In search of the origins of the English village
Just how much information has come from excavation undertaken in advance of development work? In a major survey of Anglo-Saxon settlement, John Blair has been discovering what riches lie in the archives.
It is useless for Anglo-Saxonists to deny it: Roman villas and Norman castles have a hugely greater impact on most people’s imagination than anything built in England between AD 400 and 1050. Yet small-scale works of art from the period — the Sutton Hoo and Staffordshire treasures, the Lindisfarne Gospels, the Alfred Jewel — are probably better known than individual items from either the Roman or later Medieval periods. It is widely realised that much of the substructure of the English human landscape, in its roads, land-divisions, rural settlements, and towns, was formed during the Anglo-Saxon centuries. Did these transformations of the inhabited environment really have nothing in common with the technical brilliance of the small precious objects? The answer is that they had a great deal in common, but the traces are fragile, deeply buried, and hard to decode.
Compared with the Roman, Norman, and Angevin periods, Anglo-Saxon activity lay very lightly on the landscape: houses were short-lived and timber, boundaries were marked by fences or relatively slight ditches, and household goods were made largely of textile, wood, and leather. Only the major churches were built of stone. But lightness and impermanence do not equate with simplicity, let alone crudeness. Recent comparable cultures, especially in Scandinavia and Eastern Europe, show that archaeologically fugitive materials can be vehicles for great artistry and sophistication: why should our tiny surviving sample of Anglo-Saxon works in metal, parchment, and stone be so different from the lost majority executed in perishable substances?
Making grey literature less grey
Thanks to the transformation of English archaeology over the past three decades, the loss is starting to look slightly less absolute. The huge expansion in developer-funded rescue excavation, an outcome of changes to the planning regime during the Thatcher era, has penetrated areas previously almost untouched by the trowel, notably the still-occupied cores of historic villages and small towns.
The quantity of raw primary evidence recovered during the past three decades is vast, but its very abundance creates severe problems of access. Few people have time to trawl journals and monographs systematically, and much of the material is not even available in print. Planning consents require the production of a basic report for the ‘client’ (normally a developer), a copy of which is then supposed to be placed in a public repository. The route to publication in a journal or monograph, though, is much longer and highly selective. This ‘grey literature’ is available (at least in theory, though obtaining it can be laborious in practice), but accumulation of data has run far ahead of analysis.
It was to exploit these untapped riches that the Leverhulme Trust awarded me a three-year Major Research Fellowship to assemble and analyse the evidence for English settlement and landscape from AD 600-1100. Now, as this unique and wonderful phase in my life is ending, I can look back on a project that has been fruitful beyond all possible expectations.
My first task — to cover the published literature — was formidable enough in itself, and took most of the first year. Then, as stacks of ‘grey literature’ reports built up on the floor of my room and I started to trawl through them, I could really feel that I was travelling into the unknown. Digging units are hard-pressed even to meet planning requirements, and often do not have time to contextualise their sites: it is amazing how unsuspected and startling implications jumped out after the simple exercise of superimposing the trenches on the first edition Ordnance Survey map! As I progressed, I realised that the new material was making a difference not just in quantity, but in the fundamental range of questions that Anglo-Saxon settlement archaeology could answer: at last we were getting a quantifiable and representative sample.
But the most memorable explorations were on the open road, as I travelled the length and breadth of England, visited sites and landscapes, and interviewed local specialists. In some ways, the interviews were the most revealing exercise of all. Archaeologists in commercial digging units and local planning departments are overworked people, but they are keenly interested in how their material fits into a wider picture, and pleased when academics take an interest. Everyone was enormously helpful, and their tolerance when I insisted on pinning them down (‘Are you sure you don’t have any 9th-century settlements? If you don’t, why don’t you?’) opened up many unanswered and sometimes unasked questions.
New light on an old problem
So what have I learnt from all this activity? It has opened many new lines of enquiry that will keep me busy for the foreseeable future. Here I shall follow just one of these lines, and revisit a long-running debate: when did the English village originate?
The outlines of that debate are very well known. Through the late 19th to mid 20th centuries, it was widely assumed that the nucleated row-plan village, and especially the compact and structured variety of it found in the Midlands, was integral to the collective nature of Germanic society and imported by the Anglo-Saxon invaders. But when actual settlements from that period were found and excavated, starting with E.T. Leeds’ work at Sutton Courtenay, the realisation dawned that these were very different from later villages: more diffuse, less organised, and less stable.
Since then we have learnt a good deal about 5th- to 7th-century settlements, and excellent work on them has been published (notably Helena Hamerow’s recent Rural Settlements and Society in Anglo-Saxon England). While debate continues on the extent to which these settlements were structured or stable, everyone agrees that whatever they were like, they were very different from Midland villages as we know them.
But in that case, where did the row-plan village come from? The notion of a ‘mid-Saxon shuffle’ became fashionable in the 1970s, and my own research confirms that major changes did indeed happen in the 7th to 8th centuries. The problem is that the new settlements of this era show no sign of village rows and house-plots either. So Vikings in their turn became the fashion of the day: at Wharram Percy, for instance, the basic framework of the Medieval village was for a time ascribed to the later 9th century. Once again, though, hard evidence remained stubbornly elusive.
After my own investigations, I believe that this question is now settled: the ‘classic Midland village’, with linear house-plots and houses grouped tightly along street-frontages, was introduced no earlier than the 11th century, and probably after the Norman Conquest. This will not come as news to a handful of specialists, nor to the many local archaeologists who keep track of discoveries in their own areas, but the point needs to be emphasised as it is still not widely understood. Time and again, the boundary ditches of village tofts and crofts represent a new phase of planning c.1050- 1200. In no clear case, and in only occasional ambiguous ones, can linear house-plot configurations be dated to any earlier period. Anglo-Saxon settlements were not like this, even though they often later evolved into the villages that we know.
Another big step forward has been to clarify regional contrasts in building culture. It was exciting to discover where settlement remains are found, but just as exciting to discover where they are not found. My reading of the excavation reports, and especially my discussions with local archaeologists, make it clear that during c.650-850, the ‘ordinary’ settlements visible to us concentrate almost exclusively in what I am calling the ‘Anglo-Saxon building culture province’: a zone of eastern England comprising the east Midland counties, Lincolnshire, and Norfolk — essentially the river-catchment basin of the Wash — together with parts of east Yorkshire. Within that zone, excavations both in and around existing villages regularly identify buildings, boundary ditch-systems, and associated pottery and finds.
It has long been known that the same zones are rich in metal-detected finds, but what can now be recognised is a broad-based and remarkably prosperous culture expressed both in timber architecture and in lavish personal possessions. Outside this zone, a larger area of central and southern England used the furnished burial rite up to c.600 and then, during c.600- 630, acquired the princely barrow-burials and the complexes of monumental timber halls that briefly displayed the competitive ostentation of emerging dynasties (see CA 265). But the houses, farms, and villages of people below aristocratic and high-monastic status are as invisible as in the British-occupied areas further west. In other words, the zone of extravagant display and the zone of visible ‘ordinary’ settlement are mutually exclusive.
In the archaeology-rich eastern zone, settlements were often planned and structured with precision and careful artifice, though they were very unlike later row-plan villages. Instead, they seem to have comprised extensive groups of spaced-out farmsteads within planned frameworks. Most remarkable is the now-conclusive evidence for technically precise grid-planning in many of these places, with settlements laid out using a standard module of four perches. (But what was a perch? It now seems that the standard was 15 modern feet in most Anglian areas and Kent, but 18ft in Wessex. So was the later perch, with its oddly clumsy length of 16 ½ft, a deliberate compromise?) All gridded settlements so far recognised lie within the date-ranges 600-800 on the one hand, and 950-1050 on the other: periods that correlate so closely with the two great eras of high-monastic learning as to suggest a literate source, probably from thecontinuing methods and the much-transcribed treatises of the Roman agrimensores.
From this wealth of data, four sites can serve to illustrate major themes. Two are published, but can now be seen in a new light. The other two are still unpublished, and give a glimpse of the riches lurking in ‘grey literature’ and excavation archives.
West Fen Road, Ely
A monastically planned settlement and its afterlife
This large area of mid to late Anglo-Saxon settlement near Ely, Cambridgeshire, excavated by the Cambridge Archaeological Unit and published by Richard Mortimer, Roderick Regan, and Sam Lucy, is already well known. Further work by Northamptonshire Archaeology has shown just how large the 8th- to 9th-century settled area was, extending for several hundred metres. The settlement also has an unusually clear context, since its position shows that it was peripheral to the great royal nunnery of Ely. An important question for me was whether this unusually well-reported site is typical of what I have seen on many other east Midland sites in much smaller glimpses.
The combination of a regular layout with a relatively poor life-style had already suggested to the excavators that this was a service settlement, laid out under monastic supervision but occupied by lower-status dependents. What can now be added is that at least part of the original late 7th-century settlement was grid-planned, using the short-perch module, in one-perch ‘boxes’ partly demarcated by ditches.
Within the framework of relict Roman enclosures were two superimposed phases of gridding in quick succession, each of them containing occasional small and rather flimsy timber buildings. Thereafter, the settlement was remarkably stable, showing no Viking-period hiatus, and developing through into the later Middle Ages. But in the 9th to 11th centuries it became markedly less regular in layout, and acquired a group of curvilinear paddocks or stock enclosures.
West Fen Road may be a good illustration of how formal grid-planning was introduced through educated monastic circles. But the east Midlands have now produced several fragments of such grids from c.650-850 underlying villages: can they really all be monastic? It is at any rate interesting to note that West Fen Road’s drift to a less regular form after 850 coincided with the decline and collapse of the high-monastic culture. The rather slight, spaced-out buildings may indeed have been the homes of monastic servants, but the 10th- to 11th-century phase does not look so different in kind from ‘ordinary’ late Anglo-Saxon settlements (notably Stotfold, which we will visit shortly).
The problem is rather similar to the one that has dogged definition of some high-status sites: either monastically planned settlements were very widespread, or they had a powerful and widespread influence on the design of secular settlements. At any rate, we can start to see a continuum between categories of place that were all radically different from later row-plan villages.
Settlement, planning and ritual in the heart of Mercia
Catholme, Staffordshire, in the Trent valley, takes us from the abundant settlements of the ‘Anglo-Saxon building culture province’ to a zone that was politically central but archaeologically marginal. It is well known as virtually the only coherent mid-Saxon settlement so far excavated in the Mercian heartland, and also as a settlement that seemed to show unusual stability during c.600-900. This is another case worth revisiting, with valuable help from local archaeologist Gavin Kinsley: much excavation and research has happened since Catholme was dug in the 1970s, and problems of its layout and chronology may now be closer to solution.
First, a Bronze Age barrow on the east edge of the settlement, crest-sited above the Trent floodplain, has tended to be forgotten, even though the report makes clear that it was respected by Medieval ridge and furrow. This is one of several prehistoric monuments in the vicinity including a henge to the north in a field with the suggestive name ‘Spilpits’ (Old English spel-pyts, ‘speech-pits’) pointing to an assembly-site. An adjoining field was called Arlow, which (allowing for the modification of the first syllable from Old English ea to Scandinavian ar) is clearly ea-hlaw: ‘river-barrow’! Though it cannot be proved that this referred to the barrow on the settlement site, it is a most appropriate description of a mound that would have loomed over the Trent, creating a landmark for travellers by boat.
Second, Catholme too is gridded in short perches. This is slightly less obvious than at some other sites, but when a grid is imposed on the plan it becomes clear that some buildings, though not all, conform to its strict rectilinearity and, in several cases, to the short-perch spacings. This discovery clarifies the previously unresolved phasing of the settlement: there was evidently a single gridded phase which came relatively late in the sequence of excavated structures, perhaps c.680-730. It also gives the settlement a somewhat more formal aspect. We can now see that it centred on two large buildings within a sub-rectangular enclosure. This ensemble formed a kind of forecourt to the barrow, which would have dominated the skyline for anyone approaching from the Roman Ryknield Street to the west. Although later enclosure has confused the picture, it seems likely that the settlement grid was part of a much larger one, in ten-perch blocks, laid out along the terrace-edge.
So what kind of place was mid-Saxon Catholme? Although it has tended to be seen as a potentially ‘typical’ site, it stood at the very heart of Mercia, just below the Tame-Trent confluence and at a nexus of land and water routes between Lichfield, Tamworth, Burton-upon-Trent, and Repton. Could it instead have been a rather special place — maybe a zone of solemn assembly, enhanced by ancient associations of the prehistoric monuments?
Its status did not outlast the Mercian supremacy. A few radiocarbon dates suggesting continued activity up to c.900 may have been over-emphasised: they derive from hollows in the fills of features, and discrete pits, rather than visible structures. Looked at another way, this gives a very intriguing glimpse of the kind of settlement that archaeology misses: someone was living at 9th-century Catholme, and presumably had buildings of some kind. Identifying the ‘building culture province’ leaves one wondering what happened in the rest of England, with its invisible settlements. Does the reversion of Catholme, from a formal gridded complex of timber buildings to a settlement so fugitive that we cannot see it, illustrate the consolidation of a cultural frontier?
The ‘late Anglo-Saxon village’ revealed
Site after site, two thoughts kept returning to me as I waded through reports: ‘Is this representative, or in some way exceptional?’, and ‘If only they’d dug more than just these tiny trenches!’ And then I found the Holy Grail: a thoroughly ordinary late Anglo-Saxon settlement that had been excavated comprehensively, and on a large scale. When I visited the Bedford office of Albion Archaeology, I met Wesley Keir, who is writing up a 600m strip of open excavation along the south side of the village of Stotfold, Bedfordshire. Because of its relative simplicity and short life (the area was first occupied around 950 and deserted soon after 1000), and the panoramic view that it gives of a late Anglo-Saxon settlement landscape, Stotfold is a uniquely clear guide to a mainstream settlement form that can usually only be glimpsed in tiny fragments under built-up villages.
Stotfold’s local context is rather unusual, however. It appears on the earliest maps as quite a complex settlement, with a probable block of short-perch gridding to the northeast, beside the church and manor-house. Roads and boundaries defined a precisely circular zone, about half a mile in diameter and containing the excavated site, with the church standing precisely on its perimeter. Old English stod-fald means ‘stud-fold’, so it seems possible that this circular feature had indeed been a horse-breeding enclosure, possibly attached to the nearby royal centre at Hitchin. Did the dismantling of the stud around 950 release it for housing development?
The excavated strip contained groups of ditched enclosures, associated with homesteads typically comprising a domestic range and one or more outbuildings. The homesteads were spaced out, from west to east, at intervals of roughly 100m to 150m. This cannot be called ‘dispersed settlement’: the homesteads were purposefully organised in relation to each other within a coherent framework. But if it was ‘nucleated settlement’, it was very different from standard later villages, and far less intensive. Countless small evaluations in or near villages have found traces of what look like similar occupation densities; Stotfold explains why archaeologists often find ditches but only occasionally find buildings.
Around 1000, the central section was replanned on a rectilinear layout. Since this is the largest area of Anglo-Saxon grid-planning ever excavated, it is frustrating that the grid is only loosely adhered to: the precise rectilinearity is obvious, but the presumed module of four short perches is only occasionally visible (and could certainly not have been inferred from this case on its own). But the example is fascinating for another reason: the gridding involved replacing one farmstead and its enclosures with two nearidentical farmsteads, 140m apart, framed in the same two rows of grid-squares. Was this a holding split between heirs, who brought in the surveyors to grid it and divide it up equally? The case recalls the entry for Shalford, Surrey, in Domesday Book: ‘Two brothers held it in the time of King Edward. Each of them had his own house, but they lived in the same courtyard.’
A new pre-Conquest castle
An 11th-century newcomer to the village scene was the strongly fortified private residence. There has been a long and fierce debate on how to define a ‘castle’, and on how many of them — if any — existed in England before 1066. Hitherto there have been two main contenders for pre-Conquest castles: the oval earthwork at Goltho, Lincolnshire, which was extensively excavated but leaves some problems of phasing and dating; and a ditch on the later castle site at Sulgrave, Northamptonshire, where the stratigraphy was more straightforward but the excavation done on a very small scale.
A memorable moment, as I leafed rather wearily through a heap of unrewarding printouts, was to encounter an evaluation by Archaeological Solutions Ltd at Fowlmere, Cambridgeshire, and realise that I had stumbled across another Goltho. The excavation included two large parallel ditches, each up to 5m wide. A small amount of pottery in the fills, and some overlying features, suggest a date of c.950-1050. While short of conclusive, the features look very unlike any Norman castle. The ditches had a slightly curved configuration, and I realised that they must represent the west side of an oval enclosure, of which the east side is reflected in a curved road that still survives in the village plan.
The defended enclosure thus outlined was oval in plan, like Goltho in shape and only slightly larger. When I compared the three sites at the same scale, I noticed that Sulgrave (where the fortification overlies a very clear case of a gridded village using the module of four short perches) also has a curving road to the south, outlining a similar oval enclosure of which the excavated ditch section evidently formed part. Was I starting to see a standard pattern for the ‘late Anglo-Saxon castle’?
Then came another surprise. Visible on maps of Fowlmere — in fact still there today as a treegrown earthwork — is yet another oval ditched and banked enclosure called Round Moat. This lies some 200m south-east of the excavated enclosure, with the parish church suggestively sited half-way between them. Minor excavations at Round Moat have failed to date its origins, and no firm assumptions can be made. Nonetheless, it is virtually identical in size, shape, and scale to the earthworks of the late Anglo-Saxon defended enclosure at Goltho. The spatial relationship between the Fowlmere earthwork and the excavated site is intriguing to say the least. Was this a pair of forts guarding an important road from both sides? Or more likely a manorial division, after which two heirs each built a castle with a shared church between?
Pending further work, the possibility must remain open that Round Moat at Fowlmere is our best surviving example of a late Anglo- Saxon defended residence. In any case, the excavated enclosure there strengthens the view that this kind of fortification was not an occasional anomaly, but a mode of aristocratic residence that gained popularity — at any rate in the east Midlands — during c.1000-1050.
Retrospect and prospects
As well as greatly enriching knowledge in matters of detail, the new evidence changes how we see early English settlement in some fundamental ways. The contrast between the eastern zones, which produce extensive physical evidence, and the rest of England, which does not, is startling. This boundary cuts across currently accepted ways of defining regional diversity. For the last decade, the starting-point for understanding English regionality has been the settlement atlas of Brian Roberts and Stuart Wrathmell, which classifies zones according to patterns of nucleated and dispersed settlement shown on early 19th-century maps. Dominating England in this scheme is the ‘Central Province’, the zone of classic Midland nucleated villages and open fields.
My ‘Anglo-Saxon building culture province’ is quite distinct from this ‘Central Province’, being aligned much more towards the east Midlands and east coast. The difference is not, however, so total as to exclude the possibility that, between say 1000 and 1200, the one morphed into the other, perhaps as population growth caused a shift to the intensive farming of claylands emphasised by Tom Williamson in his work on common-field origins. Resolving this problem is now fundamental to understanding the geography and regional character of late- and post-Medieval England.
Equally startling is the degree of organisation now being revealed in central to eastern England. Quite astonishingly large areas of grid-planning can be recognised there, and interpretations of the layout of fields, farms, and settlements in all later periods will need to reckon with it. Here, inscribed across the countryside on a huge scale, is the same technically precise articulation of space that we see miniaturised in the Sutton Hoo jewellery and on the pages of gospel-books.
Cases like Stotfold represent a settlement pattern that was neither fully nucleated nor fully dispersed, but comprised extensive, low-density but structured groups of farmsteads spaced out at intervals of 100m-150m. How did that work in terms of the agrarian economy? The question re-ignites another very old debate: the origins of open field-systems. We know from boundary descriptions in charters that subdivided fields of some kind were common in central to southern England by c.950-1000, but whether they already supported the intensive farming regime of Medieval open-field communities is another matter.
Fieldwork (notably Richard Jones’s for the Whittlewood Project) suggests household manure, which leaves abraded pottery as its trace-element, was not usually spread broadly across the open fields during c.900-1100, in contrast to later centuries. Conversely, the ‘grey literature’ reports show that abraded pottery of just this period is found abundantly in the boundary ditches of the spaced-out settlements. This suggests that each farmstead was surrounded by its small but intensively manured ‘infield’, the arable further from the settlement only being cropped intermittently. The intensification of the open fields after 1100 may have been a crucial stage in the definition of what would emerge as the ‘Central Province’. Finally, what about social status? In the past, I have written confidently about the controlling hand of royally endowed aristocrats after 950, but how that worked on the ground may need some re-thinking. On the whole, excavation seems to be showing ‘manor-houses’ that were not imposed from outside, but developed upwards from below: typically, one in a series of substantial farmsteads was enlarged at the expense of the others. That, I suspect, is what happened on well-known sites such as Raunds Furnells and Goltho, and it may equally underlie the defensive enclosure(s) at Fowlmere.
In fact, this has been staring us in the face, as a famous passage written c.1000 describes a prospering yeoman farmer who, having acquired five hides of land, a church, a ‘fortress-gate’, and other attributes, was ‘thenceforth worthy to be called a thegn’. At least in the east Midlands, local ‘manorial lordship’ may often have emerged inside village communities rather than being imposed on them from the outside. I started this project with some scepticism about the developer-funding regime. I am now convinced that, although current practice is far from perfect, the gains hugely outweigh the losses. In the world as we have it, there is no other way in which such a huge quantity of data could have been recovered from such a wide range of contexts, making it possible for the first time to ask and answer major questions about regional diversity, change over time, and the relationships of settlements with each other and with the landscapes around them. Not all the excavation is of the highest quality, but most of it is good enough to be useful. The biggest current problems are with the archiving and dissemination of the data, where standards and procedures need urgent improvement. It is worrying that many reports are only available through the kindness of the commercial bodies that produced them, and that a good many of them are not picked up by searches of Historic Environment Records. But the shortcomings are with systems, not with people. What I most remember from my enquiries and travels is the overwhelming helpfulness of the digging profession: everyone was enthusiastic, everyone entered into the debates that I was pursuing, nobody refused to contribute. Even — or perhaps especially — in these very difficult times, that must make us cheerful about the future of English archaeology.
This article appeared in issue 291 of Current Archaeology.