Towers in the north
While reconstruction drawings can give some sense of how impressive the timber roundhouses might once have been, it is perhaps only with the broch towers of the north and west that we can gain a real sense of the visual impact of such buildings
18 Vie broch lower of Mousa in Shetland is one of the best preserved prehistoric buildings in Britain, standing close to its original height at around 13m tall.
and of the central role of the roundhouse in Iron Age life. Yet brochs and duns have tended to be studied in isolation, long regarded as defensive retreats from the petty warfare and piracy thought to have plagued regions remote from the mainstream of the Celtic world. It is only in recent years that brochs and duns have become widely recognized as houses, albeit stoutly built and defensible ones, rather than simply as occasional refuges.
Broch towers
In their final form, broch towers represent masterpieces of roundhouse architecture. Mousa in Shetland, the archetypal broch tower, still stands 13.3m (43ft) tall, its grim, grey outer shell punctured only by a single door, its scale and intricacy far exceeding any more recent drystone structure (18).
Despite regional variations, broch-building involved certain well-established principles, best illustrated at Dun Carloway in Lewis where the broch structure has been neatly exposed by its partial collapse. Like all broch towers. Dun Carloway was formed of two concentric dry-stone walls, bonded by rows of slabs. These slabs formed a series of superimposed galleries linked by stairs within the walls, while openings in the inner wall gave access to upper timber floors that have long since disappeared (19).
Other architectural elaborations augment this basic design. Ledges, or 'scarcements' projecting from the inner wall, for example, would have supported timber floors and rafters. Perhaps most intriguingly, vertical rows of small openings known as 'voids' ascend the interiors of many broch towers. In the case of Mousa there are four such rows, the function of which has never been adequately explained. One common theory is that they served to ease the weight of stone above the
18 Vie broch lower of Mousa in Shetland is one of the best preserved prehistoric buildings in Britain, standing close to its original height at around 13m tall.
19 (top left) Ilie partial collapse of the broch tower o f Dun Tclve in Gleticlg has exposed the galleries that run between its walls.
20 The broch tower o f Dim Bcag in Skye with the Cuillin \ fountains forming a powerful backdrop.
21 This artist's cut-away drawing shows one possible interpretation of daily life in the brocli tower of Dun Carloway in Lewis. The ground floor is used for animal shelter and for storage, while domestic activity is concentrated on the upper floors.
entrances to cells and galleries, yet this seems an unlikely justification for the creation of what were potentially dangerous weak points in the structure of the building.
John Hope, an architect with an interest in prehistoric buildings, has suggested an explanation of the way in which these various distinctive features might have combined to produce a functioning building. In his model, the galleries would have prevented wind-driven rain and snow from percolating through the inner wall. In winter, the rows of voids would have let the warm air from the hearth circulate in the galleries, keeping them dry and warm, and protecting the living space inside the tower from the extremes of weather. In summer, the same principle would have spread heat from the sun around the building.
John Hope's model neatly solves some of the long-standing puzzles of broch design. For example, all broch towers were provided with galleries and stairs.Yet only at Mousa is there proper access all the way to the wall-head. At Dun Carloway, for example, the upper galleries are far too narrow for an adult to pass through, and were certainly not intended to give access to the roof. But in Hope's reconstruction their primary purpose was simply to let air circulate, even though some of the lower, wider galleries were probably also used to allow passage between floors.
Another important observation made by John Hope is that the inner walls at ground floor level tend to be rather less well finished than those above. At some sites, like Dun Carloway, the ground floor is markedly uneven, with large outcrops of rock projecting in places. This suggests that the main living space may have been sited at first-floor level while the ground floor was used to overwinter animals and store agricultural produce, much as has been suggested for ring-ditch houses in the south (21). Indeed, this might explain why the walls of many broch towers are solid-based, with the galleries beginning only at first-floor level, since this is the level at which the hearth would have burned.
Early Atlantic roundhouses
With their wealth of architectural detail, it is easy to see why the well-preserved broch towers have captured the imagination of archaeologists for so long. Spectacular broch towers like Mousa and Dun Carloway, however, represent the culmination of centuries of development of the Atlantic roundhouse tradition, which had its origins in rather humbler buildings: houses that were essentially stone translations of the southern roundhouse form.
Strangely, in Orkney and Shetland, later to become the broch heartlands, there seems to have been no real tradition of Bronze Age hut circle construction. Instead there was a continuation of much older, cellular building styles with their roots in the Neolithic. Small oval cellular buildings found widely across Shetland, for example, perpetuated structural forms found more than 1000 years earlier at sites like Stanydale and Scord of Brouster, and in Orkney at Skara Brae. But this apparent conservatism was to change radically from around 700 BC with the appearance of the first roundhouses.
The first identifiable ancestors of the broch towers were a series of thick-walled dry-stone roundhouses that appeared in Orkney sometime around 600 BC. Like the crannogs further south, these were mostly isolated farmsteads, the best known being that at Bu, excavated in the late 1970s. Although their walls were massively thick (some were even periodically enlarged by adding extra skins of masonry), they seem to have lacked the distinctive features, such as galleries, cells and stairs, associated with later broch towers.
The Bu roundhouse floor was divided by tall flagstones into a series of separate rooms. However, rather than the simple radial plan found in roundhouses further south, the maze of interconnecting rooms seems to reflect the cellular interiors of older buildings in Orkney and Shetland, only now squeezed inside a circular shell of masonry. Despite outward appearances, then, daily life probably carried on much as they had done before.
Yet why adopt the roundhouse form at all if it was simply to be used to cloak a traditional cellular interior? Roundhouses have significant disadvantages in the wind-chilled northern climate. For ?
example, while the cells of older buildings could be easily roofed with stone slabs or stumps of driftwood, the new roundhouses demanded more substantial timbers, not to mention rather more advanced carpentry skills.The new buildings were also substantially taller (the walls of the early roundhouse at St Boniface in Orkney standing at least 3m (10ft) high), and thus, unlike the more squat buildings of earlier times, exposed to the full blast of the Orcadian winds.
It is difficult to escape the conclusion that this wilful disregard for scarce timber, ease of construction and environmental suitability was completely intentional and that even the earliest Atlantic roundhouses represent ostentatious symbols of status.
Towards towers
Over the ensuing centuries we can begin to detect the gradual elaboration of these Atlantic roundhouses and their adoption over much of the Highlands and Islands. While probably never attaining the grandeur of the later towers, new, 'complex' roundhouses began to incorporate novel architectural features. The design of the roundhouse at Crosskirk in Caithness, for example, smacks of innovation and experiment.This structure, probably built before 400 BC, has cells and a staircase built into its wall, even though the instability" of its clay core would have prevented any tower-like superstructure.
Like the roundhouse form itself, galleries can be viewed as a regional reflection of a wider phenomenon. Galleries also appear, for example, in the broadly contemporary
22 (above) Broch villages like those at Gumess (A) ami Howe (B) in Orkney seem to represent the architectural embodiment o f social control. The cellular homes of lower-ranking families huddle in the shadow of the central tower which dominates the community, just as the earlier roundhouses dominated the farming landscape.
23 '¡here has been a long-running debate over the chronology °J
Atlantic roundhouse construction. Some authorities believed that the most sophisticated broch towers were not built until the first century BC and that many wi re even later. Others believed that architectural complexity began to appear as early as 400 BC with the development of full broch towers before 200 BC. Recent excavations at the site of Old Scatness in Shetland have now provided scientific datum evidence which supports the latter position, since the central broch tower seems to have been built significantly before 200 BC.
developed hut circles of Sutherland, while the paved ditches in southern ring-ditch houses may have served a similar function and. like broch galleries, seem to have marked off areas around the periphery of the roundhouse for some specialized use. Souterrains. curving subterranean cellars found in settlements in many parts of Scotland, may be yet another manifestation of this architectural trend (see chapter 5).
The emergence of broch towers from this background of architectural development is best seen at Howe of Howe in Orkney, the only such site yet to have been subject to near-total excavation.The first roundhouse at Howe was built around 400 BC, although even that overlay an earlier village and an already ancient Neolithic chambered tomb.This structure occupied a small enclosure and had walls at least 4m (13ft) thick. Other details, however, are hard to reconstruct, because some time before 200 BC the house was largely dismantled and incorporated within the fabric of a new, complex roundhouse.
The second roundhouse was an imposing structure with walls 3.5m (11ft) thick. It may even have been a broch tower although it did not survive high enough for us to be sure. Within the walls two staircases led to upper timber floors, while the entrance passage gave access to two small cells.Yet impressive as it no doubt was, this second roundhouse was superseded, between 200 BC and AI) 100, by the next and grandest phase in the settlement's history: the building of a full-blown broch tower with, clustered around its base, a planned village of subsidiary houses (22).
Broch villages
There were probably at least twenty broch villages in Orkney alone, and still more in Caithness (22). Probably the most impressive of all is at Gurness in Orkney where the symmetry and order of the settlement exceeds even that at Howe. But the village at Howe remains the only one to have been extensively excavated in modern times, and the only one where the time depth of the settlement can yet be fully appreciated.
The 5.5m (18ft) thick walls of the broch tower at Howe suggest that the centrepiece of the village was an exceptionally tall and imposing structure, enveloping the foundations of the earlier roundhouses. Around it was laid out a series of cellular stone buildings, each with its own yard, filling virtually all the remaining space within the enclosure. Standard fittings, such as cupboards and ovens, within each of these buildings served to reinforce the integrity of the villages design.
- 24 Gurness: broth ion rr and village from the air
Broch villages were powerful statements of social control. Whereas before, numerous landholding families had each expressed their own local independence and status through the early roundhouses, the very architecture of broch villages now graded the inhabitants according to rank, with the head family firmly ensconced at the centre of village life. It is easy to imagine how a child growing up in the cellular sprawl of the Gurness or Howe villages inevitably would have come to regard the authority of the broch-dwellers as a fixed and unchangeable part of the natural order of things. The tower would have been central to daily life, its constant maintenance, repair and re-roofing periodically reinforcing social roles and obligations among the villagers.
Gradually, it seems, society in the north and west had become more centralized and certain people, like the head families at Howe and Gurness, had become dominant over their peers. Their already rather grand houses were rebuilt as elaborate towers with the homes of dependent families gathered around the base of their walls. In chapter 8 we will see just how powerful some of these tower-dwelling families may have become.
25 Vie solid slab construction o f the internal fittings at Gurness enables us to reconstruct much of the detailed plan o f the building. In other areas these features would have been made of timber and have long since disappeared.
26 11 heelhouses derive iheir »ante front their distinetive slwpe. These plans arc front varioits excauated examples in tltc Hestern Isles:A. Sollns, Sörth l ist. Ii. Cnip, Lewis,
C. Kilplieder, Sontli l 'ist,
Wheel houses
However, broch villages did not emerge everywhere. In Shetland and the Western Isles, despite the isolated grandeur of towers like Mousa and Dun Carloway, broch villages seem to have been entirely absent. Instead, by around the last century BC. when the Orcadian broch villages were occupied, many of the less elaborate Atlantic roundhouses had tallen into ruin. In some areas they appear to have been usurped as the standard settlement form of the day by an entirely novel type of building: the wheelhouse.
Wheelhouses are so called because of their distinctive ground plan in which a series of regularly spaced stone piers radiates from the centre like the spokes of a wheel (26). These buildings are common in the Western Isles and are also found in Shetland, where a particularly fine series of wheelhouses was built around the abandoned 'complex' roundhouse at Jarlshof. Remarkably, however, despite more than a century of intensive archaeological exploration, none are yet known in Orkney. Indeed it seems that the areas where broch villages are found and those where wheelhouses occur may have been mutually exclusive.
The recent excavation of a well-preserved wheelhouse settlement at Cnip in Lewis has shown something of the architectural skill of the wheelhouse builders (27).The main wheelhouse at Cnip was built by first lining the sides of a great pit in the coastal sand dunes with a single skin of walling. Next the narrow stone pillars or piers (the spokes of the wheel) were set in place, rising from a single stone's thickness to support corbelled stone roofs over each of the cells. This was an intricate operation involving enormous skill in dry-stone building. Yet it was not approached in a wholly pragmatic way. During the building process numerous offerings, such as pottery vessels, animal skulls and joints of meat, were carefully placed within the fabric of the walls and below the floors (see chapter 7).
Once the stone-built parts of the building were complete, a timber roof was erected over the central area. Even in a small wheelhouse like that at Cnip, this roof rose to around 6m (20ft) above the hearth that burned in the middle of the house. While from the outside little more than a simple conical roof would have been visible, poking above the sand dune surface, this would have concealed a refined and accomplished architectural design.
Yet. unlike broch towers, wheelhouses were eminently practical as well as monumental buildings. The short span of the central area meant that no long timbers were needed for the roof. Because they were usually dug into sandhills or the ruins of older buildings, most wheelhouses were sheltered, warm and well insulated.
While broch towers were outwardly monumental, an unwelcoming edifice of featureless masonry, wheelhouses were designed to be impressive on the inside. It is easy to imagine how imposing the elegant piers and high roof would have appeared to a visitor emerging from the long narrow
- -':> .«At
entrance passage into the hearth-lit interior. Indeed, the nature of wheelhouse architecture suggests that people were now welcomed in each other's homes and implies rather more social integration than had been evident for many centuries.
This openness, along with the social order inherent in contemporary broch villages, perhaps reflects the emergence of more stable and cohesive societies throughout the north and west by the first century BC. Farmers in the islands clearly had some confidence that their homes were unlikely to be attacked, or their lands taken by force. It is hard to see how land rights and the social order could be guaranteed without the existence of some overriding authority, presumably vested in a social elite. It is even possible, as we shall see in chapter 8, that political power over all of these island groups rested with a handful of individuals: perhaps Orcadian chiefs or kings resident in broch villages like Gurness.
27 Four stages in the construction of the wheelhouse at Cnip on the west coast of Lewis:
1. Building stone is stacked in a great pit dug into the coastal sand dune.
2. '¡lie lower walls and piers are built up quickly (before the edges of the pit subside) and joined by lintels at around shoulder height. Offerings are placed in the walls and floor.
3. Hadi bay is corbelled over to form a solid ring o f masonry. Lintels cover the long entrance passage.
4. Finally the stone ring and lintels are capped with clay and turf to render them watertight and a conical thatched roof is raised over the open central area.
Post a comment