Death of Harthacnut in 1042
Emma fleeing England with Edward and Alfred, following the invasion of Sweyn Forkbeard
Whatever Cnut died of, it wasn’t old age. Contemporaries were agreed that he had been very young at the time of his conquest of England in 1016, which has led modern historians to place his date of birth at some point in the last decade of the first millennium. Thus when the king died in the autumn of 1035, he was probably around forty years old (a thirteenth-century Scandinavian source says he was thirty-seven). According to William of Jumièges, he had been seriously ill for some time, and this statement finds some support in a charter that Cnut gave to the monks of Sherborne Abbey in Dorset in 1035, asking for their daily prayers to help him gain the heavenly kingdom. It was at Shaftesbury, just fifteen miles from Sherborne, that the king had died on 12 November.
Given his Viking ancestry, and the bloodshed that had accompanied his conquest, Cnut’s anxiety to enter heaven rather than Valhalla may strike some as surprising. But in fact the Danish royal house had been converted two generations earlier, and Cnut himself had been baptized as a child (his baptismal name was Lambert). Indeed, the point of the famous story about the king and the waves, as originally told, was not to illustrate his stupidity, but rather to prove what a good Christian he had been. ‘Let all the world know’, says a damp Cnut, having conspicuously failed to stop the tide from rising, ‘that the power of kings is empty and worthless, and there is no king worthy of the name save Him by whose will heaven, earth and sea obey eternal laws.’
Cnut had in fact been famous for such acts of ostentatious piety. Having conquered England and dispatched his opponents in the traditional Viking manner, the king had sought to convince his remaining subjects that his rule was legitimate, and this meant, above all, demonstrating that it was approved by God. In 1027, for example, Cnut had gone on pilgrimage to Rome. He had also attempted to salve the wounds inflicted in the course of the Danish takeover – for example, by having the bones of Ælfheah, the murdered archbishop of Canterbury, moved from St Paul’s Cathedral in London to a new shrine at Canterbury; by causing a church to be built on the site of the battlefield where his opponent, Edmund Ironside, had been defeated; and by visiting Edmund’s tomb at Glastonbury, where he honoured the late king’s memory by presenting a cloak embroidered with pictures of peacocks. The giving of such valuable objects was also typical, and helped Cnut secure a good reputation at home and abroad. ‘When we saw the present you sent us,’ wrote the bishop of Chartres, responding to the king’s gift of some beautifully decorated books, ‘we were amazed at your knowledge as well as your faith … you, whom we had heard to be a pagan prince, we now know to be not only a Christian, but also a most generous donor to God’s servants.’
There was nothing incongruous, therefore, when Cnut was eventually laid to rest in Winchester, in the cathedral known as the Old Minster, alongside the bones of St Swithin and several earlier kings of England and Wessex. His reign – almost twenty years long, as the Anglo-Saxon Chronicle noted at the time – had been a success, largely because he had striven to observe and maintain English traditions. Even the few novelties that were once ascribed to Cnut are now reckoned not to have been novel at all. Once, for example, he was thought to have introduced a new breed of warriors from Scandinavia, his ‘housecarls’, to serve as a separate standing army. But closer examination suggests that the housecarls were no different from the household warriors maintained by his English predecessors. Cnut did have a standing army of sorts, since he maintained a permanent fleet of ships with paid Danish crews. But here he was simply following the example of Æthelred the Unready, who had maintained just such a fleet from 1012, and had introduced a new national tax to pay for it. The only difference was that Æthelred’s fleet had been bigger.
Nevertheless, for all Cnut’s determination to portray himself as a traditional Old English king, his reign had altered English society dramatically. Or rather, that society had been altered in the tumultuous period up to and including his conquest.
English society in the eleventh century was highly stratified. We know that there were approximately two million people living in England at the end of the century, and that the population was rising all the time, so there must have been rather fewer than that number at the century’s start. At a fundamental level, these people were divided into two categories: the free and the unfree.
Although many books on the Anglo-Saxons do not say much about it, more than ten per cent of England’s population were slaves. Slavery was a widespread institution in early medieval Europe, and the sale and export of slaves was one of the main motors of the economy. Since the ninth century the trade’s most outstanding exponents had been the Vikings, whose warfare was predicated for the most part on seizing young men and women as merchandise, to be sold either at home in Scandinavia or – very commonly – to Arab merchants in the Middle East. England was one of their principal hunting grounds, so individuals abducted from the coasts of Devon, Wales or Northumbria might eventually find themselves labouring under a desert sun to construct a caliph’s palace, or members of a sultan’s harem.
Slaves were similarly used in England for hard labour and sexual gratification, to judge from contemporary comments. Male slaves were generally used as agricultural workers, and something of the nature of their condition is captured in a celebrated passage written by Ælfric, a late tenth-century abbot of Eynsham, which imagines the speech of an unfree ploughman:
I go out at daybreak, goading the oxen to the field, and I join them to the plough; there is not a winter so harsh that I dare lurk at home for fear of my master. But after yoking the oxen and securing the ploughshare and coulter to the plough, throughout the whole day I must plough a full acre or more … I must fill the stall of the oxen with hay and supply them with water and carry their dung outside. Oh! Oh! The work is hard. Yes, the work is hard, because I am not free.
The ploughman had good reason to fear his master. Slaves were regarded not as people but as chattels, and as such could be punished like animals, by branding or castration. They could even be killed – stoned to death by other slaves if they were male, burnt to death if they were female.9 The purposes for which female slaves were kept are not entirely certain. Many of them were no doubt used as domestics or dairymaids, but several sources suggest that women were also purchased for sexual purposes. In the early eleventh century, shortly before Cnut’s conquest, Bishop Wulfstan of Worcester delivered a famous sermon to the English people, lambasting them for their manifold sins. Certain Englishmen, he said,
club together to buy a woman between them as a joint purchase, and practise foul sin with that one woman, one after another, just like dogs, who do not care about filth; and then sell God’s creature for a price out of the country into the power of strangers.
Above the slaves were the remaining ninety or so per cent of the population who were free. The vast majority of the people in this category were classed as ceorls (or churls), a term we might translate as peasants. They too in most instances worked the land, and most of the time the land they worked was their own. In some areas of England they were less free than in others, because lords had started to insist that they were tenants who ought to perform labour services. But ceorls, unlike slaves, were no one’s property.
Above the ceorls were the nobility, a class that included approximately 4,000 to 5,000 people, or just 0.25 per cent of the total population. The nobility were distinguished from the people below them chiefly by virtue of owning a lot more land. An anonymous tract on status, written in the first quarter of the eleventh century, explains that it was possible for a ceorl to prosper and become a thegn (or thane). But he needed to have a suitably noble residence, with a gatehouse and bell-tower, and at least five hides of land – a hide being roughly 120 acres. This was crucial – it was insufficient simply to strut about in fancy armour. ‘Even if he prospers so that he possesses a helmet and a coat of mail and gold-plated sword,’ the tract continues, ‘if he has not the land, he is still a ceorl.’
To be a noble it was also deemed necessary to have a connection of some kind with the king. For the great majority of thegns this may simply have entailed fulfilling some minor role in royal government – administering a local court or assisting in the collection of national taxes. But for a select few it meant serving the king personally – riding in his household, as the tract explains, or going on special missions. According to a twelfth-century source, the minimum property requirement for entry into this charmed circle of ‘king’s thegns’ was forty hides of land, and based on this figure it has been calculated that there were only around ninety such men in England.
Lastly, at the very apex of aristocratic society, there were the ealdormen. These were the individuals who ran entire regions in the name of the king – East Anglia, for example, or Northumbria. As the king’s immediate deputies in these regions, they presided, twice a year, over the shire courts, handing down judgements of life and death, while in times of war they led royal armies. Because their commands had been created by the kings of Wessex as they had extended their power across England in the course of the tenth century, most ealdormen were themselves descended from the ancient royal line, and related to each other by ties of kinship and marriage.
This society – slaves, ceorls, thegns and ealdormen – had been severely shaken by the Danish invasions in the decades prior to Cnut’s conquest. Naturally the population as a whole had suffered as Viking armies hacked their way across the landscape. ‘There has been devastation and famine, burning and bloodshed in every district again and again’, lamented Bishop Wulfstan in his sermon of 1014. Some slaves, he complained, had run away, abandoning Christianity to become Vikings (and who, wonders the modern reader, can blame them?). Some thegns, who had once fancied themselves brave and strong, had been forced to watch while Vikings had gang-raped their wives and daughters. And all the while the invaders had been doing as they had always done and seizing people to sell overseas. ‘Often two or three seamen drive the droves of Christian men from sea to sea, out through this people, huddled together, as a public shame to us all …We pay them continually and they rob us daily; they ravage and they burn, plunder and rob and carry on board.’
But while everyone suffered from the invasions, no section of society suffered more than the upper ranks of the English aristocracy. Consider, in the first instance, the fate of the ealdormen. The elderly Birhtnoth had been the first of them to fall, dying during the Battle of Maldon in 991; four of his fellow ealdormen had perished during the struggle against Cnut in 1016, and almost all the remainder had been killed the following year as part of the new king’s notorious purge. Then there were the high-ranking thegns, many of whom appear to have met similarly bloody ends: the Anglo-Saxon Chronicle contains frequent references to the large numbers of nobles slain, and this testimony is confirmed by the lack of continuity between the thegns who witness Cnut’s charters and those who attest the acts of his predecessors. Two and a half decades of fighting, in other words, had all but wiped out the highest echelons of the English nobility.
Unsurprisingly, Cnut chose in the first instance to fill England’s depleted aristocratic ranks with Scandinavians. The rank and file of his army had gone home soon after the conquest, satisfied with their share of the great tribute that the new king had exacted at the start of his reign (and, in some cases, raising runestones back home in Scandinavia to celebrate their winnings). But at the highest level, in place of the fallen ealdormen, Cnut appointed a new set of Nordic provincial governors. The greatest of all his supporters, Thorkell the Tall, he placed in charge of East Anglia, while his brother-in-law, Erik, was given the responsibility of ruling Northumbria. Smaller commands were created elsewhere in England for the king’s other captains and kinsmen: a trio of shires in the west Midlands, for example, went to Hakon, Hrani and Eilífr. In their own Norse tongue men of such exalted rank were known as jarls, and the new term was swiftly adopted in the conquered country. England, latterly governed by ealdormen, was henceforth governed by earls.
There was, however, a striking exception to Cnut’s general policy of promoting his Scandinavian friends and family. From the very start of his reign, one of the king’s foremost advisers was Godwine, an Englishman of obscure origins. Probably he was the son of a Sussex thegn named Wulfnoth, an opponent of King Æthelred’s regime who had commandeered part of the royal fleet and terrorized England’s south coast. Was there, perhaps, a connection between this piracy on his father’s part and Godwine’s subsequent rise under Cnut? All we know is what we are told by a tract written in Godwine’s praise half a century later: he ‘was judged by the king himself the most cautious in counsel and the most active in war’. Soon into his reign, having succeeded to the Danish throne after the death of his brother, Harold, Cnut took his new favourite to Denmark, and there too the Englishman apparently demonstrated his indispensable wisdom and courage. The king responded by showering Godwine with honours: as early as 1018 he had been raised to the rank of earl, and not long afterwards he was drawn into the royal family by his marriage to Cnut’s sister-in-law, Gytha.
Such, indeed, was the king’s reliance on Godwine that the Englishman was soon pre-eminent even among England’s new Danish ruling class. By the early 1020s his command had been extended across the whole of southern England, and included the entirety of the ancient kingdom of Wessex. At the same time, the number of Danish earls was steadily declining. Thorkell the Tall was exiled in 1021, Erik of Northumbria died in 1023, and the following year Eilífr disappears from the record. As the decade wore on, other Scandinavians in England were redeployed to fill positions in Cnut’s expanding northern empire. Earl Ulf, for example, was sent at some point to serve as the king’s deputy in Denmark, while Earl Hakon was dispatched to govern Norway after the latter kingdom was conquered in 1028.
During this period, however, Godwine’s supremacy did not pass entirely unchallenged, for into the vacuum created by the disappearing Danes stepped another favoured Englishman. Leofric, son of Leofwine, came from an existing aristocratic family: his father had been the only ealdorman to survive Cnut’s house-clearing, albeit in reduced circumstances, his authority in the Midlands being subordinated to the region’s new Danish earls. But after his father’s death in 1023, and the eclipse of his Danish rivals, Leofric’s own star began steadily to rise. By the late 1020s he too had acquired the rank of earl, and thereafter seems to have become the principal power in the Midlands – what had once been the kingdom of Mercia. The witness-lists to royal charters show that, in the final years of Cnut’s reign, Leofric was second only to Godwine in the king’s counsels.
Thus, by the time of his death in 1035, Cnut had transformed the English aristocracy. The old guard of ealdormen – descended from royalty, close-knit and long-established – were gone, killed off in the course of the bloody Danish takeover. But gone too, for the most part, were the Danes who had initially replaced them. By the end of the reign, most of England was back under the command of Englishmen, with Earl Godwine governing Wessex and Earl Leofric in charge of Mercia; only in distant Northumbria, where Earl Siward had succeeded Earl Erik, did a Dane control an earldom of any consequence. These three earls, however, shared the common quality of being new men. Godwine’s family can be traced back only a single generation, Leofric’s no more than two, while nothing certain at all can be said about the parentage of Siward. Their rapid rise under Cnut had made them immensely powerful – probably more powerful than any English noblemen up to this point. But they lacked the ancient roots of the aristocracy that they had replaced. England’s three new earls were not linked by ties of blood or marriage. As subsequent events would show, they were not partners, but rivals.
The death of Cnut triggered a protracted and extremely bitter struggle. On the most fundamental level, the late king had provided for the succession by fathering no fewer than three healthy sons. The problem was he had fathered them by two different women.
As we’ve already seen, in the year after his conquest Cnut had married Emma – sister of Duke Richard II of Normandy, widow of King Æthelred, and mother of the future Edward the Confessor. Emma was Cnut’s official partner – his anointed queen – and she figures frequently as such in royal documents and devotional artwork. Together they had two children: a son called Harthacnut, said to have been born soon after their wedding, and a daughter, Gunhilda, who had latterly been married to the German emperor.
But some time earlier, perhaps in the course of his father’s short-lived conquest of 1013, Cnut had married another woman called Ælfgifu of Northampton. As her surname suggests, Ælfgifu came from an English family based in the Midlands. An important family: her father had for a time been the ealdorman of southern Northumbria, until he was murdered on the orders of King Æthelred. This raises the strong possibility that Cnut’s marriage to Ælfgifu had been arranged to cement an alliance with a disgruntled faction of Englishmen who had wanted to see Æthelred replaced.
Whether it was to preserve such an alliance, or simply because he enjoyed having his cake and eating it, Cnut apparently took no steps to dissolve his marriage to Ælfgifu before or after his subsequent marriage to Emma. He may have felt there was no need, for it is clear that the first match, unlike the second, had not been blessed by the Church. Whether or not this distinction mattered much to society as a whole, however, is debatable. At this date the laity regarded the Church’s involvement in marriage as an option, not a requirement. The unconsecrated match between Cnut and Ælfgifu was clearly considered as sufficiently legitimate by both parties at the time it was arranged. This in turn meant that the children it produced could be regarded as legitimate as well.
Ælfgifu had given Cnut two children, both boys, called Swein and Harold. They were probably born before the king’s second marriage in 1017 (that, at least, was Emma’s later assertion) and so were probably in their late teens or early twenties at the time of his death in 1035. We hear next to nothing about them or their mother before this date, but one fact alone indicates the high esteem in which they continued to be held. In 1030, after the death of Earl Hakon, Cnut sent Ælfgifu and Swein to Norway in order to rule there as his regents.
Did this indicate some plan for the succession? At some point before 1035 the king had similarly dispatched Harthacnut, his son by Emma, to rule on his behalf in Denmark; indeed, surviving coins show that Harthacnut had begun styling himself as king of Denmark even before his father’s death. Some later chroniclers imagined that Cnut’s intention had been to divide his empire in just such a way, with Norway going to Swein, Denmark going to Harthacnut and England passing to Ælfgifu’s other son, Harold. This, however, is probably no more than historical hindsight, for at the time of Cnut’s death there was no agreement at all.
Soon after Cnut’s death, says the Anglo-Saxon Chronicle, there was a meeting of all his counsellors in Oxford. England already had a long tradition of such assemblies: it is a mark of the kingdom’s political maturity that in times of crisis its leading men would generally come together to debate their differences rather than immediately reaching for their swords. But the decision to meet in Oxford that autumn shows how serious the situation had already become, for the town lay on the River Thames, which in turn marked the boundary between Wessex and Mercia. And, sure enough, when the meeting took place, the two earldoms were divided over the succession. ‘Earl Leofric and almost all the thegns north of the Thames’, to quote the Chronicle, wanted their next king to be Harold. But ‘Earl Godwine and all the most prominent men in Wessex’ declared in favour of Harthacnut.
Godwine was almost certainly the single most powerful man in England, but on this occasion he found the odds stacked against him. We are not told anything about the sympathies or whereabouts of Earl Siward at this crucial moment, though it is hard to imagine he was not present; possibly the Chronicle’s comment about ‘all the thegns north of the Thames’ implies that he also supported Harold. But the Chronicle does tell us that Harold’s candidacy was backed by Cnut’s mercenary fleet in London, a formidable force of several thousand men, and more than a match for the late king’s housecarls, who had apparently declared for Harthacnut. The greatest problem for Harthacnut’s supporters, however, was that their candidate was still in Denmark; Harold, by contrast, was resident in England, probably present at the Oxford meeting, and therefore in a much better position to push his claim.
At length a compromise was reached which recognized the regional split. Wessex, it was agreed, would be held in trust for Harthacnut by his mother, Emma, who was to reside at Winchester with the housecarls. The rest of England, by implication, would be held by Harold, who would also act as regent of the whole kingdom on behalf of himself and his brother. Godwine and his supporters evidently opposed this arrangement but, as the Chronicle says, ‘they could put no obstacle in the way’. Their only consolation was that no firm decision had been taken on who should be the next king: as the Chronicle’s talk of trust and regents implies, the succession was to hang fire until Harthacnut’s return.
But Harthacnut, who had his hands full in Denmark, failed to appear, and the competition between the two rival camps intensified. Each side worked to undermine the support of the other, and no one worked harder than Queen Emma. A few years later, she commissioned a highly tendentious political tract, known today as the Encomium Emmae Reginae (‘In Praise of Queen Emma’), which above all else sought to justify her behaviour during this period. It is the source of the notion, noted in the previous chapter, that her marriage to Cnut had been a consensual affair rather than a fait accompli. The Encomium also claimed, conveniently, that there had been a prenuptial agreement: Cnut had apparently sworn an oath to Emma ‘that he would never set up the son of any wife other than herself to rule after him’. Harthacnut, in other words, was the only true heir; Harold, son of Ælfgifu of Northampton, could have no legitimate claim. Emma also set out to discredit her rivals in less subtle ways. The author of the Encomium assures us that Harold was not actually a son of Cnut at all, but a changeling, taken by Ælfgifu from the bed of a servant. It was crude propaganda, but clearly believed in some quarters: the Anglo-Saxon Chronicle reports the same slur.
Not that Ælfgifu was above playing the same game. It is not entirely certain at what point she returned to England, but her regency in Norway had ended in disaster around 1034 – she and her other son, Swein, had been driven out of the country, and Swein had died not long afterwards. Ælfgifu may therefore have already been in England at the time of Cnut’s death; she was certainly back before June 1036, for at that point we catch wind of her struggle against Emma in a letter written at the imperial court in Germany. Emma had sent messengers to her daughter, Gunhilda, complaining about Ælfgifu’s activities. ‘Your wretched and wicked stepmother, wishing to deprive your brother Harthacnut of the kingdom by fraud, organized a great party for all our leading men, and, eager to corrupt them at times with entreaty and at times with money, tried to bind them with oaths to herself and her son.’ According to Emma’s messengers, Ælfgifu’s wining and dining was unsuccessful. ‘Not only did the men not give their consent to her in any such way; but of one accord they dispatched messengers to your aforesaid brother, so that he might soon return to them.’
But this seems to have been wishful thinking on Emma’s part. There was still no sign of Harthacnut, and meanwhile Harold’s power was clearly growing. We can see as much by looking at the coinage that was in circulation. The English coinage system at this time was highly sophisticated; each coin, as well as bearing the name of the king, also carried the name of the place it had been minted. This means we can not only see at a glance which coins were struck for Harold and which for Harthacnut; we can also, with more considered analysis, see how much of the country each had under his control. What we see at first is power split along the line of the Thames, as had been agreed in the meeting at Oxford. But, as time goes on, the geographical spread of Harthacnut’s coinage contracts, while that of his rival expands. Throughout 1036, it seems, support for Harold was growing stronger. At some point, he sent men to Winchester, and deprived Emma of what the Anglo-Saxon Chronicle calls ‘all King Cnut’s best valuables’ – including, perhaps, the regalia necessary for a coronation. It looked as if the queen’s grip on power, assiduously maintained through her marriage to two English kings, was about to end because of her son’s continued absence. It must have been around this point that she recalled that she had two other sons living in exile across the Channel.