For a long time the housecarls passed for England's first professional army: Danish elite warriors brought over by Cnut the Great, with their own law, their own court and regular pay, who died around their fallen king at Hastings in 1066. That picture has two weaknesses. Its law survives in a Danish text written some 150 years after Cnut, and the English evidence mostly shows something else: men who collect taxes, hold land and found monasteries. This article sets the evidence side by side and separates what eleventh-century sources can bear from what the twelfth century and modern scholarship made of them.
Old Norse húskarl means literally "house man". Karl is related to Old English ceorl and denotes a free man, not a slave. According to Foote and Wilson, whom I know only at second hand, it began as an everyday term: the manservant as opposed to the húsbóndi, the master of the house, someone in another's service by choice. Alongside that came the narrower sense that concerns us here, the retainer of a lord.
The runestone at Turinge in Södermanland (Sö 338) shows that sense at work. It commemorates a Þorsteinn who fell in battle "austr í Gǫrðum", in the east in the land of the Rus, as leader of his lið, his troop. The brothers named on it are praised as the best of men at home and on campaign, who treated their húskarlar well. A family that keeps housecarls is not a royal corps but a great man with his people. The term, then, was not tied to kings.
The word reached England as a loan. English writers use housecarl or huscarl; German has settled on Huskarl or Hauskerl. Our article on lið, hirð and leiðangr covers the other words for following a lord. This one pursues the single question the word has provoked: did eleventh-century England have a professional army?
The most vivid evidence shows housecarls not in battle but collecting tax. King Harthacnut, Cnut's son, imposed an exceptionally heavy tribute in 1041. At Worcester on 4 May two of his collectors took refuge in the upper room of a monastery tower and were killed there. Their names, Feader and Thurstan, and the date come from John of Worcester, a chronicler of the early twelfth century. The Anglo-Saxon Chronicle is briefer:
"This year also ordered Hardacnute to lay waste all Worcestershire, on account of the two servants of his household, who exacted the heavy tribute. That people slew them in the town within the minster."Anglo-Saxon Chronicle, s.a. 1041; public-domain nineteenth-century English translation as given by the Avalon Project, Yale (the page names no translator, probably Ingram/Giles)
"Servants of his household" hides the key word. Benjamin Thorpe's equally old translation renders manuscripts C and D here as "his two housecarls", and Robert Cohen's Oxford thesis also speaks of housecarls at this point, drawing on the Chronicle and John. The same goes for the note that Queen Emma, called Ælfgifu in the Chronicle, stayed at Winchester in 1035/36: one translation puts her with the king's "household", Thorpe says "housecarls", and the Old English dictionary of Bosworth and Toller cites the passage (manuscript E) as evidence for hús-carl. So the Chronicle itself knows the loanword. I have not seen the manuscripts, only these renderings. Anyone reading these passages should know that some old translations blur the term.
The reprisal was large. According to John, five earls, English and Scandinavian, set out with almost all the king's housecarls to kill the culprits and burn the city. Most inhabitants escaped in time. Cohen calls it the first recorded tax rebellion in England. Two observations matter here. Housecarls did administrative work out in the country. And when things turned serious they could be mustered as a large group. Both fit a royal household without yet proving an army in the modern sense.
The old story starts two decades earlier. Once Cnut had prevailed in 1016 (the story is told under Cnut the Great), he raised an enormous tribute in 1018: according to the Chronicle £72,000 across the country and £10,500 from London alone. Part of the army sailed back to Denmark, and 40 ships stayed with the king.
Those forty ships are well attested. Whether their crews were "housecarls" the entry does not say. That equation comes later. Sven Aggesen relates in the twelfth century that Cnut admitted only men who could show their wealth with gilded axe heads and sword hilts. The figure of 3,000 men often linked with this is attributed to him in the literature; I could not consult his Latin wording. Modern figures such as 3,000 to 4,000 bodyguards probably rest on reasoning of this kind. To my knowledge no contemporary headcount for Cnut's housecarls exists.
The glamorous image rests on one text, the Lex Castrensis sive Curiae, "the law of the retainers or of the court". In Danish it is Vederloven, from Old Danish withærlagh, in older literature also Witherlogh. According to the Danish encyclopaedia lex.dk the word means "fines" or "punishment". The text runs to about 1,540 Latin words in fourteen chapters and narrates more than it decrees.
Cnut, the story goes, imposed the rules on his men with the help of Øpy the Wise of Zealand and his son Eskil. An offender is moved down the finely graded seating order at the king's table and is thereby lastingly dishonoured. A special assembly is to settle disputes, called the Huskarlestævne in secondary literature. The king is bound too: he must be gracious and pay on time. Then comes the dramatic turn. Cnut himself kills a retainer, the band deliberates and has him prostrate himself before his throne. They absolve him, and the law is tightened: an offender is to die or, in a milder form, leave as an outcast, branded with the shameful word Nithingsorth.
In 1902 Laurence Larson built his influential account of the English royal household heavily on this report. From there the brotherhood with its court and table order passed into the handbooks.
Sven Aggesen came from the Danish high nobility; his uncle Eskil was archbishop of Lund. By his own account he wrote at the request of King Cnut VI and Archbishop Absalon. The work is dated to about 1181/82 or more broadly to the years 1182 to 1201; the accounts differ. Either way, roughly a century and a half separates it from Cnut the Great. Saxo Grammaticus took up the material as well and set it in a scene at Cnut's court.
Sven claims to be translating a rediscovered ancient law. Recent scholarship does not believe him:
Why go to the trouble? The answer lies in the Denmark of the 1180s. Valdemar I had left hard conflicts with the nobility behind him, Sven's family among them, and Cnut VI sought reconciliation. A text that presents the duties between king and court nobility as ancient law, and shows retainers sitting in judgment on a king, was politically useful at that moment. Kjær reads it as a programme for a reciprocal relationship between crown and aristocracy. According to Thomas Heebøll-Holm, Sven's phrase about the courtiers of old, "qui et nunc militari censentur nomine", those who now also bear the name of knight, shows that he had his own present in mind.
One gap remains. Lex.dk mentions, alongside the Latin versions, a Danish one from about 1180. How it relates to Sven's text I cannot decide from what I had before me. As evidence for Cnut's court around 1020 it is at any rate no better than Sven himself.
The attack on the English picture came from Nicholas Hooper. His paper "The Housecarls in England in the Eleventh Century" appeared in 1985 in the proceedings of the Battle Conference and again in 1992 in a collection on Anglo-Norman warfare. I have not read it, only reports of it, which agree on the main lines. Hooper examined above all the Chronicle and Domesday Book and found the term housecarl used with little consistency. The men can hardly be told apart from Anglo-Saxon thegns, the land-holding service nobility. They were rewarded with land or money and did not form a standing force. Better training and esprit de corps he did not rule out. His much-quoted sentence, as reported, is that it is time "to debunk the housecarl".
The English props of the law have suffered as well. Two Chronicle entries were once read as trials before a housecarl assembly. The argument depended on words such as niðing and stefn, translated in a strongly legal sense, and a popular survey in the magazine All About History now treats that reading as outdated. It does not say which entries are meant. An obvious candidate is 1049: in manuscript C "the king and all the army" declare Earl Swein a "nithing" in Thorpe's translation, while the Avalon text says "outlaw". That identification is my own guess, and I do not know the second entry.
Richard Abels, who has worked on English household warriors, sums up the position briefly: much of what we think we know about housecarls comes from late and untrustworthy evidence.
A man in Dorset shows what a housecarl could look like. In a writ, a short royal instruction, Edward the Confessor grants his huskarl Urk the shore in front of his land and everything driven onto it. The document (Sawyer 1063) is dated 1053 to 1058 and Florence Harmer judged it probably authentic. It is addressed to the bishop, Earl Harold, the sheriff and the king's thegns in Dorset.
Urk or Orc is no unknown. He is regarded as founder of Abbotsbury Abbey, drew up the rules of a guild of St Peter, and with his wife Tola, said by an older monastic history to come from Rouen, held considerable property. A charter of Cnut from 1024 (Sawyer 961) gave him seven hides at Portisham. That monastic history calls him Cnut's steward or housecarl, and the foundation date varies by authority between 1026 and 1044. This man served several kings over decades and sat on his own land. The picture fits Hooper's findings far better than a barracks.
He remains a single case. Domesday Book, according to the literature (Mack 1984, here only at second hand), names just 33 land-holding housecarls for 1066, mostly with small estates. How many others are recorded as thegns, or not at all, nobody knows. That is precisely the difficulty when two words overlap.
The pay figures look the firmest. Two things must be separated here that were long run together. The heregeld, the "army tax", began by the usual reckoning in 1012 and paid for a mercenary fleet; Cohen sees the £21,000 of 1014 as the first certain payment. The Chronicle gives the rates: under Harold Harefoot 16 ships were paid at eight marks, in 1040 Harthacnut ordered the same for 62 ships, and 1041 has £21,099 and £11,048 for 32 ships. The unit is itself a nice problem of translation: the old version says eight marks per "steersman" or "rower", while Cohen, following a newer edition, writes "to the rowlock".
The recipients are called lithsmen in the sources from 1035, literally fleet-men, and they are not simply housecarls. Cohen explicitly rejects the equation, although his reasoning was not in the part of the text available to me. According to manuscript C Edward paid off nine ships in 1049 and all the rest in 1050; manuscript E places the nine ships in 1050. According to manuscript D the tax ended in the thirty-ninth year after its introduction, which Cohen reckons as 1051. According to Hooper, as quoted by Cohen, the lithsmen then disappear "lock, stock and barrel" from the sources.
Housecarls do not disappear. Urk bears the title in a writ from after 1053, and Domesday Book still reckons with them for 1066. Their upkeep must therefore have worked differently. Cohen stresses that Domesday attests further levies until 1066, among them one for housecarls as "the king's household military retinue". I could not read that section itself. The rest is familiar: land as with Urk, gifts, a place at table. Saxo speaks of monthly wages, and according to the Danish texts a man could leave service only on New Year's Eve, the time of the king's gifts. These are voices of the twelfth century and thus weak witnesses for 1020.
M. K. Lawson belongs to the question of scale. In 1984 he wrote on the collection of Danegeld and heregeld, and he disputed with John Gillingham whether the Chronicle's huge sums are credible; Lawson defends them. Cohen estimates the heregeld under Cnut and Harold at perhaps £5,500 for 16 ships. Further articles followed in 1989 and 1990. I could not consult Lawson's monograph on Cnut, so I do not report what he says there about housecarls in detail. More on tribute is in our article on Danegeld and Maldon.
If "standing army" means a permanently paid force, separate from the land and with its own law, the answer for the housecarls on current scholarship is probably no. The law is Sven's story, the pay is in good part land, and the boundary with the thegn stays open.
A piece of the old thesis survives, though in another place. The lithsmen come closest to what was meant: permanently maintained, paid from a regular tax, traceable in numbers. Thorkell the Tall had already put ships at Æthelred's service in 1012, as our article on Thorkell shows. What stood permanently in England's eleventh-century military was a paid fleet, and it ended in 1050/51. The housecarls, by contrast, were part of a household, something earls and other magnates had too. Even popular literature now doubts they were something fundamentally new under Cnut: they resemble the household warriors of earlier English kings, and what was new was mainly their name and probably their often Scandinavian origin.
They may well have been good fighters. A man who lives off the king is available and can practise. But that remains inference, because no source describes training or drill.
On 14 October 1066 King Harold Godwinson fell on the ridge near Hastings. The course of events is told under Hastings 1066, and what came just before under Stamford Bridge. The only question here: what do we know of the housecarls on that day?
Less than the books suggest. Abels considers the sources' numbers worthless; the Carmen de Hastingae Proelio speaks of 1.2 million English. He himself reckons, in explicitly speculative terms, with perhaps 5,900 men from the shire levy (fyrd) and about 2,000 housecarls, retainers and survivors from the north. William of Malmesbury, a twelfth-century author, even thought, according to Abels, that Harold had mostly paid troops because he had alienated the fyrdmen over the spoils of Stamford Bridge.
The standard scene, a ring of mailed men round Harold's banner standing to the last man, goes back to modern reconstructions such as Stephen Morillo's. The Bayeux Tapestry shows helmeted foot soldiers in mail in the shield wall and with long two-handed axes; Abels thinks they may be meant as housecarls. They are not labelled so, and he warns that the Tapestry stresses elite warriors because its audience cared about them. In the extracts of early accounts I have seen, no distinct housecarl unit is described. That does not prove it was absent, since Norman authors had no interest in English service titles either. That Harold had his household warriors with him remains the most likely reading. For formation and weapon see the shield wall and the Viking axe.
The afterlife is thinner still. The statement that surviving housecarls strengthened the Varangian Guard in Byzantium appears in popular accounts. I found no evidence for housecarls in the narrow sense, only the general question of Englishmen in Byzantine service; on the Varangians see Miklagard and the Varangians.
What does housecarl mean? Literally "house man", from Old Norse húskarl. The word could mean a free manservant as well as the retainer of a king or magnate.
How many housecarls did Cnut have? This cannot be clearly established. The often quoted 3,000 is attributed to Sven Aggesen in the twelfth century; I have not checked his wording. Only the 40 ships of 1018 are securely attested, with no word on who sailed them.
What is the Lex Castrensis? A Latin text by Sven Aggesen from around 1180/1200 that claims to reproduce Cnut's law for his retinue. It says much about the twelfth-century Danish court and little that is certain about Cnut.
How did a housecarl differ from a thegn? A clear line is hard to draw. Both could serve the king and hold land. Hooper considers them barely separable; housecarl was at first more often the title of men of Scandinavian origin.
Did housecarls fight at Hastings? Very probably, if only because Harold had his household with him. Their number and position are modern estimates.
Lið, hirð and leiðangr · Cnut the Great · Hastings 1066 · Stamford Bridge · Danegeld and Maldon · Thorkell the Tall · Emma of Normandy · The shield wall · The Viking axe
Lið, Hird & Leiðangr · Cnut the Great · The Battle of Hastings · Viking armour: what graves, laws and sagas really tell us about shields, helmets, mail, lamellar and leather.
All History & Finds articles →
Ibn Fadlan was the secretary of an embassy sent by the Abbasid caliph al-Muqtadir from Baghdad to the Volga Bulgars in …
Read more →Around the year 965 a stranger came to the Schlei. He came from Tortosa in al-Andalus, Muslim Spain, was Jewish, a …
Read more →Few peoples of Iron Age Britain left so many traces and at the same time vanished so completely behind a single person …
Read more →Soon after the year 205, a defeated army did not march home – it sank to the bottom of a quiet lake in eastern Jutland …
Read more →