Another ticklish question concerns the tools that were used at the Durrington settlement. Among eighty thousand worked flints, Ben Chan, the project’s flint specialist, has found just one small fragment of a flint ax. The 1967 team found fragments of just two among almost ten thousand Neolithic worked flints.18 By comparison with other Neolithic sites with similar quantities of worked flints, there should have been at least sixty axes and ax fragments. There was plenty of carpentry going on at Durrington but there’s no visible trace of the axes used to chop and shape the timbers for either houses or post circles. Were axes and their broken fragments carefully collected up and dumped elsewhere? Or were axes used only by people living in other parts of the settlement not yet excavated? Neither of these explanations is at all convincing. There are three pieces of circumstantial evidence that this Stone Age culture had in fact got hold of some new technology.
When people were living at Durrington Walls, copper had already been in use in eastern Europe for thousands of years, since about 5500 BC. That part of Europe has a distinct Copper Age (the Chalcolithic) between the Neolithic and Bronze Age.19 It has been generally thought that ancient Britons changed from using stone to using bronze without a noticeable intermediate stage, though archaeologists now speak of a brief Copper Age (or Chalcolithic) in Britain between 2500 and 2200 BC.
It took an exceedingly long time for copper metallurgy to spread across Europe. By 3200 BC it had only got as far as the Alps: the man nicknamed Ötzi the Iceman by his finders, who ended up frozen into a high-altitude glacier, had a copper ax. Mysteriously, copper metallurgy then apparently took about seven hundred years to spread as far as Britain. That is painfully slow progress, only about a mile a year.
Copper axheads from Castletown Roche in Ireland. These are similar to the earliest metal axes used in Britain.
The earliest copper in Britain is found in Beaker burials, a new burial rite that appeared around 2400 BC, and many archaeologists have assumed that copper was introduced by these immigrant Beaker people, who found out after their arrival how rich in deposits of copper ore Britain and Ireland actually were. But perhaps copper tools were already in use before the Beaker immigrants arrived. Copper must have been very valuable at first, so it is unlikely that people would have casually let it fall into the kinds of places where archaeologists would find it. Copper tools would have been both highly prized and recyclable, so there’s no reason why they should have ended up in the ground.
Archaeologists have often wondered whether copper axes were actually of much use. They are very pretty, shiny objects, but were they any good for practical tasks, such as chopping down trees? Unless copper is mixed with tin to form bronze (which doesn’t seem to have happened in Britain until about 2200 BC), it’s quite soft and its edge blunts easily. Perhaps copper axes and other objects were symbolic items, more like jewelry than tools, made to impress rather than to be used.
One of the documentary crews filming our excavations around Stonehenge decided to find out. They gave Phil Harding, Time Team’s well-known local archaeologist, a hafted copper ax and a medium-sized tree to chop down. Phil had often used flint axes for felling trees and was suspicious that the copper ax wouldn’t be anywhere nearly as good. To his surprise, not only was it easier to use than its flint counterpart, being lighter, but it also cut more quickly. Stone axes have to chop with oblique strokes so the axman’s cut into the tree is very broad, resembling the chewing of a beaver. Copper axes can be used to cut at a more perpendicular angle. You just have to stop every now and again to hammer out the blunted edge.
Archaeologists can get very hot and bothered about exactly when copper metallurgy started in Britain, and there has been a reluctance by some to see it pushed back any earlier than 2400 BC, or fifty years earlier at most. There are few hard facts, but some circumstantial evidence now suggests that the use of copper in Britain does go back to before the arrival of the Beaker people. When we dug into Durrington’s henge bank, we discovered that, while some of the chalk from digging the ditch bears the marks of antler picks, two chalk blocks have long thin, V-shaped cuts into them, as if made by the chopping motion of an ax. Yet the cuts are too thin to have been formed by a stone ax. It seems likely that only a metal ax could have produced such a thin groove. We can date the ditch-digging to 2480–2460 BC, giving us a clue that someone was using a copper ax slightly earlier than expected.20
New research on the metal composition of Beaker copper in Britain also hints at a start about a century or two earlier than 2400 BC. Peter Bray, a researcher at Oxford University working on Britain’s earliest metals, suspects that some of the heavily worn copper daggers found in Beaker burials might have been heirlooms already ancient when buried as grave goods. While some daggers and axes were made from copper dug out of the mines at Ross Island in southern Ireland from about 2400 BC onward, Peter thinks that the manufacture of other daggers and axes with a slightly different chemical composition could have started earlier, perhaps a century or two before.
Another tell-tale clue is the change in the scale of tree-felling after about 2500 BC. Monuments prior to that date were constructed using timbers that were generally no thicker than 30 or so centimeters. After that date, we find monuments built of timbers up to a meter in diameter.21 This change is perfectly shown in the size difference between the timbers used in the Southern Circle’s first phase and those of its second phase. Perhaps what was special about that moment around 2500 BC was that copper axes became available in sufficient numbers in Britain to fell much larger trees with greater ease. A great gathering place, such as the Durrington Walls village, might have been just the right venue for people from different regions to discuss new innovations, show off their prized new copper tools, and swap ideas about new technology. Perhaps it actually helped to spread the use of copper across Britain.
Can we resolve the question of whether this was a seasonal gathering place or a permanently occupied settlement? Hazelnuts and apples are gathered in autumn but can be stored through the winter. The pig-culling tells us that people were here, feasting on pork, at midwinter—most likely around the midwinter solstice, given the midwinter sunrise orientation of the Southern and Northern Circles.
Was the henge ditch also dug in midwinter, when it would seem too cold to be outdoors? A regular visitor to our dig each summer was Jake Keen, an “ancient technologist” who has studied traditional rural trades, such as iron smelting, in Britain and other parts of the world. He knows, for example, how to make ropes from the bast of linden trees. He is also very interested in finding out how Neolithic people dug such deep ditches. Living on the chalkland of Cranborne Chase, south of Salisbury Plain, he has noticed that chalk is softer and easier to work when it is wet. It would have been easier to hammer in those antler picks during wet periods. Perhaps the ditch-digging and henge-building were done in the wetter months of winter, a time of slack in the agricultural year.
While there’s definite evidence for gatherings at Durrington in the winter months, there is also new evidence that people came here during the summer too. Umberto’s team of animal bone specialists has identified a second peak in pig mortality about six months after the first, indicating that there was feasting in the summertime as well as the winter. Other evidence for this summertime presence is provided by the discovery of chemical residues of dairy fats in the pots. Before our agricultural revolution, cows produced milk in quantity during the spring and summer months only, after calving. So the milk pots deposited at the front of the Southern Circle, at the end of the midsummer sunset-oriented avenue, are also evidence that people were here in the summertime. It’s reasonable to assume that they came to celebrate the midsummer solstice as well as the midwinter solstice.
Why did people come to Durrington Walls? There were the timber circles to erect, feasting and general partying, but perhaps people actually came here to work on a nearby construction site: a great stone circle just downriver. We needed to find out how closely we could date t
he stages of construction at Stonehenge, and whether any of these stages coincided with the settlement at Durrington Walls.
7
THE GREAT TRILITHON AND THE DATE OF THE SARSENS
__________
Standing stones can be dated only by what is found in the holes that they stand in. We have to be sure that whatever we’re dating was put in the hole before the stone went in, and that it was fresh when it was deposited. The best objects for dating are the antlers used as pickaxes by Neolithic builders, who had the fortunate habit of deliberately leaving behind an occasional tool in the pit they’d just dug.
The radiocarbon dates obtained from antler picks found in the holes for the Stonehenge sarsen circle and the sarsen trilithons have always been problematic because they don’t seem to make sense. The date from the great trilithon’s hole (2440–2100 BC) is later than the dates (2620–2480 BC) from another trilithon (Stones 53–54) and from the sarsen circle (Stone 1). When, in the 1990s, Ros Cleal and her team drew together all the available information about Stonehenge, these dates seemed to suggest only two possible scenarios:
The great trilithon was built after the circle of sarsens enclosing it. This would have been possible only if part of the circle was first taken down, or if part of the circle hadn’t yet been finished, thereby leaving a big enough gap through which these massive stones could have been maneuvered into the center and set upright. No one has ever been comfortable with this explanation because the engineering problem is so huge.
Or:
The dates can’t be trusted. Is there something wrong with the dates for the antler picks found in the sarsen circle and around Trilithon 53–54? Perhaps these antlers were antiques that were no good for digging but had been carefully kept for generations and thus were already old when they were deposited in the stoneholes. This is not as far-fetched as it might sound: There are many prehistoric instances of old objects being kept as curios or treasures.
Most archaeologists thought the latter explanation more likely: The great trilithon just cannot be later than the sarsen circle because of the logic of construction, so therefore the antlers that produced the earlier dates of 2620–2480 BC must have been antiques when they were buried. For those who accepted this argument, the trilithons and sarsen circle were put up in 2440–2100 BC. I’d never been happy with either interpretation, because both meant ignoring dates simply because they were inconsistent and didn’t fit expectations. It was time to go back to the records and re-examine where the antler picks in the stoneholes actually came from.
Atkinson found two antler picks near Stone 56, one of the uprights of the great trilithon. This is one of the two longest stones at Stonehenge, set 2.4 meters into the ground. He thought that the antlers came from a ramp sloping down to the stonehole.1 This ramp, Atkinson argued, would have been used to tip the foot of the stone into its hole before the sarsen was hauled upright. On re-examining the plans, though, we became very doubtful of this interpretation. The other Stonehenge sarsens don’t appear to have ramps leading into their stoneholes, so why did this stone have one? (And a sizeable one at that—the area Atkinson called the ramp is more than five meters long.)
In addition, Atkinson’s explanation of how the stone was put in place using the ramp is extremely convoluted. Megalithic stones can be erected most easily by being brought to the chosen spot on a wooden cradle and then simply hauled and levered upward to the vertical. Atkinson, however, was convinced that this huge stone must have been brought in on its side and then lifted up: that is, raised from its narrow side rather than from its wide front. At first glance this seems feasible, if awkward, but when I looked more closely at Atkinson’s plans it became evident that this scenario would have been impossible.
The ramp is simply too long—it could never have guided the stone into its hole. It supposedly enabled the builders to tilt their stone on the fulcrum of the pit’s edge, to tip its foot into the bottom of the hole. But we know exactly how long the stone is and it just couldn’t have been tipped down into the stonehole from this ramp: The measurements don’t allow this possibility because the ramp’s edge is too far away from the stonehole for the stone to tip into it.
More importantly, the ramp doesn’t actually lead to the stonehole at all.2 Atkinson’s trench, dug in 1956, did not uncover the stonehole itself—that was dug by Gowland in 1901. Atkinson’s trench stopped just 40 centimeters from Gowland’s trench and uncovered only the west end of the feature Atkinson calls the ramp in his trench. By matching what he had found to Gowland’s account of features found in the earlier trench, Atkinson extrapolated that the ramp should continue to the edge of the stonehole. Unfortunately he didn’t read Gowland’s account closely enough.
Gowland had indeed picked up the edge of what Atkinson later reckoned to be a ramp, but this feature in Gowland’s trench didn’t lead directly into the stonehole. Atkinson’s ramp is off-center to the stonehole and connects only with the north side of the hole: The stone could not have been erected from the supposed ramp.
It seems that Atkinson mistook the western end of a large and deep pit to be a sloping ramp leading to Stone 56, the west upright of the great trilithon. What Gowland, and then Atkinson, actually found was simply a very large hole.
Someone dug a huge pit into the center of Stonehenge on the west and north sides of the great trilithon—a pit as deep as the bottoms of its stones. Perhaps this was what ultimately led to their destabilization and caused the great trilithon to fall. Gowland had excavated at the foot of its only surviving upright because it too was expected to topple, and excavation was needed prior to intrusive remedial work.
The antler picks found here are not, therefore, associated with the building of a ramp for the great trilithon’s erection; they’re picks deposited in a large pit dug after the trilithon was erected. It is this pit that dates to 2440–2100 BC, not the trilithon stonehole: The picks belong to the pit but the pit doesn’t belong to the trilithon. We know that this pit was filled back up to the top in prehistory, within a century or so, because a bluestone was later set into it as part of the inner oval of bluestones, dated by an antler pick to 2270–1930 BC.
Reconstructed profile of the great trilithon Stone 56 in its pit, showing that Atkinson’s presumed ramp is too long to have been a construction ramp for erecting the stone. The numbers refer to Atkinson’s and Gowland’s trenches.
Why did the Neolithic builders dig this pit? Its north and east edges have never been located (no one has ever dug in that area) but it extended at least as far north as the Altar Stone, where William Cunnington encountered part of a deep hole. This pit must have been huge, at least 12 meters long, 5 meters wide and 2.4 meters deep. Its purpose is a complete mystery: It is far too deep to be explained as simply part of the works undertaken by the Neolithic builders during the rearrangement of the small bluestones within the center of Stonehenge. Its upper fills contained all sorts of later finds, so it was probably dug into millennia later by the Duke of Buckingham’s men in 1620 and two hundred years after that by Colt Hoare.
The dates from the antler picks have, of course, been used until now to date the great trilithon; the realization that they actually belong to another, later pit means that we can discard those dates from the construction sequence of the sarsen circle and trilithons. Instead, the date of 2620–2480 BC for the sarsen-building is the one that counts.
It’s possible that one day someone will get another date for the building of the sarsen circle and trilithons. In 1964 Atkinson found the tip of an antler pick embedded in the side of one of the sarsen stoneholes (Stone 9)—it had evidently snapped off as someone tried to lever out the chalk wall of the pit while they were digging it out. We can be sure that this pick was in use during the digging of the hole for Stone 9.
Unfortunately, this surviving piece of antler is very small: At 1.4 grams, it may provide enough collagen for dating, but the process would destroy pretty much the whole object. The tip and the piece of chalk in which it was
found are currently on display in Salisbury Museum and no one’s going to touch them until scientists are able to use smaller quantities of collagen to obtain radiocarbon dates—when it will possible to leave most of the tip intact for future researchers. It may be then, at some unknown point in the future—and even without digging any more holes—that archaeologists get another date for the building of Stonehenge.
8
MYSTERIOUS EARTHWORKS: THE LANDSCAPE OF STONEHENGE
__________
Until the 1970s, archaeologists had a tendency to study each site in isolation, rather than looking at landscapes of monuments that should be treated as a whole. Stonehenge was a prime candidate for trying out a new approach to ancient landscapes that paid more attention to the associations among sites. It is surrounded by prehistoric monuments of all kinds, and scatters in the plow soil of worked flints and pottery. In 1982, Julian Richards, at that time a project officer for the newly formed commercial company Wessex Archaeology, started a landscape survey called the Stonehenge Environs Project.1 He hoped to understand prehistoric use of the landscape over time, from the Mesolithic to the Iron Age, by systematically recovering flints and potshards from plowed fields around Stonehenge, and digging exploratory trenches into a wide range of nearby prehistoric sites to date them.
The results were spectacular. Some 102,175 pieces of worked flint were collected from the surfaces of plowed fields by teams walking in 50-meter-long transects at 25-meter intervals. This is a type of sampling. By spreading a field-walking team at consistent, regular intervals, a survey project picks up only a proportion of the total finds, and cannot accidentally select only the most interesting items from a field. From the quantities picked up, and the surface area sampled, one can extrapolate the total quantity of finds likely to be present in the topsoil.
Stonehenge—A New Understanding: Solving the Mysteries of the Greatest Stone Age Monument Page 13
