Subscribe now for full access and no adverts
Hundreds of crannogs have been identified within the lochs and inlets of Scotland, with a particular concentration in the Outer Hebrides, while examples are also known in Ireland and Wales. Many of these sites have never been scientifically investigated, and it was long thought that they had been built and used between the Iron Age and the post-medieval period. This picture began to change, though, in the 1980s, when the excavation of Eilean Domhnuill in Loch Olabhat on North Uist produced evidence suggesting that it had been created in the Neolithic period, between 3800 and 3300 BC (see CA 127). For decades, Eilean Domhnuill remained a unique case – but, from 2012 onwards, a local diver, Chris Murray, reported finding Neolithic pots scattered on loch beds around crannogs on the Isle of Lewis, and subsequent investigations confirmed that these sites were similarly early in construction (CA 325 and 354).
Further clues emerged during underwater surveys and excavations led by Professor Duncan Garrow (University of Reading) and Professor Fraser Sturt (University of Southampton) in 2016-2017, and since 2021 the Islands of Stone project (carried out by both universities, funded by the AHRC) has focused on one particular crannog in Lewis’s Loch Bhorgastail. There, the team have carried out excavations, coring, surveying, and radiocarbon dating, and the results of these investigations (recently described in a paper published in the journal Advances in Archaeological Practice; see ‘Further reading’ below), have revealed the crannog’s construction and chronology in illuminating detail.
(Re)Constructing a crannog
Thanks to the project’s findings, we now know that the first phase of the Loch Bhorgastail crannog was created over 5,000 years ago, and, although the above-water portion of the island resembles a stone cairn, its original incarnation was timber. The first clues to this construction emerged in 2021, when an initial small trench (rather unexpectedly) revealed layers of brushwood and silt beneath the crannog’s stone ‘base’. Promisingly, these remains were associated with numerous sherds of Neolithic pottery, hinting that this site, too, could have early origins – and after expanding the trench in 2023 the team confirmed that the timbers expanded far beyond the outline of the stone island.
With radiocarbon analysis now in, it appears that the crannog had begun life c.3500-3350 BC, in the form of a circular wooden platform measuring c.23m (75ft) across. This was a substantial construction, created from large numbers of felled willow and birch logs that had been laid directly on to the silts of the loch bed, forming a layer 2-3 logs thick. Timber uprights identified along the edge of this arrangement are thought to represent possible revetments intended to keep everything in place. Quantities of brushwood had then been heaped on top of the logs. This was mainly hazel, and the presence of lots of hazelnuts indicates that it had been laid down in the autumn. In the centre of the platform, the team found traces of a wattle structure, possibly some kind of architectural feature, whose supporting stakes had been driven through the earlier surface. Around 2,000 years, later during the Middle Bronze Age, a second layer of brushwood and stone was added to the island, possibly a response to rising water levels.
There was also a substantial but now-submerged stone causeway connecting the crannog to the loch shore. Its date is unknown, but it seems that some stones were in place in the Neolithic, perhaps around the edge of the island along with the timber revetments, and the stone was probably consolidated in the Bronze and Iron Ages.
Testifying to the long lifespan of this site, hundreds of pieces of pottery have been found during the project, scattered in the waters surrounding the crannog remains and representing diverse forms. The 2023 fieldwork alone produced 413 pot sherds representing at least 163 vessels. Five have been identified as Bronze Age or Iron Age jars, but the rest are typical of early Neolithic Hebridean types and include various jars, bowls, and cups. They add to 372 sherds that were recovered during 2021’s investigations, and 99 more from work on the site in preceding years. Traces of food residue are preserved on many of the fragments, while some pieces show evidence of sooting and discolouration speaking of exposure to heat after they were fired: clues that these were cooking vessels.
Dr Stephanie Blankshein, University of Southampton archaeologist and lead author of the journal paper, commented: ‘While we still don’t know exactly why these islands were built, the resources and labour required to construct them suggests, not only complex communities capable of such feats, but also the great significance of these sites. Large quantities of pottery, often still containing food residue, and worked stone found on and around the islands, suggest their use for communal activities such as cooking or feasting.’
Picture perfect
Excavating a site that extends both above and below water is one thing, but documenting its features represents a challenge in its own right – so, during the 2021 phase of fieldwork, the researchers devised and applied a new technique for using stereophotogrammetry specifically in shallow water, which forms the main focus of the new journal article.
Photogrammetry involves taking multiple 2D photographs of your subject from different angles, and then using specialist software to stitch them together to create a high-resolution, 3D digital model. It is very effective in deep water, but similar effects are notoriously difficult to achieve in the shallows, as imaging in less than 1m of water can be affected by a range of factors, including fine drifting sediments, choppy conditions, floating vegetation, and reflected or distorted light. As the team note in their paper, ‘terrestrial survey methods reach their limit at the water’s edge, whereas marine geophysical equipment often struggles in depths of less than 1m, [creating] a “white ribbon” of missing data’.
To help counteract these issues, the team designed a new approach using two small, waterproof cameras with low-light performance and a wide field of view. Locked at a set distance apart on a frame, this ‘stereo’ method provided precise overlapping of imagery to help compensate for any missing or disrupted data. The cameras were manoeuvred through the water by a diver, with positioning matched by an aerial drone. This technique allowed the team to record the crannog above and below the waterline as a single, continuous structure – providing a perspective that would not have been possible using land-based or underwater survey methods alone. The team describe the process as a portable, accessible, and cost-effective solution to a well-known frustration for archaeologists, and they hope that it will help to shed light on other, similar sites in the future. Demonstrating what can be achieved through this kind of work, they have used their results to create an interactive, annotated model of the Loch Bhorgastail crannog, which can be seen and explored at: https://skfb.ly/oQMXU.
S Blankshein, F Pedrotti, F Sturt, D Garrow (2026) ‘At the water’s edge: photogrammetry in extreme shallow-water environments’, Advances in Archaeological Practice (https://doi.org/10.1017/aap.2025.10145).
To read more about the Islands of Stone project, see http://www.crannogs.soton.ac.uk.
All images: University of Southampton
You must be logged in to post a comment.