For the first post on the many amazing things I saw on my continental tour from 3 May to 17 June, I shall be a little unconventional and start near the end. On 15 June, when I was staying in Chalon-sur-Saone in Burgundy, I went for a day trip to see the abbey of Cluny (an incredible place), which will be the subject of a future post. However it was from Cluny itself that I ventured out into the very picturesque countryside in the blazing late afternoon heat to see this place ...
 |
| The Chateau de Lourdon viewed from a hill opposite |
 |
| A view from just below the hill with yours truly |
The Chateau de Lourdon had been made known to me when, as a second year undergraduate, I read "The Transformation of the Year 1000: The Village of Lournand from Antiquity to Feudalism" (1991; originally published in French, 1989) by Guy Bois. Its a super-controversial and provocative book that I've pondered and reflected on a lot ever since I first read it. I would never say its a great work of historical scholarship or that its had a big influence on how I do history, but its resonated with me because it asks the questions that many simply take for granted - when, why and how does medieval society, as most of us are familiar with it anyway, come into being? And what immediately came before it? This is part of what ultimately pushed me towards deciding to specialise in pre-1200 medieval history, as late medievalists can essentially take the existence of religious orders, knights, serfs and the like for granted. The Chateau de Lourdon is mentioned in it several times and indeed features on its front cover. Seeing it in the flesh after all this time, therefore, felt slightly surreal.
But the Chateau de Lourdon is interesting for more than just that. It is one of the oldest, purpose-built castles in France, being first mentioned in a title deed of 888 as the castrum Lordo. When William the Pious (875 - 918), count of Auvergne and duke of Aquitaine, founded the abbey of Cluny in 910 as a flagship for a new kind of monasticism, he gave the castrum Lordo to the monks of Cluny along with his unfortified villa three miles to the southeast in the valley of the Grosne - the site of the abbey of Cluny itself. The monks maintained it and continued to fortify it, using it to administer and defend their far flung agricultural estates in the surrounding area. It had a somewhat dramatic history, experiencing armed takeovers and sieges in 1166, 1250, 1470, 1471, 1574 and 1593. Finally, in 1632, King Louis XIII of France (r.1610 - 1643) and his chief minister Cardinal Richelieu (1585 - 1642), following the advice of the Bishop of Macon, proclaimed that the Chateau de Lourdon was hazardous to maintaining public order and civil peace in the region and had it demolished. Hence why it survives in the condition it does today, as a gutted ruin.
But enough with the Chateau de Lourdon itself. Lets move on to a bigger question. Why did the the earliest castles, like the Chateau de Lourdon, emerge when they did? This is obviously not a trivial question. Castles are one of most emblematic features of medieval society, up there with knights in armour with which they are regularly paired with. As that immortal duo, they're often what gets children (especially boys) first interested in the Middle Ages. That's how I started out my trajectory towards becoming a medievalist, as all the castle-themed books, toys and games I owned by the age of six can surely attest. People of all ages love visiting castles. To give one example, Tower of London is the most visited paid-for attraction in the UK which, as of 2019, was attracting more than 2.8 million visitors annually. Edinburgh castle similarly saw 2.2 million visitors that year and Cardiff castle 502,000. In the wake of the post-pandemic rise in domestic tourism, English Heritage, which counts 66 castles among its 400 sites, saw a record-breaking 375,000 new members join in 2021. People undoubtedly visit castles for the sense of adventure, romance and mystery they get from visiting these enduring, imposing and (sometimes) very ruined monuments from 500+ years ago as has been the case since at least the eighteenth century. But we also know all too well their twofold historical importance. Firstly, their use in war, as mechanisms of defence against enemy attack or as tools of conquest and subjugation. Secondly, as the dwellings of the great and the good of medieval society that served as symbols of their power and prestige and from which they (literally) lorded it over the surrounding landscape. But their great importance to medieval society and how we connect with the medieval past today, does not mean that we can take that for granted. When did they come about, and why?
 |
| The white tower (completed 1077) in the Tower of London. In the late eleventh century, it was the largest fortress palace north of the Alps, and is now part of the UK's most visited pay-to-enter historical tourist attraction. A true symbol of William the Conqueror's power and prestige, and a means of keeping the citizens of London in line. Photo Credit: By Panoramic_view_from_Tower_of_London.jpg: MatthiasKabelderivative work: Nev1 (talk) - Panoramic_view_from_Tower_of_London.jpg, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=10565675 |
If we're being parochial and thinking about it from an English/ British perspective, the answer will most likely be these three words - "1066, Norman Conquest." Now, as we'll discuss further later, there are some scholars (especially archaeologists) who would argue that both parts of that answer are wrong. But for the time being, lets just assume its correct. After all, this is what historians studying Anglo-Saxon and Norman England and medieval British castles have traditionally thought, and many still do, and it is also what secondary school history textbooks in the UK still present as fact.
So, if we're going by the traditional view, then anyone who is at all inquisitive or whose historical horizons stretch beyond the English Channel is going to be asking "well then, how did the Normans and their French allies get into this whole castle thing? When did they figure out that building them was a good idea?
Well, as the Chateau de Lourdon demonstrates, castles of a kind had existed in France for more than a century and a half before William the Conqueror set sail for England in September 1066. The exact "when" for the genesis of the castle is, however, murky and controversial, as we'll see. But everyone would agree that by the mid-eleventh century, castles were a well-entrenched and widespread phenomenon in France broadly defined, not to mention also in Italy and Christian Spain - Germany was a different matter. Therefore, we're mainly going to be thinking about France in the broadest sense here, with passing considerations of Italy and Northern Spain and the proviso that England and Germany (so often forgotten in these discussions) show that things can be different.
So what's behind it all ...
... Not the white heat of technology ...
So what was responsible for this "rise of the castle" in Continental Europe west of the Rhine between c.880 and 1060? Certainly not because of a a revolutionary technological breakthrough, as the technologies used in the building of castles really were ancient. In Northern France and post-1066 England, the most common type of castle was the motte and bailey. The technologies required to build a motte and bailey - defensive ditches, earthen ramparts, timber palisades and artificial mounds - had been around since Prehistory. An Iron Age fortified settlement like Maiden "Castle" in Dorset is testament to this, and there are plenty of sites like it from the middle to late first millennium BC across Western and Central Europe. And lets not get started on the Romans. The durability of Roman masonry meant that it was inside the still intact wall-circuit of the old Roman fort at Pevensey in Sussex that William the Conqueror built his first castle in England immediately after he landed on 25 September 1066. And closer to the period of the emergence of the first castles, both Offa's Dyke and the Danevirke demonstrate that eighth century northern European rulers were able to mobilise the manpower and resources necessary to build timber and earthwork fortifications spanning tens or indeed hundreds of kilometres. And for the ninth century, the Anglo-Saxon Chronicle and the Annals of Saint Bertin show the West Saxon, Viking and Carolingian Frankish armies hastily throwing up all kinds of fortifications like it were no biggie. So why do things that we can legitimately call castles (fortresses-cum-private residences) appear so late on the scene?
 |
| An aerial photograph from the west of the Iron Age hillfort at Maiden Castle in Dorset made shortly after is excavation in 1937 by Major George Allen (1891 - 1940), a hero of the lost age of gentleman amateur archaeologists. |
 |
| Pevensey Castle, West Sussex. The walls of the Roman fort, built in the 290s under Emperor Diocletian to defend against Saxon pirates, form the outer circuit. In the top right corner of the fort is the medieval castle, which was built in its present form by Peter II of Savoy, the uncle of Eleanor of Provence (Henry III's queen consort). The castle's Roman and medieval defences combined were strong enough to withstand a siege by Montfortian rebels in 1264 - 1265. Photo credit: By Pevensey_Castle_aerial_view.jpg: Lieven Smitsderivative work: Hchc2009 (talk) - Pevensey_Castle_aerial_view.jpg, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=17155846 |
 |
| The Danevirke. Back in 800, a timber palisade would have stood on its crest, but wood doesn't survive the centuries as well as stone, as anyone vaguely interested in archaeology knows. Photo credit: By Joachim Müllerchen, CC BY-SA 2.0 de, https://commons.wikimedia.org/w/index.php?curid=2473885 |
The military hypothesis So why do castles come about when they do if its clearly nothing to do with technological innovation. You might be thinking "well it must have been for military reasons. Fortifications are an essential part of what makes something a castle, aren't they not?" This is essentially taps into one of the biggest controversies in castle studies. Were castles, first and foremost, fortresses designed to protect their occupants and their property from attack? From about 1900 through to 1970, most historians and archaeologists looked at castles in terms of their defensive features and military functions. These were fairly easy to prove for, say, the Norman castles in England, the Crusader castles in Syria and Palestine (studied by Lawrence of Arabia himself, amongst others) or the Welsh castles of Edward I. In the last fifty years, however, the pendulum has swung the other way. Charles Coulson and his archaeologist disciples have argued that the military function of castles has been massively overstated - most European castles built between the eleventh and sixteenth centuries never actually experienced a siege - and that they were first and foremost about the display of power and status. Historians, in turn, have argued against this approach, leading to the so-called "battle for Bodiam" in the nineties and noughties.
 |
| Bodiam Castle, East Sussex, built in 1385 by Sir Edward Dalyngrigge after receiving a licence to crenelate from Richard II, looking absolutely stunning. Whether it was built to defend against the threat of French invasion, or so an upwardly mobile old soldier could live out chivalric fantasies and impress the local gentry, is the subject of much debate, and much too late to concern us here. Photo credit: By WyrdLight.com, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=7910287 |
But for the early history of the castle, there can be little doubt that castles had plenty of military potential in the early period (c.850 - 1150 let's say), in an age when trebuchets and professional military engineers, let alone gunpowder and standing armies, had yet to enter the scene. Conversely, an eleventh century motte and bailey or stone donjon lacked much potential for leisure and courtly display. And nostalgia and the pleasures of the imagination could have no part in the creation of the early castles, for the ideology of chivalry was still in its prehistory.A well-trodden path for explaining the origins of the castle goes thus: castles emerge on the Continent in the late ninth and early tenth centuries as defensive measures during the chaos wrought by the invasions of non-Christian peoples - the Vikings, the Saracens and the Magyars. People thus put their confidence in local aristocrats for leadership and protection, instead of a royal centre too distant and ineffective to be of much help to them against the invaders. After the invasions stop, castles find a new purpose - protecting these territorial aristocrats from each other. The result is a "fragmentation of powers (fractionnement des pouvoirs as the highly influential French medievalist Marc Bloch wrote it back in 1939), which gives rise to feudalism - a society with a landed warrior aristocracy, a subordinated peasantry and vertical ties, which at once unequal and reciprocal, being the key guiding principle behind everything.
On the surface of it, this explanation appears to work for France. The Annals of Saint Bertin, written by Prudentius of Troyes (d.861) and Hincmar of Rheims (d.882), don't mention castles at all. Only public fortifications built by kings and their agents feature, and politics revolves around assemblies, episcopal synods and royal visits to palaces and hunting lodges (implied to be unfortified). By contrast, in the Annals of Flodoard, covering the period 919 - 966, these sinister things, variously called castrum, castellum or arx, that barely register in the ninth century Frankish sources, just keep cropping up all the time, often in the context of kings and aristocrats laying siege to each other in them. What explains this, if not an increased sense of insecurity? To give another example, in 820 Archbishop Ebbo of Rheims quarried stone from the old Roman city walls to repair his cathedral. Come 885, as a Viking army laid siege to Paris, Archbishop Fulk of Rheims reinforced those same city walls using stones from a derelict church. The formula is simple - as the world became a scarier place, societal attitudes to fortification changed and people sought protection not by the king's armies but by whatever means they could come up with.
At a push, this might work for Italy as well. But for England, Germany and Christian Spain, it runs into complications. England is the one most often used as a counter-example. After all, as is well-known, England suffered just as badly from Viking invasions as did West Francia/ France. Yet it did not see any castle-building. Instead, the state kept the initiative and the West Saxon kings from Alfred the Great onwards built a network of fortified towns called burhs across their kingdom, with each burh being within a five mile radius of another burh, and expanded this network as they reconquered the Danelaw. These remained tightly under the control of the king and his officials and so, as the well-hashed comparison goes, while France fragmented over the course of the tenth and early eleventh centuries a unified English kingdom was made. Germany provides a fairly similar case to that of England. Ottonian kings of East Francia/ Germany managed to provide a formidable, unified response to the Magyar threat and prevent political fragmentation. There fortress-building remained under royal supervision, exemplified in the so-called burgenordnung King Henry the Fowler issued in 926. The kingdom of Leon-Asturias in Northern Spain is likewise one which didn't follow the French pattern and kings remained much more in control of both defence and reconquest against the Muslims, even if some regions like Castile (coincidentally named for its many border castles) did enjoy a long period of de facto independence in the tenth century.
 |
| A section of the surviving defences of the burh at Wareham in Dorset, built by Alfred the Great (probably in the 880s) to defend against the Vikings. Photo Credit:By Nigel Chadwick, CC BY-SA 2.0, https://commons.wikimedia.org/w/index.php?curid=13443049. Below is a map, created in 1888 by G.T Clark, of the surviving burghal defences, with the Norman castle later built within them dotted in red. |
 |
| A modern digital reconstruction of the fortified royal palace of Werla in Lower Saxony, built by King Henry the Fowler. In 924 or 926, he and his army hid there during a Magyar invasion, and then as the Magyars retreated, slowed down by all the loot, he led a force of cavalry to ambush the baggage train, and took a Magyar prince called Zoltan prisoner. With his hostage, Henry the Fowler was then able to agree to a nine year truce with Arpad, the Magyar grand prince. Taking advantage of the time this bought him to strengthen the East Frankish/ German kingdom's defences, the Burgenordnung was issued shortly afterwards. |
From such comparisons, one has good reason to wonder whether there was anything inevitable about the French situation, given that other Western European kingdoms went down such different trajectories in response to similarly serious threats from non-Christian invading peoples. It also makes one wonder whether military defence was really the reason for the rise of the castle in France. There are, indeed, very few surviving examples of private castles (as opposed to fortifications in general) that were built in greater France between c.880 and 960 with the explicit aim of repelling Viking, Saracen or Maguar attacks. And au contraire, by the late tenth and early eleventh centuries, the threats of Viking, Saracen and Magyar invasion had been seen off, yet castles were proliferating at a higher rate than ever before. For example, in Provence a hundred castles were built in the century after 930, and in the Auvergne and Limousin regions more than 150 were erected between 970 and 1020. Anjou experienced its biggest castle-building wave in the first third of the eleventh century under Count Fulk III Nerra (970 - 1040), who had no Viking invaders to fear and was busy expanding his power into Touraine against his archnemesis, Count Odo II of Blois-Champagne. Perhaps, therefore, military defence, at least against heathen invaders, was not the primary factor behind the rise of the castle, but rather power and authority, which brings us to ...
 |
| Loches Castle, the classic stone donjon of the eleventh century built by Fulk Nerra. Photo Credit: By Lieven Smits - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=8239026 |
Castles and the Early Medieval State
So now it is apparent to us that the first castles were able to thrive when the central government was not up to its job, that brings us to a question - was there anything that was actively stopping castles from emerging before they did? The answer that seems obvious is the state, by which we mean the royal court and its officers and agents in the localities. For West Francia/ France, it is made even more obvious by the fact that we have the Edict of Pitres of 864, which we explored last time. You may recall (and if you don't, scroll down to the very bottom of the linked post) that the Edict says that anyone who has built "castles (castella), walls (munitiones) or palisades (valla)" without being granted the king's express permission to do so, will have them demolished by the king's officials in the next thirty five days, with loss of office threatened for any count who neglects this command and harsh penalties for anyone trying to obstruct it.
The thinking goes that under a fairly strong, competent and energetic king like Charles the Bald, this state monopoly on building fortifications that the Edict implies could remain effective. But after the death of Charles the Bald in 877, his successors were less effective. His eldest son, Louis II the Stammerer, lasted only two years, and his younger sons, Louis III and Carloman, together only lasted another five. Then in 884, Charles the Bald's East Frankish nephew, Charles the Fat, took leadership of the kingdom. Yet Charles proved himself to be not quite up to the task of ruling three realms (West Francia, East Francia and Italy) at once, especially when the Viking challenge was entering its most intense phase, and in 888 he died of a (likely) stress-induced stroke after news reached him of the rebellion of his nephew, Arnulf of Carinthia. The West Frankish nobility then elected, Margrave Odo of Neustria, the first non-Carolingian king in almost 150 years. Odo ended up fighting a civil war from 893 to 897 with Charles the Simple, the son of Louis the Stammerer who had been born one month after his father's death in 879, and died a broken man in 898. Charles the Simple was then deposed in 922 by Odo's brother, Margrave Robert of Neustria. Charles then defeated and killed the usurper, Robert I, at the battle of Soissons on 15 June 923, only to then be betrayed and imprisoned by his cousin, Count Herbert of Vermandois. The West Frankish nobility then elected Duke Raoul of Burgundy as their king, who spent a significant portion of his thirteen year reign in open warfare with that same Herbert of Vermandois because the latter was trying to deprive him of his right to appoint the most important cleric in the realm - the Archbishop of Rheims.
As the conventional narrative runs, by the time of Raoul's death in 936, there had been a drastic shift in the balance of power between kings and aristocrats. Charles the Simple's son, King Louis IV, who returned that year from living in exile with his maternal uncle, King Athelstan of England, spent more than half of his eighteen year reign in open civil war with his most powerful subject, Hugh the Great, the son of Robert of Neustria. Only with the help of his brother-in-law, King Otto the Great of East Francia/ Germany, was Louis IV able to finally bring Hugh the Great to heel. Louis' son, Lothar (r.954 - 986), and grandson, Louis V, though the latter only lasted a year as king before his premature death, fared somewhat better. After the Carolingian line of West Frankish kings terminated in 987 and nobility elected Hugh Capet, the son of Hugh the Great and grandson of Robert of Neustria, as king. The new king struggled to arbitrate disputes between his magnates like Count Odo II of Blois, who claimed he had no authority to intervene in their private feuds, and a lippy count once said to him "who put you in charge." Hugh's son, King Robert II the Pious (r.996 - 1031) spent most of his reign fighting to assert direct rule in Burgundy. By the time of his death, the kingdom of West Francia/ France was nothing more than a confederacy of semi-independent principalities, many which had themselves fragmented into much smaller autonomous lordships, and the king only exercised direct authority over the Paris region which he seldom left. For almost 150 years, West Frankish/ French royal authority was on a constant downwards spiral until it hit rock bottom, so the traditional way of telling it goes.
 |
| Map of France in 1030 - areas in turquoise are the only ones the king had any direct control over. |
It is thus therefore assumed that as royal power became progressively enfeebled grasping magnates took advantage of the growing power vacuum by building castles, which the central government could no longer stop them from doing, thus tipping the balance of power even further towards their end. From these castles, so this school of thought goes, they were able to secure iron control over large hereditary territories, subordinate and exploit the free peasantry, wage endemic warfare against neighbouring magnates and defy royal authority as and when they saw fit. The fact that the subordinates of these magnates were given control over castles and were capable of building their own meant that effective political power was able to devolve down to the level of castellans and simple knights in charge of a few villages.
 |
| The castle of Chateau Doue de la Fontaine in western France, built c.950 on the site of an unfortified Carolingian era aristocratic residence, once owned by King Robert I (d.923), by Count Theobald I of Blois, one of Hugh the Great's vassals. |
A remarkable document produced in Aquitaine in c.1020 called the Complaint of Hugh the Chiliarch, exemplifies this well. Almost all the action takes place within the county of Poitou, making it an extremely localised world. The count of Anjou, Fulk III Nerra (970 - 1040), appears to be essentially a foreign power, and the king of France, Robert the Pious, is never mentioned at all. The political actors that feature in it, including the certain Hugh de Lusignan for whom it was written, are almost all minor castle-holding lords constantly locked in disputes and vendettas with each other, which they pursue by building new castles and laying siege to and capturing/ destroying their opponents castles. The Count of Poitou and Duke of Aquitaine, William V (d.1030), tries to impose some measure of order over his fractious vassals, but in order to do this he has to essentially play divide and rule with them. Whichever party he sides against has to have their castles demolished or besieged.
 |
| The Chateau de Lusignan, first built in the tenth century and the principal residence of Hugh de Lusignan, representing the month of March in the Very Rich Hours of the Duke of Berry (c.1414). By the early fifteenth century, the castle had been rebuilt and modernised beyond all recognition from whatever it would have looked like in Hugh de Lusignan's day, though it still occupied the same site, and it was now in the possession of Duke Jean de Berry, uncle of Charles VI of France (r.1380 - 1422), the mad king of Shakespeare's Henry V. In the intervening period, the Lusignans had moved on to bigger and better things, like being one-time kings of Jerusalem (if you've watched Kingdom of Heaven you'll know who I'm referring to), earls of Pembroke and kings and queens of Cyprus. |
Around the same time, the counts of Barcelona, who had managed to keep public order and justice going in their principality and launch successful military expeditions as far as Cordoba, the capital of the Caliphate itself, under Count Ramon Borrell (d.1017), had their power undone by the castle-holding dependents along the Muslim frontier that they themselves had created. While by the 1060s, the counts of Barcelona had recovered some measure of control over their principality under Count Ramon Berenguer I (d.1076), all the old motors of public government they had back in the 1010s were lost. Instead, they derived their power and authority by receiving oath-takings and homage from these castle-holding lords, whose activities they could not police - the lords would be their faithful vassals and give them military service in return for their fiefs and castles. Indeed, it was in Catalonia, in the time of Ramon Berenguer that the first book of "Feudal Law" was created - the Usages of Barcelona.
 |
| The tenth century Castell de Miralles near Barcelona. Photo credit: |
By Antoni Grifol - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=18264882
 |
| Scene from the Usages of Barcelona showing Count Ramon Berenguer I and his wife Almodis de la Marche counting out 2000 ounces of gold as payment to Count William Raymond of Cerdanya in return for rights over the castle of Carcassonne. Ramon Berenguer's son, Count Berenguer Ramon the Fratricidal (d.1097), notably fought against one of the most famous knights ever, Rodrigo Diaz de Vivar, also known as El Cid, and was captured twice - indeed, I've known about Berenguer since I was 8, when I played the El Cid campaign in the 2000 Real Time Strategy game Age of Empires 2: The Conquerors where he's one of the major baddies and you have to destroy his castle in one mission. |
In the north of France, Duke William the Bastard of Normandy, who famously built a number of castles of his own, spent much of his youth laying siege to rebels' castles, destroying the castles that the Norman barons had built during his troubled years of his minority (1035 - 1047) and attacking, capturing and destroying castles belonging to enemy frontier lords. The first book of the Deeds of William the Conqueror, written by William of Poitiers in the 1070s, is full of this stuff.
 |
| Remains of the early eleventh century stone donjon at Brionne in Normandy, home to William the Bastard's first guardian during his minority, Count Gilbert of Brionne (d.1040). Photo credit: By Eponimm - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=51117438 |
 |
| William the Conqueror's forces besiege the Breton border castle of Dinan in the Bayeux Tapestry. The castle appears to have a wooden keep, a motte and an outer rampart connected by a bridge, but apparently no bailey! |
And these are all the regions that were still lucky enough to have an overarching leader. In a region like Normandy's near neighbour, Picardy, power went to whoever had the most castles, as Hariulf (1060 - 1143), a monk of the abbey of St Riquier, wrote:
"Because they were not all lords of castles, he [Hugh of Abbeville] became more powerful than the rest of his peers. For he could do what he liked without fear, relying on the protection of the castle, while others, if they tried anything, were easily overcome as they had no refuge."
A similar situation existed in the region of Berry. There the Bishop of Bourges in 1038 had to create a militia of free peasants to go around destroying castles in order to restore peace to the region. In the end it was of no avail, as the local aristocracy retaliated by sending a force of mounted knights, which the peasant militia was no match for and they all ended up slaughtered.
These kinds of anarchic situations are essentially the worst nightmare of Charles the Bald, Hincmar of Rheims and the other leading luminaries of the West Frankish kingdom involved in the making of the Edict of Pitres. Indeed, the Edict itself complains of the "many difficulties and robberies" that these mid-ninth century prototype castles are creating. To many scholars, therefore, the Edict of Pitres seems like a remarkably prescient warning of the storm that was to come, and that had Charles the Bald's successors been stronger kings, maybe the Edict of Pitres would have held, the state would have kept its monopoly on fortifications and all this mess would have been avoided. Simple processes of comparison outside France seem to validate this as well. When kingship (or whatever public power there was in the region) failed, as it did in Italy and Lotharingia at so many different points in the tenth and eleventh centuries, in Germany in the 1070s, in Galicia and Leon in the 1110s and finally in England in the 1140s, the realm would descend into a spiral of unfettered castle-building as aristocrats strove to out-do each other and achieve greater local domination.
 |
| Baronial anarchy in the Home Counties: Abinger Motte, Surrey, first built c.1100 by William FitzAnsculf, rebuilt in the 1140s and abandoned in 1153. |
Its even been argued that the memory of the Edict of Pitres was clung on to in some quarters, and inspired putative various attempts to regulate fortifications. A clause in the Customs and Judgements of the Dukes of Normandy, issued in the 1090s, appears to suggest that the Dukes claimed the right to garrison the castles belonging to their barons. Two clauses in the Laws of King Henry I, issued c.1115, forbid castellatio without licence. However, the text's author mysteriously glossed that word as "fortification of three walls", which would imply not an attempt to regulate fortified dwellings per se, but one to prevent them from being activated for war by building an additional outer rampart or ditch. Nonetheless, both of these have been interpreted as signs of Carolingian nostalgia in the Anglo-Norman realm. The Edict of Pitres is also thought to be the ultimate inspiration behind the systems of royal and princely licencing of fortifications, which first properly appear in the sources for England, France, Germany and the Low Countries in the decades around 1200. Thus, as Simon MacLean puts it, the Edict of Pitres has so often been seen by twentieth and twenty-first century historians as something of a "skeleton key" or "Rosetta Stone" to cracking the early history of the medieval castle.
Attractively simple as all of this might seem, sadly its not. For one, it rests on an incredibly dated, essentially pre-WW1, school of historical thought in which strong kings are inherently a "good thing" and powerful aristocrats are inherently a "bad thing." It is thus presumed that the two were mutually antagonistic - kings strove for a weak and pliant aristocracy so they could bring peace and order to the realm, while aristocrats wanted a weak and distant monarchy if not going it alone all together so they could rule the localities like petty tyrants. Probably the biggest revolution of in our understanding of how medieval politics worked prior to the thirteenth century has been the development, mostly in the last fifty years, of the consensus model.
 |
| Where the traditional approach to medieval political history met its ludicrous parody - in Sellars and Yeatman's tongue-in-cheek humorous narrative, everything was either a "good thing" or a "bad thing", and it was the barons'grand design to revive the "feudal amenities of sackage, carnage and wreckage", which they got in the Wars of the Roses and "staved off the Tudors for a while." |
Under this school of thought, rather than being locked in eternal struggle, kings and aristocrats needed each other. Aristocrats needed to kings to provide them with war leadership, justice and dispute arbitration, patronage and the like, while kings needed the local power, knowledge and connections of aristocrats in order to further their state-building activities. Indeed, as we saw in the post on the Edict of Pitres, which was formulated at a royal assembly - the ultimate physical manifestation of consensus politics - Charles needed the advice and close co-operation of his aristocracy for his governmental reforms to succeed, and many of those government reforms provided solutions to the problems aristocrats wanted solved. And as can be argued of both the rebellions and civil conflicts that Charles the Bald himself and his successors experienced, they were not so much because the aristocracy wanted to enfeeble the system but instead have it work better for them.
There's also the issue of whether the Edict of Pitres has itself been misinterpreted. Nowhere else in the ninth century sources do we find Carolingian kings claiming a monopoly on the right to build fortifications. Nor do any rulers after Charles the Bald claim such a right until towards the end of the twelfth century, by which the dangers of unfettered castle-building really were obvious from centuries of experience. Even then, some historians like Charles Coulson have argued that late medieval English licences to crenellate weren't so much about maintaining public order and civil peace, as they were a form of political patronage granted by the Plantagenet kings to middling landowners and their favoured servants so that they could lay claim to noble status.
Simon MacLean argues that, in fact, if the edict is read closely and in proper context, its not actually asserting the state's right to a monopoly on fortifications, or warning against impending feudal anarchy if castles continue to be built without royal permission. As he sees it, the Edict reads more like an attempt by the state to build consensus among the aristocracy and mobilise as much manpower and resources as it can towards governmental reform and common defence against the Vikings. Private building projects distracted aristocratic officials and the peasant labour force from what they really should be working on - public fortifications to defend against the Vikings. MacLean points to how Charles the Bald had to threaten to call on divine retribution in order for the fortified bridge at Pont de l'Arche to be completed in 862, and how when he claims that the state has the right to call on peasants to work on public fortifications he has to appeal to "the custom of other nations" and the Theodosian Code, not the capitularies of any previous ruler. What was going on here, MacLean suggests, were extraordinary measures that required a lot of heavy-lifting through appeals to the Roman past and consensus-building.
MacLean's arguments, contained in his very recent article "The Edict of Pitres, Carolingian Defence Against the Vikings and the Origins of the Medieval Castle", Transactions of the Royal Historical Society, (2020), might be a little too sceptical and against the grain for some, and we have yet to see how other historians will respond. But he does quite successfully demonstrate how circular much of the reasoning is on this issue. While its true that the written sources from the ninth century, seldom ever mention castles, archaeology gives a different picture. MacLean points to how many fortified sites of varying sizes that can be dated with some precision to the ninth century have been found in northern and western France. However, because of the Edict of Pitres, it is presumed that these must have been public fortifications built by the king and his officials, otherwise they would have been destroyed and left no trace. A classic example to demonstrate the adage, well known among ancient and early medieval historians, about using archaeological evidence: "Texts can lie to you. Spades don't lie, but you have to make them speak."
Similarly, the archaeological evidence tells a different story for Anglo-Saxon England, where it is thought that there were no castles until after the Norman Conquest because the West Saxon kings kept a monopoly on fortifications. This view is reinforced by a small number of canonical written sources. The Anglo-Saxon Chronicle uses Castell (evidently a word of Latin and French import, not an Old English one) to refer to the fortifications on the Welsh borders that were built in the 1040s and 1050s by Earl Ralph the Timid of Hereford, Edward the Confessor's French nephew, and his followers, and generally seems to view them as dangerous foreign innovations. Orderic Vitalis likewise says in a passage famous among Anglo-Normanists that there were "very few" castles in England before 1066. What exactly he meant by these "very few" castles is, however, not entirely certain. Did he just mean the castles that Edward the Confessor's French followers had built in the two decades before the Conquest, in which case there were very few indeed on the eve of the Battle of Hastings. Or was he perhaps alluding, with pardonable exaggeration, to something else. You see, archaeologists in the last fifty years have excavated a number of manor houses with ditches and earthwork ramparts dating to the late tenth and early eleventh centuries. The most notable examples are those at Goltho in Lincolnshire, Fowlmere in Cambridgeshire and Sulgrave in Northamptonshire.
 |
| The very overgrown earthworks of Richard's Castle in Herefordshire, a motte and bailey built by Richard Scrob, one of the Norman knights that Edward the Confessor had invited over and given lands in England to sometime before 1051. Very much one of those castles you could easily not notice on a country walk in the vicinity, unless you were actively looking for it. Photo Credit: By Raymond Perry, CC BY-SA 2.0, https://commons.wikimedia.org/w/index.php?curid=9179141 |
 |
| Ewyas Harold castle in Herefordshire, a motte and bailey built in 1048 by Osbern Pentecost. Photo Credit: By Philip Halling, CC BY-SA 2.0, https://commons.wikimedia.org/w/index.php?curid=15194102
|
 |
A highly imaginative reconstruction of what the late Anglo-Saxon manor house at Goltho looked like, made by Guy Beresford, the archaeologist who excavated it in the 1970s. Photo credit: Jonathan Jarrett. Minus the motte, it would indeed look remarkably like an early Norman castle.
Map of Fowlmere, Cambridgeshire, in 1847 showing the ditch and bank enclosure (like that at Goltho), created sometime between c.1000 and 1050, which was excavated in 2002. Image credit: John Blair. |
Some might quibble at the interpretation of these sites as proto-castles and question how typical they are. We don't know a great deal amount about pre-conquest English aristocratic residences, as indeed is also the case with ninth century Carolingian Frankish ones. But what we do know is that they were expected to be quite grand indeed. The laws of King Cnut (c.1018), as glossed by Archbishop Wulfstan of York (d.1023), say that in order for a ceorl (free peasant) to be considered a thegn (lesser nobleman), he must own five hides of land, a hall, a kitchen, a proprietary church, a bell-house and a burhgeat. The meaning of the latter two is somewhat ambiguous. But Ann Williams has convincingly suggested that the term burhgeat means fortress gate, a pretty solid indication that late Anglo-Saxon manor houses were expected to be fortified by ditches, ramparts and gates, some of which may have been built in stone. And while the number of these early eleventh century Anglo-Saxon fortified manor houses that have been excavated is currently fairly small, more may yet be discovered underneath early Norman castles, as the ones at Goltho, Fowlmere and Sulgrave were. Indeed, as Simon Forder has helpfully suggested, the fact that the Normans were able to put up so many castles in such a short space of time, with such limited manpower during a military conquest means that they must have been making extensive use of pre-existing fortified sites. Some of these obviously included Alfredian burhs (like at Wareham and Oxford), Roman forts (Pevensey) and even Iron Age hillforts (Old Sarum), but they can't have been the only ones. Thus, archaeologists, who tend to be suspicious of grand narratives (except for prehistory, where they're the ones who write them), have argued that there was a steady growth in fortified sites of all kinds in England in the period 900 - 1050, as there was everywhere else in Western and Northern Europe, and that all the Normans did in 1066 was bring in mottes and donjons, which would have made their way over to England anyway if given another generation or two. How much of this tenth and early eleventh century growth of fortifications in England was encouraged or opposed by the late Anglo-Saxon state, we are simply not in a position to know.  |
| The late Anglo-Saxon stone tower of the city defences at Oxford (founded as a royal burh by Edward the Elder sometime before 918) in the background to the left of the Norman motte (constructed in 1071 by Robert d'Oilly, sheriff of Oxfordshire), taken on a November evening back in 2019. |
 |
| Other regions we haven't spoken of much were also busy building fortified enclosures as the first millennium drew to a close. Above is a reconstructed tenth century Slavic refuge castle at Raddusch in Brandenburg, near the the present day German-Polish border. Photo Credit: By A.Savin (WikiCommons) - Own work, FAL, https://commons.wikimedia.org/w/index.php?curid=64118216. Below is a "ring castle (trelleborg)", near Slagelse on the Danish island of Zealand built c.980, possibly at the orders of King Harald Bluetooth though we can't be sure of it. Photo Credit: By Thue C. Leibrandt - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=30522746 |
"Its the economy, stupid", or the culture of power?
So maybe the frontiers of the Carolingian and Anglo-Saxon states need to be rolled back a bit. But at the same time, there's no proof that castles were part of any grand conspiracy on the part of the aristocracy to undermine the state, even if castles did ultimately contribute to temporary collapse of all public authority and the emergence of revolutionary, new forms of political organisation (what textbooks still call feudalism) everywhere in Europe by the mid-twelfth century at the latest.
One suggestion is that it comes down to how the economy works really. Under the Roman Empire, the elite prospers from owning vast landed estates worked by hordes of slaves, from holding public office in the provinces and from the patronage of the imperial court. This keeps the aristocracy very much in favour of a strong state. The Western Roman Empire of course falls apart in the fifth century and taxation has largely dried up by the early seventh century (unless you're in Spain) - whatever happens to the slave system in continental western Europe deserves a post in itself. So why aren't castles already a thing in the late Merovingian period, when kings were arguably much weaker than they were in tenth century West Francia and some of their mayors of the palace, like Ebroin, were seriously close to destroying consensus politics. The simple answer is that, after centuries economic and demographic decline arguably going back as far as the Antonine Plague (165 - 180 AD) and the Crisis of the Third Century (235 - 284), the rural economy is too poor and underdeveloped for aristocrats to go it alone. Instead, they need the state, which still has lots by way of land, moveable treasure and offices to give out as patronage. Indeed, most of the factional struggles and civil wars of the period from 656 - 721 in the Merovingian kingdoms can be explained away as different groups of aristocrats trying to gain access/ control of the patronage system. Eventually the Carolingians and their faction completely won out and took over from the Merovingians, and they went about leading the Franks into a series of expansionist wars that would give them a bigger patronage base from plunder and confiscated land. Frisia, Burgundy and Provence are conquered under Charles Martel. Alemannia, Septimania and Aquitaine follow under Pepin the Short. Then Saxony, Italy, Bavaria and the Spanish March under Charlemagne. After 803, however, pretty much all conquests come to a halt, and the Carolingians become faced with the problem of managing increasingly finite resources and the patronage base shrinks. Meanwhile, the rural economy starts to experience an upswing - there's a growth in agricultural production and population, rural and urban markets are proliferating (the Edict of Pitres would support this), there's a lot more cash going around and rural settlements are starting to become less dispersed and more concentrated around nuclei like parish churches. The conditions for building local powerbases are thus set, and there's everything to be gained from subjugating and exploiting the peasantry, which are as it stands legally free and relatively unburdened either by the state or by landlords. The aristocracy thus need the political structures of the kingdom increasingly less to guarantee their wealth and position. Therefore, they can instead start to shift towards becoming territorial princes and local lords, and castles, which they can definitely now afford, are both end and means in this quest for local domination. And so, by the middle of the eleventh century, the rural economy is booming, castles have proliferated and the formerly free peasantry are now serfs.
Perhaps one could also posit another socio-economic factor behind the rise of castles - the need for upwardly mobile local bigwigs, prospering from economic growth, to demonstrate their status against better established elites. After all, Archbishop Wulfstan made his comments about the "bell-house and burhgeat" in the context of what a prospering free peasant needed to do to claim thegnly (lesser noble) status and perhaps it worked similar on the Continent in the tenth and eleventh centuries. Indeed, one of our biggest weaknesses in all of this is that so little is known about elites in the ninth century Carolingian kingdoms below the level of the high aristocracy.
 |
| Fresco of a donor from the church of St Benedict at Mals in the South Tyrol region of Italy, dated to c.800. This man was clearly elite, but how elite exactly we can't be sure. |
Yet the problem with these kinds of explanations are many. The first, at the most abstract, theoretical level, being accusations of economic determinism. The second is that the data for economic and social change is very localised and fragmentary. For example, rural social structures for northern France very well documented in the early ninth century polyptychs, whereas in Catalonia the evidence for this stuff only really gets going after 880 but then it becomes the most prodigiously documented region of western Europe in the tenth century. With something like early medieval grain yields, our data is so inadequate that all we can really say with confidence that grain yields were higher in the twelfth century than what they'd been in the second century AD, but whatever happened in between is highly disputable. Likewise some people see economic growth and the beginnings of peasant subjection to aristocratic lordship starting as early as the seventh and eighth centuries in some regions, including the Frankish heartlands between the Loire - way too early for them to be connected in any way to the rise of castles. Likewise, we have not much real data to prove the depletion of the Carolingian state's aggregate fiscal resources. And indeed its possible to argue that the Carolingian state was a lot more dynamic in acquiring state resources than we give it credit for. It seems to have been trying to redevelop centralised taxation from the Edict of Pitres, and certainly hard to imagine that the lands a ninth century king like Charles the Bald granted out as patronage were simply the final remnants of the fiscal estates of the Western Roman emperors. Indeed, as the Edict of Pitres' coinage reform and other economic measures show, with the right kinds of mechanisms and regulations in place the Carolingian state could have successfully ridden the tiger of economic growth. And if the drying up of plunder sowed the seeds of the Carolingian state's demise, why did it take multiple generations after militarily expansion ceased at the beginning of the ninth century for it to happen. Undoubtedly local economic growth and changing social structures facilitated the rise of the castle, but fitting it all into a grand narrative creates a great many complications and pitfalls.
Finally, there's culture to consider. The martial self image of the aristocracy itself certainly doesn't explain it, as the aristocracy had been pretty thoroughly militarised since the fifth, sixth or seventh centuries, depending on where in western Europe we're talking about, and yet castles don't appear until the ninth century at the earliest. And while other societies with warrior elites have produced castles, like Feudal Japan, many others have not.
 |
| Himeji Castle in Hyogo Prefecture, Japan, built in 1333 by the samurai Akamatsu Norimura. Photo Credit: By Niko Kitsakis - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=40135622. I must confess I know absolutely nothing about how Japanese incastellamento worked. But since we're being encouraged global, I thought I'd put this there without even a hint of tokenism. |
What's maybe more to the point is specifically a change in political culture. Where do the elites see themselves as getting their power from, where do they think is the appropriate forum for exercising it and to what end? And here we might have something of an answer. The Carolingian, Ottonian and Anglo-Saxon elites (can't claim to know anything about their Asturian, Leonese and Castilian counterparts), I would argue, saw their power as coming from loyal service to kings, public office holding, participating in assemblies and the royal court, ancestry, widely spread landed interests and pushing forward the wider Carolingian project of religious and moral reform. In the case of Carolingian aristocrats, we have good evidence to show that the majority of them were literate - they were definitely not hairy, mead-swilling primitives. And while the aristocracy did often rebel against the Carolingian kings from the reign of Louis the Pious on, it was always done with the aim of either correcting the ruler or replacing him with another member of the Carolingian family. Rebellion was never about trying to weaken the power of/ break away from the royal centre. Similar patterns can also be seen with the various rebellions against the Ottonians, and, for Anglo-Saxon England, with the rebellions of the House of Godwine and the House of Leofric in the reign of Edward the Confessor. Notably, retreating into private fortified sites was an option seldom ever taken in these rebellions. Indeed, as Simon MacLean has pointed out, Carolingian aristocrats seem to have associated hiding away in fortresses with losers and cowards, and Carolingian writers very frequently describe Slavic chieftains and the leaders of other foreign peoples doing this, to contrast them unfavourably with the Franks.
It is therefore arguable that the rise of castles stemmed from a greater sense of aristocratic independence - that their power came not from the state but from something else, tied to territory, family and their status as warriors, for which the castle was the physical manifestation. Yet what exactly caused this shift in cultural mentalities we have no clear answer yet. Meanwhile, there's another factor we still haven't considered.
Military hypothesis 2.0
A final suggestion might be that the shift towards castles might have been facilitated by changes in warfare and military organisation. Castles work very well when you're dealing with fairly small enemy forces, as they're not in a good position to besiege and overwhelm you - after ravaging the countryside for a bit, they'll be forced to return home (a classic pattern in high medieval warfare). Larger forces can present more problems if you don't have a strong garrison yourself and depending on the size and defensive structures of the castle. So could armies becoming smaller and more elitist have had something to do with the rise of the castle?
Unfortunately, attempting a solution to the problem like this creates too many of its own. The period in western military history between the fall of the Western Roman Empire (c.450) and the High Middle Ages (1000 - 1300) is one of the murkiest and most neglected of them all. Among the most controversial questions in this field are "how big were armies" and "how were they raised?" Obviously the exact nature of the question, and the answers to it, depend on what bit of the period and which part of Europe we're talking about. But for the Carolingian kingdoms, the question is essentially this: "was the Carolingian army for the most part a militia of yeomen farmers, or was it all leisured landlords and their household goons." These positions are in turn closely linked to views on army size - those who think armies were typically in the low thousands will favour the latter position, while those who think they could reach the tens of thousands will favour the former. The former position would appear to be supported by the Edict of Pitres, and earlier military legislation issued under Charlemagne in the last quarter of his reign and Louis the Pious. The problem is, however, that such legislation only appears at the beginning of the ninth century, and many historians such as Guy Halsall, argue that it cannot be backprojected onto the eighth century and the time of Charlemagne's wars of expansion. They are also sceptical about the legislation's effectiveness. Historians like Matthew Innes and Timothy Reuter see Carolingian armies as small, elitist groups raised through aristocratic patronage networks, while Halsall sees the Carolingians as trying various kinds of experiments to get away from this kind of set-up that ultimately don't succeed. On the other hand, at the other end of the spectrum historians like Bernard Bachrach argue that the Carolingians had substantial professional standing forces and even larger ones of free peasant conscripts from a manpower pool of two million able-bodied adult men, and thus he sees the Carolingian kings very regularly commanding military forces of 20,000, 40,000 or even 100,000. Of course, these positions in the debate are highly dependent on how one views how the Carolingian state worked. And the debate can get very nasty and ideological - Bachrach accuses the likes of Innes, Halsall and Reuter of being "Procrustean Marxists" and "doctrinaire slaves to a primitivist agenda", while Halsall accuses Bachrach back of having neoconservative political sympathies and writing speculative fiction, not history. I haven't yet formulated positions of my own on these issues, so I'm not going to pass judgement on whether either side is right or wrong, which gets us nowhere in using trying to use changes in military organisation as an explanation for the rise of the castle.
All that can be said, is that a large, well-armed free peasant militia could indeed be highly dangerous to early castles. We already mentioned the Peace League of Bourges in 1038. But similarly, when the Saxon nobles rebelled against Emperor Henry IV of Germany in 1073, they attacked his castle of the Harzburg that he had built in 1066 with a very large force of Saxon free peasant militiamen and managed to utterly overrun and destroy it. When Henry IV defeated the Saxon rebels in 1075, his knights and mercenaries terrorised the Saxon peasantry in revenge, and the Saxon nobility became hesitant to raise such militia forces again.
 |
| A reconstruction of one of the towers of the Harzburg. Photo Credit: |
By User Schmull on de.wikipedia - Schmull, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=899439
Conclusion
Overall, after having this long debate with myself, I'm afraid to say that I've come to no firm conclusions about the origins of the castle but exhausted a lot of possibilities. Certainly, the Chateau de Lourdon doesn't appear to fit clearly into any of those theories about the origins of the castle, and itself points to a glaring problem - the scantiness of what the written sources can tell us about when, why and by whom the earliest castles were built. What we can be sure about is how important the Chateau de Lourdon was to the monks of Cluny in the display and exercise of their power and lordship in the Burgundian countryside, as well as defending their estates, a role similar to that which their other fortifications played. It was for those purposes, combined also with the demonstration of aristocratic status, that castles would have such a vital role to play in medieval society right the way through to the end of the Middle Ages.
 |
| The Tour des Fromages (eleventh to thirteenth centuries) in the Bourg de Cluny. Built by the monks to remind the townspeople who was boss around here. |
 |
| The fortified monastery precinct of Cluny |
 |
| The Castelvecchio at Verona, Italy - one of my favourite of the castles I saw on my travels. Built by Mastino II della Scala, the lord of Verona, in 1351 when he felt too unsafe to live in an unfortified urban palazzo due to his unpopularity with the citizenry. |
