Chapter Four
THE NINTH AND EARLY TENTH CENTURIES
Holding Out Against the Heathens

When, in 793, ‘the ravages of heathen men miserably destroyed God's church on Lindisfarne, with plunder and slaughter’, many contemporaries saw the Viking raids as a sign of God's direct intervention in human affairs and a punishment for their own failures to walk in paths of righteousness. There are no records that either state the Vikings’ own motives for their raids—whether lack of land or lack of silver—or give a realistic assessment of the impact that they had on western Europe. The drama of the narrative of ravaging, fighting and atrocities has tended to obscure other tensions that also affected development of the period.

Despite the chronicler's vivid description, Lindisfarne was not totally destroyed, for there still remain from it the wooden coffin and other relics of St Cuthbert, which would certainly have been lost if devastation had been absolute.1 An evocative reminder of the raid may be a tombstone from the site that seems to show a war-band; if it is correct to interpret that scene as a Viking ship's crew, the stone would be archaeological evidence of at least continued use of the area for burial purposes after 793. There is, in fact, no reason to doubt the twelfth-century record that the monks ‘continued for a long time’ to live at the abbey, only leaving the island for safer territory later in the ninth century. Physical evidence, limited as it is, gives no direct evidence of the effect of the 790s raids and their immediate successors on the northern churches, or of their abandonment at that time. Some decline in spiritual life may be shown in Jarrow by the change of use of what had been a high-status building into a metal-working craft-shop. Similarly at Whitby, where the coins and other artefacts show that occupation of the abbey site continued long into the ninth century, a late building overlies the edge of the Anglo-Saxon burial-ground, perhaps indicating that the cemetery's sanctity no longer commanded the respect that it should have enjoyed.² Such signs of physical and spiritual decline are consistent with the decline in output of major works of art and literature from these monasteries after the middle of the eighth century. Preeminence in the north had perhaps passed to the archiepiscopal centre at York, and falling standards at the older, more isolated houses are not necessarily to be associated with the initial impact of Viking raids.

Since archaeology does not suggest that the Norse plundering parties had much immediate effect upon the monasteries which were their most obvious target in England, few signs of their effect in other spheres might be expected, particularly away from the vulnerable north-east coast. Although the charters suggest that bridge and fortress-building were increasingly added to the services owed to their king by land-holders,3 the evidence of the causeways at Mersea and Oxford and the ditched enclosures at Hereford and Tamworth show that such impositions were well established before the Viking raids, at least in central England, and were probably an extension of what must have been much older duties to engage in earthwork construction, shown not only by Offa's dyke, but also by the many earlier dyke systems. The charters’ instructions are the expression of the increasing formality of royal authority. Whether any constructions were actually undertaken to protect the coast from Viking raids, as Charlemagne was doing on the other side of the Channel, is not known: no fort-building or restoration of walled Roman sites can be ascribed to the first half of the ninth century. According to the Anglo-Saxon Chronicle, it was not until the 830s that the Vikings became a threat to the whole fabric of English society, leading in the 870s to the collapse of kingdoms. These, mainly Danish, raiders were prepared to fight at sea or on land near their ships, storming in quick succession Canterbury, London, Rochester and Southampton, nowhere on or near the coast being safe. There were major churches at most of the places important enough to merit naming in the sources, but they were also for the most part known as trading-stations. The degree of physical damage done to the ports cannot be assessed— nothing at Southampton suggests wholesale burning, for instance, in contrast to evidence left by the much later French raid in 1338. What Southampton does show, however, is a considerable decline in trading activity, witnessed by the diminishing number of coins found. This is not an ideal standard by which to measure change, for even in the early years of the ninth century, when there is no other reason to doubt the port's prosperity, there were fewer coins lost, perhaps because the new broad-flan pennies were easier to handle than the ‘sceattas’. Nevertheless, such pennies as there are at Southampton date mostly from the early part of the ninth century, tailing off to almost none at its end.⁴ This is consistent with the other artefacts, none of which seem necessarily to post-date c. 850. The mid Saxon port was effectively abandoned. Other English trading-stations may have declined similarly, though the evidence is much less clear-cut, their sites having had greater subsequent use. The limited information from the Strand area of London does not contradict claims for its ninth-century disuse, however. Most significantly, the waterfront installations at the great continental harbour at Dorestat were not being repaired after c. 830; even though activity there did not cease altogether, the coin record tells a similar story of trade decline.⁵

The effect that the Vikings could have on economic systems had already been seen in Northumbria in the 790s, where their raids seem to be the best explanation for the abrupt end of the kingdom's ‘sceatta’ coinage, which could not be maintained if silver stocks had been depleted.⁶ The trading-stations suggest similar disruption, the raids creating a crisis of confidence— and physically destroying or removing any ships drawn up on their beaches. Coins did not cease to be minted, however, even if minting was episodic. Although fewer in terms of stray-loss finds than the ‘sceattas’, pennies in many ninth-century hoards (themselves a token of particular crises of confidence in the 850s and 870s) show that considerable numbers existed. The silver in these pennies became increasingly debased until the 870s, probably because the kings were manipulating the coinage to increase their revenue.⁷ If scarcity of silver bullion had been a serious long-term problem, there would not only be fewer coins, but there would not be such a large number of silver brooches, strap-ends and other ornaments found with the coin-hoards, which provide dating evidence for the many objects—more than for the eighth century—that have been found without associated coins, notably swords with a range of finely decorated hilts (4, 1). Nor was silver the only precious metal: gold was used for rings and some sword-hilts, and a few gold coins were imported, such as the munus divinum ‘solidi’ of Louis the Pious and imitations of them. All this ninth-century metal had to come from the Continent, and its quality does not suggest that trade was brought to a virtual standstill for more than very short periods, if that; there is a sprinkling of silver coins of Charles the Bald and other ninth-century Carolingian monarchs, especially in southern England. Only Northumbria stands apart from this general picture, and even there coinage was re-introduced after an interlude, although the new ‘stycas’ which were unique to that kingdom never had more than a 40 per cent silver content, and were increasingly debased to as low as 2 per cent.

4, 1. Drawings by H.Humphries of a ninth-/early tenth-century sword from Gilling West, North Yorkshire, found in a stream in 1976—finds of swords in rivers suggest deliberate ‘sacrifice’, but this one was perhaps just an accidental loss. The hilt has silver bands, wider than the iron as they would have been set over a wooden or bone handgrip. The schematic animal heads on the ends of the pommel may be a reference to the creatures that would feed on the sword's victims. The blade was pattern-welded, i.e. composed of iron rods, alternately twisted and left straight, then welded together, and etched after grinding and polishing to show up the pattern. The edges were also separate strips, with a higher carbon content for extra hardness. The pattern welding reduced brittleness, making the blade less likely to shatter. (Width of pommel: 838 mm).

Despite the decline of the international trading-stations, the coins and objects show that commerce and exchange must have continued, although there are no sites that can be said to have been primarily market centres. Royal residences and other estate centres may have retained their economic rôle, with food renders still brought to them by subjects—hence the Vikings’ concentration upon gaining possession of them, the construction of the mill at Tamworth in the 850s despite the raids and the merchant Othere's visit to King Alfred's court. The continued importance of such places is shown by the excavations at Cheddar, Somerset, where a long timber hall, perhaps two-storeyed, is a record of ninth-century expenditure at what was at any rate in the next few decades a royal palace.⁸ Its economic rôle is shown particularly by the metal-working debris found there, although it was not on the scale of the iron production at Ramsbury, which seems to have come to an end during the first half of the ninth century. Goltho, Lincolnshire, is a complex that came into use probably some time later where a spread of smithing slag and a stone hearth interpreted as ‘industrial’ suggest craft activity of some kind. The buildings associated with it are of a size to suggest no more than farmsteads, but their replacement by a ditched enclosure containing a twenty-four metre long ‘hall’ and ancillary buildings could imply that the site was for high-status use from the first. The quantity of objects associated with weaving around one

of the smaller buildings has led to its designation as a ‘weaving shed’, suggesting production on a greater than domestic scale.9

It is difficult to assess the effect of the ninth century upon rural settlements generally. Dating of excavated sites is usually imprecise because coins are rarely found and the different types of pottery have a very wide time range— if pottery exists at all, since some areas of England, particularly in the west except Cornwall, seem still to have been aceramic. Consequently, few developments in settlement patterns tend to be attributed to the ninth century specifically, although in East Anglia, Ipswich and the new Thetford wares are valuable indicators. Field-walking in parts of East Anglia has produced quantities of Thetford ware in areas around churches, suggesting that it was in the later ninth and tenth centuries that these were becoming village centres. This does, however, assume that the earlier Ipswich ware would be found if those areas had been in use during its period of manufacture, and it is not certain that that would be the case, particularly since it was not found in the excavations at Maxey, although it is known in the immediate area.¹⁰ Although there were quite substantial buildings at Maxey, they were not rebuilt and probably did not outlast the ninth century.¹¹ Nevertheless, occupation seems to have continued within the immediate vicinity. The ninth century has not yet produced evidence of the widespread abandonment of sites that was such a feature of the preceding parts of the Saxon period, and this suggests at least stabilisation of the settlement system, rather than contraction caused by Vikings or other external factors. Only in the 890s does the Chronicle record plague and, although contemporaries could not have known it, the end of the century was the beginning of a long-term climatic improvement which gave warmer summers and a longer growing season. Expansion of agriculture on the high ground in the north seems to have continued, for although Simy Folds may have been abandoned, at least for permanent occupation, Gauber High Pasture in West Yorkshire's Ribblehead had a farmstead with a building sufficiently substantial to suggest year-round occupation, and the extraordinarily fortunate find within it of four coins to prove ninth-century use.¹² In friendlier environments, Wharram Percy seems to have prospered, and on the north Cornish coast at Mawgan Porth a courtyard-farm complex was probably founded in the ninth century, or soon afterwards.¹³

These sites all stress the pre-eminent role of agriculture in the economy, and Goltho shows something also of the processing of agricultural products. There is a little information about the importance of metals also, though no new iron-smelting complexes such as the eighth century has produced have been found. The quality of iron-work available is shown by the sword blades, many of which were made of iron rods twisted and welded to achieve some malleability from phosphorus-rich ores, and now with hardened steel edges. Ores for the latter probably had to be imported, and in any individual case it is usually not possible to be sure that the complete blade is not an import, though English smiths must have been responsible for many. At any rate, the swords were now much more effective weapons than they would have been in the fifth and sixth centuries. By varying the twisting of the rods, different patterns could be achieved which, etched to bring out the brightness of the phosphorus, would produce a light-catching, gleaming blade (4, 1). The valuable hilts added to many swords show the prestige that they bestowed.¹⁴

It is not known which iron-ore deposits were being exploited in the ninth century. Other metals that were probably being extracted included tin, less a feature of the ninth-century coinage when the silver alloy was pure, but certainly added very deliberately and in some quantity in the 860s to the Wessex and Mercian pennies. The hoard of silver at Trewhiddle, Cornwall, with a range of coins that show that it was deposited no earlier than 868, may well be an indication of the importance of Cornish tin. Earlier debased pennies contained much more zinc than tin, and zinc was also a major component of the Northumbrian ‘stycas’. The ores may well have been extracted from deposits around Alston, in Cumberland: were supplies from there not available to southern England in the 860s because of Viking activity?¹⁵

Coinage debasement is one measure of the economic difficulties of the middle part of the ninth century, and other disruption can also be seen, particularly in the Church. Although they survived the 790s, the great monasteries of the north did eventually disintegrate; apart from Norham, used by the migrant Lindisfarne community, none was still in existence by the end of the ninth century. The centre and south of England also had their losses, as King Alfred lamented. The nunnery at Nazeingbury probably survived well into the ninth century, but radiocarbon dates end at 870 at the latest. At North Elmham, on the other hand, there were considerable changes, with mid-Saxon ditches and wells being filled in and reorganisation taking place from at least the end of the century, but not with any obvious interlude or change in the site's function or the status of its occupants.¹⁶ The size of the church at Brixworth and the quality and range of sculpture at Breedon-on-the-Hill show something of the physical scale and breadth of contacts of ecclesiastical establishments in the eighth and early ninth centuries: neither of those churches was to regain its prestige and wealth after the ninth century. The position is graphically shown at Canterbury, where the number of literate scribes at Christchurch can be seen from surviving manuscripts to have declined sharply in the 850s, until by 873 only one elderly writer was left, his work now pathetically inept. The Viking raid of 851 seems to have made its mark. The middle years of the ninth century at Canterbury were a great contrast to the early years, when Christchurch had been rebuilt, estates purchased and the archbishop's moneyers very active.¹⁷

Other evidence of the raids are graves recognisable as those of Vikings. A male found just outside Reading, Berkshire, had with him a horse and a sword which no Christian would have been supplied with. The sword's hilt was ornamented in a purely Scandinavian manner, now known as the ‘Gripping Beast’ style. The grave was in an area called the ‘Vastern’, an English word for stronghold, between the Thames and Kennet rivers, a likely position for the rampart recorded by Asser as constructed at Reading by the army which seized the adjacent royal residence in 870–71. If the buried man was a Viking who died during that episode, he had with him a sword on which the hilt was already three-quarters of a century old when buried, for ‘Gripping Beast’ is an eighth-century style. The hilt is fairly well worn, and could have been of some age when it went into the ground, though it does give grounds for caution in too readily associating recorded raids with particular archaeological discoveries. This can be done with greater certainty when coins are found with burials: another grave at Reading had a coffin for which a small collection of coins provides a date in the early 870s. That grave was in the churchyard of St Mary within the town, which was probably close to the site of the royal palace.18

Many Viking-style burials and objects have been found in churchyards, implying that the aliens’ wish to identify themselves in death with a place of religious significance overrode any distaste for lying alongside the forebears of their vanquished, although in some cases the Viking may have come first and the churchyard developed around him because of his remembered importance. Some important burials were in mounds that appear to have been and to have remained isolated.¹⁹ One dramatic instance of Viking use of a Christian site is Repton, where a gold ring was found close to the eighth-century crypt and whatever enlargements it had by then accrued in a burial which also had five pennies in it dated to 873–76. Furthermore, the sunken masonry structure to the west, by then apparently in a state of disrepair with stucco falling off the walls, contained a mass of bones from at least 249 individuals, an axe, knives, animal bones and five more pennies, of 872–74. These cannot but be a testimony of the presence of the Viking army recorded as being in Repton in 873–75. That army probably dug the ditch which has been found leading towards the River Trent, perhaps to create a defensive enclosure incorporating the church as a tower. A seventeenth-century record suggests that the other masonry structure had had a central burial, probably that of a Viking leader.²⁰

The Repton-based Vikings had presumably gone there because it was a known centre, and their use of the Mercian royal family's burial-place for their own leaders may have been a deliberate claim to have inherited governmental authority, since it was at that time that they ‘drove King Burgred across the sea’. The 870s saw Viking armies ‘sharing out the land…to plough and to support themselves’ in Northumbria, East Anglia and parts of Mercia. The change of government in the north can be seen by the cessation of minting of ‘stycas’, which had continued in the 860s and perhaps into the 870s, showing that some sort of central authority survived in Northumbria until then. Existing dynasties and land-ownership patterns were disrupted throughout what became known as the Danelaw, though the extent to which the territory was actually peopled by Danes or other Scandinavians is not easy to judge from archaeological evidence. Sites like Goltho and North Elmham are in the Danelaw, but do not reflect anything markedly Scandinavian in their culture sequences: the claim as negroid for the skeleton of a woman whose presence would at least have shown long-distance slave-trading such as the Vikings were famous for must, rather sadly, be discounted, as a similar person in Norwich has been shown to have had merely a protruding jaw!21 At a site lower in the social hierarchy, ‘stycas’ demonstrate that Gauber is quite likely to predate Healfdene's army's settlement, and cannot be used to support the old argument that there was such an influx of new arrivals that widespread colonisation of new land was forced upon them.

The arrival of new landlords in parts of northern England towards the end of the ninth century was once thought to be witnessed by burial markers such as the ring-headed cross at Middleton, North Yorkshire, carved with a contorted beast on one side of the shaft and a helmeted male surrounded by weapons on the other. The beast is a crude rendering of the ‘Jellinge’ style, one which is derived from the sort of creatures found also in southern England: the male is unprecedented in the English world, however, and it was argued that a Viking who had recently taken over an estate at Middleton ordered a local craftsman to make for him a Christian monument which nevertheless would represent him as though buried in the old style with his weapons. This theory received a jolt when excavations in York produced fragments of Jellinge sculptures in contexts firmly dated to the first half of the tenth century. Monuments like that at Middleton cannot therefore be used as evidence of the aspirations and the tastes of the founding fathers of a Viking aristocracy.²² What the sculptures do show is that concepts about commemoration of people and representation of ideas in northern England during the tenth and eleventh centuries were very different from those further south. It is not, however, a difference that the distinguished the whole of the Danelaw from ‘English’ England. East Anglia has no such sculpture, the aristocracy there choosing to do without stone grave-markers and crosses. This is not just because of the difficulty of getting stone, since it could have been acquired from Barnack and other limestone quarries if demand had been sufficient; it could, however, be a sign of different practices, perhaps of quicker assimilation. Different parts of the Danelaw were quite likely to have had, or rapidly to have developed, differences of this sort. Nor are there other signs of a Danish culture in the rural Danelaw; stray finds of metalwork are infrequent and as likely to be found in English as in Danelaw England, ball-headed pins being an example.²³ There were no differences in the type of agriculture practised in Denmark which can be recognised by new techniques introduced into England. Both countries seem already to have been in the process of introducing the wheeled mould-board plough, fields divided into strips and nucleated settlements. Similarities of buildings at sites like Gauber to Norse structures may simply be a common response to similar terrains.²⁴ Place-names rather than archaeology suggest Scandinavian rural settlement, and they leave unresolved the problem of the extent to which they result either from the impositions of a new aristocracy whose words came to be used by English peasants or from the influx of migrants on a large scale; the latter is perhaps indicated by the use of Scandinavian words even for everyday things like field names, not just for the names of estate units. Scandinavian personal names such as those of moneyers on coins issued at Danelaw mints could also result from the influence of an élite, unrepresentative of the peasantry. In York, indeed, Scandinavian names did not even predominate on coins until the eleventh century.25

York was used as a centre by Danish, and in the early tenth century by Norse, kings. Recent excavations there have produced some objects that are certainly imports from Scandinavia, or the Scandinavian-settled Shetlands, such as soapstone vessels; amber, used for beads, could have come from the Baltic. A few objects show a markedly more Scandinavian taste, wherever made, such as a strap-end with Borre-style interlace ornament. Scandinavian trade contacts are indicated by fragments of silk, a cowrie shell from the Red Sea or the Gulf of Aden, and a forgery of an Islamic coin, all probably brought to York from the eastern Mediterranean and beyond, first overland to the Baltic, and then across the North Sea via such places as Hedeby at the base of the Jutland peninsula. To the later ninth century are ascribed the first signs of regular activity at Coppergate (5, 3), a site outside the Roman fort area, but burials in places that were neither then nor later churchyards indicate both some instability of property use and alien customs. There is evidence for a range of crafts, including glass-working, and also of the availability of a new type of pottery, known as ‘York ware’ although its actual place of manufacture is not yet known. Broad-flan pennies were minted in York from the middle of the 890s onwards. The probability is of considerable changes and expansion at York and in its trade, with a Scandinavian presence larger than that just of kings and jarls. But there are many types of Scandinavian object as yet undiscovered in York, which would be expected if a new cultural tradition had arrived en bloc.²⁶

Other places in the Danelaw were also developing in the late ninth and early tenth centuries. Norwich seems to have expanded from a complex of scattered sites along the River Wensum (4, 2). A bank and ditch on the north bank perhaps created a small enclosure, and recent excavations have found much evidence of eighth-century activity there.²⁷ The new developments took place mainly on the south bank, however, as though new initiatives were taking place. Thetford presents a rather similar picture, a little Ipswich ware and a ‘styca’ showing use of the north bank of the River Thet early in the ninth century, but with the major expansion being the in the tenth on the south side.²⁸ Concentric ditches at Stamford indicate a small enclosure on the river there, dated to the ninth century by a coin of King Alfred, whereas most of the town developed from the tenth.²⁹ At Colchester, however, where there is a certain amount of eighth-century evidence, there is no sign of ninth- or early tenth-century resurgence although it was clearly a Viking centre in 917 when captured by the English.³⁰ It would have been the logical place for Guthrum to use as a mint for the pennies which he issued in the 880s, but he seems to have preferred London even though it was on the border of his territory. At Lincoln, coins were being minted from the 890s, and a penny of that decade at the

4, 2. Plan of Norwich, Norfolk, in the tenth to eleventh centuries, prepared by B.S. Ayers. On the north side of the River Wensum, at least part of the occupation area was enclosed by a ditch and bank. Some of the earliest, eighth-century, material has come from this area, near the river. On the south side, the open area in the north-east marks its later importance as a waterfront market zone beside a major road crossing (whether by bridge or by ford). The large number of small churches emphasises the rather scattered spread of settlement (cf. 6, 3).

Flaxengate site helps to date the limestone-cobbled road and buildings of the first phase of intensive post-Roman use.31 Actual Scandinavian activity is less recognisable at these places than at York, although Thetford, Norwich and Lincoln have produced fragments of soapstone vessels, those in Lincoln all being in the earliest levels and interpreted as treasured possessions of the first settlers. At Lincoln also can be recognised a taste for amber and jet, shared with York, and there are many hones from Norway, which now became a standard trade item throughout the Middle Ages. Reflecting political rather than social trends, early tenth-century coins minted in Lincoln bore Scandinavian designs, unlike those of the late ninth century which had copied King Alfred's pennies. Silks and sherds of Islamic and even Chinese pottery show long-distance contacts. But querns and pottery show commerce with the Rhineland also, a trade that need owe nothing to Scandinavian influence. Nor was all this development in the eastern Danelaw, for in the north-west a cellared building at Chester shows activity between the Roman fort and the River Dee: Chester was another place where a mint started to operate in the 890s.³²

These developments were not confined to the Danelaw, however. It seems to have been in the 870s that occupation within the City of London reappeared, probably replacing the Strand area which thereafter became known as ‘old wic’ —hence Aldwych; the Roman walled area has produced two 870s coins and a coin-weight of 870–80, evidence that fits well enough with the Chronicle's statement that King Alfred took possession of London in 886. Refurbishment of at least part of the Thames waterfront was taking place from the late ninth century, with logs and planks forming a ‘hard’ onto which boats could be hauled, and rows of pointed stakes which may both have prevented erosion and offered defensive protection. These timbers have yielded radiocarbon and dendrochronological dates congruent with charter grants made by King Alfred, and perhaps also with the new grid system of streets that appears along the Thames frontage, where further north in the City the street lines could be earlier, associated with the churches and the enclosures referred to as hagas in charters.³³ ‘Dark-earth’ layers or organic build-up cease to occur, as they do in other former Roman walled towns such as Gloucester, where the late ninth-century mint provides a context for the excavation evidence of renewed occupation, much of it the residues of manure from stables: a large quantity of iron tools may have been a farrier's kit-bag.³⁴ Regeneration in the ninth century is not easy to distinguish from tenth-century developments, except for the establishment of mints, and even those can be uncertain. No Oxford-minted coin of Alfred's reign is known, but that a mint did exist there seems to be indicated by coins inscribed ‘Ohsnaforda’, a name probably copied from genuine Oxford contemporaries and slightly blundered. Minting in Northampton is also possible but unproven, since no fewer than nine St Edmund Memorial coins of the turn of the century have been found in the town. There is no other evidence of moneying there until much later in the tenth century.

4, 3. Plan of Winchester, Hampshire, in the tenth to eleventh centuries, prepared by M. Biddle. Here the Roman walls confined occupation more tightly than in Norwich (4, 3), with suburbs developing along the major route-ways. Over a quarter of the town was already taken up by ecclesiastical and royal enclaves. Streams were directed through these to provide a water supply, and they also turned the mills which are recorded in Winchester's excellent documentary sources. The same factor was to cause the cloth-working and tanning industries to cluster in the north-east segment of the city.

In Wessex there were also new developments in the second half of the ninth century. Excavation in Winchester, another place which first certainly became a mint then, has shown that a new street pattern was established, the regularity of its grid layout and its conformity to a four-pole system of measurement, and the similarities to each other of the earliest street surfaces combining to suggest that it was a planned development of a single period (4, 3). Chichester has hand-made pottery attributed to the ninth century, a date supported by at least one coin find; large numbers of pits and traces of building show its more intensive use. At Exeter and Bath, however, pottery is cautiously dated to the tenth rather than to the later ninth century. In Kent, there is some evidence at Canterbury of ninth-century pottery, but it does not suggest as much activity as previously: extra-mural areas went out of use towards the end of the century, suggesting that there was no quick recovery from the Viking problems.³⁵

If places like these do not have mints or clear evidence of streets it is difficult to classify them as towns in the ninth century, even if they were to become towns thereafter. Portchester (2, 4) has produced as much evidence of ninth-century pottery and artefacts as Chichester, for instance, yet it is never considered to have been urban.³⁶ Portchester is one of the names that occur in the Burghal Hidage, a document prepared either in the reign of King Alfred or of his son, which gave a list of defended places in Wessex and the number of hides attributable to each which were to provide men for its maintenance and defence. There are different texts of the Burghal Hidage, and attendant problems of identification, but its unique interest is that it shows a planned defensive system, and names sites that were all meant to be available for use at a single time—though for how long, if ever, it was operational is unknown.³⁷ The defences at these places were not all constructed at the same time. Some, like Chisbury in Wiltshire, were hill-forts where it is not known how much if any work was done on the Iron Age ramparts to re-create a usable camp. Others, like Portchester, were walled Roman sites which probably required little attention; excavations here have shown both earlier and later occupation within the fort, and a phase in the area excavated when all structures were removed from it, perhaps to clear it for action. Some defences were neither prehistoric nor Roman, like those at Wareham where, however, the bank and ditches need not have been dug at the time that the Burghal Hidage was drawn up; the place is described by Asser in terms that suggest that it was an existing castellum when taken by a Viking army in 877; there was certainly an important church there. The walls at Winchester had received attention before the ninth century, the Roman gate on the south side having been blocked. It is not possible to recognise any refurbishment of the stonework that can be associated with the ninth century but the two parallel ditches outside the west wall certainly seem to have been dug then, for their upcast has been found and pottery in layers overlying it dates from the end of that century.

Places outside Wessex were also receiving defences, or having their existing works refurbished. The best-explored sequence is at Hereford (4, 4) where a gravel rampart overlying buildings on the west side was replaced by a turf- and-clay, timber-faced bank which seems likely to have been built to make a circuit round the river crossing: a mortared stone wall on top of their bank seems to be of the late tenth or eleventh century but there is no clear dating for the earlier phases because of the lack of associated pottery. They could be ninth-century, but there is no certainty38 More often it is ditches rather than banks that have been found, and it is their in-filling not their original excavation that can be dated. In a number of cases, the only physical evidence of early defensive circuits is streets that are claimed to be aligned on them, as in Northampton where no ditch or bank of the period has actually been seen. It cannot of course be assumed that every bank and every ditch is an English construction; the Repton ditch seems to confirm documentary records such as Asser's about Reading that the Vikings used earthwork defences too. The small enclosure found at Stamford has led to the suggestion that it could be Viking also, as could those at the other Danelaw centres. The King's Gate area at York could have had something of the same sort around it, an inner core within the larger defended area; it is not known when the existing Roman fort walls on the south and east sides were demolished, and an extension taken out to the rivers.

The actual construction of bank-and-ditch enclosures like that at Wareham would have been neither difficult nor particularly time-consuming. At Wallingford, now in Oxfordshire (4, 5), the partly visible ditch which ran round three sides of the approximately 6,250 foot-long defence (the fourth side is the River Thames, apart perhaps from a small bridge-head on the opposite bank) has been shown by excavation to be about twenty-five feet deep and eight feet wide. There is no trace of a second ditch.³⁹ Wallingford had 2,400 hides attributed to it, a figure equalled only by Winchester. If each hide had to send one man to work there for forty days, the army-service period recorded in Alfred's reign, each would have had to dig an average daily depth of about a foot from an area of about 130 square feet, hardly a crippling stint! The upcast had to be used to build the bank, of course, and there would have been timber to cut and collect. But a large team of men could easily achieve such a task in a single season; the difficult part would be to assemble them and to keep them supplied, and it is that which is a better indication of administrative ability than the mere size of the defences. Those responsible were no mean organizers and can validly be assumed to have had considerable powers of coercion— exactly as would be expected from the charters.⁴⁰

The extent to which this authoritarian structure may also be seen in the street layout in some of these centres is more difficult to judge. Even if it is accepted that deliberate planning lay behind Winchester's new roads, the applicability of that model elsewhere is uncertain, for other places do not have dating evidence. The four-pole unit of measurement can be applied to various towns, but perhaps because it remained a standard unit for some centuries.⁴¹ Even in Wessex, other towns do not have plans quite like Winchester's, with back streets close and parallel to the High Street on both sides (4, 3); nor are roads inside the perimeter of the walls invariable. The street grids are more regular than those in, for instance, the Coppergate area of York, but the extent to which contour lines and streams played a part determining topography needs to be taken into account. There does, however, seem to have been some concept in the south of what a town should look like, and a structure was imposed upon it. Direct physical imposition is shown at Hereford, where the new bank was laid out over the site of timber buildings (4, 4); whether this was done regardless of someone's property rights is unknown. The grid at Winchester probably respected the boundaries of the haga associated with the stone structure on the cemetery site at Lower Brook Street; at any rate the building itself survived to become a church. It is likely enough that agreements were reached, even if pressure to reach them was applied; Edward the Elder paid money for the properties which he needed to provide space for Winchester's New Minster, so such procedures existed then just as charters show that they had done earlier in London and Canterbury. Similar processes may sometimes have been needed to produce nucleated villages and new field systems in the countryside.

4, 4. Isometric projection drawing of the phases of occupation and defence found in excavation by R.Shoesmith and P.Rahtz on the western side of Hereford. Grain-driers, reusing some Roman stones, suggest large-scale agricultural processing, succeeded by a timber building, which in turn was overlain by a sequence of ramparts.

4, 5. Wallingford, Oxfordshire (Berkshire until 1974). The trees mark the outer perimeter line of the area defended by an earth bank and ditch, the River Thames being the fourth side. Unlike Winchester (4, 3), Wallingford has no Roman antecedents, but in the Burghal Hidage it rated the same size of garrison. In the north-east quarter, the Norman castle was extended across the town walls in the thirteenth century, blocking the north gate and causing the road to be diverted to the west. Much of the open area in the north-west was occupied by a medieval priory, but that in the south-west was never built over, an indication of lack of pressure on internal space despite considerable prosperity, particularly in the twelfth century.

As in the previous two centuries, so in the later ninth century it is difficult to establish the degree of involvement in these places of royal, rather than Church, control and interest. Nearly all can be shown to have developed from a nucleus, if not within the town, then very near by. The ownership of the Northampton ‘palace’ is not known, and might have been a property of the church of St Peter adjacent to it. Gloucester's revival could owe as much to the interest of its St Peter's church as to its extra-mural palace, Oxford to its St Frideswide's rather than to the royal palace of Headington outside its defences. Towns that were the centres of bishops’ sees were particularly likely to be a focus of Church interest. The charter given to the bishop of Worcester by the governors of Mercia, allowing him rights that included a market, shows clearly the interest of at least one churchman in the trading possibilities of his property. At present, there is very little evidence from excavations in Worcester to clarify the late Saxon period there. At Hereford, the changes to the west of the cathedral may have affected the bishop's rather than the king's property. Winchester's new street pattern may have been the bishop's plan: certainly Bishop Swithun was later credited with the construction in the 850s or 860s of a stone bridge over the River Itchen to take the principal axis road out of the town.⁴²

4, 6. The Alfred Jewel (see also cover photograph). Found in North Petherton, Somerset, in 1693, this gold, rock crystal and enamel object has been associated since its discovery with King Alfred the Great of Wessex (871–99) because of its inscription Aelfred mec heht gewyrcan— ‘Alfred ordered me to be made’ —although the royal title is not used. Various suggestions have been made about the identification of the enamelled figure: Christ personifying wisdom is one attractive possibility. The beast's head at the end holds a short nozzle, in which a gold rivet possibly held in place a wooden or ivory rod for pointing at words in a manuscript. Recent work has shown that the crystal was not new and is probably a reused Roman mount—a precious object which would have been an appropriate gift to someone of the highest status. The shape of the crystal would thus have dictated the shape of the jewel as a whole. (Photograph actual size.)

The stress laid upon kingship can certainly be seen in the coinage, for the broad-flan pennies invariably carry a royal name, apart from the few minted at Canterbury for the archbishop. Similar emphasis is hinted at by royal names on objects such as a gold finger-ring from Laverstock, Wiltshire, with ‘Aethelwulf Rex’ inscribed on it, and probably the Alfred Jewel, although it does not have a title in its inscription (4, 6).⁴³ Such things were in the gift-giving tradition of the past, but they carried an explicit message and also meant that the king was not weakening his patrimony by giving away grants of land. The defences at the burhs are another manifestation of authority, but it should be remembered that not all those sites were in royal ownership in the ninth century: Portchester certainly belonged to the bishop of Winchester. The Church's sense of authority and organisation is typified by the earliest example of an English seal-die, made for Ethilwald, bishop of East Anglia from 845 to 870.⁴⁴ The rôle of the written word in administration was growing.

The ninth century can be described as one of disruption and new impetus. The extent to which the Vikings were responsible for all the many changes can be exaggerated, but equally they should not be underwritten. York's growth as a commercial centre does seem to have been more rapid than that achieved by London or Winchester, and trade contacts that stretched as far as China put northern and eastern England into a global network. Furthermore, the disintegration of the great churches, particularly in the Danelaw, led to changes in the landowning structure which gave the laity renewed ability to accrue wealth and power. The new towns would also have provided opportunities for the ambitious amongst their tenants’ children who could not see their way to obtaining land of their own. The peopling of these places suggests mobility, and already some surplus population to be absorbed; there is no evidence that Viking raids had an adverse effect upon overall levels, whatever they did to particular targets. Without documentary evidence, much of the disruption that they caused would be recognised, but not their political impact. Only Wessex survived intact as an English kingdom, but English culture mostly survived elsewhere as well, and the trends toward a political system of royal authority over a manorially-based rural economy, eventually supported by market centres, were maintained.