This period of relative calm ended with the death of Earl Aelfgar of Mercia. This event is not recorded in any of our sources, but Aelfgar disappears sometime between 1058, when he is recorded as regaining his earldom, and 1065, when his son Edwin is recorded in his place as Earl of Mercia. If the later Vita Wulfstani can be relied on, Earl Aelfgar was among those who supported Wulfstan’s election as Bishop of Worcester, dated before 27 August 1062, and this appears to have been his last recorded act. That his death is likely to have come in 1062, possibly late in that year, is further suggested by the action taken by Harold after Christmas that year. He launched a lightning cavalry raid on King Gruffydd’s palace at Rhuddlan in North Wales. It is unlikely that Harold could have taken this action unless Aelfgar, Gruffydd’s English ally, had been removed from the scene, and such a surprise attack would achieve best results if news of Aelfgar’s death had not yet reached Gruffydd.
It is likely then that Aelfgar died late in 1062, shortly before Christmas, perhaps while attending King Edward’s Christmas court at Gloucester. Earl Harold, who was also present, then decided to strike against Gruffydd at once, hoping to catch him unawares and kill him. He gathered a cavalry force and rode swiftly over more than 100 miles to Rhuddlan on the North Wales coast. John of Worcester is the only source to mention Harold’s use of cavalry in this raid, but this is undoubtedly the only way such a surprise raid could have been carried out. Harold narrowly missed catching his quarry, who was warned at the last moment and fled by sea. Instead, he had to content himself with burning Gruffydd’s palace and ships. The figurehead and ornaments of Gruffydd’s ship were probably taken in this raid.24
The Vita Eadwardi vividly recalls this raid, which must have achieved wide renown at the time:
The enemy’s house is sacked, the girded chests
Are broached, and royal pomp exposed to loot.
In blaze of glory, ably led, the men
Return, and bring back this fine ornament:
They smashed the fleet – for Welsh control and lore
Was not the equal of the Ocean’s chiefs –
And take a prow and stern of solid gold,
Cast by the smith’s assiduous skill, and this,
With looted treasures and the hostages,
As proof of victory they give their king.25
Although ultimately a failure in its chief objective, that of catching Gruffydd, the raid provided an early demonstration of Harold’s ability for rapid military manoeuvres over long distances, which would be a feature of his later campaigns in 1066.
King Edward and Earl Harold were determined now to destroy Gruffydd and remove his threat from England’s borders once and for all. Therefore, Earl Harold planned a joint expedition with his brother, Tosti, to take place in spring 1063. Under Earl Harold’s overall command, this well-coordinated campaign began on 26 May, when Harold sailed with a naval force from Bristol and Tosti invaded with a cavalry force, presumably entering North Wales from the Chester area. Harold sailed round the Welsh coast, laying waste the land and taking hostages, while Tosti did the same by land. The use of a fleet by Harold was the key element of the plan and it was almost certainly employed in the occupation and devastation of Anglesey, the granary of North Wales and the foundation of Gruffydd’s power. Indeed, this same strategy was to be followed by King Edward I in his later conquest of Wales. The author of the Vita Eadwardi provides some tantalizing hints of the tactics which may have been used to achieve Earl Harold’s first great triumph, and confirms Harold’s command of the expedition, although he is usually more favourable to Tosti. The writer speaks of swift-moving small units ravaging in Wales and defeating any Welsh opposition encountered. This account is perhaps confirmed by Gerald of Wales, who much later speaks of Harold devastating all Wales with ‘lightly clad infantry’ forces and erecting inscribed stones, reading Hic Fuit Victor Haroldus (Here Harold was victor), to mark his frequent victories. Although writing over 150 years later, Gerald preserved much local Welsh tradition and had himself seen examples of Harold’s stone victory markers. The result of these tactics was to subdue the Welsh countryside and force Gruffydd to retreat into the mountains of Snowdonia, from whence he continued to harass the English forces.26
This devastating campaign succeeded in isolating Gruffydd from his supporters and forcing the Welsh to submit, give hostages and promise to pay tribute. John of Worcester declares that the Welsh outlawed and renounced their king. Gruffydd himself might have escaped the net yet again, except that the severity of Harold’s campaign led his own followers to lose faith in him. As a result, the Chronicle says he was slain by one of his own men ‘because of the fight he fought against Earl Harold’. On 5 August 1063 Cynan, son of Iago, killed Gruffydd and brought his head to Earl Harold, who placed it before the king. This has been seen as an indication that Harold had arranged for Gruffydd’s assassination, and this was the sign of success. This seems unlikely as Gruffydd had, to a large extent, been rendered harmless by the devastation of his lands and the undermining of his support. In addition, Harold’s preference was to treat generously an otherwise broken enemy, as he demonstrated after Stamford Bridge, rather than to fight bloodily to the bitter end. It is possible that the assassin himself may have hoped for some reward from Harold, in which case he was disappointed, or simply that this macabre gift may have been a signal to Harold of the final submission of the Welsh, and that he could now cease his raiding.27
This military triumph over the most powerful Welsh leader of the period was crowned by a political settlement which placed the dead king’s half-brothers, Bleddyn and Rhiwallon, in power in North Wales. In return, they swore oaths of loyalty both to King Edward and to Earl Harold, they gave hostages, and they agreed to pay tribute. A number of English border territories were now recovered, including the lands beyond the Dee and Archenfield, which Gruffydd had acquired through his recent aggression. In addition, some territorial advances were made into Wales, such as at Portskewet where Earl Harold set out to build a hunting lodge. Meanwhile, in South Wales several native dynasties were left to struggle for control.
In this campaign Earl Harold had revealed himself as an innovative strategist and tactician and as a great commander, and he had greatly enhanced his reputation. Wales had been subdued and a major threat to England removed. He had won a stunning victory over the only king to rule all Wales. The echoes of his victory can be found in sources dating from as late as the thirteenth century. Gerald of Wales quite reasonably attributes the early Norman successes in Wales to the defeat previously inflicted on the country by Harold. John of Salisbury less credibly reports, that so many Welshmen were killed that King Edward gave permission for Welsh women to marry Englishmen, and that Harold enacted a law to punish Welshmen travelling armed into England.28
SIX
WILLIAM OF NORMANDY
Where Harold made an oath to Duke William.1
In 1064 Harold was at the height of his power and influence. A successful military commander and the leading noble in England, with the demise of Earl Aelfgar he had no real rivals. Three of his brothers were earls and he himself was right-hand man to King Edward. However, an episode occurred about this time which was ultimately to lead to his downfall. This was his visit to Normandy and his memorable encounter with William, Duke of Normandy. No contemporary English source mentions this episode – the Chronicle record is particularly sparse at this point – thus we are dependent largely on Norman sources, although there are hints of an alternative version in one later Anglo-Norman source. The Norman sources, which are closely interrelated, were composed sometime after the event, around 1067–70, and with the specific purpose of justifying William’s claim to the English throne. We must bear this in mind when considering both their accounts of this episode and, more importantly, their interpretations of it.2
The Norman sources provide no precise date for Harold’s visit to Normandy. William of Jumieges places it somewhere between the death of King Henry of France in August
1060, and the death of King Edward in January 1066. William of Poitiers places it a little more precisely, at about the same time as the acquisition of the County of Maine, by Duke William, which had certainly been completed by 1064. This Norman dating of around 1064–5 fits well enough with the lack of any positive English evidence that Harold was anywhere else between the death of Gruffydd of Wales, in August 1063, and his giving orders for the construction of a hunting lodge at Portskewet, in July 1065. Within this period, a visit to Normandy by Harold is probably best placed in 1064, since by then Maine had been conquered. William’s expedition against Brittany, on which Harold is said to have accompanied him, probably also occurred in that year.3
According to the Norman version of events, Harold was sent to Normandy by King Edward in order to confirm the latter’s earlier promise of the succession to Duke William and to swear fealty to him. However, we have seen that it is unlikely that any such promise was given by Edward, but rather that it was probably invented and imparted to William by Robert of Jumieges, Archbishop of Canterbury, following his exile in 1052. If this was the case, could Edward nevertheless have intended to make William his heir at this later date? This is highly unlikely. In 1051 Edward had no clearly established heir, although he did have a number of potential heirs, all with better qualifications than William. Now, he had secured a suitable and established heir in the person of his nephew, Atheling Edgar, and a reserve in Harold, the son of his now deceased nephew, Earl Ralph. As a result of this change in circumstances the reasons adduced against the nomination of William as heir in 1051 apply with even greater force to any such nomination in 1064. He remained a man with only distant links to the English dynasty and little or no support in the country, although he was now secure in possession of his duchy and much more widely known and regarded than in 1051. In addition, William’s recent conquest of Maine had resulted in the imprisonment and death of Edward’s nephew, Count Walter of the Vexin. Count Walter died in suspicious circumstances while in William’s custody, allegedly by poison, something unlikely to endear him to Edward. William of Poitiers hints that Edward was close to death and this was why he now sent Harold to pledge his kingdom. There is no support for this in English sources, which show that the king was still healthy enough to go hunting in autumn 1065. The suggestion that Edward intended William as his heir in 1064 seems less credible even than the case for this in 1051.4
Since there seems to be little support for the Norman account that Harold came to Normandy to pledge the English throne to William, should we believe that Harold visited Normandy at all? Harold’s previous visit to Europe, to Flanders in 1056, also goes unrecorded in the chronicles, yet is clearly evidenced by the albeit chance survival of his name among the witnesses to a Flemish diploma. Unfortunately, there exists no similarly conclusive contemporary evidence for the Norman visit. The reference in the Vita Eadwardi to Harold’s personal study of the princes of Gaul is a little too unspecific to warrant use as confirmation. However, this lack of evidence and of any reference to the event in the English sources cannot be taken as an indication that a visit did not take place, and although the Norman accounts are clearly biased, they present facts which cannot be entirely dismissed. Thus it seems likely that the basic facts contained in their accounts, Harold’s visit to Normandy, and his oath to Duke William, must have a basis in truth, otherwise their authors would lose credibility completely. This does not mean that the causation constructed around these facts and the elaborate details added to them are necessarily accurate. It must be our task to attempt to reconstruct the most plausible sequence of cause and effect which link the main events related by these sources.5
If we accept that Harold visited Normandy, but not to pledge support for William’s succession, then some other cause must have brought him there. There exist several alternative explanations for such a journey in other sources. As stated above, the contemporary Vita Eadwardi speaks in general terms of Harold making visits in order to study the French princes. Although this is a possibility, for Duke William was undoubtedly the most prominent of these princes and a man well worth study at this time, when he was at the summit of his power, it must remain an outside one.6
The Bayeux Tapestry provides no direct information on the reason for Harold’s visit. It was probably made for William’s half-brother, Odo, Bishop of Bayeux, around 1077 and therefore presents the Norman outline of events. However, it was produced by an English workshop, which perhaps wove its own outlook and sympathies into the design. This mixed origin or perhaps the simple brevity of its textual style mean that the text provides little clarity in a number of controversial areas. Nothing specific is recorded in the accompanying text to explain the reason for Harold’s journey. He is clearly shown speaking to King Edward before his departure, but the text fails either to state that Edward sent him, or to provide any reason for the journey. There are hints later in the Tapestry of what may be a marriage alliance, but the scene which reflects this unfortunately remains the most inscrutable in the entire work.7
The suggestion that the proposal of a marriage alliance was involved is reflected in a number of later sources, usually in a secondary context, although perhaps it was originally more central. It may have been pushed into the background by the need to tie Harold’s visit more closely to the Norman claim to the throne. This required a direct sequence of events consisting of a promise of the throne in 1051, the pledge of it by oath in 1064, and the rightful succession in 1066. The suggestion of another purpose behind Harold’s visit to Normandy would obviously detract from this sequence, but the possibility certainly exists that Harold went to Normandy to arrange an alliance, including a marriage, with the new power in northern France. After all, his younger brother, Tosti, had married the sister of Baldwin V of Flanders and gained both prestige and a useful ally as a result.8
The sources are confused on the details of any proposed marriage probably because it came to nothing in the end. The later English influenced sources appear to indicate some arrangement involving the marriage of Harold’s sister, Aelfgyva, to a Norman. Eadmer implies a marriage to William himself, which is clearly absurd since he was already married, but his eldest son, Robert, was then in his early teens and perhaps a suitable match. (Marriage between men and women of widely differing ages was fairly normal at this time.) One of William’s major barons may be an alternative candidate, but this would seem an unlikely match for the sister of someone of Harold’s importance. The tone of contempt considered by many to be present in the Tapestry’s reference to the woman Aelfgyva may possibly reflect a Norman view of this woman – as an unworthy match, and of a lower status. William may have viewed Harold as of vassal status in contrast to an almost sovereign lord like himself, but Harold undoubtedly viewed them as of equal status, great lords under a king. The Tapestry itself merely refers to a mysterious ‘Aelfgyva’, who may possibly be a sister of Harold. The form of the name in the Tapestry is English but this may be a result of the Tapestry’s English origins rather than an accurate representation of the original name. The woman could equally represent an anglicized Adeliza, William’s daughter, who may have been intended as a bride for Harold as part of William’s own arrangements for securing the English throne as later related by Orderic Vitalis. The real basis behind these traditions will probably never be established, but it nevertheless represents a possible reason for Harold’s journey.9
There also exists the possibility that Harold did not intend to visit Normandy at all. A later Anglo-Norman source, William of Malmesbury, suggests that Harold was on a fishing trip when blown to shore by a storm. The Tapestry in one of its scenes may provide some evidence to support this by showing a fishing pole, although this may, in fact, be the origin of William’s story. It may simply be an incidental detail of the design, although it is absent from other scenes of sea voyages. Whether the English nobility of this period actually participated in fishing as a sport or pastime like hunting or hawking is unknown but perhaps unlikely.10
The view of William of Malmesbury that Harold arrived in Normandy by accident leads us to the general possibility that Harold could have been en route to somewhere else on the Continent, and was caught in a storm and blown off course, ending up in Normandy. It should be remembered that in 1064 Harold made landfall in Ponthieu and not Normandy at all. He may have been travelling to arrange a marriage alliance elsewhere, hence the presence of his sister Aelfgyva, as, perhaps, recorded by the Tapestry. This would resolve the problem of there being no suitable groom in Normandy itself. Whatever the exact case, although it cannot be proved that his arrival in Normandy was purely accidental, this possibility should none the less be kept in mind.11
Eadmer, whose writing dates to just before the end of the eleventh century, informs us that the reason for Harold’s journey was that he wished to secure the release of those members of his family held hostage by William since 1052. Eadmer was an English monk of Christ Church, Canterbury, in the period from before 1067 onwards. He would have been well informed about Harold and his career through the close contacts of that community with its former member and Harold’s relative the monk Aethelric, one-time candidate for the archbishopric of Canterbury and later bishop of Sussex. Full details about Aethelric’s career and connections will be found in Appendix One. Indeed, it is known that Eadmer consulted this very Bishop Aethelric when writing his Life of St Dunstan and he could therefore also have been in receipt of inside information from him when compiling his account of Harold’s journey. The need for Harold to secure the release of his relatives certainly provides a very direct link between Harold and William, which is perhaps lacking in the accounts of possible marriage alliances discussed above. These hostages had, it is true, been held for some twelve years, and Harold seems to have made no attempt to obtain their release until now. The reason for this is most probably that following his Welsh campaign and with England secure and at peace, Harold finally felt able to perform this mission personally, which had been impossible before. Alternatively, perhaps he had already sought their release but had been unsuccessful, and therefore no record remains. Eadmer’s account may represent the closest thing we have to an English account of Harold’s visit and one based on sources close to his family. It should therefore perhaps be given more credence than its late date would indicate.12
Harold Page 14