval Scandinavia. However, several references to the hanging of thieves and traitors can be found in the Heimskringla sagas, recorded by Snorri Sturluson around 1230. From the second half of the 13th century, the Scots gained increasing power in the Northern Isles, culminating in the annexation of the islands in 1468/69 (Crawford 1983). Nevertheless, Scottish law was only fully adopted in Shetland in 1611. Two years before, Patrick Stewart had been arrested upon charges of treason. After his arrest, the lawthing became a Scottish Sheriff Court (Barclay 1962:18). The court continued to reside in Scalloway until it was moved to Lerwick in the mid-18th century (Dunbar et al. 1839:203). It is generally believed that Barbara Tulloch and her daughter, who were both accused of witchcraft, are the last persons to have been executed in Shetland. The date of this event is not definitively known. Based on the sparse evidence, it may have taken place around 1680.3 After the Act of Union of 1707, the Scottish sheriffs became subject to greater direction from the Justiciary Court in Edinburgh in criminal cases (B. Smith, 5 June 2014 pers. comm.), but major reforms to the organization of Scottish justice were only introduced by the Heritable Jurisdictions Act passed in 1747 (Kilday 2007:32f.). The most important change was the abolition of hereditary sheriffships, which in Shetland had been last held by the Earls of Morton. By the mid-18th century, serious criminal cases were mostly brought before the Court of Justiciary. Public executions were finally abolished in the United Kingdom by the Capital Punishment Amendment Act of 1868. To summarize the above, historical records show that a traditional execution site existed in Scalloway at least between 1574 and the 1680s. The site was located on Gallow Hill, overlooking the town in the west. It was used for various corporal and capital punishments, including hanging and burning. There is no evidence that executions took place elsewhere in Shetland after the beginning of the 17th century, which implies that if the other tentative gallows sites were ever used, it must have been at an earlier stage. Place-Names As stated in the introduction, the main indicator for the former existence of local execution sites in Shetland is provided by place-names. A total of 15 places bearing names that have been associated with gallows have been recorded in Shetland (see Fig. 1). In light of the lack of historical evidence, the significance of the nomenclature deserves closer attention. By far most common is the name Gallow Hill, which occurs at least 7 times, namely in Dunrossness, Tingwall, Scalloway, Walls, Delting, Unst, and Fetlar. Apparently, there are also Gallow Hills at Eshaness in Northmavine (Smith 2006) and on Whalsay, but these names do not appear on Ordnance Survey maps.4 In addition, the place-name Gallows Knowe is reported near Halsagarth, a deserted croft opposite the infamous Windhouse in Mid Yell (Smith 2006), but this name is neither marked on OS maps nor is it generally known to local residents nowadays. Four place-names seem to derive from gálgi, the Old Norse word for gallows (Jakobsen 1897:118f.). The name appears as “golgo/gulga” in Sandwick (Golgo) and Northmavine (Scord o Gulga), and in the slightly more deviated form “wilga/wilgi” in Cunningsburgh (Knowe of Wilga) and North Roe (Wilgi Geos). Journal of the North Atlantic J. Coolen 2016 Special Volume 8 97 Finally, the site of Muckle Heog on Unst is referred to as Hanger Heog in antiquarian reports (Hibbert 1822:406, Low 1774). Despite its rather implausible classification as an execution site, the name more likely refers to a steep, overhanging hill (Smith 2006). It is remarkable that most of the suggested places of execution bear an English name. Reflecting 700 years of Norse control, the vast majority of placenames in Shetland are of Norse origin. English names present notable exceptions and are generally considered younger (Nicolaisen 1983). There are several possible explanations for the parallel occurrence of gallow- and gálgi-type names in Shetland. Firstly, it could be argued that the strong dominance of the gallow-type may result from the fact that it is more obvious, whereas the gálgi-type is harder to recognize. Hence, names of the latter type may not have been identified yet or may simply have been forgotten. However, there has been a longstanding interest in Norse place-names in Shetland and they have been extensively recorded. Moreover, the number of alleged places of execution in Shetland already is surprisingly high considering the size and population density of the islands. It seems unlikely that many remain to be discovered. Moreover, this argument does not explain why some of the sites kept their Norse name, while others bear an English name. Secondly, Smith (2006) has suggested that all gallows sites were originally called gálgi, as he believes their origin lies early in the Norse period. This interpretation implies that the gallow-type names were later translated. This shift could have happened only as long as the meaning of the old name was still understood, either because the gallows were still in use or because their memory was still alive (cf. O’Grady 2008:360). The English word “gallows” (OE galga or gealga) and ON gálgi are closely related in terms of etymology. Both derive from Old Germanic *galgon, meaning “pole”, which in turn can be traced back to a Proto-Indo-European word, *ghalgh-, for “branch or rod”.5 Although Old Norse gálgi is cognate with Old English g(e)alga, it is unlikely that the latter word was still used by the time Scots replaced Norn, the Norse language spoken in Orkney and Shetland. It is believed that Norn gradually died out between the 16th and 18th centuries in Shetland (Knooihuizen 2005, Ljosland 2012, Smith 1996:31–35). As a consequence of the language shift, the original meaning of many place-names was eventually lost to the locals. Place-names were transformed in various ways but direct translation of existing names was, if it happened at all, rare, and is most likely to have occurred in the 16th or 17th centuries (Nicolaisen 1983:73). Nevertheless, it is not clear why some sites obviously kept their old names, while others were renamed. The argument becomes even less convincing as both Golgo in Sandwick and the Knowe of Wilga in Cunningsburgh are associated with traditions of hangings, as we shall see later, so the alleged history of these sites was obviously not forgotten. Interestingly, Golgo bore an English name—Gallowtoon—at some point according to Stewart (1987:116), but apparently the older name was stronger than the new one in this case. This analysis leads to a third possible conclusion, namely that the gallow-type names simply refer to later places of execution than the gálgi-type names. If the interpretation of the latter names is indeed correct, the sites were apparently not in use or not considered relevant anymore by the time Scots became the dominant language. Similarly, O’Grady (2008:360) states that unchanged forms of gallow place-names in Scotland are probably post-medieval in date, but in some cases may reflect genuine execution sites. It needs to be noted that the “gallow” and the “golgo/wilga” names are not just different in terms of linguistics but also refer to different geographical entities. Of the latter group, only the Scord o Gulga, which is poorly documented and difficult to locate (Smith 2006), seems to refer to a hill, while the others are the names of a farm, an outcrop knoll on a hillside and a steeply bounded inlet respectively. As shown above, the “gallow” element only occurs with hills or mounds. Oral Traditions In a number of cases, the sinister history of the gallows sites is locally recalled in oral traditions. While folk stories and oral traditions that are attached to certain locales cannot be taken as evidence for historical events, they do give unique insights about the perception of these sites. The stories mostly deal with witches or thieves, who were eventually executed at a certain place. It is important to realize that oral traditions giving a popular explanation for certain place-names may have developed later and do not necessarily reflect the origin of the name. This is especially the case with the gallow-type names, as the name still evokes an association with hanging sites. It is interesting to note, however, that some of the “golgo/wilga”-type names are also connected with stories of executions. In those cases, the etymological origin of the place names is not evident to modern speakers, and the stories must thus be rather old. Traditions of executed witches are found in Dunrossness, Scalloway, Tingwall, and Fetlar. In Scalloway, locals still point out the former site of the stake Journal of the North Atlantic J. Coolen 2016 Special Volume 8 98 an event is very unlikely to have taken place as late as the 18th century, but the story may have a more ancient origin. The tale should perhaps be seen in the context of the numerous stories of ghosts, mythical creatures, and hidden bodies that evolve around the Windhouse. The site of the Windhouse is marked as an ancient chapel site and burial ground on the first edition OS map (Canmore ID 1251). Although there seems to be little supportive archaeological evidence, this historical function could explain the origin of some of the folk tales that surround the Windhouse. Folk tales about outlaws hiding in remote upland areas are well-known from other lands, such as Iceland (Sigurðsson 2006) where we also find a strong belief in trolls and other creatures. The name of K(a) il Hulter is very similar to that of Kit Huntling known from Orkney. Both names seem to be derived from the Old Norse kett-hyndla (cat-bitch), an evil, female hybrid creature, which—just like the outlaws— haunted inaccessible, marshy upland areas such as the Burn o’ Kithuntlins on Mainland Orkney (Towrie 1996–2014). The figure of the “ketthontla” in turn resembles the Icelandic monster Grýla (Jónasdóttir 2010:21). Hence, both the story itself and the name of its main character have parallels in other Norse countries. Traditions of dwarves, dragons, roaming lights, etc. inhabiting or appearing at prehistoric monuments or natural landmarks in remote wasteland areas are also known from other parts of Europe (Coolen 2013, with references). While their origin may lie in prehistoric beliefs about ancestral and natural spirits, the legendary creatures populating the common lands around settlements seem to have got a more demonic character under Christianization (Roymans 1995, Semple 1998, 2013:109–144). Examples from England (Reynolds 2009; Semple 2013), the Netherlands (Meurkens 2010) and Germany (Meier 1950) show that some of the places associated with supernatural creatures were indeed used as execution sites in the Middle Ages. It is doubtful whether the story of Kail Hulter has a kernel of truth. However, it is possible that a popular folk tale was connected to a historical place of execution. Archaeological Evidence No visible remains of gallows sites have hitherto been identified in Shetland with certainty. However, there are possible vestiges on the Gallow Hills of Fetlar and Tingwall. A remarkable enclosure on a small mound near the summit of the Gallow Hill in Fetlar has been interpreted as the former site of the gallows (Smith 2006). The enclosure consists of a bank or wall, which is up to ~0.4 m high and ~1 m on Gallow Hill, apparently a patch of burnt soil that never turns green.6 In Dunrossness and Fetlar, the traditions also refer to visible features. According to a note by John Stewart written in the early 1950s, a stone at the top of the Gallow Hill at Huesbreck in Dunrossness is known as Katie Cornie, named after a burnt witch.7 The folklorist James Laurenson from Fetlar described the executions of witches on the Gallow Hills of Fetlar and Scalloway in great detail and, as it seems, with lively imagination8. A story of 3 witches who were tried and executed at Tingwall was recalled by the storyteller Brucie Henderson from Yell in 1955.9 Stories of hanged thieves have been reported for the Knowe of Wilga, Gallows Knowe in Mid- Yell, and Gallow Hill at Walls. They all deal with livestock thieves, a common element in Shetland folk tales (Tulloch 2014:ch. 23). Indeed, the court records show that numerous people were accused of stealing sheep in early modern Shetland. However, the stories associated with gallows sites also contain more mystical elements that mark the alleged execution sites as liminal places on the verge of the human realm. The most interesting case is the story of K(a) il Hulter, a notorious sheep thief said to have been hanged at the Knowe of Wilga in Cunningsburgh.1. The story goes that Hulter was an outlaw who hid in a cavern near the Knowe of Wilga, the Thief’s Hole, where he managed to live undiscovered for many years as he only went about stealing sheep at night. When a boy accidentally came upon his hideout, the boy was taken captive and had to stay with Hulter for a long time. One Yule, the boy tricked his captor into going to his native farm to get butter and a cow and Hulter was captured at last. A very similar story surrounds the Thieves Knowe, a conical-shaped mound in the upland valley Skelladale near the Gallow Hill of Brae in Delting (Greig 1892:17f.). Instead of an outlaw hideout, it figures as the home of fairies and trows, who also had a fancy for fat sheep and cattle. The role of the small boy, who was taken hostage for years, is played by a fiddler, who was forced to play at the trowies’ dances but finally betrayed them in the same manner.11 The story of Kail Hulter also shows similarities to the Thief o da Neean, another sheep thief who allegedly lived in a cave near West Burrafirth but was eventually hanged at the Gallow Hill of Walls.12 Smith (2006) quotes the folklorist Laurence Williamson, who recalled a story of 2 men found guilty of stealing cattle from the lord of Windhouse. Because they refused to be sent to trial in Scalloway, they were hanged at the Gallows Knowe near Halsagarth on Yell. Williamson dated the story to the early 18th century. Smith has pointed out that such Journal of the North Atlantic J. Coolen 2016 Special Volume 8 99 B. Smith, 6 June 2011 pers. comm.). Sheepfolds in Shetland were usually round or oval shaped and located on the shoulder of a hill, along hill-dykes or at the coast, taking advantage of the landscape to funnel the sheep (Tait 2012:323f.). Since the enclosure on Gallow Hill lies in open terrain, it would have been very hard to drive sheep into it unless it had lead-fences. Moreover, the enclosure appears exceptionally large for a sheepfold. Given the limited number of well-documented gallows sites in Scotland (cf. O’Grady 2008:360ff.), it is difficult to say whether such sites were commonly surrounded by an enclosure. A rectangular enclosure of similar dimensions as the one in Fetlar is found on top of the Gallowhill of Ellon in Aberdeenshire and is also believed to mark the site of a gallows (Canmore ID 145367). However, no archaeological research has taken place yet that might provide evidence on the dating and function of that structure. While there seem to be no direct parallels to these enclosures on well-studied execution sites elsewhere in Europe, it was not unheard of at least in the early medieval period to fence off gallows sites, as recorded for example in Assen in the Netherlands (van der Sanden 2010:202f.). wide at the base (Fig. 2). It has a slightly trapezoidal outline with side lengths of 22–24 m on the north, west, and east and ~29 m on the south (Fig. 3). The bank is higher on the north side, which is also the highest part of the terrain, while it can only be vaguely observed at the deepest spot in the south. Smith (2006) noticed a possible entrance on the east side, but I observed no distinct opening while examining the site. There is a strong local tradition that the enclosure represents an old execution site. Quoting James Laurenson, Robertson and Graham (1964:51) stated that “the hole in the Gallow Hill where the gallows stood can still be seen.” However, according to Laurenson, the stones that framed the hole were later moved by poaching boys.13 At present, there is no convincing alternative interpretation of the structure on Gallow Hill. Traces of peat cutting can be found along the cliffs to the west of the enclosure but it is not obvious what role the enclosure could have played in peat-cutting activities. On the first edition of the OS 6-inches-to-the-mile map (Sheet XVII) published in 1882, the enclosure is marked as an “old sheepfold”. However, this is an unlikely interpretation (I. Tait, Shetland Museum and Archives, Lerwick, UK, and Figure 2. Enclosure on Gallow Hill, Fetlar (Shetland), which allegedly marks the site of the gallows. View towards the north. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 100 A rectangular, slightly trapezoidal structure of approximately 8 m by 6 m exists on the edge of Gallow Hill in Tingwall overlooking the valley (Coolen and Doneus 2014:27). It consists of a single row of rather large boulders, which may have been taken from a tentative burial cairn nearby. The enclosure is oriented WNW/ESE and seems to be open towards the moorland in the west. Both the enclosure and the derelict cairn are located on the outside of a supposedly medieval hill dyke. At present, it is difficult to interpret the structure. An agricultural or pastoralist function, e.g., as an animal pen, is conceivable but the possibility that it marks the site of a gallows should not be excluded. A local amateur archaeologist reports to have found pieces of charcoal at 15 cm depth in two small “test pits” inside the enclosure. 14 As mentioned above, people from Scalloway have observed a low knoll of reddish, apparently burnt soil on Gallow Hill above the town.15 However, this is likely to be a popular interpretation of a natural feature as it is improbable that a stake would eliminate the vegetation for several centuries unless larger quantities of chemicals were present in the soil. While a closer investigation of the enclosures in Fetlar and Tingwall seems worthwhile, none of the tentative gallows sites in Shetland has yielded conclusive archaeological evidence for an execution site. Of course, the absence of evidence is no evidence of absence. But for the moment at least, archaeology in its traditional sense cannot answer the posed research questions. Visibility Throughout medieval and early modern Europe, gallows were generally placed on highly visible locations to enhance the supposed deterrent effect as well as to display the power of authority (Coolen 2013). Indeed, Smith (2006) has highlighted this common element for the alleged gallows sites in Shetland. The presence of historical lookout posts on the Gallow Hills of Eshaness and Unst and on a hill near to the Gallow Hill of Walls underlines the good visibility and strategic location of the sites. Due to its topography, even local traffic in Shetland mainly took place by boat until recently. As the economy of Shetland was traditionally based on subsistence farming and fishing (Tait 2012), settlement was dispersed and concentrated along the coast. Therefore, it can be assumed that if the gallows were meant to be seen by travellers, they should have been visible from the coast. Figure 3. Satellite image showing the enclosure on Gallow Hill, Fetlar, Shetland. Source: Google Earth®. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 101 While the viewsheds do confirm the good visibility of most sites, they also reveal large differences. The total area that can be seen from each site varies from almost 3000 km² to less than 40 km² (Fig. 4). The most visible site is the Knowe of Wilga, while— surprisingly—the Gallow Hill of Scalloway is least visible. The huge difference in the visible area is mainly caused by the visibility of the open sea. As from most places in Shetland, the sea is visible from all sites. In fact, sea surfaces represent more than half of the visible area in all cases; in 10 out of 12 cases, the sea even accounts for more than 90% of the visible area. However, these numbers are somewhat misleading. Since the surface, which is visible within a certain angle, increases towards the horizon, even a small strip of the sea visible at the horizon represents a large area. Moreover, due to the curvature of the earth the distance to the horizon from higher elevated points is much larger than from a point close to sea level, which again leads to an exponential increase of the visible area towards the horizon. The total land surface visible from each site var - ies from 134 to 16 km². Again, the Gallow Hill of In order to quantify and compare the visibility of the sites, viewshed models were calculated for all of the discussed sites, using a digital terrain model (DTM) with 50-m horizontal and 1-m vertical resolution in a GIS (Coolen 2012). In each case, the observer point was set at the highest point of the respective hill with an offset of 2 m, except in Fetlar, where the viewshed represents the view from the enclosure. In some cases, the viewshed was found to be more realistic when a cumulative viewshed for several points in adjacent grid cells was calculated to reduce the influence of the gridded DTM. As we do not know the exact location of any of the gallows— except perhaps in Fetlar—the viewshed models present a best-case scenario. Besides this major flaw, we also do not have sufficient knowledge of the late medieval and early modern settlement of the surroundings of most of the sites to put the gallows in a contemporary landscape context. Therefore, the scope of the visibility analysis is rather to make an optimistic statement and classify the setting of the tentative gallows sites especially in terms of visibility from the sea. Figure 4. Diagram of the total land and sea area visible from the discussed sites. The doubtful or poorly localized sites in Mid-Yell, Whalsay, and North Roe are not included. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 102 restricted. Again, Scalloway is close to the bottom of the list, with a visible area of 6.9 km². The limited-range viewshed indicates from which areas (and consequently to whom) the gallows were meant to be visible. The visible sea surface (including inlets) in the direct surroundings greatly varies. From the Gallow Hill of Tingwall, the sea cannot be spotted within 3 km at all. On the other hand, 70% of the surface visible within 3 km from the Knowe of Wilga is covered by the sea. Apart from the latter, there are only 3 sites that are focused more towards the sea than towards land in the close range, namely the Gallow Hills of Fetlar and Walls and Watch Hill in Northmavine. The contrast between the sites is illustrated by the Gallow Hills of Fetlar and Scalloway. The former overlooks the main settlements along the south coast of the island, but the location of the enclosure suggests that it was also meant to be visible to ship traffic passing through Colgrave Sound or entering the Wick of Tresta (Fig. 6). Hence, the location seems to be directed towards the outside world. Scalloway has the lowest score, but the ranking of the other sites is completely different than the ranking in terms of overall visibility. The doubtful site of Gluss overlooks the largest land surface, followed by the Gallow Hills of Brae and Fetlar. Although it is astonishing how far one can see from most sites, it is of course unrealistic that the gallows and their involuntary companions would have been visible with the naked eye from 40 km distance. We do not know what the gallows in Shetland looked like but they are unlikely to have been monumental enough to be clearly discernible from more than 3 km away. Hence, the visible area and the visible land and sea ratio within a radius of 3 km were calculated for each site, corresponding to a maximum area of 28.3 km². It must be stressed that the inaccuracy of the viewshed caused by the resolution of the DTM plays a lar ger role at this scale. The total visible area within 3 km from each site varies from 17.3 to 6.6 km² (Fig. 5). Muckle Heog offers the most complete view of the surroundings, while the view from the Gallow Hill of Brae is most Figure 5. Diagram of the total visible land and sea area within a radius of 3 km from the discussed sites. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 103 On the other hand, the limited visibility of the Gallow Hill of Scalloway suggests that it was primarily meant to be seen from the village, the castle, and the harbor (Fig. 7). Hence, the message of the site seems to have been directed towards residents of Scalloway rather than to approaching outsiders. According to the viewshed analyses, wide visibility from a vast area or from the open sea is not a general characteristic of the tentative gallows sites. It is, however, notable that most of the sites are located near good anchorages or portages and are well visible from there. For example, the Gallow Hill at Brae offers a splendid view of Busta Voe and Mavis Grind, a narrow portage between the North Sea and the Atlantic. The Gallow Hill of Walls is relatively far from the open sea but overlooks a major part of the sheltered bay of Gruting Voe, an excellent natural harbor. The proximity to natural harbors and portages can be similarly observed for other sites with “gallow”-names and “golgo/wilga”-type names, and even applies to the rather unconvincing sites Muckle Heog (overlooking Baltasound) and the Wilgi Geos (upon the mouth of Sand Voe). While this characteristic should not be taken as evidence for the actual use of individual sites, it does give an indication on the possible context and focus of the tentative gallows sites and allows for a redefinition of the visibility factor. Logically, natural harbors that offer shelter from the open sea and portages that cut off dangerous and long passages around the islands also formed focal points for settlements. Although there is not sufficient data available to reconstruct the development and status of Figure 6. Viewshed from the enclosure on Gallow Hill, Fetlar, Shetland. DTM: Ordnance Survey. Courtesy of Crown copyright/database right 2010. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 104 between being visible from afar and dominating a local harbor and potential settlement. Gallows Sites and Medieval Administrative Units Obviously, official places of execution are always associated with a court that holds the right to inflict capital punishment. Indeed, erecting a gallows was a powerful way to demonstrate judicial authority in medieval and early modern Europe (Coolen 2013). Gallows could be makeshift structures that were constructed for the event of one execution, sometimes close to the scene of the crime (cf. Poole 2008:164f.). However, most judicial districts that held high justice had one dedicated place of execution, which often also served as gibbet, although some larger towns had separate sites for various execution methods (Maisel 193:28–31). The association of gallow hills and other execution sites with medieval moot sites and hundred boundaries has also been observed in other parts of Scotland (O’Grady 2008:360ff.) and England (Reynolds 2009). The proximity of the Gallow Hill in Tingwall to the traditional site of the law ting and the tentative succession of the Gallow Hill in Scalloway, where the court was later held, gives rise to the question whether the tentative gallows sites in Shetland also reflect individual court districts. And if so, from what period? The political and judicial organization of Shetland in the Norse period has been the topic of debate for at least a century (Clouston 1914:429ff., Donaldson 1958:130ff., Sanmark 2013, Smith 2009). It has long been recognized that he tingelement in the name of some Shetland parishes indicates a Norse origin. Fellows-Jensen (1996:26) has suggested that the thing parishes of the Figure 7. Viewshed from the tentative gallows site on Gallow Hill, Scalloway, Shetland. DTM: Ordnance Survey. Courtesy of Crown copyright/database right 2010. these harbors and settlements throughout the Middle Ages and early modern period, it seems reasonable to assume that the anchorages marked on early modern admiralty charts were also used in the medieval period. The visibility analysis implies that most of the gallows sites were visible to sailors approaching or entering harbors as well as the people that lived in the associated settlement. The location of the gallows sites often seems to have offered a compromise Journal of the North Atlantic J. Coolen 2016 Special Volume 8 105 Northern Islands date back to the 9th century, while Smith (2009) has argued that they are more likely the result of a later reorganization reflecting the increased royal control in the 13th century. Sanmark (2013:105f.) sees the location of the thing sites as supportive evidence for the latter theory but stresses that the tentative reorganization was not necessarily a complete clean sweep. It has been noted before that the ting-parishes are all located in the central part of Mainland, which lacks clear geographical boundaries (Donaldson 1958:130, Smith 2009:41f.). Delting, Nesting, Lunnasting, Aithsting, Sandsting, and Tingwall are still existing parishes, while Þvæitaþing and Rauðarþing, two districts mentioned in a document of 1321, are more difficult to locate. It is generally assumed that Þvæitaþing covered the present parish of Walls (Sanmark 2013:101f.). Rauðarþing has been located in Northmavine (Jakobsen 1897:102, Smith 2009:42), but Sanmark (2013:102) has argued that it may have covered the central part of Yell. The name Neipnating, mentioned in 16th- and 17th-century documents, probably refers to Nesting (Smith 2009:41). Based on place-names and topographical considerations, Sanmark (2013) has recently ventured to locate the associated assembly sites (Fig. 8). Gallows-names occur in Tingwall, Walls (Þvæitaþing?), D e l t i n g , a n d N o r t h m a v i n e (Rauðarþing?), while no such names are known in Sandsting, Aithsting, Nesting, and Lunnasting. The distance from the Gallow Hills to the alleged assembly sites is ~2 km in Walls and 3.5 km in Delting. Intervisibility is not given in either case. The Gallows Knowe at Halsagarth near Mid Yell, ~2.5 km from the tentative assembly site at Gardiestaing (Sanmark 2013:102) might support Sanmark’s localization of Rauðarþing, but the evidence of this particular site is very weak. Neither can any of the supposedly older “golgo/wilga” names be linked to the tentative thing-districts. The available sources do not allow for a reconstruction of the Norse thing-district boundaries. Yet, if we assume that they more or less corresponded to the later parish boundaries, only the Gallow Hills of Figure 8. Distribution of gallows place-names in respect to medieval thing districts and tentative assembly sites in Shetland (after Sanmark 2013). Districts: 1 Tingwall; 2 Sandsting; 3 Þvæitaþing; 4 Aithsting; 5 Nesting; 6 Delting; 7 Lunnasting; 8 Rauðarþing? (after Smith 2009); and 9 Rauðarþing? (after Sanmark 2013). Journal of the North Atlantic J. Coolen 2016 Special Volume 8 106 Walls and Delting seem to be located near the tentative boundary. Smith (2006) has pointed out that gallowsnames occur not far from a “Herra” place-name in Tingwall, on Fetlar and Yell. The name Herra (from Old Norse hérað, county or district) occurs 4 times in Shetland and refers to small districts or groups of farms (Jakobsen 1897:117, Smith 2009:43). It is believed that the name reflects another, possibly older type of Norse administrative unit (Sanmark 2013:99, Smith 2009:43). However, the judicial authority of the hérað districts in Shetland is not known, and there is no evidence that they had an assembly, let alone the authority to inflict capital punishment (Sanmark 2013:55). In short, it is hard to link the gallows-names to what we know about the Norse administrative units. The alleged gallows sites are neither equally distributed over the thing-districts nor regularly associated with the assembly sites or tentative parish boundaries. Indeed, as it has been argued that the gallow-names may be post-medieval, we also need to look at the later administrative or ganization. Gallows Sites and Post-Medieval Administrative Units The administrative division in the late 16th and early 17th century can be reconstructed on the basis of the court books (Barclay 1962, 1967; Donaldson 1954, 1991) and other legal documents (Balfour 1854:13–92). Most districts comprised several parishes or isles. The post-medieval administrative division seems to be a reorganization of older administrative units, as some of the districts correspond to older thing-districts. As the grouping of the parishes varies between the documents, the number of districts ranges from 9 to 13 (cf. Donaldson 1958:130ff.). Based on the historical documents, we can reconstruct at least 10 districts: • Nesting, Lunnasting, Whalsay, and Skerries. The Skerries are listed together with Fetlar, Unst, and Yell in 1576. • Fetlar • Unst • Yell • Northmavine • Delting; listed as Delting and Scatsta in 1576. • Walls, Sandness, Aithsting, and Sandsting. Walls is listed separately in 1576. The district also included Papa Stour. • Burra and Gulberwick. The district also included Quarff and Trondra but they are not always listed. Burra and Gulberwick were grouped together with Tingwall, Whiteness, Weisdale, and Bressay in 1603, and with Bressay only in 1604. • Dunrossness • Tingwall, Whiteness, Weisdale, and Bressay. Bressay is listed separately in 1576, while it is grouped together with Burra in the court book of 1604. The distribution of the gallows-names shows a remarkable correspondence with the above division (Fig. 9). Apart from the smallest district of Burra and Gulberwick, there is at least one gallows- name in each district. With the exception of Yell, every district has one Gallow Hill. The parish of Tingwall hosts 2 Gallow Hills but, as has been pointed out before, these are most likely successive sites. Furthermore, it is striking that most of the tentative gallows sites are located rather centrally within the districts. On Mainland, the gallows-names do not occur along the district boundaries. However, in some cases they do occur near parish boundaries within a district. The Gallow Hill of Walls is located near the boundary between Sandsting and Walls; the Gallow Hill in Tingwall is sited near the boundary between Tingwall and Whiteness. Apart from Scalloway, there is no obvious relation between the Gallow Hills and the venues of the district courts held in the beginning of the 17th century (Fig. 10). Some of the gallows sites are relatively close to a court site, as in Unst (2.3 km) and Delting (2.9 km), but most of them are not. This inconsistency is hardly surprising as the district courts do not seem to have dealt with capital cases. As mentioned before, the only known exception is the conviction of Christopher Johnston at Wethersta in Delting in 1602 (Donaldson 1954:18f.). The Gallow Hill of Brae is visible from Wethersta, but it is not clear whether Johnston was indeed hanged there. The local courts of the early 17th century were usually held at the manors of the fouds (bailiffs) or other officials. Although these manors were mostly located in or nearby major settlements, they do not necessarily represent the old parish center (Smith 2009:43), let alone older assembly or court sites. To summarize the above: the distribution of the gallow-type names corresponds well with the court districts that can be reconstructed based on late 16thand 17th-century documents. However, at least after AD 1602, district courts seem not to have been involved in capital cases anymore. The post-medieval administrative division is obviously the result of a re-organization of older districts, including the thing-districts discussed above. It is difficult to reconstruct when and how this reorganization took place, but the striking correlation of the gallows sites and the post-medieval parish division implies that the gallows sites can be linked to this late-medieval Journal of the North Atlantic J. Coolen 2016 Special Volume 8 107 or post-medieval restructuring. There is, however, one strain of evidence that sheds some light on the parish division between the Norse period and the reign of the Stewarts, namely the ecclesiastical parish division. Gallows Sites and Ecclesiastical Parishes Ever since the establishment of the diocese of Orkney and Shetland in the mid-11th century (Cant 1996:161), the bishops of Orkney and archdeacons of Shetland had close ties to the earls and kings who ruled the islands (Smith 2003). The ecclesiastical division of the islands was strongly connected with the administrative and judicial organization. A list of ecclesiastical “benefices” compiled by Rev. James Pitcairn between 1579 and 1612 (Goudie 1904:155–158) and John Brand’s (1701:125–146) description of ecclesiastical parishes largely correspond to the parishes mentioned in the 17th-century court books. The bishops did not have as much judicial power in Shetland as in Orkney, but some parts of Shetland fell directly under the jurisdiction of the church before Robert Stewart exchanged his abbacy in Edinburgh for the episcopal domains in Orkney and Shetland in 1568 (Barclay 1967:XVIIIff.). Indeed, Sanmark (2013:104f.) considers the proximity to churches and chapels “one of the most striking features of the Shetland thing sites.” Cant (1996) made an attempt to reconstruct the medieval church order of Shetland, which he believed to have largely developed in the 12th century. In its final phase, Shetland comprised 14 priest districts, subdivided into a total of 30 parishes (Fig. 11). Most districts, except Lunnasting, Aithsting, Sandsting, Bressay, and Burra and Gulberwick (the latter not being explicitly named as separate church district by Cant) have one Gallow Hill. Eleven of the 15 gallows-place-names occur less than 3 km from a pre- Reformation parish church, the shortest distance being 550 m in Sandwick. This applies both to the Gallow Hills as well as the Golgo/Wilga-names. In the parish of Dunrossness, we find a tentative gallows site near each of the 3 parish churches. In Northmavine, Figure 9. Distribution of gallows place-names in respect to early modern court districts. Parishes: 1 Dunrossness; 2 Burra and Gulberwick; 3 Tingwall, Whiteness, Weisdale, and Bressay; 4 Walls, Aithsting, and Sandsting; 5 Nesting, Lunnasting, and Whalsay; 6 Delting; 7 Northmavine; 8 Yell; 9 Fetlar; and 10 Unst. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 108 we also find the same number of parish churches and gallows-names, but the distances are somewhat larger and 2 out of 3 gallows-sites are rather doubtful. In some cases, the gallows-names occur near a tentative parish boundary, as in Cunningsburgh, Walls, and Delting. The same applies to Gulga in Gluss, located between St. Olaf’s in Ollaberry and St. Magnus’ (?) church in Hillswick, as well as to Muckle Heog on Unst and Gallows Knowe in Mid-Yell, although the latter are again doubtful gallows sites. The spatial correlation of gallows-place names and medieval parish churches does not necessarily imply that these were directly associated. However, as has been pointed out above, there was a strong link between the centers of ecclesiastical and secular power and jurisdiction in medieval Shetland. Therefore, the pre-Reformation parish churches may give a better indication of the political and judicial centers of the post-Norse parishes than the courts held at private manors in the 17th century. Gallows and Prehistoric Monuments One of the common features of the potential gallows-sites, which Smith (2006) identified in his initial paper, is the presence of prehistoric monuments. Most of these monuments can be identified as Neolithic and Bronze Age burial cairns. Eight out of the 15 tentative gallows sites are hilltops or hillocks with a cairn at or near the summit (Fig. 12). However, the association of these monuments and possible execution sites is not conclusive as long as the latter cannot be located with certainty. The re-use of prehistoric monuments, especially barrows, as execution sites is a widespread phenomenon in northwestern Europe (Meurkens 2010, O’Grady 2008:361f., Reynolds 2009, Whyte 2003). Under Christianization, ancient monuments became demonized, and the execution and abandonment of criminals at “heathen”, liminal places enforced their expulsion from the community. The demonization is reflected by traditions of mystical creatures, which often surround these sites. As shown above, traces of such stories are also found in Shetland. Taking this phenomenon a bit further, it even seems conceivable that in some cases at least, a prehistoric monument became Figure 10. Distribution of gallows-place names in respect to court sites mentioned in the courtbooks between 1600 and 1615. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 109 considered a gallows site by popular belief, which then resulted in the place-name. For example, this could be true for the Knowe of Wilga in Cunningsburgh. As has been shown above, the story of the outlaw Kail Hulter, who is said to have been hanged there, shows obvious similarities to stories of superhuman creatures that lived in suspicious-looking mounds in the highlands. The Knowe of Wilga, a remarkable outcrop knoll on a steep hillside, hosts the remains of a large chambered cairn (Canmore ID 934). It is possible that the traditional belief of a “trowies’ home” was simply translated into a contemporary context and that the mystical creatures, who lived and possibly died at the site, later became a human sheep thief. However, the above explanation can hardly account for all gallows-place names in Shetland, especially given their equal distribution over the islands. Moreover, the association of prehistoric monuments and later execution sites is less obvious when both occur on a large, natural hill than in the case of explicitly named barrows. The main problem in Shetland is that no execution site can thus far be localized with certainty. To assume that they may have been located on or near the prehistoric monuments would therefore be an obvious circular argument. The evidence for a gallows at Scord o Gulga in Gluss (Northmavine), for example, is very weak, and the identification of Gulga is partially based on the presence of a prehistoric cairn (Smith 2006). Moreover, it should be taken into account that Shetland is strewn with well-visible, prehistoric structures. Arguably, there are fewer hills without such structures than those with visible prehistoric remains. The alleged execution site of Sandwick is the only one that is associated with a standing stone. The stone itself does not survive as it was destroyed in the 19th century to provide building material (Smith 2006). It is also the only site where a local tradition explicitly describes the use of a possibly prehistoric monument as a gibbet. A similar case is found in Aberdeenshire. The Hanging Stone on the Gallow Hill near Rosehearty, a tall standing stone now included in a field wall, is also said to have been used to hang people (Canmore ID 20791).16 Throughout Great- Britain and Ireland, there are numerous standing stones called Hanging or Hangman’s stone (Crawford 1922, Grinsell 1985). Figure 11. Distribution of gallows-place names in respect to pre-Reformation parish churches and priest districts (after Cant 1996). Journal of the North Atlantic J. Coolen 2016 Special Volume 8 110 The name either refers to their declination or indeed to a place of execution or gibbet site. In many cases, there is a local tradition of a thief, who accidentally strangled himself on the stone as he tipped over it while carrying a stolen sheep or deer. As Crawford noted, the story probably refers to the gallows, which ended the life of many a (sheep) thief. Some of the Hanging Stones are obviously too small to have been used as a gibbet and even the larger standing stones may not have lent themselves too well for this purpose. It is more likely that the stones present old boundary markers. All across Europe, medieval execution sites were usually located at judicial boundaries (Coolen 2013). Their spatial association with ancient and distinct boundary markers such as standing stones or burial mounds served to confirm ancient boundaries and reinforce the authority of those who controlled them. Moreover, boundary markers were ultimately liminal places and therefore symbolically underlined the condemnation of those who were hanged in their vicinity. However, it is doubtful whether this was the case in Sandwick. As discussed above, Golgo lies very close to the parish church and settlement of Sandwick, and it can hardly be considered a liminal place. Unfortunately, the poor evidence of the execution site, the standing stone, and the judicial status of Sandwick in the Middle Ages makes further interpretation difficult. The apparent association of gallows-place names with prehistoric monuments might reflect a phenomenon that is also known from other parts of Britain and northwestern Europe. However, given the poor identification of possible gallows sites and the abundance of well-visible prehistoric monuments in the Shetland uplands, the nature of this association is speculative. Concluding Discussion It is difficult to draw definite conclusions about the origin of the alleged gallows sites in Shetland. In most cases, the former presence of gallows is only indicated by place-names and oral tradition. Only 1 gallows site, the Gallow Hill of Scalloway, is documented in historical records. It is likely that the establishment of an execution site in Scalloway coincides with the transfer of Shetland’s lawthing from Tingwall to nearby Scalloway, which seems to have taken place shortly after the rise of Robert Stewart in the 1570s. This scenario also implies that the Gallow Hill of Tingwall may have been in use until then. As for the other sites, the fact that Figure 12. Chambered cairn on the Knowe of Wilga, Cunningsburgh, Shetland. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 111 they are not mentioned in the court records and that capital cases were handed over to the lawthing at least from the 17th century onwards implies that they must be older. In the case of the Gallow Hills, there is little reason to doubt that they were designated as places of execution at some point. This assertion does not necessarily mean that they were frequently used, nor that they must have hosted gallows on a permanent basis. Apart from their “practical” function, gallows served to demonstrate judicial authority and autonomy, and there are examples of early modern gallows that were never or hardly ever used (Mol 2007:294f.). On the other hand, the possibility should be taken into account that executions took place near the scene of crimes, though it seems doubtful whether such an event would lead to the re-naming of prominent landscape elements. The regular distribution of Gallow Hills over the postmedieval court districts also implies that this pattern is the result of a more or ganized process. The gallows sites thus document a system of local justice, in which each district apparently had the authority to impose capital punishment. The question is when this system arose. Smith (2006) has suggested that the gallows sites date to the early Norse period. This hypothesis can be questioned for a number of reasons. Firstly, the name Gallow Hill, which occurs at least 7 times, is obviously younger than the vast majority of place-names in Shetland, which have a Norse origin. The argument that the gallow-type place-names may have been translated is hardly convincing without further evidence and in any case would suggest that the respective sites were (still) in use or known by the time Scottish English replaced Norn. Secondly, no clear relationship between the alleged places of execution and the presumed medieval assembly sites (with the exception of Tingwall) or the division of the ting-districts could be identified. By contrast, there appears to be a stronger association with pre-Reformation parish churches and the early modern administrative division. Thirdly, Norse assemblies had limited executive power, and dedicated gallows sites generally seem to have appeared late in Scandinavia (F. Iversen, Museum of Cultural History, Oslo, Norway, 22 Novemeber 2013 pers. comm.). The administrative and ecclesiastical districts, which are mentioned in early modern documents, mostly comprise several parishes, including the tentatively older ting-parishes. Clouston (1914:429–432) interpreted the division of the districts as an echo of a typical Norse division into quarters and thirds and reconstructed 12 things, each of which had 1 “lawrikman”. This interpretation, however, has not found general acceptance (Donaldson 1958:130–132, Smith 2009:42f.). As shown above, the gallow-type names are equally distributed over the early modern judicial districts. If Smith’s early dating of the gallows sites is correct, this observation could support Clouston’s model that several parishes were subordinate to 1 court. Alternatively, it is possible that the grouping of multiple parishes into districts was the result of a late medieval or early modern reorganization. It seems logical to suggest that such tentative reorganization could have taken place after the annexation of Shetland with Scotland in 1469 (Crawford 1983). It has hitherto been believed that the official annexation of the Northern Isles with Scotland, which was obviously part of a much longer process of increasing “Scottification”, did not bring about major changes to the polity in Shetland. This idea is mainly based on the fact that late 16th- and early 17th-century legal documents still reveal many elements of Norse administration (Goudie 1892) and that the local book of laws, based on the Lawmender’s code, was only officially abolished in 1611. The distribution of the gallows sites might contradict this traditional view and shed some light on the judicial organization in the poorly documented period between the annexation and the rise of the Stewarts. Nevertheless, the gálgi-type place-names indicate that gallows may have been present at an earlier stage. The distribution of the place-names in Dunrossness and Northmavine could even be taken as evidence that every parish once had an execution site, possibly before the parishes were joined into larger districts. However, given the small number of gálgi-type names and the heterogeneous nature of the sites they refer to, this interpretation goes beyond the constraints of the evidence. Similar as the “hanging”-element, which is found for example in the name Hanger Heog, gálgi-type names could refer to a steep or overhanging landscape element. Indeed, this seems to be the case at Wilgi Geos and might also apply to the Knowe of Wilga (similar both in terms of topography and archaeology to Hanger Heog) and the vanished standing stone at Golgo. In Iceland, gálgi- and gálga-names are numerous and often seem to refer to natural features rather than historical execution sites (A. Friðriksson, Institute of Archaeology, Reykjavík, Iceland, 10 May 2013 pers. comm.). The oral traditions of witch burnings, which are associated with the Gallow Hills of Scalloway, Tingwall, Fetlar, and Dunrossness, could also be considered indicative for a relatively late date for these sites. Although sorcery was considered a serious crime in the Norse laws, the first documented witch trial in Shetland took place in 1602 (Willumsen 2013:168). In mainland Scotland, persecution Journal of the North Atlantic J. Coolen 2016 Special Volume 8 112 of witches started some decades earlier (ibid.:63f.). In general, the late 16th and early 17th century marked the peak phase of European witch hunts. Compared to mainland Scotland or other European regions, persecution of witches was infrequent in Shetland, with 31 documented cases in the period 1602–1725. Of the accused witches, only 6 (or 7, including Barbara Tulloch) are known to have been executed (ibid.:168–172). Even though the traditions may not be based on historic events, they do reflect a popular identification of post-medieval execution sites. A number of gallows sites in Shetland are associated with stories of legendary thieves. Indeed, thievery seems to have been a more common crime in the medieval and early modern period, which was punished with death by hanging in severe cases. However, it has been shown that the stories of thieves that evolve around alleged gallows sites in Shetland are related to fairy beliefs rather than historical persons. Based on the current evidence, the association of places of execution with prehistoric (sepulchral) monuments in Shetland can neither be proved nor discounted. Given the numerous examples of execution sites on prehistoric barrows across northwestern Europe, it would not be surprising if this form of re-use of ancient monuments was also practiced in Shetland. As shown above, the choice of such locations reflects a symbolic punishment beyond death and stresses the liminality and marginality of the condemned. Rituals and places of execution thus also expressed and reinforced the social order and spatial cosmology of the society and helped to consolidate group identities. In this respect, the different message conveyed by the “outward” facing Gallow Hills on Unst and Fetlar on the one hand and the Gallow Hill in Scalloway, directly overlooking the village, is noteworthy. Acknowledgments This survey was made possible by a grant of the Austrian Ministry of Science and Research (BMWF) and the support of The Assembly Project – Meeting places in Northern Europe (AD 800–1500). I am very grateful to Brian Smith for his support and discussion of the gallowssites. Furthermore, I would like to thank Eileen Brooke- Freeman and Lauren Doughton of Shetland Amenity Trust for information on place-names. Finally, I would like to thank the 2 anonymous reviewers and guest editor for their valuable comments on the manuscript. Literature Cited Anderson, P.D. 1982. Robert Stewart, Earl of Orkney, Lord of Shetland, 1533–1593. John Donald, Edinburgh, UK. 245 pp. Anderson, P.D. 1996. Earl William to Earl Patrick: A survey of the history of Orkney and Shetland from 1468 to 1615. Pp. 174–185, In D.J. Waugh and B. Smith (Eds.). Shetland’s Northern Links: Language and History. Scottish Society for Northern Studies, Edinburgh, UK. 254 pp. Balfour, D. 1854. Oppressions of the Sixteenth Century in the Islands of Orkney and Zetland. From Original Documents. Maitland Club Publications 77/Abbotsford Club Publications 31. Edinburgh, UK. 129 pp. Barclay, R.S. 1962. The Court Book of Orkney and Shetland 1612–1613. Kirkwall Press, Kirkwall, UK. 104 pp. Barclay, R.S. 1967. The Court Books of Orkney and Shetland 1614–1615. Scottish History Society 4th Series, Vol. 4. Scottish History Society, Edinburgh, UK. 197 pp. Bell, A. 2010. Þingvellir: Archaeology of the Althing. M.A. Thesis. University of Iceland, Reykjavík, Iceland. Available online at http://hdl.handle. net/1946/6937. Last accessed 28 March 2016. Brand, J. 1701 [reprinted 1883]. A Brief Description of Orkney, Zetland, Pightland-Firth, and Caithness. William Brown, Edinburgh, UK. 268 pp. Cant, R.G. 1996. The medieval church in Shetland: Organisation and buildings. Pp. 159–173, In D.J. Waugh and B. Smith (Eds.). Shetland’s Northern Links: Language and History. Scottish Society for Northern Studies, Edinburgh, UK. 254 pp. Clouston, J.S. 1914. Records of the Earldom of Orkney 1299–1614. Scottish History Society, Edinburgh, UK. 515 pp. Coolen, J. 2012. A survey of Shetland’s Gallow Hills. TAP Field Report 5. Available online at http://www.khm. uio.no/english/research/projects/assembly-project/ pdf/field-report-5.pdf. Accessed 24 June 2014. Coolen, J. 2013. Places of justice and awe: The topography of gibbets and gallows in medieval and early modern northwestern and central Europe. Pp. 762–779, In A. Reynolds and K. Smith (Eds.). Archaeology of Legal Culture. World Archaeology 45/5. Routledge, Oxon, UK. 144 pp. Coolen, J., and M. Doneus. 2014. Airborne laser scanning. Pp. 19–30, In J. Coolen and N. Mehler (Eds.). Excavations and Surveys at the Law Ting Holm, Tingwall, Shetland: An Iron Age Settlement and Medieval Assembly Site. BAR British Series 592. Archaeopress, Oxford, UK. 137 pp. Coolen, J., and N. Mehler. 2014. Excavations and Surveys at the Law Ting Holm, Tingwall, Shetland: An Iron Age Settlement and Medieval Assembly Site. BAR British Series 592. Archaeopress, Oxford, UK. 137 pp. Cowie, R. 1879. Shetland: Descriptive and Historical and Topographical Description of that Country. Third Edition. Lewis Smith and Son, Aberdeen, UK. 227 pp. Crawford, B. 1983. The pledging of the islands in 1469: The historical background. Pp. 32–48, In D.J. Withrington (Ed.). Shetland and the Outside World 1469–1969. Aberdeen University Studies Series 157. Oxford University Press, Oxford, UK. 246 pp. Crawford, O.G.S. 1922. Hangman’s stones. Notes and Queries S12-XI (222):50–52. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 113 Donaldson, G. 1954. The Court Book of Shetland 1602– 1604. Scottish Record Society Edinburgh, UK. 176 pp. Donaldson, G. 1958. Shetland Life under Earl Patrick. Oliver and Boyd, Edinburgh, UK. 149 pp. Donaldson, G. 1991. The Court Book of Shetland 1615– 1629. Shetland Library, Lerwick, UK. 216 pp. Dunbar, W.H., G. Dingwall Fordyce, and J. de Maria. 1839. Reports of cases decided in the supreme courts of Scotland and in the House of Lords on appeal from Scotland, being a continuation of the Scottish Jurist Vol. XI. Anderson, Edinburgh, UK. 614 pp. Evans, R. 1996. Rituals of Retribution: Capital Punishment in Germany 1600–1987. Oxford University Press, Oxford, UK. 1014 pp. Fellows-Jensen, G. 1996. Tingwall: The significance of the name. Pp. 16-29, In D.J. Waugh and B. Smith (Eds.). Shetland’s Northern Links: Language and History. Scottish Society for Northern Studies, Edinburgh, UK. 254 pp. Goudie, G. 1892. The fouds, lawrightmen, and ranselman of Shetland parishes. Proceedings of the Society of Antiquaries of Scotland 26:189–212. Goudie, G. 1904. The Celtic and Scandinavian Antiquities of Shetland. Blackwood and Sons, Edinburgh, UK. 305 pp. Greig, P.W. 1892. Annals of a Shetland Parish: Delting. C. and A. Sandison, Lerwick, UK. 100 pp. Grinsell 1985. Hangman’s stones and their traditions. Folklore 96(2):217–222. Hibbert, S. 1822. A Description of the Shetland Islands, Comprising an Account of their Geology, Scenery, Antiquities, and Superstitions. Archibald Constable and Co, Edinburgh, UK. 616 pp. Jakobsen, J. 1897. The Dialect and Place-names of Shetland: Two Popular Lectures. T. and J. Manson, Lerwick, UK. 125 pp. Jónasdóttir, E. 2010. Grýla the mother and the murderer: Analization and discussion of Grýla, trolls, and similar creatures in folk and fairy tales. B.A. Thesis. University of Iceland, Reykjavík, Iceland. Available online at http://skemman.is/handle/1946/5611. Last accessed 28 March 2016. Kilday, A. 2007. Women and violent crime in Enlightenment Scotland. Royal Historical Society Studies in History New Series. Boydell, Woodbridge, UK. 183 pp. Knooihuizen, R. 2005.The Norn-to-Scots language shift: Another look at socio-historical evidence. Northern Studies 39:105–17. Larner, C., C.H. Lee, and H. McLachlan (Eds.). 1977. A Source-Book of Scottisch Witchcraft. University of Glasgow, Glasgow, UK. 335 pp. Ljosland, R. 2012. The establishment of the Scots language in Orkney. The New Orkney Antiquarian Journal 6:65–80. Low, G. 1774 [1879]. A Tour through the Islands of Orkney and Shetland. William Peace and Son, Kirkwall, UK. 223 pp. Maisel, W. 1993. Rechtsarchäologie Europas. Böhlau, Vienna, Austria. Meier, J. 1950. Ahnengrab und Rechtsstein. Untersuchungen zur deutschen Volkskunde und Rechtsgeschichte 2. Akademie Verlag, Berlin, Germany. 158 pp. Meurkens, L. 2010. The late medieval/Early Modern reuse of prehistoric barrows as execution sites in the southern part of the Netherlands. Journal of Archaeology in the Low Countries 2(2):5–29. Avaible online at http://dpc. uba.uva.nl/cgi/t/text/get-pdf?c=jalc;idno=0202a01. Mol, J. 2007. Gallows in Late Medieval Frisia. Pp. 263–297, In R.H. Bremmer Jr., S. Laker, and O. Vrie (Eds.). Advances in Old Frisian Philolology. Amsterdamer Beiträge zur älteren Germanistik 64. Rodopi, Amsterdam, the Netherlands. 494 pp. Nicolaisen, W. 1983.The Post-Norse Place-Names of Shetland. Pp. 69–85, In D. Withrington (Ed.). Shetland and the Outside World 1469–1969. Aberdeen University Studies Series 157. Oxford University Press, Oxford, UK. 246 pp. O’Grady, O.J.T. 2008. The setting and practice of open-air judicial assemblies in medieval Scotland: A multidisciplinary study. Ph.D. Thesis. University of Glasgow, Glasow, UK. Available online at http://theses.gla. ac.uk/506/. [PROVIDE DATE ACCESSED] Poole, S. 2008. A lasting and salutary warning: Incendiarism, rural order, and England’s last scene of crime execution. Rural History 19(2):163–177. Reynolds, A. 2009. Anglo-Saxon Deviant Burial Customs. Oxford University Press, Oxford, UK. 324 pp. Robertson, T.A., and J.J. Graham. 1964. Shetland Folk Book 4. Shetland Folk Society, Lerwick, UK. Roymans, N. 1995. The cultural biography of Urnfields and the long-term history of a mythical landscape. Archaeological Dialogues 2:2–24. Sanmark, A. 2013. Patterns of assembly: Norse thing sites in Shetland. Pp. 96–110, In A. Sanmark et al. (Eds.). Debating the Thing in the North I. Selected Papers from Workshops Organized by The Assembly Project. JONA Special Volume 5. 124 pp. Semple, S. 1998. A fear of the past: The place of the prehistoric burial mound in the ideology of Middle and Later Anglo-Saxon England. World Archaeology 30(1):109–126. Semple, S. 2013. Perceptions of the Prehistoric in Anglo- Saxon England: Religion, Ritual, and Rulership in the Landscape. Oxford University Press, Oxford, UK. 330 pp. Sigurðsson, G. 2006. Did Icelandic outlaws live alone in the mountains or did they gather into bands? Were people generally frightened of them? The Icelandic Web of Science. Availble online at http://why.is/svar. php?id=5636. Last accessed 28 March 2016. Sinclair, G. 1685. Satan’s Invisible World Discovered. Facsimile reprint 2004. Kessinger Whitefish, MT, USA. 292 pp. Smith, B. 1996. The development of the spoken and written Shetland dialect: A historian’s view. Pp. 30–43, In D.J. Waugh and B. Smith (Eds.). Shetland’s Northern Links: Language and History. Scottish Society for Northern Studies, Edinburgh, UK. 254 pp. Smith, B. 2003. Archdeacons of Shetland 1195–1567. Pp. 161–169, In S. Imsen (Ed.). Ecclesia Nidrosiensis 1153–1537: Søkelys på Nidaroskirkens og Nidarosprovinsens historie. NTNU Senter for Middelalderstudier Skrifter 15. Tapir, Trondheim, Norway. 488 pp. Journal of the North Atlantic J. Coolen 2016 Special Volume 8 114 Smith, B. 2006. Gibbets and gallows: Local rough justice in Shetland, 800–1700. Available online at http:// normblog.typepad.com/normblog/2006/09/gibbets_ and_gal.html. Last accessed 28 March 2016. Smith, B. 2009. On the nature of tings: Shetlands’s law courts from the middle ages until 1611. New Shetlander 250:37–45. Stewart, J. 1987. Shetland place-names. Shetland Library and Museum, Lerwick, UK. 353 pp. Tait, I. 2012. Shetland Vernacular Buildings 1600–1900. Shetland Times, Lerwick, UK. 598 pp. Towrie, S. 1996–2014. The ketthontla: A lost ogress? Available online at http://www.orkneyjar.com/folklore/ kedhontla.htm. Accessed 24 June 2014. Tulloch, L. 2014. Shetland Folk Tales. History Press, Stroud, UK. 273 pp. Turnbull, J. 1845. United parishes of Tingwall, Whiteness, and Weisdale. Pp. 59–70, In New Statistical Account Vol. 15. Blackwood and Sons, Edinburgh, UK. 1011 pp.. van der Sanden, W. 2010. Kuilen, palen en kisten. Archeologisch onderzoek op de Gerechtsplaats in Assen. Pp. 187–203, In W. van der Sanden, H. Luning (Eds.). Over galg en rad. Executieplaatsen in Drenthe. Erfgoed in Drenthe 2. Waanders, Zwolle, the Netherlands. 255 pp. Whyte, N. 2003. The after-life of barrows: Prehistoric monuments in the Norfolk Landscape. Landscape History 25(1):5–16. Willumsen, L.H. 2013. Witches of the North: Scotland and Finnmark. Studies in Medieval and Reformation Traditions 170. Brill, Leiden, The Netherlands. 394 pp. Endnotes 1Forum post by Brian at http://www.shetlink.com/index. php?/topic/992-gallows-hills. Posted 30 September 2006. 2The site Wilgi Geos in North-Roe was not included in Smith’s initial paper as he did not consider it a likely place of execution. However, the place-name has been recorded by the Shetland Amenity Trust’s Place-Names Project. Thus, the total number of sites discussed in this paper is 15. 3Robert Cowie (1879:94f.) dated the event to the beginning of the 18th century. However, Barbara Tulloch is referred to as a “brunt witch” in the Tingwall Kirk Session of 1693 (Smith 2006). Cowie gives the name of the daughter as Ellen King. She may be identical with Helen Stewart, who according to George Sinclair (1685:177) was burnt “a few years ago” together with her old mother. Following Larner et al. (1977:221), Willumsen (2013:168) dates the execution to 1675. 4Shetland Amenity Trust’s Place-Names Project has mapped the name Gallow Hill at Eshaness on Watch Hill, a knoll to the east of Braewick. The Gallow Hill at Whalsay may be identical with Setter Hill near Marrister. 5“Gallows” on http://www.oed.com and http://www.etymonline. com (accessed 22 July 2015). 6It has been suggested that the burnt spot may be much younger and may rather have been caused by annual bonfires that were held here until the 1990s (post by “MuckleJoannie” from 12 October 2006 at http://www. shetlink.com/index.php?/topic/992-gallows-hills/page- 2). However, accounts of the burnt area are consistent and were reported as early as 1958 (recording available at http://www.tobarandualchais.co.uk/fullrecord/35070/1. Accessed 24 June 2014). 7Shetland Archives D27/1/94/18. 8Available online at http://www.tobarandualchais.co.uk/ fullrecord/81264/1 (recorded 1978; accessed 18 June 2014) and http://www.tobarandualchais.co.uk/fullrecord/ 86149/1 (recorded 13 July 1976; accessed 18 June 2014). 9Available online at http://www.tobarandualchais.co.uk/ fullrecord/31547/1 (recorded 1955; accessed 18 June 2014). 10A written version of the story was published in the Shetland Times of 27 January 1877. 11A record of the story, told by A. Couper in 1974, is available on http://www.tobarandualchais.co.uk/fullrecord/ 72926/1 (accessed 7 June 2014). 12Post by “Marjolein” on http://www.shetlink.com/index. php?/topic/992-gallows-hills (posted 27 September 2006; accessed 11 June 2014). Others believe that the Thief o da Neean came to his end in prison at Scalloway. 13Available online at http://www.tobarandualchais.co.uk/ fullrecord/78536/1 (Recorded 21 September 1975; accessed 10 June 2014) 14Available online at http://www.youtube.com/ watch?v=N0oBrfyyQx4 (uploaded 11 February 2009; accessed 7 June 2014) and http://www.shetlink.com/ index.php?/topic/992-gallows-hills/page-2 (post by ‚brunalf‘ from 3 November 2008; accessed 7 June 2014). 15Posts by “Njugle” on http://www.shetlink.com/index. php?/topic/992-gallows-hills (posted 26 September 2006, 30 September 2006 and 11 October 2006). 16Available online at http://www.themodernantiquarian. com/site/12294/gallows_hill_hanging_stone.html (posted by drewbhoy on 12 August 2009; accessed 24 June 2014).