The Nature of Warfare in Anglo-Saxon England

More History & Heritage. Copied from HERE:

The Nature of Warfare in Anglo-Saxon England


ða þær Byrhtnoð ongan beornas trymian,
rad and rædde, rincum tæhte
hu hi sceoldon standan and þone stede healdan,
and bæd þæt hyra randas rihte heoldon
fæste mid folman, and ne forhtedon na.

Then Byrhtnoth began to arrange his warriors
gave advice as he rode told the warriors
how they should stand and hold their place,
and bade that their shields they hold correctly
fast in their fists and not be afraid.

(Battle of Maldon line 17)

An important function of the warband would have been to train the young warriors. Evidence for formal weapons training is limited; the Danish chronicler Saxo describes how the hero Gram, when young, trained with experienced swordsmen, carefully copying their methods of parrying and thrusting. Saxo also describes how Swedish warriors continued to train in swordsmanship during a long period of peace until they were so skilled that they could graze an opponent’s eyebrow with unerring aim. It should be noted, however, that Saxo, writing in the early thirteenth century, might well have been influenced by contemporary practices.

An important source of training appears to have been hunting and sports. Hunting in particular would not only have given the young warriors practice in the use of the spear and the bow but would also have accustomed them to killing. The hunting of potentially dangerous animals such as boars and bears, would also have accustomed them to a degree of personal risk taking, to working in a group and to relying on their comrades.

Sports such as wrestling, weight lifting and riding would have helped to develop the warriors’ physical strength. This would have been very important as it requires significant strength to fight effectively with either a spear or sword and a shield for an extended period. In particular, with a shield the unbalanced load and the position in which it is held require an unusual combination of muscles that are not developed in normal physical exercise.

One of the few references to the training of warriors in Anglo-Saxon literature occurs in the poem about the battle of Maldon in AD 991. We are told how Byrhtnoth the leader of the Saxon forces spent the last few minutes before the battle instructing his men how to stand in the shieldwall and how to hold their shields.

In the end, however, only through experience of battle could test whether a young man (geoguð) would in fact make a warrior (duguð). Hence the advice given in the poem ‘The Wanderer’:

Beorn sceal gebidan þonne he beot spriceð
oþ þæt collenferð cunne gearwe
hwider hreþra gehygd hweorfan wille

A man should wait before he makes a boast
until his bold spirit knows through experience
which way his heart will turn him.

(The Wanderer line 70)

Scale of warfare

Ðeofas we hatað oð vii men
from vii hloð oð xxxv
siððan bið here

Less than seven men shall be called thieves,
from seven to thirty five are a band,
more are an army.

(Laws of Ine)

It should be recognised that ‘warfare’ varied in scale, involving individual combats and small skirmishes as well as sieges and full scale battles. Indeed, to an Anglo-Saxon there would probably have been little distinction between ‘social violence’ between neighbouring communities and war, except perhaps for the rank of the combatants. Even on its largest scale warfare in Anglo-Saxon England involved only relatively small forces. The law code of Ine which defines any force over thirty five men as an army is corroborated by abortive coup of the ætheling Cyneard in AD 755 which was conducted with only eighty four warriors. Descriptions of huge losses, such as 2065 Welsh dead at the battle of Beandun (AD 614), must, therefore, be considered to be an exaggeration.

Single Combat

Ane sweordemerce gemærde
wið Myrgingumbi Fifeldore heoldon forð siþþan Engle ond Swæfe,
swa hit Offa geslog.

With one sword
the border was set with the Myrgings
by Fifeldor; Henceforth it has held
between Angle and Swaefe where Offa fought.

(Widsið line 41)

The smallest scale of ‘war’ was a single combat. This was an accepted method in Germanic society of settling a dispute between two individuals and single combat between two champions sometimes preceded a battle. The Danish chronicler Saxo described how “the valiant commanders of old avoided executing their missions at everyone’s risk where the issue could be settled by the fate of one or two”. It is unclear how widespread such duelling was in England but it is a common element in Viking sagas.

Single combats are also recorded among the continental Germanic tribes. The Roman historian Procopius records a single combat before the decisive battle between the Gothic army of Totila and the Byzantine army of Narses in AD 553. The result of the single combat does not appear to have obliged the army whose champion lost to come to terms, however. Rather the two opposing armies used the single combat as a test their ‘luck’ before choosing whether to commit to battle.

The Vikings appear to have recognised two sorts of duel, hólmganga and einvígi. Hólmganga literally means ‘island going’ since duels were traditionally fought on islands. Saxo records a duel fought on an island in the River Eider by the fourth century Anglian king Offa which is also referred to in the Anglo-Saxon poem Widsið. An island would be an obvious setting for a duel since it would make escape or external interference difficult. Also, for single combat between the champions of two armies, a river might well mark the natural boundary of the tribes’ lands. The Viking hólmganga was a formal affair fought with elaborate rules. The only account of the rules is given in Cormac’s saga:

It was the law of the hólmganga that the hide should be five ells long, with loops at its corners. Into these should be driven certain pins with heads to them, called tjosnur. He who made it ready should go to the pins in such a manner that he could see sky between his legs, holding the lobes of his ears and speaking the forewords used in the rite called ‘The Sacrifice of the tjosnur.’ Three squares should be marked round the hide, each one foot broad. At the outermost corners of the squares should be four poles, called hazels; when this is done, it is a hazelled field. Each man should have three shields, and when they were cut up he must get upon the hide if he had given way from it before, and guard himself with his weapons alone thereafter. He who had been challenged should strike the first stroke. If one was wounded so that blood fell upon the hide, he should fight no longer. If either set one foot outside the hazel poles ‘he went on his heel,’ they said; but he ‘ran’ if both feet were outside. His own man was to hold the shield before each of the fighters. The one who was wounded should pay three marks of silver to be set free.

The idea of each duellist having a second who protected him with a shield was not universal and may be a misinterpretation of the rules regarding the exchange of shields that had been damaged. Nor was the prescription to fight only to ‘first blood’ always obeyed, many duels were fought to the death.

Because of the complexity of the formal hólmganga, the Vikings also recognised a simpler version or einvígi. In Cormac’s saga, Cormac challenges Bersi to a hólmganga. Bersi replies:

“Cormac, you have challenged me to the hólmganga; instead of that, I offer to fight you in simple sword-play (einvígi). You are a young man and little tried; the hólmganga needs craft and cunning, but sword-play, man to man, is an easy game.”

This is the type of duel fought without undue formality by Thorstein and Bjarni in the Saga of Thorstein Stangarhoggr (Staff-struck), after Thorstein killed three of Bjarni’s servants in self defence:

Thorstein went outside and walked with Bjarni up to the hillock. They started fighting with determination and destroyed each other’s shield. When they had been fighting for a long time, Bjarni said to Thorstein, “I’m getting very thirsty now, I’m not so used to hard work as you are.”

“Go down to the stream then and drink,” said Thorstein.

Bjarni did so, and laid the sword down beside him. Thorstein picked it up, examined it and said “You can’t have been using this sword at Bodvarsdale.”

Bjarni said nothing, and they went back to the hillock. After they had been fighting for some time, it became obvious to Bjarni that Thorstein was a highly skilled fighter, and the outcome seemed less certain than he’d expected.

“Everything seems to go wrong for me today,” he said. “now my shoe-thong is loose.”

“Tie it up then,” said Thorstein.

When Bjarni bent down to tie it, Thorstein went into the house and brought back two shields and a sword. He joined Bjarni on the hillock and said, “Here’s a sword and a shield my father sends you. The sword shouldn’t get so easily blunted as the one you’ve been using. And I don’t want to stand here any longer with no shield to protect me against your blows. I’d very much like to stop this game now, for I’m afraid your good luck will prove stronger than my bad luck. Every man wants to save his life, and I would too, if I could.”

“There’s no point trying to talk yourself out of this,” said Bjarni. “The fight must go on.”

“I wouldn’t like to be the first to strike,” said Thorstein.

Then Bjarni struck at Thorstein, destroying his shield, and Thorstein hacked down Bjarni’s shield in return.

“That was a blow,” said Bjarni.

Thorstein replied, “Yours wasn’t any lighter.”

Bjarni said, “Your sword seems to be biting much better now than it was earlier.”

“I want to save myself from the foulest of luck if I possibly can,” said Thorstein. “It scares me to have to fight you, so I want you yourself to settle this matter between us.”

It was Bjarni’s turn to strike. Both men had lost their shields. Bjarni said, “It would be a great mistake in one stroke both to throw away good fortune and do wrong. In my opinion I’d be fully paid for my three servants if you took their place and served me faithfully.”

It is not clear whether anything as elaborate as the rules for hólmganga were ever used in England.

Duels do not always appear to have been fought one against one. The duel fought by the Anglian king Offa was against two opponents, although this may have been exceptional as it was apparently to make up for a previous incident where the Swedish king Athisl was killed in a duel by two Anglian brothers, Keti and Vigi.

It appears that it was customary in a single combat to exchange blows alternately. This occurs not only in the formal hólmganga but also in the einvígi between Thorstein and Bjarni. Saxo mentions this custom:

At the outset there was argument for a while as to which of them should make the first stroke, for in days of old when contests were arranged they did not try to exchange a rain of blows but hit at one another in a definite sequence with a gap between each turn. The strokes were infrequent but savage, with the result that it was their force, rather than their number that won acclaim. (Saxo II.56).

Saxo describes such a duel between Bjarki and Agner:

Precedence was given to Agner because of his high rank, and the account has it that he gave a blow of such might that he clove the front of Biarki’s helmet, tore the skin on his scalp and had to let go of the sword which was stuck in the eye-guards of the helmet. When Biarki’s turn came to strike, he braced his foot against a log to get a better swing to his sword and drove the knife-edged blade straight through Agner’s midriff. Some maintain that his dying mouth relaxed into a smile, a supreme disguise of his agony as he gave up the ghost. (Saxo II.56).

At other times duels appear to have been undertaken without any such formality. Saxo describes the duel between the Swedish king Athisl and Keti:

… and giving as good as his word laid on with all his strength. Ketil met this with so stout a blow of his sword that it split the king’s helmet and forced it’s way to his head. Exasperated by this wound (blood was streaming copiously from his scalp) he went for Ketil with a volley of brisk strokes and beat him to his knees. (Saxo IV.97).

The belief that blows were exchanged alternately may have begun as a literary motif based on the natural rhythm of blow and counter-blow that occurs when fighting with a sword and shield.


þa wæs feohte neh,
tir æt getohte. Wæs seo tid cumen
þæt þær fæge men feallan sceoldon.

Then was the fight nigh,
glory in battle. The time had come
when fated men should fall.

(Battle of Maldon line 103)

A battle was the decisive moment of a military expedition. Because of this battles are by far the most frequently recorded type of engagement in the historical records. It is likely that many more military expeditions did not culminate in any decisive action and were therefore not recorded. The decisive nature of battles meant they were a high risk option; the defeated force risked annihilation if they did not have a line of retreat to a defensive position. Battles, therefore, would only tend to occur when either the opposing forces were, or mistakenly believed themselves to be, of a similar size or, when a weaker force was caught by a fast moving enemy and forced to fight.

As an alternative to risking a battle a commander had a number of options. Depending on the cause of the conflict he might try and negotiate terms, such as the ceding of land or payment of tribute. Such settlements would often be accompanied by an exchange of hostages as an assurance of future good faith. Thus Oswy, king of Northumbria, attempted to buy off Penda, king of Mercia, prior to the battle of Winwaed in AD 655.

Another course of action would be to attempt to retire to a fortress or other defensible position. The enemy would then have three choices, assault the position, abandon the pursuit or lay siege. In the poem Beowulf the Swedish king Ongentheow is attacked by the army of the Geatish king Hygelac. Since he does not believe his own forces, who have just fought a battle against another Geatish army, can resist them in open battle he instead retreats to his fortress, which is on high ground and defended by earth ramparts. The defences are not strong enough, however, and the position is assaulted and Ongentheow killed.

Lastly, the army might simply disperse, making it difficult for the enemy to find them. Bede records how Oswin of Deira, realising the army of Oswy of Bernica was much stronger than his own decided not to risk an engagement at that time and so disbanded his army and sent all his men back to their homes.

Anglo-Saxons, unlike the Continental German tribes such as the Franks and the Goths, do not appear to have regularly fought on horseback. The poem about the battle of Maldon in AD 991 makes it clear that many of the warriors had ridden to the battlefield but then dismounted and fought on foot. Indeed Byrhtnoth, the leader of the Saxons, gave orders to drive away the horses so that there would be no thought of retreat. Only his own horse was kept so that he could use it to ride around the battlefield. When Byrhtnoth was killed and a large part of the army routed one of the fugitives took his horse to speed his escape. Another account, albeit even later, is the Anglo-Saxon Chronicle entry for 1055. This records how the Herefordshire fyrd, in battle with the Welsh, fled before a single spear was thrown because they had been ordered to fight on horseback, contrary to their custom (Anglos contra morem in equis pugnare jussit). There are occasional references in Anglo-Saxon histories to mounted forces, such as the Northumbrian king Ecgfrith’s expedition of horsemen (equitatui) sent against the Picts which is mentioned in the biography of Bishop Wilfrid. However, these do not make clear whether the force fought mounted or dismounted. Equitates is the general Latin term for mounted soldiers as opposed to the word ala which is used only of cavalry.

In battle both armies would mass together to form a ‘shieldwall’. To be effective the shieldwall would have to be positioned so that it could not be outflanked. The smaller, weaker force would usually have some choice where to stand and fight and would therefore choose ground where natural terrain features, such as rivers, forests, bogs etc., made movement around the flanks of the line difficult or impossible. In this way they would be able to deploy in lines several ranks deep, largely negating the enemies numerical superiority.

A high proportion of Anglo-Saxon battles are recorded as taking place close to rivers. Fords in rivers would be a natural place to intercept an enemy force moving into or through an area and would also provide an excellent defensive position with both flanks secure. Hence the many battles that are recorded as having taken place at fords: Crecganford (AD 485), Cerdicesford (AD 519), Biedcanford (AD 571) and Beoford (AD 752). Armies might also move along rivers, particularly if they had boats available to transport their supplies. The river would then have provided at least one secure flank and also a line of retreat if the enemy were not similarly equipped with boats. In this way the Viking army of Anlaf escaped annihilation following their defeat at the battle of Brunanburh in AD 937. A fork in a river could provide a strong position with both flanks secure. In AD 876 King Alfred of Wessex was forced to come to terms with a large Viking army that had taken up a strong defensive position between the River Frome and the River Tarrant in Dorset. Where high ground forced a loop in a river a natural fortress was created with only a narrow frontage with secure flanks and an advantageous position. Procopius records a Gothic army holding just such a position in a bend in the River Ister against a large army of Huns in AD 550.

If suitable terrain could not be found the weaker force would be forced to thin the line or risk being outflanked. If the shieldwall was too thin, however, it risked the enemy breaking through. The Bayeux tapestry shows men fighting in both a close formation with shields overlapping and also in a much looser formation.

A battle would begin with archers and other lightly armed troops exchanging missiles. These scouts might begin by operating forward of their own shieldwall but would be forced back behind their own troops as the armies closed. Although missile weapons might be used to thin the enemy battle line and perhaps disrupt the close packed formation, the defining moment of battle for the Anglo-Saxon warrior was when the shieldwalls met and blows were exchanged hand to hand.

The majority of the warriors would have probably been armed only with a spear and even the more wealthy warriors who owned a sword appear to have used spears initially. Gradually the opposing lines would be thinned as men were killed or wounded and fell out of the line. At this stage many of the injuries would have been relatively minor and wounded men would have often been able to escape to the rear. If one side felt itself to be at a disadvantage there might even be time to try to come to terms.

If the enemy line became disrupted the more aggressive warriors would attempt to make a break through, using their swords to close on the enemy and cut them down. If the shieldwall was penetrated and could not be reformed then the remaining warriors would be forced to abandon the position or they would be cut down from behind.

Apart from the shieldwall the only other formation known to have been used, albeit infrequently, was the wedge. This formation (Latin cuneus) was used by the Romans and is mentioned by Vegetius in his Epitoma Rei Militaris written in either the fourth or fifth century:

A wedge is the name for a mass of infantry who are attached to the line, which moves forward, narrower in front and broader behind, and breaks through the enemy lines, because a large number of men are discharging missiles into one position. (Vegetius III.19)

He goes on to say that the wedge is known as a ‘pig’s head’ (caput porcinum) by the soldiers because of the resemblance to the head of a boar.

The use of the wedge in a battle between the Romans and the Sarmatians in AD 358 was described by the Roman soldier and historian Ammianus Marcellinus.

The furious madness of this onset so angered our army that it could not brook it, and as the savages hotly menaced our emperor, … they took the form of a wedge … and scattered them with a hot charge. (Ammianus Marcellinus XVII.13.9)

The wedge is also mentioned in Viking literature and by the Danish historian Saxo. Saxo twice describes the formation, known to the Vikings as a svínfylking :

… in the first row he would put two men, four in the second, then increase the third to eight, and step up each succeeding rank by doubling the numbers of the one in the front. (Saxo I.32)

… he should divide his entire battle-line into three squadrons; each of these he should pack in twenties, but extend the middle section by a further twenty men, arranging them to form the point of a cone or pyramid, and should bend back the wings to create a receding curve on each side. When a muster was held, he should construct the files of each squadron by starting with two men at the front and adding one only to each successive row; thus he would set three in the second line, four in the third, and so on, building up the following ranks with the same uniform symmetry until the outer edge came level with the wings. Each wing must contain ten ranks. Again, behind these he was to introduce young warriors equipped with javelins; to the rear of these he should place a company of older men to reinforce their comrades, if their strength waned, with their own brand of seasoned courage; a skilful strategist would see that slingers were attached at the sides, who could stand behind the lines of their fellows to assail the enemy with shots from a long distance. Beyond these he should admit indiscriminately men of any age or class without regard for status. The final battalion he ought to separate into three prongs, as with the vanguard, and deploy them in similar proportioned ranks. The rear, though connected to the foregoing columns, might offer defence by reversing itself to face in the opposite direction. (Saxo VII.248)

In both cases the secret of the wedge formation was supposed to have been given to the Viking commander by Odin.

As is made clear by Vegetius the wedge was intended to penetrate the opposing shieldwall so that the enemy could be attacked from the rear. Both Roman and Viking authors agree that the best way to defeat this stratagem was to form up a line in an arc or V shape, known to the Romans as a forceps or forfex, so as to envelop the wedge in a pincer movement.

Other tactics and formations may have been used although there is no direct evidence. The most likely tactic would be to attempt to encircle the enemy’s flanks. The enemy’s left flank was traditionally seen as the weaker, since the men attempting to outflank it could present their shields to the enemy. The terrain would also have a significant impact, however, and either flank might in fact be vulnerable. To counter such manoeuvres, and other stratagems, groups of men would probably be held in reserve behind the shieldwall, so that they could quickly deploy wherever they were needed.

Many of the most significant factors in the ability of the shieldwall to withstand the shock of battle would be moral rather than physical. The strength of the shieldwall depended on the willingness of the warriors to hold their ground while under attack and to put their lives in jeopardy by stepping forward into the front rank from the relative safety of the rear when other men fell. In war, with few exceptions, individuals are only be able to withstand their natural instincts to run away because of the presence of their comrades, most importantly those they know well. They risk death only because the alternative was worse, the complete loss of status within their small social group from being branded a coward. The structure of the army, based on individual warbands, would have been extremely effective in developing this esprit de corps. Individuals would completely identify with their warband, having lived together in a closed community for many years. In turn the army would be held together by similar bonds between the leaders of the individual warbands and the king. If its leader fell in the battle the rest of the warband would be obliged to avenge him or to die at his side.

During the fight each side would try to rouse themselves and intimidate their enemies with gestures and shouts. The battle-cry of the Germanic tribes was known as the barritus and was described by the fourth century Roman historian Ammianus Marcellinus:

This shout in the very heat of combat rises from a low murmur and gradually grows louder, like waves dashing against the cliffs. (Ammianus Marcellinus XVI.12.43)

Vegetius states that the battle-cry should not be raised until both lines have engaged each other as the enemy are more likely to be terrified if the shock of the battle-cry accompanies the attack; only inexperienced or cowardly men call out from a distance.

Ammianus also records how the Germanic warriors would ‘sound the glories of their forefathers with wild shouts’, a practice echoed six centuries later in the poem of the battle of Maldon where Ælfwine, son of Ælfric, recounted his lineage before promising to die rather than flee from the battlefield following the death of his lord.

If the shieldwall was broken or encircled groups of warriors might try to retreat to a place where they could reform. A battle between the Swedes and the Geats is described in the poem Beowulf. Following the death of their king, Hæthcyn, early in the battle the Geats were forced to retreat and were pursued by Swedes until they reached the forest of Ravenswood where they were besieged overnight. In the morning they were only saved by the arrival of a second Geatish army, otherwise they would have been attacked again and either cut down or hung if they surrendered.

If there was nowhere to retreat to, a rout would probably ensue as individuals and small groups scattered and attempted to outrun their pursuers. In a rout such as this the worst of the slaughter would have occurred, as is described at the battle of Brunanburh:

Wesseaxe forð …
heowan herefleman hindan þearle
mecum mylenscearpan.

The West Saxons …
hewed the battle-fugitives from behind terribly
with mill-sharpened blades.

(Anglo-Saxon Chronicle entry for AD 937)


If the number of combatants and the ground were such that an effective linear defence could not be made, any engagement would naturally develop into a skirmish. In addition skirmishes between opposing groups of scouts would be the natural precursor of any battle.

Skirmishes would differ significantly from battles between two shieldwalls. In a skirmish men would tend to fight individually or in small groups rather than as a formed body. The groups would usually form a loose ‘skirmish line’ with visual contact maintained between each element. Each line would manoeuvre, attempting to gain a local superiority in numbers or position. If the enemy concentrated forces so as to threaten one part of the skirmish line it could give ground without losing integrity so that if the enemy advanced too far they would find their own flanks exposed to a counter-attack. Similarly, if the skirmish line was threatened by an outflanking manoeuvre it could simply be extended in response.

The preliminary skirmishes prior to a battle could be extremely important since, by preventing the enemy from observing their movements, the victorious army might well be able to lay an ambush. King Ecgferth of Northumbria was killed along with the greater part of his army having been lured into an ambush in a narrow mountain pass by the Picts in AD 685. This defeat was so total that Northumbria never really recovered and all the Pictish and Scottish lands that had been tributary were able to regain their independence.

It is likely that whenever possible local men would have been used as scouts since they would have had a considerable advantage because of their knowledge of the terrain. This is suggested by the obligation recorded in the post-conquest Book of Fees for the men of Cumberland to join the king’s army in Scotland, serving in the vanguard when the army was advancing and the rearguard when it returned. A similar obligation is recorded in the Doomsday Book for the town of Archenfield on the Welsh border.

This type of warfare must have placed a heavy burden on the personal courage of the men. They would have had to move and fight as individuals away from the moral support of their comrades and would have felt themselves constantly under threat due to the complex and rapidly changing nature of the battlefield. This morale aspect would have been as important, if not more so, than the actual sizes of the forces involved. A determined force that moved quickly and aggressively might be able to force their opponents to retreat even if they were numerically inferior.


Her Ælle ond Cissa ymb sæton Andredes cester ond ofslogon alle þa þe þær inne eardedon ne wearþ forþon Bret to lafe.

Here Ælla and Cyssa besieged the fortress of Anderida and slew all those who dwelt within, so that there was not on Briton remaining afterwards.(Anglo-Saxon Chronicle entry for AD 491)

Even before the intensive phase of fort building in the eighth century, undertaken in response to the threat of Viking raids, there would have been a number of fortified settlements in England. The most formidable would have been based on Roman defences, such as the Roman fort of Anderida at Pevernsey, Hampshire, which was successfully besieged in AD 491 by Ælla and Cyssa. However, in general these defences were not maintained, so that by the late seventh century the walls of Carlisle had decayed to such an extent that they were regarded as a mere curiosity. An exception was York, where excavations have shown that the Roman walls were repaired during the seventh century, although by AD 867 they had been allowed to decay again, so that when the Vikings were attacked there by the Northumbrian armies of Osberht and Ælla the city was described as having neither strong nor well-built walls.

Other settlements, particularly royal vills, would be protected by a wooden palisade. The fortress of Bamburgh, a Northumbrian royal vill, is described in the Anglo-Saxon chronicle entry for AD 547 as having been first protected by a stockade, then by a wall. Excavations at two Mercian royal vills at Hereford and Tamworth have revealed the existence of ditch and palisade defences dating from the eighth or ninth century on the same line as the tenth century ramparts. This is consistent with the imposition of the ‘common burden’ of fortress work in Mercia in the mid to late eighth century.

British kings also maintained fortified settlements defended by earthworks and palisades. Many of these were based on hillforts that predated the Roman invasion. Excavations have shown that a number of these hillforts were reoccupied in the late-Roman and post-Roman periods. The largest of these is at South Cadbury, Somerset. The extensive pre-Roman ramparts were refortified in the fifth century with a timber fighting platform dressed with stone nearly 1,200m (4000′) long and the entrance defended with a gate-tower. The enclosure contained a number of buildings including a royal hall 19m (62′) long and 10m (33′) wide. The fortress continues in use until at least the late sixth century and must have been involved in the defence of the area during the expansion of Wessex culminating in the battle of Dyrham in AD 577.

There is no evidence that South Cadbury was ever attacked, however, the Anglo-Saxon Chronicle records at two battles that took place at nearby hillforts. In AD 522 Cynric king of Wessex fought against the British at Searo byrg, believed to be the fortress of Old Sarum, Wiltshire. Four years later he again fought the British, this time at Beran byrg, Barbury Castle, Wiltshire.

Not all British fortified settlements were of this scale, however; the fortress of Mote of Mark, Strathclyde, is set on a craggy hillock and was defended by a timber reinforced stone wall. The area enclosed is only 75m (244′) at its longest and 35m (114′) at its widest. Interestingly, excavations showed evidence that the defences were destroyed by fire at some point in the forts history.

Where a settlement was not fortified with a palisade the main hall could be defended, as occurred in AD 755 (see above). The defence of a hall is also a common literary theme. A fragmentary poetic account of a famous battle at the Frisian vill of Finnsburg describes in detail how the Danish king Hnaef and his warband of sixty warriors were besieged in a hall following the resurgence of an old feud. The warriors had been feasting and were awoken by Hnaef who had heard the enemy approaching. They immediately ran to the two doors of the hall to stop the attackers getting in. The fighting was long and bloody:

Ða wæs on healle wælslihta gehlyn
sceolde cellod bord cenum on handa
banhelm berstan buruhðelu dynede

There was in the hall the sound of slaughter
curved shield boards held bravely in the hand
bone helms (skulls?) burst the hall floor resounded

(Finnsburg line 28)

The defenders held out for at least five days, until eventually a truce was arranged. In the fighting both Hnaef and the Frisian king Finn’s son were killed.

An alternative to storming a fortress was to burn it down. Bede describes how in AD 651 Penda, king of Mercia, was unable to capture Bamburgh, either by storming it or by siege, and so attempted to set fire to it. He directed his troops to gather wood from the surrounding villages, pile it alongside the city walls and set fire to it. However, before the fire could take hold on the walls, the wind turned away from the city, driving the flames back on the men kindling it so that some of them were injured and the assault failed. Njal’s Saga describes how Njal and his family were besieged in their hall following a long running blood-feud. When the besiegers found they were unable to overcome the defenders they were forced to burn the hall down.

A besieged force might still pose a significant threat and during a long siege the besiegers would be vulnerable to surprise attacks. Bede related how in AD 663 Osric, king of Deira, besieged the British king Cadwalla. The British force surprised their besiegers with a sudden sally, defeated them and killed Osric. For this reason, during extended sieges, besiegers may often have built their own defences, as the Vikings did when besieging Rochester in AD 885.

Some of the most enigmatic remains from the Anglo-Saxon period are the great linear earthworks such as Wansdyke in Wiltshire and Grim’s Dyke and Fleam Dyke in Cambridgeshire. These cannot in general be accurately dated, except for the most well known, Offa’s Dyke, which was built by the Mercian king Offa (AD 757-796). The defences ran for 240km along the English-Welsh border from the Dee estuary in the north to the river Wye in the south, using natural barriers wherever possible together with 130km of earthworks. In places the ramparts were 7m (24′) high surmounted by a palisade together with a 2m (6′) ditch. It is unlikely, however, that these defences were ever intended as linear fortifications. No Anglo-Saxon kingdom would be able to sustain sufficient forces to provide adequate defences for even a small part of their length. Rather they are likely to have served as ‘trip-wires’, ensuring that enemy raids were identified so that they could be intercepted by forces stationed in the interior of the kingdom.