Danevirke: construction phases

With all the evidence measured, examined and weighed, it is now time to put raiding into the narrative of early medieval warfare, where we now know it belongs. Many saints’ lives, chronicles and histories contain references to ‘battles’, but this is possibly because decisive set-piece actions were of more significance to chroniclers than small-scale forays. Although there are examples of indecisive battles, engaging in a battle was a highly risky strategy, as one side will be defeated and the leader could even be killed; raiding carried less chance of a catastrophic defeat, so was probably more widespread. There are clear references to raids in early medieval sources like The Anglo-Saxon Chronicle and many ‘battles’ were possibly merely successful raids. As most early medieval armies were relatively small, raiding would be within their capabilities, but mass invasion probably would not. While it is impossible to quantify the amount of raiding, even on a small scale raiding could have a widespread psychological impact (the fear of something can often have as much effect on people as the likelihood of it occurring). The dykes are evidence that some people decided to do something about the periodic raiding

The Military Nature of Dykes

While dykes may be the only solid evidence of early medieval warfare that we have in the landscape prior to the building of burhs in the ninth century, can we really be sure they relate to early medieval raiding? Although some dykes were mere boundary markers (Bwlch yr Afan, Clawdd Seri, Aelfrith’s Dyke and Bica’s Dyke, for example) most early medieval dykes look like countermeasures against raiding. A few of the longer ones may have been multifunctional, in that they countered raids as well as promoting the power of a king while bonding his kingdom together (Offa’s Dyke, Wat’s Dyke and possibly the two Wansdykes, for example).

Despite good evidence that dykes countered raids, some studies still dismiss the idea, so let us briefly recap the evidence. One of our few eyewitnesses from this period, Gildas, does say that the Britons constructed walls to scare off enemies and protect people. We have seen that some early medieval Welsh poems associate dykes with fighting. It is noticeable, for example, that when the ditches of prehistoric dykes in Norfolk were recut in the early medieval period (Bichamditch, Launditch and the Devil’s Ditch at Garboldisham are possible examples), the inner face of the ditch was near vertical and the outer side flatter. This would accentuate the face of the earthwork and might have drawn people into a killing zone. There is abundant archaeological evidence of both weaponry and bodies that have suffered injuries at the dykes (beheadings at Bokerley Dyke and Bran Ditch, a battle cemetery at Heronbridge, odd weapons from the Devil’s Ditch in Cambridgeshire, skeletons of men ‘slain in battle’ at Bedwyn Dyke and so forth). We cannot dismiss all of these finds as later execution sites or disturbed furnished graves: archaeological evidence clearly suggests that dykes were places associated with violence. If the slots found in the ditches of a least four dykes were ankle-breakers, they suggest that the earthworks were designed to repel and injure those who tried to cross them.

The scale of the banks/ditches is suggestive of military structures, especially as most give good views vital to defenders of a military feature. Most face downhill, which makes them much harder to storm but more difficult to build – on sloping ground the easiest way to construct a simple delimiting mark in the landscape is to throw the soil from the ditch downhill. The dykes often end at features like marshes, ravines, estuaries or rivers, which would hinder any attempt to outflank them; sometimes the ends curve away, so they look longer than they are. We have seen that written sources like law codes, chronicles, charters, poetry and saints’ lives all suggest this was an age of raids and warfare – the poem Y Gododdin, for example, describes a raid that was defeated, with part of the fighting happening at a dyke. There may be no battles recorded at Wansdyke, but there are battles recorded in the vicinity, including two at the barrow that possibly gave its name to the dyke. The written evidence, the physical evidence and the lack of credible alternative explanations confirm that many dykes had a military purpose.

These dykes are deliberately sited to intercept raiders. As well as lying across the path of modern roads, as we have seen there is charter evidence that numerous dykes cut routes in the Anglo-Saxon period. Charters tell us that herepaths, or army paths (routes commonly used by raiders or invaders), were cut by Wansdyke (S 711 and S 735) and Bury’s Bank (S 500). The East Hampshire dykes (especially the Froxfield earthworks) cut access along vegetation-free stony valleys while their flanks are guarded by thickly wooded clay lands. Many of the dykes in Glamorganshire seem to block routes along ridges that give access to the lowlands to the south.

The struggle against violence, in particular small-scale raids often involving cattle rustling, is a clear theme of all early medieval law codes. The collapse of the Roman Empire ended the use of professional armies in much of Europe and the militarization of the civilian population. The spears found in Anglo-Saxon graves may have had a symbolic meaning, but probably also signify a society where the need for personal protection was a daily concern. Farmers may have had good reason to fear the raids of heavily armed warriors. As very small groups of people could have built most early medieval earthworks, perhaps humdrum rural communities or groups of villages constructed dykes to deter or repel raids.

The lack of more explicit direct written evidence for dykes as defences against raiders is perhaps understandable in an age when few sources survive. Early medieval sources tend to laud victories (or heroic defeats), so as dykes were defensive rather than offensive, perhaps early medieval writers would not think farmers protecting their cattle to be worthy of record. If some dykes worked successfully as a deterrent, there may have been no fighting to record or bodies to bury; there are numerous forts and pillboxes across Britain designed to repel invasions that never materialized.

Surely raiders could simply have gone around the dykes? The answer is no, as most would have been incredibly hard to circumvent. The southern end of Giant’s Grave, for example, is at a steep gully while there is a bog to the north, and both ends of the Lower Short Ditch are at steep gullies. Many dykes are in groups; circumnavigating one would just mean an invader faced yet another. No raider could simply go around Dane’s Dyke or the Cornish dykes, as the sea or estuaries were the termini of these earthworks. The Giant’s Hedge, for example, terminates below the lowest fordable point of the estuaries at either end.

The ends of many dykes were probably guarded by woodlands. Although medieval forests were more open than modern woods, as large mammals like deer (more numerous in medieval times) would keep undergrowth clear, navigating through any wood (or marsh) in good order is not easy. To a historian with an Ordnance Survey map it is obvious how to circumnavigate a dyke, but if early medieval invaders approached even a very short dyke where trees, marsh or a rise in the ground obscured the ends, they would not know how to go around it without sending out patrols. Even if a raider could go around a dyke, this would cause delay and possibly involve the splitting up of the invading force to reconnoitre a route. When looking for easy pickings, raiders would probably go elsewhere.

The best example of dykes cutting routeways is probably the Cambridgeshire Dykes, which seem to block access to East Anglia along the Icknield Way. They lie across a narrow band of chalk about 3.1 miles (5km) wide, which runs south-west–north-east, and flanked by what were then fens on the north-west side and what is thought to have been ancient woodland on chalky boulder clay to the south-east. An enemy who successfully circumvented one of the earthworks would then be faced with the problem of getting past the next.

Previous studies have often failed to see the significance of these huge earthworks in the story of early medieval warfare. We now need to look at raiding and warfare in detail, fitting these earthworks into the narrative.


There is evidence that there were large early medieval armies numbering in their thousands, like the mighty Viking army that invaded England in 866. These large armies led to big set-piece battles like Stamford Bridge and Hastings in 1066, but before 850, smaller-scale conflict was probably the norm. Engaging in battle is a highly risky strategy, as one side will be defeated and in extreme cases the leader might be killed or the kingdom could even collapse; It seems counter-intuitive, but most casualties occur in the aftermath of a battle when one side is in flight; a narrow defeat on the battlefield could lead to wholesale massacre and according to an early eleventh-century sermon, one Viking raider could make ten Anglo-Saxons defenders flee. Small-scale raiding that carried less chance of a catastrophic defeat was probably more widespread and more likely to be within the capability of early medieval leaders.

People were not constantly attacking their neighbours in the early medieval period and there were mechanisms in place to prevent uncontrolled violence. Raiding, though, did occur and such low-intensity conflict (or at least the fear of it) was probably widespread enough to be a major stimulant in the construction of most dykes. Perhaps by using evidence from early medieval Britain and elsewhere we can recreate the mechanics of a typical raid, then discuss how a dyke could counter such a threat. The period under study was one that saw fundamental changes (Britain in ad400 was very different from the situation in ad850), but as we cannot accurately date the dykes, the following scenarios are broadly based on evidence appropriate to the probable peak of dyke building in the late sixth and early seventh centuries.

The collapse of the Roman Empire brought to an end the use of professional armies in much of Europe and the militarization of the civilian population. While farmers could have attacked their neighbours, they were probably usually too busy producing food to do so. Warriors would be more likely to carry out raids, although there was probably no clear division between the two classes for much of this period. Viking sagas suggest that while some made their living purely from raiding, others supplemented ways of feeding their family (farming, trading or a craft) with a bit of seasonal raiding. The leaders of raiding war bands could have been kings, or, especially in the early stages of the period, merely successful warriors; as well as choosing warriors from among their kin, the most successful leaders would attract warriors from other communities. Those who made their living from war would become well armed with shields, swords, helmets and possibly even chain mail. Although rulers did have a band of loyal warriors, thegns in Anglo-Saxon times, who were handy with a sword, many people who fought in early medieval battles or raids may have made their living from the soil. Warfare became more professional in the later medieval period, but even well-organized kingdoms like late Anglo-Saxon England would call on local farmers to make up the bulk of their army.

How people prepared for a raid is a matter of speculation, but perhaps poetry can give us some clues. A leader would gather warriors, choose a target and attack swiftly before the victims could organize their defences. Before embarking, oaths of loyalty were probably sworn and the night before we can imagine the warriors boasting about how brave they would be, alcohol possibly helping to exaggerate their ardour. In the early morning, weapons would be checked and sharpened while promises were made about how the booty was to be divided. They would mount horses and set off in the direction of their intended target. There are numerous references in Beowulf to of all these activities, for example when Beowulf prepares to meet Grendel’s mother. We do not know if a reconnaissance was made prior to an attack; if a spy was spotted the enemy would be forewarned, so perhaps scouts were not used. The disastrous outcomes of raids like that recorded in Y Gododdin suggest that intelligence was not always obtained.

The quickest and easiest way to travel to war would be on horseback. Without detailed maps of neighbouring kingdoms, raiders would probably use Roman roads and ancient ridgeways to penetrate deep into enemy territory without the fear of getting lost or making unnecessary deviations. It is noticeable that along many Roman roads, villages with names of an Anglo-Saxon derivation are located a few miles away rather than on the road; if you drive along the nearest Roman road to where I live there are no villages on the road for about 20 miles (32km). This suggests that raiders did not stray far from these routes, possibly out of fear of ambush or losing their way. It is perhaps significant that the Anglo-Saxon word rád not only meant ‘to go riding on a horse’, but also ‘to go raiding’ and ‘a road’.

As we have seen, in other cultures the ideal raid would be one that met no resistance, or failing that one which swiftly overcame any defenders. Raiders would try to make the enemy break and run (as we have said, most casualties in battle occurred when one side was in flight), but if this was not quickly achieved the attackers might beat a hasty retreat. If raiders targeted farms, the defenders would be local peasants, or ceorls, armed possibly with spears and shields as well as the improvised weapons normally used as tools, such as axes, knives or hunting bows. If raiders were confronted by armed enemies, an exchange of missiles would probably occur before handheld weapons were used at closer quarters. If the raiders targeted religious sites, their opposition would have been unarmed priests or monks. They might target the ruler of a neighbouring kingdom, hoping to catch him with only a few members of his entourage to defend him.

While the Anglo-Saxons travelled to war on horses, it is uncertain whether they fought on horseback. They did not have purpose-bred warhorses, nor iron horseshoes that could be nailed to the hooves to protect them on stony ground. They may not have had the stirrup, which is essential when using a horse as a fighting platform. The Anglo-Saxons did pursue a fleeing enemy on horseback, often for many hours after a battle, though during a raid a quick getaway was probably more advantageous than chasing after an enemy. After the raid, the attackers would gather up their stolen goods and head back home along the most direct route (probably a ridgeway or a Roman road), then spend the evening feasting, boasting and drinking in their hall. Raids ignited vendettas that triggered revenge attacks and a cycle of counter raiding; when kings emerged, they tried to curb this partly through the use of written law codes.

A raid could have various objectives: to demoralize an enemy; to reduce their ability to fight back; and to obtain booty. If raiders did try to ambush and kill the leader of a neighbouring kingdom (as happened in Wessex in 755 when Cynewulf was murdered), this might explain the large number of kings recorded as being killed in early medieval sources. The stolen goods could be cattle that raiders could herd back to their own community. The burning down of their victims’ farms and food stores would reduce their strength and ability to strike back. The raiders could take slaves (as in the case of Saint Patrick) and high-value goods (such as jewellery); the leader of the raid could use such goods to reward his followers. This largesse would attract warriors to the victor, while the victims might turn on their leaders for failing to protect them. If rape was involved (and Anglo-Saxon sources do suggest it was prevalent in periods of instability), this would further burden the raided area with unwanted young mouths to feed, who might be looked upon with suspicion as their fathers would be enemies. Finds of female brooches made from reused British and Irish metalwork in Viking-age Scandinavia has prompted the suggestion that raiding was also carried out to obtain a bride price, that is, a dowry necessary for a young man to marry. The wealthy and well educated who were less able to fight (such as priests, teachers, lawyers and poets) would flee a community suffering raids, which would further destroy its culture. In numerous kingdoms of early medieval Britain rival factions or branches of royal families fought over control of the realm; perhaps the different groups targeted areas controlled by their rivals to weaken their power.

There are earthworks near Bokerley Dyke in Dorset that confirm that cattle raiding in particular was a real problem in the early medieval period (such raids are a theme of early medieval Irish legends). To the east of Bokerley Dyke (and therefore unprotected by it) are two giant cattle enclosures (soil samples from inside the banks confirm the presence of large amounts of dung). The first, Soldier’s Ring, is a 10.5-hectare polygon enclosure surrounded by double banks built near the end of Roman rule, while the other enclosure (39 hectares in size) is 3.1 miles (5km) further east at Rockbourne and overlays Roman fields. These earthworks reflect the widespread move across Britain from arable to pasture in the late and immediate post-Roman period, when the new cattle ranchers needed enclosures to protect their cattle from raiders. Cattle raids probably became such a problem that the locals decided to build the nearby Bokerley Dyke to try to control it.

Weapons Used in Warfare

As well as the finds from continental bogs already mentioned, like those from Esjbøl-North in Denmark, we have evidence from England of what weapons were used. Until their conversion to Christianity in the eighth century, the Anglo-Saxons often buried their dead with objects that symbolized their status; in half of male graves, this meant a weapon. Although I have noted above that some of these weapons may have been symbolic rather than what the dead person used in life, most look capable of causing damage in battle. We cannot know if the proportions of different types of weapons found in graves are typical of what was carried in life. Though bows and arrows are under-represented in the archaeological record, written evidence like the Battle of Maldon poem suggests they were used in battle. We have no reason to assume that the proportions used in Anglo-Saxon England were significantly different from what was found in continental bogs.

Most Anglo-Saxon furnished burials contained a spear (some lighter ones designed for throwing, while others with longer and heavier blades were undoubtedly handheld weapons), nearly half contained shields, 11 per cent swords, and a few knives or axes that may have been tools as well as weapons. Helmets and chain mail are rare. Anglo-Saxon swords were often pattern-welded, that is, bars of iron were twisted together then hammered flat into a blade, giving a surface which, if carefully polished, looks to me like metallic snakeskin. The later Vikings had better steel, so would use a single piece of metal. The reference to the sword breaking during combat when Beowulf fought the dragon may explain why some were buried with multiple weapons, as it would have been advantageous to have a back-up in such circumstances.

While it is possible that people were more likely to bury objects that were easier to replace, it does seem that a spear and shield formed the weapon combination of most people in this period. For the Picts, Scots and the Britons of Wales, Cornwall, Cumbria and Scotland there are far fewer finds to work on and less surviving literature, but it is likely that they used similar equipment. The Aberlemno stone in Scotland does show Pictish warriors using spears and shields, while the British writer Gildas does make reference to swords and spears being used in battle. If they had fought in a very different style using vastly different weapons from the Anglo-Saxons, early medieval authors like Gildas and Bede, who were keen to stress the differences between the nations, would probably have noted it.

Archaeological evidence of weapon injuries that caused skeletal trauma demonstrates the effects of weapons. The seventh-/eighth-century bodies from Eccles in Kent and Heronbridge in Cheshire suggest that warriors hacked down with blows to the head from a heavy sword. The damage found to skulls at these sites confirms the evidence from furnished burials that helmets were a rarity. As has already been discussed, an Anglo-Saxon sword’s balance point is halfway down the blade; Viking swords were made from better quality steel, were lighter and had a balance point nearer the hilt. The former was designed for hacking at the upper body, the latter for thrust and parry. Perhaps early Anglo-Saxon warriors expected to attack poorly armed victims, while later Vikings often faced foes also armed with a sword. Early Anglo-Saxon warriors do seem to be partial to using the weight of their weapon to hack down their opponent, targeting the head. Raising a defender up on a dyke makes the attacker’s sword far less effective. Spear damage is less easy to detect than heavy blows to the head (especially if bones are not struck), but on display in the gallery in The Collection museum in Lincoln, where the author worked, there is on display a tibia with a spearhead embedded in it that would have caused the victim to bleed to death.


How a Dyke could Counter Raiding

In 1959, Lieutenant-Colonel Alfred Burne published a paper arguing that Offa’s Dyke was more likely to be a military structure than an agreed frontier as Fox had suggested. Burne suggested that the reason why there were English settlements west of the dyke was that as an unmanned defensive structure, the Mercians needed sufficient warning to man the earthwork during a Welsh attack, while the other dykes of the Welsh borders were forward and rear lines of an integrated defensive system. Burne was not an archaeologist, but he was a respected military historian who knew about warfare. If we accept that raiding characterized early medieval warfare, we must examine how a dyke could prevent enemy forays entering the heart of a community.

Dykes could probably function in various ways in a society subject to raiding or the fear of raids. Finding a hostile raiding party in your territory and then intercepting it before it got away would be difficult. Such a raid would need to be ambushed before it could strike, so vulnerable communities blocked routeways with dykes. Dykes could deter attack by being so monumental in size that a potential attacker would deem the force needed to overcome the earthwork would outweigh any benefit from doing so. Dykes could also provide a fighting platform from which defenders could defeat raiders (monumental dykes that failed to deter an attack could obviously also be a place where defenders could make a stand). Communities could dig smaller dykes that were not a visible deterrent, but from which they planned to ambush attackers. Raiders resting before returning home could also dig an earthwork across a neck of land to defend a discrete area (like a peninsula) against counterattack. Similarly, defenders could make a peninsula defensible by use of an earthwork in order to have a refuge during a period when their community was under incessant attack; these last two functions potentially would give rise to very similar earthworks. If we examine these different functions in turn we can see how dykes fitted these four various scenarios.

The largest dykes are the most likely candidates for earthworks designed simply to deter attackers. The Devil’s Ditch, for example, is monumental in scale, rising up to 5m above the otherwise flat Cambridgeshire countryside. Kingdoms and communities to the west would undoubtedly have noticed such an earthwork, so while it would not have taken an attacker by surprise, it would certainly have made potential raiders think twice before attacking East Anglia. Equally, the sheer length of Offa’s Dyke meant that potential Welsh raiders must have been aware of it and they would have known that a king who could build an earthwork on such a scale was likely to have the resources to punish any attack on his territory. When the Welsh Annals record Offa devastating the British in 784, it was possibly a retaliation against Welsh raiding.

If these dykes failed to deter an attack, they could also have provided a platform to defeat raiders. Anyone who has tried to scale the Devil’s Ditch is surely painfully aware of how, if it was manned, it would be hard to overrun. Richard Muir, writing of the Devil’s Ditch, made the bizarre claim that: ‘invaders, charging in a column, would have overrun the dyke with ease. Despite the formidable appearance of the dyke, our tests showed that a fit young man could run from the outer edge of the ditch over the crest in less than half a minute.’ This is nonsense. It is very easy to lose one’s step on the bank and roll down to the bottom. As Allcroft put it: ‘Wet or dry, its smooth steep slope is as slippery as ice, so you are fain to go up on all fours, if at all.’ Even a slight blow from a defender would cause an attacker to tumble. I doubt that Muir’s runner was encumbered by shield, sword and spear. Today, the dykes are covered with grass and though we have evidence turf was used to stabilize the banks on some dykes, the ditches at least would have been muddy trenches. In damp weather, the soil would have clung to the attackers’ feet as they scrambled through the bottom of the ditch, slowing and tiring them as they tried to gain traction up the slope to where the defender lay in wait.

We can probably also dismiss Muir’s claim that the dyke would require too large a force to man, as the marvellous view from the top means that a relatively small force guarding it could easily send men along the bank to block any attempt to outflank them by raiders attacking at more than one point. Such a defensive strategy would not work so well with Offa’s Dyke, as its sheer length would surely make it easier for raiders to creep across an unguarded point. Perhaps Offa’s Dyke was deliberately set back from the border so that it could not be overrun by surprise (which is why English place names are found to the west) and some scholars have suggested that mounted guards could have patrolled it. Hill suggested that 100 mounted men in three shifts could patrol Offa’s Dyke, while beacons could summon defenders from nearby villages when they spotted Welsh raiders.

Mark Bell interestingly suggested that dykes were built by strong groups as a defence against more diffuse groups that they could not control; for example, we know that the stable kingdoms of China built walls against the Mongolian nomads to the north. This theory may explain why many dykes cluster round the powerful and aggressive kingdom of Mercia. Using this argument, perhaps we can suggest that powerful Mercian kings built the dykes of the Welsh borders to face pastoral Welshmen to the west, or perhaps powerful East Anglian rulers may have built the Cambridgeshire Dykes against diffuse groups to the west, which historians often refer to as Middle Angles (groups that were later absorbed into Mercia).

As the majority of dykes were much smaller than Offa’s or the Devil’s Ditch, they were less likely to deter attack, but communities may have built them as stop lines where raiders could be defeated (dykes like Grey Ditch in Derbyshire, Rowe Ditch in Herefordshire, Pear Wood to the north-west of London, for example). Most dykes are some way back from an actual frontier, so raiders could not easily overrun them, allowing defenders to assemble on the earthwork and plan their strategy before the attackers arrived. Many dykes cut Roman roads, or what charters tellingly refer to as herepaths or army paths, the very routes taken by raiders and invaders. The dykes in Glamorganshire, for example, seem to block ridges that give access from the uplands to the coastal plains (in an early medieval context this would mean keeping warriors from Brycheiniog out of Glywysing), often cutting ridges at narrow bottlenecks. Many of the smaller dykes required very few people to build them and some ‘rough dykes’ (which is the original meaning of the name of Rowe Ditch) were probably temporary measures thrown up at comparatively little notice. The need to build new dykes quickly to counter the threat of raids is possibly why the builders did not bother with a palisade. Raiders who successfully raided an unguarded community may have been surprised by a newly constructed earthwork blocking their progress when they had traversed the same route with ease the previous year.

We know from the complete lack of archaeological evidence of forts, watchtowers or fortified gateways that early medieval dykes were not permanently garrisoned, but, as scholars have suggested when discussing Offa’s Dyke, watchmen (like those mentioned in Beowulf or the speculators recorded by Gildas) may have patrolled them. These watchmen could then use beacons, flags, horns or messengers to warn the local people of the attack (if the raiders were burning farms as they came, the smoke may have made other warnings unnecessary). We know from later Anglo-Saxon documents that there was a system in each shire whereby the local lords could call on the services of their tenants to fight invaders. This was called the fyrd and earlier similar local organizations possibly existed across early medieval Britain. As the local men and the watchmen gathered at the earthwork, messengers could have sought out the local ruler and his warriors.

Previous scholars who have examined the dykes have not discussed the psychological influence of a dyke during a fight. It is likely that the defenders would be a group of (possibly terrified) locals lined up along the earthwork, whereas the attackers would more likely be warriors who derived their wealth and status from raiding. Morale is incredibly important in any battle; the local farmers protecting their land would be susceptible to reacting with panic and the biggest problem with inexperienced troops is that they tend to flee in the face of a determined attack. The attackers would probably travel on horseback to raid and dykes are very effective against cavalry, so would have inhibited the ability of mounted raiders to rout the defenders quickly. As the majority of casualties in battle happen when one side is in an uncontrolled flight, a ditch in front of the dyke which keeps the enemy at a safe distance would be a comfort to the defenders, while the bank would be a safety zone from which the defenders would be reluctant to flee. We know from the Battle of Hastings that manoeuvres like having the cavalry feign a retreat could draw out defenders from a secure position, upsetting defensive formations like a shield wall. Being on a dyke would discourage defenders from leaving their position – who would want to leave the security of an elevated position with good views for a flat field with hostile enemies, possibly mounted, pursuing you?

Perhaps the decapitated burials found at some dykes may have been the remains of defeated defenders who panicked or were overwhelmed by the superior numbers or skills of raiders. The defences of burhs later helped comparatively amateur defenders to see off attacks by semi-professional Viking raiders. In modern battles, most conscript soldiers never use their weapon; it is likely that in an early medieval context the defenders would probably hope that the enemy would simply go home. It probably takes four times as many troops to storm a defended position as to hold it, so the raiders would have to outnumber the attackers considerably before they dared to attack.

If the raiders decided to press on with their attack, a dyke would have many advantages for the defenders. The ditch would initially keep the raiders at a distance. Missile weapons (arrows, throwing axes and javelins) would potentially drive off an attacker without the need for the more terrifying prospects of close-quarter fighting. The defenders, if local farmers, would be less likely to own a shield and modern re-enactors state that it is difficult to wield a shield as well as the rather heavy Anglo-Saxon spear, so perhaps they did not form a shield wall on top of the dykes. Projectile weapons are far less effective when thrown or fired uphill, as the loss of momentum makes them easier to dodge and they will do less damage if they hit. If raiders were warriors who made their living from war, they were more likely to be armed with swords and the downward sweep of a heavy early medieval sword would be hard to manage against a defender raised up on a bank.

A dyke could also help to defeat as well as repel an attack. While the raiders, deep in hostile territory, were sustaining casualties assaulting the earthwork, messengers warning of the attack could be bringing more men to the aid of the defenders. If the defenders felt they had sufficient numbers, they could even use the earthwork to destroy the attackers, for example the men on the dyke could send out forays of more mobile and better-trained troops round the flanks of the raiders, then crush them against the dyke. The skulls found at dykes may not have been overwhelmed defenders, but the remains of a destroyed attacking force; perhaps their heads were mounted on stakes to deter other attackers. There are examples from early medieval sources across Europe that suggest displaying the body of an executed transgressor was considered normal. Individual earthworks have particular characteristics that would help to defeat an attacker. Minchinhampton Bulwarks cuts a ridgeway and the ends are located where the land slopes down to a valley; rather than being straight, the ends curve forward to form a reverse ‘C’ shape, effectively drawing raiders into the centre where the defenders will outflank them. Bokerley Dyke cuts a Roman road, but runs parallel to it for some distance, so that it can be used as a missile platform against attackers approaching from the north.

If the raiders sensibly avoided a frontal assault, even so outflanking the dyke may not have been as easy as modern fieldwork sometimes suggests. With a good map and stout boots on a peaceful sunny day it may be relatively simple to hike round the end, but it would not be so easy during a raid. As most early medieval dykes faced downhill and pollen or other environmental evidence suggest they ran across open country, defenders would have had a good panorama of the ground in front. This would not only give them good warning of attack, but during the fighting it would allow them to react quickly to any outflanking manoeuvre by the attackers. If the raiders tried to go around one dyke and rejoined the road, they may have faced another dyke, as many are found in groups, like those on Crookham Common. As the ends of many dykes are not obvious, a raider either had to send out patrols or guess which way was the best route round the earthwork. If a dyke deflected raiders off the herepath they were following so that they had to ford a large river or cross a deep gully, this would provide an ideal opportunity for an ambush. If a raider entered the wood or marsh on the flanks of a dyke, he would be entering an environment where the locals hunted and they could pick off the stragglers. As losses mounted (as raiders were wounded, killed or deserted), the leader of the raid would soon be forced to return to safety. It is likely that many raids were made at night, making it even harder to see the ends of the dyke or to navigate once the raiders had left the road.

The areas where there are no dykes are possibly ones where there was no need to build linear defences against raiders. The lack of dykes in central Mercia might be because the chronicles that record the hostile actions of Mercian kings like Offa and Penda, as well as archaeological evidence such as the Staffordshire Hoard, suggest that the Mercians did more raiding than their neighbours and it is only on their western frontier that they were being raided themselves. The border subkingdoms recorded in the Tribal Hidage that surrounded the core of the Mercian kingdom may have absorbed raids from other kingdoms. There are no dykes in north-west Wales, as mountains and tidal rivers that are hard to ford block access into the kingdom of Gwynedd. Big forests lay to the north of both Essex and Sussex (two shires devoid of early medieval dykes) and large tidal rivers or marshes cut their coastlines where defenders could ambush invaders as they attempted to ford these water obstacles. There are no dykes in the Highlands of Scotland, as there are few land routes worth cutting and most raiders would travel by boat.

The dykes built to protect headlands obviously do not block routeways, so perhaps some of these were dug as a defended beachhead by attackers on raids that required at least one overnight stop in hostile territory. We know that the Viking raiders used dykes to protect themselves from counterattack by digging earthworks across a narrow neck of land to make a safe haven, for example Coombe Bank near Reading. Perhaps Park Pale in Yorkshire is a Viking beachhead dug by Vikings attacking the kingdom of Northumbria. It is possible that Anglo-Saxon raiders constructed earthworks to protect themselves when resting during a raiding expedition, either before they settled in Britain or after. On the east coast of England Dane’s Dyke is the only earthwork that looks like a possible beachhead constructed by raiders from abroad and the twelfth-century Symeon of Durham claims that a seventh-century Anglo-Saxon king landed there. The source is unfortunately rather late, but the site is extremely well chosen and the earthwork’s massive scale makes it unlikely that it was a hurriedly built defence for a group of raiders.

There are problems with the idea that prior to the arrival of the Vikings dykes were built by raiders, but they might have been dug to defend against them. Large cliffs guard the seaward side of Dane’s Dyke and very late Roman signal towers along the coast to the north could have provided a warning of raiders, thus allowing the locals to gather their families and animals before retreating behind the dyke. Post-Roman finds and bodies with evidence of a violent death found at these signal stations support the idea that they may have acted with a fortified refuge at Dane’s Dyke to protect against raiders in the late and immediate post-Roman period. The earthwork at Heronbridge may have been a refuge or bridgehead for a marauding Northumbrian army, though it looks very well made to be a hurried defensive measure (in particular the careful reuse of Roman material in the revetment). With no gateways or signs of internal structures it does not resemble a fort, but looks like a beachhead defending a fording point. There are no dykes on the west coast of Britain that look like beachheads for Scottish or other Irish raiders; those in Cornwall are surely too long for a hastily erected defence. Therefore, though the Vikings may have used dykes as beachheads, there is little evidence that earlier raiders did, perhaps because their raids were swiftly concluded.

The Cornish dykes do not block routeways, but they do demark headlands, perhaps to defend against raiding, not from the sea but overland. In 815, The Anglo-Saxon Chronicle records the West Saxon King Egbert raiding Cornwall from east to west. The shorter dykes (like the one that delimited Stepper Point near Padstow) could have acted as refuges for large numbers of people and cattle. The larger dykes (like the Giant’s Hedge that runs between the Fowey and the Looe rivers) could be stop lines set back from a vulnerable border. The area it delimits is too large to just be a refuge, so it was probably designed to defend the core of a Cornish subkingdom, which is why it only covers a small part of the Hundred of West Wivelshire, yet it must have utilized labour from a larger area. The Cornish dykes are set back from the Tamar (and therefore the West Saxons) and beacons, for example St Agnes Beacon (a superb viewpoint enclosed by a dyke called Bolster Bank and visible from as far afield as Camborne), could have warned people to retreat behind the dyke. Similarly, at Tintagel a large ditch, probably early medieval in date, defended the settlement on the cliff-fringed peninsula from landward attack. Raiders could not outflank the Giant’s Hedge as it ended below the lowest fordable point on the Fowey and Looe estuaries. Peasant levies will often run in the face of raiders, but a dyke would give them confidence and a fixed point from which to make a stand, while raiders, always looking for easy targets, would leave a manned dyke alone. After the West Saxon raiders went home, the people could rebuild their ravaged farms.

Raiding was probably not always the norm in early medieval Cornwall. The archaeological record suggests a move away from the fortified settlements that characterized the Iron Age in Cornwall, suggesting that there must have been times of stability in the early medieval period. The dykes were temporary measures set back from the border built during times of crisis when the West Saxons threatened. As they rapidly fell out of use, locals soon forgot their names and stories of giants grew up to explain the origins of the Cornish dykes.

By not defending the Anglo-British border as the Britons of Dorset possibly had tried to do at Bokerley Dyke, the West Saxon kings could rampage along the spine of Cornwall, demonstrating their martial might while the Cornish were safe behind their dykes, having avoided being destroyed in a decisive battle. The Anglicization that occurred as Wessex expanded into Devon, Somerset and Dorset never occurred in Cornwall. Uniquely in south-west Britain, the inhabitants of Cornwall claim to possess a Celtic identity. Brythonic place names (especially in the central and western parts) abound in the Cornish entries of the Domesday Book and a Brythonic tongue was still the vernacular in western parts into the seventeenth century. The maritime links Cornwall maintained with Brittany up until the Reformation perhaps bolstered a Brythonic culture, but the south coast of Devon is as easy to reach by sea from Brittany, while north Devon and Somerset are only a short voyage from Wales. Control of Cornwall was certainly an attractive prospect to the kings of Wessex, as its mineral wealth had attracted merchants from as far away as the eastern Mediterranean from Phoenician times until the reign of the sixth-century Byzantine Emperor Justinian. The Tamar was never an impassable barrier and the English place names found in eastern Cornwall probably reflect West Saxon colonization of those parts of Cornwall nearest to Devon. The dykes may crucially have provided refuges, allowing Cornish society to weather the aggressive early stages of Anglo-Saxon expansion and so maintain their own identity.

The Anglo-Saxons: Hostages, Oaths, Treaties and Treachery I

Bargains made and broken involving the exchange of hostages and the swearing of oaths were such an important part of Anglo-Saxon warfare that scarcely an event was recorded without such an accompaniment. It is the glue that held the model of the political world together. By looking at the nature of such agreements it can be shown that the familiar tools used to cement agreements varied wildly in their effectiveness. The study of this one phenomenon alone can explain so much about Anglo-Saxon history.

There were a number of ways in which the leaders of early Medieval England could seek to cement an agreement or alliance. For Christian parties there was the baptismal sponsorship or god-parenting arrangement. Also, there was the marriage alliance, particularly effective if the leader in question had several available beautiful sisters at his disposal, as did King Athelstan at the beginning of his reign (924–39). Athelstan was also a master of the fostering ploy as well. He fostered Haakon of Norway as part of a peace agreement made by his father. But not everyone was blessed with the power and political tools of the mighty King Athelstan. For most leaders, the keeping of an enemy to an agreement was done with the sometimes gritty and risky method of the hostage exchange.

The Anglo-Saxon era is littered with a macabre history of the fate of hostages. It is a history that makes for unpalatable reading for modern minds. The making of a binding agreement with an enemy by exchanging hostages and swearing oaths might seem fairly watertight, but all too often one side or the other (frequently the pagan Danes) treated those who they had exchanged with little regard. And so it was with the swearing of oaths. Such oaths were only worth something if they were sworn on relics or holy items that actually meant something to the oath taker.

Hostage exchanges had long been used throughout Early Medieval Europe. The Old English word for a hostage was ‘gisl’. This word is similar to the Irish ‘giall’ and to the Welsh ‘gwystl’. A hostage could be of a noble background, kept at the court of his captors to ensure the good behaviour of his own master, the captors’ enemy. They could also be members of a fighting force whose leader had been coerced to come to terms with his enemy.

The Anglo-Saxon Chronicle first mentions a major use of hostages in the period covered here in an entry for 874 after the Danes had driven out King Burgred of Mercia (852–74). Here, the scribe says:

they [the Danes] granted the kingdom of Mercia to be held by Ceolwulf, a foolish king’s thegn, and he swore them oaths and granted them hostages, that it should be ready for them whichever day they might want it, and he himself should be ready with all who would follow him, at the service of the raiding army.

Exactly who Ceolwulf’s hostages were is unknown, but it is likely they were valuable to him. These men may even have been chosen by the Danes themselves, who were holding all the cards at this time. Contemporary observers such as Asser saw this whole thing as a ‘wretched’ agreement, a comment that reveals the likely effectiveness of the arrangement.

The next mention of hostages is in 876. This time, with Alfred cornering his enemy at Wareham, the impetus was apparently with him. The Anglo-Saxon Chronicle says that the king made peace with the Danes, possibly implying a cash payment (as the chronicler Æthelweard also implies), but that the Danes ‘granted him as hostages the most distinguished men who were next to the king in the raiding army, and they swore him oaths on the sacred ring, which earlier they would not do to any nation, that they would quickly go from his kingdom . . .’. Alfred had learned from his own dealings with the Danes and from what he had heard about similar arrangements in Francia. The Danes did not care much about keeping an oath sworn on Christian relics. This time they had sworn an oath on their own holy ring, possibly an arm ring of the type associated with Thor. But if Alfred thought this was enough to bind his enemies to their agreement, he was to be mistaken. The king had given his own hostages to the Danes in what was an exchange as opposed to a one-sided agreement. Asser tells us that during one night the Danes left Wareham and killed all their hostages, breaking the treaty and headed for Exeter. We can only guess what Alfred did with the men he had received as hostages on hearing this news.

Guthrum, the Danish leader in all these negotiations, made it to Exeter with his mounted army and sat there confident that he had played a master stroke and had got himself out of a tight corner at Wareham. Alfred’s men had not spotted his night exodus, and had played catch-up to no avail arriving at the gates of Exeter when the Danes were already safely ensconced. Guthrum was also expecting a fleet of 120 ships to arrive and aid his bid, but when these vessels left Poole Harbour, rounding the headland off Swanage, they all succumbed to a storm and were lost. Consequently, Guthrum was once again on the back foot with an English army at his door. And with this development we see yet another turn in the story of the hostage exchange.

The hostage ploy seems now to have favoured Alfred. It may seem surprising that there was an exchange as opposed to a one-sided Anglo-Saxon arrangement given what had happened to the English hostages at the hands of the Danes at Wareham, but conditions were not yet perfect for Alfred, he was just in the ascendancy. This time, the Danes granted him ‘as many hostages as he wished to have’, implying that the king was able to pick them. What followed this Exeter agreement, again sworn on oath, was the departure of the Danes from Wessex and their subsequent settlement of parts of Mercia which the puppet English King Ceolwulf had held open for them.

Clearly, hostage negotiations in the ninth century were a bloody and dangerous game. Each phase in a campaign seems to have involved an upping of the stakes for both sides. Alfred’s subsequent misfortunes in the wilderness of the Somerset marshes are well documented for the year 878. However, his famous victory against the Danes at Edington was so decisive that it led to a further development in the art of the hostage negotiation. A fortnight passed with the English camped outside of the Danish camp to which their army had fled after its defeat. Starving, cold and fearful, the Danes came to the English seeking surrender on terms more onerous than ever before. They would give hostages again, just as many as the king wanted, and this time they would demand none in return. No such arrangement had ever been made before. This was as close as a Viking army of the ninth century could get to abject defeat in a campaign. It was followed by the baptism of Guthrum (now to be given the English name Athelstan) at Aller and an additional ceremony at Wedmore some weeks later. Guthrum would rise from the baptismal waters as a Christian leader in a Christian land. The Danes would indeed leave Wessex, providing Alfred with breathing space to rebuild an expanded kingdom.

It happened again in 884–5 at Rochester, but this time the mention of hostages is different. A Viking force, which had been terrorising Francia for the opening years of the 880s, came from Boulogne to try its luck in Alfred’s new kingdom. It came expecting to wreck the place. They had brought many horses with them, fine mounts from Francia. But they had brought something else, too. Their army consisted of an unknown number of Frankish hostages. Alfred cut a dashing figure by 885. His army and fortification system were organised to cope with such an event as this new invasion. The garrison at Rochester proudly withstood. Alfred came to relieve the men of the town. The Vikings, as a result, fled back to Francia without either their horses or their hostages, two vital accessories of warfare. The fate of these hostages is unknown, but they are most likely to have been truly rescued by the king of the Anglo-Saxons. For those Danes who had chosen to stay in England after their defeat at the hands of Alfred, hostages were once again exchanged, perhaps in the old style. But also in the old style, these Vikings twice broke their agreement and sent raids into the wooded heartlands of southern England. Some things, it seemed, could not change.

The agreement between Alfred and Guthrum, who was now ruler of East Anglia, was bound by a treaty referred to as the Treaty of Wedmore. It effectively divided England into a Danish and an English-controlled half. From now, to the north of Watling Street which stretched from London to Chester, there would be a land that would be under Danish-inspired law, a land that ultimately became known as the Danelaw. Alfred and his family were left with the rump of English Mercia and Wessex from which to provide the platform for a re-conquest. The surviving copy of the document that outlines the treaty probably refers to the settlement arrived at between the two sides after a period between around 880–6 when the countryside to the north of London was very much up for grabs. Here again, after much bloodshed and double dealing, we have mention of hostages:

And we all agreed on the day when the oaths were sworn that no slaves or freemen might go over to the army without permission, any more than any of theirs to us. If, however, it happens that from necessity any one of them wishes to have traffic with us–or we with them–for cattle and for goods, it is to be permitted on this condition, that hostages shall be given as a pledge of peace and as evidence whereby it is known that no fraud is intended.

So, in this new Anglo-Danish world, hostages could be given as surety against fraudulent trading activity. We can only imagine how many people were dragged across Watling Street from one side to the other to provide confidence in a trading deal.

As Alfred progressed with his grand military and ecclesiastical reforms, the years went by with no recorded deals involving hostages. It was not until the return of the Vikings in 892 that the campaigns began again in earnest. A Viking army with its 250 ships came to Appledore, and new leader Hæsten’s 80 ships came to Milton Regis threatening to cut off a giant corner of the country from the English king. Moreover, although Guthrum was now dead, the Danish-led armies of East Anglia and Northumbria were full of confidence and prepared to help their cousins assault the king of the Anglo-Saxons once again. This is why Alfred needed yet another hostage arrangement.

Alfred managed to secure oaths from the Northumbrians and the East Anglian Danes not to attack him. He procured six ‘prime’ hostages from the East Anglian Danes, although we are not told how. We are informed, however, that the deal was not kept to, as it seems the English-based Danes kept helping their cousins in the south east of the country by aiding their raiding and foraging activities during this uneasy stand-off. The fate of the ‘prime’ six is not recorded.

The Appledore Danes, laden with booty from their periodic raids, soon attempted to push north to find a ford across the Thames with the intention of making it to Essex to join with other Viking ships there. It ended in disaster for them, with Alfred’s son the ætheling Edward (soon to be king himself) intercepting and defeating them at Farnham. Edward drove them across the Thames at a place where there was no ford. Soon they found refuge of sorts on a river island near Iver in Buckinghamshire known as Thorney. Edward began a protracted siege which came to an end when the English system of military rotation meant that that besieger’s supplies had run out and their time in the field was up. Unfortunately, this all happened before King Alfred was able to relieve his son. We are told, however, by the chronicler Æthelweard that the Danes who had been surrounded and starved at Thorney Island and whose leader was wounded did exchange hostages and agree to leave the kingdom. To Essex they went, to join up with Hæsten’s force which had now relocated to Benfleet.

Alfred, at some stage during this campaign decided to attempt to bring Hæsten to heel through the offer of baptism, with the king himself standing sponsor to one of the Viking’s sons and Æthelred, Lord of the Mercians to the other. We know less about it than we do of the arrangements made with Guthrum. Hæsten’s base at Benfleet was eventually captured by the English in a siege that saw women and children, ships and money seized and brought to London. Also seized by the besieging Æthelred, Lord of the Mercians, were Hæsten’s sons. The Dane had been out on a raid when all this happened. The young men were sent to Alfred’s court, probably in the hope they would prove to be very valuable bargaining chips. But Alfred sent them straight back to Hæsten delivering a message of extraordinary mercy for these times. Perhaps it is fair to say that Alfred knew he had his enemy defeated, but he seems not to have been able to hold his own godson as a hostage.

The Vikings’ subsequent movements to Buttington and Chester both ended in sieges which saw great suffering for the enemies of Alfred. At Buttington, Æthelweard tells us that ‘they [the Danes] did not refuse hostages, they promised to leave that region’. In 894 Æthelnoth, Alfred’s trusted ealdorman from Somerset, visited York in an ambassadorial capacity with a view to persuading the Northumbrian Danes to cease their pillaging of parts of Rutland held by the English under the Alfred and Guthrum treaty, but again we do not know how this deal was cemented. What we do know is that by the time the Danes had been outwitted by Alfred’s clever fortifications surrounding their ships at Hertford in 895, the Danes, under a new leader following Hæsten’s death, moved out to Bridgnorth from where they dispersed never to reform again.

The historic record falls quiet in respect of hostages towards the end of Alfred’s reign. When Edward took the throne after the death of his father in 899, there was a revolt against him from Æthelwold, the son of Alfred’s brother Æthelred I. This revolt, which is documented below, involved the pretender to power taking a nun against the king’s leave before stealing away from the watchful eyes of Edward’s army in a dash to Northumbria. It is not known if the nun was a hostage, had been kidnapped or was a willing participant. However, to take a nun from a nunnery without the bishop or king’s permission was a ‘criminal’ offence and clearly Æthelwold was using this as a way of defying Edward.

The next reference to hostages is only by inference. In 906 Edward is said to have ‘confirmed the peace’ at Tiddingford with the Northumbrian and East Anglian Danes. No mention is made of an exchange, but from what we have observed already it is an agreement likely to have involved them. We can be more certain of the next mention, however. Edward, by 915 had succeeded in defeating the Danes and had embarked on his great re-conquest of the midlands. But in this year, a Viking force had come from Brittany led by Jarls Ohtor and Hroald (sic) and it headed up the Severn raiding as it went. The force went into Wales and took a bishop named Cameleac, Bishop in Archenfield, as hostage. Edward then paid 40 pounds to ransom him back, one of the few recorded arrangements of this kind. As the Vikings headed to Archenfield for richer pickings, they were met by the men of Hereford and Gloucester and nearby forts. The resulting battle saw Ohtor and Hroald defeated with Hroald perishing and Ohtor’s brother also being killed. The Vikings were subsequently driven to an ‘enclosure’ where they were besieged until they gave hostages in return for leaving the kingdom. It sounds as one sided as some of the later Alfredian agreements, but the reality was that the force hung around for some time on the Severn Estuary embarking on sporadic raids before eventually leaving for Ireland.

Edward’s sister, the famous Æthelflæd, sometimes called ‘the Lady of the Mercians’, had her part to play in the history of the hostage. It was she who in 916 sent a force to Brecon Mere in Wales to break down the fortification there and in doing so took the wife of the Welsh king of Brycheiniog ‘as one of thirty-four’, says the Anglo-Saxon Chronicle. There would be many more organised strategic conquests of such citadels as Danish Derby and Leicester, the key boroughs of the Danelaw. Just before her death in 918, the chronicler tells us, Æthelflæd after taking Leicester had even got the leaders of Viking York to swear oaths that they would be at her disposal. They did not keep to this, of course, but the fact that the agreement happened at all is testimony to the increasing power being displayed by Æthelflæd and her brother Edward south of the Humber. As more Scandinavians poured into the northern parts of Britain, the Danish forces of yesteryear, now settled in England, began to see the sense in aligning themselves with the resurgent Anglo-Saxon monarchy in the south. In fact, before he died in 924 Edward the Elder had secured the allegiance of the Scots, a new Norse leader at York, the lords of Bamburgh, the king of the Strathclyde Britons and the Northumbrians of all cultures.

At Æthelflæd’s death an event took place that has received a number of interpretations. The Mercian Register of the Anglo-Saxon Chronicle in an entry for 919 records it thus: ‘Here also, the daughter of Æthelred, Lord of the Mercians, was deprived of all control in Mercia, and was led into Wessex three weeks before Christmas; she was called Ælfwynn.’ The most likely interpretation of this is that it was no simple hostage taking, but a political abduction. Edward the Elder, who had achieved so much in alliance with Ælfwynn’s mother (his own sister), could not afford for the spectre of Mercian independence to raise its head just when it looked like the Kingdom of the English was finally taking shape under his own leadership. In Ælfwynn many traditional-minded Mercian noblemen may have seen a standard bearer for that independence. So, the king spirited away his own niece to avoid this happening. The Kingdom of the English was growing in confidence.

King Athelstan’s (924–39) use of human resources was a far cry from the desperate hostage exchanges of the Alfredian campaigns. A kingdom of England was very much in the making now, and Athelstan’s negotiations were on an international level, binding his kingdom into the fortunes of, among others, the great Holy Roman Empire controlled by the Ottonian dynasty. In the sparse entries for Athelstan’s reign in the Anglo-Saxon Chronicle there are hints at hostages, however. In 934 for example, Athelstan’s combined campaigns by land and sea in Scotland resulted in the Scottish King Constantine having to give his own son as a hostage.

There is no mention of hostages in the years preceding Athelstan’s great Battle of Brunanburh of 937. During these years, an imperious Athelstan, able to draw an army from the length and breadth of his new kingdom did indeed hold imperial-style councils at places like Cirencester, calling in tribute from the Welsh kings which amounted to huge quantities of gold, silver, oxen, hounds and hawks. But the Scots and the Norse-Irish led a confederacy against Athelstan at Brunanburh, at a battle site that is still to be identified with certainty. Once again, even in their crushing defeat there is no mention of hostages from the lords of the north, just a mention of them fleeing the field for home.

It would seem that Athelstan’s victories might have paved the way for a Golden Age in Anglo-Saxon kingship, but this was yet to come. Olaf Guthfrithson, the enemy who had caused the great king so much trouble, swept back into the picture after Athelstan’s death on 939. A complex marriage alliance in the midlands with the daughter of a Danish jarl, and some military campaigning in the north, saw the Dane in the ascendancy again. It was Athelstan’s younger brother, King Edmund (939–46) who had to re-capture the five Danish boroughs of the midlands for the southern English crown. In 943 a new enemy, Olaf Sihtricson, had taken Tamworth from the English and had fled to Leicester. In time-honoured style, we are told this was the same year in which this Olaf received baptism with Edmund as his sponsor. But the records of the tenth century are more complicated than those of other eras. There are allusions to further submissions in York, for example, but we are not told how these arrangements were cemented.

The Anglo-Saxons: Hostages, Oaths, Treaties and Treachery II

Battle of Stamford Bridge, 1066. The English army under King Harold Godwinson fights off an invading Norwegian force led by King Harald Hardrada.

On Edmund’s death in 946, he was succeeded by another brother Eadred (946–56), who in the first year of his reign conquered Northumbria and was granted oaths by the Scots ‘that they would do all that he wanted’. More oaths then, but no details. At Tanshelf, in 947 Archbishop Wulfstan of York and the councillors of the Northumbrians are said to have pledged themselves to the king, but ‘within a short while they belied both pledge and oaths also’. The reason for this recalcitrance lay in the arrival of a famous Viking leader, one Eric Bloodaxe whose control of the Viking kingdom of York quickly became legendary. Eric, who held Northumbria from an apparent promise made by Athelstan, was eventually defeated, murdered by one of his own men after the grim campaigning of Eadred in the area.

The brief reign of Edwy, king in Wessex (955–9) records no hostage exchanges, but plenty of domestic politicking. Edwy’s grip on power subsided as his brother the young Edgar (959–75) gained acceptance in both Mercia and Northumbria and eventually Wessex itself. Edgar’s reign over England was one of peace bought by the threat to any enemy of overwhelming military force. A huge and energetic navy and a vast land army were enough for Edgar to concentrate on matters such as monastic reform and other political issues. In an act of overwhelming symbolism the king took an army to Chester. Here, at the edge of the old Roman world, King Edgar symbolically took the helm of a rowing vessel and was rowed by up to eight subservient leaders on a boat along the River Dee to the monastery of St John the Baptist. Oaths were sworn, and the promise from all these rulers was that they would be faithful to the king of the English and support him both on land and at sea. These men were Malcolm, king of the Cumbrians; Kenneth, king of the Scots; Maccus, king of ‘Many Islands’; Dunfal (Dunmail), king of Strathclyde; Siferth; Jacob (Iago of Gwynedd); Huwal (Jacob’s nephew and enemy); and one Juchil. Edgar’s passing remark after this event would ring in the ears of regional kings for centuries. Any of his successors, he said, could pride himself on being king of the English having such subservience beneath him.

On Edgar’s death in 975, Anglo-Saxon England was at its peak of power and influence. He was succeeded by his son Edward (975–9) whose reign leaves no record of hostage giving or oath taking. However, after Edward’s infamous murder at Corfe Castle and the rise to power of his half brother Æthelred II (979–1016) we enter a period of history whereby the role of the hostage once again leaps out from the pages of the chronicles and histories.

Æthelred’s reign was a little over eighteen months old when the Scandinavian raiders returned to England. Southampton was targeted first and we are told that many of the inhabitants were killed or taken prisoner. We are not told of the fate of these hostages, however. Nor are we told what happened to the people of Thanet in Kent and the region of Cheshire who also suffered similar fates. Soon, in 991, the Battle of Maldon would be played out in Essex between the local ealdorman Byrhtnoth and an invading Viking force. Here again, there is a mention of a hostage, but he is this time a Northumbrian hostage in the East Saxon Army. Pointedly, he is mentioned as having a bow, not a sword as one might expect. It is possibly an interesting glimpse into the equipment allowed for hostages when asked to participate in their captors’ campaigns.

Hostages were once again exchanged in 994 when the legendary Olaf Tryggvason and Swein of Denmark unsuccessfully attacked London. The chronicler Æthelweard, who, as we have observed, was a nobleman himself, played his own part in this negotiation as he and the Bishop of Winchester arranged for hostages to be sent to the Danish ships and for Olaf to be brought to King Æthelred for a baptism perhaps in the style of that which had happened to Guthrum all those years ago. Olaf’s interests would soon turn to Norwegian matters, but it remains the case that after this agreement with Æthelred, he never returned to England.

The later part of Æthelred’s reign saw the usage of hostages become widespread. Nothing, however, would quite match the drama of the fate of one man in particular whose fame throughout the northern Medieval world was eclipsed only by that of Thomas à Becket 150 years later. The Danish assault on Canterbury in September 1011 marked a notorious episode in the king’s reign. The Danes, when they entered the city, are supposed to have gone wild. Their rampage resulted in the capture of Ælfhere, the Archbishop of Canterbury and one Ælfweard, a king’s reeve, among others. Christchurch was plundered and countless people murdered in a long-remembered orgy of destruction.

The Danes stayed the winter in Canterbury. They suffered from illnesses brought about by the apparent unsanitary water supplies, but this much divine intervention was not enough to rid the English of their tormentors. It was, however, their key hostage whose fate attracted the attention of chroniclers. Not content with the silver offered to them to leave, they asked also for a ransom for the return of the archbishop. Ælfhere would have none of it. For his intransigence the pious archbishop was murdered at the hands of his captors in Greenwich. In a drunken frenzy, here at the termination of the Rouen to London wine trade route, one Dane pulverised the head of the archbishop with the butt end of his axe while the others hurled cattle skulls and bones at him on the hustings. The hideous event was carried out against the wishes of an observing Dane called Thorkell the Tall and it was an event that marked the beginning of a significant switch in Thorkell’s allegiance from the side of the Dane to that of the English king himself.

From the story of one famous hostage, we go to the handing over of countless people to the Danish King Swein in 1013. Swein, who had long harboured a bitter resentment against Æthelred, had sailed down the Humber and Trent to Gainsborough and called to him the men of Lindsey, those of the Danish Five Boroughs of the Danelaw (Nottingham, Stamford, Leicester, Lincoln and Derby) and Earl Uhtred of Northumbria. Their allegiance was cemented by supplies of hostages from ‘every shire’–a figure that must have been considerable. Some time after the hostages were given, Swein entrusted them to his son Cnut. It was a move that would have profound ramifications. Swein poured his forces south across Watling Street and fell upon Oxford, the townsfolk of which provided further hostages to him. Next, it was Winchester’s turn and in an embarrassing exchange for the English king, the Oxford story repeated itself. The beleaguered Æthelred and his new ally Thorkell remained in London, contemplating. It would not be long before the king would sail to Normandy, to the home of his wife Emma and into exile.

In February 1014 Swein unexpectedly died. He was succeeded as nominal king in England by his son Cnut. However, many Englishmen appealed to the exiled Æthelred to return to England and rule once again as their natural lord, although there would be conditions in the bargain. Æthelred did indeed return and not long after this he conducted a punitive campaign of destruction in Lindsey against the land of Cnut’s supporters. Cnut then decided to sail down the east coast of England to Sandwich where he dropped off those English hostages his father had passed to him, each with their hands, ears and noses cut off before they were put to shore. Then, after this grisly act, the would-be Danish king of England set sail for Denmark.

Cnut did, of course, return to England. Through the political machinations of one famously seditious Eadric Streona, a turncoat English earl, he secured a hold on power. But in an unlikely turnaround Edmund Ironside, the son of Æthelred, found himself with the support of the Danelaw (through marriage) ranged against Cnut and Eadric based in the south of England, the traditional strong areas for the family of Edmund. With a reluctant King Æthelred lying ill in London, there began a giant circle of campaigning across the country which also involved Uhtred of Northumbria, whose grisly end we have already observed. Soon Æthelred would die in London and there began a siege and subsequent campaigns in Wessex between Cnut and Edmund which eventually ended in a pyrrhic victory for the Dane.

The situation in London, if the chronicler Thietmar of Merseburg is to be believed, was full of dramatic bargaining. Emma of Normandy, Æthelred’s widow, was to hand over her sons ‘Ethmund’ (Edward) and ‘Athelstan’ (Alfred). These were her sons and heirs to the English throne, but they are both said to have escaped in a boat. A tangled tale then ensued involving the mutilation of a great many English hostages. We shall never know the truth about the fate of the hostages or their exact numbers. Thietmar has his critics when it comes to his accounts of these important events in London.

Further sieges at London and pitched battles in the countryside around Wessex between Edmund and Cnut followed this episode. Eadric Streona subsequently changed sides and when supporting Edmund bitterly betrayed him on the battlefield at Ashingdon. The agreement finally reached between Edmund and Cnut, at an island in the River Severn after they had fought each other to a standstill up to this point, saw the English king hold Wessex and the Dane hold the rest of the country including London. Once again, the agreement was sealed with hostages. There is also a hint in the sources that Edmund offered single combat to Cnut, but nothing is really known of it.

On 30 November 1016 Edmund Ironside died. The finger of suspicion surrounding his death was historically pointed at Eadric Streona, but no absolute proof has ever been offered. Cnut in an instant took the throne of the whole of England. Throughout his reign he ruthlessly dispatched his political enemies and always seemed to be aware of the potential of the popular appeal of the line of Cerdic, Æthelred’s ancient royal family. Political assassinations meant that his grip on power became stronger and there is no evidence of any significant hostage exchanges for many years. But in 1036, after Cnut died, there was one hostage whose fate was every bit as significant as that of Archbishop Ælfhere a quarter of a century earlier. Earl Godwin, a favourite English magnate of Cnut’s, whose new allegiance was very much in the camp of the new King Harold Harefoot (1036–40), was the main player in the drama of the murder of Alfred, son of Æthelred.

For the Normans, the death of Alfred would pretty much justify the entire Norman Conquest. Alfred, half Norman himself, was on his way to meet his mother Emma. Having landed at Dover and tracked his way towards Winchester, Alfred and his followers arrived at Guildford where they were met by Godwin and his men. Alfred was carried away and taken hostage to Ely where he was brutally blinded. The others were either killed or sold into slavery. It was a horrible affair. Although the Norman chroniclers never forgave Godwin for his role in the taking of Alfred, Godwin would later atone for it when Harthacnut (1040–2)–Alfred’s kinsman–came to power. As king of England, Harthacnut made Godwin stand trial for the crime and received from the powerful earl a magnificent ship complete with its warrior complement for his compensation.

The early years of Edward the Confessor’s reign (1042–66) saw no significant usage of hostages as such until 1046, when Earl Swein, the eldest son of Earl Godwin and whose new earldom bordered southern Wales, went into that country in force. He allied himself with Gruffydd ap Llewelyn, king of Gwynedd and Powys, and was granted hostages by his southern Welsh enemies. Their fate is unrecorded. The Godwin family would, however, continue to dominate the politics of King Edward’s reign, but it was with the old man Godwin that things would come to a head between king and earl and once again hostages would play their part.

There followed royal appointments of Normans, the building of castles in Herefordshire and Dover and one infamous visit to England by Eustace of Boulogne, whose men ran amok in the streets of Dover. These were just some of the reasons for the tensions between Godwin and King Edward. By 1051 Godwin had had enough of it all. He had raised a huge army from his own Wessex combined with Earl Swein’s men from Oxfordshire, Herefordshire, Somerset and Berkshire and Earl Harold’s men (the future king) from East Anglia, Essex, Huntingdon and Cambridgeshire. At Tetbury, just 15 miles away from a concerned King Edward, this force came together. A demand was sent to the king by the earl. There will be war unless the king gave him Eustace of Boulogne and the men ‘who were in the castle’, a reference possibly alluding to either the castellans of Dover or the notorious Herefordshire Normans. But against Godwin would be ranged the forces of Earl Siward of Northumbria and Earl Leofric of Mercia, plus a contingent of knights from Earl Ralph the local Norman. Once these forces had been gathered after urgent messages were sent to the north there was something of a stand-off. It seemed to those present that a battle between the finest men of England in a time of foreign interest in the English throne would be of such grave consequence that it should not be allowed to go ahead. The situation was therefore resolved with the exchange of hostages and the arrangement that there would be another meeting on 24 September between the protagonists in London.

Godwin fell back to Wessex in the meantime and the worried king took the time he had bought to raise a huge army from Earls Leofric and Siward’s lands and bring them to London for the meeting. This took around two weeks. At London the net was closing in on Godwin and his family. Godwin asked for safe passage when he was at Southwark, but this time the hostages were refused him. The upshot of the meeting was that he was outlawed along with Swein and Harold. They had to leave England. Before they did this however, the Anglo-Saxon Chronicle records that the thegns of Harold were ‘bound over’ to the king, indicating that a shift in lordship bonds of Harold’s followers was part of the punishment. So, as a ship was prepared at Chichester Harbour the Godwin family sailed off to the court of Baldwin of Flanders, all except Harold who sailed via Bristol to Ireland. At this time it is thought that Wulfnoth, Harold’s brother, and Hakon, Earl Swein’s son, were given to the king as hostages. The way was clear for a young man from Normandy–Duke William–who had been promised the throne of England to pay a visit to Edward’s kingdom.

Godwin was no fool. On 22 June 1052 he left the Yser Estuary with a small fleet and evading Edward’s forty ships at Sandwich, he landed on the Kentish coast. His former men in Kent came to him as did the ship men of Hastings and the men of Sussex and Surrey who declared they would ‘live and die with him’. The earl’s son Harold joined forces with him and between them they gathered a force large enough to intimidate the king on his own doorstep in London. Godwin had made the most triumphant of returns. In the subsequent scramble for safety, the Norman Archbishop of Canterbury Robert of Jumièges may well have taken Wulfnoth and Hakon to Normandy as he and others fled the vengeful Godwin family, handing the hostages over to Duke William.

Godwin’s subsequent death in 1053 and the rise of his son Earl Harold set the scene for the years preceding the Conquest. Harold’s hostage-taking successes on his remarkable Welsh campaigns. Harold Godwinson’s contribution to the story of oath making, however, is legendary–if a little controversial. His journey to France in 1064 saw him unexpectedly fetch-up on the shores of Count Guy’s Ponthieu. The count could not believe his luck. It seems that Harold was on his way to William of Normandy presumably with King Edward’s promise of the English throne. However, Guy’s subsequent imprisonment of Harold before he reluctantly handed him over to William amounts to a very high-profile hostage taking. Guy almost certainly was holding out for a ransom as he presumably believed that as master of these dangerous lee shores he had such a right. But William would have none of it and took the English earl on a Breton campaign with him, holding him at court and making him his man. During this period the famous oath was extracted at Bonneville from Harold on ancient relics–a scene depicted in the Bayeux Tapestry–that Harold would support Duke William’s claim to the English throne. However, we should not forget that the kinsmen of Harold were hostages in the Norman court and the riches that Harold had taken with him on his ship may well have been to entice Wulfnoth and Hakon’s captor into releasing them, another possible motive behind the journey. Nor should we miss the contemporary historian William of Poitiers’ account that the oath swearing was not in fact as one sided as history has subsequently portrayed it. Harold had asked William that he recognise all of Harold’s land holdings in the event of the death of King Edward. We will never know what was really said or done in Bonneville that year, but the sources speak of a proposed marriage of Harold’s sister to a Norman noble and of Harold taking Agatha, William’s daughter for a bride. It is suggested by the Medieval historian Eadmer that William would allow Harold to return to England with Hakon and would release Wulfnoth once Harold had masterminded the succession of William to the throne. In the event, Harold did indeed return with Hakon, leaving Wulfnoth behind.

Harold’s subsequent elevation to the throne, his brother Tostig’s banishment and alliance with Harald Sigurdsson of Norway are the stuff of history. It all led to the invasion of Northumbria by Harald and Tostig and the opening battles of the tumultuous year of 1066. After Harald and Tostig’s victory over Earls Edwin and Morcar at Fulford Gate outside York on 20 September 1066, the Anglo-Saxon Chronicle records that York gave hostages to the two victors, with the chronicler John of Worcester noting that 150 were exchanged on either side. If this is the case, it would seem Harald Sigurdsson, despite his victory, was in a mood to do business with Northumbria in his quest for the English throne. The need for further hostages was to be satisfied by a rendezvous at the junction of roads to the east of York. History would show that here at Stamford Bridge on the crossing of the Derwent the English King Harold and his men would be ahead of the allies and achieve a famous victory, but once again the issue of the hostage exchange plays centrally in the story. After the crushing and total defeat of the Norwegians at the hands of the English, the remnants who were allowed to sail away by King Harold included Earl Paul of Orkney, who dutifully left hostages behind promising, along with the son of the Norwegian king, never to return.

William the Conqueror’s usage of the hostage was no less effective. After the victory at Hastings, he extracted hostages from the remaining English nobility at Berkhamstead during a punitive campaign conducted around southern England. The submission of Edgar the Ætheling (with whom William would have a curiously long and troubled relationship) and the notable Londoners with him did not stop the countryside from being pillaged and burnt. A new era was dawning as England began to feel the effects of the Norman style of strategic warfare.

It can be seen that the history of Anglo-Saxon warfare is very well evidenced by an account of one of its chief mechanics–that of the hostage negotiation. Deals were broken, oaths sometimes meant nothing, treaties ignored. On other occasions, the method worked very well indeed. But always somewhere in the thick of it stood a hostage. Blinded, mutilated, incarcerated or sometimes just kept at court and treated well enough, the story of the hostage is the story of Anglo-Saxon warfare. We have observed many examples of the grim fate of hostages throughout the period. One is drawn to sympathise finally with a certain Æthelwine, nephew of Earl Leofric, whose mention in Hemming’s Cartulary (a list of charters and documents compiled around the time of the Norman Conquest) includes the fact that he had lost both hands while a hostage of the Danes. If our Æthelwine could ever have dictated his story, it would not have made pleasant reading for anyone.

Dark Age Warfare


655 Battle of Winwaed: Penda of Mercia was defeated by Oswiu of Northumbria. Although the battle was said to be the most important between the early northern and southern divisions of the Anglo-Saxons in Britain, few details are available. Significantly, the battle marked the effective demise of Anglo-Saxon paganism.

Battle was a high-risk strategy. It brought matters to a decision and could save the country from the horrors of rampaging armies. On the other hand, if one lost a battle one risked losing everything, including, of course, one’s life. No quarter was given to high-status prisoners in the Dark Ages. Even kings were summarily knocked on the head. A sensible commander therefore did everything possible to avoid battle unless he was confident of winning. Battles tended to happen when two forces were more or less equally matched, or thought they were, or when the commander had run out of other options. The Dark Ages have plenty of examples of desperate measures taken to avoid battle with a superior force. King Oswy of Northumbria offered to buy off his enemy King Penda in 655. A few years before, his rival King Oswin of Deira had disbanded his army and sent them home rather than face Oswy in battle.

Once battle had become inevitable, Dark Age commanders would choose their ground carefully. When Penda refused to be bought off in 655, Oswy reduced the odds by deploying in a strong position on high ground, forcing Penda’s forces to advance through a flooded river valley. A striking number of Dark Age battles were fought by fords in rivers. Perhaps the river not only secured at least one flank but enabled the army to be supplied by boats. At Brunanburh one flank of Athelstan’s army was secured by a stream and the other by a wood. Finding a short line with secure flanks enabled a smaller army to negate the enemy’s superior numbers and create several lines of defence. Another consideration was to have somewhere to retreat if things went badly. For example, at Dyrham in 577 the British commanders probably fought in front of their hillfort, retreating behind its stout walls as they were pressed back. Not that, in this case, it did them much good.

Dark Age battle tactics are difficult to reconstruct for want of evidence. The only detailed account of a real battle is Maldon, where tactical considerations went no further than standing firm. The commander ‘bade his men make a war-hedge (wihagen) with their shields and hold it fast against the foe’. Like a hedge, the line would be long, straight and thin, and bristling with thorns – a thicket of spears. The more usual name for Maldon’s war-hedge was the shield-wall (bordweall). The line would stand to receive a charge behind overlapping shields with spearpoints projecting. The advancing enemy would see a line of wood and metal, eyes glinting between helmet and shield, and the only flesh on display being the lower legs. Breaking through this human wall would be akin to breaching the walls of a fort, and one source did indeed compare the Battle of Hastings with a siege.

Since everyone, whether Saxon, Briton or Viking, adopted shield-wall tactics in battle, the challenge was how to break through. If the commander had chosen his ground well, it would be impossible to outflank him. Sometimes, perhaps, the opposing shield-walls simply advanced towards one another and fought it out. However there is evidence that Roman tactics were familiar to Dark Age commanders through tracts such as that of Vegetius, written down in the early fifth century. As a means of breaking through, Vegetius recommended the wedge, a tactic particularly favoured by the Vikings who compared it with a charging boar and called it svinfylking or ‘swine-array’. Well-trained troops would mass in front of the shield-wall in wedge formation some ten lines deep. The wedge would then charge forward, keeping formation in order to penetrate the line with great force at a narrow point. Once the wall was broken more men would flood in and the enemy would be outflanked or even attacked from behind.

The correct way to prevent this, according to Vegetius, was to ‘swallow the charge’ by receiving it in a curved formation known as the forceps. It was easier to do this in a dense formation, but of course required training and a cool commander. Both the wedge and its countermeasures depended on firmness under fire and on fighting together as a well-drilled unit. How well drilled, in fact, were Dark Age armies? No drill manuals have come down to us. On the other hand, re-enactment experience suggests that formations can be taught basic proficiency in spear-and-shield warfare very quickly. Mastery of the basic moves – open order, forming ranks, advancing from column to line and turning about (in which the shield is passed over your head) – can be learnt in a day. In terms of basic drill, levied men could be turned into soldiers in a short time. To create a soldier who would stand firm in battle was another matter. There are many instances of a Dark Age army disintegrating under pressure. Morale depended on strong leadership and a sense of comradeship. Other requirements were personal fitness, which was probably high among the yeoman class, and courage. Re-enactments have confirmed another contemporary aspect of fighting – that, as the shield-walls lock together, it helps to shout! As anyone who attends football matches or has marched in large, noisy demonstrations will know, you lose your individuality in a pack, especially when you yell with the rest.

How did Dark Age armies find one another? Although hard evidence is scarce, it seems that armies of the period were highly mobile. King Harold famously marched from London to York in twelve days at some seventeen miles per day. This implies two things: that at least the flying columns of the force were mounted and that the roads and bridges were kept in good repair. From the striking correlation of Dark Age battle sites with Roman roads and major ancient tracks like the Ridgeway, it is evident that Dark Age armies made good use of roads. Perhaps this explains how kings like Oswald and Ecgfrith could campaign far from home without maps or a compass. They simply followed the roads. They also used scouts and presumably enlisted local people as guides, though recorded instances are hard to find from this period.

Shire armies seem to have been mustered at traditional outdoor assemblies or moots, such as Swanborough Tump in Wiltshire during Alfred’s Ashdown campaign or at Egbert’s Stone somewhere on or near the border of Gloucestershire and Wiltshire at the start of Alfred’s victorious campaign of 878. The shire reeves were responsible for ensuring that the men arrived on time and properly equipped. National service lasted for sixty days and, as the levies showed time and again, not a day longer. On more than one occasion, Alfred’s commanders had to let the Danes escape because the shire levies insisted on their rights and went home. This alone might explain why Dark Age commanders often seemed anxious to get the fighting over and done with.

To maintain speed the army marched as light as possible. Heavy war gear was carried in the rear in carts or by packhorses. Towns were expected to supply the army with food and other necessities as it marched. Bede confirms this with his story of the man who escaped death at the Battle of the Trent by pretending to be a civilian ferrying food supplies to the army. We are in the dark about living conditions on the march, but it seems that armies did bring tents with them. In Egiil’s Saga, Athelstan used his city of tents to confuse the enemy about his battle deployment. The saga might be fiction, but it would make no sense to its listeners had not tents been a normal part of army life.

We know little about battle formations in the Dark Ages. Large armies were evidently divided into sub-units, serving under different lords. From the ninth century, the shire levies were led in battle by their respective reeves, as at Ringmere in 1010 when the men of Cambridgeshire and East Anglia fought in separate divisions. Men of the top social class, royalty or ealdormen, fought among their hearth-troops who were expected to defend their lord to the death. Ealdorman Byrhtnoth at Maldon probably acted in the way expected of Dark Age commanders by putting himself in a prominent position in the centre of the line where his banner would be visible to the rest of his force.

Victory or defeat in a Dark Age battle depended on moral as well as physical strength. The professional Dark Age warrior, hearth trooper or mercenary, married late, if at all. The prime of his manhood was spent in the service of his lord, and he spent his leisure in the company of men, hunting, hawking and drinking. He lived on the cusp between the fiction of the sagas and praise poetry and the hard facts of military life: the former informed him of the way he was expected to behave, the latter of how heroic ideals worked out in practice. He repaid the mead he drank and the gifts he received by absolute loyalty and devoted service. One is bound to wonder: did the Dark Age warrior fear death in battle? It has been suggested that he was a fatalist: what will be, will be, and better to die gloriously than to live dishonourably. To fall in battle was considered an honourable death. Some warriors, especially the Welsh, thought that to die in bed was a disgrace. The trouble is that we do not know these people very well except through the doubtlessly idealized form of poetry. Although they might have been expected to conform to a heroic stereotype, the Battle of Maldon shows us that there were good and bad apples in every barrel. Some did indeed live and die according to their oaths. Others, it is clear, did not.




Visigoths and Ostrogoths fight each other on the Catalaunian fields.


Ostrogoths Siege of Rome AD 537

A barbarian people whose name means “Goths of the rising sun,” or “Goths glorified by the rising sun,” or simply “East Goths,” the Ostrogoths played an important role in the history of the later Roman Empire. Identified as early as the first century by Roman writers, the Ostrogoths were at first part of a larger population of Goths that included the Visigoths. During the third century, the larger Gothic population came into contact, often violent, with the Roman Empire. Defeated by the empire, with which they then cultivated better relations, the Goths divided into eastern and western groups, the Ostrogoths and the Visigoths, and their subsequent histories diverged. For the Ostrogoths, as well as the Visigoths, history in the fourth and fifth centuries was shaped by the movements of the Huns and the rise and fall of the great Hunnish empire of Attila. In the fifth century, a reconstituted Ostrogothic tribe formed into a powerful group led by kings. The most famous and important of these kings, Theodoric the Great, participated in political life in the Eastern Roman Empire and created a successor kingdom in Italy in the late fifth and early sixth century. Despite the qualities of Theodoric and the strength of his kingdom, the Ostrogothic kingdom of Italy did not long survive the death of Theodoric. In the 530s, the great emperor Justinian sought to conquer the Western Empire, which had fallen under barbarian control in 476. For some twenty years, Justinian’s soldiers and generals fought Ostrogothic armies before finally defeating them, destroying Theodoric’s creation, and essentially eliminating the Ostrogoths as a people and a force in history.

Ancient accounts record that Gothic history began in 1490 b.c., when a Gothic king led his people in three boats from Scandinavia to the mouth of the Vistula River. Eventually the Goths moved to the area between the Don and Danube Rivers, before being forced out in the mid-third century a.d. by the Huns. The traditional accounts of the origins of the Goths by ancient historians like Jordanes, however, are not generally accepted. The origins of the Goths are no longer traced to Scandinavia but rather to Poland, where archeological discoveries place a sophisticated, but nonliterate, culture. It was from there that the Goths moved, after which move they made contact with the Roman Empire. In the third century the Goths had repeated clashes with the empire, winning some and putting the empire, already in serious straits, into even greater jeopardy. Roman emperors gradually turned the tide and nearly destroyed the Goths. In the wake of these defeats, however, tradition holds that a great king emerged, Ostrogotha, in circa 290, who founded the kingdom of the Ostrogoths. Although it is unlikely that Ostrogotha existed, it is at that point that the division of the Goths into two groups occurred.

In the fourth century the two groups, the Tervingi, or Visigoths, and Greuthingi, or Ostrogoths, had more or less come to terms with the empire. By the 370s, however, the relationship between the various Gothic groups and the empire changed as they faced the threat of the Huns. Prior to the arrival of the Huns, King Ermanaric, a member of the Amal clan, had created a substantial kingdom in eastern Europe. He led the struggle against the Huns but was defeated by them, and in 375 he sacrificed himself to the gods in the hopes of saving his people from the Huns. His successor and some of the Goths continued the struggle against the Huns for another year before they were conquered and absorbed by them. From the end of the fourth to the middle of the fifth century, the Greuthingi/Ostrogoths remained part of the Hunnish empire and fought in the armies of the greatest Hun, Attila.

After the death of Attila, however, the fortune and composition of the Ostrogoths underwent a change. Most scholars believe that the Ostrogoths of this period are unrelated to earlier groups identified as Ostrogoths. Whatever the relationship is, in the mid-fifth century under the king Valamir, an Amal, the Ostrogoths emerged from domination by the Huns. Valamir exploited the confused situation in the empire of the Huns after Attila’s death in 453 and the defeat of Attila’s successor at the Battle of Nedao in 454. Although Valamir and his Goths most likely fought with the Huns against other subject peoples, the Ostrogoths emerged as an independent people because of the collapse of the Huns not long after the battle. Valamir then faced other rivals and endured further attacks by the Huns before their ultimate demise; he died in battle against the Gepids in 468/469.

Valamir was succeeded by his brother Thiudimer, who moved his followers into Roman territory, where they became foederati (federated allies) of the empire and came into contact with another group led by the Ostrogothic king Theodoric Strabo, or the Squinter. The two groups struggled against each other for preeminence and for preference before the emperor. The empire itself, however, underwent important changes during this period. In the 470s a new emperor, Zeno, came to power in Constantinople, and the emperor in Italy was deposed and the imperial line ended by the barbarian Odovacar in 476. These changes among the Ostrogoths and within the empire had an important bearing on the future of the Ostrogothic people.

In 473 Thiudimer died and was succeeded by his son Theodoric the Amal, or later known as the Great, who had been named successor in 471. Prior to his nomination, Theodoric had spent ten years in Constantinople as a hostage of the emperor. During that period Theodoric learned a great deal about the empire and its customs and culture, even though it appears that he did not learn to write. Upon assuming power, he found himself in competition with the other Theodoric, whose followers had revolted against the emperor in 471 and again in 474. The later revolt was part of a palace coup against the new emperor, Zeno, who turned to the Amal for support. In order to ensure that neither group of Ostrogoths or their leaders became too powerful, Zeno also began to negotiate with Theodoric settled a treaty with Theodoric Strabo in 479. The hostilities between the two Theodorics were settled for a time, too, as the two closed ranks against the emperor. In 481, Strabo attacked Constantinople but failed to take it or depose the emperor. Shortly thereafter he was killed when his horse reared and threw him onto a rack of spears. Theodoric the Amal was the beneficiary of his occasional ally and rival’s death. Although Strabo was succeeded by Rechitach, his followers gradually joined with Theodoric the Amal, who had Rechitach murdered in 484.

Theodoric the Amal, or the Great, to give him his more familiar name, was able to create a great Ostrogothic power that quickly threatened the power of Emperor Zeno. The Ostrogothic king continued the struggle with Zeno, which was resolved for a time in 483, with the emperor making great concessions to the king. Indeed, Theodoric was made a Roman citizen, given the title of patrician, and awarded a consulship for the next year. The Ostrogoths were given a grant of land within the empire. But it occurred to Zeno that he could not trust the rising power of Theodoric, and he replaced him as consul, an event followed by renewed hostilities between the Ostrogoths and the empire. Theodoric’s revolt in 485 put further pressure on Zeno, who responded by offering Theodoric the opportunity to lead the assault on Odovacar, the barbarian king in Italy since 476. This assignment, which Theodoric himself had first suggested in 479, was beneficial to both king and emperor and one that Theodoric quickly accepted.

In 488–489 Theodoric led his Ostrogoths, probably numbering some 100,000 people, against Odovacar in Italy. The struggle between the two leaders lasted until 493; it was a hard fought war, with Theodoric winning the battles but unable to take his rival’s capital of Ravenna. Indeed, after losing two battles Odovacar established himself in the capital, from which he ventured out to meet Theodoric on the field of battle. Odovacar’s hand was strengthened by one of his generals, who joined Theodoric but then rejoined Odovacar, slaying the Gothic warriors who were with him. As a result Odovacar was able to take the offensive, but only for a short while, until Theodoric was reinforced by a Visigothic army. In the early 490s Theodoric gradually took control of Italy and forced Odovacar to come to terms. On February 25, 493, the two leaders agreed to terms that were to be celebrated at a great banquet. Theodoric apparently agreed to share power with his rival, but at the banquet he killed Odovacar, and Theodoric’s followers killed the followers of Odovacar in a bloody massacre that ended the war and brought control of Italy to Theodoric.

After his victory, Theodoric was hailed king of Italy, but at first he had to refuse the title in favor of patrician of Italy. The new emperor Anastasius I (r. 491–518) refused to recognize the title of king, with its implications of Theodoric’s independence, reminding him that he held power at the discretion of the emperor. Ultimately, however, Theodoric was recognized as king in Constantinople and ruled Italy until his death in 526. His reign was highly beneficial for Italy, and his relationship with the native Roman population was generally good, despite his Arianism and the Romans’ Catholicism. He preserved much of the traditional Roman administration, as had Odovacar, and cooperated with the Senate. He ensured the food supply to Italy and patronized Boethius and Cassiodorus as part of a cultural revival. He was also an active builder throughout Italy, erecting public monuments and churches as well as his famous palace and mausoleum in Ravenna. His activities were not limited to Italy, however, but included an ambitious foreign policy that saw him establish hegemony over the Vandals in Africa and the Visigoths in Spain. In competition with Clovis in northern Europe, Theodosius was able to limit the Merovingian king’s expansion into southern Gaul. Although in name only a king, Theodoric, as contemporaries admitted, ruled as effectively as any emperor.

Theodoric’s later years and the years following his death were marked by increasing turmoil, leading to the eventual fall of the Ostrogothic kingdom. This situation was due in part to changes in the Eastern Empire, as well as to mistakes on his own part. In 518 a new emperor, Justin, assumed the throne and brought an end to a period of doctrinal uncertainty in the empire. He was a Catholic Christian who promoted traditional orthodox teaching, and in 523 he prohibited Arianism in the empire. The support for orthodox teaching and stability in doctrine restored the Italian population’s faith in imperial leadership. Moreover, Theodoric was further challenged in matters of religion by the success of the Catholic Clovis against the Visigoths. His concerns were heightened by an alleged plot involving a number of senators, including his advisor Boethius. He ordered Boethius executed and at the same time imprisoned the pope, who had just returned from an embassy to Constantinople. These actions strained relations with his Roman subjects and darkened an otherwise enlightened reign.

Theodoric’s situation was worsened by his lack of a male heir, and just prior to his death he encouraged his followers to accept his widowed daughter, Amalswintha, as regent for his grandson Athalaric. At first Theodoric’s wishes were accepted, but gradually the Ostrogothic nobility turned against Amalswintha. Although she was praised for her intelligence and courage, the nobility were divided over her guidance of Athalaric and her pro-Roman foreign policy. When Athalaric reached his majority in 533, a number of nobles sought to persuade him to turn on his mother. The rebellion was nearly successful. Amalswintha requested a ship from Emperor Justinian to take her to Constantinople, but ultimately stayed and triumphed over her rivals. She married a cousin, Theodohad, in 534 to stabilize the throne, but her husband failed to remain loyal to her, and Athalaric died that same year. Her arrest and murder, which was inspired, according to the fifth-century Byzantine historian Procopius, by Justinian’s wife Theodora out of jealousy, provided the emperor with the pretext for his invasion of Italy.

Justinian’s invasion of Italy, led at first by Belisarius and later Narses, opened the final chapter of the history of the Ostrogoths. The Gothic Wars, which lasted from 534 to 552, were devastating for both Italy and the Ostrogoths. The opening phase of the war saw rapid victories and much success for the invading armies, in part because of the weakness of Theodohad. Belisarius reached Rome in 536, and Theodohad was deposed in favor of Witigis. The rise of Witigis and the arrival of a second Byzantine general, Narses, slowed imperial progress. When Narses was recalled, Belisarius went on the offensive again and may have forced Witigis to take desperate measures, which possibly included Belisarius’s acceptance of the imperial title. Although this remains uncertain, Belisarius was recalled in 540 and took the Ostrogothic king with him. In 541, Witigis was replaced as king by Totila.

Under Totila’s leadership, the Ostrogoths fought back successfully and prolonged the war for another eleven years. Totila was able to win back territory in Italy from Byzantine armies and forced the return of Belisarius in 544. In 545 Totila began a siege of Rome; he occupied it in 546, laying waste to the city in the process. Control of the city swung back and forth between the two sides for the rest of the war, which Belisarius was unable to conclude, despite putting great pressure on his rival, because of inadequate supplies and soldiers. Belisarius was recalled in 548, at his own request, and replaced by Narses two years later. Narses demanded sufficient resources to bring the war to a swift conclusion and got them. In 552 Narses won the Battle of Busta Gallorum, at which Totila was killed and organized Gothic resistance was ended. Although Totila had a successor as king and pockets of Ostrogoths resisted until 562, the Ostrogothic kingdom in Italy was crushed by the Byzantine invasion. The Ostrogoths ceased to be an independent people, and the last of the Ostrogoths were probably absorbed by the Lombards during their invasion of Italy in 568.


Amory, Patrick. People and Identity in Ostrogothic Italy, 489-554. Cambridge: Cambridge University Press, 1997.

Browning, Robert. Justinian and Theodora. Rev. ed. London: Thames and Hudson, 1987.

Burns, Thomas. The Ostrogoths: Kingship and Society. Bloomington: Indiana University Press, 1980.

—. A History of the Ostrogoths. Bloomington: University of Indiana Press, 1984.

Bury, John B. History of the Later Roman Empire: From the Death of Theodosius I to the Death of Justinian. 2 vols. 1923. Reprint, New York: Dover, 1959.

Goffart, Walter. Barbarians and Romans a. d. 418-584: The Techniques of Accommodation. Princeton: Princeton University Press, 1980.

Heather, Peter. The Goths. Oxford: Blackwell, 1996.

Jordanes. The Gothic History of Jordanes. Trans. Charles C. Mierow. New York: Barnes and Noble, 1985.

Moorhead, John. Theodoric in Italy. Oxford: Clarendon, 1992.

Procopius. History of the Wars. Trans H. B. Dewing. Cambridge, MA: Harvard University Press, 1969-1993.

Wolfram, Herwig. History of the Goths. Trans. Thomas J. Dunlap. Berkeley: University of California Press, 1988.

—. The Roman Empire and Its Germanic Peoples. Trans. Thomas J. Dunlap. Berkeley: University of California Press, 1997.

Wood, Ian. The Merovingian Kingdoms, 450-751. London: Longman, 1994.


With the collapse of the Western Roman Empire in the latter fifth century A.D. the rule of empire vanished in northern Europe, and with it for a time an integrated market economy in the Mediterranean, North Africa, and Asia. The absence of the legions to provide security in the countryside against brigands and invaders at first led to ever greater disruption of farmland, while massive fortifications, not the courage of soldiers in open battle, were seen as the more reliable defense of the cities. The lack of central taxation meant that aqueducts, terraces, bridges, and irrigation canals were not properly maintained and often abandoned, leading not merely to the loss of potable water in the cities but also to a decline in agricultural productivity as valleys silted up and terraced land eroded.

The erosion of central imperial government and the collapse of urban culture also meant an end to large standing armies. Italy, Spain, Gaul, and Britain, in the absence of authority from Rome, were convulsed by a series of invasions and migrations by Vandals, Goths, Lombards, Huns, Franks, and Germans. Yet the victorious newcomers by the sixth and seventh centuries were no longer nomadic, but often had settled permanently in Roman territory, gradually converted to Christianity, learned some Latin, and carved out petty kingdoms guided loosely by the old Roman bureaucratic and legal tradition. If the new armies of western Europe were tiny and fragmented in comparison to Rome and often ensconced in fortified castles and towns, they nonetheless continued to rely on levies of heavy infantrymen fighting in columns, not tribal swarming, when it became necessary to engage in decisive battles.

The final collapse of Rome also brought a population decline in western Europe; and economic activity was lethargic for much of the so-called Dark Ages between 500 and 800. Christianity began to encroach on public and private lands, requiring enormous acreages to support monasteries, churches, and nunneries, whose clergy in the strict economic sense were not especially productive. If the estates of the old Roman patricians were sometimes unwisely expropriated for horse raising by the aristocracy of Franks and Lombards, then similarly the church also used the harvests from scarce and precious farmland to support a vast bureaucracy and an ambitious building program. By the end of the fifth century A.D., no single kingdom from Lombard Italy to Visigoth Spain could muster an army the size of the Roman force that had been annihilated at Cannae seven hundred years earlier.

Yet the fall of Rome often spread, rather than destroyed outright, classical civilization, as the fragments of empire slowly recovered and kept alive the cultural core of the old West. Writing continued. Even literature and scientific investigation were never completely lost. Latin remained the universal script of government, religion, and law from Italy to the North Sea. The Dark Ages (the term originally referred to the dearth of written knowledge that survived about the era) were characterized not so much by the chaos of an empire fallen as by the new diffusion of much of classical culture—language, architecture, military practices, religion, and economic expertise—into northern Europe, especially Germany, France, England, Ireland, and Scandinavia.

Islam had spread in the south and east by the creation of an entirely new theocratic state; in contrast, the remnants of classical culture, fused with Christianity, advanced throughout western and northern Europe due to the collapse of the Roman Empire. “Despite the resulting turmoil and destruction,” Henry Pirenne pointed out in regard to the supposed end of Roman civilization in northern Europe during the fifth century A.D., “no new principles made their appearance either in the economic or social order, nor in the linguistic situation, nor in the existing institutions. What civilization survived was Mediterranean” (Mohammed and Charlemagne, 284).

The sixth and seventh centuries actually saw improvements. Throughout the latter decades of the Roman Empire, there had been a gradual displacement of agrarians, concentrations of huge amounts of wealth, and constant class strife in the cities. The continuance of classical culture in ancient Gaul in the sixth through eight centuries, even under radically different and troubled material conditions, often meant that local government was more responsive to rural problems than had been Rome in its last two centuries. Under the Merovingians and Carolingians there nowhere reappeared the vast numbers of slaves that had characterized Roman civilization (by the fourth century A.D. in certain parts of the empire nearly a quarter of the population had been servile). Though Roman wealth and nationhood were gone for a time from the West, the deadly military tradition of classical antiquity was nevertheless kept alive. Most of the great military discoveries in both weaponry and tactics to come in the next millennium would originate in Europe—the continuing dividends of the Western approach to the dissemination of empirical data, the scientific method, and free inquiry.

“Greek fire” emerged at Byzantium somewhere around 675. Although the exact ingredients and their ratios of mixture remain unknown to this day, the torrent of flame that was shot out of Byzantine galleys was apparently a potent fusion of naphtha, sulfur, petroleum, and quicklime that could not be extinguished by water—a nearly unquenchable toxic spume that could incinerate enemy ships in seconds. Equally ingenious as the chemistry of Greek fire was its method of delivery, which involved a keen knowledge of pumps, pressurization, and mechanical engineering. A sealed container was heated from below with fuel and bellows and injected with forced air from a pump. Then the compressed mixture was forced out another outlet into a long bronze tube. The jellied mass was ignited at the end of the barrel, resulting in a sea of continuous flame spurting out from this ancient flamethrower. Ships with such fiery contraptions allowed the small Byzantine navy mastery of the eastern Mediterranean and saved Constantinople itself on occasion—none more dramatic than Leo III’s incineration of the Islamic armada of the caliph Sulaymān in 717 in waters surrounding the capital.

Controversy surrounds the exact origins of the stirrup—it may have been originally of Asian design—but by A.D. 1000 most Western cavalrymen were employing new saddles equipped with stirrups, even if they learned of their use via the Arabs, who had copied the original designs either from the Byzantines or by trading with the Orient in the early seventh century. Under the western European kingdoms, the stirrup was envisioned not merely as an aid to horse mastery but as integral to the emergence of a new lance-bearing knight, who could for the first time absorb the shock of spearing a fixed target on the gallop without being thrown from his mount. While such lancers could never break true infantry, small corps could easily ride down isolated groups of foot soldiers during both attack and retreat. The stirrup meant not that western European militaries were dominated by heavy lancers, but that their mostly infantry armies, at key moments in the battle—when gaps appeared in enemy lines or during the rout—could send out small corps of deadly horsemen to slaughter with impunity light infantrymen and poorly organized foot soldiers.

The crossbow—in use throughout Europe by 850—was a smaller-sized derivative of the classical “belly-bow,” through substitution of a handheld crank for large torsion cables and sprockets. Scholars cite the crossbow’s deficiencies in comparison with either the later English long-bow or the Eastern composite bow, both of which had greater range and rates of fire. The crossbow, however, required far less training to use than either, did not tire the archer to the same degree as hand-pulled bows, and its smaller all-metal bolts had greater penetrating power at short ranges. Crossbow bolts alone were able to slice through the heavy chain mail of the knight, and meant that a relatively poor man without much training could kill both an aristocratic horseman and his armored mount in seconds for the cost of a tiny metal projectile. Consequently, the church often issued edicts against its use—a doomed prospect of technological repression with no heritage in the West—and finally retreated to the position that crossbows should be outlawed in all intramural wars between Christians.

Siege engines underwent constant improvement. After 1180, vast catapults were powered by counterbalances rather than torsion alone. Such trebuchets often had ten-ton counterweights and could throw stones of three hundred pounds well over one hundred yards, exceeding the delivery weight of the old Roman traction catapults fivefold, while maintaining nearly the same range. In turn, fortifications were built entirely of stone and to heights unimagined by classical engineers, replete with intricate towers, crenellations, and interior keeps. It was not merely that European castles and walls were larger and stouter than those in Africa and the Near East, they were more numerous as well, due to improvements in the cutting, transportation, and lifting of stone. Plate armor, common by 1250, was also a European specialty, ensuring that most European knights and infantrymen were far better protected than their Islamic opponents. When gunpowder was introduced from the Chinese in the fourteenth century, Europe alone was able to craft dependable and heavy cannon—Constantinople fell in 1453 through the efforts of Western-fabricated artillery—and handheld matchlock weapons in any great number. So, too, fully rigged, multisailed ships were common in European waters by 1430, and were superior to any vessels in either the Ottoman or the Chinese navies.

Key to this continuing Western ability to craft good weapons, along with fluid and innovative tactical doctrine, was the embrace of published military research, which married theory with field experience to offer pragmatic advice to commanders in the field. The late Roman handbooks of Frontinus, and to a greater extent Vegetius, were copied even throughout the Dark Ages and became a bible of sorts to many western European warlords. Rabanus Maurus, the ninth-century archbishop of Mainz, published an annotated De re militari specifically to improve Frankish warfare. For the next four hundred years, adaptations and translations of Vegetius appeared throughout Europe by Alfonso X (1252–84), Bono Gimaboni (1250), and Jean de Meung (1284).

European siegecraft itself was unmatched, precisely because it followed in the past tradition of classical poliorkētika (the arts of “polis enclosing”). Manuals such as the Mappae Clavicula instructed besiegers in the use of engines and incendiary devices. The emperors Maurice (Ars militaris) and Leo VI (Tactica) outlined Byzantine infantry and naval tactics in preparing manuals for their generals and admirals to keep the Mediterranean Sea and its harbors free from Arab fleets. In contrast, Islamic writing on war was rarely abstract or theoretical—or even practical—but more holistic and philosophical, and largely concerned itself with the proper rules and conduct of the jihad.

Among the early Franks this need to write about war and to publish manuals about its practice were in direct emulation of Roman and Greek thinkers. Military practice did not operate in a vacuum, but was closely connected to the presence of an educated elite familiar with classical ideas of military organization and weaponry. Under the Carolingians, a systematic approach was undertaken to the preservation of classical manuscripts, along with efforts to assure education in the Greco-Roman tradition:

Though defined by religion, Europe was also a community of scholars who read and wrote the same Latin language and who rescued a great part of the legacy of antiquity from irretrievable loss. In the ninth and tenth centuries, schoolmasters devised a new curriculum of studies based in part on the classics that they had rediscovered. In doing so they laid the foundations of educational practices for centuries. (P. Riché, The Carolingians, 361)

In addition, the historiographic tradition of Greece and Rome continued in the Christian East and West, especially the Hellenic and Roman propensities of Herodotus, Thucydides, Livy, and Tacitus to see history largely as the story of war and politics. Thus, Gregory of Tours (534–94, History of the Franks), Procopius (born ca. 500, History of the Wars of Justinian), Isidore of Seville (History of the Goths, written 624), and Venerable Bede (672–735, Ecclesiastical History of England ) all provided anthropological detail about various tribes as part of larger exegeses of intercultural conquests and defeats. The works of hundreds of other lesser-known chroniclers and compilers circulated throughout Europe, the sheer number of titles unmatched by anything published elsewhere.

There were numerous early Islamic historians, many of whom were candid and remarkably critical, but few saw history as really existing before the era of the Prophet (thus the maxim “Islam cancels all that was before it”). And the parameters of inquiry were limited by the Koran, whose literary and historical primacy tolerated no competition from mere mortals. Contrary to classical historiography—there seems to be little evidence of any early Arabic translation of the major Greek historians— lapses in morality, not tactical blunders or structural flaws, were cited as reasons for Islamic defeats. After Poitiers, Arab chroniclers, as would be true of Ottoman observers in the aftermath of Lepanto, attributed the Muslim slaughter to their own wickedness and impiety that had brought on the wrath of Allah.

The horse-drawn, iron-tipped plow first emerged in Europe, allowing farmland to be broken up more quickly and deeply than with the old wooden blades drawn by oxen. The ability to farm more efficiently gave Westerners greater food and opportunity than their counterparts to the south and east. By the end of the twelfth century, windmills, which were unlike anything in the Near East or Asia, appeared in England and northern Europe. With a rotating horizontal axis and a system of gears, such machines could mill wheat at rates unimagined either in classical antiquity or the contemporary non-West. Improved water wheels—more than 5,000 in eleventh-century England alone—were used not only to grind grain but to manufacture paper, cloth, and metal. The result was that Western armies were able to campaign farther from home—both because they could take greater amounts of supplies with them and because farmers could go on campaigns for longer periods. Historians often remark on the unruliness of Crusader armies, constant bickering in command, horrendous camp conditions, and the occasional imbecility of their tactics, forgetting that the transportation and supply of thousands of soldiers to the other side of the Mediterranean was a feat of logistical genius unmatched by Islamic armies of the day.

Science and technology alone did not save the smaller and more fragmented western European armies from their adversaries. The classical traditions of infantry organization and landed musters were kept alive as well. Military command and discipline followed Roman tradition, and so naturally nomenclature remained Greek and Latin. Byzantine emperors, in the manner of Macedonian lords, addressed their soldiers as systratiōtai— “comrades-in-arms.” Generals, as in classical Greece, remained stratēgoiand soldiers stratiōtai, while in the West free soldiers were milites, both pedites (foot soldiers) and equites (knights). Citizens continued to be recruited under legal and published codes of conduct—the so-called “capitularies”—with explicit rights and responsibilities.

Charles Martel’s army was not as disciplined or as large as a Roman consular army, but the manner in which its heavily armed spearmen and swordsmen were mustered, attacked on foot, and kept in rank was consistent with the classical tradition. Campaigns required the approval of assemblies, and rulers were subject to audit after battle.

By the end of the eighth century two seemingly insurmountable obstacles that had once weakened the old Roman imperial levies of the fifth and sixth centuries A.D.—the failure of Roman citizens to serve in their own armies, and the religious strictures against civic militarism and wars of conquest by the early Christian church—were beginning to erode. Augustine had composed his City of God after the sack of Rome in A.D. 410 to associate divine punishment with the sins of Romans. Even earlier, a few Christian emperors, like Gratian, had dismantled public statues and commemoration of military victory as somehow antithetical to Christ’s message of peace and forgiveness. Yet by early medieval times the earlier pacifism of the Roman church fathers like Tertullian (Ad martyres, De corona militis), Origen (Exhortatio ad martyrium, De Principiis), and Lactantius (De mortibus persecutorum) was often ignored, as the creed of the Old Testament and its idea of wars against the unbelievers regained primacy over the message of the Gospels. Thomas Aquinas, for example, could outline the conditions of “just” Christian wars, in which the cause of the conflict could make war a moral Christian enterprise. Christianity would never exhibit the martial fervor of Islam, but during the Dark Ages it more or less curbed its early pacifist pretenses and its distance from the affairs of worldly politicians. The military of Joshua and Samson, not the loving remonstrations of Jesus, was invoked to keep Islam at bay.

Franks, Lombards, Goths, and Vandals may have been tribal, and their armies were poorly organized; yet such “barbarians” nevertheless shared a general idea that as freemen of their community they were obligated to fight—and free to profit from the booty of their enemies. In that sense of civic militarism, they were more reminiscent of the old classical armies of a republican past than had been the hired imperial legionaries on Rome’s defensive frontier:

The massive reliance on citizen-soldiers in the West lowered the demands on the central government for expenditures to support the military. . . . Indeed, the flexibility of the West in building on developments that took place during the later Roman Empire resulted in immense military strengths, which, for example, proved their worth in the success for two centuries of the crusader states against overwhelming odds. (B. Bachrach, “Early Medieval Europe,” in K. Raaflaub and N. Rosenstein, eds., War and Society in the Ancient and Medieval Worlds, 294)

The legions had crumbled not because of organizational weaknesses, technological backwardness, or even problems of command and discipline, but because of the dearth of free citizens who were willing to fight for their own freedom and the values of their civilization. Such spirited warriors the barbarians had, and when they absorbed the blueprint of Roman militarism, a number of effective local Western armies arose—as the Muslims learned at Poitiers.

Second Arab Siege of Constantinople in 717–718


umayyads 750ad



Date: August 717–15 August 718.

Location: on the Sea of Marmara, modern Istanbul.

Forces Engaged:

Byzantine: unknown. Commander: Emperor Leo the Isaurian.

Muslim: 210,000. Commander: Maslama.


Defeat of Muslim forces in their first serious attempt to overpower the Byzantine Empire led to another seven centuries of Christian power in southeastern Europe.

Historical Setting

Constantine the Great established the city of Constantinople as his capital in 323. In doing so, he occupied the former city of Byzantium, which for centuries had controlled the straits separating Asia and Europe. The Sea of Marmara is flanked northeast and southwest by the Bosphorus and the Dardanelles, two narrow straits linking the Mediterranean and the Black seas. Unless one goes completely around the Black Sea, the passage from Europe into Asia Minor is across one of those straits. Therefore, Byzantium/Constantinople/Istanbul has been an extremely strategic possession for both land and naval warfare, as well as overland and maritime trade. As Rome faded and Constantinople rose in power, it became the seat of the Eastern Roman, or Byzantine, Empire.

Muhammad the Prophet founded Islam in Arabia in the seventh century. Claiming his divinely inspired teachings, the Koran, to be the successor to the Bible and the fulfillment of God’s plan for humanity, he spread his faith by both proselytization and warfare. By coincidence (or divine intervention) Muhammad arrived on the scene just as the two Middle Eastern powers, Persia and the Byzantine Empire, had fought each other to an exhausted standstill. He was therefore able to acquire massive territorial gains hand in hand with the spread of his faith. Both Persians and Byzantines suffered major losses of real estate as well as major losses of converts to Islam, who found it less oppressive than the ultraconservative Orthodox Church.

Muhammad the Prophet had a public career of ten years (622–632), then died without publicly naming a successor. His close associate Abu Bakr was elected to succeed him but ruled only two years; upon his death Omar reigned as caliph (“deputy”), the religious and political head of Islam. For ten years Omar oversaw Islam’s expansion into Byzantine territory, Persia, Syria, modern Iraq, and Egypt. It spread further still under the caliphate of Othman (644–656), ultimately stretching west to the Atlantic shore of North Africa as well as east to Armenia and Afghanistan. After he was assassinated Islam split into two major factions: the followers of Muhammad’s nephew Ali became the Shi’ites, while the supporters of the Syrian governor Muawiya started the Sunni faction. Muawiya established the Umayyad dynasty, which ruled from Damascus between 661 and 750.

Muawiya’s goal was the downfall of the Christian Byzantine Empire, for reportedly whoever was involved in capture of the capital city of Constantinople would have all his sins forgiven. Intermittently between 674 and 678 Muslim forces attempted to capture the city, by both land and sea, but the double walls protecting it proved too formidable. Muawiya settled for a peace treaty with the Byzantine emperor, which provided for an annual tribute from Damascus to Constantinople. For the next thirty years Muslim armies carried the faith as far as Spain and India, but the lure of Constantinople, the key to Europe, always beckoned. Caliph Walid (705–715) organized the forces necessary to seize the city, but died before the project began. Thus, his successor Suleiman sent men and ships to the Byzantine capital in 717.

The Byzantine Empire had suffered through a series of mediocre emperors since the last assault. Anastasius was now emperor. He came to the throne in 713 and was in the market for able soldiers to defend his realm. In his army served a general named Conon, better known as Leo the Isaurian. (He was probably from Syria rather than the Anatolian province of Isauria [modern Konia], however.) He had been a soldier since 705 and in 716 took command of the theme (district) of Anatolia. He harried the approaching Muslim army as it marched out of Syria toward Constantinople, then took the throne from Anastasius in March 717. Crowned Leo III, he immediately set about laying in as many provisions as he could for the siege he knew was coming, a daunting task for a city of perhaps half a million people. He also oversaw the repair and strengthening of the city’s two walls and the placement of weaponry to repel attacks from land or sea.

The Siege

Caliph Suleiman named Muslama as commander of his army, reportedly 80,000 men marching through Anatolia toward Constantinople. His plan was to invest the city from the western, landward side while a huge fleet blocked any supplies from reaching the city. That fleet numbered some 1,800 ships carrying a further 80,000 men under the command of a general named Suleiman, not to be confused with the caliph. The Muslim fleet was divided into two divisions: one to blockade the Dardanelles (or Hellespont) and keep any relief from coming to Constantinople from the Mediterranean, and one to hold the Bosphorus to the north, keeping out any relief from Black Sea ports. Muslama crossed the Hellespont in July 717, then divided his forces. He took command of the main body that began the siege, while sending a detachment to Adrianople to keep an eye on the Bulgars, who had been pillaging through southeastern Europe and had attacked Constantinople in 712.

Immediately upon his arrival Muslama threw an attack against the walls, but it was easily beaten back. That convinced him against undertaking a frontal assault, so he began digging trenches to prevent any breakout from the city. Most of the fighting, therefore, took place on the water. Admiral Suleiman left part of his navy at the Dardanelles, as ordered, but led the remainder northward to take up station on the Hellespont. As they approached Constantinople, however, the leading ships were caught in a swift and unfamiliar current that began to tangle them. Seizing his opportunity, Leo quickly lowered the chain that protected the Golden Horn (the upper harbor of the city) and dashed out into the Muslim fleet before they could form into line of battle. Using Greek fire, his ships quickly destroyed or captured a large number of vessels while the rest retreated. Suleiman feared sailing past the city now, for another such battle could destroy the rest of his fleet. Thus, the northern avenue for aid for a time was kept open.

The Muslim effort was off to a poor start, and soon bad news came from Damascus. Caliph Suleiman had died of a stomach ailment (probably from overeating) and Omar II, not known for his military acumen, had replaced him. For the next several months little happened except for bad luck. The winter of 717–718 was much colder than usual and snow lay on the ground for more than three months. For an army born and raised in Arabia and Egypt this was disconcerting at best, deadly at worst. Delays in the delivery of supplies from Egypt, coupled with the bad weather, meant the deaths of thousands of besieging soldiers.

The Muslims hoped to take the initiative in the spring of 718 with the arrival of a new fleet from Egypt bringing 50,000 reinforcements. The 400 ships of the fleet from Egypt slipped past the Byzantine fleet in the Golden Horn at night, thus avoiding a naval battle, and anchored at the Hellespont. That cut off the flow of supplies and would eventually have spelled the city’s doom, but Leo’s navy again saved the day. He was aided by the desertion of large numbers of crew members from the new Egyptian fleet, sailors who were Coptic Christians and had been pressed into Muslim service. Learning of the enemy fleet’s disposition, Leo launched a surprise attack in June that caught them completely unawares. The Greek fire (an unknown mixture of materials with many of the characteristics of napalm) once again caused both destruction and terror; the Christian crews deserted wholesale to the welcoming Byzantine forces. The northern blockading fleet was destroyed and Leo followed up his victory with an attack on Muslim forces on the Asian side of the Sea of Marmara, opposite the capital. That attack was so unexpected that Muslim soldiers and sailors were slaughtered by the thousands.

Leo at this point proved himself to be a diplomat as well as a general. He sent envoys to the Bulgars, who persuaded their King Tervel to attack the Muslim army from the west. In July Tervel’s soldiers drove back the Muslim holding force at Adrianople and attacked Muslama’s forces in the rear, defeating them and inflicting some 22,000 casualties. This new threat was reinforced by the rumor that a Frankish army was marching across Europe to assist their fellow Christians. The Muslims had not yet fought the Franks, but had heard tales of formidable military power. Caliph Omar decided it was time to bring the siege to a close. On 15 August 718 Muslama led the army away from Constantinople.


The defeat at Constantinople was the first disastrous loss the armies of Islam had suffered. There had been occasional defeats, but never a catastrophe such as this. Of the 210,000 Muslim soldiers and sailors who took part, it is reported that only 30,000 actually saw their homeland again. Of the more than 2,000 ships reported to have been involved, only five supposedly made it home.

Had Muslama’s armies captured the city, the route into eastern Europe would have been virtually unguarded. Little organized resistance could have been mounted against hordes of Muslim troops until they reached central Europe. Constantinople, the seat of political, religious, and economic power in the Christian East, probably would have become Islam’s capital as it did in the wake of the Muslim capture of the city in 1453. The Eastern Orthodox Church may have disappeared, with untold consequences in eastern Europe and Russia, although such did not happen in 1453. Sea power would have been completely in Muslim hands, for no European population at the time owned a significant navy. None would until the Vikings a century later. Even with the Frankish victory at Tours in France fifteen years afterward, Islam could well have become the dominant European, and therefore world, religion.

The Byzantine victory insulated Europe from Islam, but also from other outside influences. Hellenistic knowledge and culture survived and in many ways flourished in the Middle East and Africa, while Europe entered the Dark Ages. Militarily Europe was strong, but cultural progress was at a crawl. Not until the Crusades and the resulting revival of trade with the East was the old knowledge rediscovered, and the Renaissance was the result. It is interesting to speculate what Europe may have been like had Constantinople fallen seven centuries before it did.

References: J. F. C. Fuller, Military History of the Western World, vol. 1 (New York: Funk and Wagnalls, 1954); Edward Gibbon, The History of the Decline and Fall of the Roman Empire, vol. 6 (London: Methuen, 1898); Warren T. Treadgold, Byzantium and its Army, 284–1081 (Stanford: Stanford University Press, 1991).