Wild Fire or Greek Fire? Game of Thrones and medieval sieges

It is a fact that millions more people have gained a view of medieval sieges by watching Game of Thrones than by reading books by experts or learning history at school, so it is worth asking: how true a picture does the blockbuster series give?

Across the seven series so far, we have seen many bloody captures of strongly fortified cities and castles by large armies of armoured warriors, supported by archers and fleets of ships. The attack on King’s Landing (series 2, ep. 9) provided some realistic glimpses: the preparations of the citizens and garrison for battle, the taking refuge in the keep of women and children – and the exemplary execution of attendants who tried to flee. The ominous clanging of the bells gave warning of assault heralded by banging drums and the landing and drawing up of the attacking army, the bringing up of a large battering ram to batter down the main gate while scaling ladders were prepared. As everyone knows, the city was saved when the dwarf prince Tyrion inspired his men by appealing to their commitment to their families and homes to launch a sally via a postern gate which took the besiegers in the flank, and, accompanied by the timely arrival of a relieving army, saved King’s Landing.  Prior to this, the invading fleet had been consumed by the launch of the lethal ‘wildfire’ via a fireship.

Bronn lighting the wildfire with a flaming arrow. Image ©HBO.

Was ‘wild fire’ inspired by the famous ‘Greek Fire’? The real ignis graecus (mentioned in numerous Latin chronicles) was invented by Byzantine engineer Kallinikos and used to save Constantinople from Arab attack in 672 AD: its exact constituents remain unknown to this day but like ‘wild fire’, it could not be extinguished by water (hence its main use at close range against enemy ships), and among its secrets was the means of propulsion – possibly by siphon mounted on the front of ships, or on castle towers. The original formula was lost but incendiary devices continued in use, especially during the crusader period, using plentiful supplies of oil. But never again were they so decisive.

Other regular features of medieval sieges appear in GoT. The threat of hanging hostages in view of the occupants if they did not surrender, shown in the siege of Riverrun (ser. 6, ep. 7–8), maybe echoed the threat of King Stephen to kill the child who became the ‘greatest knight’, William the Marshal, if his father did not surrender Newbury (1152): in both cases the putative victims were reprieved. In other respects too a realistic scenario is painted: the besiegers erecting defences to protect their rear, building siege towers and stone throwing trebuchets, and setting camp in a vast array of pavilions flying numerous banners; defenders gathering sheaves of arrows and relying heavily on crossbows which, used from behind stone walls, could be devastating.

The siege of Riverrun. Image ©HBO.

Riverrun yielded. At other times, the attackers avoided the risk of an assault by exploiting tunnels to enter – Tyrion and Daenerys captured Castly Rock (ser. 7, ep.3) through the sewers. In 1204, King Philip Augustus’ army stormed the middle bailey of Richard the Lionheart’s ‘impregnable’ Château Gaillard in Normandy when one of his men discovered a drain outlet in the ditch that led him up into the chapel.

GoT skates over many realities of medieval warfare, particularly questions of logistics. The siege weapons materialise from nowhere when in reality they needed time for construction. All are settled quickly, whereas most medieval sieges lasted a long time, and hunger and disease killed far more people than fighting. However, many episodes convey brilliantly the drama, the sheer terror, and the brutal horror of combat. So we should recognise GoT as brilliant drama rather than history, and look elsewhere for accounts of real medieval warfare. But while we’re about it, we can only imagine what a medieval general would have given for the three fire-breathing dragons Queen Daenerys unleashed at Meereen (ser. 7, ep. 9).

Dr Peter Purton (author of History of the Early Medieval Siege and History of the Late Medieval Siege (Boydell Press 2010).

Five New Awards and £100,000 in six years

We are excited to announce five grants totalling a record £27,000 that will advance our understanding of castles. These awards mean we have reached the landmark of giving away £100,000 in grants. It has taken six years for us to do that during which time the Trust has doubled the maximum amount we can award to £10,000.

Before you read about the five projects below, be sure to subscribe to our newsletter if you haven’t already.

  • Druminnor, Aberdeenshire – Using GPR for an investigation of the 15th century core of the castle – presently under a hardcore car park. This was the original caput of the lords of Forbes. During the 15th century they were amongst the most powerful families in the North-east of Scotland.
  • Hoghton Tower, Lancashire – This project aims to form an axis of research into Hoghton Tower’s unique physical history. The main focus will be to investigate and advance the knowledge of the pre-1560 site and specifically try to test the hypothesis that the north side building may form part of the ‘original’ Hoghton Tower.
  • Laughton en le Morthen, South Yorkshire – Excavation to try to confirm the findings of the geophysical surveys the CST funded in 2018. These indicated that the castle was placed right on the top of a high status Saxon dwelling.
  • Shrewsbury, Shropshire – Geophysical survey and excavation to determine how the castle buildings were laid out in the bailey. This will be the first time the well preserved motte-and-bailey castle has been excavated. First mentioned in 1069, Shrewsbury Castle was a key point along the Anglo-Welsh border and fell into ruin following the conquest of Wales.
  • Wressle, East Yorkshire – A geophysical survey of the area to the south of the castle ruins which had been covered by the previous earthwork survey funded by the CST, to get more information about the various garden structures there, as well as other details regarding the deserted village, moat and fishponds. The 14th century castle was one of the most important castles owned by the Earls of Northumberland.

Be sure to follow us on social media where we’ll be bringing you updates from the field, with work starting this month. In the meantime, here are some of our supporters discussing our work. we hope you’re looking forward to this year’s projects as much as we are!

Going in search of Keith Marischal

In the later Middle Ages, the Keith family were some of the most important people in Scotland. Sir Robert Keith was made marischal of Scotland in 1293, a title that descended through his heirs. As marishal, Sir Robert and his successors were could hold courts during wartime and were responsible for maintaining order within the Scottish parliament. The castle at Keith Marischal, half a day’s journey from Edinburgh, was the family’s ancestral home.

For a family which was amongst Scotland’s richest in the 16th century, their seat was an important place which would have embodied their power and prestige. The great hall, the social heart of the castle, vied with royal palaces in its size. William Keith, 7th Earl Marischal was forced to sell Keith Marishcal during the Civil Wars, and, despite being an important piece of Scottish history, the castle was gradually demolished. Part of the castle survives and was incorporated into the later house built on the site, but much of Keith Marischal has vanished.

Stirling Castle’s great hall. Photo by ‘DeFacto’, licensed CC-BY-SA 4.0.

In 2017, Miles Kerr-Peterson suggested carrying out a geophysical survey to look for buried remains just north of where the house currently stands. He successfully applied to the Castle Studies Trust for funding, and in May 2018 he and Rose Geophysical Consultants visited Keith Marischal to search for the evidence in an area of 2 hectares.

Two methods were used: resistivity and ground penetrating radar (GPR). As different materials conduct electricity differently, testing the electrical resistance of the ground can be used to find features such as walls (high resistance as there is little water) or ditches (low resistance as ditches tend to hold water), and is effective to a depth of about 0.75m. GPR works by sending electromagnetic pulses into the ground and tracking how they are reflected. Part of the area north of the house is a carpark, which makes survey resistivity ineffective, but GPR can still be used.

Deciphering a geophysical survey takes a trained eye. The resistivity survey found several features, and working out what they are has been an interesting challenge. There are two features running in a mostly straight line perpendicular to the current house at the west end of the survey. The longer of the pair could be a drain, but it’s uncertain. And what are the features at the north end? The feature runs beyond the edge of the survey, so we don’t know the full shape and size of it. With trees nearby it could even be part of a root system, but the straight lines suggest it could be man-made and could be part of the lost castle.

GPR allows us to peer deeper, and to work out a rough stratigraphy of features. The survey was able to corroborate some of the anomalies found with resistivity. The pair of parallel features at the west end are visible, but the one on the right runs deeper. The fact it’s so narrow suggests it might be a drain. The GPR also found an anomaly at the north end of the survey area, lining up with the one found using resistivity. It was visible some 0.38-0.63m deep, which suggests it might be artificial rather than natural.

The results of the survey are certainly interesting. We didn’t find the extent of the lost Keith Marischal Castle, but most discoveries don’t happen overnight. Geophysics is an excellent way to identify areas of interest ahead of excavation. Without excavation, we can’t be sure about the interpretation of these features. If the anomaly at the north end of the survey is part of the lost castle, we don’t have a way of dating it without breaking out a trowel.

The survey was a vital step in the understanding Keith Marischal. Thanks to Miles and Rose Geophysical Consulting, any future excavations will know where to look. Keith Marischal has an exciting future, and the Castle Studies Trust are proud to be able to have played our part in supporting the work.

Dr Miles Kerr-Peterson is an affiliate in Scottish History at the University of Glasgow. His new book, A Protestant Lord in James VI’s Scotland George Keith, Fifth Earl Marischal, touches on the Keiths and is out now.

The projects we’re considering for 2019

The deadline for grant applications passed on 30th November. We’re going through the various projects now. Altogether the 12 projects, coming from all parts of Britain and one from Ireland, are asking for over £75,000.* They cover not only a wide period of history but also a wide range of topics. In a little more detail, here are the applications we’ve received:

Collyweston, Northamptonshire

  • Contributing towards a community excavation at the early Tudor palace commissioned by Henry VII’s mother, Margaret Beaufort. According to building accounts the palace was built around three courtyards and was the first brick building in Northamptonshire. The palace was demolished in 1640 and the site has never been excavated before.

Druminnor, Aberdeenshire

[10] Druminnor Castle - "Woops!"
Using Ground Penetrating Radar for an investigation of the 15th-century core of the castle – presently under a hardcore car park. This was the original caput of the lords of Forbes. During the 15th century they were amongst the most powerful families in the North-east of Scotland.

Hoghton Tower, Lancashire

hoghton tower
This project aims to form an axis of research into Hoghton Tower’s unique physical history. The main focus will be to investigate and advance the knowledge of the pre-1560 site and specifically try to test the hypothesis that the north side building may form part of the ‘original’ Hoghton Tower.

Lathom, Lancashire

Excavation to establish the form and location of the southern perimeter of the curtain wall of the 15th-century castle known as the Northern Court of which nothing remains above ground from the period. It was one of the most important castles in the north west of England in the late 15th and 16th centuries. The castle was besieged twice in the English Civil War and then slighted. In 2017 we funded analysis of masonry recovered in earlier excavations, which indicated architectural links with Caernarfon Castle.

Laughton en le Morthen, South Yorkshire

Photo by Mike Neid

Following on from last year’s grant, this project would undertake excavation to investigate features identified during the geophysical survey. The survey suggested that the castle was built over an Anglo-Saxon lordly residence, and the excavation would test whether there is further evidence to corroborate this.

Lewes, East Sussex

Photo by Richard Gailey, licensed CC-BY 2.0.

This research aims to answer an intriguing question: why does Lewes Castle, East Sussex, have two mottes? Do they represent a highly distinctive architectural statement, or did burial mounds of possible Romano-British or earlier origins influence the form of the 11th century fortification?

Loughmoe, County Tipperary

Castles of Munster, Loughmoe, Tipperary - geograph.org.uk - 1542634
To produce the first-ever detailed survey and structural history of the building which dates from the 15th and 16th centuries, and to determine using geophysics the extent of the original castle and whether the renaissance part had a precinct, other buildings, and gardens.

Raglan, Monmouthshire

Photo © Jeremy Cunnington

Using inferences from previous geophysical surveys to focus on key areas of the lower terraces and bowling green of the castle potentially revealing more about the clandestine political activity of the Somerset family in the late 16th and 17 centuries.

Shrewsbury, Shropshire

Shrewsbury Castle looking West
Geophysical survey and excavation concentrating on the inner bailey to in particular examine the tail of the north rampart. Despite being one of the main fortresses on the Welsh border no major excavations have taken place in the bailey. Thus the medieval plan of the enclosure, and the missing domestic ranges that should be there, are completely unknown.

Snodhill, Herefordshire

Photo © Jeremy Cunnington

Excavations in this important Welsh border fortress that was in use from just after the conquest to the English Civil War. The aim is to answer some key questions about the castle e.g. the keep’s entrance and final form, to establish the form of the North Tower and along the south side to see if that was where the entrance was.

Tarbert, Argyll

East Loch Tarbert and Tarbert Castle - geograph.org.uk - 1624617
Funding post-excavation costs of a community archaeology project. The project will be trying to see discover a number of things about this royal castle including if there was a southern entrance into the outer bailey, and what buildings there were along the north east range of the inner bailey.

Wressle, East Yorkshire

A geophysical survey of the area to the south of the castle ruins which had been covered by the previous earthwork survey funded by the CST, to get more information about the various garden structures there, as well as other details regarding the deserted village, moat and fishponds. The 14th century castle was one of the most important castles owned by the Dukes of Northumberland.

The applications have been sent to our expert assessors who will go over them. You can see how the assessment process works from our blog back in January 2016. And don’t forget to subscribe to our newsletter.

*The article was updated at 15:28, 10th December to remove Halton Castle.

The Castles that held Magic at Bay

The excitement of Halloween has quickly set upon us at the Castle Studies Trust, and we thought we would explore the connection between witchcraft and castles. Castles have a long history as the walls that confined accused witches; the prisons that kept their magic at bay.

Grab your brooms and cauldrons, we are headed to Leeds Castle in Kent. Leeds Castle was famously purchased by Queen Eleanor of Castile, wife of King Edward I, in 1278. The queen enhanced its defences and possibly commissioned the lake that surrounds the residence. In 1321, the castle saw military action when it was captured by the forces of Edward II from Margaret de Clare, Baroness Badlesmere. The winter of 1381 witnessed Anne of Bohemia, Richard II’s first wife, stay at the castle on her way to her wedding. The castle, thus, had a long history of the seat of female agency and power. It was not until the fifteenth century that the castle was used to enclose and suppress queenly authority.

Joan of Navarre, Queen of England and wife of Henry IV is the only queen of England to be imprisoned for witchcraft. In the autumn of 1419, the duke of Bedford and Henry V’s council reported a case of suspected witchcraft in the highest possible circles. Queen Joan was accused ‘of compassing the death and destruction of our lord the king in the most treasonable and horrible manner that could be devised’.[1] The queen along with her confessor, John Randolf, a friar of Shrewsbury, had according to contemporary chroniclers dabbled in sorcery and necromancy. The royal court and much of the public opinion quickly became ‘feverish with rumours of witchcraft’.[2] Queen Joan was imprisoned for nearly three years, and all her servants and property were taken from her. She was first imprisoned in Pevensey Castle in Sussex and in the last two years of her house arrest she appears to have been kept at Leeds Castle.

Leeds Castle in Kent, photo by Chensiyuan, licensed CC by-SA 4.0

Although imprisoned in the castle, Queen Joan’s surviving household accounts detail the purchase of luxury items, including minever and other furs, tartarin, silk laces, ords, and thread, sindon and Flanders linen.[3] The cash flow in the household accounts has led Alec Myers to conclude that these surely shows that the king did not believe that she had been practicing witchcraft in a treasonous way. The move to accuse and imprison Queen Joan had its obvious political and financial advantages for the king. Indeed, with the accusations of witchcraft the king was able seize all her possessions and revenues. Nevertheless, Queen Joan paid the price of her freedom for the accusation, whether it was false or not. The crime of witchcraft was used – in this particular instance – for the political manoeuvring of powerful men and the castle walls were meant to ensure the enclosure of a powerful woman.

Two of the Pendle witches, tried at Lancaster in 1612, in an illustration from William Harrison Ainsworth’s 1849 novel The Lancashire Witches. From Wiki Commons.

From Leeds Castle, we move to Lancaster Castle; a castle with medieval origins was used as a prison starting in at least the Tudor period. Nearly two centuries after Queen Joan was imprisoned, a castle would again contain the power of witches. In 1612, ten people convicted of witchcraft were held in Lancaster Castle soon to be facing the gallows. Their crimes included laming, causing madness and what was termed ‘simple’ witchcraft as well as sixteen unexplained deaths stretching back decades. Those accused included members of two major families which were headed by older widows, including Elizabeth Southernes and her two children, Elizabeth and James, and Anne Whittle and her daughter Anne Redfearne. Others were dragged into the affair: John and Jane Bulcock (a mother and son) Alice Nutter, Margaret Pearson, and Katherine Hewitt were all also involved in the trial as co-conspirators.

A watercolour by Thomas Hearne from 1778 of the west of Lancaster’s keep. The round tower next to the keep was demolished in 1796. From Wiki Commons.

Five of the ten people were tried at the castle itself with Judge Bromley presiding, accompanied by Judge Altham. The judges were assisted by Lord Gerard and Sir Richard Hoghton. The prosecutor was a former high sheriff of Lancashire, Roger Nowell, and the clerk of the court was Thomas Potts of London. A year later, Thomas Potts published his account of these events in a book entitled: The Wonderfull Discoverie of Witches in the Countie of Lancaster. Potts chose to dedicate his book to Lord Knyvett, the man who had arrested Guy Fawkes in 1605; the Gunpowder Plot still a fresh memory for many across the country. The political and religious atmosphere played a clear role in the prosecutions and convictions of those imprisoned in Lancaster Castle in 1612.

The Wonderfvll Discoverie of Witches in the Covntie of Lancaster by Thomas Potts (1613). From Wiki Commons.

The Pendle witch trials, like the imprisonment of Queen Joan, were bound up in the political, religious, and economic turmoil of the period. The role that the castle played in these persecutions may seem minimal at first glance; however, the power and significance that they held in these situation needs further investigation. The stone walls were thought strong enough to contain the magic that these people were accused of conjuring and that in itself is telling in terms of the force castles held in the minds of society. Castles did not only need to keep people out, but they were also used to keep people – and magic – from escaping.


[1] Rot. Parl., IV. 1186.

[2] Chronicles of London, ed. by C.L. Kingsford (Oxford, 1905), p. 73; Chronicle of London, 1089-1483, ed. by N.H. Nicolas and E. Tyrell (London, 1827), p. 107.

[3] See ‘The Captivity of a Royal Witch’ for the printed household account.

The art of the reconstruction

When walking round ruined castles, it can be difficult to picture how they once were. There may be missing floors, the roofs aren’t the same, and entire buildings might have vanished. Even when a structure appears to survive intact, we are usually missing the interior – how it was decorated, the furniture, even how light changed throughout the day.

Reconstructions are a wonderful tool for bringing castles back to life. This year, the Castle Studies Trust is funding one such reconstruction. Chris Jones-Jenkins is helping us understand how Ruthin Castle would have looked. You might recognise his name from 2015, when he worked on one of our very first projects, the digital model of Holt Castle, or from his various work with Cadw and English Heritage.

We caught up with Chris to learn a bit about what goes into these illustrations and how he got into reconstructions. Older illustrations, like those done by Alan Sorrell decades ago, were done by hand while a lot of reconstructions today are done on a computer. Chris has experience of both, and they are very different processes. When working by hand, you need to start off by choosing your viewpoint and deciding what it is you’re going to show. With a digital reconstruction, you have the option to change the position of where you’re looking from. Chris reconstructs everything in the image and the viewpoint might be the last thing to be settled. With Holt for example, if that had been reconstructed 20 years ago the viewpoint would have been one of the first things decided – what to show, what to discreetly tuck away, and how to balance highlighting everything you want. Instead, as Chris created the whole thing digitally it was possible to use the model to create a flythrough video.

A plan of Ruthin Castle from Thomas Pennant’s 1781 edition of ‘A Tour of Wales’, from the collection of the National Library of Wales. Image in the public domain.

A wealth of information goes into reconstructions. For Holt and Ruthin, any standing remains are taken into account and modelled, while historical evidence such as paintings or documents are used to fill in the gaps where parts of the structure are missing. Chris trained as an architect and early in his career so has an eye for how buildings are put together, complimenting his skill as an artist and years of experience with historic buildings. Will Davies and Sian Rees have been helping with the research into Ruthin Castle and how it appeared, giving Chris the best information possible to work from. There was a survey of the curtain wall, but done to the wrong scale, which kept the team on their toes.

Working with some kinds of source material you get a feel for how useful and accurate it is. For instance, in the 17th century Randle Holme made some useful plans of Holt Castle in pen and ink. He also created a view of Ruthin Castle, but for some reason his work here wasn’t as reliable. The test is that the parts of the castle that still survive weren’t well illustrated by Holme, so his work needs to be used carefully.

A lot of time and effort goes into reconstructions. It can take over six months from start to finish, with drafting, redrafting, tinkering with details, and pausing to research a particular feature or issue. Over this time, the reconstruction grows from a plain model to a textured, colourful building. Reconstructions have an element of educated guesswork, but attention to detail is important. If you misstep and include something not appropriate for the period there are eagle-eyed heritage lovers who spot this kind of thing! New information is being uncovered about Ruthin, which directly informs Chris’ work. But there are some questions, like the forms of buildings, which are only likely to be answered by excavation.

Chris has been working on reconstructions since the 1980s, and aside from the change in technology has noticed an interesting trend. Previously, there was often emphasis on showing the structure while more recently the organisations commissions reconstructions want to populate these buildings, showing daily life inside.

Talking to Chris, one thing that cropped up is that reconstruction artists don’t often get to hear what the public think about their work which is a real shame as it can make a world of difference to a site. So when going round a historic site, if you see a reconstruction that captures your imagination or helps you understand the place better, pass on that positive feedback.

The image at the top of this page shows the a work-in-progress version of the reconstruction. Subscribe to our quarterly newsletter so you don’t miss news about the finished reconstruction and our other projects.

An Introduction to the Battle of Hastings

William the Conqueror, 1066, an arrow in the eye. The Battle of Hastings is one of those events that sticks in the mind. It was a defining moment in English history, and without the Normans we wouldn’t have castles dotted up and down the country. So how much do we know about what led to this point, and what happened at Hastings?

Why did William invade England?

The Bayeux Tapestry showing William with the papal banner

The story of the Norman Conquest and the battle of Hastings goes back before 1066. In 1051, Edward the Confessor promised William, Duke of Normandy, that when he died the Norman would become king of England. Harold Godwineson – a powerful Anglo-Saxon earl – met William and swore would recognise the duke’s claim to the English throne when Edward died. But there was a twist still to come. Edward the Confessor fell ill late in 1065, and on his deathbed made Harold his heir. On hearing the news of Edward’s death and Harold’s coronation, William sent a message to the pope, asking for his permission to invade England and take the crown.

The calm before the storm

Chateau de Falaise in Normandy, where William grew up. Photo by Elisa Pictures, licensed CC by NC-ND 2.0.

Harold had fought alongside William and expected the duke to attempt an invasion. The new king of England raised an army in May 1066 and camped in the south of England, ready to fight. With the pope’s support, William built a fleet of ships for his army but bad weather prevented them from crossing the Channel. This delayed the invasion for so long, that in early September Harold disbanded his own fleet of ships.

Everything happens at once

William landed at Pevensey Bay on the morning of 28 September 1066 and fortified the nearby Roman fort. Photo by Richard Nevell, licensed CC by-SA 2.0.

Harold’s younger brother, Tostig, was in exile and had been raiding England. In September he and Harald Hadrada, king of Norway, landed 300 ships in northeast England to claim the English throne. At the battle of Fulford on 20 September they defeated an Anglo-Saxon army led by the earls of Northumbria and Mercia. Harold marched north, and on 25 September faced the invading army at Stamford Bridge in Yorkshire. Tostig and the Norwegian king were both killed in the battle, ending in victory for Harold.

While Harold’s army was recovering after a bloody battle and a long march, William’s luck changed. The weather turned, and he set sail across the Channel on 27 September, landing at Pevensey Bay in Sussex on the morning of 28 September. William and his Norman,  French, and Breton soldiers set up camp in the Roman fort at Pevensey. They began raiding the local area which happened to belong to King Harold himself. News of William’s arrival reached Harold in York on 1 October. He reacted to the news (and the insult) by marching back south, going via London to collect more soldiers.

The Battle of Hastings begins

Hastings 2006
William had moved his army from Pevensey to Hastings, a few miles away. On the morning of 14 October 1066, Harold marched his army to the ridge now known as Senlac Hill. He planned to fight on the defensive which suited strength of the shield wall. William’s forces arrived from the south. The two armies, both around 7,000 strong lined up for battle. The sound of trumpets marked the outbreak of fighting, and William made the first move by advancing his foot soldiers up the hill with the cavalry following.

Harold’s army holds strong

Battle of Hsatings 950th anniversary (121)
The shield wall was a fearsome obstacle, and hard to break. Harold’s disciplined army resists William’s first assault. As the first wave retreated, a rumour ran through the Norman ranks that William had been killed. Knowing he had to rally his soldiers or lose the battle, William rode in front of his army and lifted his helmet so they could see he lived.

William takes control

Tapisserie de Bayeux – Scène 55 : le duc Guillaume se fait reconnaître.

Seeing the Normans fall back, part of Harold’s army followed them down the hill, attempting to route them. William managed to bolster his soldiers in time to turn and face the attack coming down the hill. Outnumbered and out of formation, Harold’s men had given up the advantage of the shield wall and were defeated.

Seeing how effective this was in reducing the strength of Harold’s army, William pretended to retreat another two times. The ruse worked each time, giving William the advantage.

Harold’s defeat at Hastings

Battle Abbey Gatehouse
In the midst of the hard fought battle, King Harold was killed. The Bayeux tapestry seems to show that he was hit in the eye by an arrow, while contemporary chroniclers suggest he was hacked down in the fighting. Either way, with their leader and his brothers dead, the Anglo-Saxon army broke.

Afterwards, William marched through southeast England, capturing important towns before arriving in London. He was crowned king of England on Christmas Day 1066 in Westminster Abbey. William ordered the construction of Battle Abbey on the field where he defeated Harold, with the high altar over where Harold fell.

To secure the conquest, William and his supporters built castles across the country, especially in important towns like Lincoln and Norwich to control the area. The Norman Conquest led to the age of the castle in England.

Make your own scene in the style of the Bayeux Tapestry and share it with us on Twitter. If you want more on Hastings and its impact you can read Peter Purton’s blog post or the Oxford Dictionary of National Biography’s entries on William and Harold which were used when writing this piece.

Trial trenching at Pembroke Castle

In 2016, geophysical survey by Neil Ludlow and Dyfed Archaeological Trust (DAT) revealed the remains of long-vanished buildings, and other features, at Pembroke Castle. This work would not have been possible without the Castle Studies Trust, which funded the entire project.

Pembroke is famous for its large round keep, built by earl William Marshal, and a complex of other stone buildings is preserved in the inner ward. But the large, outer ward has been an empty space since at least the eighteenth century. Nevertheless, very dry summers had revealed ‘parchmarks’ here, showing that ruined walls lay just below the surface. These showed up particularly well during 2013 when they were photographed by Toby Driver of the Royal Commission on Ancient and Historical Monuments (Wales), and were published by Toby, with Neil Ludlow, in Archaeology in Wales.

This was the genesis of the current programme of investigation at the castle. The geophysics that followed used a combination of magnetometry, resistivity and Ground Penetrating Radar, to confirm the parchmark evidence and also show that a number of further buildings, and other features, formerly lay in the outer ward. The most exciting of these is a large building which may represent a free-standing, double-winged mansion-house from the late Middle Ages. If so, it would be quite exceptional: grand buildings of this particular kind are unusual in castles, and particularly in outer wards which are normally thought to have contained the more lowly buildings associated with everyday castle life.  It means that the outer ward, at Pembroke, may have been ‘gentrified’ – at least in the fifteenth century. This may make sense of historical accounts which place the birth of the future king Henry VII in the outer ward: it may have occurred within this very building.

What the geophysics and parchmarks seem to show is a large, central hall, flanked by two wings. If the building was anything like others of its kind, one wing will have contained the kitchen and ‘services’ – the buttery and pantry – while the other represented more private accommodation. The outer ward was partly excavated in the 1930s – sadly, without record – but a photograph shows a large, stone-lined pit, apparently leading off the hall, which may be the cess-pit mentioned in an account from the period. A square annexe showed up here in the survey work.

Just inside the castle beside Pembroke’s main gate are parchmarks (where the white tents are). They show the position of a late medieval hall. ©Neil Ludlow

While the results showed what can be achieved through geophysics – and generous grant-aid – their interpretation is, at the moment, still speculation. Only educated guesses can be made about the exact form, date and function of the building. But it has the potential to make a big contribution to the study of castles, and late medieval history, at a national level. So Neil got together with James Meek of DAT to decide how best to get answers to these questions. Only excavation could really provide the answers but, given the sensitivity of the site, which is a Scheduled Ancient Monument, and the importance of the building, minimum intervention, for now, was the preferred approach – a small excavation, primarily just to see how much has survived. This was agreed by Cadw, who granted permission for the hand-excavation of two trial trenches, one across the hall and possible cess-pit, the other across one of the wings. Overburden will be removed, and hopefully the remains of the medieval walls, and any floor-surfaces, will be revealed. It is not intended to go down any further – or at least, not in the present project which will effectively be an evaluation.

We also don’t know how much was removed during the 1930s excavation. While we know there must still be walling, as it shows up in the geophysics and parchmarks, it is possible that the medieval floors may have gone. We hope not, and we also hope that the cess-pit wasn’t emptied, as its deposits could contain valuable information on diet, health, bugs and parasites, while finds that are important to us may have been dumped there as rubbish.

Neil and DAT decided once again to approach the Castle Studies Trust who, very generously, again awarded a grant for the work. This will be supported by help in kind from Pembroke Castle Trust. One of the things also missing at Pembroke is a really accurate measured ground plan, with levels, so for the first time a full topographic survey will also be undertaken. We hope the castle can be persuaded to reveal a few more of its secrets.

Possible layout of the winged building in Pembroke’s outer ward,
based on Cothay Manor, Somerset, and others. ©Neil Ludlow

Don’t miss out on more news from our projects. We’ll have regular updates from Pembroke on our social media, and if you haven’t already please subscribe to our quarterly newsletter. We are entirely reliant on donations, so please consider giving to support our work.

How to build your castle

Review of: Charles Phillips, The Medieval Castle. Design, construction, daily life.  Haynes Publishing, 2018.  £22.99.

In the heart of a forest in Burgundy (France) a thirteenth century castle has been rising steadily from the ground since 1997. Guédelon castle was the subject of the BBC series Secrets of the castle in 2015 and, twenty years after construction started, is near completion. This new book describes what was involved in the challenge of creating a medieval castle using the Guédelon experience and relating this to the development of medieval castles in Britain.

The book is beautifully illustrated and succeeds brilliantly in demonstrating medieval building techniques, how the challenges medieval builders faced were overcome with the tools and technology at their disposal, how critical differences in types of soil, stone and timber determined what was used – in short, the great skill acquired by medieval craftsmen and now re-learnt by their modern successors (who operate under professional archaeological supervision). The tools used, the equipment created – such as enormous treadwheel cranes – how stones were cut and carved, mortar manufactured, vaults and windows installed, floors and roofs prepared and put in place: all are superbly explained in word and picture. Plumbing and heating were essential to survival and all are covered here. The daily life of those living in and around a medieval castle is superbly recreated by those working on the project (along with illustrations from medieval manuscripts, many of which will be well known to anyone already familiar with the subject).

One of the workshops where medieval techniques are used. Photo by Jean-Christophe Bruneau, Licensed CC-BY.

Working inwards from the ditch via gate, walls and bailey to the great tower and interior buildings (hall, chapel, chamber etc), Phillips takes the reader through the development of castles in England and Wales (with a rare skip across the Channel to France) between 1066 and the fifteenth century.  His descriptions are largely of well-known major castles (Dover, Tower of London, Chepstow, Edwardian castles in Wales) which are presented with clear plans and excellent photographs. But it is here that I have a problem.

Though the author states that castles were bases and symbols of power and status, he otherwise hardly acknowledges more than twenty years of castle studies that have turned traditional explanations on their head: everything, he states time and again, was done that way for military reasons. While the great tower (keep, donjon) was occasionally used as a last resort defence, it was not built for that purpose. The cross-wall in the keep at Rochester (for example) served as a defence in the final stages of King John’s siege of 1215, but that was not why it was put there. Round towers had (some) defensives advantages over rectangular towers, but that wasn’t the only reason they become fashionable, and even then mainly only in France and England. I could go on.

Despite this weakness, Phillips’ book presents an unequalled description of how thirteenth century castles were designed and erected. Guédelon was based on a style made popular by King Philip Augustus of France. If you can get to visit it, do (though you will compete with thousands of tourists and school visits). If you can’t, this book will tell you all you need to know about it.

Peter Purton (D.Phil, FSA).

Medieval Engineers: history’s forgotten builders

“King Henry II built the great tower at Dover Castle” is the kind of statement you will hear when visiting one of this country’s magnificent fortresses. But the King himself never lifted a single tool to get any castle built. While the hard manual work was done by labourers, and the finer details worked by master stone masons and carpenters hired because of their great skills, the general plan as well as the day to day running of the construction would have been overseen by an engineer-architect. Under Henry II, records survive telling us who they were (in this case, Maurice the engineer) but in most instances they are anonymous.

Dover’s great tower (the tallest building) was constructed by Maurice the engineer for Henry II. Photo by Mark Whibley, licensed CC BY-NC-ND.

This anonymity is not a surprise: the writers of medieval chronicles were interested only in the great who ruled society – kings, bishops, great lords. It was this gap that persuaded me to write my book The Medieval military engineer. From the Roman Empire to the sixteenth century (Boydell Press 2018).

Today, sappers and engineers form a key part of every state’s army, and was also true of imperial Rome. But in medieval times craftsmen were hired to carry out engineering roles and quite often the same people would have many skills, so that alongside building castles they might also design bridges, churches and cathedrals, or oversee the creation of, and sometimes operate, siege weapons. Because they were commoners, and with only a handful of exceptions, their names were not recorded throughout early medieval times. Often we only know them when (like Maurice) records start surviving showing what they were paid for their work. The great lords who also had hands-on military engineering skills were named in history, but were a tiny handful.

Many interesting questions become easier to answer as more records survive, such as how much did these engineers actually know, how they learnt their skills, how knowledge was transmitted across generations, and what part did they play as technology became more sophisticated?

The same skills that St Guthlac used to design and build a chapel (shown above) could be used to build castles.building a chapel. From the British Library’s ‘Guthlac roll’, made in the late 12th or early 13th century. Harley Roll Y.6.

Historians no longer see the years between the end of the western Roman empire and the European renaissance of the fifteenth century as one long period of ignorance, and we are more aware that change and improvement were continuous, witnessed with the development of ever more spectacular cathedrals and castles, but also mundane but vital skills such as bridge and ship building, irrigation schemes, and military equipment such as siege artillery – the trebuchet, taken up across Europe and the Islamic lands during the thirteenth century, was a game-changer, before giving way to gunpowder artillery during the fourteenth.

Change happened because people questioned existing conventions and came up with new ideas, but also because they developed the skills to put them into practice. It is time to give them the credit they are entitled to. Next time you visit a large stone castle, ask not just which lord lived there and paid for it, but who actually designed it, and admire their skills; and if despite its strength it was captured in a siege, who built the wooden engines that do not survive, or who undermined it, which decided the outcome?

Peter Purton, DPhil (Oxon), FSA.