Searching for a lost people in northern Scotland
The Picts are a fascinating but archaeologically elusive people who thrived in parts of Scotland in the 4th to 10th centuries AD. What has recent research added to this often obscure picture? Gordon Noble reports.
The Picts are a ‘lost people of Europe’ who continue to be a subject of enduring public fascination. First mentioned in late Roman sources as a collective name for troublesome, barbaric peoples living north of the Roman frontier, the Picts went on to dominate a large part of Scotland until the late 1st millennium AD. The emergence of the Pictish over-kingdom, the precursor of the kingdom of the Scots, was part of broader changes in northern Europe that laid the foundations for the modern states of Europe. Other than their enigmatic symbol stones, though, the archaeological and historical record for this region in c.AD 300-900 is diffuse and difficult – famously dubbed the ‘Problem of the Picts’.
The main Pictish powerbases were long-assumed to lie in central Scotland, but, in a seminal work of 2006, historian Alex Woolf located Fortriu – the most-cited and most-powerful Pictish kingdom – further north in the Moray Firth region. Further research has shed more light on this: in 2012, the Northern Picts Project was established at the University of Aberdeen to investigate an area stretching from Aberdeenshire to Easter Ross, covering the probable extent of Fortriu and a territory of Pictland known as Ce. Funded by a donation to the University of Aberdeen Development Trust, we have taken up the challenge of finding new archaeological features in a period with few identified sites, either in the written sources or the material record.
This unprecedented focus on the Picts was enhanced in 2017 by the Comparative Kingship project (funded by the Leverhulme Trust), and to-date the University of Aberdeen has investigated a whole series of Pictish sites in northern Scotland through large-scale excavation, survey, and targeted fieldwork. There have been some spectacular successes, not least the (re)discovery of a Pictish-period silver hoard at Gaulcross, Aberdeenshire, led by Aberdeen and the National Museum Scotland. In this article we will focus on two key elements: Pictish symbol stones and power centres.
Symbol stones are perhaps the most-celebrated element of Pictish archaeology. There are more than 200 stone monuments with symbols known from eastern and northern Scotland, and repeated attempts to decipher their meaning have been made since the 19th century. Current consensus is that this was a system that expressed names or identities of some kind, and that it was an elite form of expression found in both settlement and burial situations; providing better contexts and dating for this tradition has been a key aspect of our work.
From 2015 to 2017, the Northern Picts Project’s fieldwork targeted Dunnicaer, a towering sea stack just to the south of Aberdeen, where a series of Pictish stones were found in the 19th century. It has been suggested that their relatively simple designs (also seen in other contexts, including caves) might represent the earliest examples of the symbol system, but there has been little in the way of absolute dating.
The first stones were discovered during the gathering of building material at the site, and more examples were identified in 1832 when a group of youths found a low stone wall on the stack and threw a number of its stones into the sea. Since then, few people have visited Dunnicaer, as the site is cut off at high tide and surrounded by sheer cliff-faces – but, with the support of a professional climber, the Northern Picts team carried out three seasons
of (rather intrepid) fieldwork on the stack. This work revealed the remains of a promontory fort, with a timber-laced rampart enclosing a series of buildings (see CA 304 and 307). Much of the settlement had been lost to severe coastal erosion, but it still yielded an exciting range of finds, including Roman pottery and glass – rare imports this far north of the frontier – along with burnishing stones for metalworking.
Even more surprisingly, radio- carbon dating of samples from the fort suggests that its use began c.AD 105-225 and ended c.AD 350-450. Fort-building is rarely attested in the Roman Iron Age in Scotland, but Dunnicaer clearly flourished at this time, reaching its height in the same period as the first Roman reference tothe Picts (AD 297). While it remains impossible to directly date the symbol stones, the youths of 1832 described finding them in a wall surrounding the site, and the rampart around the southern edge of the stack which best fits that description was constructed c.AD 245-380. If the symbol stones are from this timeframe, they are much earlier than many scholars had countenanced for this tradition.
EXCAVATIONS AT ‘ROYAL’ RHYNIE
Another key focus of our project has been the Aberdeenshire village of Rhynie. Its name includes a form of the Celtic word for ‘king’, *rīg, and our work at the site suggests the surrounding valley was an elite Pictish centre from the 4th to 6th centuries AD (see CA 289). Rhynie has long been known for its notable concentration of Class I Pictish stones, and in March 1978 a particularly spectacular example was ploughed up by a local farmer at Barflat farm, just to the south of the modern village. Known as ‘Rhynie Man’, it depicts a bearded figure – possibly a pagan deity – carrying a distinctive axe that may be associated with animal sacrifice.
The field where Rhynie Man was found is home to another Pictish stone, the Craw Stane, which still stands in situ. In 1978, council archaeologist Ian Shepherd captured aerial photographs showing a series of enclosures surrounding the monument, and more than three decades later our project returned to the site to explore these features. Between 2011 and 2017, excavations by the universities of Aberdeen and Chester established that the Craw Stane stood towards the entranceway of the enclosure complex which, in an early phase, comprised ditches (and presumably banks) surrounding a low glacial knoll. A later phase saw the construction of an elaborate timber wall of oak posts and planks, inside which we found the footprints of a series of buildings and a rich array of finds hinting at a community with far-reaching connections.
As well as sherds of Late Roman wine amphorae imported from the eastern Mediterranean, there were fragments of glass drinking beakers from France, and one of the largest assemblages of metalworking production evidence known from early medieval Britain – from moulds and crucibles for making pins, to brooches and even tiny animal figurines that resemble the animals carved on Pictish stones. One of the most-remarkable finds was an iron pin shaped like the axe carried by Rhynie Man – tangible links between objects from the site and the iconography of the stones.
A few hundred metres to the north, where another of Rhynie’s carved stones (depicting a warrior) is recorded to have been recovered from a cairn, we have also found traces of a contemporary barrow cemetery. One of these mounds contained the partially preserved remains of a woman, and it is thought that two square enclosures located nearby may have been shrines or places for conducting ceremonies associated with veneration of the dead.
And what of the bigger picture? For the last three years, the University of Aberdeen has been exploring the wider environs of the Rhynie valley (funded by Historic Environment Scotland), with a particular focus on three hillforts overlooking the Barflat farm enclosure complex and cemetery: Cnoc Cailliche, Cairnmore, and Tap O’Noth. At Cnoc Cailliche, investigations revealed that this small (0.11ha) fort was constructed and inhabited c.400-200 BC, but occupation at the other two sites coincides with the Barflat complex’s use. Take, for example, Cairnmore: it is another small site (the innermost of its two stone enclosing walls surrounds c.0.2ha) where, ten years ago, an evaluation by Murray Cook suggested it had been occupied c.AD 410-630. Our excavations have now confirmed that dating – but we also revealed evidence for internal buildings, a large palisade at the edge of the inner bank, and occupation spanning the 4th to 7th centuries AD, directly overlapping with life at the Barflat site.
In contrast to its compact neighbours, Tap O’Noth is one of the most-spectacular forts in Scotland. The oblong hillfort that crowns its summit is the second highest in Scotland, and one of the best examples of a vitrified (heavily burnt) site of this kind. Moreover, the massive 16.75ha enclosure that surrounds it, scattered with hundreds of hut platforms, makes the site the second largest hillfort in all of northern Britain. This was a fascinating place to work, although excavation of the oblong fort was an exercise in extreme archaeology, with the vitrified walls and areas of the interior tackled over two gruelling seasons. Our efforts were amply rewarded, however, revealing the buckled and heavily burnt wall-faces of the vitrified fort and a well, together with dating evidence placing the site’s entire lifespan in 400-100 BC. Comprehensive radiocarbon analysis gave no hint of later reuse of the site – something that made the results from the larger fort area all the more surprising and exciting.
Due to its size and elevation, scholars have suggested that the fort was built and occupied when the climate was warmer, possibly during the Bronze Age – but last year’s excavations turned that notion on its head, with radiocarbon dates from two platforms and the rampart spanning the 3rd to 6th century AD. The rampart belongs to the later part of that range, making it the largest early medieval hillfort we know from Britain – Tap O’Noth has the potential to shake the narrative of this whole time period.
LiDAR and photogrammetry surveys suggest that many more house platforms are contained within the lower fort than previously thought – perhaps as many as 800 – making this potentially one of the most-densely occupied hillforts known in Britain. It is suggestive of an urban-scale population, and in a Pictish context we have nothing to compare this to. More hut platforms need to be tested to assess if they are all of similar dates, but it is possible Tap O’Noth enclosed a huge settlement contemporary with the Barflat complex.
WHAT’S IN A NAME?
The Tap O’Noth findings give us an unexpected and unparalleled insight into an elite Pictish landscape of the 4th to 7th century AD. After the 3rd century, settlement is exceptionally difficult to trace – something that makes the Pictish period notoriously difficult to contextualise. Compared to the hundreds if not thousands of known Iron Age roundhouses, we have no more than a handful of Pictish settlements known from the lowlands, which is why the evidence from the sites we have described above is so important.
One of the few site-types that may preserve traces of settlement are hillforts and promontory forts, but there is no clear morphological signature for a Pictish enclosure. The Rhynie valley, for example, shows the diversity of enclosed sites constructed in this period, and known and dated sites are frustratingly limited. As a result, our work has had many misses as well as hits; dozens of enclosed sites have been sampled, with the majority proving to be Iron Age rather than early medieval.
Very occasionally we have place-name or historical evidence to help target our work – such as at Bennachie, the site of a hillfort known as the Mither Tap, which we investigated last summer. Bennachie has been translated as ‘Mountain of Ce’, and Ce is mentioned in an ancient legendary section of the Pictish king-lists. Could it have been the pre-eminent site in this region? Bennachie is also possibly referred to in the two lost Gaelic sagas: Orgain Benne Ce (‘The Ravaging of Bennachie’) hints at a catastrophic battle or event at the site, while Orgain Maige Ce la Galo mac Febail (‘The Ravaging of the Plain of Ce by Galo son of Febal’) suggests further conflict in the area.
The Mither Tap hillfort consists of two large, but now collapsed, stone walls forming an upper and lower citadel surrounding a distinct granite tor that is highly visible in the surrounding landscape. The site was investigated in the 1870s by Christian Maclagan, one of Scotland’s earliest female archaeologists, and in 1881 she published a detailed plan of the fort showing the upper and lower ramparts, traces of possible roundhouses in both areas, and a well within the lower citadel.
More recently, path improvement works by Forestry Commission Scotland sparked small-scale excavations that confirmed activity at the site in the 1st millennium AD, but no extensive modern investigations had been carried out – until last June, when the Northern Picts team undertook another extreme archaeology season, trekking up the hill to evaluate the site more comprehensively. Excavation of the well, forgotten since the 19th century, exposed steps leading down to a small walled chamber and, after removing 19th century backfill, the well started functioning again, collecting water runoff from the hill in a nearmiraculous return to life. The lower citadel also yielded extensive midden deposits full of cattle, pig, and even fish bone (supplying the fort with the latter foodstuff would have been no mean feat), as well as traces of large platforms built up to create level bases for buildings.
The upper citadel held more evidence for early medieval occupation, and finds from across the site hinted at high-status metalworking. We also identified locally made pottery – a rare find from Pictish sites. Radiocarbon dating shows that the Mither Tap was in use in the 7th and 8th centuries AD; it may have taken over as one of the regional centres of this part of Pictland after the demise of Rhynie. Interestingly, towards the base of Bennachie on the north-east side of the hill lies Aberdeenshire’s most impressive Pictish cross-slab, the Maiden Stone, which is carved with an elaborate interlaced cross and a series of Pictish symbols in relief. The new dates from the Mither Tap help contextualise the landscape context of this major monument.
Since 2018, one of our main fieldwork projects has been on the remarkable promontory fort at Burghead. Although the southern portion of the site was destroyed during construction of the modern village in the 19th century, it would have originally covered c.5.5ha, and some of the site’s best-known finds include nearly 30 stone slabs carved with images of bulls, and an impressive well.
The complexity of the fort’s defences, with timber-laced ramparts over 8m wide and 6m high, was revealed during excavations in the 1860s and 1890s, but there had been relatively little in the way of modern investigations other than the work of Alan Small in the 1960s, who thought that much of the interior had been destroyed. In fact, the interior remains largely intact, comprising an upper raised citadel and a lower citadel, both of which are surrounded on their seaward side by a grass-covered rampart, and in 2015-2017, small-scale sampling by the Northern Picts Project revealed floor layers of partially intact early medieval buildings surviving within the fort.
These investigations were scaled up in 2018-2019 (as part of the Leverhulme Trust-funded Comparative Kingship project), and trenches in both the upper and lower citadels revealed further early medieval buildings with intact floor layers preserved under up to 1m of 19th-century overburden. Investigations were also carried out at the western (seaward) end of the site, which was threatened by coastal erosion: this excavation, funded by Historic Environment Scotland, showed that exceptionally well-preserved stretches of early medieval rampart survive in this area to around 3m in height – these remains show clear evidence for their destruction by fire. Timber-laced ramparts of this scale and complexity rarely survive, and the Burghead example ranks among the best in Europe.
The finds from our excavations on this site have also been exciting: dress accessories, pieces of weaponry including a sword hilt, iron tools, bone pins, and metalworking evidence. One of the most-striking discoveries, though, was a pair of Anglo-Saxon coins of Alfred the Great (r. 871-899), which were recovered from the floor layer and midden of a building. They highlight the long-distance networks that this site was able to tap into and, intriguingly, these coins had been pierced, perhaps to be worn on a necklace or bracelet.
To-date, we have obtained over 40 radiocarbon dates from Burghead, showing that it was occupied from at least the 6th century AD and was destroyed in the 10th century – a fairly obscure period when the Pictish realm had become the expansionist Gaelic kingdom of Alba. Local tradition recounts that the site was destroyed by the Vikings – can continuing work at Burghead shed any light on the fate of this major centre, and of the northern Picts in this new era?
Our research has largely involved hitherto-unrecognised or largely unexplored Pictish sites, shedding unprecedented light on a period that has been relatively neglected in archaeological studies. As well as new excavations, the University of Aberdeen has been dating samples taken during older excavations to provide new and more-robust chronologies. This archival work is as important as the new investigations, and each redating project has led to entirely new chronological frameworks being developed for major but poorly dated sites. The results outlined in this article focus on hillforts and promontory forts as that is where settlement traces appear to be best preserved, but we have also been investigating unenclosed settlements where they exist, such as the wag (a type of longhouse) settlements in Caithness.
We have been investigating ecclesiastical sites too, another poorly understood part of Pictish archaeology (although one greatly improved by Martin Carver’s excavations at Portmahomack, Easter Ross – see CA 205 and 321). Recent geophysical survey and small-scale excavation at Kinneddar, Moray, for example, has identified a major ecclesiastical site around 5km from Burghead, revealing a vallum surrounding a church site that is the findspot of one of the largest collections of Pictish early Christian sculpture and a Class I symbol stone. The vallum encloses an area of 8.6ha, and the ground-plan presents striking resemblances to other important ecclesiastical centres, particularly the famous site of Iona (CA 292).
Radiocarbon dating suggests the ecclesiastical centre at Kinneddar was in use as early as the late 6th century, with the vallum likely to date to the 7th or 8th century. We have identified a number of large vallums at other church sites in northern Scotland, at such places as Migvie, Aberdeenshire; Glamis, Perth and Kinross; and possibly Dunkeld, Perth and Kinross. We also have a range of completed or in-progress PhDs at Aberdeen examining the cemetery traditions of the Picts, Pictish period settlement, and a range of other elements of Pictish-period archaeology.
A NEW VIEW OF THE PICTS
Over the last eight years, excavation and field survey by the University of Aberdeen has led to major new insights into the Picts, discovering new power centres and hitherto-unrecognised site-types, as well as providing crucial new contextual and dating evidence for the Pictish symbol tradition. This research has revealed exceptionally rare settlement evidence too, and provided new and more-robust chronologies for Pictish-period archaeology.
As for sharing these finds with the public, the Gaulcross silver hoard featured in an exhibition, ‘Scotland’s Early Silver’, by National Museums Scotland (CA 335), and we have worked with smaller regional museums, such as the Tarbat Discovery Centre, running exhibitions of our work. The project has produced popular books on the Picts while, at Rhynie, our excavations led to the formation of a local artistic collective called ‘Rhynie Woman’ who have organised a series of public-engagement activities that ran alongside the archaeological work, such as pop-up cafes and exhibitions, a fire festival, and local school initiatives. We have been featured in the national media, such as on BBC Radio 4’s In Our Time and BBC2’s Digging for Britain, and have a prominent social-media presence. Above all, our aim is to promote new ways of engaging with the Pictish past, and to shed light on a period that for too long has been a particularly poorly illuminated part of the so-called ‘Dark Ages’.