The core of an army in Anglo-Saxon England was the king’s military household, whose members served their lord in return for reward, by the seventh century increasingly given as landholdings. Nobles and royal officers had their own households, which could be summoned to the army. Young warriors served a form of apprenticeship in these households before they married and settled on their lands. In another parallel development with Francia, changes occurred in the methods of raising armies in England in the eighth and ninth centuries. At the time when the Viking “Great Army” appeared in England there may have been a problem with increasing amounts of land being donated to the Church, both by kings and by aristocratic families. In the late seventh century Bede had claimed that this was diminishing the amount of land providing warriors in Northumbria. It may be that ecclesiastical land was held by the family in perpetuity, rather than being dependent on good service to the king. Alternatively, many earlier landholdings were also held in perpetuity, but the granting of them by charter, which made them “bookland,” enabled a lord to bequeath it to a person or persons of his choice rather than being compelled to divide it amongst his inheritors.
Charters of late eighth-century Mercia and Kent make it clear that there were three common obligations to the king: fortress work, bridge work, and military service. Church lands were obliged to provide labor services such as repair of fortifications and bridges, even if they were exempt from military service. Whether any were totally exempt is unclear. Kings like Offa of Mercia began to specify in their charters that Church lands were not exempt from military service, in order to maintain their military manpower. Offa used the obligations to create a network of forts and probably the Dyke on the frontier with Wales as well. By the ninth century these charter practices had spread to Wessex. It seems that it was around the end of the eighth and the beginning of the ninth centuries that kings began to specify the amount of military service required from a given amount of land: for instance, a Mercian charter of 801 says the owner of a thirty-hide estate should provide five men to the army, in other words one man per six hides, or more likely the owner and five men, one man per five hides. The five-hide unit is referred to in early eleventh-century tracts on social status as a designation of thegnhood. It also appears frequently in the listed military service requirements in the Domesday Book of 1086, but the variations it records in military obligation between shires tell us that this was not a universal standard throughout England. Notably, the five-hide shires largely correspond with those of Wessex and western Mercia. Given these variations, it seems unlikely that the obligations of Edward the Elder’s time were identical to those of Edward the Confessor’s, even if there was considerable continuity.
Many of the Wessex charters purporting to be ninth-century are forgeries. However, the reference made in many of these to two decimations of land by King Æthelwulf in 845 and 855 may have a basis in truth, as the ASC and Asser’s Life of Alfred also refer to the 855 decimation. The object was to provide a tenth of his lands to the Church, but probably also to give lands to thegns, who were the mainstay of his army. He, Alfred and Edward the Elder almost certainly built upon the one man per five hide requirement when they reorganized the military defenses and forces of Wessex and Mercia. Both Asser and Hincmar of Reims say that Church lands were exempt from military service in Æthelwulf’s kingdom. It is likely that the increasing Viking threat brought about some of these changes in the Anglo-Saxon kingdoms. When they overran most of Northumbria, East Anglia, and a large part of Mercia, and almost defeated Wessex as well, the Vikings gave the impetus for even more far-reaching reforms by Alfred of Wessex. The basis was the construction of a series of burhs, so that no one lived more than a day’s journey from one. Each had an administrative district that was to provide its garrison and maintain its defenses. The Burghal Hidage, a document which dates from the reign of Alfred’s son Edward the Elder, indicates that each hide was responsible for defending and maintaining approximately 4 feet of wall. This appears to correspond well to the length of many of the known burh ramparts and their districts. In all likelihood the system, which Edward and his sister Æthelflæd extended into Mercia as they reconquered land from the Danes, was little changed from that instituted by Alfred after his victory at Edington (878).
Alfred also wished to have an army available at all times. According to the ASC: “The king had divided his army into two, so that there was always half at home and half out, except for those men who had to hold the burhs.” The entry is made under 894, but the reform, if it was more than a temporary arrangement, was presumably carried out before this and after Edington. It is more likely to mean a division of the army into three, rather than the two parts normally envisaged, in which case a man liable to service would spend a third of the year on his land, a third garrisoning the burh of his district and a third with the field army. As Guy Halsall points out, apart from making a reasonable demand on the thegn’s time, this would enable the assembly of the field army straight from the burhs, rather than from innumerable individual landholdings. This would save time and mean that the men of each district were already familiar with each other and had probably trained together in preparation for field service. The ASC entry of 917 mentions burhs as assembly points for armies. While garrisoning the burhs men could also carry out fortress and bridge work. Both the ASC and Asser mention on several occasions that these men rode horses. This is additional evidence that these men were a select levy (fyrd). The levy of one man per hide of the Burghal Hidage must have been something else, perhaps called up in situations of imminent danger. Alternatively, they functioned as a “support force” for the real warriors, sequestered on the basis of the burghal system. Each man was supposed to provide twenty shillings for two months’ maintenance in the event that he was called up: this was presumbly the payment raised from the five hides for the support of the warrior (the ‘man’), perhaps on the basis of one per hide. Æthelstan’s Grately law-code states that “each man” should have two mounted men per plow (sylh). The latter would have been a heavy burden, even if the sylh referred to is equivalent to the Kentish sulung of the Domesday Book, which was two hides, rather than a single hide or even less. Again it may refer to the support for the “man” (warrior), who needed to be mobile to keep up with the army. Æthelstan needed to respond to threats further afield and more often than any of his Anglo-Saxon predecessors.
Even the warriors who remained “at home” might be called upon to fight if their lands were threatened. The Vikings were a mobile enemy who could strike at various points along the coast with little warning. In late 893 three ealdormen, Æthelred of Mercia, Æthelhelm of Wiltshire and Æthelnoth of Somerset, raised the garrisons and the “king’s thegns who were at home near the burhs” to attack a Viking army encamped at Buttington by the River Severn. The problem is that the terminology used by those who wrote our records is rarely precise enough to tell us exactly who was being called up to fight—for instance, the other folc who joined the garrison of London to defend it in 895 need not be read as “common folk,” but as men required to do service as soldiers who were not “on duty” at the time, or as commoners who aided in various capacities other than combat. London appears not to have been included in the Burghal Hidage, perhaps because it was on the boundary between former kingdoms. Despite the problems and uncertainties, we can be sure that it was the wealthier men who were expected to serve as warriors. Thus the military system required the service of a large proportion of the male population of England, but the fighting was (preferably) done by a minority. As in the case cited in Germany above, when “support troops” did encounter the enemy they were not up to the task. For instance, ceorls working on a fortification proved totally incapable of defending it against the Vikings in 893.
Whatever the exact functioning of the system, what we do know of it does not suggest a mass peasant levy. There are no signs of exemption for harvesting periods and the like, which suggests that the levy was not primarily made up of those who labored on the farms. The need to cultivate the land cannot have been the primary reason for leaving some of the men subject to the levy at home for half the year, as this would still have seriously depleted the agricultural labor force when it was most needed. The idea was presumably that some thegns remained at home to keep order and ensure that normal farming activities could be maintained without internal disorder or external attack. At the same time the system ideally provided a fighting force that was sufficiently large as a field army, but not so large that it could not be provisioned for any length of time.
Alfred’s (or Edward the Elder’s) reforms represent an attempt to exact military service from a wide class of landowners. As in the case of the contemporary Franks, in the burhs and on campaign the troops are likely to have been commanded by officers of the royal household. The reform therefore represents a significant shift from a system in which the king was largely dependent on the highest nobility, who were summoned by him and then turned up with their households and any other men they chose to bring, which they then commanded. We know that there was some opposition to Alfred’s reforms: in the 870s there had been some who preferred to submit to the Danes, and there must have been many landowners who were displeased with the measures he took after Edington. It was presumably the scale of the threat to Wessex that enabled Alfred to carry out his reforms, which may have involved seizure of Church property as well.
Alfred’s reforms did not make the king less dependent on the nobility, but created a greater bond between many of them and his court. Those who served the household directly were provided for and in return would have owed service. Others may have been granted land in return for service. All this increased the status of the king. However, as Richard Abels made clear in his work on lordship and military obligation, many nobles were “sub-contracted” to raise fyrdmen in districts where they, and not the sheriff or other exactor of royal service, had jurisdiction. One such was the Bishop of Worcester, who held one hundred of three hundred hides at Oswaldslow in this way “by a constitution of ancient times” when the Domesday Bookwas compiled. Elsewhere abbots and other churchmen, as well as thegns, had similar rights, if generally over less lands and inhabitants.
There were thus two groups of fyrdmen: those who held their land of the king by book-right and had rights of jurisdiction by royal favor, and those who held land as a loan from another lord or under his seignory. The first category of fyrdman was heavily punished for not attending a summons by the king, usually with loss of his possessions. Although the word fyrdwiteappears in the Laws of Ine, it is not clear what this means, and no reference to it as a fine for failure to perform fyrd service (or as a commutation of it) imposed by the king occurs before Cnut’s time. Thereafter the term appears only rarely, suggesting that its use was infrequent. The second category of fyrdmen was not directly the king’s concern: all that concerned him was that the lord who was obliged to provide a certain number of men per hide fulfilled this provision one way or another. Abels emphasized that the situation in Worcestershire may not have been typical of England as a whole, but there are indications that it was similar to some other shires, if not necessarily the majority. There would have been important consequences of such a system—firstly, many of the king’s most important lieutenants would have had to command their own contingents in the battle line and could not have stayed close to the king, and secondly, a considerable number of those provided by lords with their own jurisdictions may have been mercenaries, paid for to make up the quota for their hides, but not necessarily levies of the hides. Such commands would also have perpetuated the existence of regional contingents that had a distinct identity and perhaps ancient privileges as to position in the battle line. John of Salisbury claimed that the men of Kent held the right to strike the first blow, and the men of London the right to protect the king.
The “men of London” may in fact have been the housecarls. This group of warriors has attracted a great deal of attention, without any certainty as to who they actually were and how they served the king. They have been seen as a form of military brotherhood like the Jomsvikings (in any case of doubtful historicity) or as an early standing army. Alternatively, it has been suggested that there was little difference between them and other thegns. There is no doubt that the term “housecarl” is of Scandinavian origin, and that Cnut’s conquest altered the nature of the king’s household. Even in Domesday Book sixteen of the eighteen thegns recorded as landowning housecarls had Scandinavian names, but this does not necessarily mean that that a new type of guard corps was imported wholesale to help Cnut maintain his rule over a conquered land. The question may simply be one of a change in terminology, as we know that previous English kings maintained royal thegns, for which “huskarl” may have been a translation. Domesday Book may similarly have used “housecarl” simply as a synonym for royal thegn. Like Cnut, his predecessors as king had sometimes had to use their followings to enforce their rule, and some of these were Scandinavians, at least at the courts of Edgar and Æthelred II. If not quite to the same extent as in heroic poetry such as The Battle of Maldon, chroniclers such as Asser probably gave a somewhat idealized view of the relations between Anglo-Saxon kings such as Alfred and their followings, representing them in terms of gift exchange rather than the hiring of soldiers. The import of men from “outside” had advantages, as they had no link to vested interests in the kingdom.
Housecarls, perhaps like the royal thegns before them, appear to have had a special relationship with the king and his immediate entourage. For instance, the “housecarls of her son the king” protected Queen Emma in 1035, and they were used to impose taxes by King Harthacnut in Worcestershire in 1041. As such they were provided with high-quality equipment, would have been well trained and must have lived within close proximity of the court. In this sense they were an elite and may have received a royal stipend, but their status need not have differed significantly from earlier royal household thegns who served as soldiers.
Despite the great outpouring of material that stresses supposedly pan-Celtic traits in warfare, religion and culture generally, it is debatable whether the remaining Celtic-speaking regions of the Viking Era—Ireland, the Pictish and Scottish kingdoms, Strathclyde, Wales, and Brittany—had much more in common with each other than they did with their Germanic neighbors, except their linguistic affinity. The evidence for military organization in all these regions in the early Middle Ages is sparse, but there is enough to show that there were many similarities with Frankish and English kingdoms.
The military household of the Welsh kings was the teulu, rendered familia in Latin, like the households of their English and later Norman neighbors. In early sources such as the Triadsand the Gododdin, we also find gosgordd: if not the same as the teulu, it may have been the household generally as opposed to the personal bodyguard. Here the term teulu will be used to mean “household.” As in the case of other early medieval households, the size of the teulu must have varied. Literary references sometimes give numbers in the hundreds, while the Brut says that Gruffudd ap Llywelyn lost 140 men of his teulu in 1047, clearly not all of them. However, even if this was not an exaggeration, it must be noted that he was a ruler of exceptional power (in Wales) at this time. Earlier sources give lower numbers—for instance, the ASC mentions that Brocmail of Powys fled the field at Chester with his fifty men. At Mynydd Carn in 1081 Trahaearn ap Caradog was killed alongside twenty-five of his teulu when many had already died in the battle, which indicates a similar strength.
Like the households of other Viking Age rulers, the teulu could include men of all ages and social rank, including leading nobles, with the probable exception of unfree bondmen. The members had horses, although their equipment may have varied according to wealth, and many had “squires,” such as the daryanogyon (shield-bearers) who attended uchelwyr, the aristocracy. These positions became offices, and were certainly of pre-Norman period origin. The teulu could also include foreigners, hostages from the households of other nobles or rivals, and, like the Frankish households, sons of nobles sent to receive training. Gruffudd ap Cynan seems to have employed an exceptional number of Irishmen, but it is not certain that these were thought of as teulu members: for instance, they are probably the “pirates” blamed for excessive ravaging in the Life of St. Gwynllyw.
The person responsible for mustering the teulu was the penteulu (head of the teulu), who was supposed to accompany the king at all times except when sent on a specific errand. Naturally, the office of penteulu had great prestige and the laws list many privileges, including a larger dwelling than others and a share of the proceeds from judicial proceedings paid to the king. Other members were appointed subordinate commanders of the llu, the military levy, should it be summoned. The teulu maintained the tradition of the comitatus, and Welsh poetry reflects the same traditions as Old English “heroic” poetry—it was supposed to stand by its lord to the last, as indeed Trahaearn’s seems to have done at Mynydd Carn. On the other hand the Triads describe the two war-bands which abandoned their leaders on the eve of battle as earning everlasting infamy. Since the teulu was also supposed to follow its lord even if he rebelled, this gave any lord the power to mount a rebellion, with consequent disruptive effects on Welsh society. A Welsh ruler had to secure the support of his nobles (uchelwyr)—as elsewhere in Europe, whatever the laws said about a king’s rights to summon his nobles and their followings, their support had to be earned. Many Welsh rulers ran into trouble while trying to impose their will on their subject lords. In 984 Einion ab Owain of Deheubarth was killed trying to do so.
As noted, the usual word for the general levy was llu, but a variety of other terms were used for “army,” such as byddin. Usually the levy was drawn only from certain districts, but Gruffudd ap Llywelyn may have been able to assemble a host from the whole of Wales at the height of his power, if this is what John of Worcester meant by “his whole realm.” Numbers are almost impossible to assess. Historia Gruffudd vab kenan gives figures that appear realistic, the highest being in the hundreds. Davies estimates his army as something under 1,000 in 1075, as he is said to have had 160 men of Gwynedd and numerous foreigners (perhaps 500), especially Irish and Danes, and suggests that the rival armies at Mynydd Carn were each over 1,000 men. Earlier, at the height of his power in 1055, Gruffudd ap Llywelyn had enough men to trouble the English and to force Edward to assemble the host of all England.
Although Wales was never unified, it did undergo a process of reduction of the number of kingdoms in the early medieval period, as happened everywhere in Europe. Several scholars have found evidence of a change in military organization around the beginning of the Viking Period: for instance, the abandonment of some political centers, including hill-forts. In poetry of this period there are hints that free men (eillion) were being called up for service. In the laws supposedly codifed by Hywel Dda (d. 950) freemen have to attend musters and pay duties. Not all the codes imposed this duty on free men, and while some spoke of a “levy of the land,” this must in practice have been selective, as in Anglo-Saxon England. Presumably the king called up those he needed. The Brutiau claim that Maredudd ap Bleddyn gathered “about four hundred kinsmen and comrades and a teulu of theirs” for a campaign in 1118. The same source makes one of the very rare references to a general levy (of Meirionnydd) in 1110, which may even have included bondmen. With the exception of one referred to in the Life of St. Cadog (a fifth-century saint), all general levies are recorded as meeting with defeat. Bondmen had certain duties to construct encampments for the king’s men and to provide pack-horses (and supplies?), but it is unclear whether they were expected to fight or not.
In Wales mercenaries (alltud—“aliens”) were certainly used in many armies. Apart from the Irish mentioned above, John of Worcester mentions Viking mercenaries employed by Gruffudd ap Llywelyn in his attack on England in 1055. The laws also mention the billeting of armed foreigners on the local populace.
There is little evidence that Church lands were exempt from military service in Wales. There are hints that the early medieval clas church functioned in a similar way to the Church in Ireland, its senior men maintaining retinues and providing military service. Given the amount of land owned by the Church and the limited resources available to any ruler, exempting it would have made the raising of forces almost impossible. Most claims to exemption are made in later saints’ lives and many of these are dubious.