Edward now had Godwine where he wanted him and pushed his advantage, declaring him an outlaw and all his sons with him, around the beginning of October. This message was probably conveyed to Godwine by Bishop Stigand of Winchester, who appears to have acted as an intermediary in these negotiations. He was skilled in diplomacy and trusted by both sides and was to play a similar role in 1052. As a royal bishop he was one of King Edward’s servants, and as current Bishop of Winchester and previously Bishop of East Anglia he would be well known to Godwine and Harold. Edward may even have granted the Godwine family five days’ safe conduct to leave the country, in order to encourage their flight. The considerable organization involved in the exile of the family is a possible confirmation of this.32
Godwine, his wife Gytha and his sons Swein, Tosti and Gyrth all boarded ship at their estate of Bosham in Sussex, taking with them a great hoard of treasure, and crossed to Flanders. They were assured a friendly reception, not just as another group of English refugees like many before, but as a result of Tosti’s recent marriage to Count Baldwin’s sister, Judith. The store of treasure they had taken would not only provide for their own financial needs, but more importantly would pay for Flemish mercenaries to support an attempt to return to England.33
Meanwhile, Godwine’s other sons, Harold and Leofwine, went west to Bristol, avoiding interception by the forces of Bishop Ealdred of Worcester. There they took a ship, initially prepared by Swein for his own anticipated exile, and sailed from the river Avon for Ireland. The likely purpose of Harold’s journey there was also to recruit mercenaries, this time Norsemen from Dublin, to support an attempt by his family to return to power in England. After a stormy voyage, during which they lost a large number of men, the brothers reached Ireland. There they were received by Diarmait Mac Mael-Na-Mbo King of Leinster, who currently dominated the Viking city of Dublin and was to take possession of it directly in 1052. Here Harold’s first efforts at diplomacy were to prove successful. King Diarmait was to provide Harold with aid and treat him well enough for his sons subsequently to seek refuge at his court in the period after Hastings. It was possibly on this occasion that Harold collected the holy relic, the mantle of St Brigit of Kildare, which his sister Gunnhild later donated to St Donation’s in Bruges. It was probably also during his Irish sojourn that Harold had an opportunity to review the events of autumn 1051 that had led to the expulsion of his family from England and to learn important lessons that would benefit him in the future. He had seen that it was foolish for an earl, no matter how powerful, to directly challenge the king, that it was important to ensure the loyalty of one’s own followers, that the English nobility had a strong aversion to the risk of civil war, and that it was vital to consider the reactions of the other earls to any action. He realized that the ability to negotiate and compromise were essential skills for any great man.34
In England King Edward’s triumph seemed complete with the family of Godwine exiled. It was a remarkable turn of events, as emphasized by Chronicle D, which states ‘it would have seemed remarkable to everyone in England if anybody had told them that it could happen’. Edward was now free to deal out the spoils among his supporters. He probably retained most of Wessex for himself. He rewarded Archbishop Robert by finally expelling Spearhafoc from London in favour of William, a Norman priest. In return for Earl Leofric’s support, his son, Aelfgar, received most of Harold’s earldom. In return for his support, Earl Siward probably received Huntingdonshire and Cambridgeshire from Harold’s earldom. Odda of Deerhurst, a kinsman of Edward, was given an earldom in the south-west with lands taken from Godwine and Swein. Earl Ralph, Edward’s nephew, was re-established in his possession of Herefordshire, and other shires taken from Swein. The Frenchmen of Herefordshire were rewarded with the manors of Burghill and Brinsop in Herefordshire worth some £28. The suggestion that these changes were designed to secure or facilitate a Norman succession seems unlikely given the location of the grants made by Edward, which were all away from any direct Norman invasion routes. The list of those rewarded as a result of King Edward’s triumph certainly included a number of Frenchmen though not necessarily Normans, and they were not shown any particular favour, compared, for example, to the Englishmen Siward, Aelfgar and Odda.35
Freed from Godwine’s influence at last, King Edward now turned his attention to one of the main reasons for disposing of him. He repudiated Queen Edith and deprived her of all she owned, placing her in the keeping of the Abbess of Wilton. Then Edward and Archbishop Robert made arrangements for separation and divorce, since their primary concern was to provide Edward with an heir of his own, rather than to secure the succession of William of Normandy. The author of the Vita Eadwardi placed responsibility for Edith’s expulsion from court on Robert, Archbishop of Canterbury. The archbishop, as head of the English Church, would naturally have an important role to play in any royal divorce. The need for his support was a powerful reason for the king, finally, to concede the bishopric of London to Robert’s choice, William, and sacrifice his own nominee, Spearhafoc. It should be recalled here that the Vita Eadwardi was written, in part, to celebrate the marriage of Edith and Edward. The author could not therefore portray the king as having himself contemplated a divorce and hence we have the story of the honourable pretext of Edith being sent to Wilton only for her own safety. This circumstance also explains the Vita Eadwardi’s hostility to Archbishop Robert, since Edith could hardly have liked someone who had participated in an attempt to deprive her of her royal status. In both real life and as recorded in the Vita Eadwardi Robert of Jumieges was to fulfil the role of scapegoat for Edward – not only in this attempted divorce but also in the whole episode of the exile of Godwine – when it was in fact the king who had initiated both of these processes himself.36
THREE
EXILE AND RETURN
And forthwith Earl Harold came from Ireland with his ships to the mouth of the Severn near the boundary of Somerset and Devon. . . .1
Chronicle D states that it was during the exile of the Godwine family late in 1051 that William of Normandy visited King Edward, though his purpose in doing so is not recorded. This visit has been linked, although not by the Chronicle itself, to later Norman accounts of William’s claim to the English throne, and it is regarded as the likely occasion when King Edward promised the throne to William. However, a number of factors suggest that this episode may be a later interpolation in the Chronicle, made after the events of 1066. In the first place, William Douglas, the foremost authority on William of Normandy, considers such a visit extremely unlikely given the latter’s preoccupations in Normandy at the time. More importantly, the Norman writers William of Jumieges and William of Poitiers fail to record any such visit, in spite of the fact that it could obviously have been employed to reinforce William’s claim to the throne. It seems likely therefore that this visit did not in fact occur, and that the Chronicle record here has been adjusted at a later date.2
In the absense of any actual visit to England by the duke, the Norman accounts of King Edward’s designation of William as his heir stand alone as evidence in placing this event during this same period of eclipse for the Godwine family. A number of post-Conquest English sources relate a similar story, but these all derive from the same original Norman accounts. In the earliest of these, William of Jumieges states that Edward sent Robert, Archbishop of Canterbury, to William to nominate him as heir to his kingdom. This basic outline is enlarged upon by William of Poitiers, who states that this offer was made to William in gratitude for the refuge given to Edward in his early years and came with the assent of the English nobility, specifically earls Godwine, Leofric and Siward, and ‘Archbishop’ Stigand of Canterbury. William of Poitiers adds that Edward sent to William, by the agency of the same Archbishop Robert and at the same time, the son and grandson of Earl Godwine, Wulfnoth and Hakon, as hostages to guarantee the agreement. Although they give no dates, these sources clearly place the events, by association with Archbishop Robert, in the period be
tween his promotion to this office in March 1051 and his death in exile sometime between 1053 and 1055.3
Neither of these Norman sources are contemporary in the way that the Chronicle is, nor can they be described as unbiased. Indeed, both authors explicitly make clear that their purpose in writing was to justify William’s conquest of England and his succession to the English throne. No contemporary English sources support their account, which is particularly surprising given that William of Poitiers portrays the designation of William as a widely acknowledged event in England. Nevertheless, it is essential that the evidence they offer is fully considered, as this event was to have important consequences for Harold. The account of William of Jumieges, as we have seen, provides little more than a few basic facts. The account of William of Poitiers elaborates considerably on these facts, but in doing so introduces some inconsistencies, which tend to undermine confidence in the chronicler’s reliability. For example, he names ‘Archbishop’ Stigand among those who assented to William’s nomination, when the latter was in fact only Bishop of Winchester at the time.4
The problems with the Norman accounts make it very difficult for a clear narrative to be established which is consistent with other sources, although many are prepared to accept their account without such consistency. If we are to establish the full background to William of Normandy’s invasion of England in 1066 we must attempt a reconciliation between the Norman and English accounts. This involves a need to establish where they come together and where inconsistencies exist. The accounts then need to be reviewed to see which elements within them can be reconciled. This also involves the need to consider the likelihood of events which are described by the Norman writers alone, emerging from contemporary circumstances in England. If we can establish an account which allows the main events of the two traditions to be reconciled, then this is probably as close to the truth as we are ever likely to get.5
The main points of contact between the Norman and English accounts are Robert, Archbishop of Canterbury, and the hostages from Godwine’s family. In most accounts of these events the obvious time for Robert, Archbishop of Canterbury, to communicate Edward’s designation of William as his heir, as recorded by the Norman accounts, has been seen as during the course of his journey to Rome for his pallium, between his appointment in March 1051 and 29 June 1051 when he was enthroned on his return. This is a period when he is known to have been abroad, and he may have passed through Normandy. However, if this was the case, then he could not have brought Godwine’s hostages to William at the same time, as William of Poitiers clearly states. According to the contemporary Anglo-Saxon Chronicle, these hostages were given into King Edward’s hands, as a pledge for Godwine’s appearance at the London council, in September 1051, long after Robert had returned to England in June. This undermines the perceived sequence of events as established by William of Poitiers’ account, but the anachronism concerning the hostages extends beyond one of timing. It would be contrary to custom for King Edward, as donor of the kingdom, to send hostages to William, his suitor, without the latter sending hostages in return. Also, if hostages were expected from the English guarantors – Godwine, Leofric, Siward and ‘Archbishop’ Stigand, as named by William of Poitiers – then why were they only received from one of these parties? This all serves to undermine the additional details provided by William of Poitiers. In contrast, William of Jumieges provides a much briefer account which shows no such inconsistencies and therefore appears more trustworthy. They must both now be assessed against contemporary events in England to gauge whether they can provide a plausible scenario that fits these events.6
The key factor in the events of 1051, which is not dealt with by the Norman accounts, is the background provided by Edward’s childless marriage. The succession to the English kingship in the eleventh century rested on a number of factors. The first was a natural right of inheritance usually but not always through primogeniture and based on family relationship with the existing or a previous monarch. In this way, sons or sometimes brothers of the reigning monarch were recognized as potential heirs to the throne, often being given the title atheling to signify their throne-worthy status – thus the references to all of King Aethelred’s sons, Athelstan, Edmund, Eadwig, Edward and Alfred, as atheling. In cases where a king had a number of sons or brothers, the selection of a single heir to the kingdom might be secured by his designation, usually but not always given to the eldest son. Thus Cnut is said to have favoured his younger son Hardecnut before Harold ‘Harefoot’. This designation by Cnut was not in fact enforced, as we have seen above, since it was not supported by all the great men in England. The support of these men could often sway a disputed succession, as in Harold ‘Harefoot’s’ case, since royal power usually rested on such support. However, it is important to note that these men could not of themselves choose or elect a king, but only provide support to one of two or more rival candidates. In exceptional circumstances, an outsider could succeed in seizing the kingship by enforcing his authority on the country by conquest. This could usually only be achieved with the support of a foreign army, as under the Danes Swein ‘Forkbeard’ and Cnut. However, even these conquerors often sought to secure support from a portion of the great men to ease their path to the throne.7
In the circumstances of 1051, King Edward had no surviving brothers and no sons by his wife Edith. He therefore required either to find assistance in order to remove Earl Godwine, divorce Edith, and remarry in search of a son of his body, or to find an alternative heir, who satisfied the required criteria. The former might imply assistance from the relatives of a prospective bride. In these circumstances, the question must be whether or not William, Duke of Normandy, was best placed to satisfy either of these requirements in the period 1051–2.
If Edward was seeking an opportunity to remarry in order to have an heir of his own, then he needed a bride and support to remove Godwine, who could not allow Edward to repudiate his daughter Edith unchallenged. Such support would best come from someone with enough power to counter Godwine, and also with an eligible daughter. In 1051, William of Normandy was either unmarried or only recently married and therefore without eligible daughters. He did have a sister, Adelaide, but she was probably already married to Enguerrand of Ponthieu. William therefore had no eligible bride for Edward. Similarly, although he was already a powerful lord in 1051, his energies and military forces were almost entirely engaged in campaigns to ensure his own survival and security as duke. These circumstances would appear to indicate that it is unlikely that Edward sought to form an alliance with William via marriage, which might later be construed as designation of him as his heir.8
Of course, the Norman accounts did not claim that William offered any marriage alliance, but rather that he was designated as an alternative heir. We must now consider whether William met the criteria for an alternative heir for Edward in 1051. When we examine the criteria for a royal heir, we come across a number of possibles among Edward’s relatives, some with stronger claims than others. Earl Ralph, Edward’s nephew by his sister, Godgifu, and her first husband, Drogo, Count of the Vexin, was already present in England and married to an Englishwoman. He was undoubtedly the best heir available, in spite of his French background. Walter, Count of the Vexin, Ralph’s elder brother, had an equivalent claim, but was not present in England and had no direct English connections. Strictly speaking, the claim of these brothers through the female line was weak, but still much stronger than any such claim by William. However, the man with the strongest claim was Atheling Edward, the son of Edward’s older half-brother, King Edmund ‘Ironside’, although at this stage he was exiled in Europe, his exact whereabouts unknown. As an atheling and the son of a former king of England, this Edward’s claim was far stronger than those of either of the king’s French nephews. The crucial point is that any of these men were more likely as heirs than William of Normandy. William’s only tenuous connection to the English dynasty was through his great aunt Emma and he had no English royal b
lood. It is of course possible that Edward intended to abandon traditional practice altogether and appoint someone unconnected to the dynasty. However, if he chose to do so it would represent a major breach of tradition and as such it seems highly unlikely that it would have received much, if any, support in England. William was not of royal blood, a foreigner, a relatively unknown quantity in England, and still struggling to hold his own duchy at home.9
Certainly, King Edward was half-Norman on his mother’s side but he spent his formative years, up to the age of fourteen, being raised as an English atheling. He does appear to have developed a fairly cosmopolitan outlook, as demonstrated by his promotion of a number of Continental clerics and laymen. However, Edward’s cosmopolitan sympathies do not appear to have been focused solely on Normandy but instead extended to other areas including Brittany, Flanders and Lotharingia. Thus Edward’s Norman blood would not in itself appear to be enough to explain the designation of William of Normandy as his heir. Indeed, William of Poitiers himself does not seek to rely on this background alone but specifies particular reasons for Edward’s action.10
If the general circumstances were unfavourable for King Edward’s designation of William as his heir, can it be explained instead by the specific reasons provided by the Norman sources? William of Jumieges provides no reasons, but William of Poitiers provides a number of them. He states that Edward designated William as his heir in gratitude, first for providing him with refuge in Normandy in 1016 and then for securing his restoration to England in 1041. However, Edward’s gratitude to William seems unlikely for a number of reasons. It is certainly possible that King Edward felt gratitude for his refuge in Normandy under Dukes Richard II and Robert, William’s grandfather and father respectively. Indeed, Duke Robert appears to have attempted to invade England in 1033 or 1034 on Edward’s behalf, only for the attempt to end in abject failure, as a result of adverse weather conditions, and abandonment in favour of a raid on Brittany. However, any gratitude in this instance, if indeed such a fiasco warranted it, was perhaps due to Robert rather than his son. It is also possible that Edward actually resented his position as an exile, dependent on his Norman relatives, and felt that they had failed to do enough to assist him. This possibility is perhaps suggested by the lack of any major concessions to William or any other prominent Norman laymen in the early part of Edward’s reign. Indeed, it appears that Edward and his brother, Alfred, had been forced to rely on forces from outside Normandy to support their own invasion attempt in 1036. Subsequently when Edward had returned to England in 1041, it was at the invitation of his half-brother King Hardecnut. A fact accepted by William of Jumieges and at one point also by William of Poitiers. There is no contemporary evidence that Edward received any Norman aid to secure his return. In all this, therefore, it seems that there are insufficient reasons for Edward to make such a grand gesture as designating William as his heir out of gratitude to him, especially after a period of some ten years during which there appears to have been no contact between them.11
Harold Page 7