Hillforts of East Lothian: the evaluation of a prehistoric banked enclosure of White Castle

by David Connolly, Murray Cook and Hana Kdolska on behalf of Rampart Scotland

White Castle enclosure (SM756), situated near Garvald, East Lothian (Figure 1) was investigated by Rampart Scotland, under the auspices of Hillforts of East Lothian Project (hereafter HELP; Connolly, Cook and Kdolska 2021). The four seasons (2010–2013) of survey and evaluations led to a recovery of key constructional and chronological information as well as a crucial radiometric assemblage. Four broad phases of activity were proposed from the combined evidence of C14 dates, stratigraphy and partially artefactual evidence: comprising ephemeral presence in Late Neolithic/Early Bronze Age (EBA), two main phases of enclosure construction and/or reconfiguration (c 750 and 400 BC and c 366 and 100 BC), a Medieval (AD 990–1034) and post-Medieval/Early Modern (AD 1490–1632) remains.

White Catle hillfort, East Lothian © HES

Despite a significant body of research focused on Prehistoric forts, referred to here by more neutral term ‘enclosures’,  in the area of East Lothian (review in Connolly, Cook and Kdolska 2021, 12–16), their understanding is incomplete. Indeed, the majority of these prolific prehistoric monuments are characterised and dated on morphological grounds only (eg Cowley 2009, 207–208; Hunter and Carruthers 2012, 77).

The HELP Project, run as a field-school, was set-up with the overall aim to recover viable C14 dates, by means of key-hole soundings, in order to furnish the development of regionally based chronology for the site type (Hunter and Carruthers 2012, 77). While such approach has obvious drawbacks (eg Clarke 2001), the method is acknowledged as an initial and cost-effective step towards this goal (eg Haselgrove et al 2001, 5; Hunter and Carruthers 2012, 76, 94) and has been successfully applied elsewhere in Scotland (Cook 2013). Enclosures located in the catchment area of Lammermuirs were targeted as none have been previously excavated; the study area encompassing 30 sites of varied character. Two enclosures were investigated: the upstanding site of White Castle (2010-2013; Connolly, Cook and Kdolska 2021) and a cropmark site of Sheriffside (2011–2017; see Sheriffside Case Study).

Location with area subject to toporaphic survey and trench location © Rampart Scotland

White Castle enclosure (Figure 2) occupies a natural spur with water courses to three sides and is approached from the south across a natural triangular causeway, this deliberately flattened along its southwest edge to facilitate access. The spur extends from a natural mound into the mouth of the valley, dramatically dropping down to the Thorter Burn, with steep slopes on two thirds of its flanks. The location commands impressive views to north across the Lothian Plain from the eastern edge of the Pentlands to Berwick Law and beyond to Fife. In contrast, the view to south is restricted, with the site only visible from some 500m on the final approach, this providing a sudden dramatic effect on anyone travelling from that direction.

Aerial view to the south of White Castle and the Lammermuirs © D. Connolly Skyscape Survey

The enclosure is of an oval shape and multivallate design, comprising three sets of ramparts/banks penetrated by three groups of aligned entrances. The largest, most impressive, sections to the south and southwest are associated with formal access, apparently designed to both impress and control movement/access. In contrast, the steep gradients to the east, north and northeast were terraced at a minimum effort.

The archaeological works at White Castle comprised three types of survey (topographic, geophysical and condition survey) and small-scale excavations (key-hole soundings). The enhanced topographic survey of the entire site resulted in a detailed Data Terrain Model (DTM) (Figure 3) and was complemented by ground resistivity survey undertaken by the Edinburgh Archaeological Field Society. A mapping of animal burrows and bracken cover, both causing visible damage to the site, was also undertaken each year.

White Castle main features including openings © Rampart Scotland

Mindful of White Castle’s scheduled status, the project employed key-hole excavations, to minimise unnecessary disturbance, as well as avoiding substantial costs. All trenches were hand excavated and all recording procedures followed CIfA standards (CIfA 2014). The excavations targeted key stratigraphic features and, where possible, taphonomically secure, unabraded roundwood charcoal samples were recovered for C14 analysis. A total of 42 soundings (both trenches and test pits; Figure 1, 3) were dug during the eight weeks of excavations, located in the interior, exterior and the enclosing works. These amounted to less than 0.5%  of the Scheduled Area.

The topographic survey identified 20 internal features comprising 19 subcircular platforms of varied sizes, some impacting each other, and a single square structure (STR 17) built over the inner rampart (RAM 1). Only STR 17 and two of the platforms were previously recorded (RCAHMS c 1913). Further newly identified features included a new entrance group (OP 1, 9, 15) on the northwest circuit and a new double terraced outwork (TER 4), situated between the inner and middle circuits and overlying infilled ditch (DIT 1). The resistivity survey added further 7 potential platforms to the suite of internal features.

The excavations characterised the three upstanding banks/ramparts (RAM1, 2 and 3), forming the three circuits, and confirmed the double terrace (TER 4) between RAM 1 and 2. The inner two ramparts were associated with ditches (DIT 1 and DIT 2) and another short stretch of ditch (DIT 3), extending beyond RAM 3 was also defined. The excavations also identified former palisade (PAL 1), from the line of post-holes, on the northeast of the Inner Circuit, interpreted as a continuation of the line of RAM 1.

The ramparts were all of slightly different character, which suggests phased construction, with the three circuits clearly not all contemporary, as also supported by C14 evidence (Figure 4). Of the fifteen apparent openings/gaps in the circuits, nine appear to be original to the enclosure,  forming four sets of entrances to: southwest (OP 2, 8 and 10), east (OP 4 and 5), northwest (OP 1, 15 and 9) and south (OP 11). It is uncertain if all four groups of the entrances were contemporary and no gate furniture, suggesting formal gateways, was detected. 

Four of the identified internal features were selected for excavation, including three platforms (PLT 4, 8, 20) and the square structure (STR 17). Together, PLT 4 and 8 included structural remains comprising cobbled/stony surface, stone retaining bank and possible beam slots or drainage gullies/channels, suggesting their former use as platforms for timber structures. PLT 20 included a single sub-oval pit, which contained a charcoal-rich fill, this post-dating the enclosure use. STR 17,  comprising a roughly square turf banked structure, located partially over the bank of RAM 1, was confirmed as post-medieval.

Finally, the combined survey and excavations (TR19) also revealed the remains of a track, constructed of deposits of heavily compacted orange clay and pebbles, and impacted by a series of linear slots, these  interpreted as wheel ruts. Although undated, the track likely formed a portion of the historically recorded north-south route through the Lammermuirs (eg Graham 1960, 217–235; Goodman 1981, 111), which pre-dates the current tarmac road.

The excavations recovered 15 radiometric dates obtained from key features within the enclosure’s stratigraphic sequence, revealing four broad phases of activity (Figure 4). These, together with artefactual and stratigraphic evidence, suggest following:  a limited pre-enclosure activity on or in the vicinity of site in Late Neolithic/EBA; charcoal producing activity of unknown character before c 740 BC; two broad phases of enclosure building and/or redesign in Iron Age (between c 750 and 400 BC and c 366  and 100 BC); slight Medieval presence (AD 990–1034; pit PLT 20) and post-Medieval/Early Modern structure (AD 1490-1632; STR 17).

While small in scale, the White Castle excavations resulted in a number of key insights. It is obvious that White Castle underwent significant morphological changes during its lifespan, all contributing to  its present appearance as a multivallate structure. This agrees with the evidence from other excavated multivallate sites, such as Broxmouth (Armit and McKenzie 2013), which suggest that multivallation is often a result of complex chronologies rather than deliberate design (Hingley 1992, 30). Although other sub-phases remain likely undetected, on evidence, the White Castle sequence commenced with the initial construction of a univallate enclosure (RAM 1 & DIT 1), before the addition of further rampart and a ditch (RAM 2 and DIT 2), changed its character into a bivallate structure. Although undated, the subsequent addition of further bank and ditch (RAM 3, DIT 3) on the  exterior of RAM 2 may preceded or accompanied levelling of RAM 1 and the infilling of DIT 1; the latter potentially intended to create more interior space. The eventual disuse of site remains undated but was potentially in the closing centuries BC.

The deliberate enlargement and elaboration of portions of enclosing banks and the associated main southwest entrances appear to be a deliberate choice, intended to both impress and give the impression of defensibility (Haselgrove and Lowther 2000, 175) to anyone approaching the site along the route through the Whiteadder Valley. These selective aggrandisements are suggestive of the importance of physical appearances rather than functionality per se, as suggested on other prehistoric enclosures in the area (Armit and McKenzie 2013, 88, 91) and beyond (Collis 1996, 89-90; Ralston 2006, 75, 128-29).  

The enclosure was clearly associated with internal settlement, as evidenced from the multiple sub-circular platforms. However, it was not possible to determine if White Castle was occupied permanently or seasonally, or to estimate its population. The density of identified internal features (27 in total), majority of which are assumed to have been supporting timber houses and/or subsidiary structures, given the varied sizes, is certainly suggestive of intensely used site, and invites comparisons with other enclosures, such as Eildon Hill North (Owen 1992, 67-70). However, without further excavations, it is not clear if all the extant structures are contemporary.

As to the economy/function of the White Castle, its very location in Lammermuirs with documented historic link to transhumance of sheep (Robson 1977, 30-32, 449) is certainly suggestive. The possibility of seasonal summer grazing use of the upland prehistoric enclosures in Lammermuirs has been previously considered (Macinnes, 1982, 59; 1984, 193; Armit and McKenzie 2013, 506). Feasibly, White Castle community controlled access to these upland pastures, as well as guarding, both physically and visually, the movement along the well-established route through the Whiteadder Valley (Reader 2012, 346), which presently, and historically (Graham 1960, 217–235; Goodman 1981, 111) formed the most rational route from the south.  

Further Information

Connolly, D,  Cook, M and Kdolska, H 2021. White Castle: The Evaluation of an upstanding prehistoric enclosure in East Lothian. Oxford: Archaeopress Publishing Ltd. Available at: https://www.archaeopress.com/Archaeopress/Products/9781789699302

White Castle Canmore Record: https://canmore.org.uk/site/57479/white-castle

Atlas of Hillforts of Britain and Ireland: White Castle Record SC3902: https://hillforts.arch.ox.ac.uk/records/SC3902.html


Armit, I and McKenzie, J (eds) 2013 An Inherited Place: Broxmouth Hillfort and the South-East Scottish Iron Age. Edinburgh: Society of Antiquaries of Scotland.

CIfA 2014. Standard and guidance for archaeological excavation. Published December 2014, updated 1 October 2020. Available at: https://www.archaeologists.net/sites/default/files/CIfAS%26GExcavation_2.pdf

Clarke, D 2001 ‘Defining and integrating sequences in site analysis: the evidence from hillforts and other sites’, Oxford Journal of Archaeology 20(3), 293–306.

Connolly, D,  Cook, M and Kdolska, H 2021 White Castle: The Evaluation of an upstanding prehistoric enclosure in East Lothian. Oxford: Archaeopress Publishing Ltd.

Collis, J. 1981. ‘A theoretical study of hill-forts’, in Guilbert, G. (ed) Hill-fort Studies. Essays for A.H.A. Hogg. Leicester: Leicester University Press, 66-76.

Cook, M 2013 ‘Open or enclosed: settlement patterns and hillfort construction in Strathdon, Aberdeenshire, 1800 BC to AD 1000’, Proceedings of the Prehistoric Society 79, 327–52.

Cowley, D.C. 2009. The Traprain environs in a regional perspective, in Haselgrove (ed) The Traprain Law Environs Project: Fieldwork and Excavations 2000-2004. Edinburgh, Edinburgh: Society of Antiquaries of Scotland, 205-223.

Graham, A 1960 ‘More Old Roads in the Lammermuirs’, Proc Soc Antiq Scot 93 (1959-60), 217–235.

Goodman, A 1981 The Wars of the Roses: Military Activity and English Society, 1452–1497, London: Routledge.

Harding, D W (ed) 1982 Later Prehistoric Settlement in South-East Scotland. Occasional Paper No. 8. Edinburgh: University of Edinburgh Press.

Harding, D 2001 ‘Later Prehistory in South-East Scotland: A Critical Review’, Oxford Journal of Archaeology 20(4), 355–376.

Haselgrove, C (ed) 2009 The Traprain Law Environs Project: Fieldwork and Excavations 2000-2004. Edinburgh. Edinburgh: Society of Antiquaries of Scotland.

Haselgrove, C and Lowther, P 2000 ‘Excavation of the Later Prehistoric Enclosure Complex at Fishers Road, East’, in Haselgrove C and McCullagh R P J (eds) An Iron Age Coastal Community in East Lothian: the Excavation of Two Later Prehistoric Enclosure Complexes at Fishers Road, Port Seton, 1994–5. Edinburgh: Scottish Trust for Archaeological Research, 85-176.

Haselgrove C, Armit I, Champion T, Creighton J, Gwilt A, Hill J D, Hunter F and Woodward A 2001 Understanding the British Iron Age – an agenda for action. Salisbury: Prehistoric Society/Trust for Wessex Archaeology.

Hingley, R 1992 ‘Society in Scotland from 700 BC to AD 200’, Proc Soc Antiq Scot 122: 7–53.

Hunter, F and Carruthers, M (eds) 2012 Iron Age Scotland: Scottish Archaeological Research Framework Panel Report. Edinburgh: Society of Antiquaries of Scotland. Available at:                               https://scarf.scot/national/iron-age-panel-report/

Macinnes, L 1982 ‘Pattern and purpose: The settlement evidence’, in D W Harding (ed) Later Prehistoric Settlement in South-East Scotland, Occasional Paper 8, Edinburgh: University of Edinburgh Press, 57–73.

Macinnes, L 1984 ‘Settlement and economy: East Lothian and the Tyne-Forth province’, In Miket, R

and Burgess, C B (eds) Between and Beyond the Walls. Edinburgh: John Donald, 176-198.

Owen, A O 1992 ‘Eildon Hill North, Roxburgh, Borders’, in Rideout, J S,  Owen, O A and Halpin, E (eds) Hillforts of Southern Scotland. Edinburgh: AOC (Scotland) Ltd, 21–71.

Ralston, I 2006 Celtic Fortifications. Gloucestershire: Tempus Publishing Limited.

Reader, R 2012 Over the Ditch and Far Away: Investigating Broxmouth and the landscape of Southeast Scotland during the later prehistoric period. PhD Thesis. University of Bradford, Division of Archaeological and Environmental Sciences, School of Life Sciences.

Robson, M J G 1977 History and traditions of sheep-farming in the Scottish border hills: A study of customary life and practices among the sheep farming community of the central hill areas before 1900. Unpublished PhD thesis. University of Edinburgh. Available at: http://hdl.handle.net/1842/17824