Observations upon a Scene in the Bayeux Tapestry, the Battle of Hastings and the Military System of the Late Anglo-Saxon State

By M. K. Lawson

The Medieval State: Essays Presented to James Campbell (2000)

If a relative plethora of sources provides more evidence on the battle of Hastings than any other event in Anglo-Saxon history, there is surpris–ingly little of which we can be certain.* The numbers involved, the extent of the initial deployment, the tactics employed and the course of the fighting are all to a greater or lesser degree unclear, and much more debatable than the secondary literature has often been prepared to admit. As a depiction of a conflict which lasted for most of a mid–October day, for example, the Bayeux Tapestry is an inadequate record. Probably inclined to overstress the role of cavalry, it does not show the French mailed infantry mentioned by William of Poitiers,1 and offers only a limited range of scenes: Duke William’s horsemen charge the English shield-wall, other infantry fighting in loose order (including King Harold’s brothers), and a third group defending a hillock; William proves that he is still alive while his archers advance in the lower border, where, as the dead are stripped, further attacks on scattered Englishmen in the scenes above precede the death of Harold and the pursuit of his defeated army. Nevertheless, the Tapestry is a fascinating source: almost certainly seen by, and probably produced for, participants in the battle,2 such details as it offers of the fighting and of military methods and equipment seem likely to be reasonably accurate. Indeed, the scene in which the French attack English infantry on a hillock is vital. It both casts doubt on views about the scale of the battle which have prevailed for the last century, and suggests (with other evidence) that Anglo-Saxon armies fought in varied and reasonably complex ways which have never been properly acknowledged. It is true that there is much about Hastings that we can never know, and that many views of it are possible; and it is equally true that this should encourage the most open-minded of approaches. Or one might say that the battle and the capacities of the Anglo-Saxons’ military machine should be considered, like other of their activities, according to the dictum of the philosopher whose nearest approach to a positive statement was: `Not but what it may not have been, perhaps it was’.3

Preceded by the words `Hic ceciderunt simul Angli et Franci in prelio’, the men on the hillock do not wear armour; three sport moustaches and have spears and shields, while one has a beard and a spear, and two others, with neither facial hair nor weaponry, tumble down the slope apparently in death; before them stands a moustachioed warrior wielding an axe double-handed and wearing a sheathed sword, accompanied by a spear-bearing soldier who is regarding the enemy with some trepidation; to the left a clean-shaven Englishman holding a spear and wearing chain-mail, helmet and scabbard is attacking a horseman by seizing his mount’s girthstrap; further to the left a water course has brought down two more horses.4 What, then, are we being shown? Well over a hundred years ago, Freeman offered an explanation which has found little favour since. In his view, the hillock is identifiable today on the western part of the battlefield, lying just to the south of the ridge, about one mile long, upon which the conflict took place, and just to the north of a drainage area often containing much water, which now flows into New Pond. An element of a small but noticeable ridge forming part of the general fall of the ground to the south-west, it is covered by bushes and small trees, with a steep slope to the south but a much more accessible approach from the north. In Freeman’s opinion, the English deployed along most of the main ridge and stationed men upon the hillock to protect their front at a point where the approach from the south is less severe than elsewhere. Thus, while prudently giving no estimate of the number involved, he spoke of an `immoveable wedge of men’ covering `every inch’ of the ridge, and his map showed them grouped in forty-nine individual units, with housecarls in the centre and `light-armed’ on the flanks.5 There was little justification for some of this, and by the time Sir James Ramsay wrote in 1898 most of it had been rejected. The German scholar Wilhelm Spatz had by then estimated the troops on each side at not much more than 6000-7000, while J. H. Round’s refusal to accept the palisade which Freeman had posited along the English front had led to the belief that the latter must have consisted of a continuous length of shield-wall, which was, if we accept William of Poitiers’s repeated statements on the density of the English formation,6 many ranks deep. F. H. Baring guessed that it might have numbered 20,000 to 30,000 men, and, as he thought this impossible, concurred with Ramsay that the English had occupied not the bulk of the ridge at the start of the battle but simply its crest.7 On this view, of course, whatever the Tapestry’s hillock scene depicts, it would seem to have little to offer on the initial English deployment, and is not support for Freeman’s view that this extended over much of the western part of the ridge. Certainly most subsequent writers have followed Baring’s line, with the result that the modern view of the English dispositions and of the scale of the battle is not only very different to Freeman’s (in minimizing both), but wears an appearance of certainty which the actual evidence does little to warrant.8

Contemporary sources are inconclusive on the size of the force which Harold commanded at Hastings. The E version of the Anglo-Saxon Chronicle, perhaps the most contemporary of all, says that he fought before all his army had come, while the author of the D text, who had something very like E in front of him, records that the English army was great, but that William came against it unexpectedly, before it was properly organized; nevertheless, Harold and those who followed him (conceivably a hint that some did not) fought hard, and there were heavy casualties on both sides.9 This tendency to offer explanations, if not excuses, for the English defeat is taken further by John of Worcester, who took a pretty partisan view of events between thirty and almost eighty years later. Here, Harold advanced with only half his army, joined battle when only a third of what he had was in order for fighting (this being an amplification of D, which John had before him, and whose statement on the size of the army he ignored), drew his forces up in a narrow place from which many of his men retired, but still fought valiantly all day, until he was killed around twilight.10  Of the Norman writers, William of Poitiers insists that the English force, augmented by reinforcements from Denmark, was vast, a view with which William of Jumieges agrees.11 Of course, it can be dismissed as the product of a desire to magnify the achievements of Duke William, and in any case we have no way of deciding what was a large army in eleventh-century terms. Even so, there is reason to think that it occupied rather more than simply the crest of the ridge at Battle.

The Tapestry’s hillock scene seems to represent an important stage in the conflict, which its audience was expected to recognize. Moreover, it immediately precedes the depiction of Duke William proving to his men that he was still alive, which according to William of Poitiers followed the flight of the Breton foot and horse and the auxiliaries on his left wing, and was itself followed by the annihilation of several thousand of the pursuing English.12  Freeman considered the possibility that the hillock scene shows the last stand of these unfortunates (a common view since), while thinking that the tumbling horses to its left fell some distance away, in the ravine to the west of the ridge.13 An alternative is to see both horses and hillock as in the same locality, and as representing fighting which preceded rather than followed the French flight, as the Tapestry’s inclusion of them before the duke’s rallying of his men, and close analysis of the scene, in any case suggest. The defenders of the hillock are not solely `half-armed peasants’.14 While two do bear only spears, the four moustachioed warriors (all unarmoured, but three with spears and shields and one with axe and sword) look very like regular light infantry, and their moustaches `as if some special significance attached to them’.15 Indeed, it is not improb–able that what we are being shown here is one of the English `picked companies’ of the sort present at Brunanburh in 937,16 and this would also fit the way in which their light equipment suits the terrain in which they are deployed. In other words, they did not fetch up on the hillock accidentally, while pursuing the fleeing French, but were placed there from the outset, as Freeman thought. This interpretation is strengthened if we associate with them the figures and horses to the left. On this view, the Tapestry’s watercourse is that to the south of the hillock, and its representation of an armoured man seizing the girthstrap of his enemy’s horse indicates that this too was held against the French. Furthermore, if Sir David Wilson is correct in thinking that the serrated shapes shown above the water represent a `defensive work of sharpened stakes’, this would underline the determination with which the English held this area and offer a convincing explanation of why the scene’s tumbling horses are in such difficulties.17 Wace, who wrote over a century after the Conquest, mentions ditches several times, and reports that during the battle the Normans were pushed back to one which they had previously left behind them, more perishing there than elsewhere, as those who saw the dead said.18 Fortunately, the likelihood that Harold resorted to field defences hardly needs support from evidence of such doubtful reliability. In 1064 he had accompanied the Normans on campaign in Brittany, and doubtless thought long and hard about how their cavalry might best be countered: not that there need have been anything novel about his methods, for too little is known about Anglo–-Saxon fighting techniques to make arguments from silence of much weight here. Professor Leyser noted that in the late ninth century Vikings on the Continent often erected `quick and effective field fortifications, dykes fortified by stakes, palisades and advanced ditches. Time and again their enemies were hampered by these works’. In the 880s Margrave Henry of Neustria was killed when he `rode into what might today be described as a tank-trap’.19

The implications of all this for our understanding of the battle of Hastings are considerable. First, and assuming of course that the Tapestry’s hillock and watercourse have been correctly identified with those visible today on the western part of the field, the view that Harold’s deployment was limited to the crest of the ridge becomes untenable. The hillock is some quarter of a mile from the point where he fell, and the protection of it and its approaches by both light and heavy infantry and field defences increases the likelihood that there were also English troops on the flattish western part of the main ridge to the north and north-west, as Freeman thought.20  Despite statements about the density of their line, it would be rash to assume that it consisted simply of along shield-wall;21 equally, the possibility that the battle involved not several thousand men on the English side but tens of thousands must be strengthened. Another argument on these lines might run thus: Harold had been on the site long enough to build field defences, yet the D version of the Anglo-Saxon Chronicle says that his forces were taken by surprise before they were properly ordered; if he could not organize his men in time, given an area which extended beyond the ridge’s crest, might this not be because, as D says, it was indeed a great army, and great in eleventh-century terms was at least bigger than the limits laid down by Spatz and his many disciples?

There is reason to doubt their assessment of the size of the French force too, even if the relevant arguments are no more than what might be called a sequence of possibilities. A Norman ship list records the number of vessels owed to Duke William in 1066 by fourteen of his magnates, together in four cases with the numbers of milites also due. Until recently treated with circumspection, partly because it was thought to date from the twelfth century and partly, no doubt, because of the somewhat startling nature of its evidence, the list has been rehabilitated in the work of Professor Hollister and Dr. van Houts.22 The latter believes that, although extant only in a Battle Abbey copy of c. 1130-60, the document was created in about 1072 or perhaps as early as December 1067, probably at Fecamp, on the basis of information compiled just before the Conquest by monks of that abbey in William’s service. The duke’s half-brothers, Robert of Mortain and Bishop Odo, owed 120 and 100 ships respectively, others fewer, but the grand total is 776, and 280 milites. The list then states, wrongly as it seems (although its enumeration of magnates may have become truncated in transmission) that the total was 1000, and that the duke also had many other ships from certain of his men, according to their means. Now Wace says that he had been told by his father that 696 ships, including those carrying arms and harnesses, sailed with William from Saint-Valery, although he had also read (in William of Jumieges) that there were 3000 ships.23 He too, like the list, says that William Fitz Osbern contributed sixty and (unlike it) Bishop Odo forty, adding that the bishop of Le Mans provided thirty vessels and their crews. As Dr van Houts has observed, despite their agreements Wace and the list differ sufficiently to seem independent of each other, and this bolsters our confidence in their general reliability on this issue.24 By any reckoning then, Duke William’s fleet was sizeable: if we accept the list’s evidence that he was owed 1000 ships by a portion of his men and more by others, and consider the possibility that some of his continental allies supplied their own vessels, then William of Jumieges’s total of 3000 does not look completely ridiculous. Moreover, such conclusions could be supported by what is known of the anchorages employed in 1066. William of Poitiers says that the fleet assembled at the mouth of the Dives and neighbouring ports, was delayed for a month by unsuitable winds, and then blown by westerlies to Saint-Valery-sur-Somme.25  Now the Gulf of Dives may in the eleventh century have provided a very large harbour indeed, for Wace claimed that the river flowed into the sea near Bavent, which is some five miles from the town of Dives-sur-Mer, where the entrance to the English Channel is situated today;26 together with the `other ports’, it may have been able to shelter many hundreds of vessels. The same is true of the Somme estuary at Saint-Valery, and probably of the harbours around Pevensey Bay, where William landed. The presence of sand and alluvium in localities that are today well inland shows that they were once likely to flood, or were open water. One reconstruction of the coast that William found shows the bay itself with an entrance about four miles wide and penetrating inland for about six miles, while the anchorage at Bulverhythe, west of Hastings, had an entrance about two-thirds of a mile wide and a penetration of about two and a half miles.27

Of course, even if we had a reliable total for his fleet we could not deduce the number of men it carried before negotiating the quicksand of average ship capacity. Even so, this is not quite the end of the ship list’s significance. Professor Hollister has noted that William of Poitiers names seven prominent Norman lay lords from whose counsel the duke benefited in 1066, and that if we replace Richard, count of Evreux, by his son William, who fought at Hastings and is named in the list, then the seven correspond exactly with the eight (including Odo, a bishop) premier suppliers of ships according to the list. `One might almost suppose that William of Poitiers was writing with some such ship list at hand’.28Indeed one might, and one might also compare his comment that Agamemnon set out with 1000 vessels but that Duke William had more – pluribus – with the list’s total of 1000 and multas alias naves; of course, William of Poitiers may have had William of Jumieges’s 3000 in mind, even though otherwise `the two works diverge completely’ on the events of 1066.29 One might also begin to think the unthinkable in another respect: if monks of Fecamp were with William’s army keeping records (and one is said to have acted as ducal messenger to Harold before the battle),30 might this not be where William of Poitiers obtained his figures of 50,000 for the number supported by the Conqueror at Dives and 60,000 for those he commanded just before Hastings?31 Obviously, the credibility of numbers like 60,000 is desper–ately difficult to ascertain. Ferdinand Lot noted that 60, 600, 6000 and 60,000 appear frequently in medieval chronicles, nor are they absent from works of Julius Caesar 32 which William of Poitiers was eager to imitate.33  But that is not to say that he was in this case influenced by Caesar, or simply utilized random round numbers, for such numbers also appear in early medieval assessment systems. The Anglo-Saxon document known as the Tribal Hidage deals repeatedly in recurrent totals of this type, including 600, while sixty appears six times in the ship list (including one reference to milites) and is known also from the Cartae Baronum of 1166, believed to record the quotas of knights which William required from his tenants in England.34 If 60,000 was not the number of men the duke had, it could have been the total that a scribe in his service at some point thought he ought to have had.

It is very likely that the Anglo-Saxon military system utilized tens of thousands of men, at least on occasion, and perhaps in 1066. As is well known, the early tenth-century Burghal Hidage implies that West Saxon and Mercian burhs were maintained, and possibly defended, by some 27,000 men.35  The German bishop Thietmar of Merseburg heard that there were 24,000 byrnies in London in 1016, a number which he seems unlikely to have invented, since he thought it incredible.36  Even more striking, Ćthelstan required two well-horsed men from every plough.37 The `plough’ to which he refers need not have been connected with the ploughland or the ploughteam of Domesday Book, for by the 1086 assessment the former would have provided over 120,000 and the latter over 160,000 men.38 Nevertheless, his dominions must have contained many `ploughs’, whatever they were, and the probability that a levy of two men on each produced tens of thousands seems high. It is unsurprising, then, that in extreme circumstances the written sources reveal an ability to call on what look like considerable military reserves. In 920 Edward the Elder ordered fortresses to be built at Towcester and Wigingamere, the defenders of both subsequently withstanding sieges until, in the case of the former, reinforcements arrived; then, in the summer, a great force assembled from the men of nearby strongholds and took by storm the Viking defences at Tempsford, while in the autumn Colchester was captured by a great army composed of garrisons and troops from Kent, Surrey and Essex; at much the same time the Scandinavians made an unsuccessful attack on Maldon, suffering `many hundreds’ of dead at the hands of its reinforced garrison. Edward himself then led a West Saxon army (presumably one which had so far been inactive) to Passenham, while Towcester was fortified in stone; and when that army returned home another assembled for the capture of Huntingdon. It may have been the same force which then restored the defences of Colchester.39  More remarkable still, in 1016 Edmund Ironside called up all the people of England five times during his extended campaigns against Cnut.40 Of course, one might argue that the capacity to perform such feats had been allowed to lapse subsequently, but if so it had not reduced contemporaries’ opinion of the efficacy of the English navy, whose assistance was sought by both Swegen Estridsson of Denmark and the German emperor Henry III in the 1040s;41 or the Confessor’s ability to reduce the Welsh to subjection in 1063.42 In all, then, despite the battles of Fulford and Stamford Bridge, it may well be that the English admin–istrative system allowed Harold to assemble a great army at the hoar apple tree, as the D version of the Anglo-Saxon Chronicle and Norman writers claim, and that William brought against it a force transported in many hundreds of ships and also of very considerable size.

The late Anglo-Saxon state geared its people for war to an extent that the English were perhaps not to see repeated until the war of 1914-18. The defeat of the Vikings and reconquest of Danish England were major administrative and military achievements, as the Chronicle entries for the reign of Edward the Elder (noted above) demonstrate. Ćthelstan and Edgar established dominion over all Britain, while the less happy reign of Ćthelred II ended with a period of intense conflict, which ushered in Cnut’s northern empire, partly the creation of English military resources and English wealth. The reign of the Confessor was relatively peaceful, but it must have been obvious that his death was likely to result in bloodshed. As with other aspects of Anglo-Saxon life, the details of military organization, both on the battlefield and elsewhere, are largely hidden from us, but that is not to say that the English armies of 1066 were deficient in these respects, or bound to be defeated by William; indeed, if anything is clear from the sources about Hastings, it is that defeat was far from inevitable. Certainly, the Anglo-Saxon military system is very likely to have incorporated features of which we know nothing, or almost nothing. The wars of Alfred and his son, for example, often saw attacks on fortresses which may well have involved siege machines; they are mentioned on the Continent, but not in England.43  Similarly, it would be unsurprising if English military organization reflected that of the Carolingians. When Domesday Book for Berkshire tells us that a man went to the army from every five hides, there is an obvious similarity with a capitulary of Charlemagne which requests that all freemen with allods of five, four or three mansi are to attend the army.44 Yet in 806 the degree of military service imposed on the Saxons varied according to the enemy concerned, and it seems unlikely that in England the five-hide rule was the sole yardstick employed: it does not quite suit the fact that two of the Hampshire freemen killed at Hastings held the same estate (of four hides and one virgate) jointly along with another, or the varied equipment, reflecting different social ranks, found in sources like the poem on the battle of Maldon and the Bayeux Tapestry.45 Charlemagne on occasion also specified the gear that troops were to have with them, as Ćthelred II did in 1008 when he ordered that his fleet should be supplied with a helmet and byrnie by every eight hides.46  Moreover, the equipment observable on the Bayeux Tapestry shows uniformity as well as varia–tion. The shield-wall men, for example, almost all have byrnies and helmets, kite-shaped shields and spears,47 while the regularly-equipped moustachioed light infantry on the hillock are accompanied by peasants provided only with spears, and three of the rustic quartet who flee on foot at the end have clubs of similar design.48

While it could be argued that such regularities were simply convenient for the Tapestry’s designer, there is reason to think that rather more lay behind them. It is noticeable, for example, that most of the heavily-armed English fighting solely with axes and those with round shields are operating in looser order than the men in the shield-wall.49  It looks as if their weaponry sometimes reflected the functions that soldiers were expected to perform. Creating a formation of overlapping shields was perhaps easier if they were kite-shaped, that is with less width than a round shield and offering more protection to the body of a man standing side-on, and in a position to push with his shoulder behind the shield if necessary. On the other hand, the round shield with its large boss may have been more suited to men fighting in open order,50 while the space required to swing the two-handed axe suggests that the bulk of its bearers must have fought in this way too. Accordingly, one should not be too ready to assume, as might seem natural, that the Tapestry scene showing the deaths of Harold’s brothers represents a stage in the battle after the breaking of the shield–wall. It looks as though not all the English were occupied in that wall: there were men armed solely with axes, round-shield men, archers, if present in any number,51 and of course the specialist light infantry on the hillock. One might add that the latter would also have been best–suited to pursuing a defeated enemy in armies which had no cavalry, if we are safe in assuming that the companies who performed this role at Brunanburh in 937 were not on horseback.52

Of course, there is no reason why Anglo-Saxon military methods should have remained static between the late ninth and mid-eleventh centuries, and some evidence that they did not. Both the fortification of burhs in stone rather than timber,53 and Ćthelred II’s requirement that equipment should include helmets and byrnies to an apparently greater degree than had been the case earlier,54 demonstrate a willingness to innovate. The use of the shield-wall, however, changed little. In Bishop Asser’s account of the battle of Ashdown in Sir the enemy split into two divisions, forming shield-walls of equal size, whereupon the West Saxons did the same; at Edington in 878 Alfred fought cum densa testudine against the entire pagan army,55 while at Farnham in 893 Ćthelweard says that Edward the Elder engaged in an agmine denso.56  Similarly, we hear of the use of the shield-wall again at Maldon in 991, and William of Poitiers’s comments on the density of the English line at Hastings must surely refer to the same formation, which is represented, after a fashion, in the Bayeux Tapestry.57 Asser’s word for men so arrayed is testudo, which originally denoted a formation of shields used by Roman legionaries, and it was perhaps the implications of such a parallel that led Spatz to deny the use of a shield-wall at Hastings in the most important sentence in his book:

Only an army trained in protracted military exercises, such as an army made up of tactical corps would be, would be capable of forming such                     an extraordinarily difficult formation as a long, continuous shield-wall.58

Exactly so. A formation probably many ranks deep and fronted by overlapping shields must have taken time to organize, and whether it was employed defensively or, as at Ashdown, offensively, would not be easy to maintain in the stress of battle, as the dead and wounded left the ranks and had to be replaced. Nor is it likely that such formations were the sole province of the well-equipped household troops of great lords, as the Tapestry depiction of an armoured line might lead one to argue, for it is clear from The Battle of Maldon and probably from the Alfredian references that shire levies were incorporated in them too. Moreover, it is virtually inconceivable that a society to which war was so vital did not train levies to this end, a training to which there may be an oblique reference in the statement that at Maldon Ealdorman Byrhtnoth told his men how to stand and asked that they should hold their shields properly, later instructing them to form the ‘war-hedge’;59 and if as part of a national army shire levies might be expected to form a single line with others of their kind, one would think that they must have had training in that as well. How, otherwise, would different groups have known where to position themselves? As noted above, Norman complaints about the density of the English line at Hastings suggest that a great shield-wall was formed there, even if not all their troops were involved in it, and it may be that the time needed to array a sizeable army in such a fashion allowed the enemy to attack before, according to the Anglo-Saxon Chronicle D text, they were properly ordered. Hints of the sort of methods employed to create large formations might be derived from the hoplites of ancient Sparta. They fought in close order with shield and thrusting spear in a manner not dissimilar to that of the shield-wall. In Xenophon’s time a hoplite mora was made up of four lochoi, the latter having two pentekostyes of two enomotiai apiece.60 The enomotia may have contained thirty-six men, and if so the lochos had 144.61 It was the existence of sub-units which allowed the phalanx to form up and manoeuvre, as was also the case with the Roman legion, and it would be surprising if the shield-wall did not contain equivalents. Certainly it is worth noting that terms such as eored-heap, eored preat, eored-weorod, heap, scild-truma, truma and weorod may sometimes have had more precise meanings than the simple `troop’ and `company’ which is the best we can do in translating them today.62

There is little about the battle of Hastings which is certain. If it is impossible to prove Spatz, Ramsay, Baring and their many disciples wrong in believing it a relatively small-scale affair fought on a restricted hill top,63 the paucity of the evidence and the weakness of their arguments mean that it is far from clear that they were right either. The Bayeux Tapestry has limitations, but it is a reasonable supposition that the scene depicting a hillock and preceded by the words `Hic ceciderunt simul Angli et Franci in prelio’ records an important phase of the battle, which the presence of the water probably locates a considerable distance from the crest of the ridge,64 and not improbably in the vicinity of present-day New Pond and its adjacent ground; if so, the initial English deployment may have been much more extensive than has usually been acknowledged. Similarly, there is no reliable way of fixing the sizes of the armies at a few thousands rather than tens of thousands, but the Tapestry does have indications of an English force which fought in various ways, capable of being adapted to varied terrain and a varied enemy: field defences were some sort of answer to heavy cavalry, if not in this case a wholly successful one. To note these points is both to loosen the straitjacket imposed on this subject in the late nineteenth century and to bring within the bounds of possibility an important hypothesis: that the late Anglo-Saxon state, the culmination of six centuries which formed `an integral part of the history of one of the most successful human organizations there has ever been’,65 possessed a powerful and flexible military system, which reflected the wealth, administrative sophistication and intelligence to be seen in others of its works, and was itself reflected in the size and tactics of the last army it put into the field; an army overcome by an amphibious operation on an astonishing scale, and a narrow margin.

Appendix

Limitations of space prevent full consideration here of the reasons advanced by Spatz, Baring and Ramsay for limiting the armies to less than 10,000 men apiece; none carry much conviction, but here is a little of the argument. Spatz Die Schlacht von Hastings, pp. 28-9, thought that the French force must have been relatively small partly because disembarkation at Pevensey took only a day when in 1854 the landing in the Crimea of 60,000 men, just over 1000 cavalry and 128 cannon took five days; and that William could not have marched a large army some seven miles from Hastings to Battle in a brief morning period before engaging the English. It is uncertain whether he did disembark so quickly (the dates of 28 and 29 September given by the D and E versions of the Anglo-Saxon Chronicle hint at a two-day landing), and analogies of this sort are in any case dangerous as evidence of what he could or could not have done. If they are to be employed, there are others which might tend to different conclusions. For example, in 54 BC Julius Caesar, with five legions and 2000 cavalry (i.e. something over 25,000 men) in more than Boo ships, reached the British coast about midday, made an unopposed landing, chose a suitable spot for a camp and marched against the Britons shortly after midnight: De Bello Gallico, v. 8-9, ed. Rice Holmes, pp. 177-80. Furthermore, there is no certainty that William could not have marched a very large army rapidly to Battle, even if there was such a march, which is very doubtful. Spatz relied here on William of Poitiers (ed. Davis and Chibnall, pp. 116-23), who says that after an exchange of ambassadors Harold hastened his advance to take the duke by surprise, that scouts reported his approach and that William then assembled the soldiers in his camp, as most had gone out foraging. Whatever the size of his command, this leaves scant time to recall the foragers, assemble in marching order, march to Battle and deploy, all before 9 o’clock, when William of Jumieges and John of Worcester say the fighting began, even on the generous assumption that the foragers had been operating during the night: William of Jumieges, Gesta Normannorum ducum, ed. van Houts, ii. 168; John of Worcester, Chronicon, ed. Darlington and McGurk, ii. 604. In fact, to allow his army to forage if he suspected that the enemy were in the vicinity is not the sort of procedure one might have expected the Conqueror to adopt, and it is likely that we are being led astray by William of Poitiers’ love of classical allusions, which are a major problem when trying to establish the veracity of his account; see R. H. C. Davis, `William of Poitiers’. He `moves about Caesar’s Gallic Wars with the ease of a master, using its facts solely as they are relevant to his purpose’, and it is noteworthy that in 55 BC the Seventh Legion was attacked by the Britons when foraging and had to be rescued by Caesar with soldiers assembled from his camp, while in 54 BC three of his legions and all his cavalry were attacked while out foraging: De Bello Gallico, iv. 32-4, v. 17, ed. Rice Holmes, pp. 165-7, 190-1. These are not precise parallels, but William of Jumieges (ed. van Houts, ii. 168) says that Duke William, fearing a night attack, ordered his army to stand to arms from dusk to dawn, before arranging his men in three divisions and moving forward after daybreak to join battle. On balance, this is a preferable tale, and allows more time for even a very large force to march from Hastings, if that is what it did: Jumieges would sanction the interpretation that the French were near Battle the evening before the engagement, and this would make it easier to understand the Anglo-Saxon Chronicle D text’s claim that the English were attacked before they were properly organized; above, p. 78.

End Notes

* George Garnett encouraged the work upon which this essay is based; Gillian Drew, Librarian of St Paul’s School, acquired books and articles; my colleague Mr. J. R. M. Smith gave generously of his time on the subject of ancient warfare. I am grateful to them.

1. The Gesta Guillelmi of William of Poitiers, ed. R. H. C. Davis and M. Chibnall (Oxford, 1998), pp. 126-7.
2. N. P. Brooks and H. E. Walker, `The Authority and Interpretation of the Bayeux Tapestry’,Anglo-Norman Studies, 1 (1978), 1-34, 191-9, esp. 8-10; D. J. Bernstein,The Mystery of the Bayeux Tapestry (London, 1986), passim.
3. J. Campbell, `The Significance of the Anglo-Norman State in the Administrative History of Western Europe’, inEssays, pp. 171-89, at 178-9.
4. The Bayeux Tapestry, ed. D. M. Wilson (London, 1985), plates 65-7.
5. E. A. Freeman,The History of the Norman Conquest of England(6 vols, Oxford, 2nd edn, 1870-9), iii, map opposite p. 443, Pp. 471, 477.  The hillock is marked as `The English Outpost’. J. Bradbury, The Battle of Hastings (Stroud, 1998), pp. 176, 189, states (incorrectly) that the hillock is `very small’, and suggests (implausibly) that the Tapestry scene is simply another depiction of the fighting on the main ridge.
6. W. Spatz, Die Schlacht von Hastings (Berlin, 1896), pp. 30, 33; the extensive litera–ture produced between 1892 and 1898 during Round’s clash with T. A. Archer and Kate Norgate is listed by J. H. Round, `The Battle of Hastings’,Sussex Archaeological Collections, 42 (1899), 54-63 at 63; William of Poitiers, ed. Davis and Chibnall, pp. 128-9, 130-3.
7. J. H. Ramsay,The Foundations of England(2 vols, London, 1898), ii. 24-6; F. H. Baring, `On the Battle of Hastings’, appendix B of his Domesday Tables (London, 1909), pp. 217-22. Baring’s map of the field and the dispositions of both sides remains the best available on the contours of the site. Ramsey (ii. 16) assessed the French at 5000, Baring (p. 219) at 8000 to 10,000, and the English at 10,000 to 13,000, `many of them . . . rustics’.
8. For example, F. M. Stenton,Anglo-Saxon England(3rd edn, Oxford, 1971), pp. 592-5; J. Beeler, Warfare in England, 1066-1189 (New York, 1966), pp. 11, 16-17; R. Allen Brown, `The Battle of Hastings’, Anglo-Norman Studies 3 (1980), 1-21, 197-201, at 10-11; elsewhere, Brown warned that figures of about 7000 on each side `are more or less rational guesswork’: The Normans and the Norman Conquest (London, 1969), p. 150, n. 47; The Battle of Hastings: Sources and Interpretations, ed. S. Morino (Woodbridge, 1996), pp. xxii-xxx, with the statement that `the size of both armies is open to debate’, but maps showing an initial English deployment on the crest of the ridge (presumably as we `know the basic disposition of troops on either side’). For further discussion of the size of the armies, see Appendix, below, pp. 90-1.
9. Two of the Saxon Chronicles Parallel, ed. C. Plummer (2 vols, Oxford, 1892-9), i. 198-9. The date of D is uncertain. D. N. Dumville, `Some Aspects of Annalistic Writing at Canterbury in the Eleventh and Early Twelfth Centuries’,Peritia, 2 (1983), 23-57, at 34, suggests extreme outer limits of 1080 x 1130 for its compilation.
10. John of Worcester,Chronicon, ed. R.R. Darlington and P. McGurk (Oxford, 1995) ii. 604; on the dating limits, see p. lxxxi.
11. William of Poitiers, ed. Davis and Chibnall, pp. i 26-9; William of Jumieges,Gesta Normannorum ducum, ed. E.M.C. van Flouts (2 vols, Oxford, 1992-5), ii. 166.
12. William of Poitiers, ed. Davis and Chibnall, pp. 128-31.
13. Freeman,Norman Conquest, iii. 490-1; C. H. Lemmon, `The Campaign of 1066′, inThe Norman Conquest: Its Setting and Impact, ed. D. Whitelock et al. (London, 1966), p. 106. Freeman quoted William of Malmesbury’s Gesta regum at this point, which speaks of the English occupying a tumulus after a French feigned flight, and of many enemy deaths in a deep ditch: De gestis regum Anglorum libri quinque, ed. W. Stubbs, Rolls Series, 2 vols (London, 1887-9), ii. 303; but this may be Malmesbury’s own interpretation of the Tapestry. The various accounts of deaths in ditches, including the so-called Malfosse incident, are well treated in Brown, `The Battle of Hastings’, 18-21.
14. Ibid., 20.
15. The Bayeux Tapestry, ed. F. M. Stenton (London, 1957), p. 175.
16. Two of the Saxon Chronicles Parallel, ed. Plummer, i. 108. I have translated the Old Englisheorod-cist here on the assumption that the second element derives from the verb ceosan, to choose; see Plummer’s glossary, ibid., 328, and The Anglo-Saxon Chronicle, trans. M. J. Swanton (London, 1996), p. 108. The word cist can, however, also denote a company, and `mounted companies’ is an alternative: English Historical Documents, c.500-1042, ed. D. Whitelock (2nd edn, London, 1979), p. 219. Swanton states that the word eored derives from eoh, `war-horse’; see below, p. 87.
17. The Bayeux Tapestry, ed. Wilson, pp. 192-3. Naturally, it is impossible to be sure that they do not depict vegetation; Brown, `The Battle of Hastings’, n. 120, suggested `tufts of marsh grass’.
18. Wace,Roman de Rou, ed. A. J. Holden (3 vols, Paris, 1970-3), lines 6969-72, 7847-8, 8079-96; ii. 143, 176, 185-6.
19. K. J. Leyser, `Early Medieval Warfare’, in hisCommunications and Power in Medieval Europe: The Carolingian and Ottonian Centuries, ed. T. Reuter (London, 1994), pp. 29-71, at 48-9; citing, on Margrave Henry, Regino of Prum,Chronicon, ed. F. Kurze (Hanover, 1890), s.a. 887, p. 126. Worth noting, too, are Professor Leyser’s comments on the sophistication of the tactics known from continental sources.
20. Apart from Baring’s map, the most useful one is the modern Ordnance Survey Pathfinder Map, sheet no. 1290 (TQ 61/71) at 2˝ inches to 1 mile.
21. Above, p. 76.
22. C. W. Hollister, `The Greater Domesday Tenants-in-Chief, inDomesday Studies, ed. J. C. Holt (Woodbridge, 1987), pp. 219-48, at 221-6; E. M. C. van Houts, `The Ship List of William the Conqueror’,Anglo-Norman Studies, 10 (1987), 159-83. It is edited by Dr van Houts at 176.
23. Roman de Rou, ed. Holden, lines 6423-32; ii. 123; William of Jumieges,Gesta Normannorum ducum, ed. van Houts, ii. 164.
24. Roman de Rou, ed. Holden, lines 6119-21, 6163-7; ii. 112, 114; van Houts, `Ship List’, 162-3, 168. See also C. M. Gillmor, `Naval Logistics of the Cross-Channel Operation, 1066′,Anglo-Norman Studies, 7 (1984), 105-31. This article, which appeared before recent work on the ship list, argues that there is no proof that William’s magnates supplied all the vessels requested, and that building Wace’s 696 ships would have severely taxed the timber and labour resources of eleventh-century Normandy. This may be so, although it might seem doubtful whether William would have settled for much less than he wanted, and one could doubtless use similar methods to show that Caesar’s men, despite a shortage of materials, could not have produced 600 ships and twenty-eight warships over the winter of 55/54 BC: De bello gallico, v. 2, ed. T. Rice Holmes (Oxford, 1914), p. 171. However, Gillmor stresses the possibility that many of William’s vessels came from existing stock, were supplied by his allies, or built from extant supplies of seasoned timber.
25. William of Poitiers, ed. Davis and Chibnall, pp. 102-5, 108-11.
26. Roman de Rou, ed. Holden, line 2882; i. 110; Gillmor, `Naval Logistics’, 107. I have measured the distance from Bavent to Dives-sur-Mer on the Michelin map no. 231, Normandie, at a scale of 1 cm. to 2 km. Gillmor gives 4.8 km. in error. B. S. Bachrach, `Some Observations on the Military Administration of the Norman Conquest’,Anglo–Norman Studies, 8 (1985), 1-25, at 6, offers a map of the area, but with an inaccurate scale. The form of the lower Dives at this date is not, of course, known with any preci–sion; see further R. N. Sauvage, L’abbaye de Saint-Martin de Troarn au diocese de Bayeux des origines au seizieme siecle (Caen, 1911), pp. 245-52, who thought it `un vaste lac envahi par les marees’.
27. These calculations have been made from the map prepared by J. A. Williamson and printed inThe Carmen de Hastingae Proelio, ed. C. Morton and H. Muntz (Oxford,1972), p. 110.
28. William of Poitiers, ed. Davis and Chibnall, pp. 100-3; Hollister, `Greater Domesday Tenants-in-Chief ‘, p. 223.
29. William of Poitiers, ed. Davis and Chibnall, pp. 110-13; R. H. C. Davis, `William of Poitiers and his History of William the Conqueror’, inThe Writing of History in the Middle Ages, ed. R.H.C. Davis and J.M. Wallace-Hadrill (Oxford, 1981), p. 78.
30. William of Poitiers, ed. Davis and Chibnall, pp. 118-21; van Houts, `Ship List’, 168.
31. William of Poitiers, ed. Davis and Chibnall, pp. 102-3, 116-17.
32. F. Lot, L’art militaire et les armees au moyen dge en Europe et dans le Proche Orient (Paris, 1946), i. 285 n. 2; for example, De Bello Gallico, ed. Rice Holmes: 60: v. 5, p. 174; v. 23, p. 196. 600: ii, i5, p. 82; ii. 28, p. 93; iii. 22, p. 122; v. 2, p. 171. 6000: i. 27, P. 32; i. 48, P. 59; ii. 29, P. 94; iv. 37, pp. 168-9; viii. 17, p. 375. 60,000: ii, 4, p. 69; ii. 28, P. 93; v. 49, P. 219; vii. 83, p. 355. (Book viii was not, of course, written by Caesar.)

33. See Appendix, below, pp. 90-1.
34. D. Dumville, `The Tribal Hidage: An Introduction to its Texts and their History’, inThe Origins of Anglo-Saxon Kingdoms, ed. S. Bassett (Leicester, 1989), p. 227; J. H. Round,Feudal England (London, 1909), pp. 249, 251, 253.
35. The Defence of Wessex: The Burghal Hidage and Anglo-Saxon Fortifications, ed. D. Hill and A. R. Rumble (Manchester, 1996), pp. 30-1, 34-5.
36. Thietmar,Chronicon, ed. R. Holtzmann (2nd edn, Berlin, 1955), vii. 40; pp. 446-7. 37. II Ćthelstan 16:Die Gesetze der Angelsachsen ed. F. Liebermann (3 vols, Halle, 1898-1916), i. 158.
37. H. C. Darby,Domesday England(Cambridge, 1977), p. 336. And the statistics are incomplete; for ploughlands and ploughteams, see ibid., pp. 95-136.
38. Two of the Saxon Chronicles Parallel, ed. Plummer, i. 101-3.
39. Ibid., 151.
40. A point made by C. W. Hollister,Anglo-Saxon Military Institutions(Oxford, 1962), pp. 125-6. There has been considerable recent appreciation of the late Anglo-Saxon naval system: for example, M. Strickland, `Military Technology and Conquest: The Anomaly of Anglo-Saxon-England’, Anglo-Norman Studies 19 (1996), 353-82, at 373-8; N. A. M. Rodger, The Safeguard of the Sea: A Naval History of Great Britain (London, 1997), i. 14-30, including the suggestion that some of Alfred’s vessels were among the biggest `built in northern waters during the Middle Ages’. I owe this reference to James Campbell.
41. See F. Barlow,Edward the Confessor(London, 1970), pp. 270-2.
42. Capitularia regum Francorum, i, ed. A. Boretius (Hanover, 1883), no. 77, c. 10; translated by P. D. King,Charlemagne: Translated Sources (Kendal, 1987), p. 245.
43. DB i. 56c;Capitularia, ed. Boretius, no. 48, c. 2; King,Charlemagne, p. 260. See further, Hollister, Anglo-Saxon Military Institutions, pp. 42-3, and on general similarities between Carolingian and Anglo-Saxon government, J. Campbell, `Observations on English Government from the Tenth to the Twelfth Century’, Essays, pp. 155-70, at 159-66.
44. Capitularia, ed. Boretius, no. 49, c. 2; King,Charlemagne, p. 257; DB, i. Soa; The Battle of Maldon, ed. and trans. D. Scragg, in The Battle of Maldon, AD 991, ed. D. Scragg (Oxford, 1991), pp. 18-31 .
45. Capitularaa, ed. Boretius, no. 75, no. 77, c. 9; King,Charlemagne, pp. 260, 244; Two of the Saxon Chronicles Parallel, ed. Plummer, i. 138.
46. The Bayeux Tapestry, ed. Wilson, plates 61-2; one holds a small axe but has spears and a shield too, another only has a double-handed axe.
47. Ibid., plates 67, 73.
48. Ibid., plates 64-5, 70-2.
49. ‘By efficient movement of his shield arm he could ward off blows from above, side and to front’: I. Peirce, `Arms, Armour and Warfare in the Eleventh Century’,Anglo–Norman Studies, 10 (1987), 237-57, at 244. The hoplites of the ancient world, of course, did use round shields in forming lines of battle: J. K. Anderson, `Hoplite Weapons and Offensive Arms’, inHoplites: The Classical Greek Battle Experience, ed. V. D. Hanson (London, 1991), pp. 15-37, at 16.
50. As is well known, the Tapestry shows only one English archer, The Bayeux Tapestry, ed. Wilson, plate 62. But there is a danger of being misled by its limitations. If English shield-walls were occasionally fronted by lines of skirmishing archers who withdrew to the rear as the enemy approached, this would not have been easy to repre–sent in such a medium, especially considering the danger of confusion with the many French bowmen in the lower border. Also, the designer seems to have had relatively little interest in depicting troops of modest social origin.
51. Above, p. 79 n. 16. The long-held view that the Anglo-Saxons never used cavalry has recently begun to disintegrate; see for example, Strickland, `Military Technology’, 359-60.
52. Above, p. 84.
53. Above, p. 86; N. P. Brooks, `Arms, Status and Warfare in Late-Saxon England’, inEthelred the Unready: Papers from the Millenary Conference, ed. D. Hill (Oxford, 1978), pp. 81-103, at 85-90.
54. Asser’s Life of King Alfred, ed. W. H. Stevenson (Oxford, 1904), cc. 37, 56; pp. 28, 45. See the translation inAlfred the Great, trans. S. D. Keynes and M. Lapidge (London, 1983), PP. 79, 84.
55. The Chronicle of Ćthelweard, ed. A. Campbell (London, 1962), p. 49.
56. The Battle of Maldon, ed. Scragg, lines 101-2, 242; pp. 22, 28; above, p. 76 n. 6.
57. Spatz,Die Schlacht von Hastings, p. 45. `Nur ein in langwierigen militarischen Exerzitien geschultes, also aus taktischen Korpern zusammengesetztes Heer wdrde imstande sein, eine so ausserordentlich schwierige Formation, wie sie ein lang fortlaufender, gerader Schildwall ist, zu bilden.’ `Tactical corps’ was Spatz’s term for highly-trained troops. I am grateful to my colleague Paul Collinson for verifying the translation of this passage.
58. The Battle of Maldon, ed. Scragg, lines 17-20, 101-2; pp. 18, 22.
59. Xenophon,Spartan Society, c. 11; translated inPlutarch on Sparta, trans. R. J. A. Talbert (London, 1988), p. 178.
60. P. Connolly,Greece and Rome at War(London, 1981), pp. 39-41. But it should be stressed that this is speculation, and noted with reference to Xenophon’s remarks that `the size and precise relationship to one another of the units mentioned both here and later are obscure’: Talbert, in Plutarch on Sparta, p. 187.
61. See the entries inAn Anglo-Saxon Dictionary, ed. J. Bosworth and T. N. Toller (Oxford, 1898), andSupplement.
62. To be precise, the restriction of the battle to the crest of the ridge originated with Ramsay. Spatz gave no plan of the field, but seems to have thought that that the English deployed over the full extent of what he described as a narrow hill with a depth of roughly 150 yards and length of only about one mile:Die Schlacht von Hastings, p. 33.
63. This is not certain, as the western part of the field, which stands on Wadhurst Clay, becomes noticeably waterlogged after rain, and both the monastic fishpond to the east of the abbey terrace and another pond on the northern edge of the hillock contain water all the year round. It is therefore conceivable, if very unlikely, that there was water closer to the crest than that in the New Pond area.
64. Anglo-Saxons, ed. Campbell et al., p. 246.

This article was first published in The Medieval State: Essays Presented to James Campbell, ed. J.R. Maddicott and D.M. Palliser (London: Hambledon Press, 2000).  We thank M.K. Lawson for his permission to republish this article.