A Framework for Exploiting God's Creation - Agricultural Estates and Settlement Landscapes in Early Medieval Dumfriesshire

Christoph Otte – PhD History  

In the 1970s scholars of medieval economy proposed a model framework in which early medieval estates and settlements could be understood in northern Britain, variously known as ‘shire’ or, more commonly, ‘multiple estate’ model (Jones 1971; Barrow 1973). The ‘multiple estate’ model envisages a group of settlements, inter-connected within the boundaries of an estate. The administrative and legal centre of the estate would be the lord’s hall. The surrounding settlements would owe labour services to the lord, such as ploughing his fields or paying rent in kind. Another element of this model is a central church, in charge of the spiritual welfare of the estate population, often situated close to the lord’s hall. As a result, there have been suggestions that in many cases parish boundaries and estate boundaries overlap to considerable extent, as they both potentially address the same population (see Fig. 1; Barrow 1973, pp. 50-63; Winchester 1985, p. 90). It should be noted that this is a generalised model, and requires testing in each individual case.

Fig. 1: The model of the ‘multiple estate’.

The focus of my PhD thesis is the kingdom of Bernicia from c. 600 AD to 800 AD, which encompasses, at least partially, the area of modern Dumfriesshire (Rollason 2003, pp. 20-36 and pp. 87f). The aim of my research is to demonstrate whether the ‘multiple estate’ can be traced in the local settlement patterns of this area, or whether an alternative model might be needed. This investigation relies on a few select case studies based on parishes due to their possible connection with estate boundaries.

Fig. 2: The county of Dumfriesshire, and the parish of Lochmaben, within the southern Scottish context.

In the case study excerpt presented here, I am investigating the landscape of the parish of Lochmaben. My starting point is the pattern of place-names in Dumfriesshire in the aforementioned parish. As an example, the distribution of the Scandinavian place-name element –thveit, referring to clearings or cleared land, can be compared with that of –bý place-names, also of Scandinavian origin and referring to farms or settlements (Fellows-Jensen 1991, pp. 83-87). Both place-name elements are generally dated to the tenth or eleventh centuries and attributed to Scandinavian settlers moving into the study area from Cumbria or other parts of northern England. It is notable that –bý place-names have a more easterly bias, whereas the –thveit elements tend to occur further west.

Fig. 3: The distribution of –thveit and –bý place-names in Dumfriesshire.

This may indicate that the incoming settlers first founded –bý settlements or renamed existing ones with that element, and at a later stage cleared the more marginal lands in order to expand pasture or arable tracts, creating –thveit name settlements. It is remarkable with regard to the parish of Lochmaben that the very part of the parish in which we find –thveit place-names is also the part where the parish boundaries leave their normal route, usually defined by natural features such as rivers and upland ridges, and instead extend into the south-west.

Fig. 4: The –thveit place-names of Lochmaben.

It looks like this section of the boundaries had been super-imposed onto a pre-existing land unit, possibly as a result of the founding of the –thveit place-names. The outline of the parish of Lochmaben excluding the upland extension to the south-west, in the following referred to as proto-Lochmaben, may therefore pre-date the tenth or eleventh centuries in which these settlements may have been founded. Comparative case studies of other parishes will, of course, provide a fuller picture.

Fig. 5: The potential extent of proto-Lochmaben.

Leaving the investigation of boundaries aside, the study of estates and early medieval economy should always take into account the population it was meant to support, and the means by which the land was cultivated. Another way of approaching the question of settlement patterns is therefore to try and gain a sense of the population dimensions for the given area. For this purpose, I used the hypothetical boundaries of proto-Lochmaben. The deduction of major woodlands and water bodies - taking into account that some lochs were drained as late as the nineteenth century - from the total area of proto-Lochmaben has left me with an area of land which may potentially have been agriculturally exploited for the support of the local population in the early middle ages. These are, of course, rough estimates, as the soil quality is not at present taken into account.

Fig. 6: Calculating the usable land (arable, pasture) of proto-Lochmaben.

According to Marcel Mazoyer and Laurence Roudart (Mazoyer and Roudart 2006, pp. 242-45), a farmer with the cultivation implements used in antiquity and the early middle ages (such as an ard, a type of simple plough) could support a family of five on 33.5 ha of land (which includes arable, pasture and a small portion of woodland) in cold temperate climate, or on 61 ha of land in harsher, colder conditions. Due to uncertainties about early medieval climate and soil conditions in Dumfriesshire, I used the middle value of these estimates and applied them to the area of wood- and water-free land calculated for proto-Lochmaben, resulting in a population potential of 408 people. Distributed amongst the five settlements which I am currently prepared to date to the early medieval period between c. 600 AD and 800 AD, this would give a result of c. 81 people per settlement. This number, although obviously presenting a very rough estimate and open to some imprecision, is still remarkably close to the average of 80 people per vill estimated for some areas covered by the eleventh-century Domesday Book economic survey initiated by William the Conqueror (Maitland 1897, pp. 17-20).

At a later stage in my studies, these estimates can give an idea of how much land would be needed per settlement, which in turn might have impacted the configuration of settlements within an estate. Another point which may help illuminate the landscape is to take into account the agrarian reality of the population. The speed of the oxen which are needed to draw ploughs or ards would necessarily limit the distance between a settlement and its pertaining arable fields. It is these practical and spatial considerations which I hope to employ in creating a better understanding of the early medieval rural economy of Dumfriesshire.

Fig. 7: Stilt and head of a prehistoric ard shown in the National Museum of Scotland. Note the illustration below it showing how it would be fitted to a beam, which in turn would be drawn by the draught animal (photograph by author).

Works Cited

Barrow, Geoffrey W. S., The Kingdom of the Scots (London, 1973).

Fellows-Jensen, Gillian, ‘Scandinavians in Dumfriesshire and Galloway: The Place-Name Evidence’, in Galloway. Land and Lordship, eds. R. Oram and G. Stell (Edinburgh, 1991), pp. 77-95.

Jones, Glanville R. J., 'The Multiple Estate as a Model Framework for Tracing Early Stages in the Evolution of Rural Settlement', in L'Habitat et les Paysages Ruraux d'Europe, ed. F. Dussart (Liège, 1971), pp. 251-267.

Maitland, Frederic W., Domesday Book and Beyond (Cambridge, 1897).

Mazoyer, Marcel and Roudart, Laurence, A History of World Agriculture from the Neolithic Age to the Current Crisis (London, 2006).

Rollason, David, Northumbria, 500-1100: Creation and Destruction of a Kingdom (Cambridge, 2003).

Winchester, Angus J. L., 'The Multiple Estate: A Framework for the Evolution of Settlement in Anglo-Saxon and Scandinavian Cumbria', in The Scandinavians in Cumbria, eds. John R. Baldwin and Ian D. Whyte (Edinburgh, 1985), pp. 89-102.

Copyright information for mapping layers in Figs. 2-6:

© Crown Copyright and Database Right 2015. Ordnance Survey (OS Terrain 5 DTM [ASC geospatial data], Scale 1:10000).

© Crown Copyright and Database Right 2015. Ordnance Survey (GB National Outlines [SHAPE geospatial data], Scale 1:250000).

© Crown Copyright and Database Right 2015. Ordnance Survey (Boundary-Line [SHAPE geospatial data], Scale 1:10000).

© Crown Copyright and Database Right 2015. Ordnance Survey (1:25 000 Raster [TIFF geospatial data], Scale 1:25000).

(Mis-)Identities, Tinted Glasses and Riddles: Roman and Natives in South-West Scotland

Alessandra Turrini - PhD Archaeology 

My research looks at native societies and at how they change at the end of the Long Iron Age: in simple terms, that’s the period from just before the Romans arrived, to just after they left. In this post, I will present part of my research from the region of Dumfriesshire, showing the inconsistencies between written sources and archaeology for the Roman Iron Age in Dumfriesshire, as well as the tantalising riddles that still remain unanswered.

To keep within the word limit, let us consider just two sources: Ptolemy’s Geography and Tacitus’s Agricola. The former is the description of a map for the known world, the latter the biography of the author’s father in law. They are both concerned in what is known to Romanists as the Flavian period: that is, the late first century AD. The Geography lists three tribes who live in south-west Scotland, the Selgovae, the Novantae and the Damnoni; and for each tribe it lists a number of key sites, or πόλεις: literally, the word means city, but a looser interpretation of ‘focal site’ is probably better (Ptolemy 1843, p.70 (3.7–9)). Building on this text, we could extrapolate three key factors about south-west Scotland in this period: a) there are cultural subdivision in the native social landscape; b) these subdivisions are reflected in physical landscape subdivision; c) each group has at least a degree of social organisation, with common sites of importance present in each territory.

The Agricola paints a not quite as flattering image: it says that the native tribes were so terrified that they did not dare to stand up the Roman soldiers, whose only issue was the weather (Tacitus 2006, 66–69 (ch.22)). If we were to extrapolate from this source at face value, we would certainly not think about a strong social identity, nor of social organisation. However, when dealing with historical sources, it is always necessary to remember their context. The Geography is based on military intelligence, probably the same one available to the Roman generals who first pushed into Scotland. It is based on an outsider’s point of view, and it may not record anything more than fleeting alliances between otherwise similar communities. The Agricola is, for all intents and purposes, a PR text to praise his main character: it would hardly be fitting of Agricola’s exploits if, after having spent the preceding chapters obliterating one tribe, the next tribe he engages with was not suitably cowered.

Map from one of the areas from my sample of Dumfriesshire, realised in ArcGIS with data from Canmore

The archaeological evidence from the region does not support a reading of centralisation for this period, showing a number of different fortifications and defended settlements potentially in synchronic use in the landscape of Dumfriesshire. It is also at odds with the cowering people that Tacitus portrayed. In fact, it shows a degree of co-operation and integration with the native people of Dumfriesshire, with the birth of an entirely new settlement system. 

Table with distances between closest polygonal settlements across my sample area. Also note the similarity between group A and the distance represented by two militaria (Roman miles)

In the Iron Age, settlements can be described as curvilinear enclosures, while during the period discussed a new system of polygonal enclosures springs up in the landscape; each settlement is located approximately two Roman miles away from the closest one, thus creating inter-visibility chains. None of the settlements in my sample have been excavated, but other similar settlements, like Carronbridge, have yielded a date of first through second century AD, so during the period the Romans were directly present in the south-west Scotland (Johnston et al. 1994). However, the identity of these settlements’ dwellers, their relationship with the rest of the native populations, and what happens to them after the second century AD, is a mystery. Polygonal settlements are more common in the Early Historic period, so the trend they started certainly continues on. There are several other polygonal settlements which do not fit in the two-mile pattern, which may or may not be contemporary with those that do. There also are examples of settlements within the two-mile pattern which have been relocated, suggesting that the system might have outlived the Romans.

Overall, the archaeological evidence is much more layered than the historical sources. It talks of a landscape inhabited by many communities, sharing the same architectural vocabulary; it also talks of change. The native populations did not seem to be afraid of the Romans: they, or at least some of them, embraced the new culture and mirrored these changes in their new homes.

Works Cited:

Johnston, D., L. Allason-Jones, S. Boardman, S. Carter, A. Clarke, B. Crone, M. Dalland, et al. 1994. “Carronbridge, Dumfries and Galloway: The Excavation of Bronze Age Cremations, Iron Age Settlements and a Roman Camp.” PSAS 124: 233–91.

Ptolemy, C. 1843. Geographia. Edited by C. F. A. Nobbe. Lipsia: Carolus Tauchnitius.

Tacitus. 2006. “Agricola.” In Agricola. Germania. Dialogus., 1–115. Cambridge: Harvard University Press.

http://canmore.rcahms.gov.uk