< Main Page

Chapter 1

Resettlement rather than Settlement

When the Roman Army abandoned the provinces of Britannia in the early fifth century, this was part of a crisis of the Western Empire resulting from internal dissensions and incursions of barbarians from the east. Despite the dramatic terms in which most of the contemporary accounts are couched, much of the life of Roman Gaul continued as before. Barbarians had long held high offices in Roman administration and in the army; the army itself had long been composed of non-Italians; but Roman traditions and Roman ways of life were always triumphant. It would be wrong to say that things went back to normal but the population recovered from the shocks of the incursions sufficiently to continue some form of Roman life-style albeit under new rulers with the Roman Church and its cathedrals in the Gallic civitates capitals persisting with the authentic Roman mores and language. But what of Britain whose southern and eastern provinces were just as Romanised as those across the Channel in Gaul?

Roman Britain also came to an end not with a bang but with a whimper. In 401/2 the barbarian general Stilicho withdrew legionary troops from Britain to reinforce the war against the Goths who were intent on invading Italy, leaving the remainder of the army in Britain to set up leaders of their own. Two of these were murdered and the third, Constantine, in order to demonstrate his support for the emperor, led more troops across the Channel in 407 to Gaul, leaving perhaps only a vestige of the army in the British provinces. At this point the civilians took charge, retired what Roman governors Constantine had left in the country, and decided to order their own affairs, informing the authorities in Rome of their decision. Zosimus, a Byzantine historian, whose history is very often a confused account, tells us that the Western emperor Honorius replied, addressing instructions to the chief citizens in the cities of Britain to carry on, presumably until he could sort matters out in Gaul.  We might have doubts about this letter and its timing since Zosimus in Constantinople was producing his history of Roman emperors around the year 500, ninety years after these events in Britain and, in addition, he does not mention that Honorius had already issued an edict in 406 ordering the provincials to take up arms to protect themselves. Judging from the rest of his work, Zosimus, the sole source for the existence of the letter, was quite capable of muddling up the two events.

This situation is typical of the historical evidence for this early period. Our main British historical sources for early events in Britain are Gildas, a monk, probably from the west of England, who wrote a religious tract for the times in the middle of the sixth century which includes, incidentally, some doubtful historical statements, Bede, in 731, who relied on Gildas, amplified his statements and, in the crucial period 449-538 provided no dateable events at all, while in the ninth century, Nennius, in his own words, jumbled together a collection of documents. Additionally, in the ninth century, we have the Anglo-Saxon Chronicles. If we place any firm reliance on these gentlemen and the earlier entries in the Chronicles, we are left totally confused particularly when we try to combine their statements with the growing evidence of archaeology.

It does seem likely that the British leaders of the cantons or civitas, the administrative divisions of the provinces, each containing a town as its capital, with the encouragement of the Western Emperor, took control of the situation after Constantine abandoned the provinces. Two previous leaders elected by the army in Britain had both lasted only a short time and now the third had deserted the provinces with the bulk of the military so one can understand how the British aristocracy must have become disenchanted with the situation and took the law into their own hands. With the permission of the Western emperor, either in 406 or in the letter, they could do this quite properly, legal propriety being an important consideration in the Roman world.  It may be that some military officers were left in post along Hadrian’s Wall with the remnants of their garrisons and took charge of their local areas. Perhaps they were too old or too rooted in Britain to want to follow Constantine. In course of time perhaps their descendents developed into minor rulers (Wilmott and Wilson)

In the accounts by Gildas and Bede, we are told of barbarian threats to Britain by a variety of peoples - Angles, Saxons, Jutes and others who eventually overrun the country - but it is proper at this point to give warning that the question of the colonisation of Britain by these people has become the subject of one of the latest archaeological controversies.  On one side alongside Bede are ranged the proponents of the traditional view of a substantial influx by the Germanic tribes. On the other are those who see the evidence of changes in material culture (burial, village settlement, art-style etc) as indigenous and simply consistent with contemporary developments on the Continent.

A survey of Y chromosomes in the British Isles suggests that the Anglo-Saxons and the rest of Bede’s interlopers failed to leave much of a genetic stamp on the UK. It shows that the Britons weren’t pushed to the fringes of Scotland and Wales by invading barbarians but remained in England and central Ireland. Some Saxons and Danes left their mark in central and eastern England and mainland Scotland while biological traces of Norwegian migrants show up in the northern British Isles, including Orkney. (Dept of Biology, University College). However, whoever these strangers were, they were few, and it is impossible to know when their ancestors settled in Britain. In the case of the Saxons, it could have been at any time during the Roman period or even earlier or they might have been descendants of a handful of individuals employed   as the ‘federate’ which we are told were Saxon immigrants who had been present in Britain during the late-Roman period, employed by the Roman authorities as mercenaries to guard the east coast against maritime raiders and perhaps by the civitates leaders as security guards in the late-period fortified towns. These 'federate' would have been settled on the land and expected to take up arms against their cousins when pirate ships appeared on the horizon. Late-Roman-period inhumations with grave-goods that could have belonged to such people have been found at Caistor in East Anglia and at York.

The homelands of these immigrants would have been the coastal regions of the southern North Sea. This was an area that was low lying and regularly subject to inundations of the sea so the inhabitants lived on patches of high ground along this marshy coast. Their settlements were known as terpen. During the first century AD the inhabitants of the area were the Frisians, relatives of other tribes in the area including the Saxons who during the fourth and fifth century were gradually encroaching upon the Frisian coastlands from the north and east. Also encroaching was the sea which threatened not only the farming lands but some of the settlements on their elevated sites. Ezinge was one of eleven terpen settlements along the left bank of the Hunze river that were threatened at this time when the climatic situation was at its most extreme. Around 400 there appears to have been abandonment of the place. Its inhabitants could simply have decamped to another mound but there are authorities who are in favour of a migration across the North Sea to Britain

How could they have come across?  By sea, obviously, but can we say anything about the vessels that carried them and their animals? A number of ships dating from the general period we are concerned with, from about 300 to 600, have been found in excavations in various places. The Nydam vessels, sunk in a bog in what is now North Germany, the Gredstedbro boat from the same area and the Sutton Hoo vessel fall roughly into this category. .Built of oak, the surviving Nydam vessel is an open rowing boat, clinker-built with iron fastenings, while a pine boat, found at the same site and destroyed during the nineteenth century, was smaller and perhaps used for river transport. Less flexible but quicker and more economical in the building was the Gredstedbro boat, also a rowing boat, with a worn keel that demonstrates that she was frequently beached on landing-places that in those days were commonly shelving river banks. Best known is the Sutton Hoo ship of which nothing survives but the iron cleats but whose construction has been ascertained from study of the sand mound in which she was encased and of which a half-scale model has been made and adjusted to carry a lighter sail (Haywood). All three were open rowing boats which, it is argued, cannot have crossed the North Sea direct from Schleswig and must have made circuitous journeys inshore of the Frisian islands. Basil Greenhill in his book ‘The Archaeology of Boats and Ships’ makes this comment ‘The hardships in such journeys can be imagined. Only the fittest of these people, who had already lived through the rigours of infancy and childhood in the Dark Ages, can have survived to reach Britain…’  If  there were migrants, they would not only have been men; such vessels would have carried family and animals and household goods and provisions as well. Perhaps twenty people, an extended family, would have made up the passengers and crew, most having to row. To carry twenty thousand migrants to Britain, if there were that many, still a fairly insignificant addition to the population of the country, would have needed a thousand ships, few of which, if any, are likely to have made the return journey. This points up one of the difficulties inherent in the ‘invasion’ theory

On many of the sites in this part of the Continent outside the Roman borders, grubenhauser (sunken-floor buildings, or SFBs) have appeared in excavation and on many sites are more numerous than ground-level buildings. They materialize as holes in the ground in which postholes can sometimes be detected that would have supported a ridge roof so that the building would have looked rather like a ridge tent. In Britain they are often referred to by archaeologists as ‘grub-huts’. They were common amongst a variety of peoples in Europe as far east as the Ukraine and as far south as Switzerland and the French Alps from at least as early as the fourth century BC (at Ochterging in Germany), and through the first millenium, outside the Imperial frontiers, down to the twelfth century. They appear on various site-types from peasant villages to the royal Ottonian palace at Tilleda in Germany. Archaeologists have found that those in the eastern part of this range, in an area from the Rhine to the Dnieper and from the North Sea to the Danube, were used exclusively as dwellings and contained hearths or clay ovens. Cold winter winds sweeping across the plains combined with poverty might well have led to folk developing a semi-subterranean, cheap and easily constructed protection against the conditions. To the west they seemed not to have been dwellings but working huts ancillary to a residence although  cool storage is another possility. Perhaps the evidence from the Carolingian site of Morken in Germany suggesting that they could have been used as dairies should be borne in mind but the best evidence is for their use for textile manufacture. Spinning, weaving as well as dairying would be activities suited to semi-underground situations where humidity would help to prevent thread breaking easily and the more even temperature would favour the dairying.

Why are they not found inside the Imperial frontier? Probably because Roman textiles were made on a commercial scale and the work was housed in factory-type buildings. After the end of the Roman period, grubenhauser start to appear inside the erstwhile empire as at St Martin-de-Mondeville in Normandy, Neerharen-Rekem in Belgium and Sézegnin in Switzerland where, during the fourth century, small Roman villa sites were succeeded by emplacements containing grubenhauser and post-built houses. Pliny in his Natural History records that in Germany women worked in underground huts at spinning and weaving and various seventeenth to nineteenth-century texts report the survival in Champagne of huts sunk into the ground and covered with turfs where women met in the evenings to spin and weave so that domestic industry is a reasonable explanation for their use after the breakdown of commercial production.

Experiments in reconstruction of sunken-floor huts have been carried out. Although the results are only pastiches of the original, a good deal of information has been learnt simply by carrying out the exercise.  At West Stow Display Ground in Suffolk this has been done on the actual excavation spots and at Bede’s World, the museum at Jarrow, a two-post example has been reconstructed as part of the display. Another, rarer type of grubenhaus has four or six posts, either one at each corner or with an extra one at each end of the central axis. At Jarrow, the two-post example has one post at each end on the central axis supporting a ridge-beam 2.80m above the ground with ash-poles sloping down from this support to  sill-beams (beams laid flat on the ground) on each side. This ridge-tent-like roof was thatched with heather tied on to willow wattles lashed to the rafters. End walls are made of horizontal oak planks pegged to triangular frames fitted into the roof lines. At one end there is a door made from the same materials and inside it a set of steps leading down into pit.  There is often argument about the existence of a suspended floor over the pit and evidence can be marshalled on each side. Certainly some did not have such floors as an example from (Figure 1) shows where a whole line of loom weights lay in order on the bottom of the pit. If they had fallen from the loom onto a suspended floor, and then onto the bottom of the pit as the floor decayed, they could not have landed in a neat line.

With the abandonment of villas and their subsequent decay and the collapse of the economic system based on large commercial estates, small agricultural settlements appear in south and eastern England where these villas had been situated. Not villages as we would know them but simply collections of residences. Resumably these are the houses of erstwhile workers forced to find other accommodation as the villa buildings fell down. The manufacture of clothing would have been necessary and the ‘grub hut’ a natural corollary to the activity as it was across the Channel.

If there was any movement westward of the newcomers to Britain, it is likely to have involved a peaceful mingling and integrated settlement amongst the indigenous British rather than the bloody conquest described by Bede. Now that the Romano-British economic system had broken down and the villas could no longer operate as commercial, large-scale, agricultural producers, there would have been plenty of surplus land available for prospective peasant farmers to occupy. Perhaps the British landowners shared or rented land with incomers if the practice described as 'hospitalitas' on the Continent was promulgated and practised in Britain. The rule of hospitalitas was imposed on major landowners in what remained of the western Roman Empire after AD418, obliging them to cede part of their land to migrants. In France it is possible that the site at St Martin-de-Mondeville (Basse Normandie) is evidence of this practice and in England one might expect to find similar evidence if there were sufficient incomers for it to have been relevant. Re-allocation of land to peasant farmers seems to have resulted in the establishment of new settlements as most of the excavated early post-Roman hamlets are on virgin sites. Does this rather strengthen the suggestion that these people had previously been accommodated on the abandoned villa sites?

On the Continent it is possible to find settlements where continuity of activity extends from the Roman to the post-Roman period and most of these sequences began with a Roman villa that probably indicates that the people involved were originally part of the villa personnel. At St Martin-de-Mondeville, a Roman villa gave way by the fourth century to a cluster of sunken huts and halls. Later still it was the site of a medieval village. Excavations elsewhere in France and Belgium produce the same evidence of a mutation from Roman site to post-Roman settlement without any change in the manner of subsistence. In England few post-Roman sites have as yet produced evidence of Romano-British ancestry but something of the sort happened at Rivenhall (see Chapter 12) in Essex where it seems that a villa barn continued to be used during the fifth century and a little later a hall was put up and settlement continued close by into modern times. The excavator suggests that the principal medieval manor can be equated with the Roman villa estate. (Rodwell)

In Essex, we have a site at Mucking on the windy north shore of the estuary of the River Thames which began around 400 (Jones) Initial excavation produced a large number of sunken huts nowadays suggested as a base for Germanic mercenaries settled there to guard the Thames estuary against the incursion of settlers or raiders from across the and was said to provide a link with Ezinge because of the coincidence with the chronology and the lack of above-ground buildings in the Dutch site. Mucking proved to be one of the most interesting early post-Roman excavations in eastern England.  It was situated on a river terrace above the Thames and was excavated in difficult conditions over a number of years in the 1970s and early 1980s in advance of gravel digging. The site included two cemeteries that provided at the time a rare opportunity to examine graves alongside the associated settlement. Final assessment of the structures on site is of 203 grubenhauser and 53 above-ground buildings.

Mucking dates from the early fifth but by the end of the seventh century AD the settlement picture there had changed from a scatter of buildings to one of dispersed farmsteads.  If this site really was the early home of generations of incomers from Frisia as some claim then 256 structures built over a period of some 250-300 years is not a very high rate of immigration. As a postscript to what has been said above and to introduce a personal note, it struck the writer, as he helped to excavate sunken huts at Mucking in the teeth of the biting easterly wind in the Thames estuary and the encroaching gravel-digging machines, that grubenhauser were unlikely to have been used for human accommodation.

Not far away is West Stow in Suffolk (West) which also produced timber halls and sunken huts. The whole site, covering some two hectares, was excavated so it is the only settlement of its sort about which we have full knowledge. Its above-ground huts  have been suggested as family halls or farm houses surrounded by a number of smaller buildings for various domestic or agricultural uses.. The excavator believes that the grubenhauser were actually cellared structures rather than sunken-floored buildings and the simulations which have been constructed on the site include an example of each interpretation. Some cellars are covered with planked floors on which clay hearths are built. Finds on the site included bone tools, bone combs, pottery sherds, bronze pins, buckles, bucket fittings, iron knives, saws drills and saws spindle-whorls and bone weaving pins. There is evidence that at least one sunken-floored hut was devoted to weaving since a hundred loom-weights were found inside it. Animal bones were those of cattle, sheep and goats while pigs were apparently kept in neighbouring woodland. Three hundred metres away was the cemetery in which some of the dead were accompanied by a spear and a shield, women with brooches and beads, other bodies were cremated. The original settlement lasted from cAD450 to cAD650.  At the end of the sixth century and in the first decades of the next, people began to drift away from the village and built houses in the area where the present church stands. This ‘settlement drift’ is a feature of a good many early villages in eastern England and elsewhere. Fieldwork in the adjacent River Lark valley has suggested that during the fifth century that there may have been as many as one early settlement in every modern parish.

One uses the term ‘hall’ to describe a number of different above-ground rectangular buildings of the period.  At Thirlings (Miket and O’Brien) in Northumberland some twelve such buildings were recognised, varying in size from 21 to 96 square metres in floor area and one of these with a floor area of 74 square metres (say c5m by c15m) was selected for reconstruction at the Bede’s World Museum. Thirty-five wall posts were used, each approximately 300mm in diameter and were made from stripped oak stems cut to 3.10m in length and were planted directly into the ground in post-holes, a method described by archaeologists as ‘earth-fast’. On top of them were fastened wall-plates, made of square-sectioned oaken baulks fastened on top of them with undersides 2.0 metres from the ground surface. End wall frameworks had to be higher to stretch up to the sloping pitch of the roof, at its apex some four metres from the ground. To construct the walls, augers were used to bore holes in the wall timbers and into these holes frameworks of hazel and willow wattles were fixed leaving a door-space in the middle of each side. Wall covering was made from puddled clay, straw and manure, a material known as daub which cracks as it dries and is re-patched and lime-washed as the final operation.

Inside, three sets of posts carry the roof structure. Three towards the west end and two sets of two equally spaced in the rest of the building carry collars on which the purlins rest. In the middle of each collar is a king post that rises to support the ridge-beam of the roof.  Common rafters made from ash poles which rest on the ridge beam are notched at their lower ends into the outside edge of the wall-plates. Windows have been let in high up in the western end wall. The roof is covered with a reed thatch on top of a foundation of wattles of hazel and willow. There is no chimney, the smoke percolates through the thatch keeping it clear of vermin.

Materials used were thirty tons of green oak from trees aged from 50-60 years, 1800 bundles of reeds and 125 bundles of sedge, ash-poles and hazel rods, while some 25 tons of boulder clay were used for daub. This represents a sizable investment for the time and suggests that in order to provide the materials for the house and eleven other similarly constructed buildings, the local  oak woods were managed, as they must have been for a thousand years  to produce building materials and, later, ship timbers. Sometimes the woodmen would manage the trees by tying down a branch so that in the years to come it would form a useful curved beam. Otherwise, naturally curved branches would be sought out for certain jobs like making ploughs. Natural joints too would be preferred to creating artificial joints. Tools used were wedges for splitting logs while shaping was done with adzes and axes, as little wood as possible being removed so as to conserve strength while the bast fibres under the bark were employed to make ropes and the bark saved for the tanners. Bow saws were employed for certain jobs.  Green timber was always used, with oak being the commonest. Waste wood became charcoal and thin strips of wood and hazel or willow withies were woven into baskets and containers for which we today use the ubiquitous plastic.

Excavations on waterlogged sites both in Britain and on the Continent demonstrate how much the people of this time depended on timber. A list of wooden articles (treen) includes furniture, boats, boxes, tubs, butter churns, cheese presses, bowls, plates, cups and spoons as well as tools and weapons. Carpenters had a large and varied too-chest that included the axes, adzes and saws as well as planes, draw-knives, wedges, hammers, chisels, awls, mallets and gouges and they would be kept in workable condition with whetstones or grindstones. Woodworkers would also perhaps have used a pole-lathe for shaping bowls and cups.

It is difficult to over-estimate the importance of woodland at the time. So much was made from timber since metal could only be used for small articles and everything of any size had to be a timber construction and this was to be so for many hundreds of years up to the Industrial Revolution.. An early instance of the importance of woodland is given in the laws of King Ine, put together in the late-seventh century, among which is one penalizing anyone who destroys a neighbour’s trees by fire ‘he shall pay sixty shillings because fire is a thief’.  In Domesday we hear of woodlands ‘for making the fences’, ‘for house-building’, ‘for fuel’ and of particular kinds of wood oak, ash and willow. Very often woodland is described according to its capacity for acting as swine-pasture or by the rents paid for its use for this purpose, known as pannage. The large numbers, up to 2000 animals mentioned in some accounts, must mean that they are examples of commercial pig farming.

Also at Bede’s World, another timber hall was reconstructed, the model in this case being a building excavated at Hartlepool (Daniels). This one was smaller, 5.1m by 3.4m, a floor area of some fifteen square metres. Again, the wall-posts were earth-fast. Four corner posts and in each long wall eight posts with the two central ones rather larger in scantling to give support for fixing a door on one side and a window on the other. The end walls had six wall-posts each all the same height as the other posts. On top of the four walls were placed wall-plates on which two tie-beams rested in the centre of the roof. On each tie-beam is mounted a king-post rising some 1.65m and they support a short ridge-beam and from the each corner-post diagonals rise to join their respective ends of the ridge thus forming a hipped roof. Common rafters were ash-poles and willow wattling was tied onto them to support the straw thatch.

These roof reconstructions are reminiscent of those that remain to us from the  medieval period and, if they are reasonably correct, they emphasize the continuity in timber construction which probably had its origin thousands of years before. What we see in these pastiches is a fairly plain representation of how the buildings may have appeared in the past but we remember the delight our ancestors took in decoration of all sorts and we can be sure that the houses were far more lively than they seem to us.

Halls, such as these reconstructions, were made comfortable by hangings around the walls to keep out the draughts. These were surely embroidered since embroidery was one of the outstanding female skills of the period whose latest greatest achievement in our eyes is the Bayeux Tapestry. Roof spaces were very useful areas where food could be hung for smoking in the dusky area under the thatch. Salt was expensive and smoking was easier and in many ways a more palatable method of preserving. Planks placed across the tie-beams would have provided extra storage with the advantage that the greater the weight in the roof area, the more stable the building. Whether some of the timbers were carved is a moot point. We know how ships could be decorated in this way but carving house beams might have the effect of weakening the structure. Even so, it is possible that the non-load-bearing timbers were treated so or painted to give more character to the interior ambiance.

West Heslerton in Yorkshire is an early settlement and cemetery dating from cAD450 to cAD850 and an excavation (Powlesland) examined an area of some four hectares. This important high-class settlement was initially established around a spring and by 500 it had grown to cover an area of around 25 hectares, incorporating individual zones devoted perhaps to housing, craft industry and agricultural processing, and a hundred and fifty years later the place was in decline. Over 200 buildings have been excavated, both above-ground buildings (halls) and grubenhauser. More than a quarter of a million animal bones were recovered together with imported glass vessels, pottery, metalwork and Niedermendig lava quernstones from the Frankish Rhineland. Native products also included metal objects, bone artefacts, pots and textiles. The settlement was finally deserted by about 850, the inhabitants decamping perhaps to the present village 500 metres away in the same manner as the people of West Stow moved in their late phase.

An associated cemetery was also excavated and forensic examination has provided corroborative evidence for the gene survey described previously. The teeth of 24 skeletons were examined for the isotope values that are ingrained in children’s teeth by the local environment before the age of twelve.  Only one individual was thought to have been from Continental Europe i.e. an immigrant, perhaps a Merovingian Frank. Of the British people, thirteen came from across the Pennines and the rest from the surrounding area. (University of Durham) What is also interesting about this report is that there seems to have been considerable movement of the population at that time, folk within the country were more mobile than the traditional view we have of them rooted in their settlements. The Vale of Pickering was a well populated and rich agricultural area at that time so that one can understand why it would attract people from elsewhere.

Remains of a large settlement have been found near Melton Mowbray (Neil Fynn) in Leicestershire that covered at least 13 hectares. In the area examined fourteen hall houses have been found together with eighteen grubenhauser and numerous hearths that probably represent a group of farmhouses. Dating suggests the settlement originated in the fifth century and survived at least until the seventh. Finds are typical of sites of this period, including, loomweights, spindle-whorls, bone combs, bone pins, iron knife blades, copper-alloy brooches and pottery.

Chalton in Hampshire (Cunliffe) was a settlement on the chalk dating from the sixth and seventh centuries AD. In the ten per cent of the area of the site that was excavated, the buildings belonged to four successive periods of development. Halls of different sizes were found, some trench-built and one grubenhauser. Finds included grass-tempered and sandy pottery, iron objects, part of a hanging-bowl, a common prestige possession of the time, and animal bones. When the site was abandoned, people probably moved to smaller hamlets, one of which is now the modern village of Chalton.

Catholme in Staffordshire (Losco-Bradley) was an extensive settlement with three building types (sunken-floor huts, rectangular halls and long-houses that, as the name suggests, have curved sides, sometimes described as boat-shaped) which date from the sixth century and later. About two hectares of the site was excavated in the centre of which was an enclosure containing a large boat-shaped building and subsidiary structures that was approached by a road. This has been interpreted as the farm of a leading inhabitant. In several of the grubenhauser there was evidence of spinning and weaving. Catholme was longer-lived than most early settlements, surviving into the tenth century and displaying a feature that is becoming apparent in other sites as more excavation is carried out and that is the number of individual farms in separate enclosures accompanied by sunken huts and structures raised off the ground out of the reach of rats. The count at Catholme is seven different farms during the four centuries of occupation.

Near Lowestoft in Suffolk (Boulton) is a settlement that is dated from the sixth to the eighth centuries and is situated within a Roman field system so far the remains of one hut have been excavated and traces of others found together with nine grubenhauser containing hand-made pottery, some with stamped decoration, loom-weights and iron slag. No hearths have been found in the huts. The evidence available suggests that  the grubenhauser were working huts, one housing a loom, but the associated halls are still to be found.

Small scale examinations of early post-Roman settlements have taken place elsewhere, mainly in East Anglia where over two dozen have been sampled. Most of these partial investigations have produced sunken-floor buildings that are easier to pick up with modern survey instruments than post-fast structures. The locations of these settlements were short-lived, their activities being moved on to another part of the same farm estate or perhaps to another close by. There is no reason why agricultural workers should not have been mobile at the time; there was nothing in the nature of a village or church to tie them to a particular farmstead and, as far as we know, they had had no loyalty to a particular spot even though earlier on during the Romano-British period many may have been tied to the villa estate on which they laboured. With little in the way of possessions, they could roam where they wished looking for agricultural work and they would be especially welcome if they had some special agricultural skill or even something quite else like story-telling or fortune-telling, both important items in the social life of the time.

With the disappearance of the Roman army, the villa owners, the major farmers, had lost the main customer for their food and material products, and they faced an unexpected economic disaster. Pollen evidence from the north demonstrates how badly affected such producers were  Amongst the villa products had been wool for the Imperial weaving-mills, the gynecea, some of which, according to the Notitia Dignitatum were located at ‘Venta’, either Venta Icenorum (Caistor-by-Norwich) or Venta Belgarum (Winchester) or possibly at both places. No doubt there were similar civilian mills located in other places which disappeared with the demise of the towns and the more innovative amongst the sheep rearers saw that they had one product with a future if they marketed, instead of the raw wool, cloth, which was always in demand. This explains the sudden appearance of grubenhauser, weaving sheds in other words, clustered around the new timber hall residences that had taken the place of the tumbled down villas. Thus the ‘grub huts’ are not the evidence of Anglo-Saxon incursions but a sign of business initiative on the part of British farmers.

In Kent, the River Darenth has a number of Romano-British sites along the valley bottom and post-Roman cemeteries uphill of them. Thanks to excavations on these sites we know that, unusually, during the fifth century, the dilapidated villa buildings continued to be partly utilised and wooden buildings were erected close by. Do these belong to the erstwhile workers on the estate who were coalescing a new settlement close to the old villa nuclei? If so, one wonders why it did not happen more often so a site of a Romano-British villa in excavation is routinely found stratified beside/beneath a later settlement. These Kentish sites share common features of construction methods and construction materials, types of buildings and finds. Most seem to have come to an end when people slipped away to other nearby locations for reasons that are not apparent to us today.

Settlements of the Roman period ancestral to post-Roman sites are presently being identified in the Army Training Area on Salisbury Plain. It has been suggested that when excavation is allowed to take place on them some of the characteristics we have been discussing above that we are accustomed to think of as post-Roman may appear in them and provide chronological links between the two periods. At the moment we know too little about independent native faming settlements in southern Britain during the Roman period to make meaningful comparisons..

Smaller timber-built halls and sunken huts were common elements in the repertoire of the early English farm-builder but we also have examples of long-houses which contain internal supports in the form of rows of pillars that support the roof, like a  basilican church. Occasionally they have curved sides as at Chigborough Farm in the Blackwater valley in Essex  (Wallis and Waughman). What they were for is not yet clear. Too big for residents unless they were shared with the livestock, perhaps they stored farm produce like wool.

Introduction
main page
The Evidence of early Post-Roman Burials