by Paul Elliott
6
The Roundhouse
When several [Celts] dine together, they sit in a circle; but the mightiest among them, distinguished above the others for skill in war, or family connexions, or wealth, sits in the middle…
Athenaeus 4.151
Visiting a reconstructed roundhouse is a special event; standing inside that circular domestic space you can imagine the repetitive sound of grain being ground on the millstone, the hissing and crackling of the fire, perhaps even the snores of family members, sleeping in the dark. I’ve been lucky enough to spend several weekends with friends living in a roundhouse at the Ryedale Folk Museum, North Yorkshire. Of a typical Iron Age size, the house sits between a wheat field and a small wetland area. A tree-lined field opposite the entrance has held Soay sheep and Iron Age pigs, at different times.
Cooking, working, and sleeping inside the house gave a flavour of what it must have been like to call a roundhouse home. Each experience provided an insight: cooking in poor light, sleeping in straw-lined beds while a mouse scurried past, grinding enough flour for the daily bread, and managing the fire to maximise light and flame while minimizing smoke.
Any discussion of prehistoric cookery has to address the importance of the Celtic roundhouse, for it was not just kitchen and dining hall, but also storeroom and farmhouse. From around 2000 BC roundhouses began to replace the earlier rectangular structures that had been built during the Neolithic. At the same time, circular Beaker barrows began to proliferate and it may have been their design that influenced the new shape of housing. Within 300 years, the arrangement of Bronze Age houses become almost standardized, with many examples from across the British Isles displaying similar floor plans, probably reflecting a shared belief system or way of life. British farmers felt an affinity with the roundhouse that lasted up until the arrival of the Saxons and Angles, a period of history that spans 2,400 years (equivalent to almost eighty human generations).
The Structures
Roundhouses varied in diameter, but a large number of excavated examples are somewhere between 5 and 7 metres. Typically, a low circular wall was made of wattle and daub and above it stood a conical roof of thatch or turf that was supported by a spoked arrangement of timbers. Small houses were able to rely on this spoked-wheel design of rafters to support the weight of the roof. For larger buildings, more substantial support was needed, typically in the form of a ring of sturdy posts within the floorplan of the roundhouse. With the weight now borne by both the outer wall and the inner ring of posts, this type of house is termed a double-ring structure. Triple-ring houses have also been discovered that had recourse to yet another ring of posts within the centre of the house, allowing for a greater diameter and a commensurate increase in living space.
Not all of the outer walls were built of wattle fences and strengthened with daub, some were constructed from banks of earth while others were formed from drystone; methods and materials varied from one region to another according to local tradition and available resources. One of the remarkable features of roundhouse architecture is the orientation of their single, timber-framed doorways. The vast majority of entrances face south-east or east, which correlates to the direction of the morning sun. A great deal has been written on the association of the roundhouse with a sun cult or a cosmological belief in a sun-orientated world view. Other archaeologists see the orientation of the doorway more as a practical matter. There were no smoke holes or chimneys in the roof and no windows in the walls; consequently, these buildings would have been dark and smoky places in which to live.
There are few isolated examples of excavated roundhouses, more commonly they are found in clusters of between two and ten, indicative of a small hamlet or a community of related families. While some of these communities were surrounded by earthen banks and ditches or by timber fences, there are plenty of examples that remained ‘unenclosed’. The settlement at Mucking, which overlooked the Thames, dates from the Late Bronze Age and there, an impressive double-banked circular enclosure surrounded two roundhouses. Meanwhile at Blackpatch, Sussex, five roundhouses from the same period sat within an enclosure marked out by a stout fence. Dartmoor, rich in building stone, yielded a stone-walled enclosure at Shaugh Moor. Inside this 50-m-diameter settlement sat five stone-walled houses, some complete with cobbled yards. Not all of these structures served as dwellings, some are smaller than others and, at Shaugh Moor, pottery finds and the analysis of phosphate levels left behind by animal manure indicate that while the largest house was used for human habitation, the others were a mixture of outhouses and byres. Even then, phosphate research showed that the family home had been shared with farm animals from time to time.
Bronze Age roundhouses at Winterbourne Stoke, Wiltshire. (S. Garrett)
Late Iron Age farmstead, from Berwick Down, Wiltshire. This is a ‘classic’ single-phase farm with roundhouse, granaries, enclosure ditch, spacious farm yard, and working pits. (G. J. Wainwright)
Light and Dark
Without windows, the main source of light for the inhabitants of a roundhouse would have been the door. Some houses were constructed with a porch or roofed extension over the doorway that provided much needed shelter from wind and rain. A door of wood or leather would have been closed at night, but left open during the day. Post-built structures (those with at least one inner ring of supporting posts) may have had a gap between the thatched roof and the top of the outer wall, which would have allowed additional light to spill into the house in the early morning. This gap was hidden from unwanted bad weather by the overhanging eaves.
Obviously the best lit area of the house was directly in front of the doorway. However, the movement of the sun throughout the day meant that the illuminated areas of the house changed. Let us say that the doorway faced due south; the early morning sun shone from the south-east and would have lit up the area directly in front of the doorway as well as the north-western portion of the room. By mid-morning the area lit by the sun had shifted from left to right and now the shaft of sunlight illuminated both the centre of the room and backspace, opposite the doorway. Shadows lengthened in the afternoon, and the sun’s light shifted over to the north-eastern side of the roundhouse. By late afternoon or evening, a fire would have been needed to provide any useful light. Throughout the day the darkest parts of the house were the left and right peripheries at the side of the room, as well the roof-space, above.
Discussion of light patterns within the roundhouse is not theoretical, but practical. There were many activities that had to be conducted indoors. Spinning and weaving were carried out, grain had to be ground and baked as bread. Cloth was sewn, clothes were mended or made, meat and vegetables had to be prepared and cooked over the fire, flint tools needed fabricating, animal hides were cleaned, treated and stretched, baskets were woven, and bone, antler, and horn was worked to create tools and other artefacts. Finally there was a need for socializing, resting, and sleeping.
If the weather permitted, it is likely that many activities were conducted outside the house, but wet or cold weather would quickly have driven these activities inside. Harsh winter weather will have forced the inhabitants to shut the door and so rely on light from the fire for their daily activities. Lamps of butter or animal fat might have been used to supply light, but this was a food resource ‘going up in smoke’. More than likely, prehistoric farmers relied on flames from the fire for light during the evening.
Entranceways into roundhouses generally focussed on the east and south-east to maximise the amount of sunlight available. Additionally, with a prevailing wind in Britain that blows from the west-south-west, the east facing doorways provided a level of shelter. Differences in latitude can be traced in the ground plans of many roundhouses, with more northerly houses in the Highlands of Scotland facing south-east, those in southern Scotland and northern England facing mainly east-south-east and roundhouses further south orientated more towards the east. The further nort
h the communities lived, the fewer daylight hours were available to them in winter. Doorway orientations were tailored to maximise the provision of winter daylight.
The Hearth
No one would deny that the hearth formed the centre of life in a prehistoric roundhouse, and it is therefore fitting that in most houses it is precisely where the hearth can be found—at its centre. Some examples are found elsewhere, but the overwhelming majority are located at the centre point, or (less commonly) slightly forward, towards the door. In practical terms this is the most obvious location, since it provides an even distribution of light and heat. In addition, the risk of an accidental fire is reduced, since a fire near to the periphery might easily spread to the wattle wall, or the thatched roof, which is at its lowest as it passes over the outer wall. Conversely, the highest point of the roof, its apex, sits exactly over the centre of the roundhouse and sparks travelling upwards lose their heat well before they reach the vulnerable thatch.
Visitors to the house at Ryedale Folk Museum often comment on the lack of a chimney or smoke hole. This design feature is based on several decades’ worth of experimental building. Some of the first replica roundhouses that were constructed in the 1950s included such a feature. However, in the same way that a fierce draught can be created in a small house by opening both the front and back doors simultaneously, the smoke hole drew up the hot air from the fire and the sparks with it. More than one of these experimental houses was burned down after the thatch caught fire. Instead, the smoke and hot air rises into the conical roof space, with smoke seeping slowly out of the thatch. Sparks that are carried up enter this oxygen-less zone and are extinguished with virtually no risk to the roof. On waking up within the roundhouse in the early hours of the morning, I noticed that the air was very smoky and that clothing and bedding smelt strongly of smoke. All the occupants were coughing, despite having little problem with smoke the day before. I suspect that the dying fire and the lack of any upward air current allowed the smoke that had built up within the roof space to fall back into the living area. Was the thatch too thick? Should it have been thinner in order to let the smoke out more rapidly?
Hearths are usually nothing more than circular or rectangular areas marked out by high-sided stone slabs. These retained the ashes and prevented burning logs from rolling across the roundhouse floor. It is not a camp fire; the prehistoric hearth was a major area of domestic activity and had to be big enough to support several cook pots at once. Likewise, it had to be big enough to allow three or four people to stir food, move pots, and tend the fire. When my colleagues and I put on a cooking demonstration for the public, one of us kept the fire going, while another made bean cakes and was frying them on a flat stone in one corner of the hearth. Meanwhile another cook baked eggs in the ashes and yet another tended the spit, upon which a duck was being slowly roasted. We discovered that different types of fire were required simultaneously, with hot ashes being enough to bake the eggs, but flames needed to roast the duck. Our bean cake fryer needed only very moderate heat to bring his flat cooking stone to the temperature that he needed. In this way the hearth served many purposes at once, a little like the Victorian kitchen range that used a single coal fire to heat a water boiler, an oven, and one or more stoves.
Baskets of dry wood and/or charcoal would have been kept adjacent to the hearth perhaps along with a number of benches. These benches might be fixed or moveable, but experience has shown that they are not placed too close to the fire. Access is always needed to manage the fire as well as whatever might be cooking on it, and so a decent ‘kneeling space’ is best left around the edge of the hearth.
Internal Arrangement
Rarely do roundhouses survive in the ground as anything more than infilled post-holes and stained areas of soil. It is difficult, then, to determine whether a roundhouse may or may not have had a second floor, used perhaps for sleeping and for storage. This is an idea first developed by Diana Reynolds in 1982. Double-ring structures do have the structural integrity to support a second floor and approximately 20–30 per cent of their total volume lies above the ring beam up inside the roof, which essentially becomes ‘dead space’. Were full, or partial, attics used? Double and triple ring structures were certainly strong enough to bear the weight, but it is interesting to note that although smaller (7–9 metre diameter) single-ring houses were quite able to support a roof without any internal posts, some were found with a second inner ring of posts. This was overbuilding, and whether it took place for structural security, status, or the support of an upper floor, we cannot know. Wooden steps or ladders would have allowed access to this upper floor.
Finds recovered from prehistoric roundhouses indicate that different activities took place in different areas of the house. The well-lit space around the hearth and in front of the doorway seems to have been the focus of craft activities and food preparation. Leather and bone-working, sewing and weaving all require decent light levels. Talking, sleeping, resting, and storage need very little light and were therefore relegated to the periphery of the roundhouse, to the outer edges and the backspace, behind the hearth. It was here that some measure of privacy could be had. This front-versus-back division is often observed amongst the modern roundhouse dwellers of today, people like the Galla and Dorze from Ethiopia, as well as the Kipsigis from Kenya.
Pits were a common feature of roundhouses and were used, it seems, mainly for storage. These pits were commonly dug at the front of the house, or else at the back of the structure. In a house from South Shields, pits and grain processing debris were found between the hearth and the doorway, while bedding material was found at the back. Meanwhile, a house at Tormore on the Isle of Arran contained a pair of post holes near the doorway, indicative of a vertical weaving loom and there was evidence of flint knapping and both crop and wood processing in the area. Tormore revealed that grain and timber had been stored towards the back of the roundhouse.
Roundhouse using cultures today place their furniture around the periphery of the house, leaving the central area fairly clear. Studies have shown that this cluttered periphery is typically around 1.7 metres in width and covers more area than the central space. In one of these 10-metre-diameter houses, the periphery accounted for around 62 per cent of the total floor space, with the central area making up the remaining 38 per cent. We can assume similar patterns of use in British prehistory, where the central area around the hearth and the well-lit front space were important for work, while the periphery was utilised mainly for living and storage. At Greaves Ash, Northumberland, a roundhouse was discovered that featured a paved periphery, clearly marking out the importance of this space to the inhabitants. At the Ryedale Folk Museum, two double beds have been constructed in the Iron Age roundhouse at the eleven o’clock and one o’clock positions (as one enters the house at the six o’clock position). Both wooden beds are straw-filled wooden boxes, able to hold two or three sleepers with ease and covered by a platform supported on four posts. This platform serves as a place for the storage of baskets.
The circular periphery, running around the inside of the roundhouse wall, would have held beds, straw-filled mattresses (‘paillasses’), firewood, equipment and tools, food, shields and weaponry, and also animals. Although circular stables and byres are a common feature of Celtic farms, there is evidence that some roundhouses were inhabited by both humans and cattle. In southern Scotland and parts of England phosphate analysis has shown that cattle were very likely to have been kept inside many of the region’s roundhouses over winter. Studies of phosphate levels within the peripheral areas of roundhouses at Dalnaglar in Perthshire, Shaugh Moor in Dartmoor, Morl y Gerddi in Gwynedd, and Lintchie Gutter in Lanarkshire indicated the stalling of animals there. Families may have slept within other parts of the periphery or, alternatively, on a second floor. This may not represent the entire herd, rather it would have been younger animals and their mothers, or animals that needed to be kept close for milking.
<
br /> Around half of the modern roundhouse using cultures today stall their animals (mainly cows, calves, and goats) inside their houses. Larger roundhouses like the double and triple-ring structures, are more likely to be used in this way, i.e. bigger houses were more likely to accommodate some of the family’s animals. Typically the animals are just stalled overnight and kept behind a partition, but one tribe, the Kenyan Kipsigis, let their sheep settle freely while the family sleeps on an upper floor. The practice of bringing animals into the house is common throughout Africa, just as it was in Medieval Britain. The advantages of doing this in winter are raised temperatures within the roundhouse and easy access to the animals for the collection of dung, or for the milking of cows and goats.
Being warm and dry, roundhouses were the perfect place for the storage of equipment, tools, firewood, and foodstuffs. People using roundhouses today store their belongings and food supplies in a variety of ways: in string bags hung from the rafters, on racks suspended from the roof, on a raised attic or upper floor, on shelves or pegs mounted on the outer wall, or on hooks driven into the inner posts. Warm air and smoke rising up into the roof space meant that prehistoric farmers could use this area to store food on a medium or long term basis. Smoke will have inhibited the growth of bacteria, prolonging the life of meat and cheese and the dry warm air will have kept crops (like grains, beans, and peas) dry and free from damp. Firewood was most likely dried out under the over-hanging eaves of the house (as the Kenyan Kikuyu do today), or beneath a purpose-built wood shelter. At regular intervals, perhaps even daily, firewood was transferred indoors so that it was thoroughly dry before burning.