Introduction

The Battle of Assandun was fought on 18th October 1016 between Danish invaders under King Cnut and a proto-English army under King Edmund Ironside. After a string of inconclusive earlier battles, this one was a resounding victory for Cnut. Edmund fled west. Cnut chased him down and defeated him in a final battle, probably in the Forest of Dean. Edmund was forced into a division of his realm with Cnut taking most of modern England north of the Thames. Edmund died a few weeks later, probably assassinated, ceding the rest of his realm to Cnut.

Historians think that the Battle of Assandun was fought at Ashingdon in southeast Essex or at Ashdon in northwest Essex. This paper explains why we think it is at least as likely to have been fought at Essendon in modern Hertfordshire.

Battlefield clues

Original battlefield location clues, infrastructure aside, are in six contemporary accounts: the Anglo-Saxon Chronicle, John of Worcester (Chronicon ex Chronicis), William of Malmesbury (Gesta Regum Anglorum), Henry of Huntingdon (Historia Anglorum), the Encomium and Knytlinga’s Saga. The last of these only has a brief poem about the battle, but it contains one unique and interesting clue (11 below).

1. The Anglo-Saxon Chronicle says that the followed the last of three
2. All the English accounts bar Malmesbury say that the Danes camped on the Isle of Sheppey before their third raid into Mercia. The Encomium says that the Danes wintered at a camp on the Isle of Sheppey then passes straight onto the Battle of Assandun with no mention of raids into Mercia or anywhere else.
3. All the English accounts bar Malmesbury say that the Danes advanced through ‘Essex’ before their third raid into Mercia.
4. The Anglo-Saxon Chronicle says that the Danes were returning to their ships after plundering Mercia when they were pursued by the English and ‘overtaken’, thereby leading to the battle. The Encomium says that the Danes left their ships to face the English enemy when they heard news of Edmund’s approach.
5. The Anglo-Saxon Chronicle says that the initial encounter was at ‘Assandun’, John of Worcester at ‘Assandun’, William of Malmesbury at ‘Assandunam’, Henry of Huntingdon at ‘Esesdune’, the Encomium at ‘Æsceneduno’, and Knytlinga Saga at ‘Assatun’.
6. The Anglo-Saxon Chronicle and John of Worcester say that Assandun was a hill, and its ‘dun’ suffix often means ‘hill’.
7. John of Worcester says that Cnut “very slowly brought his men down to a level ground” before being attacked by the English.
8. The Encomium and all the English accounts bar Malmesbury say that the Edmund was betrayed by Ealdorman Edric Streona.
9. The Encomium says that the battle started at the ninth hour of the day and lasted until nightfall. If they left their ships earlier that day, as it implies, the battlefield was probably no more than a three-hour march from the Danish fleet.
10. The Anglo-Saxon Chronicle says that Cnut commissioned a stone minster at ‘Assandune’ as a memorial to those killed in the battle, and that it was consecrated in his presence in 1020.
11. Knytlinga’s Saga says that the battlefield was north of ‘Dana skoga’, ‘Danes Wood’.
12. Alfred’s Battle of Ashdown in 871 was fought somewhere that sounds like Assandun.

Not much to go on. Only Clue 5, the battlefield’s name, is specific enough to identify candidates. The Anglo-Saxon Chronicle says that Assandun was in ‘East Seaxan’, Henry of Huntingdon says it was in ‘Estsaxe’. This province is always interpreted to mean the modern county of Essex, so both the orthodox battlefield candidates – Ashdon and Ashingdon - were hills in Essex whose names might derive from Assandun. Perhaps the other clues can be used to discount one or both of the candidates, or at least to rank their merits.

Development of the orthodox narrative

Holinshed proposed Ashdon (then ‘Ashdone’) as the battlefield in his 1577 ‘Chronicles of England’. There can be no ambiguity because he says it is three miles from Saffron Walden. Ten years later, William Camden proposed Ashingdon in his ‘Britannia: Chorographical’. Or, at least, everyone has assumed he was proposing Ashingdon because he says it was near the “pretty proper towne” of Rayleigh. But he says that the battlefield was named ‘Ashdowne’ not Ashingdon. He provides no other details, so there is a possibility that he agreed with Holinshed that the battle was fought at Ashdon but was confused about its location.

Renowned Norman Conquest historian Edward Freeman took up Ashingdon’s case in the 1860s: “As for the battle of Assandun, I have no doubt that the modern Ashington is the true site.” His main evidence is an “exact” match between Ashingdon’s geography and the battlefield described by John of Worcester. But that description is so vague that it also matches Ashdon and dozens of other hills in Essex. Freeman mentions that ‘Canewdon’, a hill near Ashingdon, might mean ‘Cnut’s Hill’, confessing that the evidence is tenuous. He categorically states that the Danes moored in the Crouch estuary, without any sort of justification or supporting evidence, then argues circularly that one reason to think the battlefield was at Ashingdon is that it was adjacent to the Crouch. Perhaps his only valid evidence is that the etymological transition from ‘Assandun’ to ‘Ashingdon’ would be analogous to other places in England including ‘Abbandun’, ‘Huntandun’ and ‘Ethandun’ which became Abingdon, Huntingdon and Edington. 

A H Burne favoured Ashingdon too, mainly because its “spelling is practically the same as the Anglo-Saxon Chronicle”. Only it isn’t. He says that the Anglo-Saxon Chronicle spelling is ‘Assingdon’ but it is actually ‘Assandun’ (or a declension of it) in all Anglo-Saxon Chronicle recensions. He notes that etymology expert Percy Reaney, writing in ‘The Place-Names of Essex’, agrees that ‘Assandun’ could easily transition to ‘Ashingdon’ but that “it cannot lie behind the forms Ashdon”. Domesday’s Ashdon spelling of Ascenduna and the Encomium’s of Æscenduno are close enough to suggest that Reaney might be wrong. Burne reports possible physical evidence of a battle in Ashingdon: “There have been found in the churchyard at Ashingdon parts of a shield, a spear and also a silver penny with Canute’s head”. The evidence has been lost, so it is impossible to know whether it was relevant, or even real.

According to the English accounts, Cnut was attacked as he returned from his third raid into Mercia. Burne tried to work out the events. He notes that Cnut moored in the Orwell before his second raid into Mercia. It is the northernmost of the major North Sea Essex estuaries, so Burne reasons that Cnut probably plundered north Mercia. If so, he argues, Cnut probably plundered south Mercia on his third raid into Mercia. Burne, like Freeman, arbitrarily decides that the Crouch estuary was the ideal place to moor and claims that both banks had similar strategic merits. This being so, he says that: 1) The raiding party was dropped off at Burnham-on-Crouch on the north bank and made their way west to Mercia; 2) The fleet guard took the ships across the estuary so to camp at Canewdon - he points out that the Ordnance Survey labelled it as ‘Supposed site of Canute’s camp’, which was true then, though the label has subsequently been removed; 3) The returning raiders, with the English army hot on their heels, joined the fleet guard at Canewdon; 4) The English camped at Ashingdon; 5) The battle happened between the camps. As far as we can see, it is baseless speculation with no supporting evidence, fundamentally flawed insofar as the returning raiders would not have been able to get to Ashingdon or Canewdon by land because they were surrounded by marshland. In the unlikely event that the Danes chose to drag their plunder cross country, they would surely have arranged to be collected by ship from modern Battlesbridge which was on a marsh free route.

H  B Swete weighed in behind Ashdon in the 1880s. He noted that Ashdon is named Ascenduna in Domesday, almost identical to the Encomium’s spelling of Æsceneduno. He also spotted that Ralf Baignard was Ascenduna’s tenant-in-chief in Domesday and that records from St Pancras Priory in Lewes (Cotton MS Vespasian F. XV) show that this same Baignard family made a series of gifts to the Priory, including the benefice of the church at ‘Essendon’. The manuscript says that this ‘Essendon’ was also known as ‘Assendun’ and ‘Asshedon’, so Swete concludes that ‘Asshedon’, and hence ‘Ashdon’, were cognates of ‘Assandun’. While he was vicar of Ashdon’s church, Swete monitored work which exposed Saxon foundations. He speculates these might have been from Cnut’s minster. He also commissioned the excavation of a nearby mound to find a huge pile of large animal bones and oyster shells, far too many for local consumption. He speculates that they may have been discarded by Cnut’s army.

Miller Christy claimed to have incontrovertible proof that the battle was fought at Ashdon in his 1926 paper for the Journal of the British Archaeological Association. It turned out to be a refined version of Swete’s, with the addition that the Saxon church at Hadstock, two miles from Ashdon, is of the right date and construction to have been Cnut’s minster, and it was dedicated to St Botolph, with whom Cnut had an affinity. Subsequent dendrochronology dating has shown that the current Hadstock church is too late to be Cnut’s minster. It is on the site of an earlier Saxon church but the mortar dates that church to be 200 years too early to be associated with the battle. Burne reckons it is too far from Ashdon to be relevant anyway.

Patricia Croxton-Smith updated Ashdon’s case in 2002. Her only valid new argument is that ‘Assandun’ probably meant ‘the hill of the ash trees’, and that Ash trees like chalky land like that around Ashdon but not the London clay around Ashingdon. She has a bunch of invalid arguments too. She says Liber Eliensis describes how the monks of Ely bore the body of Bishop Eadnoth back to Ely on the night of the battle, which is possible from Ashdon but not from Ashingdon. It is bogus because Liber Eliensis does not say or imply any timescale. It might have taken weeks to return his body, so the battle could have been anywhere. She says that she found evidence contradicting Freeman’s assertion that the Danes “reached their ships” before the battle, but he says no such thing. She says, “a local county history of Cambs reports that weapons were found in Red Field in the 1850s when the railway cutting was dug”, but the original report  does not mention weapons. Instead, it says: “The remains of a Roman villa were discovered in 1825 in a field separated by the river Granta from the site of Barham Priory, and in 1862, when excavating for the railway from Cambridge to Sudbury, the workmen met with the remains of numerous skeletons in this field at a depth of 3 feet from the surface.” Even this might be faulty because the original evidence is lost, and no archaeology has been found at the location since. A 2015 geophysical survey and a 2017 archaeological dig failed to find evidence of military action near Ashdon.

In  our opinion, the evidence supporting both the orthodox battlefield candidates is weak, faulty or irrelevant. Even the refutations are mostly weak or faulty. The only useful conclusion from the longstanding balanced support for two battlefield candidates is that they each match the battlefield name clue, that there is insufficient supporting evidence to win over supporters of the other theory, and that there is no knockout evidence to refute either of them. Neither Ashingdon nor Ashdon look like promising battlefields to us, for logistical reasons.

Medieval armies moved around on Roman roads. Ashdon was three miles north of Margary 300, five miles south of the Via Devana (Margary 24), and five miles east of Margary 21b (Figure 1). It sounds like a communications hub, close enough to several Roman roads to be a plausible battle location, but we think not. One reason is that the ASC says Edmund pursued the Danes and ‘overtook’ or ‘caught’ them, implying that both armies were on a Roman road. Also, the Danes must have had dozens of carts carrying armour, tents and booty. They would not have left a paved Roman without good reason. It is possible that they went to plunder Ashdon, but it would have been poor pickings lacking even a church if it was the battlefield. It is possible that the Danes left the Roman road when they saw the English coming because they found somewhere better to defend, but that would not have been Ashdon, on a downhill spur that backed onto a boggy stream. Medieval events are often associated with the nearest named settlement, so they might be several miles from the batttlefield, but Ashdon would not have been the nearest named settlement to anywhere on a Roman road because Saffron Walden, Chesterford and Linton were on those Roman roads. Ashdon is therefore inconsistent with the English accounts.

Ashdon is not better a match to the Encomium. It omits the Danes plundering activities. Instead, it cuts to the chase, just saying that the Danes were at their ships when they got news of Edmund’s approach, so they left their ships to face him. From Ashdon, the nearest navigable stretch of estuary that intersected a Roman road would have been Colchester, so the fleet was at least 20 miles away. That is more than a day’s march, so they would have had to take tents and food. It seems unlikely to us that the Danes would encumber themselves and march twenty miles to defend a miserable downhill spur like Ashdon when there were dozens of better hills on the way.

Ashingdon, four miles northeast of Raleigh (Figure 1), is a much better match for the Encomium, just a few hundred metres from a navigable mooring place, but it makes no logistical sense, 15 miles from the nearest Roman road. Burne has caused a certain amount of confusion by referring to this part of the Crouch estuary as “Burnham roadstead”, giving the impression that it is close to a Roman road, but he means in the nautical sense: “a place less enclosed than a harbour where ships may ride at anchor”, according to Websters Dictionary. In our opinion, the Danes would not have chosen to haul armour, tents and booty across one unpaved mile.

Moreover, the land route from Ashingdon to Mercia was tough and indirect. Ashingdon is labelled with a white A on Figure 2. It sits on a peninsula between the Crouch and the Thames. The bridge at Battlesbridge was not there in the 11th century, so the land route to Mercia would have been southwest along the Rayleigh escarpment, across the marshy isthmus at Benfleet, around the south of Basildon, northwest to join the Pye Road at Brentwood, thence to Colchester and Margary 32 to Braughing. It seems unlikely that the Danes could have taken this route because it crossed twenty miles of unmetalled unpaved ground, a marsh between Benfleet and Basildon, and several inclines that were too steep for carts. Burne presumably realised this when he proposed that the raiding party was dropped off on the north bank at Burnham-upon-Crouch, but they would still have had to return to the south bank for the battle to have taken place at Ashingdon.

Both banks of the Crouch between Burnham-upon-Crouch and Hullbridge were lined by 1km of tidal marshland. The southern bank has been partially drained by the Old Fleet River, but it is still boggy today. It seems unlikely that the Danes would have moored their fleet in a marsh to camp at Canewdon. Even if they did, they would surely have collected the raiding party from Battlesbridge or Burnham-upon-Crouch, which would preclude a south bank battlefield, at Ashingdon or elsewhere.

If Burne is right that the Danes plundered southern Mercia on their third raid, and we think he is right for reasons that we explain below, they had better mooring options. In principle, they could have rowed up the Thames into Mercia, but that would have taken them past Edmund’s stronghold in London. Next best would be one of the north Thames estuaries downstream of London. The Danes had moored in all three of these estuaries on pre-Cnut raids into Mercia. Each was less than half the marching distance of the North Sea estuaries, and entirely on Roman roads. But if Cnut moored in a north Thames estuary, the battle could not have happened at Ashdon or Ashingdon, each more than 20 miles in the wrong direction, and there are no other candidates in modern Essex.

A battlefield in Hertfordshire

It has always been assumed that the Assandun battlefield must be in modern Essex because the Anglo-Saxon Chronicle and Henry of Huntingdon seem to be specifically saying so. The former (Whitelock translation): “When the king learnt that the [Danish] army had gone inland, for the fifth time he collected all the English nation; and pursued them and overtook them in Essex at the hill which is called Assandun”. The latter (Greenway): “Cnut with all the Danish armies gathered in Essex at Ashingdon. And so the fiercest and final battle was fought”. But these references to Essex might not mean modern Essex.

Essex was not always synonymous with the modern county. Ancient counties that lack a ‘shire’ suffix - Essex, Kent, Sussex, East Anglia, Northumbria – were effectively kingdoms in the so-called ‘Heptarchy’ before England was unified by Alfred the Great. Essex was then known as ‘East Seaxan’, the land in the east that was occupied by ethnic Saxons. It was bounded by the Thames to the south and by the Colne to the west. That land encompassed the southeast of modern Hertfordshire and all of modern Middlesex. Its bounds are described in ‘The Place-Names of Essex’ and depicted by Bartholomew (Figure 3).

East-Seaxan had an eventful early history, getting conquered by Mercia, Wessex, Danes, Wessex again, Danes again, then becoming part of Edward the Confessor’s united England, before getting conquered for good, along with the rest of England, by the Normans. It ceased being a kingdom in 825 when it was conquered by Ecgberht, King of Wessex. The current county boundary, without Middlesex and Hertfordshire, was probably established by Alfred. Yet the original boundary was not lost. It retained a vestigial Diocesan meaning as the ‘East Saxon See’ into the 18th century.

Modern Middlesex and southeast Hertfordshire had more complications. They were subjugated in the 8th century by Mercia whose kings regularly granted out land therein. But East-Seaxan kings also granted land in modern Middlesex and Hertfordshire. Barbara Yorke explains: “When East Saxon kings granted land in Hertfordshire or Middlesex, they frequently made reference to their foreign [Mercian] overlords whereas in Essex they granted land freely.”  In other words, she is saying that modern Middlesex and southeast Hertfordshire were vassal parts of East-Seaxan with Mercian overlords.

So, what did the Anglo-Saxon Chronicle and Henry of Huntingdon mean by East-Seaxan? There are four viable possibilities: 1) The post-Alfred county of Essex; 2) A vestigial geographic term for the historic East-Seaxan kingdom; 3) The eastern lands inhabited by Saxons; or 4) The Diocesan East Saxon See. The first would exclude Middlesex and southeast Hertfordshire, the other three would include them. There is no evidence that the Assandun references to ‘East Seaxan’ and ‘Estsexe’ referred to the post-Alfred ‘shire’ of Essex, so the battle could have been fought in modern Middlesex or southeast Hertfordshire.

Mooring place of the Danish fleet

Most  historians assume that the Danes moored in a North Sea estuary, because that is the only way that they could have got to Ashdon or Ashingdon. It is a circular argument. The only supporting evidence in the contemporary accounts is Forester’s translation of John of Worcester: “Canute with his forces crossed the river into Essex”. The river is assumed to be the Thames because they were camped on the Isle of Sheppey. John of Worcester does not say that the Danes went upstream, so most historians assume that they did not. If they did not go upstream, they must have moored in a North Sea estuary. Forester’s translation is unreliable.

The Latin original says: “Canutus suas copias in East-Saxoniam trajecit”. No mention of a river. ‘trajecit’ has a rare niche meaning ‘he crossed’, without any implication that a river is being crossed, but more usually means ‘he marched’. Stevenson translates: “Canute led his forces into Estsaxe”, McGurk: “Cnut sent his troops into Essex”. So, there is no evidence that the Danes moored in a North Sea estuary. Indeed, there is no hint in the contemporary accounts where they moored.

Burne’s conclusion that the Danes plundered south Mercia on their pre-Assandun raid could be significant if right, but his reasoning is unconvincing. He reckons that they plundered south Mercia on their pre-Assandun third Mercia raid because they had plundered north Mercia on their second raid. His only evidence is that they moored in the Orwell before their second raid. His geography is awry. Mercia was some 140 miles from south to north. The Orwell is the northernmost of the Essex estuaries, but it was only about 40 miles north of the southern border, just about level with the southernmost parts of central Mercia. If the Danes had intended to raid north Mercia on their second raid, they would have moored in a Lincolnshire estuary. Even so, we think they did raid southern and western Mercia before Assandun.

The Anglo-Saxon Chronicle describes Cnut’s first Mercia raid (Whitelock): “[Cnut] crossed into Mercia at Cricklade”, near Swindon. They must have headed north because they “ravaged and burned” Warwickshire. King Ethelred commanded his son and heir Edmund to repel the invaders. Edmund led the English army to York where they joined Uhtred’s Northumbrians and headed back south. Their men were reluctant to fight without King Ethelred being present, so they plundered northwest Mercia instead. Cnut, presumably realising that Northumbria had been left undefended, skirted around the English army, ravaging through Buckinghamshire, Bedfordshire, Huntingdonshire, Northamptonshire, Lincolnshire, and Nottinghamshire until he reached York. Uhtred ceded Northumbria to Cnut. Cnut’s second Mercia raid probably avoided the places they plundered in the first. They moored in the River Orwell which suggests that their main target was Cambridgeshire and the parts of eastern Mercia that got away lightly in their rush to get to York.

Cnut’s pre-Brunanburh third raid into Mercia probably targeted the shires that were unscathed from the first two Danish raids and unscathed by Uhtred and Edmund. This means Leicestershire, Derbyshire, Staffordshire, Oxfordshire, Gloucestershire, and Worcestershire. The first three were traversed by Watling Street, the last three by Akeman Street. These major Roman trunk roads met at St Albans in the southeast corner of Mercia. It is the most likely gateway via which the Danes entered and left Mercia on their third raid.

If the Danes plundered southern Mercia on their third raid, the North Sea estuaries were unnecessarily distant. In principle, they could have rowed up the Thames into south Mercia, but it passed Edmund’s stronghold in London. The north Thames estuaries downstream of London – the Lea, Roding and Benfleet - would have been a good compromise, less than half the distance to Mercia compared to the North Sea estuaries, and entirely on Roman roads. But if the Danes moored in a north Thames estuary, the battle could not have happened at Ashdon or Ashingdon.

Pre-Cnut Vikings had moored in all three of the major north Thames estuaries downstream of London before inland raids, but the route from the Roding and Benfleet passed within a mile of Edmund’s stronghold in London, and they were twice as far from Mercia as the River Lea. In our opinion, the River Lea is the most likely mooring place for Cnut’s fleet, and it reveals a promising new battlefield candidate.

A battlefield at Essendon in Hertfordshire

The Anglo-Saxon Chronicle says that the Danes were attacked on their way back to their ships. The Encomium says that they left their ships to face the English. In either case, they would have been on a Roman road, or adjacent to one, on a route between the ships and Mercia. If Cnut moored in the River Lea, that route would have taken the Danes close to Essendon in modern Hertfordshire. It fits the battlefield location clues at least as well as Ashdon or Ashingdon, without their logistical drawbacks. It is a credible third battlefield candidate.

Cnut would have moored where a Roman road came close to the river. Four Roman roads intersected with the Lea at three different places (see Figure 4). Ermine Street crossed the Lea at Ware. Margary 214 was adjacent to the Lea at Hertford and Ware. Margary 212 and Margary 213 terminated at Cheshunt. Ermine Street can be discounted because it did not go to Mercia. Margary 212 and Margary 214 went to St Albans, Margary 213 went to Dunstable, the gateways to south Mercia. It therefore seems likely to us that the Danes moored at Herford, Ware or Cheshunt.

Margary 214 did not pass within four miles of Essendon. The Danes might have branched onto Margary 220, thereby passing close to Essendon, but there is no obvious reason they would have done. Therefore, Essendon is more likely to have been the battlefield if the Danes moored at Cheshunt. We think it probable because Cheshunt had several important advantages over Ware and Hertford.

Cheshunt is downstream of Ware and Hertford, where the river is 24m lower. Cheshunt was easily navigable, just 6m above the tidal limit, and just one weir. In principle, the Lea was navigable to Hertford in the 9th century, but it would have needed a lot of portage being 30m above the tidal limit and eight weirs.

Cnut would probably have liked to moor his fleet next to a fortification, to help with the defence of the ships while the raiding party was away. Hertford would have been attractive, with burh fortresses on both sides of the river built by Edward the Elder in 912. Ware is thought to have had a fortress, built by Hæsten in 895. Cheshunt too. According to ‘Place names of Hertfordshire’, the first syllable of Cheshunt’s name derives from Old English ‘ceastre’, meaning ‘fortification’, typically of Roman origin. Cheshunt was at the end of two Roman roads, which suggests it probably had some sort of fortification. We suspect that it was reinforced by Alfred the Great to trap Hæsten’s Danes.

The only evidence for the location of Hæsten’s fortress is in the Anglo-Saxon Chronicle which says that it was “XX Mila buran Lunden-byrig”, usually translated as ‘twenty miles above the City of London’. We think that it was using Roman miles which would place Hæsten’s fortress no further north than Hoddesdon. The ASC (Whitelock) goes on to say that: “One day the king rode up along the river, and examined where the river could be obstructed, so that they could not bring the ships out. And then this was carried out : two fortresses were made on the two sides of the river.” John of Worcester says that Alfred built piers out into the river to trap Hæsten’s fleet upstream. Henry of Huntingdon says that Alfred split the river into multiple channels, each too shallow for Hæsten’s ships. Cheshunt would be the obvious place for all these obstructions because it is just south of a constriction where the flood plane widens to 4km.

Hæsten was forced to abandon his ships and flee inland because Alfred blocked the River Lea was downstream. Cnut would surely have been conscious of a similar fate, encouraging him to moor at the lowest practical intersection with a Roman road. as far downstream as he could. That means Cheshunt. He would have used Margary 212 to St Albans and then taken Watling Street and/or Akeman Street (red lines on Figure 4) to plunder Mercia. If so, the Danes passed within spitting distance of Essendon.

Matching the battlefield candidates to the battlefield location clues

We explain above that Essendon is a better match to the logistical battlefield location clues than Ashingdon or Ashdon. It is interesting to compare the three candidates against the other battlefield location clues.

Etymology

Place names were spelled phonetically in medieval times. The contemporary accounts have a few stabs at the battlefield name: ‘Assandun’, ‘Esesdune’, ‘Æsceneduno’ and ‘Assatunis’.

Philologist Percy Reaney thinks that the ‘Ass’ part Assandun probably derives from Old English ‘æsc’, meaning ‘ash tree’ and pronounced ‘aysh’. If he is right, the ‘Ass’ part of Assundun was pronounced ‘Æsc’. Reaney reckons that roughly three-quarters of Old English ‘Æ’ ligature placenames in Essex – at least those where the Æ did not immediately precede a hard consonant – evolved into Middle English ‘A’. The rest evolved into Middle English ‘E’. So, it is more likely that Assandun’s ‘Æsc’ would become ‘Ass’ or ‘Ash’, as in Ashdon and Ashingdon, but not unlikely it might become ‘Ess’ or ‘Esh’, as in Essendon. According to the British Place-Name Gazetteer, it is similarly likely that the last two syllables would evolve into ‘ingdon’, ‘endon’ or ‘don’. So, there are plausible name transitions for all three candidates.

Reaney and Whitelock mention another possibility, that ‘Assandun’ and ‘Æscesdune’ might derive from Old English ‘Æssa’s dun’, where Æssa is a personal name and ‘dun’ means ‘hill’. Knutsdrapa’s Assatunis is an Old Norse genitive singular declension, which supports this conjecture. If Assandun was ‘Æssa’s dun’, it is equally likely to have evolved into ‘Essendon’ or ‘Assingdon’ (Ashingdon’s name until the 19th century) or ‘Ashdon’.

John of Worcester has a different take, saying that ‘Assandun’ meant ‘mons asini’, the ‘hill of asses’. It would have had a hard ‘A’, favouring a transition to Ashingdon or Ashdon rather than to Essendon, but it does not seem likely. Donkeys were rare in Saxon England, and they were too valuable to roam free over hills. Swete reckons that Assandun’s name might have been a spiritual dedication because asses were sacred to some pagans, but this original would be incredibly rare. It is much more likely, as Reaney says, that John of Worcester was confused.

We should return to Swete’s argument about Essendon church. As he says, the Baignard family gifted Essendon church to the St Pancras Priory in the 13th century, and the gift manuscript says that it was also known as ‘Assenden’, ‘Assendon’ and ‘Asshedon’. His point was that the Baignard family held the manor of Ashdon (then known as Ascenduna) in Domesday. He reasons that the gift therefore probably referred to the church at Ashdon in Essex and that it was previously known as ‘Assendon’. If so, he argues, the Battle of Assandun was probably fought at Ashdon in Essex. There is a twist.

Hertingfordbury manor in Hertfordshire was also held by the Baignard family in Domesday. Essendon does not have its own Domesday entry. It was midway between Hertingfordbury and Hatfield, both Domesday manors, so Essendon was probably in one or the other. The River Lea runs between the modern towns of Hertingfordbury and Essendon. The ‘ford’ part of Hertingfordbury’s name suggests it spanned the River Lea, so it seems likely that modern Essendon was in Hertingfordbury manor, held by the Baignard family. If so, they could have gifted Essendon church to St Pancras Priory. Its primary name was Essendon, which hints it is more likely. If the gifted church was Essendon’s, it is evidence that the battle was there too.

We have investigated the St Pancras papers for years without finding compelling evidence that the Baignard’s gifted church was Essendon or Ashdon. They never say or imply whether it was in Hertfordshire or Essex, and it might not be useful anyway because, as a religious institution, its idea of ‘Essex’ would probably have been the East Saxon See. There are a dozen or more references to Essendon in Charters, Pipe Rolls, Feet Fines and the like. Some references are associated with ‘Newenham’ (Newnham) in Essex, but others are associated with Bayford or Epping in Hertfordshire. Most of the references to Essendon are in the Essex volumes of the Pipe Rolls or Feet Fines without an alien county note, which suggests that Essendon – and therefore Assandun - is an alias of Ashdon. But those volumes also have separate references to Asshedon, or cognate, which suggests the opposite.

Even though they are 25 miles apart, it is possible that joint references to Essendon and Newenham referred to Essendon in Herts and Newenham in Essex because they were both owned by the Baignard family until the mid-13th century. They could have gifted land in both manors in a single writ. While this is possible, we have never found any specific supporting evidence and if this is so, our only explanation for why most of the references to Essendon are in Essex records is that there were two places named Essendon. The jury is out.

Note that Wikipedia says that the Battle of Assandun is also known as the ‘Battle of Essendune’. This sounds like good supporting evidence that the battle was fought at Essendon, but we suspect it is a red herring. Wikipedia cites Earnest Frederick Smith as its source, and he does indeed say: “after the Battle of Essendune in 1016, Canute and Edmund Ironsides met on a small island in the Severn”, but he gives no source and the battle is only of incidental relevance to his book’s main subject, Tewkesbury Abbey, so he probably did very little research into it. The only previous reference to a ‘Battle of Essendune’ is in Rev Frederick George Lee’s paper about St Mary’s Ashingdon in Buckinghamshire when referring to Alfred’s ‘Battle of Ashdown’ in 871. We suspect that Smith got confused.

Our conclusion is that the names Ashingdon, Ashdon and Essendon all derived from somewhere sounding like Assandun. It seems that Essendon was also known as Ashdon and/or that Ashdon was also known as Essendon in medieval times. Ashingdon, on the other hand, is the same form as Abingdon, Huntingdon, Edington and others with a proven transition from a name like Assandun. Perhaps, then, it is marginally more likely that Assandun would evolve into Ashdon or Ashingdon, but all three candidates are consistent with the etymology.

Danes’ Wood and ash woods

Knytlinga Saga says that the battlefield was north of Dane’s Wood. Ashingdon is only five miles north of the Thames. The area is too small to have had a significant wood, especially not one with such a portentous name as Dane’s Wood. Unsurprisingly, then, the only Domesday woodland listed in that area is a copse in Rochford, too insignificant to be named Dane’s Wood, if it had a name at all.

Essendon and Ashdon were north of the Forest of Essex which covered much of Essex and Hertfordshire south of Margary 32. There is no evidence that it was ever known as Dane’s Wood, but it was one of the biggest woods in the Danelaw, so it might have been.

There is a ‘Dane Wood’ ten miles from Essendon. It is to Essendon’s northeast rather than to its south, but the 1883 Ordnance Survey map shows it was part of a bigger wood between Much Hadham and Hatfield that arced south of Essendon. There is no evidence that it was known as ‘Dane’s Wood’, but it would have encompassed a manor house named ‘The Danes’ 1km southeast of Essendon. That manor house is only Georgian, but it might have taken its name from an eponymous woodland, albeit there is no evidence that it did.

In short, Ashingdon was not north of Dane’s Wood, or any other significant woodland, while Ashdon and Essendon were north of a huge wood, albeit there is no evidence that it was known as Dane’s Wood. Essendon was north of a nearer wood for which there is evidence of links to the Danes, but it also lacks evidence that it was ever known as Dane’s Wood.

We mention above that Assandun probably meant ‘hill of ash trees’, and ash trees like chalky soil. Essendon is even more chalky than Ashdon. Its soil is full of flint, and it is surrounded by old chalk excavations, including ‘Ashfield Farm’ less than 1km away. ‘Ashen Grove’ and ‘Ashendene Farm’ are nearby. Ash trees would have been at least as common at Essendon as at Ashdon, and far more common than at Ashingdon.

Cnut’s minster

Cnut commissioned a minster church near the Battle of Assandun battlefield that was consecrated in 1020. St Andrews Ashingdon and St Mary’s Ashdon are not old enough. According to Historic England, St Andrews is: “Said to be on the site of the Church built by King Canute AD 1020”, but there is no evidence of it. Swete reports that he saw earlier foundations beneath St Mary’s Ashdon when repairs were being made, but they were not dated and have not been seen since. Miller Christy was convinced that Cnut’s minster became St Botolph’s Church at Hadstock, two miles from Ashdon. Its door was dendrochronology dated to 1035, making it at least 15 years too young. The entire area was intensively surveyed and excavated in 2005, to reveal no evidence of battles or occupation in the 11th century.

Essendon’s surviving church is no older than Ashdon’s or Ashingdon’s. But, as we mention above, Essendon was probably in Hertingfordbury manor in Saxon times. The Hertfordshire VCH explains, “In many instances the church lies a short distance from the village and adjoins the court or hall which in almost all Hertfordshire parishes retains the Anglo-Saxon title of ‘bury’.” In other words, Hertingfordbury probably had a significant Saxon era church that was outside the settlement. There is no reason it could not have been at Essendon and if it was not, it could have been near enough to take its name.

In summary, there are no churches of the right age or construction to have been Cnut’s Assandun minster at Ashdon, Ashingdon or Essendon. There is anecdotal evidence of possible a Saxon church at Ashdon and Ashingdon. There was probably a Saxon era church at or near Essendon. It is possible that any of these might have been Cnut’s minster. None of them seem any more compelling than the others.

The Battle of Ashdown

In 871, Alfred fought the Great Heathen Army in the Battle of Ashdown. According to the Anglo-Saxon Chronicle, the battlefield was at ‘Æscesdune’. It looks similar to Assandun, even more similar to the Enconium’s ‘Æsceneduno’, and allowing for transliteration, similar to Henry of Huntingdon’s ‘Esesdune’. Moreover, the St Pancras Priory papers say that ‘Assendon’ was also known as ‘Asshedon’. Perhaps, both battles were fought at the same place. If so, the Battle of Ashdown’s location clues could perhaps be co-opted to find the Assandun battlefield.

The Anglo-Saxon Chronicle says that the ‘Battle of Æscesdune’ was four days after the same belligerents fought at Reading and two weeks before they fought at ‘Basing’ in modern Basingstoke. There are three more references to Æscesdune in the Anglo-Saxon Chronicle. In 648, Cenwealh King of Wessex gave three thousand hides of land near Æscesdune to a kinsman. In 661, Cenwealh fought Wulfhere at ‘Posentesbyrig’ and was chased to Æscesdune. In 1006, the Danes went to Wallingford, spent “one night at Cholsey, and then turned along Æscendune to Cuckamsley Barrow”.

Historians interpret these clues to mean that Alfred’s Æscesdune was Ashdown in Berkshire, mainly because it was near to Cholsey. It has never been seriously considered that Æscesdune might have referred to Ashingdon or Ashdon, 80 and 100 miles away respectively, because the Danes could not have marched to either of them from Reading in three days. But they could have marched 45 miles from Reading to Essendon in three days.

This is not to say that it is likely that the Battle of Ashdown was fought at Essendon, but it might have been whereas it could not have been fought at Ashingdon or Ashdon. If the Battle of Ashdown was fought at Essendon, it would resolve all the Battle of Assandun names used in the contemporary accounts: Assandun, Esesdune, Æsceneduno and Assatunis.

Summary

The battlefield location evidence is sparce and vague, but it matches Essendon in modern Hertfordshire at least as well as Ashingdon or Ashdon. Moreover, Essendon is consistent with all the battlefield location clues, such as they are, whereas Ashingdon and Ashdon are each inconsistent with at least two of them.

The meaning of 'offerde'

One frustration is that the English accounts and the Encomium seem to contradict each other’s core events. The Encomium says that the Danes got news of Edmund’s approach while they were at their ships and left to fight. The English accounts are always interpreted to be saying that the English ‘pursued’ and ‘overtook’ the Danes as they made their way back to their ships. Whitelock’s translation is above. Garmonsway has: “he [Edmund] called up all the people of England and followed them [the Danes] up, overtaking them in Essex at the hill called Ashingdon”.

It impacts the battlefield location search: 1) If the Encomium is right, the battlefield would have been within a few hours march of the Danish camp, whereas if the English accounts are right, it could have been days away; 2) If the Encomium is right, Cnut had time to choose an advantageous battlefield, whereas if the English accounts are right, he may have been forced to compromise.

The Encomium looks unambiguous. We suspect that the Anglo-Saxon Chronicle is being mistranslated. The relevant Old English statement from the D recension is: “offerde hi innon Eastseaxan æt þære dune þe man hæt Assandun”. The other recensions also use the term ‘offerde’. It is always translated as ‘overtook’, which is understandable because Old English ‘ofer’ is the root of the modern English word ‘over’, but that cannot be what it meant. There is no possibility that the English followed the Danes and overtook them, at least in the modern sense, for the obvious reason that they were on the same road, and it was too narrow for armies to pass each other without engaging.

Bosworth-Toller’s entry for ‘offerde’ has: “to come up with those who are pursued, to overtake, to get near enough to attack, to reach and attack”. Its only example of the first two meanings is this passage from the Anglo-Saxon Chronicle. It seems to us that its third and fourth meanings are more usual, and either of those meanings would make the Anglo-Saxon Chronicle consistent with the Encomium.

In other words, we think that the English accounts are trying to say that Edmund followed the Danes’ route back to their ships, perhaps a day or two behind. When he got close enough to attack (‘offerde’), the Danes left their ships to confront him at Assandun. It would place the battlefield within an hour or two’s march of the Danish camp. This would be consistent with a battlefield at Essendon.

Battlefield geography clues

Two clues have not yet been considered: Clue 7 that Cnut “very slowly brought his men down to a level ground” before being attacked by the English, and Clue 8 that the English were betrayed by Ealdorman Edric Streona.

Clue 7 is always interpreted to mean that both armies were on rising ground and that the Danes descended to level ground between the armies. But Edmund had the bigger army, and he was the aggressor. Cnut had no obvious incentive to move his army to disadvantageous terrain. This is especially true if the Encomium is right that the Danes left their ships after choosing to fight, because Cnut would also have chosen the battlefield. If it difficult to believe that he chose to occupy level terrain between two hills.

There is another possibility. John of Worcester does not say that Cnut advanced to level ground, so he might have fallen back to level ground. There are several plausible reasons he might have done this:

1. Cnut was taken by surprise, realised that the English army was too powerful so decided to bolt for his ships, but got caught on level ground by Edmund’s rapid advance.
2. Cnut was taken by surprise, realised that the hill occupied by his army was vulnerable to flanking, so fell back to defend a narrow – albeit level - plain.
3. Both armies were on a ridge (a type of hill) when they first sighted each other, but the English were higher up the crest, so Cnut reasoned that it would be better to fall back to level ground than to fight on falling ground.
4. Cnut was drawing the English into a trap.
5. Cnut knew that Edric Streona would betray Edmund by withdrawing his flank during the battle and he wanted to be on the perfect terrain to take advantage of it.

All of these are feasible. None of them fit the geography at Ashdon or Ashingdon, but they all match the geography near Essendon, and any combination of them are plausible.

Revised battle narrative

If, as we conjecture, the Battle of Assandun was fought at Essendon in Essex and the Anglo-Saxon Chronicle’s ‘offerde’ meant ‘to get near enough to attack’, as we explain above, the events described in the contemporary accounts come to life.

The pre-engagement is as described in the contemporary accounts. The Danes returned to their ships after plundering Mercia. Their fleet was moored in the River Lea, probably at Cheshunt, so they took the Margary 212 Roman road from St Albans. Edmund and the English army were in London waiting for them to return. Edmund was notified that the Danes had crossed north of London. The English army left the following morning. It was only 13 miles from St Albans to Cheshunt on a paved road, so the Danes made it back to their ships the previous day.

Edmund had a choice of two routes: 1) North up Margary 220 then east on Margary 212; or 2) East on Margary 3a (which Margary referred to as the ‘Great Road’) then north up Ermine Street. Ermine Street ran parallel and close to the River Lea meaning that it was on the same elevation as the foe, overlooked all the way by higher ground to the west and lined by boggy ground to the east, making it dangerously ambush prone. The other route, on Margary 220 and 212, looks far preferable. They ran along ridge crests, making it impossible to be caught on adverse terrain. Edmund might even have hoped that the Danes could be taken by surprise, because the 212 descended from a ridge to the River Lea, potentially allowing Edmund to trap them on their ships. Moreover, it matches the ASC’s statement that Edmund ‘pursued’ the Danes because they had used the 212 the previous day.

Cnut received the news of the English army’s departure soon after they left London. He decided to fight. It is interesting to speculate why. Of course, if he could defeat the English army and kill Edmund, he would take the richest country in Europe, so he had a huge incentive, but it would not be much use if he lost. He would have been told that Edmund had more men, and that the English were coming on Margary 220 and 212, which provided no outstanding ambush opportunities or defensive locations. Perhaps Cnut was forced to fight because Edmund had blockaded the Lea. Perhaps he was confident in his men’s ability to defeat a bigger force. More likely, we think, the English accounts are right that he had done a deal with Edmund’s traitorous confidant Edric Streona to withdraw his flank at a crucial moment in the battle.

The only clues about the battle’s exact location are the battlefield’s name and John of Worcester’s passage about Cnut leading his army down to level ground. Medieval battles were almost never fought at settlements, so they were usually named after their nearest named settlement. This implies that the engagement and battle were within a mile or two of Essendon.

The most straightforward scenario is that the armies engaged on Margary 212, where it ran parallel to the modern Woodfield Lane and Tylers Causeway, two miles south of modern Essendon. But this would not account for Cnut leading his army to level ground. If they were on this stretch of Margary 212, they were already on the level crest of the ridge. Perhaps, John of Worcester was mistaken, but it is an unusual, maybe unprecedented, tactic to fabricate. There is a possible explanation at Essendon.

First, we have to note that Margary generally joined his Roman road discoveries by straight lines, not unreasonably because the Romans liked to build straight roads. In the case of the 212, the road evidence was near Brookman’s Park and at Cheshunt, so he drew in a tentative straight road between the two (shown as a faint black line on Figure 5 and Figure 6). He did the same with Margary 214 between Dunstable and Cheshunt (shown by a slightly more northerly faint black line). In practice, it is implausible that there would have been two Roman roads within a 100m or each other, and almost as unlikely that they would cross the steep-sided valley near Cuffley. We guess that they joined south of Essendon to follow the ridgeway down to Cheshunt, as shown by the dotted line on Figure 5 and Figure 6.

If Margary 212 did follow the ridgeway between Essendon and Cheshunt, the geography is consistent with Streona’s treachery. We list some reasons that Cnut might have fallen back to lower ground in the Battlefield geography section above. The last two would apply at Essendon. Cnut knew that Streona would withdraw his flank just as the armies engaged, so he backed away from the English army luring them down the ridge crest.

So, Essendon was the nearest named settlement to the initial engagement. It means that the English were heading east on Margary 212, and that they were at Essendon when they spotted the Danes along the road at Tylers Causeway. The ridgeway is level between Essendon and Epping Green. It then drops 10m to a wide level plain between Ashendene Farm and the Ponsbourne Tunnel. Cnut slowly fell back to Ashen Grove - marked with a red X on Figure 7 - luring the English down the slope. Just as the armies were about to make contact, Streona ordered his flank to flee, allowing his men to surge through the gap and occupy the slope. The English position was hopeless, so they fled, at great loss of life.

Er, ‘Ashendene’!? Could we have done an Inspector Clouseau by fingering the right battlefield for the wrong reasons? We think not but it is an eery coincidence.

The Battle of Assundun at Essendon location clues are few, vague and ambiguous. It is the least compelling of our lost battlefield studies, but the battle was fought somewhere and, in our opinion, it was not fought at Ashdon or Ashingdon. Given something of a beggar’s choice between Ashingdon, Ashdon and Essendon, the latter is overwhelmingly the most likely battlefield because it is the only one that is consistent with all the clues.

As always, we are happy to hear from anyone that has extra evidence, or that can help narrow down the battlefield location or who can refute anything we say. Please contact us on our usual email address: momentousbritain@outlook.com