The Use and Reuse of Stone Circles

by Richard Bradley

  Such arguments apply to nearly all the sites reported here. They are equally relevant to those examined during earlier fieldwork. The present project supplies several examples of this process. The centre of the circle at Hillhead became a cremation pyre in the later Bronze Age, and the existing monument at Croftmoraig was occupied by a timber roundhouse during the same period. In turn it was replaced by an oval setting of monoliths enclosed by a wall. Something similar happened on the Hill of Tuach where a setting of six monoliths was surrounded by a considerable bank and ditch. It is possible that the central court of the ring cairn at Waulkmill was reused at about the same time, but the monument achieved a greater significance during the Roman Iron Age when it became a cemetery.

  There is comparable evidence from the monuments excavated during earlier phases of fieldwork. Both the passage graves at Balnuaran of Clava contained secondary deposits of cremated bone dating from the Late Bronze Age (Bradley 2000, 119–20). The interior of the ring cairn at Newton Petty included the same material (Bradley 2000, chapter 6), and so did the recumbent stone circle at Tomnaverie (Bradley 2005, 157–58). The earthwork enclosure at Lairg was reused in the Early Iron Age, as a posthole in its entrance has a date of 751–241 BC (Bradley 2011, 153).

  The Pictish period was equally important. The principal ring cairn at Balnuaran of Clava was associated with a cremation burial which dates from the later first millennium AD (Bradley 2000, 56–59), and a decorated standing stone was discovered outside the henge at Broomend of Crichie (Bradley 2011, 10). These observations form part of a wider pattern.

  Later Bronze Age reuse

  Richard Bradley

  This section concerns the ‘later’ Bronze Age. The term is used deliberately as some of the dating evidence is imprecise and spans the conventional distinction between the Middle and Late Bronze Ages. Several radiocarbon dates cut across the period boundary, while the kind of pottery associated with renewed use at particular monuments was made in both these phases.

  The secondary reuse of older monuments took at least two forms (Bradley 2011, 169–70). There were sites at which new structures of stone or wood were erected during the Middle Bronze Age, and others where the only evidence consists of cremated bone. This distinction may be too precise, for the latest stone structures – the kerb cairns inside the larger circle at Temple Wood, those at Claggan and Stromtoller, the monument at Cairnwell, and possibly the inner stone setting at Croftmoraig – were all associated with deposits of burnt bone. Such structures were built between about 1600 and 1000 BC, and most of them between approximately 1500 and 1200 BC. These monuments were considered in some detail in Chapter 7.

  Deposits of cremated bone have a longer history and their chronology extends well into the Late Bronze Age. It is clear that the reuse of older sites continued after the last buildings had been erected. They were associated with burnt human bone until about 800 BC. After that time there could have been an interval before any new structures were erected, and those at Laikenbuie, Clava and Forvie may not have been built until the Early Iron Age. Although these monuments share architectural elements in common with older features, their construction need not have been a continuous process.

  Figure 8.1. Secondary cremation pyres at the Hillhead and Old Keig stone circles. Information on Old Keig from Childe 1934 and Welfare 2011.

  Figure 8.2. Outline plan of Moncreiffe with details of the two trenches containing burnt material and cremated bone. They are compared with the secondary features excavated in the centre of Hillhead.

  Most of the cremated bone from secondary deposits at stone monuments dates from the Late Bronze Age. This is particularly true at Balnuaran of Clava, Newton of Petty and Tomnaverie (Bradley 2000 and 2005). All three monuments have several radiocarbon dates, but there is similar evidence from sites where less material has been analysed. They include stone circles at Hillhead, Fullerton, Castle Fraser, Garrol Wood, Gownie, Old Keig, Croftmoraig, a four poster at Fortingall and the henge monument at North Mains (Bradley 2011, 169–70) (Fig. 8.1). The pottery associated with secondary cremations at Loanhead of Daviot might be of the same age (Kilbride-Jones 1935). It is not always possible to work out how these deposits had formed. Early sources record the presence of charcoal, burnt soil and even small settings of boulders, but it is rarely possible to establish whether the monuments were reused as pyres. Hillhead provides a rare exception, and here the secondary features were in the centre of an existing ring cairn. That may form part of a wider pattern, as similar positions were associated with deposits of burnt material at Loanhead of Daviot, Newton of Petty, Tomnaverie and Old Keig – all of them sites with detailed excavation reports (Bradley 2005, 100–102).

  Something similar may have happened at Moncreiffe where Margaret Stewart (1985) identified two stone-lined trenches inside the stone circle (Fig. 8.2). They contained charcoal and burnt bone, and there was another just outside the monument. In this case there is a problem for she interpreted these features as furnaces associated with metalworking. This is entirely plausible as a chisel and several pieces of scrap metal came from the excavation. With them were tuyeres and part of a crucible. Tomnaverie may be relevant here as a Middle Bronze Age axe is recorded from the site, although its context is not known (Coles 1964, 132). Like Moncreiffe, the Early Bronze Age site at Loanhead of Daviot was reused by a smith, for it was associated with the mould for making a Late Bronze Age sword (Kilbride-Jones 1935; 1936). The recumbent stone circle of Tillycoutrie illustrates a similar association because a hoard dating from the end of the Bronze Age was found just outside the monument (O’Connor 2007).

  Two of the recently excavated monuments were enclosed at the end of their periods of use. The inner stone setting at Croftmoraig was contemporary with the construction of the perimeter wall. The dates from this enclosure overlap with the terminus post quem for the rebuilding of the henge monument at Pullyhour (Bradley 2011, 133). Just as its entrance had been partly blocked, a gap in the outer boundary at Croftmoraig seems to have been closed by a decorated stone which may have been of some antiquity.

  During the Late Bronze Age a small stone circle on the Hill of Tuach was also enclosed, but in this case by a massive earthwork. It was constructed at a time when there is evidence of settlement in the surrounding area. Such a late date had been unexpected, but is paralleled by new work at another Scottish henge, Quarrywood at Elgin (Gordon 2012). Again the place had a longer history, for geophysical survey has located a stone circle in the interior. Radiocarbon dates from the bank place its construction in the later Bronze Age. The enclosure on the Hill of Tuach can also be compared with the circular earthworks considered in Chapter 7, for its entrance appears to have been blocked. In common with a similar earthwork at Migdale, this part of the monument was associated with lithic artefacts. There were standing stones in equivalent positions at both sites.

  A final characteristic of some of the monuments was the way in which they were sealed by a deposit of quartz when their period of use was over. That is shown at Hillhead and most probably at Croftmoraig where numerous pieces of worked and broken quartz were spread across the interior. There are references to a similar practice in the antiquarian literature, but the closest parallels come from north Wales and the Island of Lewis. At the Druids’ Circle large boulders, lumps and smaller fragments of quartz were scattered inside an embanked stone circle and covered the positions of several Early Bronze Age burials (Griffiths 1960). At Olcote near to Calanais a kerb cairn built between 1700 and 1500 BC was covered by a similar deposit, but in this case the evidence is more complicated since worked quartz had been associated with an older settlement on the same site and some of the artefacts were embedded in the core of the monument. Even so, the excavators concluded that when the cairn had been built it was covered by pieces of ‘crudely smashed quartz’ (Warren and Neighbour 2004). The excavation produced more than 15,000 pieces. That is similar to what happened at Hillhead where the total was close to 25,000.

  Stone cir
cles and roundhouses

  Richard Bradley

  It may be no accident that the earthwork enclosure at Tuach was built at a time when there was settlement in the vicinity. A similar situation arose at Lairg where a small circular enclosure associated with two cremation burials was identified in the midst of a group of later prehistoric buildings (Bradley 2011, 147–50). Their proportions were so similar that they could not be told apart on the basis of surface evidence. That is not surprising as the dimensions of the last stone circles and henge monuments were like those of domestic dwellings. This observation is important in view of some recent discoveries.

  One of the most thoroughly excavated recumbent stone circles was at Strichen (Phillips et al. 2006) (Fig. 8.3). When the project took place in the 1980s it had an unexpected result. The stone setting there was obviously earlier than the remains of a substantial timber building. That ran counter to the assumptions considered in Chapter 1, and the problem became even more severe when samples from the wooden structure produced dates in the Early Iron Age: at one sigma 600–400 BC and 800–390 BC. Some years later a similar building was excavated beside another recumbent stone circle, the Candle Stane (Fig. 8.4). It presented still more difficulties (Cameron 1999). It obviously dated from the Iron Age, but the site saw two periods of activity, between about 700 BC and 400 BC, and between 400 BC and 100 BC. Despite its lengthy history, it was not associated with any artefacts and was situated on top of a hill that might have been unsuitable for everyday occupation. Moreover, its footprint seemed to echo that of the stone setting alongside it. Perhaps it had played a specialised role in the past. More recently, the earthwork of another later prehistoric roundhouse has been identified inside a stone setting of the same type at Loudon Wood (Welfare 2011, 161).

  Figure 8.3. The evidence for later timber buildings inside the stone circles at Croftmoraig and Strichen. Information from Phillips et al. 2006; Welfare 2011; Piggott and Simpson 1971; and an unpublished drawing in NMRS (Canmore ID 24891 – DP 137220).

  In the light of these observations the timber structure at Croftmoraig took on a new significance. It had always posed problems, but its character became much more apparent in the light of a previously unknown plan by one of the excavators, Stuart Piggott, which forms part of the archive of the 1965 fieldwork held by NMRS. The central area includes three distinct elements, all of which predated the erection of the oval setting of monoliths. There was a shallow hollow, described by the excavators as a ‘ditch’, two parallel slots which had obviously held upright posts, and a ring of postholes. The relationship between the post sockets and the ditch was difficult to resolve, but another problem was that the published site plan showed every subsoil feature in the same format. Piggott’s drawing records their depths, suggesting that there had originally been two circles of wooden uprights, one replacing the other. Additional elements were depicted in the published survey, but they were much shallower and it is unlikely that they had ever held timbers. The bona fide postholes represent most of the outline of a building with a porch to its east. They define the positions of roof supports or the course of the outer wall; the question is discussed in Chapter 10. Traces of this circuit were absent in only two places: in a disturbed area recorded in the 1965 excavation, and where there are fallen monoliths. The ‘ditch’ followed the same course as the posts and extended around roughly half the perimeter. Since 1965 similar structures have come to light at settlements where they are described as ‘ring ditch houses’. At one time they were thought to be a purely Iron Age phenomenon, but more recent excavations have shown that their chronology extended back as far as the Middle Bronze Age. Far from being a distinctive architectural ‘type’, the form of these buildings may result from the stalling of animals or other activities. The hollowed area is the result of erosion and is not a structural feature. The timber structure at Croftmoraig had the same characteristics and is similar to domestic dwellings at Lairg and Kintore (McCullagh and Tipping 1998; Cook and Dunbar 2008).

  Figure 8.4. Plans of a recumbent stone circle and a large roundhouse at the Candle Stane. Information from Cameron 1999 and Welfare 2011.

  In some ways the identification of a roundhouse in the centre of the Croftmoraig stone circle raises the same problems as the buildings at Strichen and Loudon Wood. Why were timber structures of a kind associated with settlements erected in such unusual locations? They would have been difficult to access since the standing stones remained in place, and the sites themselves were poorly suited to everyday activities. At Strichen a decorated stone, which may have covered an Early Bronze Age burial, was incorporated in the foundations of the Iron Age building (Phillips et al. 2006). Even if these relationships can be explained in mundane terms, that can hardly apply to the roundhouse at Croftmoraig, for, after an initial phase in which the building was repaired, its place was taken by an oval setting of monoliths. They were laid out on the same footprint, but their long axis was directed towards the southwest. The perimeter wall at Croftmoraig was constructed at about the same time. It is unhelpful to make a categorical distinction between ritual and daily life, but this is enough to suggest that the building (and presumably its occupants) played a specialised role.

  The discovery of timber buildings inside older stone circles has further implications which recall important themes in contemporary research.

  Matthew Grove (2010) has studied the sizes of Irish stone circles and has considered how many people are likely to have used them. His paper has been so influential that it is quoted as a case study in a textbook on archaeological theory (Hodder 2012, 108–109). It contends that human communities are organised into units of similar extent to those recognised by the anthropologist Robin Dunbar among other primates. At the same time the sizes of human groups are reflected by the scale of the monuments they build. Thus the smallest structures were able to accommodate single households, whilst the largest might be associated with larger gatherings. Grove distinguished four thresholds in the areas of Irish stone circles and associated them with events conducted on an increasing scale. Unfortunately these particular monuments are poorly dated, and his findings can be interpreted in another way. Like their Scottish counterparts, the smallest monuments were among the latest examples. Far being contemporary with one another, the scale of these buildings changed over time. The largest date from the Late Neolithic period, and the smallest from the Middle or Late Bronze Ages. It was this last group that Grove associates with the household, and that could be why a few of them contain wooden structures similar to domestic dwellings.

  Figure 8.5. Distribution of earlier prehistoric monuments with evidence of Roman Iron Age reuse.

  Roundhouses have also been studied by Rachel Pope (2007), who challenges the popular notion that they were directed towards the south or east because of the significant role played by the sunrise in ancient systems of belief. She rightly argues that such alignments are by no means exact and may have changed over time. In any case there could have been practical reasons for illuminating the interiors of these dwellings. The timber buildings associated with Scottish stone circles are no exception, but in this case there is a different problem, for the orientations of the wooden structures are at odds with those of the standing stones, which face between south and southwest. As argued in Chapter 7, they can be directed towards the dark side of the sky (Bradley 2016; Henty 2014). This raises an intriguing problem. If the orientations of stone circles of this kind were influenced by cosmological principles, surely the same should apply to the wooden houses erected inside them. Again it would be unwise to distinguish between the requirements of ritual practice and those of domesticity.

  This idea is consistent with a growing body of evidence that during later prehistory ritual activities focused on the settlement and the house rather than specialised monuments. During the late second and early first millennia BC the construction of stone circles, cairns and henges lapsed. It was at the same time that more durable houses were constructed. The excavators of t
he settlement at Kintore observed that it was then that specialised deposits became more closely associated with the domestic world (Cook and Dunbar 2008, 365–69). Under these circumstances it seems hardly surprising that buildings indistinguishable from other roundhouses were erected on the sites of older stone circles. There was no longer a separate style of public architecture. The evidence from Strichen illustrates this point. To some extent the same applies to the structural sequence at Croftmoraig, but in this case the site of the timber building was commemorated by a setting of standing stones.

  Roman Iron Age activity at stone monuments in north-east Scotland

  Fraser Hunter and Richard Bradley

  An unexpected feature of the Waulkmill stone circle was the way in which it was chosen as the site of a Roman Iron Age cemetery. The evidence is compelling. A token cremation with a radiocarbon date in the third or fourth century AD was placed in the exact centre of the monument, and its position was marked by a setting of quartzite cobbles. Another cremation was a short distance outside the stone circle. To some extent the axes of the individual graves reflected the course of its perimeter. Both contained large flakes which seem to have been detached from the standing stones. In one instance they were incorporated in the filling of a grave containing a penannular brooch, gaming pieces and amulets. In the other, they were associated with a massive wooden coffin. At least two more burials were found in the nineteenth century. They seem to have been further to the east along the same low ridge as the stone circle. Although it would be easy to regard this as a special case, the antiquarian literature suggests that there were comparable deposits at other sites in north-east Scotland (Fig. 8.5).