The Rochester Roman road and the Battle of Hastings

5. The Norman ‘sea camp’ was at Hæstingaceastre

Three Norman camps are described in the contemporary accounts:

1. A bridgehead camp near the Norman landing site where they stayed for a few days
2. The main Norman camp where William and his barons spent nearly a month either side of the Battle of Hastings
3. A battle camp where the Normans assembled and dressed before the battle

Our interest here is in the second Norman camp, the one that Carmen refers to as their ‘sea camp’.

William of Jumièges, William of Poitiers and Orderic say that the second camp and its wooden kit-fortress were at ‘Hastingas’. The Chronicle of Battle Abbey says that William took his men to a “port named Hastinges” where “having secured an appropriate place ... he built a fortress of wood”. Bayeux Tapestry panel 45 depicts the second Norman camp with the caption: “ISTE JUSSIT UT FODERETUR CASTELLUM AT HESTENGA CEASTRA”, meaning: “He [William] ordered that a castle be dug at Hestenga Ceastra”. The D recension of the Anglo-Saxon Chronicle says that: “soon after his [William’s] landing was effected, they built a fortress at Hæstingaport”. John of Worcester says the second fortress was at ‘Heastingam’, then that after the battle “William, however, returned to Heastingam”. Carmen says that, after the battle, William “returned to his castra marina” (‘sea camp’). It goes on to say that he then spent fourteen days in his “camp at Hastinges Portus”.

‘Heastinga’ is an alternative phonetic spelling of Old English ‘Hæstinga’. ‘Hastingas’ is a Latin phonetic transliteration of Old English ‘Hæstinga’ and of Norman Romanz ‘Hastinges’, so the root names are etymological cognates. ‘Hastinges Portus’ is the Norman Romanz translation of Hæstingaport. ‘Hestenga Ceastra’ is an alternative spelling of Hæstingaceastre. All these references are to the place where William built his second fortress and where he made his second camp, so the compound names are geographic cognates of the root names.

There are two more reasons to think that Hæstinga, Hæstingaceastre and Hæstingaport were linked. Firstly, John of Worcester refers to them all as Heastinga, as if they are cognates. Secondly, Hæstingaceastre was one of 36 ‘Grately Code’ places in England that were licensed by Æthelstan as mints. Coins from that mint are stamped with an abbreviation of ‘Hæstingaceastre’ or ‘Hæstingaport’ or ‘Hæstinga’ or ‘Winchelsea’, as if they are cognates.

This is compelling evidence that the second Norman camp, the ‘sea camp’, was at Hæstingaceastre.

6. Therefore, the Normans made their ‘sea camp’ at modern Winchelsea after landing in the Brede estuary

If the Norman sea camp was at Hæstingaceastre (Statement 5) and Hæstingaceastre was at modern Winchelsea (Statement 4), then the Norman ‘sea camp’ was at modern Winchelsea.

According to Wace, the Normans headed for a port and landed in a nearby estuary. The implication, although Wace does no say so explicitly, is that the port was at the mouth of the estuary in which they landed. According to the Anglo-Saxon Chronicle, the Normans made their second camp at Hæstingaport. If the Norman ‘sea camp’ was at modern Winchelsea, the port was Old Winchelsea, a major port on a shingle island a mile or so south of modern Winchelsea, at the mouth of the Brede estuary. It implies that the Normans landed in the Brede estuary.

Wace and Carmen state that the Norman fleet landed in a sheltered estuary. Wace elaborates that they landed together on an estuary strand adjacent to a firm level plain. The only place in the region that matches this description and that was long enough to land more than half the Norman fleet is the north bank of the Brede estuary (see ‘The Landing’ section of our book).

7. Medieval armies moved on metalled roads where possible

Charles Barrett pointed out over a hundred years ago that: “Nearly all the battles on English soil have been fought either across one or other of the old Roman roads, or in close proximity thereto.” Mike Bishop elaborates in his chapter about the post-Roman use of Roman roads: “The Romans may have gone, but armies still needed roads. In 1993, N.J. Higham published a map illustrating the association of Anglo-Saxon battlefield sites with Roman roads. Although it went almost uncommented in his text, the significance of his observation is obvious: early medieval armies used Roman roads to move around.” They are saying that medieval armies moved on Roman roads unless they had no alternative.

Some disagree. Medieval settlements were interconnected by ‘country roads’, unmetalled earthen tracks. They formed a nationwide network. Some scholars have suggested that medieval armies might have used these byways instead of Roman roads. We think this is implausible if they had the choice. One major factor is the army wagon train. Medieval country roads were primarily used by walkers, individual riders, and packhorses. Hooves and wheel pressure quickly turned earthen tracks into rutted, impassable sludge, entirely unsuitable for heavily loaded wheeled traffic except during the driest part of summer. Daniel Defoe described Wealden country roads as so poor that he saw six oxen struggling to pull a small carriage carrying an elderly lady to church. An army wagon train, much heavier and longer, would have been unmanageable on such byways.

Military logistics expert Robert Evans, Head of the Army Historic Branch, estimates that Harold’s army would have needed 100 ox-drawn wagons just for the tents of his barons and huscarls, with additional wagons required for armour, weapons, and provisions. Medieval armies also travelled with blacksmiths, carpenters, weaponsmiths, cartwrights, wheelers, cooks, and their tools. Harold’s wagon train therefore probably consisted of 150 or more heavily loaded wagons. Such a convoy could not have traversed unmetalled byways from London to the Hastings Peninsula quickly enough to arrive in time for the Battle of Hastings.

Barrett identifies two additional reasons why medieval armies used Roman roads when possible. First, the role of commissariat had not been invented. Medieval armies foraged for provisions along their route: Roman roads typically passed near large, well-established farms, while byways usually traversed sparsely populated areas with small farms. Indeed, Barrett lists several examples of medieval armies that starved after deviating from Roman roads. Second, medieval barons expected comfortable accommodation while on campaign. Settlements with adequate facilities were commonplace along Roman roads, rare along byways.

Barrett overlooks a third critical reason: access to ale. Unprocessed water was unsafe to drink, as evidenced by half of William’s army contracting dysentery during the siege of Dover (as reported by Poitiers). Ale was the primary beverage, even for children, and the quantities required by an army made transporting it impractical. Only settlements along Roman roads could supply ale in sufficient quantity.

Woodland presented additional challenges for medieval armies. Away from rivers and the coast, much of the country was blanketed with mature deciduous woodlands. Among these forests, the Andredsweald was the largest in southern England, some 120 miles east-west by 60 miles. It was situated between London and the Hastings Peninsula, and therefore had to be traversed by the English army. Nearly every historian writing about the Battle of Hastings assumes that the English army used forest tracks to cross the Andredsweald. Edward Foord was probably the first, depicting Harold’s route to be roughly following that of the modern A21 (Figure 4).

Our blog examining the development of the orthodox Battle of Hastings narrative (here), reproduces more than thirty engagement diagrams from Battle of Hastings analyses. Without explicitly saying so, they nearly all seem to agree with Foord that the English crossed the Andredsweald on the route of the A21, indicated by labelling the road through Battle ‘To London’ or ‘To Tonbridge’. It is, of course, anachronistic because the route of the A21 was only cleared for the construction of the Hastings to Flimwell Turnpike in the 1750s.

Harold’s only choice was between the Rochester Roman road on the one hand and unmetalled byways, including unmetalled forest tracks through the Andredsweald, on the other. It is implausible that the English wagon train could have traversed the Andredsweald on forest tracks and still arrive in time for the battle. Not only would they have had all the byway problems listed above, but they would keep getting jammed. Horse collars and pivoting front axles had not yet been invented. Without the horse collar, heavy wagons were pulled by yoked oxen, and without pivoting front axles, wagons were unable to make sharp turns. Even though the average gap between tree trunks in mature deciduous woodland might have been around eight meters, wagons would frequently become stuck in ruts or jammed against trees. In such cases, the only solution would be to lift the wagon around the obstacle, perhaps having first partially unloaded it, an exhausting and slow process.

Provisions would compound the logistical difficulties because food and ale for the men and fodder for the oxen would have required an additional 20 wagons a day. The impracticality of feeding thousands of men and hundreds of oxen on a painful slog through the Andredsweald on forest tracks suggests that it did not happen.

The difficulty of navigating medieval byways, especially those in mature deciduous woodland, can be illustrated by the royal itineraries of Norman and Plantagenet kings. Even with relatively small trains of just a few dozen wagons, they moved almost exclusively on Roman roads and never through extensive woodland. Coastal routes were the only notable exceptions, and these were likely traversed by sea rather than by overland byways.

So, as Barrett, Higham, and Bishop say, medieval armies rarely used unpaved byways and only when Roman roads were unavailable. This was not an issue for the English army marching to the Hastings theatre of war because they could use Margary 13, and it seems certain to us that they did.

8. Therefore, the Battle of Hastings was not fought at the orthodox Battle Abbey location

If the English army arrived at the theatre of war on the Rochester Roman road (Statement 7) and it terminated at modern Winchelsea (Statement 1), there are no credible circumstances in which they could have ended up fighting the Battle of Hastings at the orthodox Battle Abbey battlefield location (A on Figure 5). To explain why, we must plug these statements into the orthodox engagement scenario, which is to say that the Normans were in their sea camp at modern Hastings (H), while Harold was either trying a surprise attack on the Norman camp or was trying to blockade the Normans on the Hastings Peninsula. The orthodox English camp is at Caldbec Hill (CH), although some think they camped at Battle (A).

1. The orthodox engagement scenario is inconsistent with a surprise attack

The only evidence for Harold’s decision to lead the English army onto the Hastings Ridge comes from Poitiers and Jumièges, both of whom claim he was attempting a surprise attack. However, it seems implausible that Harold might believe he could catch William by surprise when Wace and Poitiers report that they had been exchanging messages during Harold’s journey to the theatre of war. Moreover, it is inconceivable that he thought William’s scouts would have missed his army streaming over Sedlescombe bridge and laboriously ascending the slope to the Hastings Ridge.

A surprise attack is also inconsistent with the orthodox battlefield location. If Harold intended to attack the orthodox Norman camp, he would not have headed west through Petley Wood after crossing the Brede because the woodland would impede his wagon train and the route heads 3km in the opposite direction from the Norman position. Any alternative route onto the Ridge would have brought Harold to the Hastings side of Battle. For a battle to have been fought at the orthodox battlefield, Harold would need to have led the English army away from the Norman camp, the exact opposite of what a surprise attack entails. It has been suggested that Harold intended a surprise attack then realised his folly and decided to fall back to defend the ridge at Battle. It seems implausible for someone as thoughtful and meticulous as Harold, and if he was in that position, he would have fallen back to Margary 13 to defend the bridge at Sedlescombe.

2. The orthodox engagement scenario is inconsistent with a blockade

The improbability of an English surprise attack leads many modern scholars to believe that Harold’s strategy was to blockade the Normans on the Hastings Peninsula, but this is inconsistent with the geography (see Figure 5): Neither of the orthodox English camps would block Norman egress from the Hastings Peninsula because they are not within two miles of Margary 13.

Indeed, implausibly if Harold was trying a blockade, the orthodox engagement scenario assumes that the English abandoned Sedlescombe bridge, the one strategic location where a blockade would have worked, to move to a barren, disadvantageous position on the Hastings Ridge. The Normans, unencumbered by a wagon train, would have marched down to Margary 13 from their orthodox camp at modern Hastings to blockade Sedlescombe bridge themselves, trapping the English on the Hastings Peninsula and cutting off their supplies and reinforcements. In effect, the orthodox engagement scenario means that the English voluntarily placed themselves in a self-imposed siege for no reason.

We are convinced that Harold did intend to blockade the Normans on the Hastings Peninsula by blockading Margary 13 at Sedlescombe bridge – as we explain the ‘Harold’s tactics’ section of our book linked at footnote 1 - rather than by blockading the Hastings Ridge at Battle.  

3. The orthodox engagement scenario is militarily implausible

The only credible way for Harold’s wagon train to get onto the Hastings Peninsula was to cross the Brede at Sedlescombe. The bridge was probably only one cart wide, but the fluvial Brede was only a few metres wide. Perhaps Harold added a jury bridge to accelerate the crossing. Even so, it would still have taken several hours for the English army to cross the Brede, all the while disorganised and unarmoured on disadvantageous terrain with nowhere safe to retreat. William would need to have been implausibly naïve or inept to have missed the opportunity to attack Harold while he was so vulnerable. Harold would need to have been implausibly ignorant to think he might. Even if William allowed the English to cross the Brede, Harold’s problems were not over.

The northern slopes of the Hastings Ridge - from Petley Wood in the west to Coghurst Wood in the east - were covered in dense woodland. Moving ox-drawn fixed axle wagons through dense woodland was slow and difficult, as we explain in Statement 7. It would have been doubly difficult for Harold to get his wagon train onto the Hastings Ridge because of the steep slope. In principle, the steepness could be mitigated by an oblique climb, but on the north slope of the Hastings Ridge, this would mean crossing more woodland and more streams, so it would not help much.

The woodland was cut by a dozen or so streams draining runoff from the Hastings Ridge and there were Anglo-Saxon settlements at Baldslow on the Ridge and, probably, at Telham. There might have had partially cleared forest tracks from these settlements down to Margary 13. But all these routes would have been soft and rut prone, slow going for an army wagon train climbing a steep hill.

Whichever way Harold might have wanted to climb onto the Hastings Ridge would have taken hours, if not days, all the while leaving the English disorganised, unarmoured and on a downslope. Again, William would need to have been implausibly naïve or inept to have missed the opportunity to attack Harold while he was so vulnerable. Harold would need to have been implausibly ignorant to think he might.  

4. The orthodox engagement scenario is inconsistent with William’s tactics

The geography is inconsistent with the orthodox Norman attack. If the Normans were at modern Hastings and the English were at the orthodox Battle Abbey battlefield, William would not have attacked on the traditional route by marching northwest along the Ridge. The Normans were not encumbered by a wagon train, so they could negotiate woodland and steep slopes with relative ease. Their obvious and best plan of attack would have been to descend to the Rochester Roman road, blockade Sedlescombe bridge to trap the enemy on the Hastings Peninsula, march through Petley Wood to get onto the Hastings Ridge north of Caldbec Hill, then attack modern Battle downhill from the northwest (route shown by cyan dots on Figure 5). It is militarily implausible that William would attack up the disadvantageous slope south of Battle Abbey as tradition dictates, when he could easily have attacked on level ground from the north and east.

The orthodox Battle Abbey battlefield is even less consistent with our proposed engagement scenario. If the Normans were camped at modern Winchelsea (Statement 5) and Harold was arriving on Margary 13 (Inference 6 and Statement 7), there is no credible reason that Harold or William would have climbed onto the Hastings Ridge, and no credible reason they would have gone within two miles of the orthodox Battle Abbey battlefield.

In our opinion, these inconsistencies make it implausible that the Battle of Hastings was fought at the orthodox Battle Abbey location, and we list another 25 non-Roman road reasons to think so in the ‘Traditional Battlefield’ section of our book.

9. Therefore, the Battle of Hastings was fought near the Rochester Roman road between Cripps Corner and Winchelsea