Castles
in a Hurry

The Whole Thing in One Page

The castle in your head is a stone object: towers, battlements and a drawbridge. Its natural condition is siege. Between attacks, it waits.

A medieval castle was less tidy and more useful. It was a defended residence or headquarters for people exercising control over land, routes, rents, justice and force. Some were royal centres. Some belonged to military orders. Some were small timber enclosures. Many were rebuilt so often that no single date describes them. The wall was one part of a working institution.

That institution converted resources into controlled access. A steep site slowed approach. A ditch exposed anyone crossing it. Walls raised defenders above attackers. Towers let them watch and shoot along the wall face. Gatehouses forced movement through narrow, observed spaces. Inner wards preserved another position if an outer line failed. Stored grain, protected water, loyal people and hope of relief turned masonry into time. Without them, a high wall was an expensive trap.

Building one began before the first stone. A patron needed authority over the site, money or obligations to command labour, timber for scaffolds and roofs, earth or stone for mass, lime for mortar, iron for tools and fittings, carts, animals, boats, skilled masons, carpenters, smiths and many labourers. Timber could be raised quickly, repaired and altered. Stone resisted fire and weather better and announced wealth, but demanded quarries, transport, kilns, seasons and maintenance. The familiar sequence from wood to stone was never universal. Builders used what the ground, threat, budget and timetable allowed.

Attackers faced the same institution from the other side. They might blockade roads, bargain with the garrison, surprise a gate, climb at night, batter masonry, shoot into the enclosure, undermine a tower or wait for hunger, thirst, disease and fear to work. Direct assault could be necessary and was often murderous. Many sieges ended through surrender, betrayal or relief. The decisive question was seldom whether a wall could be broken in theory. It was whether the attackers could sustain pressure longer than the defenders could sustain resistance.

Inside, the castle remained a household. Food arrived, accounts were kept, bread baked, ale brewed, horses shod, fires fed, rooms cleaned, guests ranked, petitions heard and children raised. The great hall displayed hierarchy while kitchens, wells, stores, stables, chapels and latrines kept the place functioning. Women of the ruling household could manage estates and domestic administration or command during an owner's absence. Servants, craftspeople, clerks and soldiers supplied most of the daily motion.

Castles changed because the balance around them changed. Larger rulers could finance longer operations, heavier artillery and permanent administration. Guns did not erase castles overnight, but they raised the price of resisting the strongest armies. Some walls were lowered or thickened, some residences expanded, some sites became forts, prisons or offices, and many were abandoned or deliberately damaged. Weak government, disputed frontiers and civil war could make old strongholds useful again. There was no single end date.

A castle was therefore neither a home with decorative battlements nor a war machine containing inconvenient bedrooms. It was an attempt to make authority durable by controlling movement, supplies and time. Its towers are the surviving shell. The real castle was the organised human world inside and around them.

That is the book.

Why You Should Care

Stand outside a ruined castle and look away from it. The first lesson is usually behind you.

A road bends towards a gate. A river narrows below. A town presses against one wall while open country spreads from another. A ridge exposes anyone approaching. A harbour or landing place lies within reach. The picturesque view was once a map of movement. The castle occupied the place where traffic, food, information, law and violence could be watched or interrupted.

That changes what you see. Towers stop being a collection of medieval shapes and become answers to practical questions. Where can an attacker stand unseen? Which wall can be reached by a ladder? Can a cart turn before the gate? Where does rainwater go? If the well fails, how long can the garrison last? Which door separates a public hall from private chambers? Why does the kitchen sit where smoke, fire and delivery can be managed? A ruin becomes legible when you ask what had to move through it.

Castles also make power unusually concrete. Medieval lordship could be dispersed through rights, rents, oaths, courts and land. A castle gathered those claims into one visible place. It stored records and food, housed armed followers, received tenants, imprisoned opponents and gave an owner somewhere difficult to remove him from. To the household inside, it could mean shelter, status and employment. To people outside, it could mean a court, market and protection, or tolls, labour demands and a garrison able to enforce both. The same gate could welcome a feast and close against a village.

Their construction reveals how large projects worked before modern machinery. A major site became a temporary town of quarrymen, diggers, carpenters, lime burners, smiths, carters, boat crews, accountants and supervisors. Stone walls advertise the patron, but the achievement belongs to organised supply. Harlech employed nearly 950 workers during the busiest summer of Edward I's campaign, a seasonal peak rather than a normal medieval workforce. At Conwy, the castle and town walls rose in about four and a half years because royal government concentrated money, material, transport and skill on an occupied landscape. Speed was administrative before it was architectural.

Sieges correct another instinct. We remember impact: ladders rising, stones falling, a breach opening. Much of the work happened before or around that moment. Attackers gathered intelligence, cut communications, protected their camp, found food, built engines, dug and tried to prevent relief. Defenders rationed, repaired, watched, negotiated and judged loyalty. Both sides depended on people and supplies under threat. A wall increased the cost of attack. It never cancelled weather, disease, politics or hunger.

Then there is life inside. Empty ruins magnify stone and erase timber, plaster, colour, furniture, smell and people. Conwy's exterior was rendered, so it could appear pale rather than the bare grey seen now. Chambers had shutters, hangings, beds, chests and fires when status and circumstance allowed. Great households moved between residences, carrying plate, textiles, officials and servants. Local staff kept the castle functioning between visits. They often knew it more continuously than the lord who owned it.

This book concentrates on the medieval European castle, especially the unusually well-documented traditions of Britain and France, while using boundary cases from neighbouring regions where they change the model. It does not treat every fortified palace, citadel, walled town or stronghold as the same institution. Nor does one British ruin represent a continent. Designs, legal categories and social arrangements varied, and the surviving sample has been altered by demolition, restoration and tourism.

Those limits sharpen the subject. The question is how a fortified household or headquarters turned terrain, labour and authority into control, how opponents tried to reverse that conversion, and what daily life cost the people who kept it working.

The Core Ideas

A Castle Was a Claim Made Physical

A castle was a building with an argument attached: this place, its approaches and the land around it answer to someone who expects to remain.

The argument separates a castle from neighbouring forms. A fort may house troops without being an elite residence. A palace may house a ruler without serious fortification. A walled town defends a community larger than one household. Medieval words did not police these categories neatly, and modern labels travel badly across languages. The useful working definition is a fortified high-status residence or headquarters tied to lordship, government or command.

That definition has boundary cases. Crac des Chevaliers in the eastern Mediterranean was held and rebuilt by the Hospitaller military order rather than inherited as one family's home. Royal castles might be occupied mainly by officers. Some small lordly sites offered little comfort. These cases do not empty the category. They show that residence could belong to an institution as well as a dynasty, and that castle life ranged from household government to permanent garrison.

Fortified settlements and defended residences long predated castles. Roman forts, hillforts, town walls and royal enclosures supplied techniques and reusable sites. What became common in parts of western Europe from the tenth and eleventh centuries was a more local concentration: aristocrats, rulers and corporate elites building strong residences from which they could hold land, followers and rights in a divided political order. The castle belonged to a landscape of powers smaller than an empire and stronger than a village.

The Norman conquest of England makes the mechanism visible. A few castles existed before 1066, but conquest multiplied them. Earth-and-timber works rose near towns, road junctions, river crossings and existing centres of administration. At some sites houses were destroyed and streets reordered. The castle protected incoming lords while inserting a new strongpoint into an occupied community. Security for one group meant displacement for another.

Motte-and-bailey is the best-known form. A motte was a steep artificial or adapted mound topped by a tower or defended summit. The bailey was an enclosed working court containing buildings, animals, stores and people. A ringwork achieved a related result with an embanked enclosure rather than a dominant mound. These were not crude sketches awaiting stone. Earth absorbed impact, timber could be worked fast, and height changed the geometry of approach.

The building extended its owner's reach in peace. A gate controlled entry. A hall received guests, tenants and officials. A chamber protected the household, seal and valuables. Stores supported residents and soldiers. A chapel placed rule within a sacred order. Courts, prisons and accounts turned lordship into local experience. A settlement might grow beside the walls because people sought work, trade and protection, while the owner gained tolls and tenants. The relationship was unequal. A castle could shelter a district from raiders and give the same lord the means to extract rent or punish resistance. Protection and coercion were capacities of one institution.

Appearance worked before battle. Height, banners, dressed stone, painted surfaces and elaborate gates announced resources. This did not make display the opposite of defence. A patron who could raise a tower was advertising the capacity to command labour and armed people. Others planned around the possibility that it could be held.

Castle power still depended on relationships. A king might demand custody, install a constable, order demolition or require surrender during war. English licences to crenellate were neither universal permits nor a complete register, and their issue could express favour or negotiation as much as military control. The owner also needed the surrounding economy. Roads brought tenants and produce. Mills, woods, fisheries and fields supplied revenue. A market or borough could enlarge income while the castle protected and supervised it. The strongpoint was local because its resources were local.

The first mental shift is therefore from object to jurisdiction. Ask who entered, who commanded, who supplied the place, whose disputes were heard there and whose movement it constrained. Battlements were the visible edge of a larger claim.

Defence Bought Time

No builder could make a castle impossible to take. The practical aim was to make capture slow, costly, exposed and uncertain enough that the attack failed, relief arrived or negotiation became attractive.

The site supplied the first layer. A steep hill tired climbers and limited positions for engines. A river or marsh blocked some approaches and carried supplies along others. Coastal access could keep a garrison connected after roads were cut. Clear ground denied concealment. The site also imposed weaknesses: poor water, weak rock, cramped summits, mineable slopes or nearby heights from which later artillery could fire down. Terrain offered a bargain, not free strength.

Earthworks reshaped it. A ditch forced attackers to descend and climb under observation. Spoil could form a bank. A motte created elevation with labour rather than geology. A moat might be wet, dry, tidal or unreliable. Its main effect was broken access. An attacker carrying a ladder, shield or ram crossed awkward ground while defenders worked from above.

Walls turned height into separation. Thickness helped against impact and mining. Height complicated ladders, though excessive height increased cost and risk of collapse. A wall-walk moved defenders around the circuit. Battlements supplied cover. Timber hoardings or stone machicolations sometimes extended positions beyond the face. Arrow loops allowed shooting from relative safety but restricted vision. Every protection imposed a limit.

Projecting towers corrected blind ground at the wall base. They let defenders observe and shoot along adjacent curtains. Round or polygonal forms removed some vulnerable corners, but no shape solved every problem. Towers also contained stairs, chambers, stores and latrines. A defensive projection was expensive floor space.

The gate remained necessary and dangerous. It had to pass carts, horses, servants and guests, then resist concentrated force. Builders narrowed, turned and layered the route with bridges, doors, bars, portcullises and flanking towers. Yet an elaborate gatehouse gathered people and machinery around the opening the enemy wanted. It could fail through fire, surprise or betrayal because it was busy.

Depth preserved options. An outer ward might fall while an inner circuit held. A keep could offer another position, although many keeps were prestigious residences rather than routine last refuges. Concentric layouts nested defensive lines, sometimes allowing the inner wall to command the outer. Many castles acquired such layers through rebuilding rather than arriving as perfect diagrams. Defenders could close doors behind a retreat, move along protected passages and use towers as partly independent positions. They could also make sorties, fire the enemy's works or attack a mine. Walls gave a smaller force positions from which to act.

People completed the geometry. A small peacetime staff could not man every loop and tower. In crisis the owner needed warning, reinforcements, food, water, weapons and authority. Defenders watched at night, repaired damage, controlled fire and guarded against treachery. Too many refugees consumed stores; too few soldiers left gaps. Water was among the hardest reserves to improvise. Wells, cisterns and protected routes mattered because grain could be rationed while thirst closed the timetable quickly. Fire was the other permanent threat. Timber roofs, doors, floors and stores made the stone shell combustible from within.

Defence was accumulated delay. Each ditch, gate, tower, well and barrel imposed another task or bought another day. Their value depended on what time could bring: relief, winter, political settlement, the attacker's bankruptcy or the defender's own starvation. The castle did not cancel capture. It charged the attacker for every hour.

Stone Began as a Supply Chain

Before a castle could resist an enemy, its builders had to overcome distance, weight, weather, money and the site.

The first decisions were political. Who owned the ground? Could houses or fields be taken? Which road, port, river or settlement would support the work? Was speed more valuable than permanence? A conquest castle could begin with ditch, bank and palisade. A royal residence might develop for decades. A town castle had to fit around streets, walls and property claims. The plan began as a negotiation with power.

Surveying translated purpose into ground. Builders set out ditches, walls and levels with cords, rods, plumb lines, squares and sighting methods. Master masons used dimensions, templates and experience, adjusting to rock and older fabric. Few complete working drawings survive, so the route from a patron's wish to built form is reconstructed from accounts, marks, archaeology and the structure itself. Scarcity of drawings does not mean absence of design.

Earth supplied immediate mass. Diggers cut ditches and moved soil in baskets, carts and barrows. Timber formed palisades, gates, bridges, floors, roofs, scaffolding, cranes and the centring beneath arches. It decayed and burned, but heavy carpentry could be formidable. Timber remained indispensable after walls became stone.

Stone enlarged the chain. Quarrymen exposed beds, split blocks and rough-shaped them. Local rubble filled cores while better stone formed faces, corners, arches and stairs. Material could come from the site, nearby quarries, dismantled Roman buildings or distant sources chosen for quality. Boats moved bulk efficiently. Carts punished roads and animals. A wall preserved a map of transport.

Mortar required limestone or chalk, fuel, kilns, water and sand. Quicklime had to be slaked and mixed with care. Masons raised faces and core together on timber scaffolds, tying the two skins through a packed interior rather than pouring one uniform mass. Openings, stairs and drains had to be formed as the wall rose. Carpenters made roofs and lifting devices. Smiths repaired tools and forged fittings. Plumbers, tilers, glaziers, painters and plasterers finished prestigious spaces. Labourers dug, carried, mixed and lifted. Clerks recorded wages and purchases.

The site became a temporary settlement. Workers needed food, drink, shelter, tools and pay. Managers scheduled deliveries so specialists did not wait beside empty scaffolds. Frost, rain, harvest demand, war and cash shortage interrupted work. Construction proceeded in campaigns, leaving usable sections and temporary closures. Maintenance began before completion because roofs leaked, drains blocked and earth settled.

Labour was not organised by one medieval rule. Major royal accounts reveal paid skilled workers, day labourers, contractors and supply purchases. Other projects used customary obligations, requisitioned carts, military service or coercive demands on local populations. Surviving wage rolls illuminate the labour they counted and understate people whose work reached the site through estates, households and compelled service. Neither a modern free labour market nor a single mass of enslaved peasants describes the whole picture.

Edward I's programme in north Wales shows both concentration and limit. At Harlech, nearly 950 workers were recorded during the busiest summer of 1286: 227 masons, 115 quarrymen, 30 smiths, 22 carpenters and 546 labourers. This was a seasonal peak within an exceptional royal conquest programme, not a typical castle crew. The castle and town walls at Conwy rose chiefly between 1283 and 1287. Surviving accounts do not separate their costs, so the recorded expenditure of about £15,000 belongs to the integrated project. Water access helped supply both. Beaumaris, begun in 1295, remained incomplete when money and war priorities shifted. The most ordered plan still answered to a treasury.

Stone did not replace timber in one upgrade. Some early castles used stone from the outset; some timber defences endured. Stone buildings kept wooden roofs, floors, doors and galleries. Choice followed material, speed, threat and meaning.

The finished wall conceals the movement that made it. Before stone could control a landscape, a political organisation had to make hundreds of people, animals and loads arrive in the right order.

A Siege Was a Contest of Systems

The dramatic image begins when a stone flies. The siege began when an army decided to stay.

That decision created a second settlement outside the walls. Attackers needed camps, guards, food, fodder, tools, timber, smiths, engineers, scouts and routes home. They had to prevent sorties, watch for relief and limit disease. An army large enough to surround a castle could exhaust the countryside faster than the garrison consumed its stores. Time injured both sides.

A castle could be bypassed if it did not control a vital road, port or political symbol. Yet a hostile garrison behind an army threatened wagons, messages and legitimacy. Attackers often negotiated. A garrison might surrender after a fixed period without relief. Defenders could receive safe conduct or face harsh treatment depending on status, politics and the conventions of that war. A negotiated gate spared lives and preserved the place.

Surprise shortened the contest. Spies studied routines. Bribery reached gatekeepers. Climbers used an overlooked wall, postern or drain. The more complicated the entrance, the more people had to operate it correctly. Repeated behaviour could be studied.

Escalade brought ladders or ropes to the wall while missile troops suppressed defenders. Ditches broke formations, height exposed climbers and the top narrowed movement. Success required speed, numbers, distraction or weak defence. It was possible and dangerous, not the natural opening move.

Rams, picks and stone-throwing engines attacked gates, roofs, parapets and masonry. A counterweight trebuchet could throw heavy projectiles, but range, rate and accuracy varied with machine, crew, ammunition and target. Bombardment might create a breach, start fires or persuade the garrison that worse was coming. It did not erase thick walls on command.

Mining attacked below. Workers dug beneath a wall or tower, propped the cavity with timber and burned the supports. Defenders listened, dug countermines, flooded workings or fought underground. At Rochester in 1215, King John's forces mined a corner of the keep and used fat from forty pigs in the burning operation, according to a royal order. The corner collapsed. Internal masonry preserved another position, and hunger finished the siege.

Blockade attacked the castle's connections. Roads and rivers were watched, external mills and wells denied, messages intercepted. Defenders rationed and searched for reserves. Hunger weakened judgement, spread illness and turned rank into allocation. At Kenilworth in 1266, extensive water defences and a large garrison supported resistance for 172 days. The royal host drew supplies from a wide area and assembled engines. Disease, dwindling food and negotiation ended the stand-off. The count marks the documented span from 25 June to 13 December; the forms and intensity of action changed across those 172 days.

Surrender terms mattered because both sides expected future wars. A garrison that yielded after accepted resistance might leave safely or preserve property. One taken by storm risked harsher treatment. Practice was inconsistent and often brutal, but the possibility made bargaining part of military technique.

Relief changed the value of another day. A garrison expecting allies might endure damage; the same people might surrender a sound castle once hope vanished. Information could matter as much as grain. Capture did not finish the calculation. The victor had to decide whether to repair, occupy, ransom, return or slight the place. Deliberate damage concentrated on towers, gates or wall sections whose loss made renewed defence expensive, though later quarrying and decay can resemble military destruction.

Masonry shaped the options. Supply, intelligence, labour, morale, politics and time decided which option held. An attacker could break a wall and lose the campaign if disease, debt or relief broke the camp first. Defenders could preserve every tower and still surrender because food, loyalty or political purpose had gone.

The Household Made the Fortress Work

A castle at peace was a yard, office, lodging house, workshop, stable, court and high-status home sharing one address.

The great hall gave the household a public centre. Meals, ceremonies, business and judgement happened there. Seating made rank visible. Servants carried food, clerks presented accounts, petitioners waited and guests watched the owner's resources perform. The hall impressed because it organised people, not because it was empty.

Privacy increased by degrees. Chambers held beds, chests, textiles, documents and valuables. Access through a sequence of rooms made proximity a privilege. Separate suites could serve household members. Keeps, gatehouses and ranges all contained accommodation. The idea that everyone slept in one smoky hall mistakes one public room for the whole plan.

Great households travelled between estates because supplies were easier to consume near their source and political presence had to be distributed. Bedding, plate, clothing, records, chapel equipment and servants moved with them. A smaller staff under a constable kept the castle between visits. At such itinerant residences, the owner could be less permanent than the porter.

Goodrich makes this rhythm visible through the accounts of Joan de Valence. Between May 1296 and September 1297 she stayed at fifteen residences, while her travelling household varied between roughly 122 and 196 people. At Goodrich in late 1296 she had around 120 servants with her. Bakers went ahead to prepare bread, officers managed transport and permanent staff maintained custody. When she arrived in November, the accounts record eighty-one pigs killed for the stay. These numbers describe one wealthy household during recorded visits, not average castle life. Their value is that they put people, appetite and movement back into rooms now empty.

Kitchens needed hearths, ovens, water, drains, tables and stores. Their fire risk affected location. Bakehouses, brewhouses, gardens, fishponds and dovecotes might sit within the castle or wider estate. Grain, meat, fish, cheese, ale, wine and spices arrived through purchase, rent and household production. A feast was the end of transport and accounting.

Water came through wells, cisterns, springs, pipes, gutters and vessels. It served people, animals, cooking, cleaning and building. Deep wells were expensive and did not guarantee abundance. Wastewater and rain needed drains. Latrines emptied through shafts, pits, ditches or channels according to site. They offered comfort to some and cleaning work to others.

Occupied rooms could have hearths or fireplaces, shutters, glazing, hangings, beds and timber fittings. Service spaces and wall-walks remained colder. Candles and fuel cost money. The experience ranged from royal chamber to wet night watch.

Noble women could manage estates and households, receive tenants and guests, supervise patronage and command in an owner's absence. Their authority remained constrained by inheritance, law and male kin. Children grew inside networks of nurses, tutors, chaplains, servants and relatives. Stewards, chamberlains, cooks, grooms, porters, clerks, guards and craftspeople supplied daily motion, and many lived beyond the wall. Hierarchy organised labour through access. The porter knew who could pass. Chamber staff controlled private rooms. The steward translated estates into expenditure. The constable held custody for an absent owner. These offices made the household governable and created points at which incompetence, theft or divided loyalty could damage it.

War changed the population. Soldiers and refugees crowded rooms. Stores replaced hospitality. Livestock entered. Privacy contracted and sanitation worsened. The household's knowledge of keys, wells, routes and supplies became military capacity.

Domestic and military histories were one system. Halls built loyalty, estates supplied food, servants carried messages and marriage created allies. Bare walls could not improvise that organisation after an enemy arrived.

Design Had No Final Answer

There was no perfect castle plan because every solution created another target, and every patron wanted more than defence.

Earth-and-timber castles answered speed. A motte gave height and a bailey enclosed work. Stone replacement could preserve the site while altering it. A shell keep ringed a summit. A rectangular great tower concentrated residence, storage and command. Massive keeps projected authority, but corners, limited flanking and crowded circulation encouraged other arrangements.

Curtain walls and projecting towers distributed defence around an enclosure. Round towers removed an external right angle and widened some views. Gatehouses became larger and more layered. Domestic ranges spread around wards. Existing earthworks, rock, property boundaries and older buildings constrained each addition, which is why accumulated castles rarely resemble clean diagrams.

Dover joined strategy and royal theatre. Henry II's late twelfth-century works occupied the height above the Channel and centred on a great tower with ceremonial and domestic rooms. Inner and outer circuits deepened defence. The castle guarded a route into England while receiving high-status visitors. Later tunnels and gun positions answered later threats.

Rochester showed both resilience and limit. Internal masonry preserved defensible space after one corner collapsed in 1215, yet the tower could still be isolated and starved. Later builders placed increasing weight on defended circuits, flanking towers and complex gates without abandoning keeps.

Across the eastern Mediterranean, Crac des Chevaliers records another institutional setting and another sequence. The Hospitaller Order rebuilt it from the mid twelfth century, and Mamluk rulers altered it after taking the site in 1271. Its surviving phases show that fortification developed through conflict and exchange among Byzantine, Frankish, Islamic and local traditions. It is a poor template for an English baron's residence and a strong warning against treating England's labels as European law.

In thirteenth-century Wales, Caerphilly and Edward I's later castles developed layered wall and water defences on a large scale. Conwy, Harlech, Caernarfon and Beaumaris joined enclosures, towers, gatehouses, apartments and waterborne supply. Beaumaris is celebrated for nested circuits and a tidal dock, but remained incomplete. Its ordered plan should be read beside the finances that interrupted it.

Techniques moved through war, pilgrimage, trade, dynastic travel and specialist employment. Similar problems could also produce similar answers without direct copying. Claims that one region invented concentric defence and another imported it whole exceed the evidence. The useful question is narrower: which feature appears after contact, and did local builders reproduce its mechanism or only a familiar shape?

Display remained practical. Bodiam's moat, towers and composed approach advertised status and controlled ordinary access. It did not need to defeat a royal siege train to deter lesser violence. Symbolic power was an effect patrons purchased. Comfort also changed plans. Private chambers, larger windows, gardens and ceremonial routes could expand where threat allowed. They did not announce one general transition from fortress to house. The same patron might maintain a hard border stronghold and a more comfortable inland seat. Design answered the expected enemy and the life intended between alarms.

Gunpowder widened the argument. Early guns were awkward and castles mounted them defensively. Larger bombards later threatened high masonry. Builders thickened walls, lowered profiles, added earth and made artillery platforms. These changes moved fortification towards the low gunpowder fortress.

The best design was conditional on enemy, garrison, water, money, politics, relief and intended life. A plan optimised for every threat would have been unaffordable and unpleasant. Castle architecture records choices under uncertainty, then preserves them after the uncertainty changed.

The Balance Changed When Power Scaled Up

Castles flourished where authority could be concentrated locally. They changed when rulers, cities and armies learned to concentrate resources over wider areas.

A small lord's castle raised the price of removal. A rival needed people, time, engines and political justification. Monarchs who depended on local lords had to bargain with their strongholds, demand custody or build royal castles. Kings granted castles and nobles served as constables, so this was not a permanent war between crown and aristocracy. The tension was comparative: one household could become harder to govern from elsewhere.

Rulers answered with networks. Royal castles held ports, towns and borders. Constables, garrisons, ships, roads and records connected them to taxation and law. Edward I's Welsh programme joined castles to walled settler towns and sea supply. Local strength came from a larger conquest system.

Castles also helped power scale. Their clerks, stores, courts and officers made local resources more legible and available. A royal stronghold could collect revenue, hold archives, receive officials and move supplies. The institution that protected one local authority could be absorbed into a network able to overcome private strongholds elsewhere. Centralisation was partly built through castles, not merely against them.

The same capacity altered siege. Governments became better at raising cash, requisitioning food, hiring specialists and moving artillery. A private castle that could defy a neighbour might not resist a ruler able to sustain pressure for months. Heavy guns widened the imbalance when trained crews, powder, shot and transport arrived together. A bombard stranded without animals or ammunition was less useful than an older engine with a working crew. Once rulers could move several guns, replace losses and maintain a camp, tall local walls faced an opponent operating at another scale.

Owners adapted. They added gunports, earthworks and thicker walls. Some castles became artillery forts. Others gained apartments, windows and gardens while retaining security against disorder. The word persisted across buildings whose military balance had shifted.

Old sites still mattered. During the Wars of the Roses, several important castles changed hands through allegiance, custody or short operations rather than prolonged formal siege. Seventeenth-century civil war then put many old strongholds back into service. Garrisons repaired walls and mounted cannon. After capture, deliberate slighting reduced future use. A castle could be outdated against the strongest artillery and dangerous in a local war.

At many sites, administrative and residential uses survived longer. Castles became courts, prisons, stores and offices. Gatehouses remained occupied after ranges decayed. Roof lead, timber and stone were sold or reused. Abandonment was an economic decision repeated over years. From the eighteenth century, ruins acquired another use. Artists and tourists treated broken towers as evidence of romance, ancestry or national character. Owners cleared, restored and sometimes rebuilt them according to later taste. The castle's image survived by shedding much of the household labour and local coercion that had made the institution work.

No smooth European story runs from private castle to modern state. Some rulers were strong early, some frontier societies kept local fortification useful, cities and military orders built on other foundations, and political fragmentation endured in many regions. What changed was the comparison between what one site could sustain and what an opponent could bring against it.

That comparison repays the beginning. The castle first converted local land, labour, stores and loyalty into a defensible concentration. Larger powers learned to mobilise similar resources across districts and kingdoms. They could surround the strongpoint, cut its supply, batter it, staff it or order it dismantled. The walls remained. The scale around them had changed.

How It Actually Works

Before the castle

Europe entered the early Middle Ages with many fortified places and no need to invent the wall. Roman forts still occupied roads and frontiers. Town walls enclosed populations and markets. Hilltops, royal centres, monasteries and refuges could be defended. Carolingian rulers built and occupied fortified residences. Scandinavian communities raised ring fortresses and defended settlements. The castle emerged from this crowded field rather than from an empty landscape.

Defence had also been organised at several scales. A town might maintain its walls, a ruler might occupy an old fort, and communities could retreat into defended places during raids. The castle differed by concentrating residence and command in one elite household. Its rise did not mean communal defence vanished. It added another kind of strongpoint whose owner could negotiate with kings, towns, monasteries and neighbours from behind his own barrier.

Across parts of western Europe during the centuries around 1000, the social unit inside the defence changed. A powerful household established a residence that combined protection with command over surrounding land and people. Political authority had fractured, mounted elites competed locally, and rulers often depended on men whose power came from estates and followers. A strong residence let such a household survive attack, store resources and make jurisdiction visible.

The transition was uneven. Words in charters and chronicles shift meaning. Archaeological dates are often broad. Some fortified sites belonged to rulers, some to local lords, some to communities, and many changed category over time. France, the Low Countries, the German lands, Italy, Iberia and the British Isles did not follow one timetable. By the eleventh century, however, castles had become a recognisable instrument of aristocratic and royal power across much of Latin Europe.

Conquest in earth and timber

England offers the clearest burst. Edward the Confessor's foreign associates had built a small number of castles before 1066. William's conquest turned the form into a system. Norman forces raised strongpoints to secure routes, towns and centres of rebellion, then placed trusted followers in them. The castle helped a relatively small conquering elite occupy a wealthy kingdom without spreading itself evenly across every village.

Custody mattered from the start. A trusted castellan or constable held the site, supported by troops whose service might be tied to land, wages or temporary obligation. The castle provided a secure place for horses, weapons, food and messages while forces moved elsewhere. It also created a point where local people met the new order: bringing dues, attending a court, supplying labour or seeking protection. Conquest became routine through repeated visits to the gate.

At first, speed mattered. A ditch, bank, palisade and defended enclosure used local material and large amounts of labour. Where the terrain allowed, builders raised a motte with a tower at the top and a bailey beside it. Elsewhere they made ringworks or inserted enclosures into older defences. The buildings within were timber: hall, chapel, accommodation, stores, stable and workshops. Archaeology often recovers postholes, burnt layers and ditches rather than the structures people saw.

Castles went into existing places because existing places mattered. At Lincoln, houses were cleared. At York, castles stood within and beside an important city. At Exeter, Warwick and Cambridge, fortification altered urban space. Other sites controlled river crossings, roads or estate centres. The pattern was military occupation joined to administration. Domesday Book records a transformed landholding order; castles helped keep that order enforceable.

The conquest did not end in one campaign. Rebellions, northern devastation and frontier warfare continued. Norman lords built in Wales and the Welsh March, where castle, lordship and settlement expanded together under unstable conditions. In Scotland and Ireland, castle building later accompanied forms of royal expansion and colonisation without copying the English sequence exactly. Earth-and-timber works remained useful because political control had to move faster than monumental masonry.

Stone, towers and residence

Stone could announce a longer claim. Chepstow's great tower began soon after the Norman conquest on a cliff above the River Wye. Colchester and the White Tower in London drew on Roman material and monumental forms. Their massive walls contained halls, chapels, chambers, stairs and stores. They were residences of exceptional patrons as well as military statements. A great tower gave the owner a defensible interior and a stage for rule.

The White Tower illustrates the scale of that stage. Its chapel, chambers and circulation belonged to royal occupation, while its mass dominated the eastern approach to London. Such towers used carefully cut stone around openings and corners, with rubble and mortar forming much of the wall body. Floors and roofs remained timber. The monument that appears solid was a layered construction filled with combustible, replaceable rooms.

Stone building spread through several routes. Some timber castles were rebuilt piece by piece. Some acquired a stone shell around the summit of a motte. Others added curtain walls while domestic buildings remained wooden. Patrons reused existing earthworks because the site, labour investment and local authority were already there. The result was rarely a total replacement on one date.

The twelfth century made castles central to civil war as well as royal government. During the conflict conventionally called the Anarchy, supporters of Stephen and Empress Matilda fortified positions across England. Chroniclers complained about adulterine castles, meaning sites built without accepted royal authority, though the label and totals are difficult to pin down. The political lesson was clearer: private fortification could turn contested lordship into persistent armed fact.

Henry II restored royal authority and invested in major castles. At Dover, work beginning in the late 1170s created a great tower surrounded by defensive circuits on the commanding Channel site. The tower's rooms were prepared for royal ceremony and hospitality. Defence and reception reinforced each other. A king entering through controlled gates to an elevated hall did not cease governing because nobody was shooting at him.

The model was broader than hereditary lordship. From 1142 the Hospitallers developed Crac des Chevaliers as a fortified base in the eastern Mediterranean. It passed to the Mamluks in 1271 and was altered again. The site belongs fully to crusading and regional histories outside this book's scope, but its phases matter here: military orders and conquering states could combine residence, storage, command and fortification without reproducing an English noble household.

The period also widened the design vocabulary. Curtain walls enclosed larger areas. Projecting towers improved flanking. Gatehouses grew more complex. Keeps remained, but domestic ranges could spread around wards. Castles became less easy to describe as one tower inside one fence because different functions were distributed across the plan.

The age of the siege

Royal power did not end castle warfare. It made some sieges larger.

Besiegers often built earthworks, timber shelters and fortified camps of their own. A counter-castle could watch a gate, protect engines or block relief. Trenches and barriers reduced the danger of sorties. The attacking army therefore answered architecture with temporary architecture, while engineers adjusted range, ground and cover. A siege landscape could contain more new building outside the walls than repair inside them.

In 1199 John became king of an Angevin political world stretching across England and substantial French lands. He inherited and built castles on both sides of the Channel. Château Gaillard, raised by Richard I above the Seine, demonstrated how a ruler could fortify a strategic route with layered defences. Philip Augustus captured it in 1204 after a long operation that combined isolation, engineering and assault. The fall helped open Normandy to French conquest. Even a celebrated castle worked inside a wider campaign.

John's conflict with his barons brought the lesson home. Rochester Castle was held against him in 1215. The royal army attacked the bridge and enclosure, then concentrated on the great keep. Mining collapsed one corner. The defenders withdrew behind an internal division and continued. Starvation eventually forced surrender. The siege became famous because its engineering is documented, including John's order for pigs whose fat helped burn mine supports. Its deeper value is the sequence: access, isolation, structural damage, internal fallback, hunger.

Castles also shaped terms short of destruction. Possession could be pledged during negotiations. Heirs, constables and royal officers became entangled in custody. A lord might be required to surrender a stronghold to the king during war, then receive it back. Marriage transferred castles. Minority and widowhood opened arguments over control. The building was property, office, residence and weapon, so changing hands rarely meant one thing.

The baronial conflicts of Henry III's reign produced the siege of Kenilworth in 1266. Water, extensive walls and a large garrison made the castle difficult to close down. The royal host built up engines and supplies while disease and scarcity worked on both camps. A settlement known as the Dictum of Kenilworth was issued during the wider conflict, offering rebels routes to recover lands through payment. The castle surrendered after months. Political reintegration and siege pressure developed together.

Smaller sieges followed the same logic with fewer records. A local commander might threaten crops, seize an outer bailey, display prisoners or negotiate through clergy and relatives. Weather could close roads or spoil grain. Pay could fail. A relief rumour could empty a camp before any army appeared. The famous engine was only one lever among social pressure, damaged reputation and the fear that allies had changed sides.

Castles of conquest and government

North Wales supplies an unusually complete demonstration of castles working as a conquest network. It should not be mistaken for the normal European sequence. Native princes and marcher lords built castles before Edward I. Llywelyn ab Iorwerth used strongholds such as Dolwyddelan and Dolbadarn. Anglo-Norman lords developed sites including Chepstow, Pembroke and Caerphilly. The landscape held competing castle traditions tied to conquest, resistance and regional government.

Caerphilly, begun by the marcher lord Gilbert de Clare in 1268 amid conflict with Llywelyn ap Gruffudd, used extensive water defences and layered walls. Its scale reflected a political emergency and the resources of one of the most powerful nobles in Britain. The castle altered the local balance before completion because building itself declared possession.

Edward I's wars of 1277 and 1282 to 1283 destroyed the independent power of the Welsh princes. The following programme joined castles, walled towns, roads, ports, garrisons and officials. Flint, Rhuddlan, Conwy, Caernarfon, Harlech and Beaumaris were not isolated monuments. They were nodes in occupation. Coastal or river access helped ships bring food and troops past difficult terrain and hostile land routes.

The adjoining boroughs were part of the design. Settlers received privileges and protection, while Welsh residents could face exclusion from property and office within the new towns. Markets, courts, walls and castle garrisons reinforced one another. A ship unloading grain at Conwy supported soldiers, officials and a planted urban community at once. The architecture of supply was also an architecture of colonial rule.

Master James of St George and other specialists worked across the programme, but no single designer produced every decision. Clerks, paymasters and local supervisors coordinated labour drawn from wide areas. At Harlech in the summer of 1286, records list nearly 950 workers: more than two hundred masons, over a hundred quarrymen, smiths, carpenters and hundreds of labourers. The figures turn a silent wall into a crowded industrial project.

Between 1283 and 1287, Conwy's castle and town walls rose with remarkable speed. The castle had eight towers, strong curtain walls and separate residential areas, while the adjacent walled town housed a settler community under royal protection. The site controlled a crossing and could be supplied by river. Its lime-rendered walls, banners and painted fittings would have made occupation visible from a distance.

Harlech stood on rock above the sea, with a protected route down towards water. Beaumaris, begun later on Anglesey, offered a more regular nested plan and a tidal dock. It remained unfinished as Edward's Scottish wars drew money elsewhere. Caernarfon joined military architecture to imperial imagery and administration. Together the buildings expressed a system that could mobilise at scale. They also record the coercion of conquest, land seizure and exclusion from the new boroughs.

The system was tested. During the revolt of Madog ap Llywelyn in 1294 to 1295, Edward was trapped at Conwy with restricted supplies. In the early fifteenth century, Owain Glyndŵr's rising captured several castles and used Harlech as a headquarters until royal forces recovered it. Strong coastal connections helped some royal garrisons endure. A castle designed as part of a network remained dependent on that network.

From stronghold to great house

By the later Middle Ages, castle building continued while domestic ambition became more visible. This was not a retreat into decoration. The political environment, likely enemies and forms of elite competition had changed.

At Goodrich, the Marshal family had added a strong curtain and corner towers in the thirteenth century. Later Valence work developed impressive residential ranges within the defences. Joan de Valence's household accounts from 1296 to 1297 show movement among properties, large food purchases and a staff divided across service, worship, transport and administration. Architecture and documentation meet: the rooms reveal hierarchy, the accounts reveal who filled them.

The household was not a fixed crowd. Officers could remain while the countess travelled; servants joined for a stay; messengers and tenants passed through; craftsmen arrived for repairs. Women appear in the surviving evidence as noble managers, attendants and suppliers, though official lists preserve them unevenly. The population of a castle expanded and contracted with season, business and danger, which is why one garrison figure or one meal account cannot define life there.

Other owners built or remodelled castles to stage arrival. Gatehouses framed heraldry. Moats reflected walls and separated privileged space. Gardens, parks and private chambers expanded the experience of lordship. Bodiam Castle, licensed in 1385 during concern about French raids, combined a formidable image with a residential plan. Arguments over whether it was defensible or theatrical miss the mixed requirement. It was meant to secure, house and impress at the level its owner expected to face.

Garrisons in many castles were modest during peace. A constable, porter, watchmen and household staff maintained custody. Numbers expanded under threat, sometimes with local levies or hired troops. Weapons and armour had to be stored, but the walls spent far more days enclosing accounts, meals and repairs than resisting ladders. The castle's military value lay partly in being ready enough that attack did not occur.

Guns, forts and ruins

Gunpowder artillery entered European warfare during the fourteenth century, first as one troublesome weapon among older engines. Early guns could frighten, injure and damage gates or parapets, but reliability and logistics limited them. Defenders mounted guns too. Loops were adapted, platforms strengthened and older walls repaired.

The weapon's effect depended on a train behind it. Powder ingredients, metal or stone shot, timber beds, ropes, wagons, draught animals and skilled gunners had to arrive together. Heavy pieces moved slowly and could break roads or bridges. A castle that denied a suitable firing position or forced a long supply route could still frustrate artillery whose theoretical power exceeded its practical reach.

During the fifteenth century, larger and more effective cannon changed siege calculations. The French reconquest late in the Hundred Years War used royal artillery as part of a mobile, financed campaign system. The Ottoman capture of Constantinople in 1453 displayed heavy bombardment alongside mining, assault, naval pressure and overwhelming operational preparation. Cannon mattered most when armies could manufacture, transport, supply and protect them.

Fortification changed shape. Lower, thicker walls backed by earth resisted impact better than tall exposed curtains. Rounded towers and artillery platforms provided space for guns. In Italy, angled bastions and interlocking fields of fire developed into a new fortress system. These works belong to the history of gunpowder fortification more than to the classic castle, even when older sites were absorbed into them.

Castles did not vanish. They remained residences, prisons, stores and local bases. During the sixteenth and seventeenth centuries some were modernised, some became grand houses, and some decayed. The British civil wars returned many medieval sites to combat. Raglan, Corfe, Scarborough, Pontefract and others endured sieges under cannon. After capture, deliberate damage reduced their future value.

Ruin then became the final design layer. Roofs were removed, lead sold, stone reused and vegetation established. Antiquaries measured. Owners restored selectively. Modern custodians stabilised walls and directed visitors through safe routes. The bare shell is neither the medieval building nor neutral evidence. It is what weather, politics, salvage and later taste allowed to remain.

How we know

Castles leave unusually large evidence and unusually strong distortions. Standing walls reveal masonry, sequence, circulation and alteration, but lost timber, roofs, plaster, colour, furniture and service buildings make them look stonier and emptier than they were. Archaeology recovers ditches, postholes, food waste, tools, weapons, drains and phases erased above ground. Tree-ring dating, mortar analysis and architectural comparison can narrow chronology, though reused material and repeated repair complicate it.

Written evidence is uneven. Royal building accounts may record wages, transport and materials in exceptional detail, while a small private castle leaves little beyond a charter name. Household accounts expose consumption and staff but survive for unusual owners and short periods. Chronicles favour sieges, rulers and dramatic failure. Administrative documents reveal custody, licences, rents and repairs. Images can clarify equipment while compressing time or serving moral and political aims.

The sample therefore favours stone, wealth, royal projects and places later people chose to preserve. Earth-and-timber castles, ordinary workers and routine years are harder to recover. Restoration adds another layer: a parapet or cleared courtyard may embody modern judgement. General claims in this book are bounded by region, date, patron and evidence type.

What People Get Wrong

“A castle was a fort”

A fort and a castle can look alike because both make attack difficult. The mistaken model removes the household and leaves a military shell.

Most castles housed high-status people or represented their authority. They contained halls, chambers, chapels, kitchens, stores and administrative spaces. Tenants came to courts. Officials kept accounts. Guests were received according to rank. A lord could retain or build a castle without facing an immediate siege because defence, status and local government were parts of the same position.

The distinction is not absolute. Some castles were campaign bases with little comfort. Some royal fortresses housed officers more often than rulers. Some palaces were heavily defended, and walled towns could contain castles. The useful distinction is functional rather than lexical: do not explain the whole building from the battlements. A military reading cannot account for why so much expensive space was devoted to ceremony, privacy, worship and service.

This matters because the castle's daily power operated in peace. A defended hall could enforce hierarchy without firing anything. War revealed the institution. It did not create it. Ruins and chronicles favour military features because stone and crisis survive; kitchens, paperwork and routine obedience leave quieter evidence.

“Castles progressed neatly from wood to stone”

The schoolbook sequence is attractive: quick wooden motte-and-bailey, strong stone keep, concentric masterpiece, then cannon. It turns several overlapping choices into one ladder of improvement. The sequence survives because lost timber leaves earthworks while rebuilt stone supplies the photograph.

Timber and earth were capable materials. Earthworks absorbed force and created height. Timber allowed rapid construction, complex carpentry and easy alteration. Stone resisted weather and fire better and carried prestige, but took money, skill and supply. A stone castle still relied on timber roofs, floors, doors, scaffolds, bridges and outbuildings. A few early castles began in masonry. Some earth-and-timber sites lasted for generations. Others were abandoned before any upgrade because their political purpose disappeared.

Plans also moved sideways. A shell keep, rectangular tower, enclosure castle and ringwork answered different sites and budgets. Later owners could reuse old forms or add domestic ranges without replacing defensive circuits.

Architecture follows constraints, not a technology chart. Wood was not failed stone. Stone was not automatically the rational next step. The patron purchased a combination of speed, endurance, display and maintainability. Survival then biases the sample: the stone phase remains visible while the timber buildings that stood beside it disappear.

“Attackers usually stormed the walls”

Assault dominates images because it supplies a visible climax. An opened gate could achieve more than a bloody climb.

Sieges used negotiation, blockade, surprise, bribery, bombardment, mining and threats alongside assault. Attackers cut routes and waited for hunger. Defenders bargained over relief deadlines and surrender terms. A party might climb one neglected sector at night. Miners could collapse a tower. Engines could damage roofs and parapets or make a breach. An assault often came after these methods had weakened resistance.

Storming carried special danger. Ditches broke formations, ladders channelled climbers and the wall gave defenders height and cover. A failed attack damaged morale and consumed trained soldiers. Yet waiting also cost food, wages and health, so commanders sometimes accepted the risk.

Replace the duel between wall and warrior with a contest between systems. The best question is not how the attackers got over. It is why the defenders could no longer make them stay out. Chronicles may compress weeks of digging, waiting and bargaining into the breach that supplied a satisfying ending. A surrender could also be omitted or moralised when the writer cared more about courage, treachery or divine favour than operational detail.

“Boiling oil was the standard defence”

Hot oil survives because it is vivid, cruel and easy to animate. Oil was valuable, difficult to heat safely and not the obvious substance to pour away.

Defenders could use stones, arrows, bolts, timber, hot water, heated sand, burning material and other missiles depending on what they had. They pushed ladders, fought fires, repaired damage and made sorties. Openings above gates or projecting defences did not exist for one trademark liquid. Some accounts mention oil or similar substances in particular places, but rarity is the responsible default.

The myth grew because later storytelling likes a recognisable medieval weapon and because an opening in a ceiling invites a single explanation. Real defence was opportunistic. The garrison used gravity, cover, stored material and the attacker's exposure.

The better reading turns the eye away from named gadgets. A castle did not work through a set of iconic tricks. It worked by forcing attackers into bad positions where many ordinary actions became more effective. The myth also mistakes possibility for routine: one documented use does not establish the normal contents of every gatehouse.

“Peasants built castles by brute force”

The image has thousands of anonymous peasants dragging blocks while a lord watches. It survives because walls preserve patrons' names and erase most workers.

Castle building demanded enormous physical labour, and coercion belonged to medieval government. Depending on place and period, rulers and lords could demand customary services, requisition transport, take land or impose supply obligations. Conquest projects imposed costs on people who had little power to refuse. But brute force alone could not set an arch, select quarry stone, burn lime, frame a roof, make a crane or keep deliveries moving through several seasons.

Major accounts show mixed workforces: master masons, quarrymen, carpenters, smiths, lime workers, carters, boat crews, clerks and labourers, many receiving wages. Contractors and customary obligations also appear. At Harlech's busiest recorded summer, the 546 labourers were numerous, but so were 227 masons and more than a hundred quarrymen. The exact categories belong to one exceptional royal programme, not to every castle.

The correction matters because construction was an organisation, not a feat of lifting. Skill determined sequence; administration connected distant material; ordinary labour supplied most of the movement. Wage rolls still miss household work, compelled service and people who fed or transported workers off the books. Replacing the lash with a modern building firm would be another fiction. The medieval project mixed expertise, payment, obligation and power.

“Castle life was permanently cold, dark and filthy”

Ruins teach this myth efficiently. Roofless stone feels bleak. The missing materials are the ones that made rooms liveable.

Occupied chambers could have fireplaces or hearths, shutters, glazed windows, wall hangings, painted plaster, timber screens, beds, cushions and rush or board floors. Kitchens, wells, drains and latrines managed food, water and waste with varying success. Elite residents could enjoy considerable comfort by medieval standards. Conwy's royal apartments included private rooms, chapel access, fireplaces and latrines, though the king used them rarely.

None of that makes every castle pleasant. Fuel smoked. Light cost money. Upper rooms, wall walks, service yards and garrison quarters could be miserable. Crowding during siege damaged sanitation. Servants performed the cleaning and carrying that allowed elite comfort.

Comfort was distributed, not universal. A castle could be warm in one chamber, freezing on the next watch and foul beside an uncleared shaft. Comfort existed because labour moved heat, water, food and waste through the building. Modern visitors experience an unroofed public monument; medieval residents experienced furnished rooms, service routines and sharp differences of rank.

“Cannon made castles obsolete overnight”

The first European guns did not arrive as reliable wall erasers. Early artillery varied in power, accuracy, rate and safety. Moving powder, shot and heavy pieces was an operation. Defenders adopted guns too.

Over the fourteenth and fifteenth centuries, artillery improved and some governments became better at financing and transporting it. High exposed masonry became increasingly vulnerable to sustained bombardment. Builders lowered profiles, thickened works, backed walls with earth and provided gun positions. Some old castles were adapted. Others remained useful against raiders, rebels or armies without a serious siege train.

Political use also outlasted military fashion. Castles housed courts, prisons, stores and households. Medieval sites returned to combat during the seventeenth-century civil wars in Britain, when many were repaired, besieged and then slighted.

The weapon changed the comparison, not the calendar. Cannon mattered when an organisation could obtain guns, move them, keep them supplied and protect their crews, and when terrain allowed a useful firing position. New forts inherited garrisons, controlled approaches and defended storage while changing their geometry. The classic castle did not disappear at one moment. Its military value diverged by place, opponent and purpose.

Use It

Read the approach before the skyline

When you visit a castle, resist walking straight to the highest tower. Begin at the edge of the site and reconstruct the journey in.

Find the old road, harbour, bridge or track. Notice which ground is visible from the walls and where the slope changes. Look for ditches that appear as harmless lawns, banks softened by centuries and gate routes altered for visitors. Ask where a cart could move, where horses would wait and where an attacker could place ladders or engines. A modern entrance may cut through a wall that was once continuous.

Then turn around. What did the castle watch? A river crossing, town, valley, coast, frontier, estate or route to another stronghold usually explains more than architectural vocabulary. A castle gained value by controlling relationships among places. A tower on an isolated postcard is missing its subject.

This lens transfers beyond ruins. Security often begins with approach and access rather than the protected object. The route tells you what the system expects to admit, delay, inspect and exclude.

Follow the bottlenecks and fields of view

A defensive system makes some movements easy and others costly. Trace those choices.

At the gate, count changes of direction and barriers. From a tower, look along the wall rather than only outward. From an arrow loop, notice how a narrow opening trades view for cover. At a stair or doorway, ask how many people can pass and who controls the upper landing. Inside layered wards, identify what remains if the first enclosure falls.

Do not award every feature one perfect purpose. A gatehouse may house officers and guests. A tower may provide a chamber, latrine and prestige as well as flanking fire. A moat may mark status and manage access while offering uneven military value. Good interpretation allows several functions where the fabric and context support them.

The useful question is: where does the design force a decision? Bottlenecks concentrate control, but they also concentrate failure. A blocked gate delays an enemy and traps a defender. A single well protects water and creates dependence on one source. Strength and fragility often occupy the same feature.

Ask what supplied the walls

Every impressive structure invites attention upwards. Castle logic runs downwards into quarries, woods, fields, kilns, roads, rivers and accounts.

Look for the material sequence. Is the facing stone local? Are there different building phases or reused blocks? Where could timber, sand, lime, iron and water have arrived? A harbour, river landing or service gate may explain construction speed. Tool marks and mason's marks show skilled work, but labourers moved most of the weight. Mortar and wall cores carry less glamour than carved windows and more of the building.

Then extend the question into operation. Which farms supplied grain and animals? Where was flour milled? How did fuel reach kitchens and chambers? Who repaired roofs, cleared drains and kept horses? A siege attacked these connections because the castle depended on them.

The transferable lens is severe: do not confuse a protected centre with self-sufficiency. The thicker the wall and larger the household, the more material must cross its boundary in ordinary time. Resilience comes from protected supply, reserves and alternatives, not from appearance of independence.

Separate the peacetime household from the wartime garrison

A ruin encourages one population: soldiers. Refill it twice.

On an ordinary day, place the constable, porter, clerk, cook, groom, chaplain, craft worker, messenger, cleaner, estate officer and visiting tenant. Add the lord's household when present: family, guests, servants, guards, officials, animals, baggage and furniture. Follow food from delivery to store, kitchen, hall and waste. Follow a petition from gate to officer. Follow a private visitor through rooms whose access signals rank.

Now impose siege. Bring soldiers and refugees inside. Close normal routes. Move weapons and stones to walls. Ration food. Protect water. Turn halls into stores, workshops or crowded accommodation. Reduce privacy and increase fire risk. The architecture has not changed, but the operating system has.

This distinction prevents two errors. It stops daily life becoming a permanent battle scene, and it stops domestic rooms being treated as evidence that defence was theatrical. Peacetime organisation generated the knowledge, loyalty and reserves used in war. The two populations overlapped without being identical.

Distinguish display from uselessness

A crenellated skyline, symmetrical moat or ceremonial gate can look too composed to be military. The tempting response is to expose it as theatre. That can be as crude as believing every turret was indispensable.

Display altered behaviour. A castle announced ownership, wealth, alliances and the capacity to mobilise labour. It could deter opportunistic violence and make tenants, officials and guests experience the patron's rank. A moat that a royal army could cross might still control everyday entry. A gatehouse that would fail under heavy cannon could remain effective against a raid. Security is always security against someone.

Judge the claim at the intended scale. Who was the likely attacker? How quickly could help arrive? What political permission would assault require? What damage was the owner willing to suffer? A residence can be defensible without being a frontier fortress. A wall can be symbolic without being functionless.

The wider lesson is to avoid a false choice between practical and performative design. Institutions often work partly because people believe the display. The test is whether appearance changes incentives, access or legitimacy, and what happens when a stronger opponent refuses the message.

Read every ruin as edited evidence

The building you see is the result of construction, alteration, attack, decay, salvage, excavation, restoration and visitor management. Each stage selected what survived.

Look for blocked doors, scars where roofs met walls, different masonry, reconstructed battlements and modern stairs. A clean lawn may replace service buildings and rubbish deposits. An open courtyard may once have been dense with timber ranges. A bare wall may have carried plaster, colour and hangings. The safest path may reverse medieval circulation.

Read labels for verbs. Rebuilt, restored, consolidated, cleared and reconstructed mean that later judgement entered the evidence. Ask which date the site presents. A thirteenth-century gate attached to a sixteenth-century range and a nineteenth-century parapet cannot be reduced to one authentic moment.

The lens applies to archives as well. Royal accounts preserve royal projects. Chronicles preserve crises. Stone favours wealthy patrons. Ordinary maintenance leaves weaker stories. Evidence is not false because it is selected, but the selection determines which questions it can answer.

The limits

Castle logic is useful, but it should not become a universal metaphor for security or medieval society.

Most people did not live in castles. Many regions relied more on towns, communal fortifications, fortified churches, manor houses, citadels or mobile force. A British or French sequence cannot stand for Iberia, Italy, the Baltic, the eastern Roman world or the Islamic Mediterranean. The category itself shifts across languages and scholarship.

Walls also reveal power from the owner's position. Protection for a household could mean forced labour, displacement, tolls, imprisonment and surveillance for neighbours. Admiring engineering without asking who paid and who was controlled repeats the castle's preferred story.

Finally, no design lens can reconstruct private experience cleanly. Accounts count food without telling us whether a servant was content. A chamber's fireplace proves capacity for heat, not constant comfort. Ruins invite certainty because stone feels direct. It is still evidence requiring limits.

The one thing to keep

Keep the clock.

A ruined castle looks timeless. Its original purpose was to make time unequal. A ditch lengthened the approach. Gates divided entry into stages. Towers lengthened observation. Walls protected repair. Stores bought days. A garrison held because someone expected relief, surrender terms or the attacker's exhaustion before its own reserves failed.

That changes how the building looks. The strongest feature may be a harbour rather than a keep, a well rather than a tower, an account book rather than a sword. The castle was the point where land, labour, supply and authority were organised to make another person spend more time than they could afford.

Every saved hour had a cost. Quarrymen cut, carters hauled, masons built, servants cooked, clerks counted, tenants paid and guards watched. The walls produced none of these things for themselves. When the roads closed, loyalty failed or a larger power could keep pressure on for longer, masonry exposed its dependence.

Read the towers as stored delay, then follow the people and routes that made delay useful. The ruin stops being a fantasy of permanent strength. It becomes a record of power trying to outlast its problems.

Terms

Castle. A defensible residence or command centre associated with elite authority over land, people or government. Forts, palaces, walled towns and citadels overlap with castles but are not synonyms.

Motte. A steep mound carrying a tower or defended summit. It created height quickly, though its construction, settlement and later stone loading could present serious engineering problems.

Bailey. The enclosed court beside or below a motte. It contained the working castle: hall, chapel, stables, stores, workshops, accommodation and animals beyond the tower.

Ringwork. A defended enclosure formed by a bank and ditch, usually topped in its first phase by timber. Ringworks offered rapid occupation without the dominant mound of a motte-and-bailey plan.

Keep or donjon. The principal tower or strong building of a castle, often containing prestigious rooms, stores and command space. French donjon supplied English dungeon, but the medieval donjon was not defined as an underground prison.

Shell keep. A stone wall enclosing buildings around the top of a motte. It strengthened summit defences while spreading rooms around a courtyard rather than filling one tower.

Curtain wall. The defensive wall enclosing a ward. Curtains connected towers and gates, carried wall-walks and separated the approaches. Their effectiveness depended on height, thickness, flanking and maintenance.

Enceinte. The main defensive enclosure or circuit of walls. A castle could contain more than one enceinte, creating successive zones rather than relying on a single perimeter.

Ward. An enclosed area within the castle. Outer and inner wards distributed service buildings, troops and high-status accommodation while providing another defended zone if an outer line was lost.

Battlement. A parapet with alternating raised merlons and open crenels. It gave defenders cover and shooting gaps and became such a strong status symbol that later unfortified buildings borrowed the form.

Wall-walk. The passage along the top or inner side of a defensive wall. It allowed movement between fighting positions, though broken stairs, narrow sections and separate towers could complicate circulation.

Arrow loop. A narrow opening through which an archer or crossbowman could observe and shoot from cover. Internal splays widened the angle, while the opening restricted vision and movement.

Flanking tower. A tower projecting beyond a curtain wall so defenders could see and shoot along its face. Flanking reduced dead ground at the base, one of the central problems of straight walls.

Gatehouse. The defended building controlling the main entrance. Doors, bars, portcullises, turns and flanking positions could be combined with guardrooms, offices, chambers and an imposing ceremonial front.

Barbican. An outer defence protecting or extending the approach to a gate. It ranged from a walled passage to a separate work forcing entrants through another controlled space.

Portcullis. A vertically sliding gridded barrier, usually timber and sometimes reinforced with iron, moving in grooves. It could close an entrance quickly, but needed lifting gear, maintenance and protected operators. Doors and bars remained common.

Drawbridge. A movable bridge crossing a ditch or separating parts of an entrance. Some lifted by chain or counterweight. Fixed bridges were common, so a moat does not guarantee one.

Moat. A ditch around all or part of a castle, dry or water-filled. It obstructed movement, exposed attackers and could hinder mining. Water level, geology and maintenance determined its practical effect.

Hoarding. A temporary timber gallery projecting from a wall or tower, allowing defenders to work beyond the face. Sockets in masonry may reveal where beams supported a hoarding now entirely lost.

Machicolation. A stone projection with openings through its floor, used to observe or attack the base of a wall or gate. It made part of a hoarding's role permanent.

Postern. A secondary door, usually smaller and less conspicuous than the main gate. It could support communication, supply or sorties and created an access point that needed careful guarding.

Concentric castle. A castle with nested defensive circuits, often arranged so an inner wall commands the outer. It describes related defences, not one plan or guaranteed invulnerability.

Licence to crenellate. A royal grant recorded in medieval England permitting or acknowledging fortification. Licences could carry honour and political recognition; their survival does not mean all unlicensed battlements were illegal.

Garrison. The troops assigned to hold a castle. Peacetime numbers could be small and rise sharply during war. A garrison's value depended on leadership, supplies, skill, loyalty and expectation of relief.

Constable. The officer responsible for custody and management of a castle on behalf of its owner. Duties could include commanding soldiers, supervising repairs, controlling stores and carrying out local administration.

Investment. The enclosure of a fortified place by an attacking force to restrict movement, supply and communication. Complete investment demanded enough troops to watch routes and resist relief.

Escalade. An attempt to take walls by climbing ladders, ropes or siege equipment. It could exploit surprise or weak defence but exposed attackers at ditches, wall bases and narrow entry points above.

Trebuchet. A stone-throwing engine using a sling, with traction or a counterweight providing power depending on type. Its effect depended on projectile, crew, position and sustained operation, not legendary precision.

Mine and countermine. An attacker's tunnel driven under masonry to bring it down, and the defender's attempt to intercept, flood or destroy that working. Soil, rock, timber and listening shaped the underground contest.

Garderobe. A term commonly used for a castle latrine, often built into a wall with a chute or pit. Medieval usage could be broader, including a room for clothes or belongings, so context matters.

Go Deeper

Marc Morris, Castle: A History of the Buildings That Shaped Medieval Britain (2003). Start here for a clear narrative from the Norman conquest into the seventeenth century. Morris follows six major buildings and the rulers, wars and households that changed them, so architectural development never becomes a list of features. It is written for a general reader and moves quickly. The focus is Britain, and the television-series origin gives famous sites more room than ordinary castles, but it supplies the most inviting bridge from this book to a full history. Read it first for sequence, then revisit the buildings with the social questions raised here.

John Goodall, The English Castle, 1066 to 1650 (2011). Use this as the architectural survey. Its plans, photographs and close readings show how gates, halls, chambers, chapels, walls and service routes fitted together and how one building accumulated phases. Goodall keeps warfare, governance and domestic life in the same frame. The scale and visual detail make it better for browsing beside a site visit than for reading at speed. It remains England-centred, but few books train the eye so effectively to distinguish original sequence from later alteration. Keep it open beside plans when visiting a complicated ruin.

Charles Coulson, Castles in Medieval Society: Fortresses in England, France, and Ireland in the Central Middle Ages (2003). Read this for the strongest challenge to a military-only model. Coulson uses legal and documentary evidence to place crenellation, fortification and castle ownership within status, consent, household government and political relationships. His interpretation explains why many formidable buildings were rarely attacked and why royal licences cannot be read as simple permits. It is scholarly and deliberately revisionist. Set it beside siege evidence so that a necessary correction does not become a new one-factor account.

Peter Purton, A History of the Late Medieval Siege, 1200 to 1500 (2010). Turn here when camps, engines, mines, surrender terms and artillery need full treatment. Purton compares hundreds of operations across Europe and beyond, following the spread of expertise and the slow interaction between older methods and guns. The density of cases makes it a reference work rather than a first read, and its geographical range demands attention to local context. Its reward is proportion: celebrated assaults shrink back into a military world in which waiting, supply and negotiation occupied much of the work. The earlier companion volume carries the same method from late antiquity to 1200.

Notes and Sources

Scope, definition and chronology

The book uses castle for a fortified high-status residence tied to lordship, administration or command, while recognising that medieval terms and modern scholarly definitions do not map cleanly across regions. Coulson, Pounds, Goodall, Liddiard and Creighton supply the principal debate over military function, residence, status, consent and landscape. The definition is a working model, not a claim that every site falls into one exclusive category.

Fortified settlements, elite residences, Roman sites and communal defences predated the medieval castle. The account therefore avoids a single inventor or birth date. The concentration of castles in parts of western Europe from the tenth and eleventh centuries follows the common chronology in castle scholarship, with regional variation kept explicit.

The book concentrates on Latin western and central Europe and uses England and Wales most heavily because standing remains, royal records and modern institutional research are unusually dense. France supplies a major campaign case, and Crac des Chevaliers is used only as an eastern Mediterranean boundary case. None of these settings is treated as a universal European sequence.

Norman expansion, mottes and baileys

Historic England monument descriptions support the standard components and functions of motte-and-bailey castles in England: mound, tower, adjoining enclosure, military use, aristocratic residence and administration. English Heritage site histories for Totnes, Berkhamsted, Old Sarum, York, Lincoln and related sites support the account of earth-and-timber phases and later stone replacement.

A small number of castle sites existed in England before 1066. The manuscript says that the Norman conquest multiplied and systematised castle building rather than introducing the first defensible elite residence to an empty landscape. Urban displacement and the insertion of castles into existing centres follow archaeology, Domesday evidence and the syntheses by Pounds, Goodall, Morris and Liddiard.

“Adulterine castle” is retained as a conventional term associated with Stephen's reign but no exact total is given. Surviving chroniclers used fortification within a political complaint, and modern estimates depend on definition and dating.

Defence and design

The explanations of ditches, walls, battlements, flanking towers, gates, keeps, wards, hoardings and machicolations follow Goodall, Kenyon, Pounds and standing-building evidence. No component is assigned one universal function. Keeps could combine residence, storage, ceremony and defence; gatehouses could contain prestigious accommodation; water defences varied in depth, reliability and purpose.

The account rejects a clean timber-to-stone ladder. Earth and timber remained structurally useful, stone castles retained extensive carpentry, and patrons rebuilt selectively. English Heritage's Castles Through Time, Cadw's Castles of Wales and the site histories for Totnes and Chepstow support the examples.

UNESCO's documentation for Crac des Chevaliers and Qal'at Salah El-Din supports the limited eastern Mediterranean comparison: Hospitaller building from 1142 to 1271, later Mamluk work, and the site's value as evidence of changing fortification through cultural and military exchange. The manuscript uses Crac as a boundary case, not as a template for Latin Europe or as a substitute for a history of the Crusades.

Licence-to-crenellate interpretation follows Charles Coulson's documentary work. Licences began as royal instruments but were neither consistently sought nor a complete register of fortified building. Their honorific and political meanings prevent their use as a simple measure of military permission.

The closing account of scale is a synthesis of Pounds, Coulson, Goodall, Purton and the royal case studies rather than a universal measured sequence. Castles could support custody, records, revenue, supply and local government, while larger political organisations could mobilise longer siege efforts and artillery. The manuscript therefore treats castles as tools through which wider power could grow as well as strongpoints it might later overcome, and it rejects a single crown-against-nobility timetable.

Construction

Goodall, Pounds, Morris and site accounts supply the construction sequence: setting out, quarrying, rubble cores and dressed work, lime burning and slaking, timber scaffolding, lifting, carpentry, ironwork, roofing and finishing. The scarcity of complete working drawings is stated without claiming that medieval builders worked without design. Templates, geometry, practical knowledge and on-site adjustment are all retained.

Construction labour varied by patron, region and obligation. Royal accounts expose wage labour, skilled trades, contracting, transport and procurement unusually well. Customary service, requisition and coercive demands also occurred, while estate and household work may be absent from formal rolls. The body therefore rejects both the anonymous-peasant stereotype and a modern free-labour model.

Cadw's Castles of Wales supplies the Harlech workforce for summer 1286: 227 masons, 115 quarrymen, 30 smiths, 22 carpenters and 546 labourers, nearly 950 people in total. This is explicitly treated as a seasonal peak in an exceptional royal programme, not an average workforce. Cadw describes Conwy's castle and town walls as built between 1283 and 1287. Their surviving accounts do not separate the two costs, so the roughly £15,000 figure is presented for the integrated project rather than the castle alone. No modern monetary equivalent is used because purchasing power, labour share and fiscal capacity are incompatible measures.

Beaumaris began in 1295 and remained incomplete as resources shifted. Cadw supplies the tidal access, concentric arrangement and interrupted programme. The manuscript avoids reducing incompletion to one bankruptcy event.

Siege methods and outcomes

Purton's two-volume history is the main comparative authority for blockade, negotiation, assault, artillery, mining, camps, specialist labour and the gradual effect of gunpowder. The account does not claim that all sieges ended through starvation or that assault was rare everywhere. It states the narrower conclusion that siege commanders selected among several methods and often combined them.

English Heritage's Rochester history supports the 1215 sequence: attack on the enclosure and keep, mining beneath the south-east corner, royal provision of fat from forty pigs to burn supports, structural collapse, continued defence and final surrender under hunger. The pig figure derives from a contemporary royal order, not later folklore.

English Heritage's Kenilworth research dates the siege from 25 June to 13 December 1266, a span of 172 days, and supports the royal supply effort, engines, large rebel garrison, water defences, disease, diminishing food and negotiated surrender. The body identifies the span precisely and avoids treating every day as the same form or intensity of action.

English Heritage's survey of siege myths supports the rarity of hot oil as a standard defensive substance and the use of stones, hot water, sand and other available materials. The body allows exceptional use without making the absence claim universal.

Château Gaillard and the French conquest of Normandy follow Purton, Morris and standard political histories. The account uses the case to show that a strong castle operated within a campaign and does not repeat disputed details of every assault episode.

Household, work and ordinary operation

The principal evidence for Joan de Valence at Goodrich comes from surviving household accounts interpreted by English Heritage. They document fifteen residences used between May 1296 and September 1297, a travelling household varying roughly from 122 to 196 people, around 120 servants at Goodrich in late 1296, permanent staff, named occupations and the killing of eighty-one pigs before the November stay. These figures describe one wealthy household and short recorded periods. The body labels that limitation rather than treating them as average castle life.

Goodrich standing remains and English Heritage research support the hall hierarchy, chambers, fireplaces, window seats, washbasins, latrines and service routes. Cadw's Conwy research supports the royal apartments, separate access patterns, chapel, watching chamber, latrines, lime render and rare royal occupation. The claim that the owner could be less permanent than the porter is an interpretive summary of itinerant lordship and permanent castle custody, not a measured universal rule.

Women's management of castles and estates follows documentary scholarship and named cases including Joan de Valence and Eleanor de Montfort. The book states capability and documented responsibility without claiming equal legal authority or one role across class and region.

The treatment of heat, light, sanitation and furnishing follows material evidence and household studies. It distinguishes the capacity of elite rooms from constant conditions and from the experience of servants, soldiers and temporary wartime occupants. “Garderobe” is qualified because medieval usage was wider than latrine alone.

Edward I's Welsh castles

Cadw and the UNESCO World Heritage documentation are the main institutional sources for Conwy, Caernarfon, Harlech and Beaumaris. The castles and associated walled towns were built within conquest and colonial settlement, not as detached achievements in military architecture. Coastal and river access mattered to supply, while Welsh resistance repeatedly tested the network.

Conwy's castle and town walls were built chiefly from 1283 to 1287. The manuscript uses the documented speed, eight towers, river supply and royal apartments. Cadw also records Edward I's difficult stay during the 1294 revolt; the text does not generalise that episode into permanent shortage.

The Harlech labour figures are a seasonal peak. They do not establish the average workforce across the whole project. Master James of St George had a major supervisory and design role, but the text avoids attributing every feature or every Edwardian castle to one architect.

Display, landscape and mixed purpose

Coulson, Liddiard and Creighton underpin the correction to a military-only interpretation. Their work shows that castles shaped landscapes, processions, jurisdiction and elite identity. Purton and the material record supply the counterweight: siege remained a real design pressure. The manuscript rejects both reduction to war machine and reduction to symbolic theatre.

Bodiam is used as a mixed case. Its licence dates to 1385, in a period of concern about French attack, while its landscape, residential plan and composed image supported status. The book does not decide that it was either useless militarily or a front-line fortress.

Gunpowder, adaptation and decline

Royal Armouries research and Purton support the documented appearance of European guns in the fourteenth century and their uneven early performance. The account distinguishes invention from operational dominance. Fifteenth-century artillery's growing effect is linked to improved weapons and to some governments becoming able to finance, move and supply them. It does not make state growth uniform or inevitable.

The capture of Constantinople in 1453 is used only to show combined operations: bombardment, mining, naval pressure, assault and large-scale preparation. The Ottoman campaign belongs fully to other titles; it is not presented as one cannon ending the medieval castle.

Cadw and English Heritage site histories support continued administrative, residential and military use into the early modern period, including Civil War repair, siege and slighting. Ruin is treated as a sequence of deliberate damage, salvage, lost maintenance, weather and later restoration rather than proof of medieval failure.

Evidence limits

Standing stone overrepresents wealthy and durable castles. Earth-and-timber works, routine service, women's labour, poorer households and peaceful years survive less visibly. Royal accounts are unusually detailed and therefore dangerous as averages. Chronicles select conflict. Restored ruins incorporate later choices. These limitations shape the manuscript's repeated restriction of examples by place, date and evidence type.

No invented dialogue, scene, weather, gesture or private motive appears. Named anecdotes are documented or phrased at the level the sources support. Numbers are retained only where their denominator and period are visible.

Bibliography

Modern works

Coulson, Charles. Castles in Medieval Society: Fortresses in England, France, and Ireland in the Central Middle Ages. Oxford: Oxford University Press, 2003.

Creighton, Oliver H. Castles and Landscapes: Power, Community and Fortification in Medieval England. London: Equinox, 2002.

Goodall, John. The English Castle, 1066 to 1650. New Haven and London: Yale University Press for the Paul Mellon Centre for Studies in British Art, 2011.

Kenyon, John R. Medieval Fortifications. Leicester: Leicester University Press, 1990.

Liddiard, Robert. Castles in Context: Power, Symbolism and Landscape, 1066 to 1500. Macclesfield: Windgather Press, 2005.

Morris, Marc. Castle: A History of the Buildings That Shaped Medieval Britain. London: Windmill Books, 2012. First published 2003.

Pounds, N. J. G. The Medieval Castle in England and Wales: A Social and Political History. Cambridge: Cambridge University Press, 1990.

Purton, Peter. A History of the Early Medieval Siege, c. 450 to 1200. Woodbridge: Boydell Press, 2009.

Purton, Peter. A History of the Late Medieval Siege, 1200 to 1500. Woodbridge: Boydell Press, 2010.

Institutional and site research

Cadw. Castles of Wales; The Castles and Town Walls of Edward I; histories and interpretation for Beaumaris, Caernarfon, Conwy, Harlech and Caerphilly. Rechecked 4 September 2026.

English Heritage. Castles Through Time; Medieval Castles; site histories and research for Dover, Goodrich, Kenilworth, Rochester, Totnes, Berkhamsted, Christchurch and related properties. Rechecked 4 September 2026.

Historic England. National Heritage List descriptions and scheduling assessments for motte, ringwork and motte-and-bailey castles. Rechecked 4 September 2026.

UNESCO World Heritage Centre. Castles and Town Walls of King Edward in Gwynedd; Crac des Chevaliers and Qal'at Salah El-Din. Site descriptions and documentation. Rechecked 4 September 2026.

Royal Armouries. Research and collection interpretation on medieval siege engines, gunpowder artillery and late medieval warfare. Rechecked 4 September 2026.