Did Harold really get it in the eye?

‘You had your 1917 in 1066,’ a Russian diplomat was once said to have told his British counterpart. The ruling class of England, and much of the rest of Britain, was re-created by the Norman Conquest. Most of the nobly born have at one time or another sought to find progenitors among the Companions of the Conqueror, and the words ‘noble’, ‘gentle’ and ‘aristocrat’ themselves come from French. Within two decades, the Conquest had been commemorated by two astonishing historical artefacts. The Bayeux Tapestry (pre-1082) must be the only artistic masterpiece that is also a crucial source for a major historical event. Domesday Book (1086) provides a cross-section of a society centuries before such information could be extracted from any other source, in England or anywhere else. Then, and since, it has furnished the ruling classes with their title-deeds.

A little over a century ago, the Battle of Hastings was the subject of a scholarly dispute of a virulence not seen again until the Storm over the Gentry or the Condition of England Question in the 1960s. One protagonist, the polymathic E.A. Freeman, echoed some famous words of Macaulay, celebrating the ‘cause for which Harold died on the field and Waltheof on the scaffold’ (Waltheof was the last survivor of the Old English aristocracy: he was executed for treason in 1076 after a rebellion which, according to the sources, was fomented by leading members of the Norman nobility). The other, the acid-penned J.H. Round, found the patronage denied to him by the academic circles that favoured his opponent by supplying the aristocracy with lineages going back to 1066. Soaring effortlessly above the mêlée, F.W. Maitland, a greater historian than either (or anyone then or since), asked to be updated on the progress of ‘the battle’ from his winter resort in the Canaries. We know more about what happened on the field of Hastings that October Saturday than about any battle anywhere since Ammianus Marcellinus chronicled the destruction of the East Roman army by the Goths nearly 700 years before. Yet the price paid for good sources at this stage of history is that they rarely agree.

Take King Harold’s arrow in the eye, a central icon of English history. We have a number of accounts of his death, all of which differ. What may well be the earliest version (though not all historians agree) has him cut down by William himself, together with the author’s local lord and his nephew. This is the least likely to be true. The second earliest source contradicts all the others by recording Harold’s fall in the battle’s first phase. This is almost certainly because the source or his copyist misread the Latin abbreviation for ‘last’. Within 15 years, a writer as far away as Montecassino reported that Harold had been hit in the eye. He is echoed (hardly copied) by a classicising poet writing for the Conqueror’s daughter, who has him killed by an arrow. Contemporary English texts have nothing, but the two best historians of the early 12th century, William of Malmesbury and Henry of Huntingdon, say that Harold’s eye was pierced by an arrow, one making this his death-wound, while the other has him sinking to the ground, where he is finished off by horsemen; William adds that a cavalryman was disgraced for hacking at the corpse’s thigh. The trouble is that both these writers may have seen or known of the Bayeux Tapestry’s portrayal of the event.

The tapestry’s last main scene shows, from left to right, cavalry charging an infantry group comprising a soldier brandishing a spear, a white-bearded standard-bearer who is also shown biting the dust, and a third – the central and most prominent figure – clutching a missile one end of which disappears behind his nose-guard. Then come a horseman striking at the knee of a prostrate man who seems to have let go of an axe, and finally another group of infantry falling victim to cavalry. Along the top of the scene runs the legend ‘Here King Harold is killed’; the word ‘Harold’ is bisected by the helmet of the missile-clutching figure, and the words ‘interfectus est’ are squeezed in above the prone axeman.

Fifty years ago, it began to be argued that the real Harold was the fallen man beneath the words ‘is killed’, and not the central figure, whether or not he has been hit by ” an arrow. This view was dismissed by David Wilson in his superb colour facsimile of 1985. It turns out, however, that the scene as we know it is the result of a restoration in the earlier 19th century. The areas in which work was done included, crucially, the flights on the end of the missile in or near the central figure’s eye and most of the words ‘interfectus est’: the scene as reproduced by two antiquarians a century earlier has a flightless missile extending well beyond the clutching hand. M.K. Lawson decides that the missile was in fact a spear, most of whose length has perished down the ages; that the central figure, not struck by but brandishing it, was merely one of the bodyguard defending the king and his standard; and that the inscription over the fallen figure reads ‘is struck to the ground (in terra iactus est)’. Accordingly, the tapestry does not provide independent and early evidence of Harold’s eye-wound, but lends support to William of Malmesbury’s story that he was struck on the thigh when already dead. Clearly, we’re never going to know for sure, but given the artist’s compositional technique, and that even in the early 18th century the missile in question was less than half the length of the tapestry’s many other spears, with no point anywhere to be seen, the likelihood remains that both the standing and lying figures are Harold; hence William of Malmesbury got it right on both counts, whether or not he derived his knowledge from the tapestry.

The most important thing Lawson does in his study of the battle is to remove one of the few certainties of the last century. Medieval armies, it has been believed, were more or less axiomatically small, because, unlike the Romans, their rulers couldn’t organise anything bigger. That assumption is put in question by the topography of the battlefield at Hastings. The English army is known to have been marshalled along the ridge where the ruins of Battle Abbey now stand, its high altar having marked the spot where Harold fell beside his standard. According to the premier Norman source, William of Poitiers, the English lines were so densely packed that the dead stayed on their feet. To defend the whole ridge in such strength against Normans advancing from the south, there would need to have been 25,000 soldiers, a very large number by ‘medieval standards’.

Freeman’s view was that the army was that size, but egged on by a German scholar for whom it was a point of dogma that medieval armies rarely approached the totals led by Caesar or Moltke, his critics reduced the English line by rejigging the contours. The right flank that was exposed if the ridge was only partly defended was bent back at an angle to the main front, so as to exploit a gully. Closer examination of the site than anyone seems to have attempted before – though to have done so at any time since the 1890s would have required no more than a day trip from Charing Cross Station – leads Lawson to believe that no general in his right mind, or not already in deep trouble, would have settled for such a deployment. By juxtaposing details from the tapestry with well-chosen photographs (including a millennium map aerial view), he shows that the scene of floundering warhorses, when the Normans must have come closest to defeat, probably took place to the ridge’s south-west. The water shown here may well be the ‘sandy brook’ that would have given the field the name Senlac, taken by Freeman from the later but richly detailed account of the Anglo-Norman Orderic Vitalis. We may even glimpse the stakes of the defensive ‘palisade’ recorded in the yet later but even richer verse romance by Wace (also credited by Freeman though scornfully demolished by Round).

The English ‘state’ was eminently capable of organising an army of the size Freeman suggested. The evidence allows, or even encourages us to postulate an elaborate system of assessing the military liability of the kingdom’s various administrative units. Armies would have contained lightly armed elements levied from the general populace, like those shown on the tapestry giving Norman horsemen a hard time, but also heavy infantry – and, when not in defensive mode, cavalry – recruited from men ranked as gentry, whose resources were comparable with those of a later knight or esquire. The government hired professionals, too, whether for the royal bodyguard of ‘housecarls’, or for the navy: the English fleet patrolled the south coast through the summer of 1066, disbanding only when it seemed that the Normans were not going to arrive. Lawson turns the tables on the many historians who have followed William of Malmesbury in believing that English resources were overstretched by the double whammy of a Norwegian attack from the north just before the Normans set sail: given that year’s multiple emergencies, ‘the more that becomes known about the ways in which the English armies assembled and fought . . . the more it may become apparent that it saw the mobilisation of the country and its resources for war to an extent that was not to be repeated until the total wars of the 20th century.’

Lawson’s book is as important for the possibilities it raises as for the conclusions he is sometimes hesitant to draw. Among the questions he doesn’t explicitly answer is why the Normans won. It wasn’t, he shows, a victory for horse over foot, as traditional military commentators have supposed. William of Poitiers and the tapestry may suggest that, but they were speaking to a military and social elite that fought in the saddle. French cavalry is no more likely to have been able to override or outflank the English on their ridge than at Waterloo. Nor was their enticement of an over-hasty counter-charge by a ‘feigned retreat’ (if they made one) any more decisive. As David Crouch points out, most battles in this era were as good as over inside an hour. But Hastings/Senlac lasted all day, as the earliest English source records and most others imply. Everything points to a ding-dong struggle between infantry. The English were not just shot down by arrows but worn down by superior fitness and experience. That is what Crouch argues in a two-page account of the battle which manages to say hardly less about its course than Lawson’s whole book. He makes the highly plausible suggestion that there was a pause in the fighting in late morning, after which the duke regrouped his forces and began an attritional assault, as was perhaps his original plan, abandoned in favour of an early strike. Lawson argues convincingly that the Norman army was around the same size as the English one. Crouch gives, remarkably, a list of ships contributed by the notables of Normandy, the total coming to 1000. Julius Caesar says that he landed 25,000 horse and foot from 800 ships and Norman totals may well have been similar.

Crouch shows very clearly how the interminable campaigning of the decades before 1066 prepared William and his army for their fateful encounter in Sussex. If Harold underestimated what he was up against, he was in good company: the Byzantine emperor, the Saljuq sultan and the emir of Mosul all made the same error. Hastings pales into insignificance beside what the French-speaking soldiery of the First Crusade achieved in taking Jerusalem 33 years later: as big a military surprise as any since Marathon or before Valmy, and likewise the herald of a new age in warfare. It was not so much that northern French chivalry perfected the technique of the mounted charge with lance at rest (only one or two combatants in the tapestry do anything like that), or that this was a freshly ‘feudal’ society confronting ploughboys carrying superannuated kit. More to the point was the all-round professionalism and esprit honed by constant conflict all over 11th-century France. For more than a century this fighting style carried almost all before it, until in 1187, in Palestine, the opposition began to get its act together.

Crouch’s book tells the story of William’s family, one of the two Norman families that made spectacularly good in this age of opportunities. The other rose from obscurity to rule Sicily, the richest and best organised kingdom in the Latin world. William’s family had been dukes for five generations by the time he was born; there were a series of similar families in different parts of the one-time Carolingian empire, but few prospered for so long. There was always something special about the Normans, whether it was the wealth of their seaside province – their coinage, like the English, was unusually high quality – or that, like their Capetian rivals in Paris, they had a large number of sons, even if these were often bastards, like William himself (Henry I, last of the line proper, had a dozen illegitimate sons and one legitimate one, who was drowned with most of the court’s young bloods in a wreck Orderic describes brilliantly). The Normans’ genius lay in targeting the richest and most prestigious corporations in the 11th-century bull-market, then squeezing them for all they were worth and more. They knew how to use the organisations they captured: not their least typical creation was what could be thought of as the West’s first civil service college, the cathedral chapter and school ” of Salisbury.

Among Crouch’s other virtues is that he does not neglect the conquered English, one of whom saved William’s life in a rare defeat, while another, scion of a 1066 survivor, carried Henry’s banner in one of his main victories. English support was arguably crucial when each of the Conqueror’s younger sons saw off their elder brother, Robert, and his Norman backers. Doubts arise, however. One is counter-factual: had all three of William’s sons died within seven years of his death, like Cnut’s, the throne would presumably have been disputed between his half-brother, another Robert, and 1066’s legitimist candidate, Edgar, whose heirs were his Scottish nephews, the sons of Malcolm Canmore. What might then have become of the Norman Conquest – and would we have had an Anglo-Scottish union half a millennium early? More serious is Crouch’s account of William’s grandson, King Stephen, who counts as the last Norman, though no more Norman in descent than Henry II ‘Plantagenet’, and strictly the first and last of the line of Blois. Crouch might have made more of the late W.L. Warren’s argument that Stephen gave his nobles their head, with such apparently lamentable results, because fiscal acquisitiveness was not the done thing in the French world he came from. Crouch sensitively perceives that William and his sons may not have enjoyed being kings, which was why they were so keen on hunting. Our power-obsessed age should make more of the fact that rule is a great strain.

The most striking monument by far to ‘Norman’ government, and to its fiscal appetite, is, of course, Domesday Book. The only satisfactory answer to the familiar question, ‘what was it for?’, is that it was meant to maximise the kingdom’s revenues. Its immediate context was the Conqueror’s hiring of a larger assault force than the Anglo-Saxon Chronicler had ever seen. Among its engineers was probably Ranulf Flambard. He became the chief hatchet-man to William II, who succeeded his father within a year of the great survey, and most of the abuses that Henry I promised to remedy when he took over from William II had to do with the remorseless exploitation of the estates ‘held of the king’ by clerical and lay magnates. Domesday Book came without question to be seen as the ‘Register of Title’ which Maitland for one denied it could have been in conception. But only two of the seven Domesday commissions seem to have seen it as their business to adjudge disputed claims, although the survey rarely fails to record the ‘value’ (meaning, probably, the value to its lord in money rent and marketable produce) of any one manor. Contested titles were decided no doubt because it was necessary to know exactly who was answerable for that ‘value’ to their royal lord.

Domesday Book is the first document made by a government in England, or anywhere else, to have remained in government hands. It is actually not one book but (at least) two. Most counties south of the Tees and Ribble – though not London or Winchester – were covered in a mighty first volume, very largely written by one scribe, who thoughtfully adjusted formulae and lay-out as he ploughed through successive county returns. The commission for East Anglia, however, went into such irrepressible detail that its return was left unedited as Volume II. There is a similar ‘circuit return’ from the South-West – compiled, we now know, by Salisbury’s trained clerks; and on top of that, a number of copies of what seem to be earlier stages in the survey. The transition ‘From Memory to Written Record’, charted by Michael Clanchy a quarter-century ago, has no more eloquent harbinger.

There hadn’t been anything like it before; or if there had, the Anglo-Saxon Chronicler had never heard of it. Yet the system it documents so thoroughly was not introduced by the survey, or at any time in the recent past. Shires and ‘hundreds’ (subdivisions) were well over a century old. The assessment of land in ‘hides’ (that is, notional units equivalent to what would support a farmer’s household) is at least as ancient as Bede’s time, and almost certainly older: that it has counterparts in ancient Irish laws would imply origins in the ‘Indo-European’ era – whatever that was. European government, that is to say, could from remote antiquity make the sort of calculation underlying Domesday’s figures, even if it had never before reproduced them in this form. The Domesday Book heralds the bureaucratic eruption that is one of the hallmarks of the 12th-century Renaissance, but that did not make the gain in government efficacy commensurate with the undoubted boon for historians. On the contrary, because it was produced on the cusp between administration by memory and by record, it shows what governments could do before documentation gave historians a chance to see them at work.

Domesday has been lucky in its publishers. A noble edition in 1783 by Abraham Farley, printer to George III, was followed by a notably good collotype facsimile in the later 19th century. In the 1970s and 1980s, Phillimore produced a serviceable county by county text/translation with adequate maps and decent indexes. But all this is put into the shade by the gigantic Alecto edition underway since 1986, surely one of the most remarkable publishing achievements ever. For Book I there are two huge cases of facsimiles (with the same quiring as the original), two of translation and one apiece of maps and indexes (of places and persons or corporations but not topics) for each county seriatim; then there is a set of county volumes adding an introduction to indexes and translation as with the cased set. For Volume II, we have one fat volume of facsimiles and maps, and one of introduction and translation for each eastern county. There is also an introductory volume of essays on the survey, complete with a glossary of (most, not all) technicalities. The catch, of course, is that the cost is too high for all but the best-endowed libraries. So Alecto and Penguin have combined to produce a single portmanteau volume, by the simple expedient of spatchcocking all the translations, plus the index (of places, not persons, let alone topics) and glossary (but not maps or introduction) into one binding. The intended audience is presumably the interested common reader. Which is where reservations begin.

A common reader might begin by looking up his or her own home and birthplace in the index. I accordingly find on p.718 (original folio 265) that Neston was in Willaston (Wirral) Hundred, Cheshire, and ‘the church held and holds [it], and William [holds] of it. There is a third part of 2 hides paying geld. There is land for 1 plough. It rendered and renders at farm 17s4d.’ The next entry tells me that William also held Raby from that church, and on the page before, I discover that the church was Chester Cathedral. Turning to the glossary (for some reason after the index), I find that the hide was ‘the standard unit of assessment to tax, especially geld’, was ‘notionally the amount of land that would support a household’ and was ‘divided into four virgates’; that geld was ‘the English land-tax . . . assessed on the hide’; that ‘number of ploughlands may (1) estimate the arable capacity of an estate in terms of the number of eight-ox plough-teams needed to work it; or (2) record an assessment of the dues required from the estate’; and that ‘at farm’ means that it was ‘in effect, leased at a specific rent in return for which the tenant, known as the "farmer", received the profits of the estate or office’.

I have been dipping into Domesday Book and explaining it to students for thirty years and have some idea what all this amounts to, but if common readers are puzzled by ‘17s4d’, they will receive no help from the glossary. The index of personal names in the Cheshire volume of the Phillimore edition would tell them that this William is probably the one who held two hides of Tarvin from the Bishop of Chester (which were ‘laid waste’, so that the land was worth barely half what it had been in 1066); that he also held from Chester Cathedral, Middle Aston and Clifton in the next-door hundred; and that what may be the same man held lands from the great Earl Hugh of Chester and his son, and was either (as W. Malbank) Hugh’s tenant for the Nantwich salt-works, or else (as W. fitz Nigel) held the other two-thirds of the two Neston hides directly from the king; all of which would be illuminated by Phillimore’s maps. The experience might incline them to think that they are better off with Phillimore, however superior Alecto’s translation.

Not that the other two books discussed here do much better when it comes to maps. Crouch has two, one marking areas of tenth-century Norman settlement but little else, the other showing ‘Normandy in King Stephen’s reign’, which omits numerous places named in the text and is inexplicably sited 117 pages before the chapter on Stephen. More bizarrely yet, Lawson has colour plates of Freeman’s (admirable) maps and of the British Geological Survey of the Hastings area, but nothing so plain as modern maps of southern England and northern France. Pre-modern government did amazing things without maps. That is no reason to expect us to do the same.