The earliest excavated building
Lying within Trench J, at the northwestern end of the site, Structure Five held the distinction of being both the earliest excavated building at the Ness and the largest, non-funerary Early Neolithic construction found in Orkney to date.
Built around 3300BC, Five pre-dated the large piered structures in Trench P (Structures One, Eight and Twelve) by approximately 200 years.
Four centuries lay between it and Structure Ten – the last major building in Trench P, raised around 2900BC.
Structure Five was first revealed in 2005 and excavated until 2008. Trench J was reopened in 2017, when work in and around the building resumed.
The excavated section of Structure Five was sub-rectangular with an oval north-eastern end. Orientated north-east to south-west, the building stood beside, and was probably contemporary with, the northern boundary wall, or “Great Wall of Brodgar”.
The exposed elements of the building measured 16 metres long and 7.5 metres at the widest point. Geophysical survey suggested another 4.5 metres lay beyond the south-western trench edge, making the building over 20 metres long.
Initially, Five’s early date was based on its architecture, which was very similar to other known early Neolithic buildings, namely the Knap of Howar, in Papa Westray, and the Braes of Ha’Breck, in Wyre – both of which date from around 3300-3000BC.
This was confirmed in 2021, following the discovery of round-bottomed pottery in the building’s south-western end. More round-based pottery was found in the lower deposits of this area over subsequent years. This ceramic style generally preceded the appearance of Grooved Ware pottery around 3200BC.
The original building
Like all the Ness of Brodgar buildings, Structure Five went through multiple phases of use.
Although early, Five was very carefully built, with every element clearly planned. It began life as a rectangular structure, measuring over 14 metres long and 7.5 metres wide.
Although the later Ness buildings used midden for wall core, Structure Five had small stones and redeposited boulder clay in its double-skinned walls.
Access to the interior was by a single entrance in the south-eastern side wall. Inside was huge, 1.95-metre-long, rectangular hearth with stone furniture features lining the walls.
Unlike the later buildings, the interior of was not divided by orthostats or piers.
It began life as a rectangular building, measuring over 14 metres long and 7.5 metres wide. While the later Ness buildings used midden for wall core, Structure Five had small stones and redeposited boulder clay in its double-skinned walls.
At this stage, the interior was not divided by orthostats or piers.
3d model of Structure Five’s stone-based hearth…
…and the underlying channel.
Furniture features
In the excavated section of Structure Five’s primary section were the remains of eight stone furniture features against the interior wall faces. The upper sections were long gone, making interpretation difficult but, in 2022, it was suggested at least some might be variants of Skara Brae-type “dressers”.
“Dressers” have been encountered at the Ness before and have become iconic symbols of Skara Brae, where visitors can still see complete examples. There they were built to the same design and placed in the same position — directly opposite the house entrance.
The term “dresser” is a hangover from the 19th century antiquarians, who basically saw the stone edifices as simple display cabinets – a place to put your best pottery and other prized possessions on show.
But the significance and role of the “dressers” has been questioned over the years. Were they more than just a set of shelves? Their presence in Structure Ten, reignited the debate.
Skara Brae’s “dressers” were built against the walls, but Ten’s primary dresser was free-standing and incorporated slabs of striking red and yellow sandstone — stone deliberately brought to the site to create the dresser.
Considering the non-domestic role of Structure Ten, it is possible that these “dressers” had a function beyond storage or display. However, this raises the question of why Structure Five had at least eight.
What they represent is still open to question. Five’s north-eastern feature (pictured right), for example, was found to extend across most of the width of the building. Do we have a bench-like feature akin to those encountered in stalled cairns? But for the living?
Whatever their role, the presence of beautiful stone tools, buried in the clay floor next to each “dresser”, strongly suggested the furniture held some significance to the builders of Structure Five.
Post-holes
An absence of stone tiles inside Structure Five suggested it was roofed differently to the other, later, buildings on site. This made sense of a series of post-holes discovered at the base of its inner walls.
In 2022, two large, well-built post-holes showed the doorway was flanked by a pair of substantial timber posts. Another four were subsequently discovered, forming a rectangle approximately seven metres by four metres. These held large posts, measuring up to 0.40 metres in diameter.
Their location along the bottom of the interior wall strongly suggests they were integral to the building and probably supported a turf roof.
Excavated in 2023, a layer of black material on top of floor deposits in the building’s primary section may represent the vestiges of a dismantled or collapsed turf roof. Turf has also been suggested as the roofing material used at the nearby Barnhouse settlement.
The use of posts in its construction was a feature unique to Structure Five. Where evidence of posts was found elsewhere on site it was for hasty structural repairs, such as that encountered in Structures Eight and Twelve, due to subsidence.
Structure Forty-One – the earlier, timber building
The discovery of post-holes led to the suggestion that they represented the remains of a timber building that preceded Five.
The posts were similar in size to those encountered in the much-smaller Timber Structure One at the Wideford Hill Early Neolithic settlement. There, excavation in 2002 and 2003, not only revealed evidence of early timber buildings but a switch from wooden construction to stone around 3300BC.
In 2024, the evidence for an earlier building stacked up. Several earlier features, including two hearths multiple channels/gullies, were found to pre-date the stone walls and clay floor of Structure Five.
As a result, we now believe the posts used in Structure Five were part of an earlier, wooden building – Structure Forty-One – that stood on site. Five’s stone walls were built around the still-standing posts of the original hall, after its timber walls were dismantled. Inside, a new hearth was constructed, new drainage channels dug and a firm clay floor laid from wall to wall.
The 3300BC date proposed for Five fits the timber-stone transition period suggested for other Orcadian sites. The incorporation of substantial wooden posts into Structure Five may have been more than just structural. It not only connected the stone building to its timber predecessor, but perhaps enhanced Five’s grandeur as well as emphasising the builders’ status.
If a dwindling supply led to the transition from wood to stone, that would have seen timber become a more valuable resource. This may have contributed significantly to the prestige of procuring, transporting and using the large posts within a structure.
That impact may also have been enhanced if they were part of an earlier, significant building. With this in mind, and given the quantity of stone artwork across the Ness site, it is worth pondering whether Five’s internal posts were decorated or carved.
Early occupation
Inside Structure Five, the depth of the occupation deposits varied.
In some areas they were extremely shallow, suggesting the building was not in use for long. Alternatively, we may have a situation where the floor was being kept extremely clean – as was noted in other buildings – or different areas were being used differently and perhaps had specific roles.
Ongoing studies of the floor deposits by XRF and micromorph analysis should clarify this.
Structure Five extended
During the primary section’s life an oval extension was tacked on to its north-eastern end. This five-metre-long addition was probably separated from the primary occupation area by its original north-eastern end wall, which now partitioned Structure Five into two “rooms”.
Again, it is difficult to tell because of stone-robbing, but it seems likely that a doorway was inserted into the dividing wall’s eastern end to provide access between the two sections.
Around the same time, a stone porch-like feature was added outside Five’s original entrance as part of an operation to rebuild the south-eastern wall.
At the northern, curved end, a large drain was also inserted into the wall, probably around the time the timber was replaced by divisional orthostats inside.
The discovery of a post-hole at the end of the 2024 season suggests the extension also incorporated timber posts in its construction.
Given the volume of ashy material encountered, it also had a central hearth – although what form it took is unclear. The hearth was replaced twice: once with the change from posts to orthostats, and again more later, perhaps around the time the extension’s outer wall face was rebuilt.
Dismantled and decommissioned
Not long after the addition of the extension, Five’s original, rectangular section may have suffered a calamity, resulting in its decommissioning and partial dismantling.
Not long after the addition of the extension, however, Five’s original, rectangular section may have suffered a collapse, resulting in its decommissioning and partial dismantling.
The timber posts were carefully removed, suggesting the building’s roof had already gone. Artefacts were placed at the bottom of the empty post-holes, which were then backfilled. Each “dresser” was dismantled and the stone taken away – apart from the one against the south-east wall, where the “shelf” was broken.
The wooden posts in the extension were also removed and radial, upright stone slabs inserted to divide up the interior. A new doorway was also inserted in the rebuilt north-eastern extension after the original entrance, a few metres to the south-west, was blocked.
An arc-shaped wall, constructed from large boulders and flagstones, was built to contain this blocking material and to separate the former doorway from the rest of the building’s remains.
The careful removal of the timber posts again suggests that wood was a significant, and valuable, resource. It also implies it was destined for re-use in another construction, but where and how is not known.
Occupation continues
With the primary section dismantled, occupation continued within the north-eastern extension, which saw some poor-quality rebuilding and remodelling. Some of this later reconstruction sat on top of the detritus associated with the destruction/robbing of the “Great Wall”.
Structure Five had been raised on natural glacial till – the original ground surface 5,000 years ago. As a result, the building’s floor was lower than the surrounding ground level, which meant there were problems with water ingress.
This accounts for the clay sealing found on the lower courses of Five’s external walls – a feature also encountered at the Early Neolithic settlement site at Smerquoy, at the base of Wideford Hill.
The clay barrier replaced a drainage gully cut into the till outside the original building’s walls. How well it worked is open to question given the later addition of the drain at the north-eastern tip.
In the area of the original rectangular building, amid the rubble, sections of walling still stood. These were robbed to provide building material for Structure Thirty-Two, which also incorporated part of Five’s still-standing north-western wall into its fabric.
Occupation deposits in the north-eastern end were much deeper and complex than those in Five’s primary section, meaning the extension was in use for a much longer period of time.
The presence of Grooved Ware pottery in the later phases of occupation confirms Five continued to be used after the construction of the piered buildings around c3200BC – something confirmed by the fact that part of the rebuilt extension stood on top of rubble relating to the ‘Great Wall’s’ demolition.
At the end of its life, Structure Five’s sole surviving entrance was deliberately blocked – probably a deliberate act of closure/decommissioning.
Where the final abandonment fits into the Ness of Brodgar’s timeline is unclear, but it seems likely the building was still in use after the construction of the first phase of piered buildings (e.g. Structures Seventeen, Eighteen and Twenty-Eight) in Trench P and perhaps as late as the final phase of piered buildings (Structures One, Eight, Fourteen and Twelve) around c3100BC.
This may also go some way to explain why no major buildings were constructed on top of Structure Five. The area continued to be used, however, as shown by the multiple hearths and features encountered within the midden and rubble overlying its remains.
3d models
- Trench J (August 2024)
- Round-bottomed clay pot from north-eastern extension. (July 2024)
- Structure Five and the ‘Great Wall’ (September 2023)
- ‘Dresser’ remains in Structure Five (July 2023)
- Structure Five and the ‘Great Wall’ (August 2022)
- Worked whalebone artefact (July 2022)
- Structure Five (July 2021)
- Trench J (August 2019)
- Trench J – end of season (August 26, 2018)
- Trench J (August 7, 2018)
Notes
- [1] Card, N. and Edmonds, M. (2020) Setting boundaries. In In Card, N., Edmonds, M. and Mitchell, A. (eds) The Ness of Brodgar: As it Stands. The Orcadian: Kirkwall.
- [2] Card, N. and Edmonds, M. (2020) Early Buildings. In Card, N., Edmonds, M. and Mitchell, A. (eds) The Ness of Brodgar: As it Stands. The Orcadian: Kirkwall.
