Bayeux Tapestry scene showing Norman cavalry attacking the Anglo-Saxon shield wall at Hastings

Hastings 1066: We Know Who Won. Do We Know Where They Fought?

Iron Standard — Update 2 · 7 September 2026. This revision reworks the source genealogy, tests the method as well as the competing battlefield theories, and records more clearly where the evidence stops. The conclusion remains provisional.

Battle Abbey has marked the accepted site of Hastings for nearly nine centuries. Yet the ground has still not yielded a securely identified object from the fighting of 14 October 1066. Rival sites have therefore multiplied. The interesting question is no longer whether tradition can be challenged. It is whether any alternative explains the surviving evidence better.

If you stand among the ruins of Battle Abbey, the story is presented with unusual precision. Harold died here; William won here; the high altar of the abbey church was placed over the spot where the last English king of the old order fell. It is one of the most powerful pieces of historical landscape in Britain.

There is also an awkward fact beneath it. No weapon, arrowhead, shield fitting, mass grave or other object has yet been securely identified as belonging to the battle of 14 October 1066 at Battle Abbey.

That matters. A major battle involving thousands of men ought, instinctively, to leave something. But the absence is not the same thing as proof that the battle happened elsewhere, and too much of the argument around Hastings has depended on confusing those two statements.

So this investigation began again, not with English Heritage and not with any alternative-site campaign, but with the earliest surviving written evidence. Then we asked a second question that is just as important: are those sources really independent of one another, or are several writers simply repeating the same story?

The abbey tradition is older than the abbey chronicle

The late-twelfth-century Chronicle of Battle Abbey is not a source anyone should read innocently. By the time it was composed, the abbey was defending privileges, arguing with the bishops of Chichester and constructing an institutional memory in which William the Conqueror’s wishes mattered enormously. Some charters used in those disputes were forged. Any claim that conveniently magnified the abbey’s royal foundation deserves suspicion.

That has encouraged an attractive counter-argument: perhaps the monks created the battlefield story later to make their house more important.

The problem is chronology.

The Anglo-Saxon Chronicle’s obituary of William I, under 1087, says that William founded a monastery on the place where God had granted him the conquest of England. However we translate the exact phrasing, the association between the abbey and the victory existed long before the late Battle Abbey chronicle. By the early twelfth century William of Malmesbury, John of Worcester, Orderic Vitalis and Henry of Huntingdon were all connecting the abbey with the fighting or with Harold’s death. An early life of William produced at Battle Abbey does the same.

That does not give us five independent witnesses. Medieval writers borrowed. Monasteries exchanged texts and stories. Royal foundation traditions travelled. A single early Norman account could, in principle, sit behind several apparently separate chroniclers. Counting authors is therefore a poor substitute for tracing where their information came from.

Even after allowing for that, however, the alternative-site case has a harder job than is sometimes admitted. It is not enough to discredit a twelfth-century monk. A rival theory must explain why the battlefield association was already established within a generation of William’s death and why no comparably early surviving tradition places the main battle somewhere else.

That is not proof of Battle Abbey. It is simply evidence that has to be beaten.

Early does not mean infallible

There is an obvious danger at the other extreme. Because a tradition is early, historians can start treating it as though it were a witness statement.

It was not.

William had every reason to bind a religious foundation to the event that gave him a kingdom. The abbey had every reason to preserve that connection. Later chroniclers had every reason to repeat a story attached to a royal foundation. Memory can become official very quickly, especially after a conquest.

The physical siting of the abbey is nevertheless curious. This was not the easiest piece of Sussex on which to build a major monastery. The crest and slope forced awkward engineering and terracing. That is consistent with a king insisting on a meaningful spot rather than monks selecting the most convenient building ground. It is circumstantial, and it should stay circumstantial.

What would be far better is archaeology.

So far, archaeology has refused to settle the argument.

The silence in the ground

English Heritage does not hide the problem. No object securely tied to Hastings has been recovered from the accepted battlefield. Time Team’s 2013 work, involving trenches, metal detecting and landscape analysis under battlefield archaeologist Glenn Foard, did not produce the hoped-for Norman or English military assemblage.

That is the strongest piece of negative evidence against the traditional site.

It is also easy to overstate. Iron corrodes. The local soils are not kind to metal. Medieval battlefields were stripped of useful equipment almost immediately. Bodies were buried or removed. The abbey then altered the most important part of the ridge through construction, rebuilding, drainage and centuries of occupation. Much of the possible eastern battlefield is now under the town.

None of those explanations should be used as a universal escape hatch. If every failed search can simply be blamed on corrosion, later building or incomplete coverage, then the traditional site becomes unfalsifiable. That would be bad history.

The fairer point is that the alternatives suffer from the same absence. Crowhurst, Caldbec and Heathfield have not produced an accepted battlefield assemblage either. At the moment, archaeology has failed to confirm Battle; it has not confirmed somewhere else.

The neat battlefield is probably the wrong battlefield

The landscape south of the abbey is often pictured as a clean uphill approach: William in the south, Harold on the ridge, two armies facing one another almost squarely. That reconstruction is convenient for diagrams and visitors. It may also be too tidy.

LiDAR and terrain work associated with the 2013 investigation identified former watercourses and wetter ground in the valley. One interpretation was that William’s approach was more south-easterly and that substantial fighting may have taken place perhaps two hundred metres east of the later abbey church, around the Upper and Lower Lake area.

That does not move Hastings away from Battle. It changes the shape of the fight.

If the Norman approach was oblique, the English line may have run farther east than the familiar reconstruction. Men could have fought across ground now occupied by streets and buildings. A battle that lasted most of a day did not need to fit inside the boundaries of a modern heritage property.

Christopher Macdonald Hewitt’s GIS work is useful here because it pushes against another convenient assumption: that a thousand years of erosion has made the modern terrain almost meaningless. His modelling suggested the broad form of the landscape may have changed less than some arguments require. The present ground cannot be treated as a perfect copy of 1066, but nor can it simply be wished away whenever it causes a theory difficulty.

Crowhurst

Nick Austin’s Crowhurst theory moves the principal battlefield roughly three miles south. It has the attraction of bringing the fight closer to William’s coastal approach and has been built through years of work on roads, local topography, the Malfosse tradition and claimed finds.

The claimed material evidence is where the theory must ultimately stand or fall. Objects described as possible helmet fragments and other military traces have attracted attention, but they have not become an archaeologically accepted 1066 assemblage. Time Team examined parts of the Crowhurst case and did not uncover the decisive Norman military context that would justify overturning the early Battle tradition.

That does not make Crowhurst foolish. It makes it unproved.

The most difficult question remains documentary. If the principal fighting took place three miles away, why was William’s abbey being associated with the battlefield so early? Royal propaganda is one answer. Memorial rather than literal siting is another. A loose medieval use of “on the battlefield” is possible too. But those are explanations for inconvenient evidence, not evidence in themselves.

Crowhurst needs something positive enough to outweigh the early tradition. It has not yet produced it.

Caldbec Hill

Caldbec Hill, about a mile north of the abbey, starts from a soldier’s objection. Why would Harold, exhausted after the northern campaign and relying on infantry, abandon stronger high ground for a lower ridge?

John Grehan and Martin Mace have pressed that point, and it deserves to be taken seriously. Caldbec also becomes entangled with the Anglo-Saxon Chronicle’s mysterious haran apuldran, usually translated as the “hoary apple tree”, where Harold is said to have assembled his force before William came upon him while it was not fully deployed.

The trouble is that the tree has vanished and its function is unclear. It may have been the battlefield, a muster point, a landmark on a route, or something known to contemporaries whose geography is lost to us. Moving an entire battle because a modern hill looks tactically attractive risks turning military common sense into location evidence.

Commanders do not always occupy the best contour on a modern map. Roads, timing, reconnaissance, supply, enemy movement and incomplete information all interfere. Caldbec is therefore a worthwhile alternative reconstruction, but it has not yet solved the documentary or archaeological problem.

Heathfield Down

The Heathfield proposal deserves more attention than most distant alternatives because Rebecca Welshman and Simon Coleman have developed it through peer-reviewed research rather than simply local tradition. Their case places the battlefield roughly ten miles north-west of Battle Abbey and combines routeways, military geography, the hoary apple tree reference, medieval place-name evidence and later references to Heathfield or Heathfield Down.

The medieval place-name Horeapeltre is provocative. So is the later name Slaughter Common. But place-names are treacherous evidence. Similar names can arise independently, and later folklore often grows around places precisely because people already believe a battle happened there.

The strongest version of the Heathfield argument is methodological rather than archaeological: perhaps the apparently multiple early references to Battle Abbey all descend from one official Norman foundation story. If so, their numerical weight has been exaggerated.

That is possible. It is also not yet demonstrated.

Heathfield currently lacks the material signature required to overcome the distance from Battle and the early abbey tradition. A coherent, securely dated eleventh-century battlefield scatter there would change the argument dramatically. Until then, it remains a serious hypothesis rather than a replacement location.

The source genealogy problem

This is the point at which our own method needs challenging.

It would be easy to make the traditional case look stronger by lining up the Anglo-Saxon Chronicle, William of Malmesbury, John of Worcester, Orderic Vitalis, Henry of Huntingdon and Battle Abbey itself, then treating each as a vote. That is not how medieval evidence works.

Some of these texts may share written or oral ancestors. Some authors may simply have accepted an already established royal story. We do not possess a complete map of their dependencies. For that reason the article does not assign a numerical probability based on how many chroniclers repeat the claim.

But the opposite reconstruction carries assumptions too. To move the battle away from Battle, we must suppose that a false location tradition arose very early, travelled successfully through several textual environments, attached itself to a conspicuously awkward abbey site and displaced any genuine local memory so thoroughly that no clear competing location entered the surviving early record.

That can happen. Conquests produce official memory. Monasteries manufacture legitimacy. Texts copy one another. What we still lack is direct evidence showing that this happened here.

Harold and the high altar

Three different claims are usually bundled together: the battle was fought in the Battle area; the abbey stands on part of the battlefield; and the high altar marks the exact place where Harold was killed or found.

They should not carry the same evidential weight.

The first two have an early documentary tradition and a militarily credible landscape behind them. The third is much more precise. Early-twelfth-century writers do associate the church and Harold’s body, so the altar story is not a Victorian invention. But we have no body, no eleventh-century grave beneath the altar, and no secure account of Harold’s manner of death. Even the famous arrow-in-the-eye reading of the Bayeux Tapestry remains disputed.

The altar tradition is old enough to respect. It is not strong enough to turn into a grid reference.

Where the evidence leaves us

The Battle ridge still works as a battlefield. It offers an infantry defender high ground and observation over southern and south-eastern approaches. Wet ground in the lower areas could restrict movement without making an engagement impossible. The eastern ground allows a wider and more complicated battle than the visitor landscape implies. William could reach it from his coastal base without asking either commander to behave irrationally.

Crowhurst shortens the Norman approach but has not overcome the documentary evidence. Caldbec gives Harold attractive defensive ground but relies heavily on reconstructing his intentions from terrain. Heathfield offers an intriguing route and place-name case but must explain both the ten-mile separation and the rapid appearance of the Battle Abbey tradition.

That is why the conventional site survives this investigation. Not because English Heritage says so, and not because medieval monks should be believed automatically. It survives because the alternatives have not yet produced evidence strong enough to displace it.

Our present grading is therefore deliberately uneven. Battle Abbey lying within the battlefield is probable to very probable. The high altar marking Harold’s exact death position is plausible but unproved. The familiar straight Norman assault due north across the modern open slope is probably over-simplified. Crowhurst is currently weak as the principal battlefield; Caldbec is plausible enough to keep testing; Heathfield is a serious but unproved alternative. The exact footprint of the battle remains unresolved.

That last point may matter more than the argument over a single pin on a map. “Battle Abbey” and “the battlefield” were never necessarily the same small patch of ground. The fighting may have spread well beyond the later monastic precinct and into areas now built over.

What would change our mind?

Material evidence would change it fastest. A securely recorded concentration of eleventh-century military artefacts at Crowhurst, Caldbec or Heathfield would force a reassessment. A mass burial securely dated to the battle, a coherent projectile scatter, horse equipment or weapon fragments in a convincing context would outweigh another century of arguments about what a chronicler “must have meant”.

The traditional case could also be damaged by environmental work showing that the Battle ridge could not have supported the deployments described in the sources, or by an early and demonstrably independent text locating the fighting elsewhere.

Conversely, a securely dated military assemblage east or south-east of the abbey would not simply “prove English Heritage right”; it might redraw the accepted battlefield and confirm that generations of diagrams have been too narrow.

For now the best answer is not dramatic. The battle probably did happen at Battle. We are much less certain that it happened in quite the shape visitors have been shown.

SCHOLARLY CHALLENGE

Iron Regiment welcomes substantive examination of the evidence and methodology used in this investigation. Historians, archaeologists, researchers and other specialists are invited to challenge our primary-source readings, source relationships, topographical assumptions and reconstructions. Where appropriate, and with permission, significant commentary may be summarised or referenced in a future update.

Related Regimental Record investigations

From the collection: Hastings 1066 Medieval Battle T-Shirt.

Sources and further reading

Revision record

Update 2 — 7 September 2026: Full editorial rewrite under the expanded Iron Standard. Added explicit source-genealogy testing, methodological self-critique, scholarly challenge, clearer evidential gaps and a less formulaic narrative structure. The principal conclusion did not change.

Iron Standard Investigation. We do not repeat the historical record; we interrogate it. Where the evidence is uncertain, we say so.

Back to blog