The potential of multidisciplinary approaches to the early medieval period has long been recognised but the challenge of integrating this with a coherent archaeological programme has not yet been realized. Much of this has taken place through doctoral theses. For instance, Christopher Aliaga-Kelly (1986) reconsidered the evidence for Anglo-Saxon occupation in south east Scotland, while Ian Smith (1990) undertook a major reassessment of the historical and archaeological sources for early medieval settlement in the Tweed basin but sadly died in 1994. The major outcome of Smith’s work was demonstrating the long-term continuity of settlement and landholding throughout the formation of British and Northumbrian kingdoms. This is seen most clearly in the retention of British-language place-names for Northumbrian royal and monastic centres, such as Dynbaer (Dunbar) and Mailros (Melrose). Smith and Aliaga-Kelly’s works were notable for their use of both historical and archaeological sources to recreate historic land use.
Smith observed that the line of the Roman road of Dere Street remained a persistent boundary in subsequent centuries. He found that the evidence for Anglo-Saxon place-names and archaeology largely remained to the east of it, a conclusion that has not much changed in the intervening decades (Blackwell 2018, 99). Using this insight, he further postulated the sub-Roman Tweed basin to have been divided into three putative kingdoms: Anglo-Saxon Bernaccia (Berwickshire) to the east of Dere Street, and to west, British Godeu (Tweeddale) and Calchvynyd (Kelso/Roxburghshire). However, this model has been less influential.
Smith and others have also explored the possibility that the numerous upstanding linear earthworks of south east Scotland could have been erected or reused as borders between the British and Northumbrian kingdoms (Barber 1999, 79). The largest of these, referred to as the Catrail, has been proposed as a continuous linear earthwork some 86 km in length, but is in fact a series of discontinuous earthworks (RCAHMS 1956, Vol.2, 479–80, fig. 613). Part of the supposed Catrail is in fact a wholly separate earthwork referred to as the Picts’ Work Ditch in Selkirkshire (RCAHMS 1957, 120, 127). Where dated, these and other earthworks appear to belong to the end of the first millennium BC (Smith 1990, 315–6; Barber 1999). That does not exclude their use as boundaries in later centuries, but the evidence for their active use as territorial boundaries in the early medieval period is currently still negative.
O’Brien (2002) used similar multidisciplinary methods to argue that early medieval north Northumberland had been divided into shires each with its own royal manor during the early medieval period. The historical basis for the ‘shiring’ of south east Scotland is well established, but the written evidence only dates back to the 11th century, and then only in exceptional circumstances (Barrow 1973). Smith’s (1990) detailed case study in the Manor Valley still serves as a useful model, as does an exercise in reconstructing ‘multiple estates’ in the Lothians combining map regression, place-names and the locations of churches and long cist cemeteries (Proudfoot and Aliaga-Kelly 1997). However, as these studies acknowledge, there are obvious gaps in the evidence: long cist burials date to a period well before the establishment of ‘shires’ (SESARF 8.4.3 Burial traditions), and their relevance to later territorial boundaries is still speculative.
Therefore any link between early cemeteries and territorial boundaries is bound to be complex. For the earliest part of the period, where there is a real dearth of settlement evidence, community formation can be glimpsed partly through burial evidence. Maldonado (2011) showed that at 5th to 7th-century sites like Thornybank (MEL8391) Lasswade and the Catstane, there was evidence that long cist cemeteries built up in clusters, almost like family plots, rather than in long, managed rows like modern cemeteries. It suggests that scattered rural communities gathered together only infrequently for rituals such as funerals. At Thornybank and Catstane, the cemeteries were aligned on prehistoric monuments which will have been prominent local landmarks ideally suited for such gatherings. That would make these places ‘central’ to a community, but ‘peripheral’ from their settlements.
Katherine Forsyth (2005) observed that cemeteries associated with Latin-inscribed stones were more convincingly associated with territorial boundaries. The Latin-inscribed Catstane stands at an important crossing place over the River Almond, as well as being at the edge of the later medieval parish of Kirkliston. The place-name Kirkliston seems to incorporate a British cognate of old Welsh lys meaning court, suggesting the cemetery might be marking the entrance to an important royal district. The Yarrow stone, also associated with early burials, may also mark the entrance to an estate centre placed along an important routeway between the Annan and Tweed valleys (Smith 1990, 288-90).
With regard to the possibility that these early cemeteries were assembly places, it is interesting to note the occurrence of rotary stone querns for grinding cereal into flour at several long cist cemeteries, not all of them in cists, while at Lasswade, an enigmatic burning layer was cut through by later burials (eg Cowie 1980; Henshall 1958; Perry 2000, 283). This suggests early cemeteries were used for multiple kinds of gatherings aside from just funerals, as has been argued elsewhere in Britain and Ireland (eg Shiner 2021).
Overall, it is perhaps more realistic to try and track small-scale estate organisation over time before attempting to extrapolate out to the level of entire kingdoms, with the excavations at Castle Park, Dunbar and Auldhame for East Lothian, and The Hirsel for the lower Tweed area, as good examples tracking the evidence from one site over a long period of time. At Dunbar, the royal estate showed lots of overall continuity throughout the period with regard to subsistence and consumption (SESARF 8.2.6 Food production), but also dramatic changes to the site’s layout on the ground, indicating that the placement of burials and halls shifted very rapidly (and occasionally disappeared) at different parts of the medieval period. Excavations therefore support this model of a tribute-based economy holding stable for centuries despite changes in political power (Perry 2000). Excavations revealing food production at the Grassmarket and Burdiehouse have been argued to be the kinds of satellite settlements which supplied renders to the nearby power centre, in this case presumably Edinburgh (MacIver and Paton 2023; McMeekin 2010, 111).
By the 11th century, it is clear that communities had begun to agglomerate into small settlements, though there is no evidence for the emergence of towns in this area until the 12th century. However, the proliferation of Romanesque church architecture (SESARF 8.4.6 Later sculpture and architecture) and the expansion of existing burial grounds (as at Dunbar in the 11th to 12th century) suggests some clustering of settlement around this time. The origins of towns and burghs is discussed further in chapter 8.3.5 Pre-burghal markets and townships.
Overall, there is considerable potential to carry out archaeological work on the organisation of communities across the SESARF area through large-scale landscape reconstruction in combination with targeted excavation. One good model might be a recent multidisciplinary project based on the catchment of the Whiteadder, extending from the Lammermuirs in East Lothian to the Tweed in the Scottish Borders. It incorporated a complete LIDAR survey and citizen science to help identify new sites by trial trenching (https://whiteadder.aocarchaeology.com/).