We planned our first season of fieldwork for 2004. If all went well, this could be the first of seven years’ work. Stan Rawlins agreed to move his cattle for a month so that we could dig up a small part of his field. We applied for permission to English Heritage, English Nature, and the Environment Agency (which had to be consulted on any works within eight meters of the river’s edge). And we also applied for grants. We needed £4,000 in addition to the support provided by our various universities. Labor wasn’t going to cost much, as this was to be a student-training excavation. Digging on a world-famous Neolithic henge is as good as it gets for student archaeologists, so we knew we’d have no trouble finding a big enough team, but we had to raise money to feed, house, and transport them, as well as to buy all the necessary equipment. We also needed to pay some supervisors: With so many inexperienced students on site, we had to recruit a handful of unusually patient professional archaeologists to make sure things went smoothly.
Although £4,000 would see us through four weeks’ digging, it would not be enough to pay for the many months of processing and analysis back in the lab, nor the costs of the fieldwork planned from 2005 onward. What we really needed was a very large grant so that we could plan well into the future and avoid scrabbling for small pockets of cash every season. Until we got long-term funding, English Heritage would not be satisfied that the project was sufficiently resourced for it to run for more than a season. But what institution would take the risk of providing such funding for a project that hadn’t even started? We had a good theory, and a good research design, but no one knew if we’d find anything at all.
Things looked brighter when the Royal Archaeological Institute announced a national competition to award a prize of £100,000 to the best archaeological project proposal. Our brand-new Stonehenge Riverside Project made it through to the final three, but then we received a dismal letter informing us that the RAI found our proposal “too speculative.” In fact, they decided not to award the money to anybody. We realized that what we were attempting did not have the support of many senior archaeologists. In some experts’ minds our plans were clearly wrong-headed: Why were we going to dig outside the Durrington Walls henge, when surely everyone knew that the place to look was inside?
Still, we ignored the disappointment and raised enough money to dig in 2004. That August we excavated a large trench near the henge’s east entrance and two smaller trenches on the steep slope leading down to the river. None of the geophysics results or contour surveying prepared us for what we would find. Down by the river, Josh’s team hacked deep into layers of sediment accumulated from centuries of plow movement, only to find that the gully in the riverbank was of comparatively recent origin, created during the last two thousand years. It had nothing to do with access from the Neolithic henge to the River Avon.
Colin and Julian had more luck nearer the henge. Part of their trench had been positioned over the highly magnetic spots picked up by the magnetometer. These turned out to be Neolithic pits filled with animal bones, broken pottery, worked flints, and ashes. Even better, there were about forty pits—many more than indicated by the magnetometer results.
Some pits were filled with finds and others were almost empty. The emptier ones were grouped in small clusters, and we realized that they had been dug to extract the chalk sub-soil. We normally picture chalk as a soft white rock whose surface is covered in fissures and cracks. However, these Neolithic pits were dug into a yellow claylike chalk residue that had collected within a wide and deep gully more than twenty thousand years ago, during the last ice age. The glaciers never reached Salisbury Plain but the front of the ice sheet may have been only fifty miles to the north. Conditions on the plain were therefore extremely cold. Within this periglacial environment, areas of chalk were heavily weathered and eroded—turning some of the chalk rock into fine silt and clay, carried away by seasonal meltwater and deposited within newly formed gullies and valleys. This “soliflucted” chalk is perfect material for making daub, the clay that has been used for millennia together with wattles (wooden rods and branches) to build house walls.
A pit excavated in 2004 at Durrington Walls. It contained animal bones which are the remains of feasting, as well as flint tools and pottery and a human femur. The black and white photographic rods are used to indicate scale, with each alternating color representing 10-cm intervals.
These extraction pits filled the northern half of the trench but, disappointingly, the southern half—where we expected to find signs of an avenue running eastward between the entrance and the river—had uncovered an area that had been scoured by thousands of years of erosion and plowing. We found the bottoms of a few pits but little else. Even the ditch found by the geophysicists turned out not to be one side of an avenue; it was an Iron Age land boundary constructed two thousand years after the henge itself.
Although we were disappointed at not finding an avenue, the huge amount of Neolithic rubbish in some of the pits showed that we were not far away from something intriguing. There were lots of flint arrowheads, of the type known as “oblique” arrowheads.
The fragment of human femur from the pit shown in the previous figure. Note the impact scar in the center of the bone (below the 3cm mark in the scale) caused by a flint arrowhead.
Bows and arrows were standard equipment for Neolithic people. They were probably used for hunting but were also used as weapons, together with wooden clubs. Recent analyses of human remains dating to the Neolithic period seem to indicate that Neolithic people suffered more injuries from being whacked on the head than from being shot with arrows; about one in fifteen people buried in tombs of this period had been clubbed on the head, fatally in half the cases.12 In contrast, we have fewer than a dozen instances across Neolithic Britain of skeletons with arrowheads in them. However, these are probably the tip of the iceberg. In identifying an individual as having been shot, archaeologists tend to recognize only those cases where the arrowhead fortuitously hit solid bone. More than a hundred arrowheads have been found in Neolithic tombs; many of them might have entered these burial places buried deep in the soft tissues of archery victims rather than as gifts for the dead.13
One of our pits, filled with animal bones, contained a battered human leg bone. Unlike the animal bones, deposited fresh as food waste, this femur had evidently been knocking around for some time, possibly a century or more, before it ended up in the pit. We wondered if it might be a grisly trophy of combat. Its surface was perforated by two small depressions, pronounced by our human bone specialists to be the results of trauma caused by arrow impacts. There were no broken-off tips of flint arrowheads in the wounds, but this leg bone was evidence that the Stonehenge people were not entirely peaceful.
A chisel arrowhead from Bluestonehenge. This type of arrowhead was in use in the Middle Neolithic (3400–2600 BCE).
As well as providing evidence of warfare and violence, arrowheads are also useful to the archaeologist as indicators of chronology. Their shapes changed over time. The earliest farmers had different arrowheads from Britain’s hunter-gatherers. During the first thousand years of farming, these were leaf-shaped points. From around 3400 BC to 2600 BC, people adopted a new fashion of arrowhead with a flat blade, called a chisel arrowhead—these could cause a wider wound than the leaf shapes and would be more likely to remain behind when the arrow was pulled from the body of its victim. Between 2600 BC and 2200 BC, these were themselves replaced by oblique arrowheads, with a distinctive asymmetrical, pointed triangular form. One side of an oblique arrowhead is longer than the other, and culminates in a tang. Some are beautifully made, flaked to produce long ripples across their surfaces.
We found almost four hundred oblique arrowheads at Durrington Walls, a surprisingly large number. They were almost as numerous as flint scrapers, normally the most common tool on Neolithic settlements, used for scraping fat and bristle off animal hides. Another surprise was the very small number of bone points for piercing and stretching animal hides. Instead, ther
e were bone and stone beads and lots of bone pins, several inches long and polished smooth, used to secure either clothing or hair. Clearly this was a different kind of settlement from the norm, since the everyday tool types used in daily farming life were so few. The large sizes of the broken pot shards and the fact that some of the animal bones had been dumped with soft tissue still holding them in articulation showed that this was rubbish collected after some sort of feast. The rubbish had been buried soon after each feasting event but, judging by the gnaw-marks on some of the bones and some preserved dog feces, not until after the dogs had rummaged through the party refuse.
The find that caught most people’s attention was a flint phallus. The flint of Wessex formed about eighty million years ago, as seams within chalk, derived from the calcium in tiny animals’ bodies decaying at the bottom of a long-vanished tropical sea. It seems that the silicate components of their bodies collected in seams at certain levels within the chalk and formed flint in all kinds of shapes, sometimes flat, sometimes nodular. It also formed occasional small spheres and elongated knobs. These bizarre shapes have sometimes been further exaggerated during their formation by the impressions of prehistoric shells and sea urchins.
An oblique arrowhead from the ditch of the Stonehenge Avenue. This type of arrowhead was in use in the Late Neolithic/Chalcolithic (2600–2200 BCE).
Prehistoric man (for I suspect it was he rather than she) seems to have had an eye for these natural oddities. Back in the 1920s a local archaeologist and antique dealer, Mr. A. D. Passmore, excavated a round barrow (a Bronze Age burial mound) close to Stonehenge; its central grave pit contained not a prehistoric skeleton but a collection of strangely shaped flint nodules.14 From the bottom of a Neolithic flint mine in Sussex came an impressively large penis-shaped flint nodule, its glans formed by a fossil echinoid, or sea urchin.15 At other Neolithic sites archaeologists have found phalli carved out of chalk. There is even one from Stonehenge. They are generally unattached, anatomically speaking, but a rare find of a carved wooden figurine—the so-called god-dolly—from a bog in the Somerset Levels portrays a hermaphroditic individual with both penis and vulva.
Our Durrington Walls phallus was a natural nodule whose glans was formed by a fossil shell, with a possibly worked groove at the tip. It might have been nothing more than a chance geological quirk—more in the excavators’ imagination than noticed by prehistoric people—except that its context provided evidence for careful selection. This phallic object lay near one end of a narrow Neolithic pit, in close proximity to two testicle-sized, natural flint balls. Close to it, a small flint slab covered another unusual natural flint lying on the sloping side of the pit. This flint had a natural hole through its center and came to be called the “pelvis flint” by the diggers because of its shape. Perhaps we had uncovered a prehistoric sex-education lesson.
The flint phallus (and two flint balls) from a pit beneath the avenue at Durrington Walls.
Phalli are known from many different cultures, ancient and modern. The Romans, for example, used to provide them with wings and considered them symbols of luck and fertility. Perhaps they were similarly treated as fertility objects in the Neolithic. Back at the campsite our find soon took on the nickname of the “Durrington Dong,” and gave its name to the project’s improvised cocktail of the season—gin, Campari, and ginger beer.
4
PUTTING THE TRENCH IN THE RIGHT PLACE
__________
Working out where we should dig in the next excavation season in 2005 was difficult. All our trenches in 2004 had been in the wrong places. Only the trench halfway between the river and the henge had found Neolithic remains, and even that trench was not quite where it should have been. Had we known in advance how severe the erosion was immediately upslope from the riverside, we would have positioned the trench westward, further up the slope, where preservation was better. In 2005 we could undo the mistake and put the trench in the right place.
This might be our only chance to look for what we had predicted in our theory about Durrington Walls. We had to open a trench that would produce undisputed evidence of concentrated human settlement. The pits that we found in the first year’s digging were full of feasting rubbish but that wasn’t enough proof – they could be the remains of some one-off event, like Coneybury. If we didn’t find an avenue or better evidence that Durrington Walls had been a place of the living, we’d have to pack up and go home for good.
If there were an avenue or roadway outside Durrington Walls, perhaps a new trench into the better-preserved zone, south of where we had found the pits, would pay dividends. If we found an avenue here, then we were on the right track; if not, Ramilisonina’s theory could be dismissed. It was also worth extending the trench northward beyond the group of pits. Kate Welham’s geophysics team had found an anomaly in this area. They had detected it using an earth-resistance survey, which measures the resistance to electrical current below ground. High resistance is created by dense and dry features, such as walls, whereas low resistance is found in damper areas, such as the fills of ditches and pits. The anomaly north of our 2004 trench was a large circular area of high resistance; it could be a small henge or a round burial mound.
One of the most exciting moments of archaeological excavation is the removal of plowsoil by mechanical excavator to see what lies beneath. However much research has been done above ground to work out what is there, the mechanical digger always reveals surprises. Machine stripping of topsoil is a delicate task: Take off too little, and what remains has to be laboriously removed by hand; take off too much, and you’ve destroyed valuable archaeological layers. It all depends on the sharp eyes of the archaeologist and the skilled hand of the digger driver.
The 2004 season had shown us that well-preserved Neolithic remains lay immediately at the base of very shallow plowsoil, buried as little as 0.15 meters (six inches) deep. As I watched the hired digger prepare the trenches in the summer of 2005, the driver gently removed the topsoil to reveal a layer of black soil filled with animal bones and potshards. This was an entire ground surface that had been preserved for more than four thousand years—a complete surprise. Not only was the ground strewn with settlement waste (what archaeologists call “midden”), but it had also survived four millennia of weathering and erosion.
The chalklands of Wessex are famous for their archaeological remains of all periods, but normally an excavator sees only the lower parts of features, such as pits, ditches, and graves, that have been cut into the chalk bedrock. The uppermost layers have usually been destroyed over time by natural and man-made processes. Rainwater percolating through soil on to the surface of the chalk reacts with the calcium to form a weak hydrochloric acid that eats away at the upper layers of chalk; meanwhile plowing not only breaks the surface to accelerate this weathering but also actively erodes the chalk. Most of the uplands of Wessex have been plowed since prehistoric times, and studies have shown that nearly a meter of chalk has been removed. We therefore rarely see a view of the Neolithic at the original ground level—this has vanished completely—but are instead usually looking at things that were once below ground. We now know that, even at Durrington Walls, parts of the henge interior have lost at least half a meter of chalk since it was first constructed. So how had the ground surface in our new trench been preserved?
Ground-surface preservation can be expected beneath substantial earthworks, such as a henge bank, but our new trench was at least 10 meters from the edge of Durrington Walls’ bank. It soon became clear that the bank had been quarried into a couple of centuries ago by people in search of chalk to puddle into cob (a traditional material for making walls) or lime (to spread on the more acidic soils of the river’s floodplain). The quarrymen, who left behind a button and pieces of their clay pipes, had spread chalk rubble well beyond the edges of the henge bank, inadvertently sealing the Neolithic ground surface beneath.
It was this circular-shaped spread of more recent rubble that had shown up on the resistivity pl
ot. It had provided protection from nineteenth- and twentieth-century plowing, but how had the old surface survived earlier erosion? We later discovered that this small part of the field, in front of the henge entrance, had for some reason been spared the plowing that had commenced in the Iron Age. Evidence of the farming landscape of the Iron Age can still be seen today: The hedge lines of Iron Age fields are preserved at Durrington as lynchets—ridges still visible in grassland today throughout Britain—formed by plowsoil accumulating against a hedge line on the upslope and by plow erosion against the downslope of the hedge.
As soon as the digger was off the trench, we troweled off the last of the protective chalk lumps, and could then see a blackened surface that extended beyond the limits of the trench. Within it were heaps of burned flint, low mounds of piled-up chalk (the up-cast from Neolithic pits) and spreads of ashy soil. The southernmost of these ash spreads extended just as far as the edge of our 2004 trench. As we troweled over the top of it, we could see that the thin layer of ash lay on top of a square area, just over two meters across, of solid chalk plaster, at the center of which was a circular depression. This was where some of the soliflucted chalk extracted from the Neolithic pits we found in 2004 had been deposited. Both Colin and I recognized what we were looking at: This was a house floor, not very different from the ones that we’d been digging for the last fifteen years in Orkney and the Outer Hebrides. More than that, it was very similar in size and plan to the Neolithic houses that Colin knew so well from Orkney.
The floor of House 547 at Durrington Walls. The white rectangular area is the chalk-plaster floor in the center of the house. Within it, the dark circular area is the hearth. A line of stakeholes, showing where the wattle-and-daub wall once stood, surrounds the house.
Stonehenge—A New Understanding: Solving the Mysteries of the Greatest Stone Age Monument Page 7