Some Bayeux Tapestry Landscape Secrets

 

The Bayeux Tapestry is the most studied medieval artifact in the world. Shirley Ann Brown’s bibliography lists no less than 307 books and reference papers about it, and she only covers those published by reputable experts before 1988. They drill down into minutiae about the Tapestry's origin, patron, embroidery, embroiderers, physical history, armour, weapons, heraldry, ships, horses, buildings, margins, and much else. It seems unlikely that there could be a useful stone that has been left unturned. But there is at least one: Analysis of its geography is almost non-existent. We will start to address the balance.

History or allegory?

Some aspects of the Tapestry’s attention to detail are legendary. Hundreds of horses, for instance, are accurately tacked down to their bridle rings. The troops too, more than 600 of them, almost all depicted with spears or bows, the English with moustaches, the Normans with undercuts. The Tapestry’s clothing and armour are the standard references for the period. There are some amazing technical details, not least a sailor depth sounding on Panel 5, and the practical application of medieval woodworking tools on Panel 35. As far as we know, these human scenes and artisan techniques are pinpoint accurate.

It is a different story with the landscapes, by which we mean the buildings and geography. Analysis of the Tapestry’s geography is completely absent, other than to cause confusion about whether its hills are natural or mottes. It is little better with the buildings. According to Norman Denny, “The Tapestry ‘suggests’ buildings but does not show them as they were”. He reckons that Westminster Abbey (Panel 26) is the only building on the entire Tapestry where an attempt has been made at an accurate depiction.

David Bernstein explains that the Tapestry relies heavily on copies from psalters. He has identified half a dozen buildings that were copied from known psalters, as well as all the building decorations, tree designs, and some of its best-known set-piece scenes. It is not unreasonable then to think, as most do, that the other buildings and set-pieces were copied from psalters that have subsequently been lost, or were invented by the artist.

The only evidence is thought to corroborate the Tapestry’s lack of landscape fidelity, insofar as the buildings and geography contradict the orthodox invasion narrative. The orthodox first Norman camp at modern Pevensey (Panel 42) should depict the huge Roman fortress of Anderida, but instead depicts a flimsy stone gazebo. The orthodox second Norman camp (Panel 45) is supposed to be at an uninhabited iron-age fort at modern Hastings, but is depicted with an obviously Saxon timber tower. Panel 48 shows the Norman knights supposedly leaving their orthodox camp at modern Hastings to attack Harold, but it depicts William standing beside an elaborate Romanesque church. Domesday lists no manors, let alone manors rich enough to have elaborate Romanesque churches, between modern Hastings and the orthodox battlefield. Panel 53 should show the English fighting parallel to a ridge crest but instead shows them defending both sides of a hill.

So, the Tapestry is thought to be mainly allegorical. If it is allegory, there is no point in trying to match the buildings and geography to real places and real events, which is why no one has tried. We are more optimistic. It seems unlikely to us that the embroiderers would take such extreme care with hundreds of horses, troops, ship and artisanal scenes, but not a handful of buildings and set-pieces. There are accepted excuses.

Most historians believe that the Tapestry was embroidered by Canterbury nuns whose lack of worldly experience meant that they had to be advised by a ‘master artist’. Perhaps, then, he was too junior to have been present at many of the Tapestry’s scenes, or forgot them, so they had to be invented or copied. Denny has an alternative theory that the Tapestry was designed and embroidered by English artisans who would not have been present at any of the events, so the artisanal scenes are accurate while the buildings, geography and set-pieces were invented or copied. His main evidence is that the Tapestry uses only the two most basic embroidery stitches, and it contains no gold thread, which makes it too rudimentary to have been made by Canterbury nuns.

These theories are plausible, but incomplete as excuses. The embroiderers did not skimp on detail with the buildings and set-pieces. It would have been easier to depict them accurately, if they had access to the someone that was there. The Tapestry was embroidered in Canterbury, one of the most populous towns in England. William immediately commissioned a fortress there. He gave Canterbury and much of Kent to Odo, the Tapestry’s patron. Odo and his entourage and many battle participants must have been present in Canterbury during the time that the Tapestry was being made. If the master artist forgot the scenes in England or was not present, he would surely have asked one of these battle participants for details.

Dives provides the solution, we think. Most of the invaders entered the Tapestry’s narrative at Dives. Nearly all the buildings and set-pieces that are known to have been copied from psalters precede Panel 34 at Dives. The exceptions are Odo blessing a meal at the first Norman camp on Panel 43 and William, Robert and Odo sitting under a gazebo at the first Norman camp on Panel 44. Odo was the Tapestry’s patron. We suspect that both these scenes were invented to appease his ego. If we are right, invasion participants provided designs for all the historical scenes after Panel 34, while earlier scenes were copied from psalters in Canterbury libraries.

Our simple explanation for the contradictions between the Tapestry and the orthodox invasion narrative is that the orthodox narrative is completely wrong. We explain our alternative narrative in our blog ‘The Battle of Hastings … at Sedlescombe’ - here. To summarise, we believe that the Normans landed near Penevesellum on the north bank of the Brede estuary, that they temporarily camped nearby, that they moved to a second camp at modern Winchelsea where they stayed for two weeks before fighting Harold at Sedlescombe. The Tapestry is consistent with our revised narrative.

Therefore, by and large, the Tapestry accurately depicts historical scenes in England, including the geography. Armed with this knowledge, it unveils some new secrets. We will share these insights below.

Bobbles

The key to understanding the Tapestry’s geography is its baseline. There are three types: plain, bobbly and water. Examples of each can be seen on Panel 22 at Bayeux and Panel 36 at Dives below.

Panel 22 (above) shows horses walking on a bobbly baseline, and a fortress on a hill on a plain baseline. Panel 36 (below) shows men wading in a water baseline, and armour carriers on bobbly baseline.

What do these different baselines mean? Water is self-explanatory. We interpret the bobbly baseline to represent undulating ground, like fields, heathland, scrubland and dunes. The plain baseline represents everything else, usually the ground under hills and buildings. We interpret this to mean that the plain baseline represents level or gently rising ground. Many buildings are adjacent to a section of plain baseline on one side or both, which might imply that they were accessed by road.

The image above shows the baseline types across a hugely compressed Tapestry: Red = bobbly baseline; Cyan = water; Green = plain baseline; White = anomalous plain baseline. There are seven anomalies.

Three anomalies - shown in white above - have a plain baseline with no hills or buildings above. The first is on Panel 2 depicting Harold arriving at Bosham. The second is on Panels 40 through 44 at the Norman landing site. The third is on Panel 58 where the English are fleeing.  We think these were deliberately designed by the artist to divulge some unusual geography. We interpret Panel 2 to mean that Harold travelled on Stane Street. We will discuss Panels 40 to 44 and Panel 58 below.

The other four anomalies are the reverse, depicting buildings on a bobbly baseline. They are Panel 32 and 33 under Haley's Comet, Panel 36 at Falaise, Panel 42 at the first Norman camp, and Panel 47 showing a building being set alight. We interpret them to mean that the buildings are far distant, beyond fields, heath or scrub.

A few words about mottes before we start. Many analysts have said that the hills under Tapestry fortresses are mottes. Perhaps, then, the plain baseline under buildings implies they were built on mottes. We think it unlikely. Non-military buildings were not built on mottes, yet they are also on plain baselines. One purpose of a motte is to raise a fortification above the surrounding land to make it more difficult to attack. Some mottes, at Launceston for instance, are 10 metres or more high, making them look like the Tapestry’s hills. But they are rare, especially before 1066. It is implausible that all the Tapestry’s fortresses are on high mottes. Indeed, Bernstein reckons that mottes were very rare before 1066, if they existed at all, so none of the Tapestry’s buildings are likely to be on mottes. Moreover, mottes are generally levelled before being used as building foundations, whereas the hills under the Tapestry’s fortresses have rounded tops. We think the Tapestry’s fortresses are on natural hilltops.

Panel 38

Panel 38 shows the Norman fleet crossing the Channel. Its caption says that William: “... crossed the sea and came to Pevenesæ”. Nearly all historians interpret it to mean that William “came to Pevensey and landed”, but then why would it not simply say that he: “... crossed the sea and landed at Pevenesæ”? Just one account, Brevis Relatio, is usually translated to be saying so, but it uses the Latin verb ‘appello’ which is more likely to mean ‘arrived’ than ‘landed’. Three other accounts say that the Normans arrived near modern Pevensey - ‘pefenes ea’, ‘Pefnesea’ and ‘Pevenesel’ are thought to be Pevensey cognates - without suggesting that they landed:

  • ASC-D says that: “Earl William came from Normandy to pefnes ea”.
  • Benoît says that the Normans: “Arrived at Pevenesel”.
  • John of Worcester says that William: “moored his fleet at a place named Pefnesea”.

Two other accounts agree that the Normans moored near the English coast:

  • Carmen says: “On the open sea you moor offshore”.
  • Poitiers says: “having reached shallow water off the English coast, William drops anchor to wait for the rest of the fleet to catch up”.

We interpret this to mean that the Norman fleet sailed somewhere within sight of modern Pevensey (Benoît says that they arrived beneath a ‘handsome fortress’), where they moored in shallow water before sailing somewhere else to land. In other words, they did not land at modern Pevensey, or anywhere that might have been named Pevenesæ at the time, and we list 20 or so other clues in our main Sedlescombe Battlefield paper that they landed in the Brede estuary.

Panel 40

Panel 40 shows the Norman cavalry soon after landing. It is captioned: “and here the knights have hurried to Hestinga to seize food”. Nearly all historians interpret this ‘Hestinga’ to mean modern Hastings, but the Norman knights would have been lucky to bag a few seagull eggs at the end of a long narrow sea cliff. They must have gone somewhere else. Some experts think that the Tapestry’s ‘Hestinga’ referred to ‘Hæstingaport’, but a port is an equally unpromising foraging location, unlikely to offer more than a few goats and hens. The only credible meaning for the Tapestry’s ‘Hestinga’ is the Hastings Peninsula, and our main Sedlescombe Battlefield paper lists a dozen or so clues that this was so.

There is an important consequence. If the knights hurried to Hestinga and Hestinga meant the Hastings Peninsula, they did not start on the Hastings Peninsula. Instead, they must have started somewhere that was connected to the Hastings Peninsula by a firm land route. The next panel, 41, shows that the knights returned the same day with livestock, so the landing site cannot have been far from major farms on the Hastings Peninsula.

In other words, the Norman fleet did not land on the Hastings Peninsula but somewhere nearby that was connected to it by a firm land route, probably a road. We explain in our main Sedlescombe Battlefield paper that this landing place was the north bank of the Brede estuary.

Panel 41

Panel 41 shows the first Norman campsite. It is one of the anomalies we mention above, an outdoor scene that is depicted on a plain baseline. We interpret this to mean that the landing site was billiard table flat. In the background are three huts. They are the only modest buildings on the entire Tapestry. We think that the master artist put them there deliberately to say something important about the landing site geography.

The leftmost hut has stone walls with a tile roof. The rightmost hut has weatherboard walls with a tile roof. The middle hut has weatherboard walls with a plank roof. This at a time when nearly all dwellings in England had wattle and daub walls with a thatched roof, especially as here, where plentiful reeds are nearby.

Panel 41 is a perfect match for our proposed landing on the north bank of the Brede which was a 5km long pancake-flat plain of dried out salt evaporation ponds. The huts in the background would be salt-houses where concentrated brine was crystalised.

The north bank of the Brede had the only salt-plain in the entire region that was long enough to land the Norman fleet. We list 20 or so other clues that the Normans landed on the north bank of the Brede in our main Sedlescombe Battlefield paper.

Panel 42

Panel 42 and 43 shows the Normans having a meal at their first camp. There is a gazebo behind the waiters on Panel 42. It is usually interpreted to be the kit fortress that the Normans built at their first camp. It is not. Wace explains that the kit fortress was made of timber planks fixed together with pegs, whereas the Panel 42 building is made of stone. The kit fortress had palisades, whereas the Panel 42 building has open sides. The kit fortress would be unadorned and roofless, whereas the Panel 42 building has windows and cupola roofs. The kit fortress would have been assembled on flat ground, whereas the Panel 42 building is on a bobbly base with foundations.

Note that the gazebo is on a bobbly baseline whereas the kitchen and bench either side are on a plain baseline. We interpret this to mean that it is in the far distance. We suspect it is a remote Saxon lookout tower, or less likely, a tower to dry fishing nets.

Panel 45

Panel 45 shows the second Norman camp. It is one of the most enigmatic scenes on the entire Tapestry. The caption reads: “ISTE JUSSIT UT FODERETUR CASTELLUM AT HESTENGA [CEASTRA]”. Weird. We will try to unravel some of its secrets.

AT’ is a valid Latin word, but not in this context. Most translators assume it is Old English, with the same meaning as now. ‘CEASTRA’ is not a valid Latin word either. Most translators assume it is a misspelling of the Latin word ‘castra’, meaning ‘camp’. It is embroidered inside the fortress palisades, so it is often assumed to be a label for the Norman camp. However, it could be Old English too: ‘ceastra’ means ‘fortification’, typically one of a Roman origin.

The Tapestry has relatively few Old English proper nouns, but it usually tries to transliterate them into Latin. ‘Edward’ is ‘EADVVAD[I/US]’, apart from in the first scene. ‘Leofwine’ is ‘LEVVINE’. ‘Pefenesea’ is ‘PEVENESÆ’ (Latin V was consonantal when positioned between vowels, being pronounced somewhere between modern English F and V; Latin Æ was pronounced similarly to Old English ‘ea’). The only exception is Harold’s brother GYRÐ, which uses the Old English letter ‘Ð’, known as ‘eth’ and pronounced something like ‘th’, but this was probably because there is no ‘th’ sound in Latin.

If the Tapestry is being consistent, HESTENGA and CEASTRA were transliterations of Old English proper nouns. ‘ceastre’ is not a proper noun on its own, so we think it is part of the main caption. If so, HESTENGA CEASTRA is the Latin transliteration of the compound proper noun ‘Hæstinga ceastre’, the Alfredian burh fortress on the Hastings Peninsula. Perhaps, then, the Old English preposition ‘AT’ was used to indicate that an Old English proper noun follows.

There are two explanations for why the CEASTRA has been embroidered in the fortress palisades: 1) They ran out of space – note the squashed A on the end of HESTINGA; or 2) It shows that the fortress on the top of the hill is a Saxon ‘ceastre’ and not the kit fortress that the Normans are about to build.

The kit fortress at the second Norman camp is described by Poitiers, Jumièges, ASC, CBA. It is usually assumed to be the palisaded fortification depicted on the top of the hill. Thus, it is usually assumed that the men are digging a moat around the base of the fortress hill. They are not. The ditches that fill with water to become moats are where soil was removed to make a motte. They are dug before construction starts and they are immediately adjacent to the fortress walls.

The most natural translation of the Panel 45 caption is: “He ordered that a fortress be dug at Hestinga [Ceastra]”, giving the impression that the men are digging a fortress. But fortresses are not dug. So, what is going on? Wikipedia translates: “He ordered that a motte should be dug at Hastings”. This is possible (so it is a shame that Wikipedia does not give its source or its reasoning). A kit fortress would need to be assembled on flat level ground, which would be a motte, even if Bernstein is right that the word and technique had not yet been invented. On the other hand, the caption does not include any word that might mean ‘motte’.

We suspect that the confusion is caused by the word CASTELLUM. Pocket dictionaries and Google always translate it as something like ‘stronghold, fortress, refuge’. The full Oxford Latin Dictionary has more: “a. fortified settlement or garrison; b. an outpost or strong point guarding a town; c. a refuge or stronghold.” Only one of these is dug, the strong point. If the caption is accurate, the men probably are digging a motte for the kit fortress, to create a strong point to guard Hæstingaceastre. They are still digging, so the kit fortress has not yet been started, yet there is a fortress on the top of the hill. It must have been there when the Normans arrived. The only known Saxon era fortress on the Hastings Peninsula is the Alfredian burh of Hæstingaceastre. As we say above, we think the caption confirms it is Hæstingaceastre.

Why then are the men digging so far down the slope? The answer, we think, is that the scene being depicted on Panel 45 is at modern Winchelsea, where the fortifications and/or kit fortress would guard the narrow neck which was the only easy access point.

The image above shows the probable location of the Hæstingaceastre burh fortress as a green square on the top of the hill, where St Thomas's now stands. The tower is shown as a yellow square to the right, where the tower of Greyfriars now stands. The earthwork being dug on Panel 45, probably a motte for the kit fortress, is shown as a red square guarding the narrow entrance to Winchelsea. The artist’s view is shown by a magenta arrow.

Panel 48

Panel 48 shows the Normans leaving their battle camp to attack Harold. It is captioned: “Here the knights have left Hestenga and have come to the battle against King Harold”. Historians interpret this ‘Hestenga’ to mean modern Hastings, but as we say above the Normans were never at modern Hastings.

If this ‘Hestenga’ is the same as ‘Hestinga’ in Panel 40 - and Latin ‘e’ and ‘i’ are interchangeable in Latin transliterations of place names, so it probably is - then it is referring to the Hastings Peninsula. In other words, the caption is saying that the Norman cavalry left the Hastings Peninsula to attack Harold, so the battle did not happen on the Hastings Peninsula. A section in our main Sedlescombe Battlefield paper gives three other clues that the battle took place near to the Hastings Peninsula but not on it.

The Panel 48 church is fascinating. It is clearly Romanesque. Saxon churches were not of this style and were very rarely made of stone. If it is based on a real building, it would be incredibly rare, which leads many to think it was copied from a psalter. But the master artist and embroiderers would not have lacked Saxon churches to copy. They just needed to look out of their window. There is a more likely explanation.

There probably was a Romanesque church in the vicinity of the battlefield, in the form of the monastery for the Fécamps Abbey cell in Rameslie manor. And the Panel 48 building probably is a monastery because the thing that looks like a huge door between William and the tower is probably a dormitory or refectory or guest house. Rameslie manor lined both banks of the Brede. This building was probably in modern Sedlescombe, north of the Brede and adjacent to the Rochester Roman road.

Panel 49

Panel 49 depicts William approaching the battlefield. According to the caption, he is asking whether Vital has seen Harold’s army. This hill between William and Harold is quite unlike the orthodox engagement, in which the Normans were at Telham on a downhill section of the Hastings Ridge crest when they first spot Harold’s army.

In our Sedlescombe Battlefield paper we explain why the English army was were deployed at Hurst Lane in modern Sedlescombe and that William arrived from modern Winchelsea. His first view of the English army would have been as he crested Cottage Lane ridge, exactly as depicted on Panel 49.

Panel 52

Panel 52 shows an early part of the battle. It is captioned: "Here fell Leofwine and Gyrth, brothers of King Harold". The base is bobbly, which indicates fields or heathland. Yet many of the contemporary accounts say that the English shield wall was deployed on a hill and that the Normans were attacking up a steep slope. As we say at the start, Tapestry hills are invariable depicted on a plain baseline.

Panel 52’s bobbly baseline might be a mistake, but we suspect that Leofwine and Gyrth had been drawn out of the shield wall. Wace describes exactly this happening early in the battle, long before the famous feigned retreat. Perhaps Leofwine and Gyrth were part of the chasing group or, more likely we think, they were trying to call the men back into position.

Panel 53/54

Panel 53/54 shows the English being attacked on a hill. The English are fighting back-to-back with the Normans attacking from both sides. We interpret this scene to be the Norman view looking at the front of the shield wall. If so, the battlefield hill was narrow, steep-sided, flat topped and rugged. If the English were fighting back-to-back, they were deployed in a loop. We list 12 clues from the contemporary accounts in the relevant section of our Sedlescombe Battlefield paper which confirms that the English were deployed as an enclosed loop at Hurst Lane.

The artist’s view is shown as a transparent arrow on the diagram above. It happens to be where there is an unusual 10m high knoll, upon which he was presumably standing. The magenta ‘wedge’ shows the English shield wall troop deployment.

The Tapestry depicts English troops with moustaches, so two English men at the bottom of the hill were English. Presumably, they too had been drawn out of the shield wall by some ruse or another.

Panel 58

Panel 58 is the last. It is captioned: “and the English have turned in flight”. The English have some sort of double-decker escape route. This is the scene at the Sedlescombe iron ore pits through which the English fled towards the Rochester Roman road. The bottom row of Englishmen were fleeing along the ditch base while the top row were fleeing along the far rim of the ditch.

The diagram above shows the English flight route (magenta dots), from the Hurst Lane battlefield (magenta loop) to the Rochester Roman road (black line). The artist's view was from the red arrow, looking across one of the huge iron ore pits with the English fleeing west across the other side. Of course, the English were fleeing to the artists left whereas the Tapestry depicts them fleeing right, but that it is because the Tapestry's comic strip timeline progresses to the right.

Bibliography

David Berstein; The Mystery of the Bayeux Tapestry; Weidenfeld and Nicolson; 1986

Francis Birrell; Guide to the Bayeux tapestry; V&A Museum; 1921

Andrew Bridgeford; 1066 : The Hidden History of the Bayeux Tapestry; Walker; 2005

Norman Denny; The Story of the Norman Conquest, 1066; Parkwest; 1966

Charles Gibbs-Smith; Bayeux Tapestry; Phaidon; 1973

David Musgrove & Michael Lewis; The Story of the Bayeux Tapestry; Thames and Hudson Ltd; 2021

Lucien Musset; The Bayeux Tapestry; Boydell Press; 2005

Trevor Rowley; An Archaeological Study of the Bayeux Tapestry; Pen and Sword Books; 2016

Brian Williams & Brenda Williams; Secrets of the Bayeux Tapestry; Pitkin; 2008

Arthur Wright; Decoding the Bayeux Tapestry; Frontline Books; 2019