Rapid erosion has revealed spectacular Neolithic and Bronze Age archaeology on the coast of Westray, Orkney. Contemporary with the Ness of Brodgar’s religious monuments but with a domestic focus, what can this settlement tell us about daily life in prehistoric Orkney? Hazel Moore and Graeme Wilson explain.
Overlooking the North Atlantic on the island of Westray, the Links of Noltland has an impressive prehistoric landscape stretching over 4ha. Comprising the well-preserved remains of over 20 buildings — including Neolithic structures contemporary with, and comparable to, the famous ‘village’ at Skara Brae — together with extensive middens, field systems, and a cemetery, the site is revolutionising knowledge of Neolithic and Bronze Age Orkney.
Noltland’s wealth of archaeological features is in danger of being lost, however. Facing into the wind and exposed to almost constant salt spray, the site is at severe risk of erosion, with the dune system that has protected it for millennia rapidly depleting. Designated a Scheduled Ancient Monument and a property in the care of Scottish ministers, managed on their behalf by Historic Scotland, the settlement has been closely monitored for change since the 1980s. By 2005 it was clear that — for reasons still not fully understood — the scale of erosion was accelerating at an unprecedented level. Matters had become urgent.
In response, Historic Scotland launched a rolling campaign of assessment and conservation works, and rescue excavations undertaken by EASE Archaeology, directed by the authors and project-managed by Historic Scotland Senior Archaeologist Richard Strachan. Since 2006 these have revealed a large number of hitherto unsuspected Neolithic and Bronze Age remains, with a highlight of the most recent season being the discovery of two carved stone figurines similar to the celebrated ‘Orkney Venus’ (CA 236). Local soil conditions favour the preservation of skeletal material, meaning that there is a large amount of animal bone available for study, providing valuable opportunities to examine husbandry and butchery practices. Such bones, from both wild and domesticated animals, were crafted into a wide range of tools and decorative objects, from beads and elaborate dress pins to mattocks, awls, polishers, and points. Bone-working debris is rarely encountered on archaeological sites of this period, and study of this material is providing a glimpse of manufacturing processes employed by prehistoric craftsmen.
While the Neolithic remains at the Links of Noltland bear comparison with those at Skara Brae on Orkney Mainland in terms of both age and architecture, at Noltland erosion of the ground surface over such a large area has permitted a far more extensive investigation of the site’s hinterland, making it possible to examine the settlement’s evolution over a long duration. The inhabitants saw dramatic changes to both their built and natural environments during the settlement’s lifespan, which began in the 3rd millennium BC and endured into the Bronze Age.
The earliest occupational evidence revealed by our excavations so far was a farmstead, dated to at least 2800-2500 BC. Set inside a stone-walled enclosure, this complex stood on a ridge surrounded by cultivated fields, with its finest, and perhaps earliest, building (Structure 10) in the centre, constructed from neatly coursed quarried stone. Over time, more buildings were added and older elements were modified and reused, creating a series of closely packed rooms and passages. We investigated the interiors of two (Structures 10 and 19) during the most recent phase of work. These represent two of the larger rooms, both rectilinear in form, with upright stones used to divide up internal space.
In Structure 10 we found structural modifications suggesting that the building had seen later phases of occupation, with tantalising glimpses of the original walls visible beneath. These were extremely well constructed, and it appears that the interior was deliberately backfilled with midden material and rubble prior to this later use. Among the infill material we discovered a carved stone ball — an enigmatic type of prehistoric artefact found mostly in Scotland, with five unearthed at Skara Brae alone. While the purpose of these objects is still open to debate — suggestions range from ceremonial use to a projectile for taking down wild animals — this was a significant find, since only a few have been discovered in secure archaeological contexts. Most come to light as stray finds. Structure 19, by contrast, appears to be one of the later buildings within the complex. While radiocarbon-dating results are still awaited, the structure proved to have been built over the original enclosure wall. It was entered via a narrow entrance passage that opened into a central floor area surrounded by peripheral recesses or ‘box-beds’, which were separated from the main space with upright slabs. One measured over 2m in length. Opposite the entrance were the footings of a dresser similar to those seen in houses at Skara Brae, while later floor layers inside the structure produced very large quantities of decorated Grooved Ware pottery dating to the early 3rd millennium BC, together with a wide range of stone and bone tools.
The settlement is surrounded by extensive contemporary middens. These are directly responsible for the preservation of many of the site’s buildings, absorbing the structures as they fell out of use. They are also proving a productive source of new information about the settlement. While Noltland’s house-proud inhabitants generally kept their floors clean, discarding few artefacts inside the structures, the middens hold a vast amount of material, in places reaching over 1m in depth.
We have excavated considerable areas of midden to expose the structures hidden beneath. In so doing, we recognised that at Noltland these were not merely refuse heaps — they were used for a variety of activities, including animal butchery and craftworking. Stone pathways lead through the deposits, while specific areas appear to have been reserved for specific activities. There are butchery zones, for example, where we found rough ‘skaill knives’ made from split beach pebbles, as well as worked flints that would have been used to dismember animals. Elsewhere, caches of tools such as bone mattocks and bead-making debris suggest that, in addition to sourcing their raw materials, bone implements were also being manufactured here.
Close analysis of this wealth of discarded material has provided many details of life at Neolithic Noltland. We now know that the inhabitants were predominantly cattle farmers, but also kept sheep; that they had access to abundant wild resources including numerous species of bird and fish, together with deer, marine mammals, and shellfish; that they cultivated barley, and that their dogs regularly came to gnaw at meaty scraps of bone. We can reconstruct other aspects of their world from the farming and craftworking tools that they left behind, alongside decorative items such as dress pins and beads, as well as worked shell.
One of the most exciting aspects of the midden investigations has been the discovery of bizarre ‘compositions’ — these are consciously and sometimes elaborately arranged groups of materials. In one instance, a scallop shell was placed between the horns of a sheep skull, while a flint tool was set inside the skull. In another composition, numerous animal jawbones were arranged together, perhaps votive offerings associated with the killing and butchering of animals. We have also noticed composite items of bone and clay, equally tantalising, but regrettably less well preserved.
Outside the clustered farmstead, several other Neolithic buildings have been identified during our work, including two houses with a cruciform interior. The first (Structure 9), located just outside the enclosure wall, had 28 cattle skulls, two of which have been dated to the mid 3rd millennium BC. Deliberately placed within its foundations, they would have been an important gesture from this community of cattle farmers. Standing further apart, the other building (Structure 7) — home to the site’s second dresser — seems to have been enclosed by a series of ‘casement’ walls — concentric ‘skins’ of stone, producing hugely thick structures.
The most complete building to have been excavated at Noltland so far is a subterranean house and annex (Structure 18), isolated from the other structures and of very different construction. Dubbed the ‘Grobust house’ after the bay it overlooks, the structure was built in a large pit cut into a sand dune and comprises two unequalsized rooms joined by a passage. While Noltland’s other buildings are freestanding, the Grobust house has revetted dry-stone walls. In places still standing up to 1.1m high, these are probably preserved to almost roof height.
Discovered and partially excavated in the late 1970s by Dr David Clarke of the National Museums of Scotland — who revealed that part of the building at the end of its life may have been deliberately filled with soil, from which large numbers of flint tools, worked bone, pottery, and stone objects, were recovered — there was a lag of over 30 years before work resumed. This was the main focus of our 2012 excavation, during which the last remnants of infill were removed. With the interior of the house fully uncovered for the first time, we were finally able to explore its entire layout.
Lying beyond a narrow passageway lined with lintels, the larger of the two rooms had a series of alcoves arranged around a central floor with a stone kerbed hearth, as well as a shelved recess opposite the doorway — the same location as freestanding ‘dressers’ seen in Structures 7 and 19, and at Skara Brae. Few artefacts were found here, with the exception of caches of flint, including several arrowheads, which were recovered from around the central hearth.
In this room we also found a neat stack of stone slabs, and several large whalebone rib fragments, possibly dismantled roofing materials. In Orkney’s near treeless environment, whalebone would have provided valuable material for rafters and structural elements. It seems likely that the roof of this room had been deliberately removed prior to its abandonment. Although neither the entrance passage nor the smaller room appear to have been physically ‘decommissioned’ in this way, further clues to the building being ritually closed came from its entrance passage, where 18 cattle skulls and six sheep skulls were placed on the floor, together with two more sheep skulls set in bar holes on either side of the door check. Perhaps mirroring the cattle skulls found in the foundations of Structure 9, this gesture could have marked the end of the building’s use.
Across the passage was a smaller room with a little alcove opening off one side. The room originally had two entrances, although one was later blocked up, and its limited floor space and partially corbelled walling would have made it somewhat cramped and difficult to move about — though it may not have been intended as living space. Quantities of animal bone, mostly from sheep or goats, and much of it still articulated, were strewn across the floor, together with large, unworn fragments of pottery that do not seem to have been significantly disturbed since their deposition. These findings could indicate that the room served a specialised function — something certainly suggested by the discovery of a small stone figurine amid this material. This was the third carving depicting a person to be found at Noltland, following the ‘Orkney Venus’ (or ‘Westray Wife’) found in 2009, and a second, ceramic, figure dubbed the ‘Headless Husband’ in 2010 (CA 247). It was not, however, the last enigmatic development at the Grobust house.
During the final weeks of excavation, we unblocked an aperture in the wall of the room where the figurine was found, revealing a small and hitherto unsuspected cell lying behind, robustly constructed from quarried stone bonded with yellow clay. Its function was not readily apparent, as the aperture was too small to admit an adult and, in any case, had been blocked early during the building’s lifespan. Our work took another exciting turn as an object set into the floor nearby was revealed to be yet another figurine, somewhat larger than those found previously. The surface detail is worn, but its features are unmistakeably similar to those of the ‘Orkney Venus’, with two eyes, a possible nose, and a probable ‘eyebrow’ motif. It is hoped more details will become visible during conservation.
This remarkable object and the ‘hidden’ cell in which it was housed are so far unparalleled in Scotland and their interpretation will doubtlessly invite much speculation. What seems clear, however, is that they represent aspects of a belief system that is intimately associated with the settlement’s domestic life. As with the use of cattle skulls as foundation and closing deposits, these gestures would have been obscured from public view after deposition, and may have been known about only by the occupants and their associates. In these respects, they provide an interesting and very personal counterpoint to the conspicuous monumentality of the approximately contemporary ritual complex at the Ness of Brodgar (CA 241).
Bronze Age inundation
Noltland’s spectacular archaeology is not limited to Neolithic remains: the site is also home to the most extensive Bronze Age settlement yet discovered in Orkney, with some 12 buildings identified to date, the largest an oval structure measuring 11m by 9m. These are particularly vulnerable to erosion, as unlike the earlier structures they were not normally covered with midden material or deliberately filled in after abandonment. They have also given us important insights into life in prehistoric Orkney — not least because we have found the burials of dozens of the settlement’s inhabitants.
Bronze Age buildings at Noltland generally occur in pairs or small groups, each cluster consisting of a house and one or more annexes. These are often arranged with their entrances facing each other across a paved pathway — a form commonly encountered in Shetland, but not widely seen elsewhere in Orkney. Similarities in the Grobust house’s layout might suggest it represents a forerunner to the site’s more developed Bronze Age buildings. In each, the largest building of the pair contains a central hearth, surrounded by either raised or paved floor areas. The annex buildings are more varied, some divided into stalls, perhaps for storage or housing animals, while others contained one or more stone-lined tanks. In two cases, large pits cut into the floor and lined with stone tools have been tentatively identified as grain-storage pits.
New building styles were emerging at this time, and artefacts recovered from the settlement also reflect the introduction of new ideas. Although fewer objects have been found compared to the Neolithic part of the site, flint and worked bone are noticeably less common, while new materials such as steatite, imported from Shetland, make their first appearance. The middens of this period are also different in their makeup, showing dramatic changes to the local diet. Although animal bone is still present, discarded food waste is predominantly composed of shell. There are also few man-made objects. It is not yet clear whether this paucity of artefacts is linked to a decline in the community’s material circumstances or whether survival from this period has simply been poorer, but it is apparent that this was a time of significant local environmental change. Analysis of the remains of cultivated field systems illustrates this vividly, revealing farmers battling to consolidate their land through intensive composting in the face of massive incursions of marine sand. Whereas in the Neolithic period sand cover was limited to low-lying areas and not apparently mobile, by the Bronze Age it had begun to arrive in ever greater quantities, swamping whole fields and wiping out crops. The effect of this on the community’s ability to maintain their farms and to provide sufficient food for themselves and their livestock is still being investigated. The massive increase of shellfish in the inhabitants’ diet may, though, suggest that wild resources were becoming increasingly important.
Further opportunities to study the health of Noltland’s prehistoric population came with the discovery of a cemetery, contemporary with the Bronze Age settlement, containing at least 65 burials ranging in date from c.1700 BC-1400 BC. Individuals of both sexes and all ages have been found here, with cremation and inhumation burials occurring throughout the period that the cemetery was in use. There are also several instances of multiple individuals being interred in a single grave. In one instance, a total of 15 inhumations were identified in a stone-lined cist. These were successive rather than contemporary burials, indicating that the grave had been reopened on several occasions, perhaps serving as a family vault. Several of its occupants — particularly young children and babies — were accompanied by grave goods, including whole or fragmentary ceramic pots, and objects made from shell or bone. The burial ground revealed a wide range of funerary practices taking place at this time, but, perhaps surprisingly, it did not corroborate ideas of a community struggling to feed themselves in a hostile environment. Most of the individuals interred here for which data are available appear to have been healthy and not to have suffered from diseases associated with poor diet. Clearly life at Noltland at this time was challenging, but not insurmountably so.
This was not to last, however, as there is no evidence for settlement on the site surviving beyond the Late Bronze Age. The continued influx of sand made cultivation increasingly difficult, if not impossible, and under these conditions it is not surprising that the inhabitants decided to abandon their homes and move on. About half a mile away, on the coast at Queena Howe, lies an Iron Age settlement comprising a broch around which a village developed. It is tempting to wonder whether this represents the successful transplanting of the Noltland community to a more forgiving home.