Journal of the North Atlantic 16 R. Rennell 2015 Special Volume 9 Introduction: The Outer Hebridean Iron Age The Outer Hebrides are a group of islands located off the west coast of Scotland (Fig. 1). In the Outer Hebridean islands and across Atlantic Scotland more generally, the Iron Age is defined by the use of decorated pottery and the emergence of monumental domestic architecture in the form of brochs, duns, and wheelhouses. Brochs are large, dry-stone built roundhouses, associated with long sequences of occupation spanning the second half of the first millennium BC into the first few centuries AD. These buildings are associated with tower-like proportions and a range of specific architectural features, including concentric walling, intra-mural galleries and stairs, scarcement ledges for secondary flooring, and long, narrow entrance passages (Fig. 2; Armit 2003). The term dun, within the context of the Atlantic Scottish Iron Age, is traditionally associated with smaller drystone built roundhouses that lack evidence for the complex architectural features described above. An alternative classification system uses the idiom “Atlantic roundhouse” as an umbrella term for a range of monumental Iron Age roundhouse types that includes both brochs and duns (Armit 1992). However, the precise classification of these sites is a subject of fierce Re-engaging with the Iron Age Landscapes of the Outer Hebrides Rebecca Rennell* Abstract - This paper explores Iron Age landscapes and places in the Outer Hebrides. The Outer Hebrides are a group of islands where the Iron Age is defined chiefly by the distribution of monumental settlement architecture. The research behind this paper was motivated by the observation that experiential or sensory landscape archaeology had been largely neglected within British Iron Age archaeology and in the study of the Outer Hebridean Iron Age more specifically, despite being comparatively well developed in the context of Neolithic and Bronze Age research. Although there are a variety of perspectives on the Outer Hebridean Iron Age within current literature, they all rely upon the premise that this society was structured primarily around differences in monumental domestic architecture. This paper offers an alternative narrative for the Outer Hebridean Iron Age, structured specifically around an understanding of landscape and place. Four principal landscape settings are identified for Iron Age sites: lowland coastal, inland islet, upland, and coastal headland. These places are associated with a range of distinct experiences, and I argue that they provided locales in which dwelling would by necessity have functioned very differently. This paper concludes by examining some of the assumptions previously made about the landscape location of Iron Age sites and in doing so questions some of the dominant interpretations of the Outer Hebridean Iron Age. Special Volume 9:16–34 2010 Hebridean Archaeology Forum Journal of the North Atlantic *Lews Castle College, University of the Highlands and Islands, Benbecula Campus, Lionacleit, Isle of Benbecula, Outer Hebrides, HS6 5PJ; rebecca.rennell@uhi.ac.uk. 2015 Figure 1. The Outer Hebrides archipelago, showing the main islands; Lewis, Harris, North Uist, Benbecula, South Uist and Barra. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 17 debate, and remains as yet unresolved (Armit 1997a, Harding 2004, Sharples 2006, Sharples and Parker Pearson 1997, Sharples et al. 2004). Across Orkney, a number of simple, thick-walled roundhouse structures, associated with Early Iron Age dates (ca. 800 BC–400 BC), have been interpreted as native precursors to the broch tradition (Ballin-Smith 1994, Hedges 1987, Renfrew 1979, Sharples 1984). However, the Early Iron Age remains fairly elusive within the Outer Hebrides, and here the emergence of monumental domestic architecture appears to be a largely Middle Iron Age phenomenon (ca. 200 BC– AD 400) (Armit 1990a, 1991). Both brochs and duns are found widely across Atlantic Scotland, while wheelhouses have a more discrete distribution, as yet identified only in the Outer Hebrides and Shetland. Wheelhouses are also drystone built roundhouses, but are characterized by radial piers which sub-divide the interior roundhouse space into small bays, set around a central area (Fig. 3). The majority of wheelhouses were revetted into the ground so that they would have been occupied as semi-subterranean buildings. Although not externally monumental in the manner of the broch or dun, their elaborate internal architecture suggests that wheelhouses were also impressive, monumental buildings of their time. There are examples of wheelhouses built into earlier broch-type structures, both in the Outer Hebrides (Armit 1998) and in Shetland (Hamilton 1956), suggesting that wheelhouses replaced brochs as the primary form of Iron Age dwelling. However, radiocarbon sequences continue to extend the dates associated with wheelhouse occupation from perhaps as early as the 4th century BC (Barber 2003) into the 5th centuries AD (Sharples 2012). Although questions remain over the reliability of the earliest dates in this sequence (ibid.), it has become apparent that there was a significant degree of overlap between brochs and wheelhouses in the Outer Hebrides, where they are perhaps best viewed as broadly contemporary Middle Iron Age settlements (Parker Pearson and Sharples 1999:359). The construction of brochs, duns, and wheelhouses marks a change in attitudes to the home and domestic architecture and is the culmination of a process whereby monumental expression shifts from more overtly “ritual” contexts to increasingly “domestic” spheres of life (Armit and Finlayson 1992:670). A similar trend towards elaborate and conspicuous Iron Age settlements can be identified Figure 2. Dun Carloway, Lewis. A “classic” Iron Age broch.. Journal of the North Atlantic 18 R. Rennell 2015 Special Volume 9 within wider narratives of British prehistory (Haselgrove and Pope 2007). Traditionally, brochs have been interpreted as defensive structures (Childe 1935:204), a view still popularly held (Blythe 2005). However, over the last twenty years, purely defensive explanations for these sites have been increasingly challenged (Parker Pearson and Sharples 1999:350), reflecting wider trends within the discipline (Fitzpatrick 1997; Hill 1989, 1995; Parker Pearson 1996), and it is now more readily suggested that the monumentality of these sites would have represented social power, over and above the need to defend territory (Armit 1997b:249), perhaps legitimized through associations with earlier monumental styles of architecture (Hingley 1996, 1999, 2005; MacDonald 2008). Environmental change (Sharples 2006), population pressures (Armit 1990a), and economic intensification (Dockrill 2002) have all been cited as potential catalysts for this sudden desire for monumental displays of power. Largely as a consequence of the visual dominance and impressive monumentality associated with Middle Iron Age settlements, research in the Outer Hebrides has tended to view and interpret Iron Age society primarily through these buildings. Furthermore, alternative interpretations of the relationship between broch and wheelhouse communities hinge largely upon debates about site classification and associated problems of chronology. For example, where sites are defined by a restricted number of characteristics (MacKie 2006), the number of “true” brochs remains small, leading to the interpretation that these were elite residences, in comparison with the more lowly wheelhouses (Parker Pearson and Sharples 1999). Conversely, a more inclusive site typology suggests that broch-type architecture may have represented a fairly standard Iron Age dwelling (Armit 2002). The Outer Hebrides offer unique preservation conditions, which provides great scope for reconstructing details of Iron Age life. For example, extensive analysis of faunal and floral remains (Bond 2002, Smith and Mulville 2004) and ceramic residues (Campbell 2000, Campbell et al. 2004) provide invaluable insight into Iron Age subsistence and food-consumption practices. However, when the data for the Outer Hebrides is considered as a whole, Figure 3. One of the wheelhouses at The Udal, North Uist. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 19 as yet there remains no clear distinction between brochs, duns, and wheelhouses in this respect, and instead these analyses point more readily towards regional differences rather than site-type trends. Nevertheless, the idea that differences in architecture imposed the primary, underlying order to Iron Age society persists, and frequently other forms of material culture are forced to fit these predefined categories. Additionally, there has been a lack of interpretations driven by an understanding of the archaeological landscape. Notable exceptions include Scott’s (1947) seminal paper “The problem of the brochs”, Noel Fojut’s (1982) work on brochs in Shetland, and Sharples and Parker Pearson’s (1997) discussion of Iron Age landscapes across South Uist. What remains absent, however, are specifically engaged studies of sensory landscapes and the types of landscape approach that are increasingly common to other periods of British prehistory. While studies of architectural space have begun to explore the meaning and underlying structure of Iron Age roundhouse dwellings (Foster 1989; Hingley 1984, 1995; Giles and Parker Pearson 1999; Parker Pearson and Richards 1994; Parker Pearson and Sharples 1999:16–21), these buildings remain detached from their landscape context. In fact, the view that Iron Age dwelling constitutes the creation of meaningful and symbolic spaces has not yet extended much beyond the roundhouse wall. The scale and presence of these sites demonstrate a considerable investment in the Iron Age house, many in use over several generations. Although rarely considered, this also involved significant investment in place and the creation of wider social landscapes. Occasionally, references are made to the experiential qualities of Iron Age site locations; the locations of Iron Age sites have been described as “liminal” (Parker Pearson et al. 2004:39) as well as “extreme” and “dramatic” (Branigan and Foster 2002:84). In the absence of any comprehensive study, however, it becomes problematic when assertions about the experiential qualities of place are used to reinforce arguments about the structure of Iron Age society. In light of these observations, my research set out to systematically investigate Iron Age places and landscapes across the Outer Hebrides, from a specifically experiential landscape approach. An Experiential Landscape Approach The term experiential landscape archaeology is used here to refer to research that explores the experiential characteristics and sensory qualities of archaeological landscapes and landscape locales (or places). As an approach, experiential landscape archaeology is grounded in social and spatial theories that emphasize the importance of human scales of experience. Although experiential approaches tend to focus on more explicitly ritual or ceremonial contexts (for example, Cummings and Whittle 2004, Cummings et al. 2011, Noble 2007), more recently the relevance of this approach to everyday or domestic settings has been increasingly explored (Hamilton and Whitehouse 2006a, b; see also Rennell 2008). The importance of place and landscape in the structuring of everyday life are ideas well developed in Pierre Bourdieu’s (1977) Theory of Practice and are encompassed in the term “dwelling” (Ingold 2000). Also of relevance is the work of sociologist Anthony Giddens (1984) and the social geography of Allen Pred (1984), both of whom develop upon Hagerstrand’s theory of “Time-geography” and his concern with the spatio-temporal character of daily life. These ideas also articulate with a range of phenomenological approaches within archaeology (Tilley 1994, 2004) in which it is emphasized that people come to know, understand, and act in the world through their very physical experience of “being- in-the-world” (Merleau-Ponty 1962). Importantly, an interest in everyday experience in the Outer Hebrides also requires that serious thought is given to the island nature of these landscapes. Despite a well-developed body of theory relating to the archaeology of islands (Broodbank 2000, Noble et al. 2008, Rainbird 2007), these ideas have yet to be fully integrated into our understanding of the Outer Hebridean Iron Age. In the context of my research, adopting an island approach meant thinking about the way in which islands are defined, perceived, and experienced and critically analyzing the appropriateness of the island or group of islands as a region of study (Rennell 2010a). These arguments about place, landscape, and islands informed the theoretical and methodological framework for my research, which combined subject- centered landscape survey and the use of GIS. The subject-centered field survey involved engaging with the Outer Hebridean landscape and investigating and recording sensory qualities of Iron Age places via a number of exploratory field practices. In order to shift the emphasis away from architectural typologies, the survey was not restricted to any particular type of Middle Iron Age site, but explored the full suite of monumental roundhouses associated with this period: brochs, duns, and wheelhouses (Fig. 4). An initial field survey was undertaken to gather information about landscape location. This Journal of the North Atlantic 20 R. Rennell 2015 Special Volume 9 approach included recording detailed descriptions of the local topography, underlying geology, soil, and vegetation, and an on-site assessment of environmental changes that might have occurred over the last two thousand years. As well as documenting details about site location, records were made about specific experiential qualities of place, in order to test some of the assertions made within the archaeological literature about the location of Iron Age sites. This data included observations about visibility of the sea, (Armit 1990b), the visibility of various environmental zones (Cunliffe 1978; Parker Pearson et al. 2004a, b), and scales of landscape visibility (Fojut 1982). A sub-sample of eight sites was then selected for more detailed survey work. This follow-up effort included field experiments designed to assess inter- visibility and inter-audibility within these landscapes and between Iron Age sites. GIS was used as a subsequent means of mapping these places and modelling aspects of visual experience.1 Continuous viewshed models were generated as a method of characterizing the potential visual experiences associated with site locales. Continuous viewshed models show the percentage of an area visible from all locations within that area (Llobera 2003). In effect, these models, based upon topographic data, provide information on the visual qualities one is afforded while moving throughout a specific landscape. A series of cumulative viewshed models (Lake et al. 1998, Wheatley 1995) and heightened viewshed models were generated to investigate the visibility of roundhouse sites and the visibility from roundhouse sites, using a series of hypothetic building heights. When assessing the visibility of Iron Age sites in the field, these experiments were restricted by the preservation of each roundhouse, which in most cases survived as mounds containing a few courses of stone walling. Consequently, the field-based research was unable to inform about the impact of these buildings as specifically monumental structures. Therefore, the purpose of the heightened and cumulative view models was to explore how the potential monumentality of these buildings affected visual experience within the landscape. The overall methodology was deliberately experimental, designed to explore and develop methods for experiential landscape archaeology and, more specifically, the potential for combining subject-centered field survey practices with the use of GIS (for a more detailed discussion of the experimental methodology used, see Rennell 2009:Chapter 5, 2012). By including the full range of Iron Age sites from across the Outer Hebrides, another unique element of this research was its geographical breadth. Similar coverage of the Iron Age material across the Outer Hebrides has been lacking since Ian Armit’s (1992) publication of later prehistoric settlement types and in relevant chapters of his later publication The Archaeology of Skye and the Western Isles (Armit 1996). Since these publications, a large number of excavations and survey projects have been published that contribute significantly to our Figure 4. Distribution map showing the location of known broch, dun and wheelhouse sites throughout the Outer Hebrides. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 21 data indicates a decline in native woodland across the islands from the early Holocene, with arboreal species rapidly replaced by heathland vegetation and the formation of peat (Birks and Masden 1979; Bohncke 1988; Edwards and Brayshay 1996, 2000; Edwards et al. 2000; Fossit 1996). By the Iron Age, the island interior would have been defined by fairly barren peat-based moorlands, dotted with small lochs, giving way to more mountainous and rocky areas on the east coast. In terms of these broad environmental zones, the Iron Age landscape was similar to the islands we find today (Fig. 5). From as early as 400 BC, Iron Age communities began building monumental roundhouses across these island environments, creating a nexus of domestic places that formed an integral part of the social landscape. A large section of the Iron Age community established their homes along the low-lying coastal machair, an area that had been densely occupied since the Late Bronze Age (Sharples et al. 2004). Iron Age communities also built roundhouses within the island interior, on islets within inland lochs. Increasing investigation suggests that islets had been important places during the Neolithic, which may have provided precedence for the establishment of Iron Age settlements in these locations. At the same time, a small number of monumental roundhouses were built within upland landscapes or on high coastal headlands. This range of locations provided locales with distinctly different experiential and sensory parameters, and dwellings in the different types of locations would have functioned in markedly different ways. In the following section, I will describe some of the results of this research, with reference to three of the case-study landscapes: the Vallay Strand, a lowland coastal landscape in North Uist (Fig. 6); Dun Bharabhat, an inland islet site at Bhaltos, Lewis (Fig. 7); and the wheelhouse at Cleitreabhal, within the upland landscapes of North Uist (Fig. 8). Lowland Coastal Landscape Within lowland coastal landscapes, Iron Age houses were built into the machair grasslands or were located on islets within the lagoons or machair lochs. In these parts of the landscape, houses tended to be built in close proximity to one another, as at the Vallay Strand. The Vallay Strand is a low-lying area of coastal machair on the north coast of North Uist, fronting a large area of inter-tidal sand flats that separate the island of Vallay from North Uist at high tide. During the Iron Age, however, this inter-tidal zone was probably defined by expansive understanding of the Outer Hebridean Iron Age. To date, these projects have been largely discussed within the context of smaller island regions: South Uist (Parker Pearson and Sharples 1999), Barra and associated islands (Branigan and Foster 2002), and the Bhaltos peninsula, Lewis (Armit 2006, Harding and Dixon 2000, Harding and Gilmour 2000). Based upon my own research, and drawing upon this vast dataset, the following section of this paper outlines an alternative narrative for the Outer Hebridean Iron Age structured around an understanding of place and landscape. A Landscape Narrative The Iron Age environment in the Outer Hebrides, forming the underlying skeleton to the social landscape, was comprised of three distinct zones: rocky coasts defined by thin acidic soils and sparse heathland vegetation, coastal machair, and interior zones of comparatively inhospitable peat-based moorland. The machair is a unique type of ecological environment, formed from wind-blown calcareous sands and comprising a number of different landscape elements, including dunes, grasslands, lagoons, and hill machair (Angus 2001). The machair systems probably began to take shape approximately 8000 years ago, formed through processes of erosion and deposition caused by strong oceanic winds and an excess of sand following deglaciation (Ritchie 1979). During the Iron Age, the machair defined large parts of the west and northern coasts of the islands and would have been the focus of agricultural activity. However, since the Iron Age, erosion and sea-level rise have caused extensive areas of machair grasslands to flood and the coastline and dunes to move progressively inland. Evidence for sea-level rise (Ritchie 1966, 1979, 1985) suggest that during the Iron Age some current islands, such as the islands of Uist, would have been connected for large periods of the tidal cycle, while others may have been linked by permanent land-bridges. As has been widely acknowledged, the island unit, although convenient for archaeologists, does not necessarily correlate with the way in which island communities interact with the spaces around them and their experiences of island life. This perspective is particularly pertinent in studies of prehistoric communities. Instead, physical barriers within the landscape are more likely to take the form of mountain ranges, such as the mountains that separate Harris from Lewis or the large hills on the east coast of Uist, and large inland water systems, such as Loch Be, also in South Uist. Pollen-core Journal of the North Atlantic 22 R. Rennell 2015 Special Volume 9 Figure 5. Key environmental zones across the Outer Hebrides based on SNH landscape character sssessment data. Figure 6. Cumulative Viewshed model for Vallay Strand. Maps A and B show the location of the Vallay Strand and the distribution of Iron Age roundhouses within this area. Map C shows the percentage of the surrounding landscape visible from each location. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 23 machair grasslands that have since flooded due to rising sea levels and erosion of the coast. Along the modern coastline of the strand are the remains of three wheelhouses inserted into the remains of earlier dun-type structures: Eilean Maleit, Garry Iochdrach, and Cnoc a’Comdhalach. These sites were all investigated by the antiquarian Erskine Beveridge (1911) in the early part of the 20th century. In addition, Eilean Maleit was partially excavated during the 1990s, confirming Beveridge’s interpretation of the site (Armit 1998). Pottery comparable with material from wheelhouses at Cnip on Isle of Lewis, Figure 7. Cumulative Viewshed model for Dun Bharabhat. Maps A and B show the location of the Bhaltos Peninsular and the distribution of Iron Age roundhouses within this area. Map C shows the percentage of the surrounding landscape visible from each location within this area. Figure 8. Cumulative Viewshed model for Cleitreabhal. Maps A and B show the location of the Cleitreabhal and the distribution of Iron Age roundhouses within this area. Map C shows the percentage of the surrounding landscape visible from each location within this area. Journal of the North Atlantic 24 R. Rennell 2015 Special Volume 9 the public and communal character of large parts of these landscapes (Fig. 9A). Alternatively, the closeness and lack of privacy may have meant that boundaries between communities required strong social expressions and concepts of territory, and land ownership might have been emphasized in ways that do not survive archaeologically. The continuous viewshed models strengthen the interpretation of these landscapes based upon field-survey data alone, indicating fairly high visibility of the surrounding landscape but also fluctuating and therefore variable scales of visibility (Fig. 9). Hypothetical, GIS-generated models of roundhouse visibility indicate that within these low-lying landscapes there was immense potential for increasing the visibility of these sites by building outwardly monumental roundhouses. Tidal cycles would have dictated any shorebased subsistence practices, such as the collection of shellfish, shore-based fishing, and the collection of seaweed for improving machair soils. Constant sand movement, deflation and accretion, and the general instability of coastal dunes would have been major concerns. The importance of agricultural land and the threat to these areas through longer-term environmental changes would have increased competition and claims to land between communities. Seasonal cycles affecting the weather would also have imprinted themselves on these places. Excavations of several lowland coastal sites have indicated that windblown sand was a persistent problem for Allasdale on Isle of Barra, and Sollas on Isle of North Uist point to occupation during the first few centuries AD (ibid.). A possible wheelhouse is also located further north still along the Vallay Stand coastline, on the small headland at Geiriscleit. This area also contains the remains of a number of earlier prehistoric and later historic sites, including a severely eroded Neolithic burial cairn located on the edge of the current high-water mark (Dunwell et al. 2003), a burnt mound and associated cellular structures at Ceann nan Clachan (Armit and Braby 2002), as well as the remnants of numerous walls and structures relating to an abandoned post-medieval settlement. Experiments in the field suggest that lowland coastal landscapes, like the Vallay Strand, would have been noisy and busy places to live. The sound of people tending to nearby crops, animals, and children and also the sound of the sea would have filled these landscapes. From outside the roundhouse, people would have been able to see other members of the community working on the machair; tending to crops, perhaps bringing animals back home for slaughter. The sounds of people and their animals at neighboring houses would have been heard on all but the windiest of days and if not audible, then activities around these places would have been highly visible. There would have been limited privacy for the occupants of these landscapes, and daily experiences of these places would have strengthened the strong social links between people, emphasizing Figure 9. Panoramic photographs from: (A) Cnoc a’Comhdhalach, Vallay Strand, North Uist; (B) Dun Bharabhat, Bhaltos Peninsular, Lewis; and (C) Cleitreabhal, North Uist. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 25 have necessitated traversing a causeway, using a boat, or wading through the water—very physical experiences of the nature of separation from the wider landscape. Although not the case at Bhaltos, many islet settlements were remote or peripheral to the more densely occupied lowland coastal landscapes, and instead these places would have been dominated by moorland environments. Proximity to these environments suggest that communities living on islets were likely to have been involved in pastoral activities over other subsistence practices that we associate with the Iron Age. The machair, coast, and sea at the majority of islet sites would have been some distance away and may well have been regarded as peripheral, existing on the margins of visual and audible communication. Therefore, cultivation and care for these landscapes may not have been principal concerns for the occupants of islet sites. Hypothetical models of roundhouse visibility indicate that these buildings would only have marginally increased their visibility within the surrounding landscape by constructing monumental proportions, and in comparison with the machair landscapes, there was limited potential for creating visually impressive sites. As with lowland coastal places, islet dwelling would have encompassed dynamic characteristics— excavation at Dun Bharabhat indicated that the occupants regularly rebuilt the site in order to combat the rising water levels (Harding and Dixon 2000). Similar evidence was found at the Neolithic islet site of Eilean Dòmhnuil in North Uist (Armit 1996), indicating that these conditions have persisted throughout prehistoric occupation of these locales. Water levels will have changed as a consequence of fluctuating periods of heavy rain or drought, either on a seasonal basis or following exceptional weather conditions. These transformations would have affected the nature of these places profoundly. The roundhouse may have been cut-off from the shore for some if not lengthy periods of the time, accentuating separation and removal from the surrounding landscape. Water-based communications were also potentially an intimate part of islet dwelling in these parts of the landscape. Elsewhere in the highlands, there is considerable evidence for the use of log boats during this period (Mowat 1996), and given the suitability of these vessels to the Outer Hebridean environment (McGrail 2001), the lack of direct evidence should not preclude discussions of the social implications of boat travel within this island-based community (Farr 2006:90). If the Iron Age occupants of these roundhouses had access to log boats or other water-borne vessels, then the occupants of these sites. At Cnip, the excavators believe that adaptations to the original wheelhouse, including an extended entrance passage and guard cell, were modifications motivated by the need to combat accumulating sands within the house (Armit 2006). Similarly, at The Udal, Crawford (ND) comments on the problems of sand incursion at wheelhouse B. Iron Age occupants of these lowland coastal landscapes would therefore have been accustomed to the dynamic nature of living in these places, and this changing environment would have established a particular tempo and series of concerns central to living in these fragile machair locations. Interior Islet Landscapes By contrast, islet sites found within the island’s interior would have been characterized by the restricted nature of landscape visibility and experiences of enclosure and isolation. Lochs within the island interior, with their substantially defined banks, differ from those on the coastal machair, which are often more temporary bodies of water. The islet settlement within Loch Bharabhat is a good example of this type of site. This site was excavated during the 1980s as part of an Edinburgh University-led research project focusing on Later Prehistoric occupation at Bhaltos (Armit and Harding 1990). This project included the excavation of a broch at Loch na Beirigh (Harding and Gilmour 2000), the wheelhouses at Cnip (Armit 2006), and the dun within Loch Bharabhat (Harding and Dixon 2000). Another wheelhouse has been identified on the Traigh na Berigh machair, and a possible dun is located on the southern coast of the peninsula (Fig. 7). Although a number of broadly contemporary Iron Age sites were built in fairly close proximity to Dun Bharabhat, these other sites would not have been visible from the islet, and the sound of people at the nearby broch or wheelhouses, or people working on the machair would not have been audible (Fig. 9B). Similarly, the islet and the dun would have been concealed or hidden from the surrounding landscape, and people working in the wider landscape or approaching this place would be invisible, beyond the banks of the loch. The continuous viewshed models generated for Dun Bharabhat highlight the predominance of very low general visibility within the landscape surrounding the islet (Fig. 7). Sounds emanating from the roundhouse itself, even people shouting, would have been contained within the banks of the loch, echoing around the site and further accentuating an impression of isolation and seclusion. Access to areas beyond the islet would Journal of the North Atlantic 26 R. Rennell 2015 Special Volume 9 this would have had a profound effect upon their experiences, knowledge, and understanding of these places and the wider landscape. The possibility that islet dwellers were using boats also demands that we review the concept of islet sites as detached or cutoff from the surrounding landscape. Instead of experiencing these places as physically or experientially isolated, Iron Age people may well have regarded these places as highly connected and dynamic locations. These locations might also have provided Iron Age occupants with links to other parts of the landscape through the complex inland water systems that would have dominated these Iron Age environments. These alternative experiences and knowledge of the landscape may well have reaffirmed differences within the community between islet-dwelling and lowland coastal-dwelling sections of this Iron Age society. Upland and Coastal Headland Sites While the majority of the Iron Age community appear to have created domestic places within lowland coastal landscapes or on inland lochs, some monumental architecture of this period was placed within upland environments and on high coastal headlands. These places shared a number of experiential qualities that would have contrasted markedly with places on the low-lying coastal machair zone and the inland islets. For example, the wheelhouse at Cleitreabhal is located on the slopes of a large hill, providing extensive and uninterrupted views of the wider landscape, as well as extensive views out to sea (Fig. 9C). However, despite these visual qualities and immediate proximity, these would not have been places from where the sea was easily accessed. The cumulative viewshed model for Cleitreabhal highlights limited visibility of the local landscape in contrast with the extensive views of regional and distant locales (Fig. 8). These fairly anomalous locations present the possibility that these sites functioned in very different, perhaps non-domestic, ways despite apparent similarity in architecture. The relationship between the wheelhouse at Cleitreabhal and the earlier chambered cairn (Scott 1935) perhaps indicates that this site had a special meaning or function with Iron Age society. Discussion The narrative presented above has attempted to convey how Iron Age people might have engaged with the landscapes and places that they inhabited. I have offered an additional schema for investigating the structure of Iron Age society based on the sensory qualities of Iron Age places and landscapes and argue that these places provided locales in which dwelling would by necessity have functioned very differently. The results of the field survey and GIS-based modelling are summarized in Tables 1 and 2. The results of this research also questions some of the assumptions previously made about the landscape location of Iron Age sites and offers an alternative perspective on this Iron Age society. Iron Age sites, environment, and subsistence practices Parker Pearson and Sharples (1999:363) suggest that wheelhouse-based communities were closely tied to arable cultivation. In contrast, they envisage a broch-dwelling elite living in comparatively “marginal” areas of the landscape and engaged more immediately with pastoralist practices. Currently, however, there is an absence of evidence for brochs, Table 2. Summary of GIS results by landscape location. Lowland Coastal Islet Upland Coastal Headland Continuous viewshed models General landscape visibility High Low to mod Low High Variation in landscape visibility Mod to high Low to mod Low High Heightend viewshed models Significant Limited N/A Moderate Cumulative viewshed models Significant Limited N/A Significant Table 1. Summary of field survey results by landscape location. Lowland Coastal Inland Islet Upland Coastal Headland Inter-visibility between Iron Age site and surrounding landscape (>5 km) Mod Low Low Mod Inter-audibility between Iron Age site and surrounding landscape (>5 km) High Low Low Low Inter-visibility with nearby sites (>5 km) High Low N/A Mod Inter-audibility with nearby sites (>5 km) High Low N/A Low General description of landscape locale (>5 km) Local scale views, Intimate, Divided Open, exposed variable experiences hidden Journal of the North Atlantic R. Rennell 2015 Special Volume 9 27 other complex roundhouses, duns, or wheelhouses differing in terms of the scale of agricultural activities. Wheelhouses, it is rightly observed, tend to be located on the machair, which would have provided the most suitable soils for cultivation; historically, the machair has been the focus of agriculture across the islands (Boyd and Boyd 1990, Lawson 2004) and was also likely the prime location for agricultural activities from as early as the Neolithic (Armit and Finlayson 1992, Mills et al. 2004). However, brochs and duns are also closely associated with these types of landscape, and across the Outer Hebrides as a whole, a direct correlation between architectural typology and environmental/economic landscape zones was not found to be the case. In fact, a principal outcome of this research is the observation that types of place and landscape cross-cut architectural classification systems (Table 3). The types of place identified as “lowland coastal” tended to be sites established in close proximity to fertile machair grasslands, suggesting that agricultural practices would have been important to the subsistence practices of these communities. My research also highlighted strong sensory relationships between roundhouses in these areas and what would have been cultivatable machair surrounding these places. Many Iron Age houses (e.g., wheelhouses) were constructed so that they were physically embedded in the machair soil. Perhaps similarly, at Dun Mhulan midden material was allowed to accumulate outside the doorway of the broch, eventually reaching up to 2m in height (Parker Pearson and Sharples 1999). Midden material from these houses was almost certainly used to fertilize and improve the soils. These practices would have emphasized the conceptual links between the home and the agricultural potential of the surrounding landscape. In contrast, sites located on inland islets would have been distant from the focal points of Iron Age agriculture and in this respect would have been relatively remote and isolated from these other areas of occupation. Instead, it is likely that the occupants of these sites would have focused on pastoral subsistence practices—moving cattle and sheep among grazing land and perhaps bringing them into the islet- dwelling for winter byre or slaughter as suggested by some excavations (Harding and Dixon 2000, Parker Pearson and Sharples 1999). Historically, islets have been used to segregate sections of grazing herds at certain times of year, hence place names within the Outer Hebrides such as “Soay”, meaning “Goat Island” (Lawson 2004), and this nomenclature suggests other ways in which islet landscapes might have served a pastoral-based community. Alternatively, the occupants of these sites would have had stronger links with the moorland and potentially, via interconnecting waterways, with the east coast and moorland areas even further afield. Living in these different parts of the landscape, the tempo and catalysts for environmental change would have been profoundly different to those affecting communities dwelling within the lowland coastal landscapes. These different subsistence practices and environmental concerns would have distinguished these communities and may have encouraged the development of distinct senses of identity. Importantly, this research suggests that differences within Iron Age society were not necessarily structured solely around choices in architecture, but also potentially related to the types of landscapes in which people made their homes and carried out their daily lives. Although Parker Pearson and Sharples’ (1999) interpretation represents an accurate assessment of the Iron Age landscape across South Uist, this model of Iron Age society is far less convincing when transposed to neighboring island regions, such as North Uist and Lewis. In particular, across North Uist, the west coast distribution of sites is not as emphasized as it is in South Uist, and here the distribution of brochs and wheelhouses does not betray this distinct relationship with these environmental zones. Although wheelhouses have not been identified on inland islets, they are found in the uplands as well as low-lying coastal landscapes, while brochs and duns were built in all of the four types of landscape location, suggesting that subsistence practices cross-cut existing classifications of Iron Age site types. Island landscapes As well as affording specific associations with the various environmental zones across the Outer Hebrides, the locations of Iron Age sites also likely Table 3. Experience of place according to landscape location. Uninterrupted views Good visibility No. of sites of the sea and of local Sense of Atlantic roundhouse surveyed wider landscape context landscape enclosure Broch Dun Wheelhouse Upland 16 X X X X Coastal headland 10 X X X - Lowland coastal 32 - X X X X Inland islet 94 - - X X X - Journal of the North Atlantic 28 R. Rennell 2015 Special Volume 9 presumably focused towards more local concerns. Both interpretations can be reconciled within this study of place and landscape, suggesting that Iron Age society was complex and potentially incorporated varying social perspectives, identities, and social practices. The contrast between occupation of the low-lying Atlantic coasts and dwelling within the more easterly, island moorlands adds a further dimension to Iron Age settlement patterns. A prevailing characteristic of the Outer Hebridean island geography is the vital contrast between the Atlantic west coast and the Minch coast on the east and the potential these areas of the island landscape would have afforded Iron Age communities. Generally, the east coast would have provided the principal areas for sheltered and safe anchorage and thus the opportunity for seafaring practices, perhaps relating to subsistence such as deep-sea fishing, although evidence for this is largely absent from the archaeological record, or as means of communicating and accessing wider areas for social reasons. In contrast, the west coast of the Outer Hebrides would have provided minimal opportunities for accessing the sea by boat, but greater potential for a range of shore-based subsistence activities such as collecting shellfish and gathering seaweed and driftwood. The east and west coasts would also have afforded distinctly different opportunities for sensory experience. Across the Atlantic, there is little between the Outer Hebridean islands and the American continent, and this setting in combination with a prevailing westerly wind, makes for a distinctively exposed and windswept western coastline. In comparison, while the landscape of the east coast is far more rugged, the sea across the Minch is less volatile, and this coastline is significantly more sheltered. The major townships across the islands—Castlebay, Lochboisdale, Lochmaddy, Tarbet, and Stornoway—are all located on the east coast. Today, these places function as important harbors for the fishing industry as well as serving the major ferries between the islands and the mainland. In light of these differences, the density of Iron Age occupation on the western coastline, within what has been defined as lowland coastal landscapes, prompts a number of questions. These sites appear not to have been located in places where people could easily view or monitor the sea. Neither were these places associated with natural bays or harbors, and it is unlikely then that experiences of living in these landscapes provided an intimate link with the sea, beyond the shore and coastline. Nevertheless, the sea and coast would have been within accessible distances. People would have been able to hear the sea as well as sea birds, and the smell of conveyed distinct experiences and relationships with the wider island landscape. In particular, there appears to have been a contrast between Iron Age places dominated by views of the immediate locality and with an inward/landward focus (primarily lowland coastal and islet sites) and a minority of places with views of the regional and distant landscapes and specifically outward/seaward looking perspectives (coastal headland and upland sites) (Table 4). In the case of the former, we might envisage a specifically local understanding and knowledge of landscape, with everyday experiences reinforcing local, perhaps intra-island identities. Similarly, daily activities might have been increasingly focused within the immediate landscape, perhaps relating to increasing investment in the immediate locality and the intensification of localized subsistence practices. In comparison, a smaller section of this Iron Age society, living in places with wider views of the regional landscape and perhaps extensive views out to sea, might have associated themselves with a larger social group—perhaps through seafaring activities developing inter rather than intra-island concerns and identities. Within coastal headland and upland landscapes, daily activities had the possibility of taking place within a wider area, perhaps involving subsistence practices taking place further afield or the procurement or movement of materials from “other” places. Similarly contrasting interpretations have been drawn from more traditional analysis of Iron Age material. For example, both decorated ceramics and monumental domestic architecture indicate a high degree of cultural contact between the Outer Hebrides and other island regions within Atlantic Scotland during this period. Henderson describes how contact enabled this area to become a “recognisable zone, prone to simulating itself, (and) creating broad similarities over long distances” (Henderson 2000:150), suggesting an active relationship between island communities and perhaps a sense of shared inter-island regional identity. Alternatively, Armit comments on the relative dearth of imported Iron Age items, pottery or otherwise, within the Outer Hebrides and highlights distinct sequences in the development of monumental architectures that distinguishes the Outer Hebrides from wider Iron Age processes, as indicative of a progressively “inward looking island people” (Armit 1996) Table 4. Contrasting scales of island experience. Local experiences Regional experiences Land Sea Intra-island identity Inter-island identity Local material culture Regional material culture Journal of the North Atlantic R. Rennell 2015 Special Volume 9 29 to the themes raised in association with an “Island Approach”, it appears that islet sites, like islands more generally, can be understood to combine elements of isolation and connectivity. These locations might also have provided Iron Age occupants of these sites with links to other parts of the landscape through the complex loch and sea-loch systems that would have dominated these Iron Age environments. These alternative experiences and knowledge of the landscape may well have reaffirmed differences between islet-dwelling and lowland coastal-dwelling sections of this Iron Age society. Monumental architecture and place This investigation into place and landscape experience also questions some assumptions frequently made about how Iron Age communities engaged with monumental architecture. While brochs were constructed to be highly visible and imposing from the outside, wheelhouse sites were comparatively modest buildings in this respect, yet similarly monumental when viewed (or experienced) from within. It has therefore been suggested, with reference specifically to the location of Iron Age sites on the Bhaltos peninsula, that wheelhouse sites were hidden within the landscape, while Atlantic roundhouses were built within more prominent places (Armit 2006:256). However, my research suggests that the locations of Iron Age sites provided quite different experiences. Wheelhouse sites, predominantly located within lowland coastal landscapes, were not found to be within “hidden” locales. Instead the field survey indicates that these places were highly communal, easily accessible, and relatively exposed places in the landscape. Brochs were found within a wider variety of locales; however, sites built on islets within freshwater lochs, such as Dun Bharabhat on the Bhaltos peninsula, tended to be concealed within the landscape. Analyzing the visibility of roundhouses using varying height models within a GIS suggests that even built to a height of 10 m (unlikely dimensions, given the small diameter of Dun Bharabhat) these places would still not have achieved the prominence that has hitherto been described. This finding suggests that brochs, when built on islets like Dun Bharabhat for example, did not always function as visually imposing sites. In contrast, brochs built within lowland coastal and coastal headland settings would have had much greater potential for visual prominence. If the establishment of monumental domestic architecture during the Iron Age is regarded as a symbolic means of legitimizing rights to land, demonstrating ownership, and local identity (Armit 1997b), how do these interpretations of place further inform our understanding of Iron Age society? Perseaweed and salt in the air would have been a constant reminder of the sea’s proximity. Shell middens associated with lowland coastal sites produce high proportions of whelk (winkle) and limpet shells that could have been collected by Iron Age communities from these westerly shores. Evidence for deep-sea fish is minimal across all Iron Age sites. The fish bone assemblages from the Middle and Late Iron Age sites at Dun Mhulan and Bornais, for example, are dominated by small saithe, suggesting smallscale subsistence fishing practice, probably carried out largely from the shore (Cerón-Carrasco 1999, Ingrem 2012). However, a general lack of fish-bone analysis and minimal strategies for fish-bone retrieval during excavation, mean that these results require some further attention. On the basis of location, however, it is unlikely that these communities had an intimate relationship with the sea, beyond the confines of the shore. In contrast, communities located on the east coast would have had greater potential for accessing the sea, involvement in sea-based as opposed to shore-based subsistence practices, and contact with mainland communities. Table 5 summarizes some of the contrasting experiences, concerns, and activities associated with east and west island settlement. Water-based communications were also potentially an intimate part of islet dwelling in these parts of the landscape. Little consideration has previously been given to the use of boats in association with these sites, perhaps on the basis that causeways precluded their necessity. However, access to log boats or other water-borne vessels by Iron Age occupants of these roundhouses would have had a profound effect upon their experiences, knowledge, and understanding of these places and the wider landscape via the complex maze of inland lochans, sea lochs, and the sea itself. The possibility that islet dwellers were using boats also demands that we review the concept of islet sites as separated or cut-off from the surrounding landscape. Instead of experiencing these places as physically and experientially isolated, Iron Age people might well have regarded these places as highly connected and dynamic locations. Returning Table 5. Contrasting experiences of east and west coast dwelling. West East Atlantic Minch Land/coast Sea/Water Agriculture Pastoralism Shore collection Deep sea fishing? Machair Moorland Lowland coastal Islets Public Private Links with mainland Journal of the North Atlantic 30 R. Rennell 2015 Special Volume 9 reuse of available stone. The fact that the entrance to the chamber is maintained in the construction of this site supports this interpretation. Crawford (2002:127–128) has proposed that wheelhouse sites had specifically “religious” as opposed to domestic functions within Iron Age society. He draws attention to evidence for votive deposits and to specific elements of their architecture that he describes as analogous to church or amphitheatre structures. Yet wheelhouse excavations clearly demonstrate that these sites were domestic buildings, occupied over long periods of time, associated with a range of recognizable “domestic” practices, albeit alongside distinctively “ritual” behavior (Armit 2006, Barber 2003, Campbell 1991, Parker Pearson and Zvelebil 2014, Rennell 2010b, Sharples 2012). However, it may well transpire that not all wheelhouse sites were used primarily for domestic purposes and that different uses of these buildings may have been deemed more appropriate in certain landscape settings. Conclusions This paper has presented some alternative approaches to interpreting Iron Age society in the Outer Hebrides. I have identified a number of different ways in which Iron Age sites were positioned in the landscape and have explored the variation of Iron Age experiences associated with these places. I have suggested that differences in the everyday experience of these places provided alternative perspectives on the wider landscape. I have also pointed out different ways in which monumental architecture might have been experienced, suggesting that creating visually imposing settlements was not always a primary concern. In conclusion, I argue that people’s everyday experiences and perspective of the wider social landscape were as important to Iron Age communities as similarities or differences in the use of architectural styles. Furthermore, these differing experiences affected the manner in which communities identified themselves, played out their day-to-day lives, and structured social relationships. Acknowledgments This paper is based on research carried out as part of my doctoral thesis at the Institute of Archaeology, University College London. This research was funded by the AHRC and the Graduate School at the University College London. I would like to thank my supervisors Professor Sue Hamilton and Dr. Mark Lake for helping me develop this research. I would also like to thank the two anonymous reviewers for their valuable feedback and constructive comments. haps certain Iron Age communities sought to actively harness these senses and experiences of isolation associated with islet locations in order to separate themselves from other parts of the community and to reinforce and maintain their identity. Therefore, the creation of domestic places within certain parts of the landscape might have been a strategy, alongside the establishment of distinctively elaborate and monumental architecture, for demonstrating local power. The relationship between Iron Age sites and earlier monuments may also have played a role in the establishment of monumental architecture during the Iron Age. There is growing evidence to suggest that Iron Age communities across Atlantic Scotland had a particular fascination and interest with these ancestral landscapes (Hingley 1996, 1999, 2005; MacDonald 2008; Sharples 2006). In particular, the conspicuous chambered cairn monuments of Early Neolithic communities may have had a particular significance for Iron Age people. Across Orkney, at least three recently excavated Maes-Howe type tombs revealed Early Iron Age roundhouses built within the Neolithic structures (MacDonald 2008). Elsewhere in the Outer Hebrides, two Neolithic burial tombs were reused as locations for Iron Age roundhouses, and there is evidence that a number of tombs were disturbed and perhaps re-used during this period (Hingley 1996, 1999; Sharples 2006). In fact, it has been argued that Iron Age monumental domestic roundhouses across Atlantic Scotland, brochs in particular, were built with direct reference to the architectural forms of Early Neolithic burial monuments, using links with the past to legitimize a new social order (MacDonald 2008). Iron Age places were obviously created within a landscape that was already embedded with places of meaning, significance, and culture. The burial monuments of Early Neolithic communities, for example, would have been prominent visual markers in the Iron Age landscape. Similarly, the islet settlements of very early island communities such as Eilean Dòmhnuill and Eilean an Tighe, both on North Uist, were also perhaps recognizable places within the Iron Age landscape and may well have acted as templates for Iron Age islet settlements. The wheelhouse at Cleitreabhal is another example that points to a relationship between Iron Age sites and earlier landscape features. The uniqueness of the landscape location of this wheelhouse and the sensory qualities, landscape associations, and removal from other areas of Iron Age settlement, suggest that the re-use of the early chambered cairn was a deliberate and defining factor in the construction of this roundhouse rather than a fortuitous Journal of the North Atlantic R. Rennell 2015 Special Volume 9 31 Beveridge, E. 1911. North Uist: Its Archaeology and Topography (Paperback edition 2001). Birlinn, Edinburgh, UK. Birks, H.J.B., and B.J. Madsen. 1979. Flandrian vegetational history of Little Loch Roag, Isle of Lewis, Scotland. Journal of Ecology 67:825–842. Blythe, I. 2005. A military assessment of the defensive capabilities of brochs. Pp. 246–253, In V.E. Turner, R.A. Nicholson, S.J. Dockrill, and J.M. Bond (Eds.). Tall Stories? 2 Millennia of Brochs. Shetland Amenity Trust, Lerwick, UK. Bohncke, S.J. 1988. Vegetation and habitation history of the Callanish area, Isle of Lewis, Scotland. Pp. 445–461, In H.J. Birks, et al. (Eds.). The Cultural Landscape: Past, Present, Future. Cambridge University Press, Cambridge, UK. Bond, J.M. 2002. Pictish pigs and celtic cowboys: Food and farming in the Atlantic Iron Age. Pp. 177–184, In B. Ballin Smith and I. Banks (Eds.). In the Shadow of the Brochs: The Iron Age in Scotland. Tempus, Stroud, UK. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge University Press, Cambridge, UK. Boyd, J.M., and I.L. Boyd. 1990. The Hebrides: A Natural History. Collins, London, UK. Branigan, K., and P. Foster. 2002. Barra and the Bishop's Isles: Living on the Margin. Tempus, Stroud, UK. Broodbank, C. 2000. An Island Archaeology of the Early Cyclades. Cambridge University Press, Cambridge, UK. Campbell, E. 1991. Excavations of a wheelhouse and other Iron Age structures at Sollas, North Uist by Aitkinson in 1957. Proceedings of the Society of Antiquaries of Scotland 121:117–173. Campbell, E. 2000. The raw, the cooked, and the burnt. Archaeological Dialogues 72(2):184–198. Campbell, E., R. Housley, and M. Taylor. 2004. Charred food residues from Hebridean Iron Age pottery: Analysis and dating. Pp. 65–85, In R. Housley and G. Coles (Eds.). Atlantic Connections and Adaptations: Economies, Environments, and Subsistence in Lands Bordering the North Atlantic. Oxbow, Oxford, UK. Cerón-Carrasco, R. 1999. “The fish remains from Dun Vulan, South Uist”. Pp. 118–134, In M. Parker-Pearson and N. Sharples (Eds.). Between Land and Sea: Excavations at Dun Vulan, South Uist. Sheffield Academic Press, Sheffield, UK. Childe, V.G. 1935. The Prehistory of Scotland. Kegan Paul, London, UK. Crawford, I.A. No date. The West Highlands and Islands, a View of 50 Centuries: The Udal (N. Uist) Evidence. Great Auk Press, Cambridge, UK. Crawford, I. 2002. The wheelhouse. Pp. 274–281, In B. Ballin Smith and I. Banks (Eds.). In the Shadow of the Brochs: The Iron Age in Scotland. Tempus, Stroud, UK. Cummings, V., and A. Whittle. 2004. Places of Special Virtue: Megaliths in the Neolithic landscapes of Wales. Oxbow, Oxford, UK. Literature Cited Angus, S. 2001. The Outer Hebrides: Moor and Machair. White Horse Press, Cambridge, UK. Armit, I. 1990a. Monumentality and elaboration: A case study in the Western Isles. Scottish Archaeological Review 7:84–95. Armit, I. 1990b. Brochs and Beyond in the Western Isles. Pp. 41–70, In I. Armit (Ed.). Beyond the Brochs: Changing Perspectives on the Atlantic Scottish Iron Age. Edinburgh University Press, Edinburgh, UK. Armit, I. 1991. The Atlantic Scottish Iron Age: Five levels of chronology. Proceedings of the Society of Antiquaries of Scotland 121:181–214. Armit, I. 1992. The later prehistory of the Western Isles of Scotland. British Archaeological Reports. Archaeopress, Oxford, UK. Armit, I. 1996. The Archaeology of Skye and the Western Isles. Edinburgh University Press, Edinburgh, UK. Armit, I. 1997a. Architecture and household: A response to Sharples and Parker Pearson. Pp. 254–265, In C. Haselgrove and A. Gwilt (Eds.). Reconstructing Iron Age Societies. Oxbow (Monograph 71), Oxford, UK. Armit, I. 1997b. Cultural landscapes and identities: A case study in the Scottish Iron Age. Pp. 248–254, In C. Haselgrove and A. Gwilt (Eds.). Reconstructing Iron Age Societies. Oxbow (Monograph 71), Oxford, UK. Armit, I. 1998. Re-excavation of an Iron Age wheelhouse and earlier structure at Eilean Maleit, North Uist. Proceedings of the Society of Antiquaries of Scotland 128:255–271. Armit, I. 2002. Land and freedom: Implications of Atlantic Scottish settlement patterns for Iron Age land-holding and social organisation. Pp. 15–26, In B. Ballin Smith and I. Banks (Eds.). In the Shadow of the Brochs: The Iron Age in Scotland. Tempus, Stroud, UK. Armit, I. 2003. Towers in the North: The Brochs of Scotland. Tempus, Stroud, UK. Armit, I. 2006. Anatomy of an Iron Age Roundhouse: The Cnip Wheelhouse Excavations, Lewis. Society of Antiquaries of Scotland, Edinburgh, UK. Armit, I. and A. Braby. 2002. Excavation of a burnt mound and associated structures at Ceann nan Clachan, North Uist. Proceedings of the Scoiety of Antiquaries 132:229–258. Armit, I., and B. Finlayson. 1992. Hunter-gatherers transformed: The transformation to agriculture in North- West Europe. Antiquity 66:604–676. Armit, I., and D.W. Harding. 1990. Survey and Excavation in West Lewis. Pp. 71–107, In I. Armit (Ed.). Beyond the Brochs: Changing perspectives on the Atlantic Scottish Iron Age. Edinburgh University Press, Edinburgh, UK. Ballin Smith, B. 1994. Howe: Four millenia of Orkney prehistory. Excavations 1978–1982. Society of Antiquaries of Scotland Monograph Series Number 9. Society of Antiquaries of Scotland, Edinburgh, UK. Barber, J.W. 2003. Bronze Age farms and Iron Age farm mounds of the Outer Hebrides. Scottish Trust for Archaeological Research (Monograph 4). Edinburgh, UK. Journal of the North Atlantic 32 R. Rennell 2015 Special Volume 9 Hamilton, S., and R. Whitehouse. 2006a. Three senses of dwelling: Beginning to socialise the Neolithic ditched villages of the Tavoliere, southeast Italy. Iberian Archaeology 8:159–184. Hamilton, S., and R. Whitehouse. 2006b. Phenomenology in practice: Towards a methodology for a “subjective” approach. European Journal of Archaeology 9(1):37–71. Harding, D.W. 2004. The Iron Age in Northern Britain. Routledge, London, UK. Harding, D.W., and T.N. Dixon. 2000. Dun Bharabhat, Cnip, an Iron Age settlement in West Lewis. Calanais Research Series No. 2. University of Edinburgh, Edinburgh, UK. Harding, D.W., and S. Gilmour. 2000. The Iron Age settlement at Beirgh, Riof, Isle of Lewis: Excavations 1985–1995. Calanais Research Series No. 1. University of Edinburgh, Edinburgh, UK. Haselgrove, C., and R. Pope. 2007. Characterising the Earlier Iron Age. Pp. 1–23, In C. Haselgrove and R. Pope. (Eds.). The Earlier Iron Age in Britain and Beyond. Oxbow, Oxford, UK. Hedges, J.W. 1987. Bu, Gurness ad the Brochs of Orkney. BAR Series, Oxford, UK. Henderson, J. 2000. Shared traditions? The drystone settlement records of Atlantic Scotland and Ireland: 700 BC to AD 200. Pp. 117–154, In J. Henderson (Ed.). The Prehistory and Early History of Atlantic Europe. Proceedings of the 4th European Association of Archaeologists Annual Meeting. British Archaeological Reports, Oxford, UK. Hill, J.D. 1989. Re-thinking the Iron Age. Scottish Archaeological Review 6:16–24. Hill, J.D. 1995. Ritual and Rubbish in the Iron Age of Wessex. Archaeopress, Oxford, UK. Hingley, R. 1984. The archaeology of settlement and the social significance of space. Scottish Archaeological Review 1:16–23. Hingley, R. 1995. The Iron Age in Atlantic Scotland: Searching for the meaning of the substantial house. Pp. 184–194, In J.D. Hill and G. Cumberpatch (Eds.). Different Iron Ages: Studies on the Iron Age in Temperate Europe, British Archaeological Reports, International Series 602. Archaeopress, Oxford, UK. Hingley, R. 1996. Ancestors and Identity in the Later Prehistory of Atlantic Scotland: The re-use and reinvention of Neolithic monuments and material culture. World Archaeology 28:232–243. Hingley, R. 1999. The creation of later prehistoric landscapes and the context of reuse of Neolithic and Earlier Bronze Age monuments in Britain and Ireland. Pp. 233–242, In B. Bevan (Ed.). Northern Exposure: Interpretive Devolution and the Iron Age in Britain. University of Leicester, Leicester, UK. Hingley, R. 2005. Brochs as markers in time. Pp. 97–105, In V.E. Turner, R.A. Nicholson, S.J. Dockrill, and J.M. Bond (Eds.). Tall Stories? 2 Millennia of Brochs. Shetland Amenity Trust, Lerwick, UK. Cummings, V., C. Henley, and N. Sharples. 2011. The Chambered cairns of South Uist. Pp. 118–134, In M. Parker Pearson (Ed.). From Machair to Mountains: Archaeological Survey and Excavation in South Uist. Sheffield Environmental and Archaeological Research Campaign in the Hebrides. Volume 4. Sheffield University Press, Sheffield, UK. Cunliffe, B. 1978. Iron Age Communities in Britain, 2nd Edition. Routledge, London, UK. Dockrill, S.J. 2002. Brochs, Economy, and Power. Pp. 153–162, In B. Ballin Smith and I. Banks (Eds.). In the Shadow of the Brochs: The Iron Age in Scotland. Tempus, Stroud, UK. Dunwell, A., M. Johnson, and I. Armit. 2003. Excavations at Geirisclett chambered cairn, North Uist, Western Isles. Proceedings of the Society of Antiquaries of Scotland 113:1–33. Edwards, K., and B. Brayshay. 1996. Late glacial and Holocene vegetational history of of South Uist and Barra. Pp. 13–26, In D.D. Gilbertson, M. Kent, and J. Grattan (Eds.). The Outer Hebrides: The Last 14,000 Years. Sheffield Academic Press, Sheffield, UK. Edwards, K., and B. Brayshay. 2000. The vegetational history of Barra. Pp. 310–318, In K. Branigan and P. Foster (Eds.). From Barra to Berneray: Archaeological Survey and Excavation in the Southern Isles of the outer Hebrides. Sheffield Academic Press, Sheffield, UK. Edwards K.J., Y. Mulder, T.A. Lomax, G. Whittington, and K.R. Hirons. 2000. Human environmental interactions in prehistoric landscapes: The example of the Outer Hebrides. Pp. 13–32, In D. Hooke (Ed.). Landscape: The Richest Historical Record. Society for Landscape Studies, Supplementary Series 1. Available online at http://www.landscapestudies.com/ index_files/HistoricalRecord.htm. Farr, H. 2006. Seafaring as social action. Journal of Maritime Archaeology 1:85–99. Fitzpatrick, A.P. 1997. Everyday life in Iron Age Wessex. Pp. 73–86, In A. Gwilt and C. Haselgrove (Eds.). Reconstructing Iron Age Societies. Oxbow, Oxford, UK. Fojut, N. 1982. Towards a geography of Shetland brochs. Glasgow Archaeological Journal 9:38–91. Fossit, J.A. 1996. Later Quaternary vegetation history of the Western Isles of Scotland. New Phytologist 132:171–196. Foster, S. 1989. Analysis of spatial patterns in buildings as an insight into social structure. Antiquity 63:40–50. Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Blackwell Publishers, Oxford, UK. Giles, M., and M. Parker Pearson. 1999. Learning to live in the Iron Age: Dwelling and praxis. Pp. 217–232, In B. Bevan (Ed.). Northern Exposure: Interpretive Devolution and the Iron Ages in Britain. Leicester Archaeology Monographs. Leicester, UK. Hamilton, J.R.C. 1956. Excavation at Jarlshof, Shetland. Her Majesty’s Stationery Office (HMSO), Edinburgh, UK. Journal of the North Atlantic R. Rennell 2015 Special Volume 9 33 Ingold, T. 2000.The Perception of the Environment: Essays in Livelihood, Dwelling, and Skill. Routledge, London, UK. Ingrem, C. 2012. Resource exploitation: The sea. Pp. 157–158, In N.A. Sharples (Ed.). Norse Farmstead in the Outer Hebrides: Excavations at Mound 3, Bornais, South Uist. Oxbow Books, Oxford, UK. Lake, M., P. Woodman, and S. Mithen. 1998. Tailoring GIS software for archaeological applications: an example concerning viewshed analysis. Journal of Archaeological Science 25:27–38. Lawson, B. 2004. North Uist in History and Legend. John Donald Publishers, Edinburgh, UK. Llobera, M. 2003. Extending GIS-based analysis: The concept of visualscape. International Journal of Geographic Information Science 1(17):1–25. MacDonald, K. 2008. The role of the past in the Bronze and Iron Ages of the northern and western islands of Scotland. Unpublished Ph.D. Thesis. Sheffield University, Sheffield, UK. MacKie, E. 2006. The roundhouses, brochs, and wheelhouses of Atlantic Scotland c. 700 BC–AD 500: Architecture and material culture. British Archaeological Reports. Archaeopress, Oxford, UK. McGrail, S. 2001. Boats of the World from the Stone Age to Medieval times. Oxford University Press, Oxford, UK. Mills, C.M., I. Armit, K.J. Edwards, P. Grinter, and Y. Mulder. 2004. Neolithic land-use and Environmental Degradation: a study from the Western Isles of Scotland. Antiquity 78:886–895. Merleau-Ponty, M. 1962. Phenomenology of Perception. Routledge, London, UK. Mowat, R.J.C. 1996. The Logboats of Scotland. Oxbow Books, Oxford, UK. Noble, G. 2007. Harnessing the waves: Monuments and ceremonial complexes in Orkney and beyond. Journal of Maritime Archaeology 1:100–117. Noble, G., T. Poller, J. Raven, and L. Verrill (Eds.). 2008. Scottish Odysseys: The Archaeology of Islands. Tempus, Stroud, UK. Parker Pearson, M. 1996. Food, fertility, and front doors in the first millennium BC. Pp. 17–132, In T. Champion and J.R. Collis (Eds.). The Iron Age in Britain and Ireland: Recent Trends. Sheffield University Press, Sheffield, UK. Parker Pearson, M., and C. Richards. 1994. Ordering the world: Perceptions of architecture, space, and time. Pp. 1–37, In M. Parker Pearson and C. Richards (Eds.). Architecture and Order: Approaches to Social Space. Routledge, London, UK. Parker Pearson, M., and N. Sharples. 1999. Between Land and Sea: Excavations at Dun Vulan. Sheffield University Press, Sheffield, UK. Parker Pearson, M., N. Sharples, and J. Symonds. 2004a. South Uist: Archaeology and History of a Hebridean Island. Tempus, Stroud, UK. Parker Pearson, M., N. Sharples and J. Symonds. 2004b. The archaeological landscape of South Uist. Pp. 28– 47, In Atlantic Connections and Adaptations: Economies, environments and subsistence in land bordering the North Atlantic. Oxbow Books, Oxford, UK. Parker Pearson, M., and M. Zvelebil. 2014. Excavations at Cill Donnain: A Bronze Age Settlement and Iron Age Wheelhouse in South Uist. Oxbow Books, Oxford, UK. Pred, A. 1984. Place as historically contingent process: Structuration and the time-geography of becoming places. Annals of the Association of American Geographers 74(2):279–297. Rainbird, P. 2007. The Archaeology of Islands. Cambridge University Press, Cambridge, UK. Renfrew, C. 1979. Investigations in Orkney. Reports of the Research Committee of the Society of Antiquaries of London No. 38. The Society of Antiquaries of London, London, UK. Rennell, R. 2008. Exploring “everyday” places in the Iron Age landscapes of the Outer Hebrides. Pp. 43–60, In O. Davis, N. Sharples, and K. Waddington (Eds.). Changing Perspectives on the First Millennium BC: Proceedings of the Iron Age Research Student Seminar 2006. Oxbow, Oxford, UK. Rennell, R. 2009. Exploring places and landscapes of everyday experience in the Outer Hebridean Iron Age: A study of theory, method, and application in experiential landscape archaeology. Unpublished Ph.D. Thesis. University College London, London, UK. Rennell, R. 2010a. Islands, islets, experience, and identity in the Outer Hebridean Iron Age. Shima: The International Journal of Research into Island Cultures 4(1):47–64. Rennell, R. 2010b. 2010 excavation of a wheelhouse at Sloc Sabhaidh, Baile Sear. Unpublished SCAPE Report. Available online at http://www.scapetrust.org. Rennell, R. 2012. Landscape, experience, and GIS: Exploring the potential for methodological dialogue. Journal of Archaeological Method and Theory 19(4):510–525. Ritchie, W. 1966, The post-glacial rise in sea level and coastal changes in the Uists. Transactions of the Institute of British Geography 39:79–86. Ritchie, W. 1979. Machair development and chronology in the Uists and adjacent islands. Proceedings of the Royal Society of Edinburgh 77:107–122. Ritchie, W. 1985. Inter-tidal and sub-tidal organic deposits and sea-level changes in the Uists, Outer Hebrides. Scottish Journal of Geology 21:161–176. Scott, W.L. 1935. The chambered cairn of Clettraval, North Uist. Proceedings of the Society of Antiquaries of Scotland 69:480–536. Scott, W.L. 1947. The problem of the brochs. Proceedings of the Prehistoric Society 13:1–37. Sharples, N. 1984. Excavations at Pierowall Quarry, Westray, Orkney. Proceedings of the Society of Antiquaries of Scotland, 114:75–125. Sharples, N. 2006. The First (permanent) Houses: An interpretation of the monumental domestic architecture of Iron Age Orkney. Journal of Iberian Archaeology 8:281–305. Sharples, N. 2012. A late Iron Age farmstead in the Outer Hebrides: Excavations at Mound 1, Bornais, South Uist. Oxbow Books, Oxford, UK. Journal of the North Atlantic 34 R. Rennell 2015 Special Volume 9 Sharples, N., and M. Parker Pearson. 1997. Why were brochs built? Recent studies in the Iron Age of Atlantic Scotland. Pp. 254–265, In A. Gwilt and C. Haselgrove (Eds.). Reconstructing Iron Age Societies. Oxbow, Oxford, UK. Smith, H., and J. Mulville. 2004. Resource management in the Outer Hebrides: An assessment of the faunal and floral evidence from archaeological investigations. Pp. 48–64. In R.A. Housley and G. Coles (Eds.). Atlantic Connections and Adaptations: Economies, Environments, and Subsistence in Lands Bordering the North Atlantic. Oxbow Books, Oxford, UK. Tilley, C. 1994. Phenomenology of Landscape: Places, Paths, and Movement. Berg, Oxford, UK. Tilley, C. 2004. The Materiality of Stone: Explorations in Landscape Phenomenology. Berg, Oxford, UK. Wheatley, D. 1995. Cumulative Viewshed Analysis: A GIS-based method for investigating intervisibility, and its archaeological application. Pp. 171–186, In G.R. Lock and Z. Stancic (Eds.). Archaeology and Geographical Information Systems: A European Perspective. Taylor and Francis, London, UK. Endnotes 1The viewshed models were generated in GRASS GIS using a 10-m resolution DEM (www.edina.co.uk/digimap) as base data for the topographic defined line-ofsight calculations. The continuous viewshed models and cumulative viewshed models were generated using the command-line program r.cva (see Lake et al. 1998 for description). For the continuous viewsheds, view point and target mask were set to a 1500-m zone around each site, and the viewing distance to 3000 m, reflecting the maximum distance between cells within the local landscape area. Viewer height was set to 1.7 m. These models have been previously termed total viewsheds and have been applied within archaeology specifically by Llobera (2003). For the cumulative viewshed models the -f flag was used within the r.cva program, in order to specify visibility to (rather than from) the identified site locations. Binary maps of site locations were used as a target mask and a region 1500 m around the site was used as the viewpoint mask. Viewer height was set to 1.7 m. The heighten viewsheds were generated using the command-line program r.los, using viewer heights of between 1.7 m and 10 m above the ground surface.