iron age

5.9 Atlantic stone-built roundhouses: sequence, subdivision and interpretations

Armit’s sub-division of Atlantic roundhouses from Turner et al. 2005

The broch has long been a dominant feature in the study of the Scottish Iron Age, and its classification and development has excited much debate (for historiography, see MacKie 2002, 27-44; Romankiewicz 2011, 15-21).

Armit's terminology of Simple and Complex Atlantic Roundhouses (Armit 1991, revisited in Armit 2005b) is an attempt to rationalise into a related typological sequence the numerous sites known mainly from non-intrusive survey of settlement mounds that reveals only limited architectural detail. This also aimed to bridge the regional gap between things called brochs in the north and things called duns in the west, the terminological difference hiding many clear similarities (here Harding's (1984, 218-219) split of duns into those which could be roofed (dun-houses, less than c.15m) and those which could not is very useful).

The Northern Isles are considered a likely candidate for early development by most researchers (e.g. Armit 2003). Some regions (notably Orkney) appear to show a typologically clear developmental sequence from fairly simple, though sometimes substantial, roundhouses (e.g. Early Iron Age structures at Bu, Quanterness, Calf of Eday, Pierowall Quarry, Howe), through increasing architectural complexity (including intramural chambers & galleries, upper staircases, inner wall-face voids, scarcement ledges) to broch towers. Howe, in particular, shows a clear sequence of increasing architectural complexity and scale over a prolonged period of time from the Early Iron Age through the Middle Iron Age (Ballin Smith 1994; cf MacKie 1998), although there are problems with comparing poorly preserved wall foundations of earlier truncated roundhouses with later, better-preserved structures on the site. The Western Isles lacks clear parallels for the simpler Early Iron Age roundhouses that are found in the Northern Isles, and whether or not this is a genuine absence requires more work. For those seeking a western origin for this complex architecture, parallels can instead be drawn to other architectural types (block-houses, galleried duns and semi-brochs; e.g. MacKie 1992, 2002, 2008). Early dates from Old Scatness, Shetland have reinvigorated the debate (Dockrill et al 2006; MacKie 2010), which takes the absolute dating of 'true brochs' (showing complex drystone architecture) back to the third or fourth centuries BC (Dockrill et al 2006). This places a lot of evidential value on the dating of a single site, and there remains a need for broader pictures of the development of a series of sites to give a better view.

The diffusionist paradigm utilising architectural and artefactual typologies has been, and continues to be, very influential in Atlantic Iron Age studies. Further reassessment of, and sustained innovative studies of, material culture are required in order to assess and where necessary update the assumptions that have been inherited from previous work (see Atlantic Architecture & Portable material culture section below). Connectivities within the Atlantic zone have been stressed in recent synthetic reviews (Cunliffe 2001; Henderson 2007a ), and the mechanisms and meanings of this deserve further work.

In recent years there has been a move away from detailed analysis and debate concerning architectural detail and consequential definition of broch status (but see Armit 1997 a & b, Sharples & Parker Pearson 1997) This move has, for better or for worse, been in favour of attempting to address more intimate scales of human activity and practice, seeking to define the functionality of brochs; often post-structuralist, post-processual approaches have been adopted, highlighting social and symbolic factors.

In order to understand the experiences of occupants and visitors to Atlantic roundhouses, and the capacities that the people and the buildings had to exert influence and power over social relations, near and far, it is necessary to examine and understand the detail of architecture. The perennial question as to how high a particular building was is essential to any appreciation of its setting and impressiveness in the landscape. An appreciation of the three-dimensional space available within such buildings is also key, and not readily grasped from small-scale plans.

  1. The development of the complex architecture of Atlantic Scotland remains an active area of debate; new approaches to existing data provide new perspectives (e.g. Romankiewicz 2009, 2011), but the impact of the dating evidence of Scatness stresses the prime need for more, reliably dated sequences
  2. The investigation of the social motives behind the construction of such massive structures has seen important work in recent decades, but remains a key area of debate.

Chronology, temporality and biography

As with the dating of other aspects of the Iron Age (see theme 9.3) the issue of the 'plateaux' on the radiocarbon curve that reduces all dates within a wide bracket to a common blandness has presented problems. AMS coupled with Bayesian analysis are beginning to resolve these, and a range of other scientific dating techniques are becoming more widely applied, such as OSL dating and archaeomagnetism, although they are far from routine.

Hingley (2005) has suggested that substantial roundhouses created and marked a particular temporality, or sense of time amongst past communities. Duration and endurance as a mark and 'qualification' for status are also factors: a remarkable social change may be implied by the adoption of 'permanent' dwellings that were designed to outlive by far their builders (Sharples 2006) - although the idea of very long duration has been challenged (Cowley 2003). Sharples suggests the adoption of 'permanent' dwellings is a response to environmental degradation - an idea that requires demonstration. A closer reading of the fine resolution (including soil micromorphology) of floor deposits and occupational layers in tandem with more precise absolute dating may help to resolve the question of continuity or discontinuity of occupation.

The potential long endurance of buildings raises issues of lineage and inheritance (Armit 2005a) and the life-cycles of buildings have been emphasised as important aspects of their social use (Sharples 2005). The end of substantial circular buildings and the development of the generally smaller structures that succeed monumental Atlantic roundhouses have been charted (see Gilmour 2000 for the West, Hedges 1990 and Smith 1990 for the North). Exactly when and why this occurs remains a very important topic for investigation. Changes in social organisation are presumed to be the likely cause, but this requires demonstration (see theme 5.4).

  1. The important recent work on issues such as building longevity and inheritance has opened up new areas to consider, and emphasised that there are still many fresh perspectives to consider with this well-known group of material.

Use, Activity and Deposition

It is frequently assumed that Atlantic roundhouses indeed functioned as 'houses for living in' with little reference to evidence for activities that took place within them. One rich resource for interpretation is the series of soft deposits that the hard shells of Atlantic architectural spaces protect and preserve, although Armit (2006) is admirably realistic about the complex series of processes that will have lasted centuries and served to complicate and truncate the floors and floor-deposits as well as confusing any single, coherent and contemporary pattern of activity and deposition (see more generally theme 5.2 above). At Scalloway (Sharples 1998) it was argued that sudden conflagration and collapse of the organic roof led to the preservation of floor deposits frozen at a particular point in the annual cycle of activities and tasks undertaken inside. It was inferred from the evidence that overwintering animals were accommodated on the ground floor while humans domiciled in upper storeys. A similar conflagration was argued for the remains at Bharabhat, Lewis (Church 2002). Dockrill (2002) has suggested that broch towers provided a major role in the management of the local economy as centres for the redistribution of cereals.

How the generality of Atlantic roundhouses were actually used, however, remains an important research question, including whether there is a wider range of functions than the simply domestic. It is inherently unlikely that they all functioned in a similar way and provided a recognised and stable role in terms of the agricultural cycle of tasks and processes and the social continuum of changing status and activity. Research is required as to whether any standard patterns of activity can be established and whether, if this proves to be the case, these vary locally or by region.

As noted earlier, interpretations of building use should be more integrated, with excavated evidence, artefactual, ecofactual and other scientific techniques being drawn together.

Atlantic Landscapes and Housescapes

The landscapes of the Atlantic Iron Age have long been part of archaeologists' thinking; however, landscape was an implicitly assumed background to the narrower focus of study of sites and artefactual typologies. Fojut (1982, 2005a) pursued the theoretical dimensions of broch landscapes within the socio-economic sphere. A territorial model in which brochs sat within an agriculturally viable unit of land (including the coast) was developed. Dockrill and Bond (2009) have explored the relations between people and the landscape, through investigating the production and curation of soils by inhabitants of Atlantic Iron Age houses in a model emphasising the ecological marginality of Atlantic Scotland. Similar relationships between people and resources have been explored in the Western Isles (Cerón-Carrasco et al 2005). The relationship of substantial roundhouses to other contemporary buildings should also be considered. In Orkney and Caithness fairly extensive extramural complexes of buildings exist around a central 'broch-tower' (Howe phase 5 & 6); this can be misleading, as some are later accretions (e.g. most of the village at Gurness; MacKie 1994), but this is a settlement trajectory not taken in the Western Isles.

The landscape dimension needs to be addressed more explicitly. In many areas of Atlantic Scotland, although there is a lot of detailed information about monumental roundhouses and their economic and environmental relationships within landscapes, there is still a need to do more work on the political, social and symbolic aspects that almost certainly existed in the relations of Atlantic roundhouses with the landscape, other Atlantic roundhouses, and "non-Atlantic roundhouse" sites (Sharples & Parker Pearson 1997). While the extremely high density of Atlantic roundhouses/brochs in some landscapes (e.g. Rousay and Evie, Orkney; the Glenelg brochs, Highlands; the Keiss brochs, Caithness) may represent the sequential foundation of brochs over time, it may also have implications for the establishment of inheritance, lineage and generational development. Research into these dense clusterings of dramatic monuments would be very useful in investigating the complexity of inter-site relationships, successional or not, in the Atlantic Iron Age.

Why Atlantic roundhouses/brochs in the Western Isles, and the Northern Isles and Caithness respectively, have such a different relationship with non-broch forms in each area could be further explored. Parker Pearson has suggested that the nature of working the land in agricultural terms in each region was quite fundamentally different (Parker Pearson 2004) but does this environmentally /ecologically determinist approach represent the full story? Why does Orkney appear not to have typical wheelhouse structures ?

There is a need to do more work on political, social and symbolic aspects of the relations of Atlantic roundhouses with the landscape, other Atlantic roundhouses, and other settlement forms

Research into areas with dense clusterings of brochs would be very useful in investigating the complexity of inter-site relationships, successional or not

Atlantic Architecture & Portable Material Culture

In parts of Atlantic Scotland portable material culture is quantitatively very rich and diverse. Mackie (2008) has raised issues of the neglect of artefacts in recent syntheses (although aspects have been addressed, eg Topping 1987; Hallén 1994; Harding 2000a, 17-27; Heald 2001; Smith 2002). This material is still often dealt with in excavation reports as a series of parallels and provenances with unusual or exotic material picked out for most discussion. The ‘mundane’ bulk of objects need to be examined for the information they can yield. This material should be approached in an integrated fashion, placing it centrally to the interpretation of structural sequence and the lives, strategies, and identities of the Iron Age inhabitants. As Smith (2002) has pointed out even some fairly basic questions surrounding Atlantic portable material culture are unanswered and require sustained programmes of research (e.g. the fundamentals of dating and function).

In order to establish a more sophisticated and useful analysis of portable material culture in general, studies of Atlantic Scottish artefacts need to be encouraged. In particular new approaches to synthesising and integrating portable and architectural material culture have to be developed and made widely available. A start has been made to this process (e.g. Sharples (1998) at Scalloway, Shetland), although this shows its experimental nature as it makes the report very hard to use. The idea, of integrating specialist reports, is sound, and needs to be developed with further innovation, while continuing to present the basic information in a way that others can re-interrogate.

It has been suggested that the Middle Iron Age investment in impressive monumental architecture represents an emphasis upon corporate/communal (though not necessarily non-hierarchical) identities in contradistinction to the succeeding Late Iron Age period with its increased emphasis upon personal adornment, a clear formal tradition of individual burial and the construction of much more modest architectural spaces (e.g. Sharples 1998, 2003). However, there is a range of small items of portable material culture (brooches, pins, glass beads, bangles, finger and toe rings, combs etc) from the earlier period that together may represent the marking-out of individual persons and bodies (Hunter 2007b, 289-90). In addition, Roman portable material culture in Middle Iron Age contexts in Atlantic Scotland may also have served partly to fulfil this expression of personal identity.

Many aspects of Atlantic material culture merit fresh synthesis (see also theme 4.4) - little apart from pottery has seen detailed study, and pottery itself still has much to yield

Approaches to publication need to develop the ideas behind reports such as Scalloway, but not neglect the need for presentation of the material in a way others can use for different ends.

5.8 The transmission of architectural ideas over space and time

The development of the complex drystone architecture of the Atlantic has seen exensive discussion (see theme 5.9). The role of maritime links is of prime and indisputable importance here, enabling the transmission of knowledge and architectural preferences within social contexts that remain uncertain. This whole area remains one of active debate.

Edins Hall, Scottish Borders © RCAHMS

The adoption of exotic architecture has often been seen as an indication of construction and use by high-status occupants. The so-called 'southern brochs' are an excellent example of this - Macinnes' model of a network of high-status sites with occupants controlling a prestige goods economy associated with the redistribution of Roman goods is still widely accepted (Macinnes 1984). More could be done on the background to this phenomenon - there are other stone houses in the southern Scottish Iron Age, often termed duns (e.g. Stanhope (Peeblesshire), Castlehaven (Kirkcudbrightshire), Castlehill Wood (Stirlingshire);  RCAHMS 1967, 157-8; Barbour 1907; RCAHMS 1963, 81). How do these fit into the model of exotic influence? And how many more such brochs can be anticipated? The Buchlyvie broch was an unprepossessing mound before excavation (Main 1998), while the recently-discovered broch at Castle Craig, Auchterarder gave no hint of any surface presence (DES 2011, 144-5). Are all southern brochs similar, or should the Galloway ones be seen as an integral part of the Atlantic world, rather than an introduced innovation (Henderson 2007a , 165-66; Cavers 2008, 16-17)? And do they all date to the Roman Iron Age? The lack of Roman finds from some extensively-excavated examples (e.g. Edin's Hall; Dunwell 1999) raises questions over the suggested tight, Roman Iron Age chronology.

Whitslade roundhouse and souterrain, Scottish Borders © RCAHMS

Other architectural styles or concept are widespread in space and time. Souterrains vary in date from the Late Bronze Age/Early Iron Age in the Northern Isles to the Roman Iron Age south of the Forth, with a presumed floruit in the last centuries BC and first two centuries AD (Armit 1999; Miket 2002). There are differences in construction and dating across the areas of occurrence, but similarities in conception, situation and material assemblages imply links in terms of their function and behaviour. Composite ritual and storage functions (Henderson 2007a , 142-7) have been argued although such a composite functional interpretation may not be sustainable for all areas (see Dunwell & Ralston 2008 on Angus souterrains versus Carruthers on Orcadian examples).

Ring-ditch houses may be a further example of longevity accompanied by gradual transfer over territory. They appear to be present for up to a millennium north of the Tay before they are documented south of the Forth, (although there is the possibility of a visibility or research bias here, and the nature of the ring-ditch and its formation remains a key question; see theme 4.2). A parallel issue is the re-use or re-invention of crannogs across much of the first millennia BC and AD (and into the medieval period).

It is also clear that the movement of these different building traditions was accompanied by the independent growth of local 'sub-styles' wherever they passed. There are good examples of local distinctiveness within these architectural 'streams'. For example, wheelhouses in Shetland can be argued to be distinct in design from those in the Western Isles (Harding 2009 , 112-4). In turn both areas are quite distinct from Orkney, where classic wheelhouses are absent, but where there are buildings with radial partitions (e.g. at Howmae), suggesting that wheelhouse traits did penetrate the archipelago and so the total distinction drawn may be overemphasised (Harding 2006 , 74; Henderson 2007a , 160). At Scatness in Shetland wheelhouse use extends into the second half of 1st millennium AD, albeit with significant structural differences to earlier forms (Dockrill 2003, Dockrill et al. 2010).

Complex Atlantic roundhouses on Barra and North Uist are also considered distinct from their counterparts on South Uist (Armit 1997 a & b, 2002; Sharples and Parker Pearson 1997). These arguably represent local variations (perhaps autonomous or hierarchical) among the societies of the Western Isles that adopted Atlantic roundhouse architecture. The Atlantic roundhouse need not have resulted from a homogenous cultural background or social structure (Romankiewicz 2011).

  1. The mechanisms behind the spread of these phenomena remain hotly debated.
  2. Are the Galloway brochs late 'bastard forms' (Cowley 2000, 174) or part of the Atlantic mainstream (Henderson 2007a, 165-66; Cavers 2008, 16-17)?
  3. How do lowland brochs fit into their settlement landscapes, especially in relation to other stone architecture?