A teacher once asked us: “If there are 128 players in the Wimbledon Men’s Singles draw, how many matches are played?” After laboriously working it out (answer 127), someone piped up, “It’s obvious, everyone plays exactly one match in which they lose, apart from the winner”. Things look different when you know the answer. Half of Momentous Britain like to play wargames. The other half had spent forty years working out where the Battle of Hastings was fought, when one of them piped up, “It’s obvious, blah, blah, …”. Thanks a bunch.
Most of our thoughts on the Battle of Hastings explain our interpretation of accounts written soon after the battle. We know this makes it read like a series of wild guesses that lead to fanciful conclusions. Our wargaming brethren pointed out that our proposed battlefield is likely to be correct based only on the geography of the region. They also noted that the sequence of events can be verified from six relatively uncontentious statements in the contemporary accounts. If you accept this, our more speculative interpretations can be viewed in the way they are intended: To colour the events rather than to define them.
The geography has changed a lot since medieval times. In those days, the Hastings Peninsula was bounded by the sea to the south, the River Ash Bourne to the west and the River Brede to the north. The Normans were somewhere south of the Brede. The English were coming from the north. In order for the armies to engage, one or other had to circumvent or cross the Brede. Sedlescombe bridge (S on Figure 2) was at the Brede tideway. It carried the Roman road from Rochester to Westfield on the Hastings Peninsula. Downstream, the estuary was 300m wide and sandy. It had a low-tide ford (F) below Brede village. Upstream, the fluvial Brede was 5m or so wide and boggy. It had a ford at Whatlington (W). Further upstream, the river was narrow enough to be spanned by a temporary bridge. Alternatively, there was probably a route around the river on ancient ridgeways via Netherfield (N).
Figure 1: 1066 coastline showing Roman roads (black) and trackways (red)
The essential tools for this exercise are a Roman road map (Figure 1), a topographical map (Figure 2), Domesday, a primary school understanding of medieval history, and the knowledge of one conceit.
For the history. Normandy had been left undefended. William had befriended his traditional enemies in what is now northern France, in order to buy time for his invasion. Peace could not last forever. The longer William was away the greater the chance that someone would try to annex Normandy. Also, the larger the English army and the greater the attrition in the Norman army. William needed to kill Harold quickly. Defeating the English army without killing Harold would probably have led to a Fabian defence which would almost certainly have scuppered the conquest.
Domesday provides a baseline: it reports that large swathes of the Hastings Peninsula were either ‘wasted’ or lost most of their value at the time of the battle; the Normans must have plundered and/or raided there. Harold was in Westminster. In between was the 120-mile-wide primeval Andredsweald forest. Its boundary, shown in green dots on Figure 1, is implied by Domesday manors with no meadowland or ploughland. Figure 1 and Figure 2 show Roman roads in black, known ridgeways and trackways in red or white, and assumed trackways in white dots. These provided the only feasible way for armies to get around, because most of the region was low-lying bog or dense woodland.
There were two possible routes for Harold to get to the battle theatre. The easterly route takes Watling Street to Rochester, then Margary 13 to the Hastings Peninsula. The westerly route takes Margary 14 from Peckham to Uckfield, then cuts east across the Andredsweald on the LIN-129 trackway from Uckfield to Rye, branching at Heathfield to join the LIN-130 trackway towards the Hastings Peninsula.
There is no wargaming advantage in taking the westerly route, the ridgeway had no settlements for supplies and it would be difficult for carts. If it was typical of most ancient forest ridgeways, it would have become overgrown when the Romans left because there was no longer a central power to organise infrastructure maintenance. Moreover, the western route provides no opportunity to muster with troops coming from Kent or with huscarls arriving by boat up the Rother. Wargamers would take the quick and simple easterly route. It does mean crossing the Rother by ferry, but the King could requisition every boat and oarsman in the region to help.
A wargamer playing the English would have an obvious starting place: To create a forward operations base close to the Hastings Peninsula, from where they could dispatch scouts, monitor enemy troop movements, create a picture of the enemy strength and position, and devise a plan of siege or attack.
The obvious place for a forward operations base is the last hill on the Rochester road before the Hastings Peninsula. We will refer to it as Great Sanders ridge (G). It was good camp terrain: woody, close to running water, big enough for the English army, and protected on two sides. It overlooked all three of the Brede crossing points and the Hastings Ridge along which the Normans would have to march if they wanted to circumvent the Brede. It seems to be a safe distance from the nearest possible Norman position.
Figure 2: East Sussex geography and topology
But then there is the conceit. When Harold chose the English camp location, he thought the Norman army was weak and footbound. He had no reason to fear a Norman attack on Great Sanders ridge. It was 5km from safety at the Rother. The Normans had to be least 2km further away because they were south of the Brede. If Harold decided to withdraw, he would have thought there was no way the Normans could catch 2km uphill over 5km on foot. In reality, they had several thousand experienced knights with trained war horses that could canter to Cripps Corner (C) before the English could run there, thereby cutting off a retreat and catching the English in the open. This conceit is not an invention. It is explained in Roman de Rou. We will return to it in the main text.
If the wargamers are right, the real battle narrative is as simple as it could possibly be. The English camped on Great Sanders ridge, at what Harold thought to be a safe distance from the Norman army. To his horror, he discovered the next morning that William had a huge cavalry that put them in range of attack and prevented an organised withdrawal. The English army was effectively trapped. William spent that day scouting the English camp and devising a plan of attack. He attacked the English camp at dawn the following day. Unlike other battle theories, this does not rely on idiotic battle tactics, suicidal surprise attacks or camping at naff defensive positions. It just needs Harold to make one minor intelligence error.
For any sceptics thinking that a similar argument could be made for any hill in the vicinity, we disagree. The nearest potential camps to the north of Great Sanders were on the Udimore and Isthmus ridges. If the English were on the ridge, there would have been no way for the Normans to cut off their retreat by occupying it. The nearest potential camps to the south of Great Sanders were on the Hastings Peninsula. From a wargaming perspective, it would be unthinkable for the English to cross the Brede, either to make camp or to attack the Norman camp, before the Norman position had been scouted. Quite apart from the ambush risk at the crossing points, it probably had no food; Domesday says that there was very little farmland between the Brede and the Hastings Ridge, and the Normans had been foraging there for two weeks.
Information about the invasion comes from 12 contemporary accounts, plus the Bayeux Tapestry and the Domesday Book. It is standard practice is to refer to them as the ‘primary sources’. For the sake of brevity, we will sometimes use these abbreviations:
- ASC = Anglo-Saxon Chronicle (reasonably contemporary with events)
- ASC-C, ASC-D = The versions of ASC that cover the invasion
- Tapestry = Bayeux Tapestry; finished c1077
- Benoît = Chronique des Ducs de Normandie; Benoît de St-Maure; c1170
- Carmen = Carmen de Hastingae Proelio; c1067
- CBA = Chronicle of Battle Abbey; c1170
- Chronicon = Chronicon ex Chronicis; John of Worcester; c1125
- CKE = Gesta Regum Anglorum; William of Malmesbury; c1135
- Domesday = Domesday Book; 1086
- Huntingdon = Historia Anglorum; 1129
- Orderic = Historia Ecclesiastica; Orderic Vitalis; c1125
- Wace = Roman de Rou; Master Wace; c1160
- Warenne Chronicle = Chronicon monasterii de Hida iuxta Winton; c1200
- WJ = Gesta Normannorum Ducum; William of Jumieges; c1070
- WP = Gesta Guillelmi; William of Poitiers; c1072
Six primary source statements summarise our argument. We believe that the first four verify the simple battle scenario explained above, and the last two pinpoint the battle location.
- The Anglo-Saxon Chronicle says that: “[Harold] came against him at the estuary of Appledore”. Appledore was on the Rother estuary, so the English army must have come down the Roman road from Rochester.
- Bayeux Tapestry Panel 48 says: “Here the knights have left Hestenga and have come to the battle against King Harold". We believe the Tapestry’s Hestenga referred to the Hastings Peninsula. Thus, the Tapestry is saying that the Normans left the Hastings Peninsula to attack the English. This means the River Brede was between the rival camps.
- Carmen says that when William’s monk messenger returns to the Norman camp, William asks him: “Where is the King?” The messenger replies: “Not far off. You can see his Standards”. The only pairs of camp locations on opposing sides of the Brede where this would be possible are: Woodmans Green (1E) and Canadia (1N); the ridge below Great Sanders (2E) and Cottage Lane (2N); Cackle Street (3E) and Starlings Hill (3N).
- Wace says that Harold and Gyrth ride off alone to scout the Norman camp at dawn on the day before battle. They: “rode on, viewing and examining the ground, till from a hill where they stood they could see those of Normans, who were near”. This would not be possible from Woodman’s Green or Cackle Street. It is possible if they rode south from Great Sanders ridge, probably to Balcombe Green (B).
X marks the spot. Two more statements reveal the exact battlefield location:
- Carmen says that when the Normans started their attack: “Suddenly the forest poured forth troops of men, and from the hiding places of the woods a host dashed forward ... there was a hill near to the forest ... they seized possession of this place for the battle". We interpret this to mean that the English left their woody camp and occupied a nearby hill for the battle. The only other hills in the vicinity were four spurs pointing south from the Great Sanders ridge. If the English camp was at Great Sanders ridge, the English must have occupied one of these spurs for the battle.
- John of Worcester says that the English were: “Drawn up in a narrow place” for the battle. The only hill or ridge near to the Great Sanders ridge that could be described as a narrow place is the mid-east spur (X) along which Hurst Lane now runs. We think that is where the battle happened.
We are the first to acknowledge that the contemporary accounts are unclear, ambiguous and less than fully trustworthy. However, we think that these six statements are among the most reliable and straightforward (although even these have conflicting translations and interpretations). If we are right about these six, we are confident that most of our other battle related interpretations are right too. They always seem to be the least convoluted and the easiest for a medieval reader to understand. In our opinion, they pinpoint the battlefield at Hurst Lane near Sedlescombe, as depicted in the introduction blog.
Figure 3: 1066 Hastings Peninsula coastline with landing site candidates
A similar wargame analysis can be applied to the landing. In retrospect, we now think it does not really matter where the Normans landed, but we spend a lot of time talking about it, so it is interesting to use wargaming ideas to narrow down the possibilities.
First, did the Normans land on the seacoast or inland? A wargamer would prefer inland, because it splits the defence. The idea would be to anchor line-astern in the middle of the estuary or inlet, in order to give no clue about the preferred landing side. If the defenders are on one bank, land on the other. If the defenders are on both banks, land on the side with the least defenders or with the worse defences. In the very worst case, an inland landing has to face half the defenders compared to a sea shore landing. This process can be repeated, shuffling to-and-fro across the river as the defenders work their way around to the other side, gradually wearing them out and whittling them down.
Which estuary or inlet? Figure 4 is our assessment of the Hastings Peninsula coastline in 1066. The wargaming idea here is to gain as much time as possible to fortify a bridgehead by landing near the mouth of a long waterway. Any defenders that happen to be on the other bank have to go as far as possible inland to get to the nearest crossing point and back on the other side.
There are three plausible estuaries and inlets around the Hastings Peninsula where this would apply: the Brede estuary, Combe Haven and the Ash Bourne estuary. Hooe Haven is a fourth possible landing site. It was too short for the mid-stream anchoring ploy, but it might have had a secret advantage we will discuss in the main text. The others are much of a muchness. The Brede had the longest way around (17 miles), but the best trackways. Combe Haven had the shortest way around (11 miles), but the worst trackways. In all three cases we reckon it would have taken roughly 6 hours to get from the mouth on one bank to the bank opposite. Six hours should be enough time to fortify a bridgehead. The defenders would be exhausted by the time they arrived. And they could not have arrived much before nightfall, which would have given another 12 hours to fortify the bridgehead.
There are two other important factors to consider in establishing a bridgehead: the defensive qualities of the terrain and the distance to the nearest defensive garrisons. Defensively the three candidates are similar. The nearest garrisons to the Hastings Peninsula were Pevensey to the west and Lympne to the east. As it happens, probably because of Tostig’s attack in Northumbria, they were unmanned, but William was not to know this when he was planning the invasion.
William’s preference, based mainly on it being furthest from any garrison (i.e. it is midway between Pevensey and Lympne), should be the Brede or Combe Havrn. The other two are skewed towards Pevensey. Assuming in each case that the Normans landed on the furthest bank from Pevensey, if the garrison had been manned, we estimate that it would have taken them six hours to get to Hooe, eight hours to get to Wilting or ten hours to get to Cock Marling at the most distant mouth of the Ash Bourne, Combe Haven and the Brede respectively. There was only ten hours of daylight at that time of year.
From a wargaming perspective, knowing nothing about contemporary accounts, William’s preference should have been to land in the Brede, with the north bank being marginally preferable to the south bank.