An entire Neolithic settlement, predating Skara Brae, has been found on the tiny Orkney Island of Wyre. Replete with unusual features, it is set to rewrite the story of Orkney’s first farmers, as Daniel Lee and Antonia Thomas explain.
Small places are often overlooked. This is certainly true of the island of Wyre. Lying at the heart of the Orkney archipelago, it is one of Scotland’s smallest inhabited islands. Covering just c.2.5km x 1.5km, it has a population of only 19. Even the majority of Orcadians have never visited and it tends to be missed off the usual tourist trail. Despite a wealth of Norse and post-Medieval history almost nothing was known about Wyre’s more ancient past.
In 2006, our Orkney Research Centre for Archaeology (ORCA) team undertook a walkover survey of the entire island. Immediately, we identified numerous sites, including kelp pits and animal enclosures, and we even found a previously unrecorded Bronze Age barrow. But then we came across the recently ploughed ‘Field 13’ in the Ha’Breck area – unlucky for some, but not for us!This is in striking contrast to Wyre’s much larger neighbours: the Mainland and Rousay. Dubbed the ‘Egypt of the North’, Rousay has one of the densest concentrations of prehistoric sites in Europe. Meanwhile, Orkney’s largest island, the Mainland, is home to some of the archipelago’s most famous sites, including Skara Brae, arguably Northern Europe’s best preserved Neolithic village. But has the focus of archaeological fieldwork simply biased the pattern of known sites? Could Wyre have anything equally ancient? We decided to investigate.
Lucky Field 13
The field surface was littered with hundreds of individual finds including pottery, coarse stone tools, flint, burnt bone and pumice. Some fieldwalkers were even fortunate enough to find a polished stone axe or macehead. The finds were certainly prehistoric, and appeared to be Neolithic. We decided to follow up with a geophysical survey across the densest part of the artefact scatter (an area measuring c.100m by 50m). The results showed several strongly-defined anomalies, which we suspected signified ancient buildings.
Could this be another Skara Brae? In 2007 we excavated a series of evaluation trenches across these anomalies to find out.
We placed our first trench (Trench A) over one of the strongest geophysical anomalies, which was also where we found the densest concentration of surface finds. This trench revealed a rich black midden (or waste dump) full of flint tools, Early Neolithic round-based pottery, and Skaill knives – flaked beach cobbles used as butchery tools. A hazelnut shell from this midden level provided an Early Neolithic radiocarbon date of 3300-3100BC.
In the final few days of excavation we found a line of large, well-laid flagstones, which appeared to form a pathway. Digging on, we found they led into a beautifully constructed stone entrance, complete with a threshold stone. It was the doorway to a Neolithic house. Incredibly, despite extensive stone robbing in prehistory and modern deep ploughing, 2-3 courses of stonework survived in parts of this house. The Neolithic builders had cleared the ground of topsoil down to the glacial till. They then cut the outline of a building into this natural clay, laying the foundation for a rectangular stone-built structure with rounded internal corners, and measuring approximately 8m x 4m (internally).
Large central upright stone slabs, known as orthostats, project from a slight pinch in the side walls, dividing the house into two. This is a classic Early Neolithic style, reminiscent of the Knap of Howar on the island of Papa Westray, a remarkably well-preserved farmstead that, until the late 1990s, was the only known Early Neolithic house in Orkney (see Box 1). So, far from finding a ‘new’ Skara Brae, we had discovered something older. Constructed during the 4th millennium BC it predated Skara Brae by several centuries, and was thus home to some of Orkney’s first farmers.
Continued excavation of the house revealed further orthostats creating a bay in the northern half, plus stone ‘furniture’ in the southern half. An axial post-hole in the centre of each of the two parts of the house suggests that the roof was supported by a central ridge pole that formed a frame, which sloped down to the stone wall heads. If Knap of Howar is anything to go by, these walls could have been 1.5 – 2m high. The northern post-hole had been reworked at least three times, suggesting that the building was maintained for a considerable length of time. In the last phase, an elaborate square stone-lined box was used to foot the post. This was left to rot in situ, revealing it was 20cm in diameter – not large timber in comparison to the rest of the UK, but a sizeable tree for Orkney and certainly enough to have supported a thatched roof.
A Neolithic granary?
Working on, we soon discovered that House 3 was not any old farmstead. The first (and earliest) phase of use in its northern part was particularly interesting: spread across the floor were several thick layers of charred material, 70mm deep in places, comprising tens of thousands of barley grains. We had found one of the largest assemblages of Neolithic cereal inScotland (currently under analysis by Rosie Bishop as part of her doctoral research at the Universityof Durham). The intensity of grain-production evidence in the house is so unusual it leads us to suspect that the building may have had a more agricultural function–a granary perhaps?
Whatever the case, there was something else unusual about this house: mixed into the burnt grain layers were several sizable chunks of wood charcoal, suggesting that a large fire had ripped through the building. The heat was intense enough to redden the underlying glacial till within the central northern part of the house. Yet both the huge quantities of grain and the fire seem to have been confined to the northern half of the building, while the southern portion – with its series of shallow pits and scoops cut into the floor – was largely unaffected, and continued to be used after the fire.
Dramatic as the conflagration must have been, it did not mark the end of life of the building. Neither was the structure cleaned out. Rather, the charred grain was pushed to the sides of the room, spilling into open pits and gullies, while the scoop hearth was sealed with a large flagstone slab. A quern rubber was placed on top on this charred grain layer, and the area affected by the fire was covered with a thick layer of mixed clay. Thus all evidence of the charred remains was hidden, while the quern was left visible above the new floor surface. The occupants then placed a small polished stone axe (one of the 15 recovered so far from the site) next to the quern and built-up further levelling layers and floor deposits. We found several further quern fragments elsewhere across the site, often broken and deliberately placed within closing deposits.
Why this ‘strange’ activity? All the evidence, from the vast quantities of grain involved to the actions of the farmers in the aftermath, makes us suspect that the fire was started intentionally. It was probably a ritual act, rather than the result of a series of accidents. Indeed, some Early Neolithic houses inIrelandalso appear to have been deliberately burnt down, suggesting buildings had a life cycle of their own, perhaps linked to that of their owners. Or perhaps something else was going on – a religious or social act reflecting the importance of grain in the lives of these early farmers?
A settlement emerges
The ‘granary’ did not stand alone, and we discovered another stone-built structure adjoining its southern wall. Labelled House 5, it consists of two rooms: one large and one small (with internal dimensions of 4.8m x 3.2m and 2.2m x 1.32m). The larger room has an internal opening connecting it with House 3, as well as an external entrance in its eastern wall. At the southern end of the main room, a narrow opening, set between two stone-built piers, leads into the smaller cell, giving the impression of a more ‘private’ space. A rectangular stone-lined hearth lay at the heart of the main room. Curiously, the hearth’s basal slab overlay a stone-lined drain that runs downslope before disgorging through an opening in the wall of the south cell. But why would a drain begin under a hearth? We can find no logical explanation.
As for the connection between it and the neighbouring ‘granary’, although House 5 is stratigraphically later, the two buildings appear to have overlapped for a time. The outer (eastern) wall of the ‘granary’ was given an additional ‘skin’ when House 5 was constructed. Moreover, though we have not yet obtained any radiocarbon dates for House 5, the finds all indicate a broadly similar date to the ‘granary’.
Both houses were robbed out in prehistory. The entrance of the ‘granary’ was sealed with redeposited natural clay before large amounts of black midden were brought in and deposited over the paving to the east of the house to form a levelling layer. To our surprise, when we excavated those midden deposits, we found the remains of a further house. Unlike the isolated Knap of Howar farmstead, we were on our way to discovering a whole settlement! But this house (number 4 on the plan) proved different to its neighbours: instead of stone walls, we exposed 14 post-holes – it was a timber building: a remarkable discovery for the Neolithic of Orkney, where this period is always assumed to have been dominated by stone architecture.
This is an extract, but the full feature is published in CA 268, on sale June 1.