126.96.36.199 Prehistoric and Roman sites
The place-names in our historical sources invariably record the elite settlements of their times: the power centres and major churches. Yet the archaeological traces of these early settlements are still very limited. The reuse or continuation of occupation of Iron Age hillforts into the early medieval period is most clearly seen at Traprain Law, the power base of the Votadini in the Roman Iron Age, and new work on the silver hoard deposited there has pushed the dating into the mid-5th century (Hunter et al 2022). The discovery of a child’s long cist grave on the summit also suggests an early medieval burial ground and potentially a centre of occupation at the site (Hunter et al 2019). As part of the Leverhulme-funded Comparative Kingship project carried out by the University of Aberdeen, excavations undertaken in 2022 on Eildon Hill North showed the hillfort was primarily occupied in the Late Bronze Age but there appears to be extensive reuse of the fort in the Roman Iron Age and early medieval periods (Noble in prep).
Roman forts rarely remained occupied after their military functions ceased. Despite extensive excavations, there are only few traces of early medieval activity within Roman forts in the area (Masser 2006; Maldonado 2015, 227–8; Blackwell 2018, 107–8). However, a notable exception now comes from Cramond, where a mass grave from the infill of the derelict Roman bathhouse latrine was confirmed to date to the 6 to 7th centuries (see SESARF Chapter 8.6). The bathhouse was outside the fort itself, some 200m north of the modern church. This unique and violent deposit tells us, amongst other things, that the ruins of old Roman forts and outlying structures were still clearly visible in the early medieval period. Two other stray finds from the churchyard at Cramond are also suggestive of early medieval activity: an enamelled mount of 8th to 9th century type (NMS X.FC 302), and a rare finger ring inscribed with Anglo-Saxon runes possibly dating to the 9th or 10th century (NMS X.NJ 19), both suggest later occupation, if not necessarily ecclesiastical in nature. In the Borders, the discovery of a small group of long cist burials on Abbey Knowe, a natural hillock overlooking the Roman fort at Lyne, suggest there was an even earlier settlement nearby than that associated with a chapel attested at Lyne since at least the 12th century.
The Atlas of Hillforts in Britain and Ireland database lists only six sites with confirmed evidence for early medieval occupation in the SESARF area: Edinburgh Castle (Driscoll and Yeoman 1997); Auldhame (reused as a monastery and long cist cemetery: Crone and Hindmarch 2016); Castle Park, Dunbar (Perry 2000); St Abb’s Head, Coldingham (Alcock et al 1986); Rubers Law (which incorporates reused Roman masonry, but is otherwise undated); and Moat Knowe, Buchtrig (proposed as a nuclear fort but undated). Dalmahoy, Edinburgh is also suggested to be of ‘nuclear’ early medieval form, but the Atlas rules it out on the basis of lack of evidence (see below). Early medieval occupation is now confirmed through radiocarbon dating at Tinnis Fort near Biggar, and possible evidence has been found at both Kidlaw Fort (MEL785) and Craig’s Quarry (MEL1343). The fort at Craig’s Quarry (MEL1343) had several long cists excavated in the 1950s. Although it is just outside the study area, the Great Enclosure of Yeavering, Northumberland is a related site with confirmed early medieval royal settlement (see below).
Hillforts in neighbouring regions, including Dumbarton Rock on the Clyde and Bamburgh in Northumberland, were early capitals of their respective kingdoms, and it is often assumed that Edinburgh was the equivalent for the Lothians. The earliest record of it is from a siege in 638, which is usually seen as the formal Northumbrian takeover of the area from the Britons. Severely truncated early medieval layers were detected in excavations at Edinburgh Castle, but not enough survived to make out any of the early defences (Driscoll and Yeoman 1997). Early medieval finds were also few, but enough to establish early medieval occupation, including a 7th to 8th century decorated bone comb and a spearhead.
In Edinburgh as elsewhere, it is easier to discern occupation from the 7th century onwards, when there are diagnostic finds such as styca coins and bone combs. The Northumbrian settlement at Dunbar, described as an urbs in the 7th century, is presumed to have been occupied by the Britons prior to the Northumbrian invasions. However, the sequence of settlement is more secure from the 7th century with finds such as a continental sceat coin and a gold-and-garnet pectoral cross (NMS X.1997.529). The precise nature of the settlement, with a possible minster church, is still debatable, but was certainly of considerable status throughout this period and beyond (Perry 2000; Alcock 2003, 212–7).
Other tantalising finds hint at early medieval phases at unexcavated hillforts in the region. A gold stud with Anglo-Saxon parallels (NMS X.FE 86) and ceramic moulds for metalworking from Dalmahoy (Stevenson 1948–49; Blackwell 2018, 206–7), and an ‘annular twist’ glass bead from Denholm Hill, Borders (NMS X.FJ 120), provide evidence for Northumbrian-period occupation. They may not only have been used as forts. Defended enclosures were reused as cemeteries in the early medieval period, as seen at Castle Park, Dunbar and Auldhame. Castle Dykes may be another example, since a “great number of human skeletons” in cists were found during clearance of the promontory fort at in 1831.
188.8.131.52 Hall complexes
Rather more well-attested in the study area is the use of lowland enclosures as power centres. The villa regis or royal centre of the Northumbrian King Eadwine (r. 616–633) was named by Bede as Ad gefrin (Yeavering). While it stands just outside the SESARF study area, it provides an important model for unenclosed high-status settlement in the area. Excavations by Brian Hope-Taylor (1977) revealed a sequence of timber structures and burials, in a landscape dominated by prehistoric monuments which were reoccupied from the 6th century. Dating was difficult owing to the lack of diagnostic finds. At its height, the settlement included a series of timber halls, a unique timber ‘grandstand’, a palisaded ‘Great Enclosure’ and a possible beam or stapol (a pre-Christian carved timber post) in a square enclosure which formed the focus for cultic activity. In a later phase, a possible minster church with an enclosed cemetery was added to the site. It seems likely that the settlement predated and post-dated the reign of Eadwine.
There are at least two and as many as five more postulated royal hall complexes similar to that at Yeavering in the SESARF study area. Three are known only as cropmarks and remain unexcavated, so plausibly contain numerous superimposed phases of settlement. Two are in the Scottish Borders region, at Sprouston on the Tweed near Kelso, and Philiphaugh on the Ettrick Water near Selkirk (Smith 1992). Sprouston has a series of timber hall-like structures, rectangular post-built structures, sunken-featured buildings and a cemetery of at least 380 graves. Another complex of cropmarks at Philiphaugh revealed several more rectangular structures and square enclosures, as well as another very large, enclosed cemetery.
Both seem to be parallel counterparts to the hall complex at Yeavering, and presumably fall within the 6th to 7th centuries, with possible extended use represented by the burial grounds. All three complexes include large, palisaded enclosures which were presumably used for large gatherings and/or livestock enclosures and have been interpreted as places where food renders were collected and redistributed alongside other assembly-related functions. Stray finds have not yet shed much light on either site; a silver ingot from near Sprouston might indicate activity continuing into the Viking Age (NMS X.2005.6). A blue glass lobed bead of 6th to 7th century type from near Philiphaugh supports the presumed date of the settlement, but otherwise has no known context (Blackwell 2018, 99).
Two further examples of possible hall-complexes are in East Lothian, at Whitekirk and Aberlady. At Whitekirk, cropmarks revealed two rectangular, hall-like structures with annexes, amidst other rectilinear enclosures of unknown date. Even if all of these are early medieval structures, the Whitekirk site is difficult to compare with the sprawling hall-complexes at Yeavering, Sprouston and Philiphaugh, and it may instead be that this is a farmstead or other ancillary settlement serving the monastic centre at nearby Tyninghame (Lowe 1999). At Aberlady, geophysical survey in the area of Kilspindie Castle (now the Glebe Field) had indicated the presence of two superimposed halls. Trial excavations in 2016 (Tulloch and Davies 1999; Murray 2016) revealed these to be stone-built structures, including a large, paved area 40 x 20m across which was dated to after the 7th–9th centuries. It was associated with bone and metal finds characteristic of the Northumbrian period, including a 9th-century styca coin and decorated bone combs (see SESARF 8.3.3 Coinage and 8.3.5 Pre-Burghal Markets). These two examples show how little we know about the nature of early medieval settlement without large-scale excavations with rigorous dating programmes in place.
Finally, a 5th possible hall-complex may have been glimpsed in excavations at Castle Park, Dunbar (MEL1913). The settlement of Dunbar was referred to in Northumbrian sources as an urbs regis, a related term referring to royal centres, including hillforts. A mortar-mixer dated to the late 8th to 9th centuries indicates the presence of a mortared stone building of some significance, even though it was difficult to link conclusively to any particular structure. The only stone-built hall (or less likely, a minster church) was difficult to date closely as only a short section of wall was glimpsed in excavations, but it seems to have been abandoned in the 9th century, based on a single coin of Eanred (837–41) from a later posthole (Perry 2000, 73). The excavations here captured only part of a much wider important royal settlement, but whether it could be defined as a hall-complex remains up for debate.
A cautionary tale in this regard is the presumed hall of Doon Hill, East Lothian. A putative early medieval palisaded enclosure and hall complex here has long been part of the discussion, situated at the heart of the distribution of Anglo-Saxon material culture and place-names. Excavations by Hope-Taylor built up an elaborate narrative of elite takeover, in which an early British timber hall on the site of a Neolithic mortuary complex was destroyed by fire and replaced by an ‘Anglian’ hall with its own palisaded enclosure, providing a historical model for the conquest of the Britons more generally. These events were dated to the mid-7th century and presumed to be the residence of a thane or other nobleman tributary to a king (Hope-Taylor 1977; Smith 1990; Alcock 2003).
All of this was based on comparisons with Yeavering, despite the lack of any diagnostic early medieval finds (only Neolithic ceramic was encountered) and the absence of any absolute dating. Sadly for fans of a tidy narrative, a recent reassessment backed with C14 dates and a critical review of the excavation archive has overturned the interpretation of this site; the hall and its associated structures are resolutely prehistoric (Ralston 2019). This makes the exploration of cropmark hall complexes like Sprouston and Philiphaugh that much more desirable, and new dating evidence from ongoing excavations at Yeavering are eagerly awaited.