It is time to pause before we look at the actual conflict. We have already made use of most of the sources which give accounts of the battle.1 Use of Wace, for example, has often been preceded by some modifying remark to warn of his relative unreliability. Some sources are clearly invaluable, and our whole look at the period depends on them. But now we come to the crux of our present business, the battle itself. Any modern account depends not only on the contemporary and near-contemporary sources, but also on interpretation of them. History is not a precise science. We never have perfect materials on which to work. The degrees between good, reliable material and difficult, unreliable material are many and slight along the way. Some works themselves are rather like the curate’s egg, good in parts. What each historian chooses to use or to disregard makes his own view individual. One historian will disagree with another, but there is no absolute right and wrong. However much care we take, we can get it wrong. Indeed, we do not truly know if we get it right or wrong; we can only do our best. It is necessary to make a careful evaluation of sources but, in the end, interpretation of them is subjective rather than objective, since none of us knows the absolute truth of what happened in the past.
The battle of Hastings is well covered as medieval battles go, and we know much more about it than most conflicts. It was quickly recognised as a major event, and was treated as such – headline news for any chronicler writing on the period. The major lack in the Hastings sources is of an eyewitness account. This is a serious gap, and we have no way of filling it. No one who was on the field of Hastings has left us an account of what happened. All our narratives are therefore at second hand. Even then, in nearly all cases, we can only guess at where the chronicler obtained his information, and how much reliance can be placed upon it. Our primary concern though must be to try and gauge which chroniclers were best placed to receive accounts from participants and give good information.
We also must try to determine the viewpoint of the writer, since we know how much this affects his account. We try to detect bias and partisanship, and it is often apparent; medieval writers made little attempt to be neutral as modern journalists sometimes pretend to be. And they were all human; each one lived a life which gave a particular view to the events at Hastings. One would be a monk, hearing accounts from knightly guests; one would be a chaplain in a noble or royal household, listening to the table chit-chat of battles past; another would grow up in a house where the elders told tales of valour in days gone by. In a way, the openness of these opinions is an advantage, because the bias is often clear, and allows us to counter it. At Hastings, the most obvious bias would be whether the writer was pro-Norman or pro-Saxon. As it happens, nearly all the sources for the battle are Norman in viewpoint, and this creates a problem in trying to make a fair balance.
The other major consideration (we do not have time to go beyond this in our discussion) is the date at which the writer was putting quill to vellum. Obviously the nearer to the event the more valuable the account tends to be. Unfortunately again, precision on dating is not always possible. We usually begin from undated manuscripts, sometimes only copies without originals, sometimes only printed copies with the originals lost. Dating manuscripts is a whole science on its own, and dating works where we do not possess the autograph work by the writer is even more difficult. That said, we can usually come to some conclusion, an approximate date which gives at least a suggestion as to the likely value of a source.
Let us then survey briefly our major sources, and try to give some indication of their main values and weaknesses. There is still a certain amount of dating debate over the two main Norman chronicle accounts, but they are both eleventh century. In most people’s estimation, the major source for Hastings and the Conquest is William of Poitiers’ Gesta Gulielmi Ducis Normannorum et Regis Anglorum (Deeds of William, Duke of the Normans and King of the English).2 The original manuscript belonging to Sir John Cotton was copied and printed by Duchesne in 1619 and never seen again. The original may have been burnt in a fire in 1731. For dating we must rely on evidence within the contents of the account. There is a good modern edition of this by Raymonde Foreville, with a French translation, and sections are printed with English translations in all modern collections of sources relating to the Conquest and the battle. The main reason for recognising its value is twofold: it is the most detailed account of events that we have, and it is by someone who was in a position to be well informed. William of Poitiers may be treated as virtually the mouthpiece of the Conqueror. It was also written down early, Foreville believes by 1074, most others would agree by 1077. We know a little about the writer: he came from a noble family in or near Préaux, related to the Beaumonts, and had apparently early in life been trained in military discipline. But his sister became an abbess, and he became a priest at about the age of thirty. He had been born in Normandy in about 1028 and for some time studied in Poitiers, hence his toponym. He had legal knowledge and was for a time Archdeacon of Lisieux. He entered the household of the Conqueror as a chaplain, and this of course is why his work has such value. He did not cross to England in 1066, but he did come at some slightly later date. He saw the Conqueror as a hero who could virtually do no wrong, which is where we treat him with some circumspection. What we have goes up to 1067, but his account up to 1075 is probably used and preserved by Orderic Vitalis.
The second important Norman source is the Gesta Normannorum Ducum (Deeds of the Dukes of the Normans) by William of Jumièges.3 In contrast to William of Poitiers, this work does not concentrate on the Conqueror. It is a history of all the dukes. The early section is almost a direct copy from the work of Dudo of St-Quentin, which makes one wonder about William of Jumièges’ historical acumen. But it does come up to his own period, and he was writing in the early 1070s. He was a monk at Jumièges, a great Norman monastery with old links to England in its associations. His work was added to by other writers in the versions we have, in particular by Orderic Vitalis and Robert of Torigny. This has been known for some time, and there are two modern editions. The best and the most recent is by Elisabeth van Houts, and has an English translation with the Latin, and clear indications of whether the work we are reading is by William of Jumièges or one of the interpolators.
The chief English source for the Conquest is the Anglo-Saxon Chronicle, which is actually several different versions of a work begun in the time of Alfred the Great.4 It is a year by year account of events. Additions to the annals were kept in several monasteries, and so different versions of the Chronicle developed. This became more complex over time, as one house borrowed a version from another house and then began to make its own additions. There are five main versions, known as A, B, C, D and E. The one which gives the fullest account of the Conquest is version D, the manuscript for which is written in a late eleventh-century hand. Both D and E may come from a version which had been made in York, sometimes called the northern recension. They are pretty well the same until 1031. D then continues with its interest predominantly in the north, probably still being kept at York. It continues till 1079. From 1031 E was probably being written up at Peterborough; it continued later than other versions of the Chronicle, until 1154. These versions, D and E, are therefore mainly from the point of view of northern and eastern England, where Scandinavian influence was strongest, not in the heartland of Godwin power. The writers are Benedictine monks but, so far as we know, without the major contacts that benefited William of Poitiers. Nevertheless, the D version in particular is valuable material as the main English view of what happened, and told in Old English. The easiest version of the Chronicle to use is that edited by Dorothy Whitelock and others, which places the versions side by side in columns so that comparison is easy; but for the original Old English one needs to go to other editions.
John of Worcester may be taken in conjunction with the Chronicle, since its earlier part is almost a Latin version of it.5 The Anglo-Saxon material used to be called the work of Florence of Worcester, but his recent editor argues that we should call the work John of Worcester’s. Here we have another Benedictine house, Worcester, keeping an annal. The value of the Worcester account is that although close to version D on the Conquest, it is at least another English view of events.
The other major source for the Conquest is that invaluable and unique embroidery and document in one, the Bayeux Tapestry.6 We all know that it is really an embroidery and not a tapestry as such, but it would be pedantic to call it by anything other than its familiar name. Those who view it in its present setting are often surprised that it is only 20 inches high, but also marvel at its 230 foot length, which cannot be appreciated in full in the usual book reproductions in separate plates. Indeed, the way the artist has designed the scenes to move fluently along from one event to the next is masterly. The backing is of bleached linen, and the embroidery is in five main colours of wool, with three less used colours. An artist sketched the scenes, and the embroiderers filled in the outlines with laid and couched work, stem stitch and outline stitch. It is made of eight sections pieced cunningly together so that one has to search hard to find the joins. The original end is lost, probably because of the manner in which it used to be kept rolled up. It is thought that perhaps 9 feet are lost, and that the missing section may have contained scenes of William’s entry into London and his coronation. The latter would be a fitting conclusion, since we have the Confessor on his throne at the start and Harold’s coronation in the centre. Apart from the main narrative beginning with Harold’s trip to Normandy and ending with the English flight after Hastings, there are top and bottom margins, which sometimes add to the main story, sometimes retail myths and fables, and also provide delightful illustrations of such things as harrowing, scaring birds, hunting and boat-building. It is now widely accepted that the work was made for Odo, Bishop of Bayeux. It is possible that some of the minor figures on the Tapestry – Wadard, Vital and Turold – were tenants of Odo, who soon became earl of Kent.
The Tapestry may have been specially made for display in Bayeux Cathedral at its dedication in 1077. Some critics have thought certain scenes on the Tapestry too bawdy to be intended for such an ecclesiastical setting, but this seems to come from modern rather than medieval sensibilities. Certainly, an inventory of 1476 shows that the Tapestry was at the cathedral then, and was put on view annually. Some efforts have been made to devalue the Tapestry. A recent suggestion was that the ‘kebabs’ were too modern and that the work dates from the nineteenth century. But although it is clear there were some repairs done to the original, we can still feel safe that it is an early and valuable contribution to our knowledge. Its particular value is that it gives pictorial versions, and therefore information not otherwise available. Indeed, it is not merely visual evidence but a great work of art.
The Tapestry also has a written legend, a brief account of the events portrayed. From its tone this is almost certainly the work of a Norman, or at least a pro-Norman, though it does have one or two interesting touches of English sympathy, such as the rescue carried out by Harold on the Brittany expedition. It completely ignores the invasion by Hardrada. The Tapestry magnifies the role of Odo of Bayeux, who barely appears in the chronicle accounts.
It was probably made in England and worked by English embroiderers, perhaps at Canterbury: some of the scenes seem to have been adapted from manuscript art in Canterbury works. English working seems evident in such places as the use of a crossed d in names and in the English version of the Hastings place-name.
A printed copy was published in 1730 by Montfaucon, the drawings done by A. Benoît. The original was nearly lost during the French Revolution, when it had to be rescued from being used to cover a wagon. It was taken to Paris, and returned for exhibition at the Hôtel de Ville in Bayeux in 1812. It is known that repairs were done in 1842, which can be recognised from the use of different colours in the wool, and restoration marks on the linen. Modern examination has included such details as stitch marks which suggest how the original looked.
Then there are a number of sources to which we choose to give a secondary place, either because there is some question mark over their reliability and/or because they are late in time compared to the sources already mentioned. The most interesting of this group is the Carmen de Hastingae Proelio (The Song of the Battle of Hastings).7 This is a long poem about the battle. It is very detailed and many historians have considered it a prime source, and some still do. To a degree the jury is still out on the Carmen, though all would agree it has some value. The manuscript was rediscovered in 1826 by G.H. Pertz. There was no title on the manuscript, indeed its subject matter is really ‘The Norman Conquest’ rather than simply the battle. It was written by an educated person, with plenty of biblical and classical references. It has been thought that this was a work mentioned in the twelfth century by Orderic Vitalis, a poem about the Battle of Hastings by Guy, Bishop of Amiens. This being so, it would be an early source, earlier than William of Poitiers. But the surviving manuscripts, one main and one tributary, which come from Trier, have been dated to about AD 1100 from the handwriting. If written by Guy, then it was by a respected and important noble, who came to England with the Conqueror’s wife a couple of years after Hastings. The poem retains a mysterious dedication with initial and not names: ‘L … W … salutat’ – which could be either L greets W, or W greets L. Those who name Guy as the author fill in ‘Lanfrancum Wido salutat’ (Guy greets Lanfranc (Archbishop of Canterbury, 1070-89)).
But R.H.C. Davis made a serious attack on the attribution of the work to Bishop Guy. In an article, in English Historical Review, he suggested that too much of the poem is of a style and content that would fit with a later date. The present author found that argument convincing and still does; others have been less sure. Davis argued that the poem is not as hostile to Harold as Orderic had suggested Guy was. He thought that the Carmen borrowed from William of Poitiers, though others believe it is the other way round. The most convincing argument is that the Carmen introduces legendary and incredible material which could only be later. This includes the story of Taillefer, the giant who opens the conflict. He appears in none of the accredited early sources, and the tale has the touch of legend about it. Then there is the killing of Harold, by four men, identified as Duke William, Eustace of Boulogne, Hugh, the heir to Ponthieu, and Walter Giffard. The first, presumably William, cleaved through Harold’s breast, the second smote off his head, the third pierced his belly with a lance and the fourth cut off his thigh and carried away the leg. If this were true it is not credible that the main Norman sources would have ignored the fact that William actually participated in the killing of Harold. We can agree with Davis that this, and other examples he quotes, is ‘literary embellishment’. One possibility, though probably beyond proof, is that the Carmen is a work of about 1100 which is either based on the poem by Guy of Amiens or is an embellishment of it. This would leave it with importance, but the need for care must be stressed. It seems that as our knowledge stands at present it would be unwise to give the Carmen the credence that we give to William of Poitiers, the Anglo-Saxon Chronicle, or the Bayeux Tapestry.
Another difficult work to assess, though its author and date are known, is the Roman de Rou by Wace.8 This is undoubtedly a late work, as Wace was not born until about 1100. He was born in the Channel Islands, though the handwriting for his native island could be interpreted as either Jersey or Guernsey, we are not sure which. He at least tells us the source of his information: people he spoke to who had witnessed the events. He said, ‘I talk to rich men who have rents and money, it is for them that the book is made.’ He described himself as a ‘vaslet’ or varlet, which it is thought might mean that, like William of Poitiers, he had some knightly training in his youth. A recent article shows that he had a good knowledge of warfare, which gives his work value for our purposes. He was educated at Caen in Normandy, and in the French realm, later returning to Caen. He was patronised by Henry II, who gave him a prebend at Bayeux cathedral. He held this post for nineteen years, so it is nearly certain that he was familiar with the Tapestry. He was a prolific writer, and his works included verse romances, one called the Roman de Brut and another the Roman de Rou. Rou is a version of the name Rollo or Rolf, the Viking leader who became the first ruler of the new Normandy, so the work was a kind of verse history of the dukes. He probably wrote it in the second half of the twelfth century, and died in 1184. The problem with the Roman de Rou, apart from its latish date, is that it is a romance. Wace was a literary writer, he was looking for effect, he liked a good story and was not always fussy about accuracy or borrowing from one situation to enliven another. He is the sort of medieval author who is most difficult for historians to use: too useful and too lively to ignore, but too risky to trust. The use of Wace in this book is to allow passages about which there is no serious concern, but to have great caution with any lines which have the feel of legend or invention about them. As said before, history is not an exact science.
Finally, we need to consider a group of twelfth-century historians who covered the history of the battle. By this time, the Norman Conquest was well established, and its significant consequences were apparent. This coloured views of events, and William’s position is usually seen as the correct one: the winner is always right in history, as some would say. We do not have time or space to look at all the later sources which deal with Hastings, so this is simply a selection of those which seem the most useful or important. It should be said that, because of their dating, they must generally be given second place behind the earlier sources when trying to assess their relative significance. But often they will confirm what appears in the earlier sources; usually, of course, they are based on one or more of them, and sometimes on sources which are lost – and this will give an added value.
William of Malmesbury was a major twelfth-century English historian. His De Gestis Regum Anglorum (Deeds of the Kings of the English) includes material on the period of the Conquest.9 William was a Benedictine monk, but he was unusual in that he explains to us something of his methods, telling us that he travelled about in search of documents. He was also able to read Old English as well as French and Latin. He wrote a vast number of works, and was a stylish and lively historian. He was probably born in the last decade of the eleventh century, and is thought to be of mixed Norman and English parentage. He says he collected historical information at his own expense, which suggests that his family was fairly wealthy. He became librarian at Malmesbury, and perhaps precentor. William is selected because he was an outstanding writer rather than because he gives especially important information.
Orderic Vitalis, like William of Malmesbury, was a product of the Conquest, son of a Norman priest and an English mother. He was born at Atcham near Shrewsbury in 1075. His father may have been in the household of Robert of Montgomery. Orderic received some local education and was then sent as a boy of ten to train to be a monk in the Norman house of St-Evroult. There he stayed for the rest of his life, until 1141 or 1142. An early work consisted of his additions to the chronicle of William of Jumièges. He then spent some thirty years working on his great opus, the Ecclesiastical History, which was a long and rambling work, whose intention and structure changed with the years. As a result, it is full of the sort of titbits which make history fun, tales of people he knew or heard about, the occasional scandal.10 Orderic comes across in his writing as a likeable man who enjoyed life, and his work has a human touch which some medieval chronicles lack. We have Orderic’s work in his own beautifully neat script. His English background gives his work interest from its attitude. He is the only one to give us the name for the battle location as Senlac, and he has some criticisms of the Conqueror’s actions.
A third English historian of note is Henry of Huntingdon.11 He was Archdeacon in Huntingdon, a secular cleric rather than a monk, a father of children as well as the son of a priest, a man of the world. He was well travelled and visited Bec in Normandy, and Rome. He had a historian’s, almost a journalist’s nose for information, and wrote ‘there is nothing in this world more exciting than accurately to investigate and trace out the course of worldly affairs’. He tells us that he used ‘compilations of the chronicles preserved in ancient libraries’. His Historia Anglorum is a great wide-ranging history of the English. His outlook has a strong East Anglian slant. He was born in about 1080, began writing in about 1133 and, after adding new work to his original effort, brought it down to 1154. It was a popular work of its day with many known copies. One of his virtues was that he could read Old English, and used the documents to which he had access. Henry’s work is now available in one of the first-rate Oxford editions, with Latin alongside an English translation.
Finally, because it has a special significance, we must consider the value of the Chronicle of Battle Abbey.12 This too is now in an excellent modern Oxford edition. The chronicle is undoubtedly late in date, after 1155, but it was written in the abbey built on the site of the battle - so it tells us – its very name depending upon the event. It was written by a monk with a great interest in the law. The trouble is that his modern editor has caught him out. He used documents forged in the abbey to make a case, and there is little doubt that some of his claims are false. The difficulty is to know if others are true. He was certainly trying to boost the importance of his abbey. The chronicle consists of two separate texts, of which the first is an account of the Norman Conquest, and both mention the abbey’s foundation. The main value of the chronicle is its local knowledge. It gives detail not known elsewhere: the name of Hedgland on Telham Hill, and the story of the Malfosse. The modern editor of the chronicle has shown the problems over this location, and we shall look at them in the following chapter. The even more basic problem is that if this chronicler is unreliable, can we trust his story of the abbey’s foundation on the spot where Harold was killed? This also we shall return to. The uncertainty over theCarmen is added to by the fact that there are incidents which only appear, other than in the Carmen, in Wace and the Battle chronicle. We seem to be looking at three sources which are all beginning to enlarge on original facts with dubious tales.
There are numerous other sources, but we must call a halt. We conclude that William of Poitiers is our primary source of information, that he is followed by several valuable early works in William of Jumièges, the Anglo-Saxon Chronicle, and the Bayeux Tapestry. All the other sources, to some extent, depend upon these early versions. The only other account which may be early and original is the Carmen, but our decision is to relegate this to the second division of sources. Here it joins forces with later and less trustworthy accounts, depending either upon hearsay, third-hand material or invention. The difficulty is that these include some of the most detailed works and some of the most lively, Wace and the Carmen; and some by the best historical writers, William of Malmesbury, Henry of Huntingdon and Orderic Vitalis. We must pick our way between their accounts. In the end, our objective is the truth. We cannot be certain we find it, but we must be certain that our attempt is honest. Such is the historian’s task, every historian’s task, from professorial academic to humble student.
1. A useful general guide is A. Gransden, Historical Writing in England c. 550 to c. 1307, London, 1974.
2. William of Poitiers, ed. Foreville. Translations of part of the work with the battle may be found in D.C. Douglas and G.W. Greenaway (eds), English Historical Documents, ii, 1042–1189, 2nd edn, London, 1981; and Brown (ed.), The Norman Conquest, London, 1984 – both of these have a good selection of sources of the Conquest.
3. William of Jumièges, ed. van Houts.
4. Whitelock et al. (eds), Anglo-Saxon Chronicle. The ‘Collaborative Edition’ (eds D. Dumville and S. Keynes) is at present incomplete, but will become the foremost academic edition of the Anglo-Saxon Chronicle with the text in Old English. Of the volumes published to date, vol. vi (ed. G.P. Cubbins) of the D manuscript is the most valuable for events of the Conquest.
5. John of Worcester, eds Darlington and McGurk.
6. See Bayeux Tapestry which has excellent colour photos. See also F.M. Stenton (ed.), The Bayeux Tapestry, 2nd edn, London, 1965; Thorpe (ed.), The Bayeux Tapestry, which also has a translation of part of William of Poitiers. N.P. Brooks and H.E. Walker, ‘The authority and interpretation of the Bayeux Tapestry’, ANS, iii, 1980, pp. 1–21.
7. Carmen, eds Morton and Muntz; Davis, ‘Carmen’, pp. 241–61.
8. Wace, ed. Holden; Wace, ed. Taylor; Bennett, ‘Wace and warfare’.
9. William of Malmesbury, De Gestis Regum Anglorum, ed. W. Stubbs, RS no. 90, 2 vols, London, 1887–9; William of Malmesbury, ed. Giles.
10. Orderic Vitalis, ed. Chibnall; M. Chibnall, The World of Orderic Vitalis, Oxford, 1984.
11. Henry of Huntingdon, Historia Anglorum, ed. D. Greenway, Oxford, 1996; Henry of Huntingdon, ed. Arnold; Henry of Huntingdon, Chronicle, ed. T. Forester, London, 1853; N.F. Partner, Serious Entertainments, Chicago, 1977.
12. E. Searle (ed.), The Chronicle of Battle Abbey, Oxford, 1980.