The Use and Reuse of Stone Circles

by Richard Bradley

  His argument is not unlike the interpretation of Roman activity at older stone circles and related monuments, for it involved a similar appeal to tradition during a period when local communities were exposed to unfamiliar practices and beliefs. In that case they developed their own accommodation with a wider world and did not explicitly contest its values. Perhaps similar pressures were experienced at other times of rapid change – the Early Iron Age, for example, or the later Bronze Age – but it will take much more research to show whether the same approach to the evidence is likely to be productive.

  CHAPTER NINE

  The Extent of Variation: four stone circles in Cromar in the light of recent fieldwork

  Richard Bradley

  The recumbent stone circles of north-east Scotland occur in such large numbers that it has been difficult to interpret their distribution. The problem is discussed by John Barnatt and Aubrey Burl, who share a similar approach. Barnatt views these sites as ‘local foci’ (1989, 175–76) and Burl suggests that they were associated with small territories extending for 10–15 square km (2000, 220–22).

  There are problems with this reconstruction. The distribution of such monuments is imperfectly known. Only a year after excavation had recovered the plan of the destroyed structure at Waulkmill, a previously unrecorded example was identified at Hillhead, only 4 km away. At the same time Adam Welfare’s research demonstrated that freestanding ring cairns occurred in the same parts of north-east Scotland as the recumbent stone circles and extended well beyond their distribution towards the south (2011, chapter 1). Some were substantial structures, like Monandavan in the Howe of Cromar which was 15 m in diameter. A better preserved monument is at Milestonehill East (RCAHMS 2007, 51) (Fig. 9.1). They may have lacked rings of uprights, but that could have been the only important difference between them. Chapter 7 argued that the monoliths, recumbent stones and flankers were all a secondary development and that they were not erected until particular sites went out of use. Up to that point these monuments were little different from the others.

  What distinguished them were probably their sizes. This question was considered by James Kenworthy in 1972. The figures presented in his article suggest that in most regions it was the larger ring cairns which were more likely to be enclosed by stone circles. The smaller examples were less often modified in this way. The main source of variation was their external diameter; the dimensions of the internal court did not appear to have much influence. The most important factor could have been the amount of effort expended in building these monuments. In turn that influenced their chances of survival because the recumbent and the flankers were sometimes too large to move. Such factors affect the distribution of monuments on which Barnatt’s and Burl’s studies were based.

  There is another problem in postulating a uniform system of territories, each of them dominated by a single structure. Although recumbent stone circles are treated as a distinctive style of architecture, Welfare’s account places as much weight on their diversity as the features they shared. What is needed is a new study of the monuments within one small region. That is possible in Cromar where three stone circles have been excavated over the last two decades. A fourth has been investigated by field survey. The account that follows considers the stone circles at Hillhead, Waulkmill and Tomnaverie, together with the surface remains of the Blue Cairn at Ladieswell. These sites occupy a compact area measuring 4 × 10 km (Fig. 9.2).

  There are significant differences between the places where they were built. The positions of 72 recumbent stone circles were considered in Welfare’s survey. Their heights above sea level vary from 30 m to 362 m (Welfare 2011, 31–32). When he was writing, the Blue Cairn was the second highest member of the group with an elevation of 285 m, but the discovery of a similar monument at Hillhead changes the pattern, for its height is 345 m. Tomnaverie is much lower, at 180 m, whilst Waulkmill is close to the 150 m contour. On the other hand, the dating evidence from two of the excavated monuments is broadly consistent, although it seems as if the larger stone circle at Hillhead was built after its neighbour at Tomnaverie.

  Figure 9.1. (above) Distribution of ring cairns with and without stone circles in north-east Scotland; (below) Outline plans of the recumbent stone circle at Tilliefourie and the ring cairn at Milestone, both of have which had small courts at their centre. Information from RCAHMS 2007 and Welfare 2011.

  Important contrasts can be seen in two other ways. Both Hillhead and the Blue Cairn are located on the sides of the basin which contains the other two sites. Tomnaverie occupies a hilltop commanding a view in all directions, but Waulkmill is at one end of a low ridge overlooked by higher ground. The basin once contained a loch which was drained in the nineteenth century. Both Tomnaverie and Waulkmill would have been within a few hundred metres of the water.

  In some respects the Blue Cairn compares with Hillhead, and Waulkmill with Tomnaverie (Fig. 9.3). That is indicated by the scales on which they were built. The newly discovered monument at Hillhead has a diameter of 26 m and is one of the largest of its kind – the full range of variation extends from 30 m to 13 m but most of them are between 25 m and 15 m across (Welfare 2011, 74–76). The Blue Cairn is not much smaller, with an external diameter of 23 m. Both the monuments at lower altitudes were about the same size. The ring of monoliths at Tomnaverie was 17.5 m in diameter and the estimated diameter of Waulkmill is 18 m. Such figures can be misleading, for the most striking feature of these sites would have been the areas enclosed by the circles. At Hillhead it was 530 square m; the equivalent figure for the Blue Cairn is 414 square m. The monuments at Tomnaverie and Waulkmill were considerably smaller. In each case they enclosed approximately 250 square m – under half the estimate for Hillhead. Following Matthew Grove’s interpretation summarised in Chapter 8, each pair of monuments might have been used by a different number of people.

  Figure 9.2. The stone circles in the Howe of Cromar discussed in this chapter (contours at 100 m intervals).

  Those inside the basin were smaller and it took less effort to construct them. The examples at its outer limits must have made greater demands. That is not consistent with the model favoured by Burl and Barnett since it is the less elaborate structures which are associated with the most productive farmland. The larger ones, on the other hand, were established in marginal locations which should have supported fewer people. It may be because Hillhead was outside the expected distribution of recumbent stone circles that it escaped notice until recently.

  Despite the differences between each pair of monuments, the stone circles of Cromar were built using the same techniques. The most striking connection is perhaps their most unusual feature. It has long been recognised that the outer boundary of the Blue Cairn combines a ring of kerbstones with the circle of monoliths that would normally be built outside it. It also has a recumbent stone on the SSW, but there is no sign of the usual flankers (Welfare 2011, 15). It is as if the successive phases observed on other sites have been telescoped to form a single structure. In 2012, the same arrangement was identified during the excavation at Waulkmill. The connection between these sites goes even further. Most recumbent stone circles contain 12 or so monoliths, although there is a wider range of variation and those found to the northeast in Buchan are significantly smaller (Welfare 2011, 93–96). Two sites in Cromar depart from the usual pattern. Welfare estimates that uprights were built into the kerb of the Blue Cairn at intervals of about 3 m. That would involve the erection of up to 24 standing stones – twice the usual number in this region. In the excavated segment at Waulkmill the monoliths were about 3.5 m apart, suggesting an original total of 15. By contrast, Tomnaverie included the expected number of 12. Without more excavation it is impossible to reconstruct the original form of the monument at Hillhead.

  At Waulkmill and Hillhead there were traces of an inner court. Nineteenth century excavation at the Blue Cairn located a similar feature, but it did not occur at Tomnaverie where the summit of the hill outcropped and the mate
rial of the cairn was piled around it – that is why no inner kerb was built there. These structures had very different proportions. At Hillhead the inner court was 15 m in diameter and had an area of 176 square m. At Waulkmill, it measured only 4 m across and was 12.5 square m in extent – just over 7% of that figure. The central area within the circle at Tomnaverie was only a little larger – approximately 5 × 6 m – whilst the court found during antiquarian investigation at the Blue Cairn is said to have been 3.65 m in diameter (Welfare 2011, 314–16). It follows that the primary monuments at these four sites might have accommodated different numbers of people, and for that reason it could be misleading to treat them as a single class of monument. The largest groups could have visited Hillhead – assuming these spaces were not reserved for the dead. The site stands out compared to most of the recumbent stone circles in the wider region where the diameter of the court was normally between 4 and 6 m. There were some important exceptions. The sites at North Strone and Cothiemuir Wood had equally large internal courts (Welfare 2011, 88; Bradley 2005, 53–67). Similarly, considerable open areas are found at monuments in Buchan, but in this region a circle of monoliths was erected on top of a perimeter wall (Fig. 9.4). At Aikey Brae it happened during a secondary phase (Bradley 2005, 77–82).

  Figure 9.3. Outline plans of four stone circles in the Howe of Cromar.

  Within the Howe of Cromar the unexpected similarity between Waulkmill and the Blue Cairn has an important implication. Small-scale excavation at Waulkmill did not find any evidence that the stone sockets were deeper towards the south or southwest, where the flankers normally occur in a recumbent stone circle. On the other hand, its siting in the landscape resembles the local settings of structures of that type and it commands exactly the same kind of long distance view. Comparison with the Blue Cairn suggests another possibility, for on that site the recumbent was placed on the ground surface and was not framed by a pair of standing stones. If the Waulkmill monument was built in the same manner, there is nothing to preclude the presence of a recumbent stone which was later destroyed. Indeed, it may account for the number and size of the flakes incorporated in the fillings of the Roman graves, for they were in precisely the position where a structure of that kind would be built.

  Figure 9.4. Outline plans of two monuments in north-east Scotland with large open courts at their centre. Information from Welfare 2011.

  Tomnaverie shares some features with Hillhead. Their outer kerbs were constructed in the same way. At both sites large upright stones – some of them slabs – were wedged in position between two groups of boulders. There was also a rubble platform outside the kerb, although at Hillhead it was much slighter than its counterpart at Tomnaverie (Bradley 2005, 22–23). In other respects the structures in Cromar differ from one another. The inner court at Waulkmill must have been bounded by a kerb of upright slabs held in place by a foundation trench. This has no equivalent at the other excavated sites, although it is paralleled at Sundayswells not far to the east (Welfare 2011, 228). It is known that the Blue Cairn also had a central court, but there is no information on how it was constructed. At Tomnaverie, however, the granite summit of the hill was exposed in the centre of the monument, making it unnecessary to construct an inner kerb.

  Other characteristics are shared between some of these monuments. At Tomnaverie there was a certain amount of patterning in the selection of raw materials for use in the kerb and the stone circle. This involved their shapes and surface textures (Bradley 2005, 28–33). There was much clearer evidence that the flankers were a different colour from the recumbent stone. According to Adam Welfare (2011, 314–16), the stones used at the Blue Cairn may have been selected for their visual appearance, but the monument has not been excavated in modern times. On the other hand, there does seem to have been a striking contrast between the rocks apparently employed as flankers at Hillhead and the colour and lithology of the recumbent stone.

  Far more striking is the evidence for patterning in the surface of the primary kerb at Tomnaverie. The radial divisions in its structure do not have any equivalent in the ring cairn at Hillhead, but the concentric arcs identified during the 1999–2000 excavation at Tomnaverie could have delimited low steps in the surface of the monument. This possibility was not investigated during the previous project, but at both sites their configuration emphasises the south-western aspect of the monument. That is particularly important as pollen analysis at Tomnaverie showed that the structure was constructed in an open landscape (Bradley 2005, 46–47). Like the monument at Hillhead, it was aligned on a conspicuous mountain. In each case they could have been where people stood to observe the distant horizon and the sky.

  If Hillhead and Tomnaverie were directed towards Lochnagar or the high ground running south from Morven, the Blue Cairn and Waulkmill seem to have been orientated towards the south or SSW. In neither case is the alignment especially clear. At Waulkmill it is uncertain whether the circle had a recumbent stone at its entrance, but the distinctive peak of Mount Keen appears on the distant horizon. At the Blue Cairn, Laidieswell, it seems as if the monument faced Roar Hill across the low ground on the edge of the basin. Both places were out of sight of Lochnagar and neither was obviously directed towards a single landmark. In the reverse direction Hillhead was in a position where the summit of Mither Tap was just visible on the far horizon. That reflects its unusual position on the watershed of the two main rivers in this region. Taken together, it seems as if the stone circles of Cromar were integrated into a larger area whose limits were marked by mountains on the horizon. Their sites were directed towards the most prominent local peaks just as the monuments themselves were visible from one another.

  A number of these links connect sites of different sizes or construction. To that extent they support the argument that all four monuments belonged to the same architectural tradition. At the same time, the examples in Cromar were on the edge of the overall distribution of recumbent stone circles and this may help to explain why they had some unusual features. Two of these buildings were exceptionally large and occupied much higher ground than is usual for this kind of monument. At the same time they have other unfamiliar features. One was the entrance at Hillhead. Another was the massive cairn with a recumbent stone at its edge at Ladieswell. Further to the southeast a series of monuments close to the North Sea share this characteristic and feature a recumbent stone without any circle of monoliths (Barclay and Ruggles 1999). Another unusual structure was the stepped ring cairn at Hillhead. Still more striking is the exceptional size of the open area in the centre of that site. Although something similar is seen further to the northeast, a closer comparison may be with the earthwork of a henge. The internal area of Broomend of Crichie was a little larger – it measured 20 × 17.5 m – but despite the different ways in which they were built, both monuments are of the same order of magnitude. In fact the earthwork enclosure may have been linked to a recumbent stone circle, now destroyed, by an avenue of paired standing stones. The histories of these two monuments could have overlapped. The ring cairn at Hillhead has a terminus post quem of 2200–2000 BC. At Broomend of Crichie a similar estimate for the construction of the bank is less precise but falls between 2140 and 1740 BC (Bradley 2011, 60–61). Another link between them is that Hillhead possessed an entrance. Such unusual features suggest that structures found towards the limits of the distribution of recumbent stone circles had less in common than the others and that the people who built them were more disposed to experiment.

  How were these monuments related to the prehistoric pattern of settlement? There are four radiocarbon dates from Hillhead that suggest activity on the site some time before the monument was built. The same could well apply to the dates for the construction of Tomnaverie. In neither case was there any evidence of domestic occupation, but the discovery of what may have been a disturbed Beaker grave just outside the stone circle at Hillhead suggests that these sites already played a specialised role. So do the earliest radiocarbon dates from that site an
d Tomnaverie.

  After the excavation at Tomnaverie the wider setting of the monuments was investigated by fieldwalking (Bradley 2005, 87–92). In retrospect the results of this work are subject to two limitations. The first was unavoidable. By chance, subsequent excavation at Waulkmill found a pit associated with Carinated Bowls. Although the pottery was abundant, this feature did not contain any lithic artefacts. Since prehistoric pottery is friable, it may be that some of the occupation sites in the study area would not have been indicated by surface finds. At the same time the area examined on the ground extended from the shores of the former loch up the sides of the basin as far as the 350 m contour. In principle it included the site at Hillhead but none of the nearby fields were ploughed when the work took place. In fact the distribution of surface finds seemed to end at an elevation of about 300 m. The new project shows that this was an artefact of modern land use. With that limitation, the area that was investigated included Hillhead, Tomnaverie and Waulkmill, but did not extend to the west as far as the Blue Cairn.

  The project revealed five concentrations of artefacts, as well as 32 fields (out of a total of 86) with an above-average density of surface finds of worked flint and quartz. Almost without exception they were distributed on south-facing slopes at heights of 150 m to 300 m. The clearest concentrations were in the lower part of this range. There were few diagnostic artefacts, but most of them could have been contemporary with the use of the excavated monuments. To a large extent the distribution of surface finds avoided the position of these structures, suggesting that they had been built on the margins of the ancient landscape. The same interpretation was suggested by pollen analysis at Tomnaverie (Bradley 2005, 46–47). A similar project which focused on the recumbent stone circle at Cothiemuir Wood 20 km to the northeast produced the same result (Bradley 2005, 92–97).