Designed to enchant: the great dolmens of Neolithic northern Europe

When you look at a monument like Pentre Ifan in Pembrokeshire, what do you see? Back in 1951, writing in A Guide to the Prehistoric and Roman Monuments in England and Wales, Jacquetta Hawkes saw a massive wedge-shaped capstone almost ‘floating in the air’, held aloft by three slender orthostats. She observed that ‘the structure is in reality given a slightly fantastic air by the very narrow points on which the capstone rests; it appears improbable that so little can support so much’. Current visitors to the site, however, are met by an information board that states that ‘what we see before us today are the bare bones of a burial chamber that would originally have been covered with an earthen mound’ (phrasing that is echoed by the entry for the monument on Cadw’s website, https://cadw.gov.wales/visit/places-to-visit/pentre-ifan-burial-chamber). In ‘Megaliths, Memory and the Power of Stones’ (2010), moreover, Chris Scarre characterises Pentre Ifan as ‘the denuded chamber’ from a mound that would have been radically different in appearance from the present monument.

This view – that all dolmens are the relics of monuments designed for burial – has led archaeologists to invent a whole raft of new typological terms, including megalithic tomb, sub-megalithic tomb, demi-dolmen, chambered tomb, portal tomb, portal dolmen, Type 1 and Type 2 passage grave, and gallery grave. Some of these are further subdivided into ‘simple’, ‘classic’, and ‘devolved’ forms, while some structures have defied categorisation altogether. This shift in terminology has had the effect of turning attention away from the monument’s form to consideration of its function, with the sole apparent purpose of the structure being to create a stone chamber to hold human remains.

This bewildering array of analytical terms has gone well beyond typology, and has also been harnessed to the task of tracing cultural influences. The perceived similarities and differences between the different tomb types has been used to suggest cultural connections between the various Neolithic peoples of Britain, Ireland, Scandinavia, France, Portugal, and Spain; to suggest trade and migration patterns and routes; and to construct chronologies of monument-building and ‘Neolithisation’ (the adoption of Neolithic lifestyles). A separate strand of analysis, prevalent in the 1930s, concerned the shape and form of any covering mound – long or round – and the implications for cultural diffusion and chronology.

Not everyone has been convinced by these approaches, however. Tatjana Kytmannow, in Portal Tombs in the Landscape (2008), argued that the typological boundaries were fluid and that ‘it is rather easy to turn a simple dolmen of the Carrowmore variety [named after the Carrowmore Megalithic Complex in Co. Sligo] into a portal tomb and vice versa’. Meanwhile, Elizabeth Shee Twohig, in Irish Megalithic Tombs (2004), says that ‘in the south [of Ireland] many sites classified as portal tombs seem never to have had a portal feature’. The same can be said of Cornish ‘quoits’ which, according to Kytmannow, ‘obviously owe a lot to portal tombs [but] lack portals’.

Barbara Bender, in Stone Worlds (2007), objected to ‘the dry typologies into which megalithic structures in all their infinite subtle variation have to be squeezed’, while Tim Darvill and Geoff Wainwright, in their chapter on prehistory in the Pembrokeshire County History (2016; see CA 324), came close to rejecting all these sub-categories, along with the endless arguments over which monuments belong to which categories, offering the terms ‘raised stone’ or ‘propped stone’ as alternatives.

Dolmens on display

Now Vicki Cummings and Colin Richards, in their recently published re-examination of megalithic architecture in northern Europe (see ‘Further reading’ at the end), have taken the final step, asking us to recognise that some dolmens (their preferred term) were never encased within mounds or cairns at all – rather, their purpose was first and foremost to create a display. The authors describe dolmens in terms of wonder, enchantment, power, and dramatic effect; they are spectacular, amazing (in the true sense of causing astonishment), and more than a little bit magical – all terms that we are familiar with in the work of people interested in mysticism and the occult, but not often used by archaeologists.

Nevertheless, the authors present a well-argued case that we are doing the builders of these extraordinary monuments a disservice in not recognising dolmens as astonishing works of architecture that still have the power to make us stand in wonder some 6,000 years after they were first conceived and built. With approval, they quote Andrew Fleming, who wrote as long ago as 1973 (in ‘Tombs for the Living’, in the journal Man) that ‘it seems quite clear that these tombs, far from being merely containers for the dead, were quite deliberately designed to rivet the attention of living individuals’.

Surely, you would think, the matter should be easy enough to resolve through archaeological excavation, but unfortunately there has been little enough scientific investigation of such sites. Furthermore, when archaeologists have looked for evidence, they have found considerable disturbance from people digging in the past for fairy gold, buried treasure, or passages to the underworld – and, more prosaically, from continual ploughing, which significantly affects the ground beneath and around dolmens. Where excavation has taken place, the results are often ambiguous and difficult to interpret, as Frances Lynch discovered at Carreg Samson, Pembrokeshire, in 1968. Frances had hoped to find in situ deposits in the internal chamber area to help with dating, but instead discovered that the soil had been extensively disturbed, containing only a few smashed sherds from a single pot, some flint flakes, and a few small fragments of bone, from which it was not possible to obtain a radiocarbon date.

The authors also cite the example of Gunderslevholm, one of 364 recorded dolmens in the Zealand region of Denmark, of which 202 are covered in mounds. Did the remainder have mounds that have disappeared over time due to stone-robbing and erosion? At Gunderslevholm, the excavators found that the dolmen chamber was surrounded by a spread of boulders and stones up to 1m in height, held in place by kerb stones. Some have argued that this material represents the remains of a cairn that once covered the entire structure. Others have argued that the capstone would always have remained visible and that the mound would only have reached its base. Others still have argued that the stones exposed by the excavators once formed a platform around the orthostats and capstone, which remained freestanding and visible. A fourth possibility is that this is a multi-phase monument and that any enclosing mound could have been a later work.

Eye-catching designs?

Recent fieldwork elsewhere has looked for evidence of enclosing mounds: at Kit’s Coty House in Kent; at St Lythans (Maes y Felin) in South Glamorgan; at Carreg Coetan in Pembrokeshire; and at Trethevy Quoit in Cornwall. These investigations have uncovered intriguing evidence traces of a plinth-like ring of carefully laid stones encircling and framing the dolmen. At Trethevy, the freshly quarried greenstone was used to create a platform with larger blocks at the base and smaller broken pieces at the top – surely designed to be seen, because the colour and brightness of the greenstone would have provided a visual contrast with the dark granite slabs of the dolmen. At Carreg Coetan, the diameter of the ring (8-10m) meant that any theoretical mound would have had a very steep slope to cover the capstone completely – not impossible, but unlikely.

At Dyffryn Ardudwy, Gwynedd, the 2ft-high plinth was found to have been formed mainly of large water-rolled stones, specially selected and carried some distance. Similar low rings of rounded or angular stone have been found in Ireland and at Trollasten in southern Sweden, where some 6,000 pottery sherds and hundreds of broken flint axes (dating from the early and middle Neolithic) were found on the surface and within the stones. From this evidence, it could be argued that these plinths formed part of the architecture of the monument, and that they were designed to frame the dolmen, perhaps serving as an offering platform and functioning as a barrier that wrapped around the central dolmen to keep onlookers at a prescribed distance.

The Dyffryn Ardudwy dolmen illustrates another key point in the argument for free-standing dolmens. A second dolmen stands a short distance to the east, and this is enclosed by an extant cairn. The fact that some tombs do have substantial surviving mounds while others do not needs an explanation. It is difficult to imagine a process that could have made some monuments more prone to the loss of a mound or cairn than others, and it is far more likely that such dolmens were intended to be visible. There is, however, evidence at Gunderslevholm and at other sites in northern and western Europe that some free-standing megalithic structures constructed in the early Neolithic period were subsumed and encased in a substantial mound or cairn at a much later date – a point to which we will return.

Extravagant capstones

Is there another way of telling whether a dolmen was intended to be seen or to be covered? One obvious clue is the scale of the dolmen capstone, which is often disproportionately large and far more substantial than would be necessary to form the roof of a burial chamber. It seems that size does matter in the design of dolmens, as if the builders were aiming to present the monument in a way that almost defies belief, marrying a massive stone block with orthostats that are often so slender that they scarcely seem capable of supporting so great a weight.

Among those dolmens celebrated for the great bulk of their capstones is the 40-tonne Dolmen de Crucuno in Brittany; the 80-tonne lump of rhyolite at the site of the collapsed Garn Turne in Pembrokeshire (CA 286); and the biggest example in north-west Europe: the 169-tonne granite erratic used as the capstone at Brownshill, Co. Carlow, in Ireland.

The fact that many capstones slope at an angle exaggerates the drama of the display – and adding further to the sense that the capstone is precariously poised is the fact that (as Frances Lynch discovered to her alarm at Carreg Samson) the orthostats have no post-holes or packing stones: the carefully sited uprights are held in place entirely by the weight of the capstone.

There are other features, too, that suggest that the dolmen’s primary function was extravagant display. Ireland has a number of double capstone sites, such as the two immense blocks of stone at Aughnacliff, in Co. Longford (where the larger upper stone is teeteringly balanced on the lower), and the elegantly stepped dolmen at Knockeen, Co. Waterford. The capstones themselves seem to have been selected for their unusual shape, beautiful colouring, texture, striations, and natural inclusions. In parts of Scandinavia and Ireland once covered by ice sheets, for example, dolmens frequently employ glacial erratics, chosen as exceptional objects due to the exoticism of their anomalous geology, which contrasts with the colours and textures of the surrounding landscape.

Constructing a dolmen

Many dolmens are derived from naturally split boulders, with a flat underside and a rounded or craggy upper surface – in one instance, in Jutland, Denmark, two adjacent dolmens have capstones from the two halves of a single split boulder. There is no evidence to suggest that the boulders were divided by the monument-builders, but many dolmens with exceptionally flat lower surfaces have marks consistent with pounding and pecking to remove slight irregularities. In some cases, the dolmen is erected directly above the site of the boulder: at Carreg Samson, Frances Lynch found that the whole dolmen sat within a ‘large amorphous pit’, while W F Grimes, when excavating Pentre Ifan in 1945, discovered a similar ‘great pit’ underlying the monument. Both of these pits are now interpreted as extraction hollows: the depressions left after the capstones were excavated and raised on their orthostats.

The authors call this ‘in situ’ dolmen construction, and although no comprehensive geophysical survey has been undertaken to establish whether the practice was common, it does seem from the evidence we have that many dolmens were constructed within the capstone extraction site. This may have been done by excavating around and beneath the intended capstone and then using wooden levers, wedges, and props to raise the stone to the desired height, before digging the orthostat pits and securing the uprights with packing stones. This method was employed in experimental archaeology in the mid-1980s at Bougon in Nouvelle-Aquitaine, south-western France, where the techniques used to reconstruct megalith building-technology are the subject of a display at the regional archaeology museum.

Only when the intended capstone was released from the ground would the underside become visible, and if this needed shaping to attain a flat surface, the megalith might have been laid on its side for dressing. Just such a boulder, resting on its side in an extraction pit, can be seen at Carn Alw in Pembrokeshire – in this case, the boulder was abandoned and never lifted on to the expected stone supports.

Monumental messages

In considering the kind of message that dolmens conveyed to those who encountered them, it is important to remember that many of them date to the beginnings of the Neolithic in north-western Europe – or, in other words, from the final centuries of the 5th millennium and into the early to middle centuries of the 4th millennium BC. Recent evidence from ancient DNA shows that Neolithic practices were introduced by successive small migrations of people travelling northwards through France and Germany and up the Atlantic coast, and between Britain and Ireland – a complex process with no simple single narrative. Dolmens were thus being built at a time of substantial social change, when the nomadic lifestyle of the Mesolithic was giving way to settled farming communities, and we have to ask: what sort of interactions took place between Neolithic newcomers and hunter-gatherer groups? And what sort of landscapes might these migrants have encountered when they arrived as strangers in a strange land?

Monuments in the Making reminds us that the dolmens we see today often sit in splendid isolation at the centre of a manicured lawn, surrounded by little but heather, bracken, and gorse. In their original context, however, they were located in boulder-strewn post-glacial landscapes covered in thick forest. Creating open land for grazing and crop-cultivation involved removing trees and shifting boulders. Dolmen-building, though resulting in monuments that evoke a sense of awe and wonder, was simply an extension of the everyday activity of raising and moving large stones from the land.

Dolmens originally stood in landscapes yet to be cleared, surrounded by woodland and rivers of stone. They are a marker in that landscape, and a statement of intent in a land that had no ancestral significance for those who built them. Perhaps, then, dolmens were a way of staking a claim by constructing a showy monument that said: ‘We are here, and we are clever people with the muscle, technology, and social organisation to create sophisticated structures from the materials all around us.’

Relevant here is the work of anthropologist Michael Scott, who has concluded from his study of Melanesian cosmologies that wonder-evoking monuments like dolmens have ‘great potential for generating and grounding claims underpinning embryonic power structures in new and/or turbulent situations’. Supporting this view is Martin Bell’s suggestion (albeit yet to be fully researched and demonstrated) in Making One’s Way in the World (2020; CA 367) that Neolithic monuments were constructed at nodes and meeting points of the local paths and long-distance trackways of Mesolithic and Neolithic northern Europe – that is, built to be encountered, not erected in secluded and secret locations.

Neolithic migrants arrived in new worlds in which they had no direct ancestral roots, worlds that were ‘devoid of traditions and not appreciably manipulated by the actions of their forebears’ (to quote the final chapter of Monuments in the Making). That may be so, but these dolmen-building communities encountered crags, cliffs, tors, and rock outcrops that, while being entirely natural features, could easily have been interpreted as the works of an earlier race of gods or giants.

The authors quote Chris Tilley’s A Phenomenology of Landscape (1994) to the effect that physically striking natural rock formations ‘would be great sources of symbolic potency and power, signifying a wide range of enduring relationships between people, the land, time, and space’. No doubt the Mesolithic people already familiar with this landscape would have had their own stories to explain the mythological significance and cosmological import of these special places. Tilley concludes: ‘in elevating large stones, these people were emulating the work of a super-ancestral past’. One might go further and suggest that, in situations where Mesolithic and Neolithic people came into conflict (numerous examples of this are cited in the book), monument-building on this scale could have been a way of appropriating the past and claiming direct continuity with it (much as early medieval leaders liked to claim continuity with the Roman world).

Wonderful places

The distribution map of dolmens supports the theory that they were inspired by natural geological phenomena: these monuments are clustered in and around landscapes with extraordinary rock formations, such as the tors of south-west England, the Preseli Hills of Pembrokeshire, the Mourne Mountains of north-eastern Ireland, and the Comeragh Mountains of Co. Waterford.

Yet Monuments in the Making concludes on a slightly forlorn note. Mention has already been made of the later encapsulation of free-standing dolmens within substantial cairns or mounds and additional chambers, which took place at a number of sites in Britain, Ireland, France, and Scandinavia in the middle of the 4th millennium BC. At this point, dolmens do indeed become chambered tombs. The authors characterise this as ‘an end to wonder’, as wonder-places diminish in potency through familiarity and as large parts of the landscape become settled and domesticated. A monument-type that arose out of the encounters of new people with wondrous places had by this time served its purpose and was now redundant and mundane.

Wonder, however, rarely truly degrades – instead, it finds new forms of expression. Towards the end of the 4th millennium, those rocky outcrops that have been linked to narratives about origins, ancestral significance, ownership, and belonging were being quarried for materials that were used to create new monuments, including stone circles and standing stones. What more evidence could there be of the continuing potency of the rock formations of the Preseli Hills or the sarsens of the Marlborough Downs than their redeployment at Stonehenge (CA 311, 345, and 367), translocating one wonder-place to another?

Further reading
Vicki Cummings and Colin Richards, Monuments in the Making: raising the great dolmens in early Neolithic northern Europe, Oxbow Books, £39.95, ISBN 978-1911188438.

All IMAGES: V Cummings and C Richards unless otherwise stated..