top of page

SAF 1999 

Circular Arguments: the Archaeology of Roundhouses

Scottish Archaeological Forum Conference
23- 24 October 1999 at the University of Glasgow 



The conference reviewed the evidence, past and present, for round houses in Scotland during later prehistory. It explored our understanding of the construction, use and destruction of such sites. New approaches to the data were examined and their influence on future research strategies discussed. Methodologies with regard to the interpretation of results and reconstruction work were also addressed.



Scottish round houses: the early chapters. Inside and outside Iron Age Europe

Prof. Ian Ralston, Edinburgh University

This paper will attempt to set the scene by means of a general review of the place of circular architecture in the study of later Scottish prehistory. It will identify some of the milestones in the development of this field of enquiry, using Robert Munro's Prehistoric Scotland, published exactly a century ago, as a baseline. Some comments will be offered on the occurrence of round houses in the temperate European Iron Age outwith Britain.

The End of a Round House

Niall Sharples, Cardiff University

This lecture intends to present some new information concerning the occupation of round houses in the Atlantic seaboard and to discuss the nature of the transformations which mark the end of these houses in the Late Iron Age. The first aspect relates to the recent excavations at the site of Bornais on South Uist. These revealed a round house which has been converted, after a massive conflagration, into a sub-rectangular building. There is good evidence for the use of this round house but the principal interest lies in its conversion into a rectangular building. Such conversions are known from other sites in the Western Isles most noticeably Cnip and appears to be part of a wholesale reorganisation of domestic space that was underway during the Late Iron Age throughout Atlantic Scotland. The transformation of round houses in each region does, however, appear to be negotiated in quite different ways in each region. These transformations indicate the various ways social control is being renegotiated within each region and contrast with evidence suggesting growing cohesion within the Atlantic Province.

All around the south-west: a review of round houses in SW Scotland

Gavin MacGregor, GUARD, Glasgow University

This paper considers the current state of knowledge of round houses in SW Scotland. The interpretation of the social use of space, rather than a concern with form, is recognised as an important avenue to the understanding of round houses in this region. Therefore, although acknowledging the contribution of aerial photography, the paper focuses on the results of recent excavations. While considerable advances in our knowledge of round houses in SW Scotland have been made over the past ten years, this has largely been the result of developer funded excavation. Consequently, I stress that there is a need for further research excavation of well preserved examples in order to clarify many of the issues relating to the interpretation of round houses in the south west of Scotland.

Short and brutish? The life of a round house on Buiston Crannog, Ayrshire

Dr. Anne Crone, AOC Ltd

Excavations at Buiston crannog have revealed occupation deposits dating to the late 6th - mid 7th century AD. Settlement on the crannog consisted, at any one time, of a single round house within a defensive perimeter and evidence for the construction, duration and condition of these houses will be presented. The waterlogged conditions on the crannog have preserved a wealth of organic detail which means that we can clothe the superstructure more fully than on 'dryland' sites. The settlement has been closely dated through the dendrochronological analysis of structural timbers which has highlighted the short lifespan of the houses. The preservation of insect remains and food debris has provided insights into the conditions in which these structures were inhabited.

Off at a tangent: a review of round houses in NE Scotland

Derek Alexander, CFA, Edinburgh University

The enormous size of the study area, which extends from Fife in the south, to the Dornoch Firth in the north, obviously causes problems for such a review and perhaps inevitably variety in structural form will be the main conclusion. The nature of the database is certainly diverse. By the end of the last century, survey work had recognised the detail recoverable in the upland landscapes. This survey work has been taken to its logical conclusion in the detailed mapping undertaken by RCAHMS in NE Perthshire which has recorded one of the densest concentrations of later prehistoric settlement remains in Scotland. Further survey work in other areas, the Angus glens for example, is starting to provide similar results. Numbers of the upland sites were excavated from the 1930s-60s and some remain classic excavations - such as Scotstarvit, Fife. Dating, until recently, relied on artefacts, often scarce on round house sites, and on attempts at structural typology. More recent excavation of upland sites has produced a growing body of radiocarbon dates which confirm the time depth visible in the surveyed and excavated remains. Throughout the 1980s and 90s excavation has increased on plough truncated sites in the lowlands, most of which appeared as cropmarks, although there are notable exceptions. This increase in the data set in the lowlands provides a complimentary picture to that in the uplands.

Atlantic Scotland: evidence and controversy

Simon Gilmour, Edinburgh University

This paper is split into two related parts. The first will very briefly review several of the current controversial aspects of round houses in Iron Age Atlantic Scotland. These centre mainly on classification, chronology and function and are directly relevant to the second part of the paper: the evidence for round houses in Argyll. Although a major part of Atlantic Scotland, study of the drystone structures here has been difficult due to their classification as heterogeneous 'duns' and their perceived 1st millennium AD dating. An alternative classification system, based on that devised for the Outer Hebrides indicates that Argyll round houses are comparable in function, form and date to those elsewhere in Atlantic Scotland. Indeed their study as part of this wider continuum raises interesting possibilities in terms of Iron Age round house origins and development.

Floor formation and structured deposition: Cnip wheelhouse, Lewis

Dr. Ian Armit, Belfast University

The wheelhouse complex at Cnip, Lewis, was occupied from the last centuries BC until around the 3rd century AD, during which time it underwent a number of major structural alterations. There is evidence that, throughout the occupation of the settlement, the floors of the various buildings were kept scrupulously clean and free from the build-up of archaeologically-recoverable floor deposits. From time to time, however, this routine practice was disrupted by the deliberate creation of new floors, usually by the importation of clean white shell sand. These laid floors buried and thus preserved the pre-existing floor deposits. The vast majority of the excavated deposits within the buildings, therefore, owe their survival to these non-routine events. Functionalist explanations for the creation of these floors appear weak, and it can be argued that the burial of old floors should be seen as a deliberate and conscious act analogous to other forms of structured deposition, including foundation and closure deposits, associated with Iron Age round houses in general and wheelhouses in particular.

Round houses on South Uist

Dr. Mike Parker-Pearson, Sheffield University & Niall Sharples

This paper will focus on the Late Bronze Age/ Early Iron Age at Cladh Hallan on the machair of South Uist where recent excavation has recovered the remains of five round houses with a total of 13 intact floor surfaces. Reference will be made to an unusual group of deposits under the floor of one house whose eight floors indicate that it was inhabited for more than 500 years. This excellent preservation will enable us to reconstruct the way it was used in any one phase and over its long term sequence of occupation. The chronological and architectural sequence of the site spans at least half of the 1st millennium BC and will help the understanding of the context within which brochs and wheelhouses were later to develop in the Western Isles.

The botany of Brochs.....20 years on

Mike Church, Edinburgh University

" The botany of brochs" by Prof. Jim Dickson was published in the 1979 Scottish Archaeological Forum conference proceedings entitled "Early man in the Scottish landscape". The paper outlined the plant remains recovered from the excavations of the Iron Age complex at Crosskirk, one of the first detailed archaeobotanical investigations from a site in Atlantic Scotland. This paper will review the archaeobotanical evidence from Atlantic round house sites excavated over the past 20 years. The nature of the database will first be examined before highlighting a number of key themes concerning plant management and procurement that recur throughout the region in the Iron Age. These include the nature of arable agriculture, timber procurement and the management of heathlands. Possible research directions for the future will also be examined.

Ethnography and the roundhouse

Rachel Pope, Durham University

As part of research into the prehistoric round house in North Britain, an attempt is made to clarify the author's position regarding the use of ethnographic analogy in round house studies. The paper begins by investigating the use and abuse of ethnography in Iron Age studies and its relationship with the new narratives of the Postmodern period. The use of analogy is then considered regarding its form and validity as an archaeological technique. The paper moves on to present results from a survey of the ethnographic literature and the modern-day round houses of twenty traditional African groups. Several common aspects of the round house are identified in the ethnographic literature, factors concerned with: a) the natural environment; b) the human environment; c) socio-economic concerns; and d) ideo-cultural concerns. The paper concludes by discussing the role of ethnographic analogy in the study of the round house and how we might consider revising our approaches in the near future.

Approaches to artefacts from houses: methodologies, theories and factoids

Fraser Hunter, National Museums of Scotland

Too often there is a failure to make the most of artefacts and integrate them into wider interpretations of the site. This paper will explore two themes connected to this. It will look firstly at the question of deposition within houses, and how far current models of cosmologically structured deposition are sustainable in a Scottish context. Secondly, it will explore possible methodologies for studying assemblages, show their potential, and outline areas of research required to improve our use of artefacts.

SE Scotland: review of evidence

Andrew Dunwell, CFA, Edinburgh University

This presentation will examine the character and development of later prehistoric round house forms, focusing upon south-east Scotland but drawing in evidence from surrounding areas. Given time constraints I do not aim to present a regional settlement history, but will concentrate largely upon the round houses themselves. The review will take as a starting point the state of research as defined in a series of important volumes published in the early 1980s, and will examine the impact of subsequent research in furthering themes popular at that time and in developing new lines of enquiry. Much important new data has been gathered in recent years, principally through excavation, the significance of which has yet to be synthesised at the regional level. This evidence, along with changing perceptions of how later prehistoric societies and households were organised, combine to indicate some potential future areas of study.

Roman-period round houses and Romanisation

Richard Hingley, Durham University

It has usually been considered that round houses over the south of Britain ceased to be built fairly soon after the Roman invasion in the middle of the first century AD. Under this approach it is considered that households across the south of the province of Britannia turned away from building round houses during the first century and adopted various types of rectangular house as their basic building types. This approach fits with the dominant 'progressive' approach to the study of 'Romanisation' in Roman archaeology. Under this progressive approach, all within the 'civilised' part of the province adopted 'Roman' material culture after the conquest. As a result, the idea of the rectangular house, having been introduced to the south east, spread gradually north and west during the period of Roman rule. However, across much of the north of Britain and parts of the west the natives did not adopt rectangular architecture, thereby indicating a continuity of primitive/uncivilised behaviour in the so-called 'military zone' of Roman Britain. This paper will contend that this progressive approach is incorrect. Round houses do not die out in the south in the first century of Roman rule but continue to be fairly common throughout the period of Roman rule. The continuing tradition of round house building in the south has been ignored by Romanists as a result of the progressive interpretations inherent in Romanisation theory. It is argued that the continuity of round house building in the south requires a re-think of the dominant approach to social change in Roman Britain.

Deer Park Farms: building a round house and what to do with it

Chris Lynn, DOE, Belfast

The paper will begin with a brief resumé of the evidence for round houses predating the earthwork of the Iron Age sanctuary at Navan Fort, Co Armagh. These large buildings (diameter typically 14m) may have been used for ceremonial purposes in the second century BC. Although they were associated with 'normal' occupation material and had central hearths, they did not have internal post-rings and always occurred as sets of three concentric wall-slots used in sequence - middle, outer, inner- with the middle and outer slots in fact being robber trenches. It appears that the walls of the structures may have been renewed twice while keeping the roofs intact. The skull of a Barbary Ape was found in one of the later wall-slots.

There is a gap in the Irish evidence of at least half a millennium. The paper will continue with a brief outline of the evidence for round houses associated with raths dating from about the 7th to 9th centuries AD and note that there is evidence for a progressive change to universal use of rectangular buildings in Ireland before AD 1000.

The dating evidence is better at the beginning than the end of the period. The main bulk of the paper will explain how to build a wattle-woven house of the 7th-8th centuries AD, based on waterlogged remains of 7m diameter wicker structures found in the lower levels of a rath in Deer Park Farms, Glenarm, Co. Antrim (Scotland visible from nearby on a clear day, 22 miles away!). The pushed-over wall of one of the houses survived to a height of 2.5m. These houses, often built as figure-of-eight units, are the typical round houses of Early Christian Ireland. Differences in house design apparent in excavated dry sites can largely be explained by post-depositional factors. The houses were built with damp-proof courses, they had cavity wall insulation and were woven using a simple but effective and strong basketry technique. There seems to have been no structural istinction between 'wall' and 'roof'. Details of the houses and internal finds accord with the contemporary literary evidence and some obscure features mentioned in the literature are easily explained by the actual structures.

Realising the round house: inside Archaeolink

Hilary Murray, Archaeolink

Realising the round house at Archaeolink Prehistory Park in Aberdeenshire has had two meanings; the act of completing the reconstruction itself and the gradual glimpses of the reality of the house derived from nearly 3 years of working, cooking, eating and on occasions, sleeping in it. Such insights, although limited by the confines of my own time and experience still have some value as they relate to practical realities, such as the distribution of light and dark and their effect on functions within the building. The reconstruction, based on an unusual plan with only 4 internal roof support posts, is discussed and the validity and use of partitions, hearths and lofts, examined. Consideration is given to the fittings we might expect in such a building and to the possibility of art within the house.

Loch Tay crannogs and circularity

Dr Nick Dixon, Edinburgh University

Crannogs are generally described as 'circular' or 'sub-circular' islands or submerged mounds. While many are seen from aerial photographs, and from the plans of early excavations, to be clearly circular it is possible that others, which do not appear to be so regular, have irregular shapes because of substantial structural modifications carried out during the period of occupation. Circular structures are known to be strong and, in the case of the crannogs of Scotland, this would be a considerable benefit as they have to withstand the effects of changing water levels, wind-generated waves and significant gales blowing down long fetches of open water. It is also easier to build circular structures, rather than rectangular ones, especially when construction is being carried out in, or over, open water. Trying to measure straight lines and right angles would be particularly difficult. The high proportion of circular buildings in prehistoric Britain speaks for their popularity and use for a wide range of functions. These issues are discussed with reference to the results of survey and excavation from various sites. The issues of ease of building and strength of structure are considered with regard to the construction of a full-sized crannog built recently in Loch Tay, and the effect upon it of four years of winter gales.

bottom of page