Pool Neolithic settlement, Sanday

By Sigurd Towrie

Regular readers will have seen us mention Pool and Toftsness before. These were the locations of two Neolithic settlements excavated in the island of Sanday in the 1980s.

Today, we’ll briefly look at Pool on the island’s south-western coast.

The excavation, between 1983 and 1988, concentrated on an eroding settlement mound that was found to span millennia of activity. After Neolithic occupation, which the excavators suggested began in the early to mid-fourth millennium BC [1], it went out of use until the Iron Age, after which it became a Pictish/Norse settlement.

The eroding mound, however, was just part of a more extensive area of settlement and the earliest evidence of Neolithic activity did not come from the eroding mound but from deposits some 50 metres to the north-east.

Because these deposits, within an anomaly approximately 40 metres in diameter, were not under threat, they were only partially investigated. This confirmed they were part of the same settlement sequence as the eroding mound.

Sub-circular, with a diameter of 70-80m, roughly half of the mound was investigated but this was “only a fraction of the extensive remains” in the area. [2]

An 80-metre-long spread of archaeological features and a fourth site to the east were left undisturbed.

Excavation suggested three main phases of activity with two episodes of sand blow, the latest of which may have seen the site abandoned at least once. [3]

The coastal site was “almost certainly” an inland Neolithic settlement, which like Skara Brae, was separated from the sea by an area of land and an inshore lagoon. [3]

Although thermoluminescence dates from the earliest Neolithic phase placed it in the middle of the fourth millennium BC [2] doubt has been cast on these. [4]

With no radiocarbon dates for the earliest activity, the most that can be said is that the excavated area was in use before 3100-2890BC [4]. It should be noted, however, that the earliest occupation evidence did not come from the main mound but from the unexcavated area to the north-east. [2]

The Neolithic layers at Pool contained 14 structures in three suggested phases.  The earliest contained sherds of round-based pottery and the remains of three poorly preserved, small structures. [4]

The second phase saw a “slight shift in geographical focus” but no structural remains were identified. This period was punctuated by two layers of sand, interpreted as episodes of sand blow that covered the settlement.

There is, however, a difference of opinion as to the impact these had on occupation.

MacSween et al suggested the sand layers related to sandstorms after which “the site may have ceased to be occupied” [4]. However, despite describing the site as being twice “completely subsumed under major sand blow events” [1], to the excavators, the sand horizons were “unlikely to have had any major effect” because there was no evidence they were “catastrophic to the continuity of the settlement”. [2]

Although the question of occupation breaks in the second Neolithic phase remains unresolved, the second sand blow was followed by another shift in settlement focus and the construction of three, more substantial, buildings, of which only two were excavated.

The end of the second phase was marked by a clear, and lengthy, hiatus – between 115 and 235 years [4] – before a return to the site saw the construction of eight structures of a “more monumental type of architecture” [2] and the introduction of a “significantly different” type of Grooved Ware pottery. [4]

Among the finds from the largest surviving building on site, Structure Eight, were shaped stone objects, akin to those found at the Quoyness chambered cairn in Sanday, and a fragment of geometrically incised stone. An example from Pool and two from Quoyness are shown below.

At the end of its life, Structure Eight was deliberately blocked up and infilled. The uppermost layers of the filling material contained deer antlers, which appeared to have been deliberately deposited in a manner suggesting “the building had been formally decommissioned.” [5]

The Neolithic occupation at Pool appears to have ended around 2400BC, after which there was no activity until the Iron Age. [4]

Although the radiocarbon dating programme at Pool suffered setbacks, a chronology for the site was proposed. Although no dates were available for its founding, the second phase began in 3100-2890BC and ended in 2805-2735BC. [4]

The third, and final, phase began in 2625-2545BC and ended, along with Neolithic occupation of the site, in 2455-2370BC. From these dates, it is suggested that the second phase lasted between 195 and 360 years and the third, 105 to 250 years. [4]

After three phases of Neolithic activity, the excavators found no discernible reason for Pool’s abandonment [1] – although only sections of the final Neolithic phase had survived later Iron Age re-occupation. [2]

That said, there was no evidence for “decline nor catastrophe” and “neither environmental nor climatic factors” were responsible. Instead, it was suggested the inhabitants simply left for “move favoured parts” of the island. [1]

Although entirely possible, there is considerable archaeology around this single mound. Because of the partial nature of the excavation, are we simply seeing sections of a larger settlement going in and out of use over time?

Like Skara Brae, it is impossible to say how much of the seaward side of the Pool Neolithic settlement have been lost and the presence of Neolithic deposits to the north-east suggests there are more chapters in Pool’s biography.

The Sanday site has intriguing parallels with Barnhouse, with an apparent shift to monumental architecture in the run up to its hypothetical demise.

The last Neolithic building in the excavated mound, Structure Fourteen, was described as “undoubtedly tomb-like” although there was no evidence indicate that a funereal or mortuary role. [1]

Iron Age disturbance meant little of the building had survived, but it had been constructed on a platform and incorporated parts of an earlier structure. [2]

The partial remains make interpretation difficult but are we looking at another monumental building – such as those at the Ness of Brodgar and Barnhouse’s Structure Eight – rather than a chambered cairn?

Whatever it was, Structure Fourteen remained a focus right through until the later years of viking settlement:

Notes

• [1] Hunter, J. R. (2007) The Neolithic. In Hunter, J. R., Bond, J. M. and Smith, A. N. (eds) Excavations at Pool, Sanday, a Multi-period Settlement from Neolithic to Late Norse Times: Investigations in Sanday, Orkney Volume 1. The Orcadian, in association with Historic Scotland: Kirkwall, pp. 9–19.
• [2] Hunter, J. R.(2007a) Introduction to the Investigations. In Hunter, J. R., Bond, J. M. and Smith, A. N. (eds) Excavations at Pool, Sanday, a Multi-period Settlement from Neolithic to Late Norse Times: Investigations in Sanday, Orkney Volume 1. The Orcadian, in association with Historic Scotland: Kirkwall, pp. 9–19.
• [3] Hunter, J. R., Bond, J. M. and Smith, A. N.(2007) Overview. In Hunter, J. R., Bond, J. M. and Smith, A. N. (eds) Excavations at Pool, Sanday, a Multi-period Settlement from Neolithic to Late Norse Times: Investigations in Sanday, Orkney Volume 1. The Orcadian, in association with Historic Scotland: Kirkwall, pp. 515–522.
• [4] MacSween, A., Hunter, J., Sheridan, A., Bond, J., Ramsey, C.B., Reimer, P., Bayliss, A., Griffiths, S. and Whittle, A. (2015) Refining the chronology of the Neolithic settlement at Pool, Sanday, Orkney: implications for the emergence and development of Grooved Ware. In Proceedings of the Prehistoric Society (Vol. 81, pp. 283–310). Cambridge University Press.
• [5] Hunter, J.R. (2000) Pool, Sanday and a sequence for the Orcadian Neolithic. Neolithic Orkney in its European context, pp.117-25.