An excerpt from
Castles, Battles, and Bombs
How Economics Explains Military History
Jurgen Brauer and Hubert van Tuyll
The Case of the Medieval Castle
|Castle||Years building||Cost (£)|
Edward I’s annual income over twenty-five years ranged from less than £25,000 to over £100,000, with an average, including a variety of special taxes and parliamentary subsidies, of around £67,500. Unfortunately for Edward, not to mention the taxpayers, the fortifications outran his finances. Caernarfon, the most spectacular site, never was completely finished inside. “In Wales, Edward’s great castle at Beaumaris remained unfinished, a striking witness to his financial problems.” The castle’s walls only reached half their intended height. The treasury could no longer pay the wages of the work crews at Beaumaris. Even heavy levies in Ireland could not make up the difference. As parliament’s great magnates became unexcited about disgorging endless funds for war, a constitutional crisis resulted. No British monarch would ever attempt castle building on this scale again. Perhaps later rulers, or their advisors, were all too aware of the crisis that occurred in 1296–1297. Although not the sole cause, the magnificent walls of Wales perhaps were the single biggest factor:
Now this financial crisis was not altogether of sudden growth: for although it had been hastened since 1294 by the coincidence of almost simultaneous wars with French and Welsh and Scots, it was largely the slow product of the costly enterprises of Edward’s reign as a whole. Among those enterprises, none had more steadily devoured his treasure than his castle-building in Wales.
The nobles’ love of the king’s masonry was doubtless tempered by their awareness that projects on this scale were no longer within their own grasp. A handful of nonroyals in England and on the Continent could afford big walls. Gilbert de Clare built the enormous castle at Caerphilly. In France, the counts of Champagne bought castles from financially strapped nobles. Of two hundred English baronies in the twelfth century, only 35 percent had castles, others making do with fortified houses. Between 1154 and 1214 the number of baronial castles actually fell, from 225 to 179, while royal fortresses more than doubled in number, from 45 to 93. In an age when royal income never fell below £10,000, only seven barons were earning more than £400, the average was about £200, and at least twenty had incomes of less than £20. An up-to-date castle at the end of the twelfth century cost about £1,000 although a basic stone castle could be built for £350.
The nobles were at a disadvantage in the castle arms race with regard to both money and legal powers. The king had the greater ability to pay and also the greater ability to make people fulfill his wishes. The king could force a shire to cover significant military expense. The nobles were in a relatively poor position. While the king had begun in England, as elsewhere, as a primus inter pares, he was now very much primus. Yet as Edward discovered during the crisis of 1296–1297, he was definitely not free of fiscal constraints and hence remained somewhat dependent on the wishes of his nobles.
New weapons often increase the capability of the individual soldier, but this can be offset by growing development and construction (or manufacturing) costs. In this respect, the stone castle was fairly typical. The number of soldiers needed to defend it might be small, but the number of people needed to build it was quite large. In the days of the motte and bailey castle, the labor required for building a very basic fort was small. A hundred men working for a month could build a modest one, and if they labored about three months, a substantial mound could be constructed. The stone castle was another matter. The tower constructed at Langeais required about 83,000 “average work days” to construct, ten times what a substantial motte and bailey needed. Langeais was built in two seasons. If the workforce was able to labor six months each year, there would have had to be some three hundred laborers continuously on site. The labor of an additional 1,000 to 1,200 farmworkers would have been needed to sustain the builders.
Edward I’s castles required much more (fig. 2.1). None of his fortifications were completed in less than five building seasons (six to seven months each), and three took much longer. Many of the works were done at the same time. After 1277, Builth, Aberystwyth, Flint, and Rhuddlan were being built at the same time; after 1283, Conwy, Caernarfon, and Harlech were going up together. The number of workers at each castle varied enormously. At Builth only a hundred might be present at any one time, but at Beaumaris the total exceeded 3,000. In 1283–1284, about 4,000 men were employed at Conwy, Caernarfon, and Harlech, while 3,500 were hired in 1295 to work on Beaumaris and repair Caernarfon (which was damaged in an uprising). The population in Edward I’s day was about three to four million people, so that the use of 4,000 workers was not a negligible number (an equivalent number of Americans today would be about 300,000). Workers had to be brought in from faraway shires. “In other words, we seem to be driven to the conclusion that the numbers of workmen employed by Edward I on his Welsh castles were sufficiently great to constitute a noticeable draught upon the total pool of mobile labour that was available in the England of his day.” That, of course, was the rub. With much of the population engaged in subsistence agriculture, the pool of labor on which to draw was not massive. Designers and masons were hired from the Continent. This reduced the labor drain in England (although at a financial cost) yet the castle-building program may, nevertheless, have absorbed a tenth of the national workforce. The impact was greater than army recruitment would have been. “The recruitment of men for the army may have had similar effects on occasion, but probably more serious was the call for skilled workmen to build the great Welsh castles, which must have had its effect on the building industry.” While that conclusion cannot be accepted in its entirety—church construction may have contributed to a surplus of available stone masons—the burden was still huge. The true experts in the business, the architects and designers who knew the military ins and outs of castles, were greatly valued and greatly compensated, excepting the unfortunate architect of Ivry (France) who was put to death so that he could not reveal its secrets.
Building a castle was only a first step. Castles had to be adapted, modified, improved, or even replaced if they became obsolete. Improvements in siege warfare required design changes. The square stone keep was sufficiently expensive that Norman castle-owners sought simply to reinforce their motte and bailey castles to hold costs down. Increasing costs of fortifications drove increases in military expenditure throughout the eleventh century. The crusaders further stimulated the process by bringing back innovations from the Middle East. The motte and bailey castles mutated from inexpensive palisaded hills to costly stone fortifications as powerful concentric walls replaced the palisade. The outer “curtain wall” became more sophisticated in the twelfth and thirteenth centuries as towers were added for flank fire. By the time castle building reached its peak toward the end of the thirteenth century, the two-wall system, with a “killing ground” in between, had become a standard design. This did nothing to constrain the continuing expansion of costs. There was increasingly no limit on how much could be spent, how many improvements made, and how many refinements introduced.
Another cost, harder to calculate, was that the castle suffered from inherent weaknesses. No castle could hold out forever. Even Galliard, perhaps the most expensive castle of the twelfth century, fell to the French. No defense is impregnable. “The best that a besieged castle could hope for was to raise the price of victory to a point which the besiegers would be unwilling to pay.” That price could be exacted in various ways. Storming a castle was one of the most dangerous activities any medieval fighter could undertake because repelling such an attack was exactly what the stone walls were designed for. A besieging army often suffered more than the besieged from lack of shelter, the spread of disease, periodic food shortages, and the threat of a relief force appearing in the rear. Once feudal service requirements expired, the besieging troops would have to be paid, making long sieges unlikely.
Regarding pay and length of service, the castellan had two advantages. Garrisons tended to be small. Admittedly, their numbers, like those of armies, are hard to calculate because medieval chroniclers were notoriously bad with numbers. Nevertheless, it is rare to find evidence of large garrisons, and sometimes these were raided to flesh out field armies. A second advantage was that the soldiers in a surrounded keep could hardly leave, and no pleasant result could ensue from stopping fighting, unless the garrison in its entirety decided that its obligation to serve had expired, which occasionally did happen. Yet the maintenance of garrisons also presented many problems. Defending a castle was not a task for amateurs. Professionals had to be hired. Because many of these were essentially mercenaries, their loyalty was sometimes doubtful and mutiny was a serious risk. Gatehouses were constructed to provide a safe place for the castellan to live and also to allow him to control entry and exit. But the loyalty of the castellan was also a problem for rulers. Moreover, the paid or “stipendiary” troops had to be convinced that their ultimate lord was loyal to them. If they were besieged and their lord made no effort to relieve them, they might surrender as they then “had no natural lord.”
Even if the cost of garrisoning castles were overcome, inherent weaknesses remained. A castle can do little to prevent an attacker from choosing his moment. “No matter how strong, an isolated castle was vulnerable . . . A network of castles supported by a field army was a very different proposition, but only a few kings could afford such a combination.” Improved construction techniques could make the siege longer and more difficult, but not impossible. The early motte and bailey castle was highly vulnerable because the various parts of the defense could not support each other. That was solved by the more modern stone castles, but the impregnable fort was never built. Galliard was considered so, but it fell five years after its completion. Its fall illustrated a major conundrum for the builders. A well designed castle could deter an attack but not if its location was truly vital. Then the attacker would find a way. Any castle would eventually succumb to starvation.
That a castle could be taken was perhaps no more significant than that a field army could be beaten. More significant to evaluating the opportunity cost of choosing a castle over an army is that a castle cannot maneuver. In fighting Edward I, the Welsh frequently bypassed the king’s castles, attacking them only when they chose. In turn, Edward did not spend more than ten days besieging Welsh castles. The English did construct castles during the conquest of Ireland but these merely became targets for successful destruction raids by the Irish. Spreading castles over the countryside during a conquest sometimes created more resistance than it could contain. The most expensive fortifications simply could not fulfill the strategic functions assigned them. Historian Michael Prestwich judged Edward I’s castles to be “the most magnificent series of fortifications to be built in all of Medieval Europe,” even as he questioned the wisdom of Edward’s strategy regarding the resources required to maintain them. If castles were stupendously expensive, ultimately vulnerable, and prevented the build-up of field armies (the main opportunity cost), why build them? The advantages must have been thought more than enough to offset these drawbacks.
Historiography of medieval warfare has dispelled the notion that commanders and rulers were simpletons. Medieval rulers were keenly aware of their choices and attending costs. Knowing the cost of castles, they continued to build them. Even when short of money, they continued to build them, risking confrontation with vassals and taxpayers over the bill. To leaders of the age, the benefit of permanent walls outweighed the cost.
This was based on a very important consideration: the multiplicity of goals met by building castles. The castle was anything but a stone ring within which to hide from conflict. The castle dominated the landscape, figuratively and literally.
The castle was also a storehouse for munitions, an advanced headquarters, an observation post in troubled areas, home of a lord, and a place where he could be secure from attacks by his enemies. Royal castles could in times of emergency act as havens for the king’s field army, or supply the men to raise a new army if the field army was defeated. In the event of an invasion castles drew off large numbers of men from the invading force, which had to capture or at least contain the castles being left in the rear or the flanks to maintain supply and communication lines . . . [It was] not a place of refuge, but a centre of military power.
This multiplicity of roles helps explain why castles were the “main bones of contention . . . and focal points” of struggles. As a center of government and war in a warlike age, the castle was a symbol. The great royal castles of Wales were more than Edward I’s way of controlling the Welsh; they were a way of reminding the Welsh that he was there. Evidence suggests that Edward understood this perfectly well. External architecture was used as a means of impressing the people, such as at Caernarfon. Its expense was magnified by its unique construction: Edward intended its walls to look like Constantinople’s. Good locations were exploited without regard to expense, such as at Conwy, where the castle overlooks most of the town. In especially vulnerable locations, such as Beaumaris, on remote Anglesey, Edward’s engineers incorporated every possible feature—double walls, moat, access to the ocean, a profusion of towers—that could add to its strength. Construction of Edward’s castles frequently included huge round towers that were doubly advantageous because they were harder to assault and gave the castle an especially imposing visage. This method of building was expensive because for structural reasons the walls consisted of multiple layers of materials, as erosion and age now reveal. The appearance of the castles was more than a matter of aesthetics or ego. As is shown by the elaborate and complicated interiors, these buildings were also the residences of high officials—and sometimes the monarch himself. The castle was wasted expense if it were mere vanity, but sensible expenditure if viewed from the perspective of occupation politics.
For the expansion-minded magnate the castle was the best, perhaps only, means of laying permanent claim to disputed, threatened, or rebellious land. “The great value of the castle was that it held down the land; without taking it there could be no secure mastery of the land that it dominated.” Controlling territory required removing your rival’s castles, as well as having your own castles there. This strategy can be seen in action in Germany, France, and Britain. In Germany, the Duke of Swabia advanced down the Rhine, building one castle after the other, using each one to subjugate the country before continuing to the next spot. Fulk Nerra laid the basis of the Angevin dynasty in France by building substantial castles across his territories, spacing them so that mounted forces could easily move from one to the other. Edward I’s needs were slightly different from Fulk’s as he had to contend with entire hostile populations but his techniques were, nevertheless, similar. Each Welsh castle was built no more than a day’s march from its nearest neighbor, and they were also located where they could be easily supplied. Although the criticisms of the amounts spent seem just, the revolutionary spirit in Wales has led one author to conclude that a “policy of energetic castle-building was essential.” Perhaps it is no accident that Edward kept Wales but lost Scotland, where he did not engage in much building. His strategy there has been described as “less impressive, consisting as it did of sending northwards large armies with the goal of bringing the Scots to battle.”
The Welsh were not the only medieval Europeans to look askance at the forbidding fortresses in their midst. The temptation to use the castle’s advantage to abuse surrounding inhabitants was too great. According to the Anglo-Saxon Chronicle, early medieval castles were used to imprison and torture people who might have goods. The Normans were long thought to have gained supremacy in parts of the Mediterranean because of the supremacy of their mounted warriors, but it now appears that their most successful strategy was to take a castle and use it as a base from which to terrorize the surrounding district into submission. When Alfonso VI captured Toledo in 1085, he placed a garrison at Aledo, far to the south of his effective rule, where it functioned as “a thorn in Muslim flesh” until 1092.
Normans, Anglo-Saxons, Anglo-Normans, and Spaniards were able to maintain castles for years among hostile peoples. A field army could hardly have done so. It might be a better instrument for short-term conquest, but permanent occupation was another matter. An army could rescue a besieged castle, but a castle could shelter a small army. In an age where capacity to stay in the field was limited, this gave the castle a crucial advantage. The historical literature is filled with examples of medieval commanders avoiding battles, and some never fought any at all. With a castle as a handy retreat, there was no need to fight unless victory was certain, in which case the adversary would not fight. Clausewitz’s dictum that “no battle can take place unless by mutual consent” is unusually accurate for this age. Most invaders used a single line of advance and could easily be evaded. When Edward III (r. 1327–1377) invaded Scotland, the Scots either avoided battle or took up strong defensive positions the king was unwilling to assault. Even when multiple lines of advance made complete avoidance impossible, the weaker force could retreat into its castles and wait for the fighting season to end. Fulk Nerra avoided battle whenever possible. Even when it was not clear that the enemy was stronger, fighting from behind walls was tempting.
Fighting from behind walls deterred direct assault, and this might tempt us to think of this kind of war as “low cost.” It was not. The suffering among ordinary folk was magnified in two ways. Obviously a successful storming was followed by massive misdeeds. More often, it lengthened the war. According to J. F. Verbruggen, the lawyer and pamphleteer Pierre Dubois
lamented that the enemies of Philip the Fair were unwilling to fight battles any more because the royal army was too strong, and castles and fortified towns made it impossible to finish a war quickly and successfully. Against such fortresses “your army of splendid knights usually has to fight out a lengthy siege.”
Longer wars spelled higher tax bills, more frequent military service, more raids, property seizures, burning of hovels and possessions, and the privations of being besieged, even if the result was not a storming or a subsequent slaughter. In fact, a type of conflict developed that was relatively low in direct military cost but devastating for the populace. Feuding, as it was called, involved plunder, looting, and devastation; in other words, warfare aimed directly at civilians. Absent a carelessly opened gate, the castle was immune to these attacks. It could, however, serve as a place to store the loot.
Retreating behind walls might also imply that a cost of reliance on castles was the sacrificing of offensive strategy. Again, this was not quite true. Yes, castle spending inevitably meant less money for the field troops. But defending a castle was not quite the same as defensive warfare. A castle could have both offensive and defensive purposes, and very few castle builders thought solely in terms of defense. From the moment of their construction, castles “threatened neighbors, and . . . often served as bases for attack”; they “could menace enemies as well as protect friends.” Rarely was the defense of a castle a passive operation. “The overriding thought in castle strategy was not passive defence but action and destruction.” Moving across the drawbridge actually strengthened the “retreating” force. “Shutting oneself up in a castle was not an attempt to avoid conflict, but a manouevre to make the enemy fight at a disadvantage. . . . The defence had an enormous advantage. To an invader time would be vital.” Depending on distance, essentially three cases existed in which the castle garrison threatened an enemy. First, sallies might be made against the immediate besiegers. Second, raids from the castle could threaten areas within a day’s round trip or so. An example of this can be seen in the castle at Le Puiset, taken by Louis VI in 1111; no one dared to come within 8–10 miles of the walls. And third, the castle’s troops could be launched on a major operation against a neighboring principality. It is no coincidence that the most offensive-minded leaders of the age often withdrew behind stone walls. The castle was not purely defensive.
Raids and major operations could be contained by siege. A brief siege would be meaningless, as the castle’s troops could return to their offensive ways as soon as it ended. To contain an aggressive prince, it follows that his garrisons had to be besieged to the point that they would surrender rather than face a worse experience. Siege warfare became the operational center of medieval warfare, not just to overcome the defense but to prevent the fortified enemy from engaging in offense. The castle became the geographic center of warfare. In civil wars the focus on castles was even greater than normal because there was more emphasis on establishing and maintaining control of territory. Even the English campaigns in Scotland revolved around castles. The English capture of Stirling (1304) ended resistance more effectively than the famous Battle of Falkirk (1298), and the English disaster at Bannockburn (1314) was a direct result of the need to relieve the garrison at Stirling. Robert Bruce attempted to make successful sieges more decisive by destroying captured castles but even such extreme “decastelization,” practiced elsewhere as well, still depended on a successful siege to begin with.
If we accept that siege was central to medieval warfare, why its lack of study until the 1990s, especially since the methods and conventions of siege warfare did not change very rapidly? A straightforward answer is that a generation of military historians preferred to emphasize medieval open-field battle. Historians were nearly as affected as people at large by the image of the heroic knight on horseback. The armored cavalry of the High Middle Ages can hardly be disregarded, of course, but it played no role in siege warfare. A knight could fight for or against a castle, but hardly while horsed. He could climb ladders, fire weapons, pour boiling oil, or give orders to so do, but this fighting, while often more dangerous than the charge with lance and sword, would produce much less epic poetry. The horse in such operations was rarely a gallant, elegant steed, but rather the sturdy and ugly draft horse that pulled the battering rams and other heavy equipment. Sometimes that work was given to the even less glamorous ox.
Glamorous or not, medieval warfare required sieges. The only alternative was rapid raiding, so rapid that counterattacks by castle troops were irrelevant. This was unlikely as the most important points on the map were the most likely to be occupied by a castle. Sieges, while difficult and expensive, were inevitable.
The business of laying siege to a castle was a highly complex and costly affair and was not undertaken lightly. Frequently such sieges involved so many men and so much materiel that no other action could be undertaken by the army involved and once committed to a siege there arose the problems of guarding against an attack by a relieving force, maintaining adequate supplies of food and forage, and epidemic disease caused by the concentration of a large body of men in a small area for a long time.
This seems to have been the case throughout the Middle Ages. Castles might be primitive, but taking them was often beyond “the incoherent and ad hoc armies of Western Europe.” Even the simple motte and bailey castle was rarely captured. A documentary and archeological study of some one hundred and fifty such castles in Belgium has found scarce references or evidence of successful attacks. Only five are definitely known, although the baileys were sometimes burned, and one castle was captured by two knights and twenty peasants when the gate was left open.
The single stone keep or donjon also posed extraordinary problems. The donjon of Coucy (built 1223–1230) was so strong that its barons defied royal commands for two centuries! That these castles were smaller than the motte and bailey variety was no balm for putative conquerors. “Long passive sieges were usually futile, because the small garrisons required to defend these keeps used up their vast stores very slowly and used little water.” Even large castles, if properly designed, could be defended by small groups. Caernarfon survived two sieges (1403–1404) and inflicted three hundred casualties even though its garrison was down to twenty-eight men. In 1216 a substantial French army required fifteen days to capture a castle held by thirteen men; Harlech resisted the rebellion of 1294–1295 with a garrison of only twenty. The famous crusader castle Krak des Chevaliers, defended by a skeleton garrison, survived twelve assaults and was finally taken only by ruse in 1271. Château Galliard had more troops—one hundred and forty at the time of its surrender—and was only lost after a six-month siege, a five-week assault, an inadvertent failure to defend a window, which allowed the attackers in, and no serious relief attempt. For the attacker, the situation did not improve over time. “By 1300, the odds of a strong force successfully forcing a well-built and well-defended fortress were less than they had been three centuries earlier.” Gunpowder changed things, of course, but seven seventeenth-century cannon could not reduce the twelfth-century keep at Rochester.
The attacker faced severe costs. Money supplies sometimes succumbed before the defense did. Stephen (r. 1135–1154) spent £10,000 besieging Exeter Castle, a sum five times his estimated annual income. Indeed, a successful siege could be financially disastrous. Henry III (r. 1216–1272) captured Kenilworth in 1266 through negotiation, “but the siege had been cripplingly expensive, absorbing the income of ten English counties.” The personnel cost constituted the real culprit. On both sides specialists were needed. A long siege quickly raised expenses because feudal levies could only be required to serve for periods set by tradition and law, and once these were exhausted, they had to be paid. The cost of personnel—to ruler and soldier—was high. Estimates of the numbers needed for a successful storming range from 4:1 to 10:1, and this seems conservative in view of some castles’ successful defense against odds of 50:1. For the soldier, the attack was extremely hazardous and even after a breach of the wall it was not always easy to get troops to assault, sometimes requiring leading nobles and rulers to lead from the front. This could easily backfire. In 1088 a king was leading an attack on a gate when a woman described as “female in sex but not in spirit” threw a millstone on his head, bringing his reign to a premature end.
Consequently, the attackers often opted for the slow process. But time itself was a cost. The besieging army could do nothing else and decisive victory could prove elusive. Stephen attacked rebellious baronial castles one by one, capturing a number through starvation, but at his death much of his kingdom was still in rebel hands. Delays could result in failure and the collapse of morale. To such psychological problems the garrison troops were vulnerable as well, but their mighty blocks of masonry were not. In a word, the nub of the problem was that the resources invested by an attacker were much greater than those of the defender.
The disproportionate resources needed to take a castle in comparison to defending it, once it had been built of course, was one of the key reasons for the constant rebellions of the nobility that characterized medieval politics. The nobles often could defy their lords with impunity, at least in the short term, behind the wall of their castles.”
The result was an endurance contest that often ended without a decisive fight. No definitive numbers exist for successful versus unsuccessful sieges. That the failure rate was very high is undisputed; indeed, the majority of successes were the result of negotiation, not storming. William the Conqueror allegedly never failed to take a castle, but he did so mostly by agreement. Even after a three-year siege of a castle in Normandy, the famous duke offered terms because he “had been forced to recognize that taking the castle was too demanding in time and resources.” And this medieval ruler, more than most of his contemporaries, was able to assemble and maintain the obvious alternative to the castle, a trained and reliable field army.
Whether the castle could have survived much longer once gunpowder became prevalent is actually a moot point. Its very strength was a barrier to the growth of great national governments. Monarchs regarded private castles as an inherent threat. Legal and other measures were taken to eliminate them. In Britain, the Tudors were particularly effective in eliminating great noble castles as part of a well designed program to establish the state’s monopoly on violence. In France, Louis XIII probably destroyed more castles than he built. This trend was actually a paean to the military virtues of the castle.
Copyright notice: Excerpt from pages 45–66 of Castles, Battles, and Bombs: How Economics Explains Military History by Jurgen Brauer and Hubert van Tuyll, published by the University of Chicago Press. ©2008 by The University of Chicago. All rights reserved. This text may be used and shared in accordance with the fair-use provisions of U.S. copyright law, and it may be archived and redistributed in electronic form, provided that this entire notice, including copyright information, is carried and provided that the University of Chicago Press is notified and no fee is charged for access. Archiving, redistribution, or republication of this text on other terms, in any medium, requires the consent of the University of Chicago Press. (Footnotes and other references included in the book may have been removed from this online version of the text.)
Jurgen Brauer and Hubert van Tuyll
Castles, Battles, and Bombs: How Economics Explains Military History
©2008, 420 pages, 23 line drawings, 21 tables
Cloth $29.00 ISBN: 978-0-226-07163-3 (ISBN-10: 0-226-07163-4)
For information on purchasing the book—from bookstores or here online—please go to the webpage for Castles, Battles, and Bombs.