Stonehenge—A New Understanding: Solving the Mysteries of the Greatest Stone Age Monument

by Mike Parker Pearson

The smaller postholes around the outside of Cunnington’s structure were very like some found by Julian Thomas in the center of Durrington Walls: the postholes of the circular palisades around the two “special” houses inside the henge. The very large internal postholes of Cunnington’s structure were too deep and too wide to have been roof supports. With each of these posts likely to have been more than 50 centimeters in diameter, they would probably have stood at least 5 meters high. Josh realized that the structure was not a house but a tower, surrounded by a fence. The four large posts probably supported a wooden platform. With its entrance aligned on the midwinter solstice sunrise, this tower is likely to have had some ceremonial purpose. The view from its platform, if it had one, would have included grandstand coverage of the river below.

  North of Cunnington’s structure, Josh discovered another, similar array of postholes. Here four postholes formed a square arrangement and had two pits to the east of them; the layout was identical to that of the first structure, although slightly smaller, and only a few of the ring of little postholes had survived. The posts of this second structure had been set about half a meter shallower but were nonetheless massive.

  These timbers had been left to decay in situ and, as with the Southern Circle, pits had been dug into the tops of them and filled with cattle and pig bones—and, in one case, with the skull of a wolf or large dog. One of the two pits at the front turned out to be another posthole. In its fill there was a lump of daub, indicating that the oval enclosure fence had probably been faced with daub like the walls of the nearby houses at Durrington Walls. A radiocarbon date has established that this timber structure was probably of the same date as the settlement at Durrington.

  South of Cunnington’s structure, there was a third arrangement of postholes, this one much smaller and consisting of a square of holes about 2.5 meters apart. There was no sign of a surrounding palisade fence, and the fills of these postholes showed that the posts had been pulled out and not left to decay. How many more of these Neolithic timber structures had lined the cliff top above the river? And what were they for?

  We don’t know what happened to most of the dead during the Neolithic period. During the Early Neolithic some people were buried in long barrows, but certainly not everyone—there are nothing like enough barrows to contain the whole population. There is almost no trace of what happened to the dead of the later Neolithic. Some people were cremated, and the ashes then buried.12 Perhaps others were scattered after cremation; some groups may have used funerary rites that are unfamiliar to us and that have left almost no archaeological trace. Josh thinks that the timber towers south of Woodhenge were raised platforms for exposing the remains of the dead. Corpses could have been left for their flesh to be picked clean by birds, and the bones then collected up and placed in the river. In that way, the spirits and the bones of the dead could begin their journey downriver toward Stonehenge. Whatever the purpose of the wooden towers, we have no doubt that the people of Stonehenge regarded this stretch of the river as very important in their spiritual as well as everyday lives.

  We hoped to find out whether anything had been regularly deposited into the Avon during prehistory. In 2007, we therefore dug a couple of trenches into the paleochannel below Durrington Walls. Because of the Environment Agency’s rule about no disturbance closer than eight meters to the riverbank, we could not examine the bit that we really wanted to get at. We should have liked to investigate the deeper part of the former river channel, where the Durrington Avenue would have ended. Anything heavy thrown into the river from here, such as stone artifacts or even pots and human bones, might have sunk into the mud at the bottom of the channel.

  We were limited to the eastern edge of the paleochannel, and here we found only a few burned sticks and a worked flint. Further downstream, where the channel had widened in prehistoric times to form a reed swamp, there were no artifacts at all in our trench, but the deep sequence of layers contained pollen grains that would give us a complete vegetational history of the locality. We could see the bottom layers, formed after the Ice Age when hunter-gatherers lived along the river’s margins. Then there were layers of reeds that built up during the Neolithic and Early Bronze Age; these would have provided the roofing material for the Neolithic houses. The reed layers were abruptly terminated by deposits of gray soil, washed into the river from the Middle Bronze Age (about 1500 BC) onward, as the intensity of farming dramatically increased and rain flushed loosened plowsoil into the river.

  Examining plant pollen is a standard archaeological technique for reconstructing prehistoric vegetational environments. Although the plants themselves rot away, leaving no trace, their pollen can survive for millennia. Ancient pollen survives only in damp places and can sometimes be retrieved by coring a hole through layers of waterlogged soil; pollen is, of course, entirely invisible to the naked eye, so the extracted core of soil is sent for examination by a specialist who identifies the various species present.

  Mike and Charly’s colleague Rob Scaife has analyzed the microscopic pollen grains from the Avon paleochannel’s different layers to find out which species of tree and shrub grew in this area through time. He’d previously studied a sequence from a spot 300 meters upstream from Durrington Walls, and these two studies led him to a surprising discovery. The landscape of 8000–7000 BC, inhabited by Mesolithic hunter-gatherers, was only lightly wooded with beech, pine, hazel, and oak trees; by the time of the Neolithic and Bronze Age, the area was largely devoid of trees and shrubs.13 Instead, the vegetation was dominated by grasses, with the addition of sedges and marsh plants on the floodplain. Perhaps this vegetational development of Salisbury Plain since the Ice Age—from light woodland to largely treeless grassland—can be put down to woodland clearance by the earliest Neolithic farmers. Mike, Charly and Rob have also found a similar sequence on the high chalkland of Cranborne Chase, twenty miles to the south; here, however, the change probably took place without human interference.

  Mike Allen has reached a similar conclusion about the lack of woodland on Salisbury Plain by using a quite different type of evidence. Mike is an expert on snails: When snails are retrieved from archaeological layers, specialists like him can look at which species are present and thereby reconstruct what sort of vegetation was present.14 When most people think of snails they imagine the big, fat garden variety, a centimeter or more across, but actually the majority of snails are extremely small and have to be identified by species with a microscope.

  As far as archaeologists are concerned, there are two types of snail. Some are catholic, in the sense of wide-ranging or broad: These snails are adapted to live in many different and varied habitats, which makes them useless for our purposes. The types of snail that are useful to us are those that are niche specific; some species can live only in open grassland, others only in shady woodland, and so on. Through years of patient research, examining snails from prehistoric ground surfaces preserved beneath burial mounds, Mike has discovered that during the Neolithic and Bronze Age the area around Stonehenge was not inhabited by woodland-dwelling snails. It looks as if there never was a dense Neolithic forest on the chalk uplands of Salisbury Plain.

  After the Ice Age, growth of trees on the downland began later than in surrounding lowlands, initially with just birch and pine being present. After 8000–7000 BC, the warming climate supported light deciduous woodland of oak, elm, and hazel, gradually replacing the cold-temperature species. The woodland thinned out during the millennium before 3000 BC when Stonehenge was built; the open, grassy plain that we see today was already formed five thousand years ago, partly through clearance and grazing by early farmers and their animals, and partly because much of this landscape was already open. Of course, away from Stonehenge, the river valleys, lowlands, and lower slopes of the chalk were forested and, even long after the time Stonehenge was built, the Avon’s floodplain still supported alder and hazel woodland even though the downlands around it were largely treeless. Only a few stands of woodland or single
trees of oak and linden were scattered over this open, chalkland landscape and, as we have learned from our excavations of prehistoric tree holes, these lone sentinels might well have had special significance for the people of the Neolithic.

  10

  THE DRUIDS AND STONEHENGE

  __________

  Anyone planning to dig at Stonehenge has to jump through a lot of hoops. Permission is granted in the name of the Secretary of State for Culture, but the advice given by English Heritage to the Department of Culture, Media and Sport (or “Arts and Darts,” as it’s sometimes called) decides the matter. English Heritage’s opinion is formed from the views of its regional inspector and regional manager, with input from the management team for Stonehenge. Before the recommendation is sent to DCMS, a final decision is taken by English Heritage’s chief executive and the English Heritage Advisory Committee (EHAC), which comprises not civil servants but rather the great and the good in matters of historic buildings, ancient monuments, and archaeology.

  EHAC must have been taken aback when it received not one but two applications for digs at Stonehenge in 2008. The first application was for the work planned by Professors Geoff Wainwright and Tim Darvill. Geoff had retired from his job as Chief Archaeologist at English Heritage and was now President of the Society of Antiquaries. After an illustrious career, including digging on virtually every henge in Wessex other than Stonehenge, he was fulfilling the dream of a lifetime. Tim and Geoff had been working in the hills of west Wales, searching for the source of the Preseli bluestones.1 Now they planned to dig a trench within Stonehenge to try to date the double arc of bluestones there.2

  This double arc was thought by Atkinson to be the partial remains of two concentric circles of bluestones. He called the outer arc the Q Holes and the inner arc the R Holes. Atkinson’s excavations in 1954 had found that one of the stoneholes in the sarsen circle (Stone 3) cut into the edge of one of the Q Holes (Q Hole 4). Wherever one pit cuts another, we know that the second pit is later than the first. What Atkinson found meant that the Q and R Holes had to be earlier than the sarsen circle. Atkinson had excavated some twenty Q and R Holes but found nothing that could be used to radiocarbon-date their construction. Geoff and Tim hoped they would be luckier.

  The Stonehenge Riverside Project made the second of that year’s applications. I had invited Julian Richards and Mike Pitts to join us. Julian’s Stonehenge Environs Project had made a major contribution to understanding Stonehenge,3 and he was hugely enthusiastic about the opportunity to dig at Stonehenge itself. Mike also has a considerable amount of knowledge about Stonehenge, having dug there—outside the wire—in 1980 in advance of the cable trench.4 We wanted to recover a large quantity of cremated human bones dug up by Hawley in the 1920s. These cremated remains were found primarily in the pits known as Aubrey Holes; this circle of fifty-six pits lies inside the Stonehenge ditch and dates to the first phase of activity at Stonehenge. In 1935 these remains had been reburied, for safe keeping, in Aubrey Hole 7 by archaeologists Robert Newall and William Young. It seems that they could find no museum prepared to take the cremated bones, probably because back then, long before the development of today’s methods and techniques of analysis, there appeared to be no possibility of ever learning anything from them.5

  The committee members of EHAC must have had a difficult decision to make: Should they refuse both applications, on the grounds that Stonehenge’s below-ground remains should never be disturbed, or grant permission to just one project, or agree to both? Fortunately the committee was happy to see both proposals go forward. Each project had a long and fruitful track record over the previous five years, Tim and Geoff in the Preselis and ourselves around Stonehenge. The excavations were supported by meticulous research designs, and would have little impact on preserved remains because of the small sizes of the proposed trenches. Plus, these quite tiny trenches would be placed in or adjacent to areas already dug by earlier archaeologists.

  Tim and Geoff planned to re-open one of Atkinson’s trenches, between the sarsen circle and the trilithons, and then extend for about two meters into untouched deposits. The trench was positioned in the southeast area of the monument (between Stones 10 and 35; see figure on page 29), on top of the line of Q Holes, immediately outside the stones of the bluestone circle.6 Atkinson had recorded half of a Q Hole in the end of his trench here; Tim and Geoff would be able to dig out its other half, and they hoped to find another Q Hole to its west.7 This end of Atkinson’s trench had been full of postholes, so there was a good chance that they might find more of these and be able to date them.

  A plan of Stonehenge showing the locations of the excavated cremation burials (black circles) in the Aubrey Holes and ditch.

  By 1956, Atkinson had mapped out the extent of the Q and R Holes in the northeast and east areas of Stonehenge.8 He must surely have looked very hard for a similar double arc of stoneholes on the west and southwest sides, but there is only a single arc in this area. The Q and R Holes, and their place in the story of Stonehenge, can be difficult to follow, particularly because these are stoneholes, not stones. Briefly, the Q and R Holes are where bluestones once stood. The bluestones were pulled out of these holes around 2400 BC and then re-erected, in much the same arrangement, as a bluestone circle. The Q and R Holes are mostly underneath where the bluestone circle stands today; nobody knows why the prehistoric builders went to the trouble of removing and then re-erecting these stones.

  Some archaeologists think that the rings of Q and R Holes were never finished. Others think that an incomplete double arc was all that was intended. Another possibility, which seems to me more likely, is that as far as the builders were concerned the arrangement was complete, given the symmetry in the arrangement of bluestones in a double arc on the northeast side and single stones on the southwest side.

  Atkinson’s other major discovery was that Q Hole 4 appeared to be cut by the stonehole containing one of the stones of the sarsen circle (Stone 3).9 Thus the Q and R Holes were thought to be earlier than the sarsen circle, and probably earlier than the sarsen trilithons. Atkinson appeared to have discovered a separate phase of construction at Stonehenge, between the earliest phase of bank and ditch and the sarsen circle. The arrangement of small bluestones belonging to this phase was itself later modified by the erection of the sarsens and the re-arrangement of the bluestones into a bluestone circle and bluestone oval at the center of the monument. If the Q and R Holes were the first bluestone construction at Stonehenge, then if Tim and Geoff could find dating evidence in one of the Q Holes this would, they reckoned, reveal when the bluestones were brought from Wales.

  Their dig started in March 2008, with cameramen jostling to film Tim and Geoff cutting the first sods. There were daily updates on the web, a live feed to a screen in a tent in the visitors’ parking lot, a blog, and web forums. Mike Pitts, Julian Richards, and I visited them after their first week’s digging. Tim told us that he had spotted that the cuts of the features in the end of Atkinson’s trench were not doing what they should have done. He could see that the Q Hole actually cut the hole for the fallen stone of the sarsen circle. According to Atkinson, the relationship between these two pits should have been the other way around. We were all mystified.

  Tim and Geoff were unlucky: There was no antler pick in the Q Hole. Given that more than twenty Q and R Holes had already been dug out by Hawley and Atkinson without a sniff of one, it had always been a long shot; it seems it just wasn’t Neolithic practice to leave picks in these particular pits. The situation was, however, salvageable. Geoff and Tim hoped to collect enough pieces of charcoal and bone to be able to radiocarbon date groups of such items from particular layers. By the end of the excavation, their hard-working flotation team had accumulated a modest collection of charcoal lumps, along with a few pieces of animal bone and a human tooth.

  There was a lot of pottery, too. Atkinson’s excavation team hadn’t sieved the soil and had missed some sizeable shards of Beaker pottery. What really surpr
ised Tim and Geoff, though, was the quantity of Roman pottery. Just like the Cuckoo Stone, this place had been a magnet for people in the Roman period. They didn’t come just to look at the standing stones, but actually dug holes into the site. Even deep in the Q Hole, Geoff and Tim found a shard of Roman pottery. It was no wonder that the stratigraphy was mixed up, as we’d seen on our site visit. These holes had been dug into and refilled long after they were first used in the third millennium BC.

  By the end of the environmental processing, Tim and Geoff had enough pieces of charcoal to submit several samples from each context. These came from a sarsen circle stonehole (Stone 10), from two stoneholes belonging to the bluestone circle, and from a Q Hole. Cut by the Q Hole, and therefore earlier than it, were two small postholes. Only one had a piece of charcoal in it. This was the first opportunity in many years to look carefully at postholes in Stonehenge. Like the others found by Hawley and Atkinson, these two were less than half a meter across. Curiously, they had no post pipes and no packing, nor was there any indication that posts had been withdrawn. Their fills were utterly homogeneous. Tim even wondered whether they had really held posts.

  Tim Darvill (left), Geoff Wainwright (center), and Miles Russell (right) excavating at Stonehenge in 2008.

  After a cold few weeks of gray skies and rain (and some unseasonable snow), Tim and Geoff finished off and filled in their trench. They would now have to wait months for the results of the radiocarbon dating. Mike Allen was advising them on this and, well aware of the amount of previous disturbance, set out a strategy to follow:

  First of all, identify the tree species from which the charcoal came, and then select only those pieces that derived from small roundwood. Dating charcoal formed from heartwood is not much use; if a heartwood sample happens to have come from a large tree, it will have already been centuries old when it was burned. Such a date only tells us how old the tree was, not the date of the event at which it was burned.