Caballus et Caballarius in Medieval Warfare

By Bernard S. Bachrach

From: The Study of Chivalry: Resources and Approaches (1988)

The little poem "For Want of a Nail" has for centuries conveyed to children a glimpse of the fundamental technical underpinnings of the chivalric world.1 These grubby details were, of course, a commonplace to the mounted fighting man of the Middle Ages, whose life or death often depended upon his equipment and the health and training of his horse. Knowledge in these matters was grasped by the greater nobles who served an apprenticeship which has come to be identified with the roles of page and squire; men of lesser status continued throughout their careers to care for their own mounts and pack horses. By contrast, the creators of chivalric literature and their modern explicators have largely ignored the less than romantic aspects of chivalric and military life. Maxims such as "an army travels on its belly" condition medieval as all other warfare. The purpose, in part, of this study is to discuss some of the basic biological, veterinary, and ecological factors concerning horses and their use which, through the collective experience, written and oral, of Western tradition and practice, were well understood by medieval military men.


From the vast corpus of surviving medieval illustrations and sculptures that are easily studied through the thousands of photographs that have been assembled by the Index of Christian Art, many useful observations can be made concerning medieval war horses. Among the most forthcoming sources for many aspects of military life, including the horses of the period before the First Crusade, is the Bayeux Tapestry which was designed and executed sometime before 1082 and depicts the Norman invasion and conquest of England in 1066.2

William the Conqueror's war horses, as depicted in the Bayeux Tapestry, were a breed of rather large and heavy equines.  The animal's head, as compared to the rest of his body, seems to have been disproportionately small, but it was rather thick and its nose was full. The crest was rounded. The neck, shoulders, and chest were very fully developed and probably thickly muscled rather than simply fat. The croup was markedly arched, while the haunches and thighs were heavy and give the impression of great strength. The mane was relatively short, and the tail was long and scraggly.3  These horses appear to have been considerably more bulky and stronger than the lean Arab breed with its dishface and small, wedge-shaped head that was found primarily in the Muslim East.4 European war horses also may be differentiated from the ponies of Central Asia and the South Russian Steppe used by the Huns and later by the Mongols. These animals had great hooked heads, manes hanging below their knees, drawn bellies, lean rumps, and bushy tails.5 The ponies used in Scandinavia and later in Iceland much more closely resembled the horses of the Steppe, especially in size, than they did the war horses used by William and his contemporaries.6

Although the designer of the Bayeux Tapestry generally cannot be relied upon to have provided lifelike proportions when executing fortifications and ships in relation to men and horses,7 his efforts in regard to the rider and his mount often strike the viewer as reasonably realistic. One such case is particularly good and depicts a mounted fighting man from William's army engaging a housecarl who is wielding the often-celebrated two-handed Danish axe. These elite foot soldiers who served in King Harold's personal military entourage were renowned for their great strength and size.8  The animal's height at the withers (once allowance is made for his bent front legs) is 9/11 of that of the housecarl as measured from the point of his helmet to the soles of his feet.

Thus if the horse measured 15 hands, the housecarl would have stood 73.3 in. to the helmet point, and if the horse measured 16 hands, the housecarl would have stood 77.3 in. to the helmet point. Since the top of the warrior's head reached only to within 3 or 4 in. of the point in helmets of the type depicted in the tapestry, the housecarl under discussion here would have been somewhere between about 5 ft. 10 in. and 6 ft. 3 in. tall. A further measurement is of some importance in this context; i.e., the shaft of the Danish axe is generally agreed by archeologists to have been approximately 42 in. in length.9 With this as a base figure, the housecarl would have stood about 78 in. to his helmet point, and the horse of the Norman soldier against whom he is fighting would appear to have been just slightly more than 16 hands.

It may be observed parenthetically that most of the foreground scenes in the tapestry represent the proportions of men, horses, and weapons in a manner that is similar.10  For example, the archaeological evidence gives us an average length for swords of about 36 in., while spears are estimated to be in the 6.5-7.5 ft. range. The latter calculations for shaft length (the shafts themselves usually do not survive) are made in relation to the size and weight of the iron spearheads (which survive in great numbers from throughout western Europe) as these ratios are seen to effect the balance and ballistic characteristics of the weapon as a whole.11  In general, what we know of the skeletal remains of early medieval warriors enables us to conclude that these ratios and ranges for height are essentially sound. There is no need to resort to examples of "giants" such as Harold Hadrada, who was probably about 6 ft. 8 in. tall, in order to justify these data.12

The above estimation of horse size made on the basis of the evidence provided in the Bayeux Tapestry is affirmed by many manuscript illustrations and sculptures from throughout the Middle Ages.13  However, of more importance than multiplying manuscript references in this context is the archeological record from pre-crusade Europe in which horses in the 15-16 hand range have been identified on the basis of skeletal evidence in sufficient numbers proportional to those smaller in size to indicate that the larger animals were not aberrations.14  From written accounts it is also clear that large horses were bred for military purposes. For example, the Franks in the ninth century were celebrated for breeding such very large horses that some are reported to have been so tall - perhaps with a modicum of exaggeration - that they posed difficulties to those who would mount them.15

In light of the conformation of the horses depicted in the Bayeux Tapestry, their weight should be estimated conservatively to have been in the 1,500 lb. range and surely no less than 1,300 lb. For purposes of comparison, this places William's war horse as slightly larger but not heavier than the cavalry mounts of the British army in the later nineteenth century. Light cavalry mounts averaged 14.3 hands and heavy cavalry horses 15.3 hands.16 The loads carried by these horses in the nineteenth century were not markedly different from those of medieval war horses.17

Our estimates concerning the weight of medieval war horses, and William's animals in particular, are given some additional support by the fact that horses can comfortably carry 20% of their own weight under normal conditions; however, when conditions become more difficult as a result of excessive exertion, extreme weather, or rough terrain, the comfort index decreases dramatically.18  At the Battle of Hastings, William's war horses were required to carry live loads which in some cases surely exceeded 250 lb., i.e., a physically well-developed warrior, who was likely to have been overweight by modern standards, armed with a sword, spear, and shield, and dressed in a coat of mail, probably some sort of mail "trousers," and a helmet.19  In addition, the horse carried a saddle, saddle blanket, and an assortment of other harness under particularly trying conditions. For example, at Hastings the horsemen made repeated uphill charges while under missile attack against a formidable enemy throughout a battle that lasted from ten to eleven hours; the battle was preceded by a 7 mile ride and ended for some only after they had carried out the hot pursuit of Harold's retreating army.20

Yet another means of calculating the size and strength of war horses is by contrast with other types of horses used in medieval Europe. Firmly institutionalized distinctions between horses with varying functions were widely established even during the very early Middle Ages. Indeed, from Germanic law codes such as the Lex Alamannorum, the Lex Baiwariorum, the Lex Gundobada, and the Lex Ribuaria, a general principle was established that a war horse had three times the value of a pack horse, and a riding horse was worth twice as much as a pack horse.21 Under Anglo-Norman usage in the post-Conquest era, pack horses were restricted to loads of 200 lb.22  The pack horse depicted in the Bayeux Tapestry may perhaps seem slightly small but nevertheless conveys the impression that such animals were far less formidable than war horses. A pack horse limited to carrying loads of 200 lb. would not exceed 1,000 lb. in body weight and probably was no more than 12 hands tall.23

The horses of the medieval West were, in general, stall-fed breeds which required, as such animals still require, a food ration that is approximately half grain (usually barley or spelt, but sometimes oats) and half hay.24 The portion of hay could be substituted for by grass but at a ratio of 3:1 because of the high water content of the grass as compared to the dry hay.25 However, there should be no illusion that the domesticated horses of western Europe could be sustained in good condition only on hay and/or grass without the proper grain ration. Indeed, Vegetius, writing in the mid-fifth century, emphasized for the readers of his tract on veterinary medicine the long-standing Western awareness of the disastrous effects that the lack of a satisfactory grain ration had upon the domesticated horse. He stressed this point by comparing the negative effects a lack of grain had upon the horses of the Romans as compared to the much smaller grass-fed ponies used by the Huns.26 The Vikings' small ponies, whose descendants can still be seen in Iceland, appear to have been more similar to the horses of the Steppe, especially in size, than to those used by the armored fighting men of medieval Europe.27

Not every type of grain was considered suitable for feeding a war horse, or even, perhaps, a riding horse. Victor Vitensis, a later contemporary of Vegetius, points out that some types of barley or perhaps spelt were of such poor quality that they were commonly regarded as fit only for pack horses.28  That grain was considered a regular part of the diet of the war horse in the West is well illustrated by an episode in which Muslim scouts in Spain who were seeking information concerning the whereabouts of a Carolingian mounted column picked up its trail by examining equine feces that they found floating in a nearby river. When the scouts discovered undigested particles of grain in the feces, they concluded that the enemy horsemen were not far away. The scouts' reasoning appears to have been based upon the fact that wild horses, which habitually are smaller and lighter than domesticated ones, do not eat grain and that, since the feces had not thoroughly been dissolved in the current of the river, it had been dropped in the water only a short time earlier.29  Stories such as these give a sense of life to the ubiquitous documentary references that underscore the problems encountered in trying to find sufficient grain for the feeding of one's horses.

William the Conqueror's war horses as described above may be considered to have been in the same general size and weight range as war horses used in medieval Europe both before and after the First Crusade. Since these animals would be required to carry live loads of at least 250 lb. under difficult conditions, it is extremely unlikely that many war horses were smaller than 15 hands and weighed less than 1,300 lb. Animals in the 1,300-1,500 lb. range require, on average, more than 25 lb. of feed per day. Of this total requirement, about half, as mentioned above, had to be either oats or barley (spelt generally is an adequate substitute for the latter depending upon quality) and half hay. This combination sustains the nutritional and bulk needs of a stall-fed horse used for mounted military service and weighing about 1,300 lb. Larger horses required, and still require, more feed in proportion to their weight, but, for the sake of convenience and so as to avoid the possibility of exaggeration, we will use these minimum figures which are, in effect, a lower limit.30

Although the grain component is not subject to more than nominal variation without causing serious damage to the horse, part or even all of the hay component may be substituted for, for short periods of time, by green grass which may be grazed in situ or cut and brought to the horse's feeding place. If the entire hay component were substituted for by grass, each horse of the type under consideration here would have had to have consumed approximately 40 lb. of green grass per day in addition to the 12 lb. of grain discussed above.

When the war horse was in service, either for the purpose of transport–ing men or while on campaign, it could not obtain any significant part of its forage by grazing because of the great length of time that is required for this process. A horse takes from five to ten minutes to eat a pound of grain from a nosebag and from fifteen to twenty minutes to eat a pound of prepared clean hay from a feed box. Grazing green grass in a meadow requires approximately twenty to thirty minutes per pound, depending upon the length of the grass and its distribution within the field. Thus, if a horse were to graze a full 40 lb. of grass, the better part of the animal's waking hours would be spent in feeding; this, of course, would also include some two hours to consume the grain ration of 12 lb. and does not include the two hours required for watering.31  In general, it must be concluded that grain and hay were at hand when the horse was in service. When this was not the case, large amounts of time had to be spent in foraging (see below), which was usually a haphazard exercise, or in grazing, which seriously limited the range of the horse and the military usefulness of the mounted troops. Living off the land was a risky business.32

In addition to the grain and hay ration discussed above, a horse of the type under discussion here must have approximately 8-12 gallons of water per day, depending upon the work being done and the nature of the weather. This water by and large must be fresh and cool. Warm, brackish, or otherwise tainted water, provided that it is not toxic, may be used in small amounts in an emergency for several days without causing harm to the animal, but not on a regular basis.33  Of course, in most military operations, commanders had the routes thoroughly reconnoitered and camps established so as to assure the availability of sufficient water. For example, William the Conqueror's horses gathered for the invasion of England required in the neighborhood of 25,000 gallons of fresh water per day. When he encamped at Dives-sur-Mer for a month in August 1066, the total water requirement was at least 750,000 gallons.34  Despite the care taken by most commanders, especially successful ones, at times the system failed. For example, during the first crusade the author of the Gesta Francorum observes: "We pursued them through a land that was deserted, waterless, and uninhabitable . . . . We survived wretchedly, but we lost most of our horses so that many of our mounted-troops had to continue on as foot soldiers . . . ."35  Such errors were not commonplace, but, as in the orientation of the contemporary media toward crime and death, these occasional disasters gain a prominent place in the chronicles.

The task of supplying food and water for horses while on military campaign was not a simple one because of the primitive nature of the land-transport technology which prevailed in pre0crusade Europe. Four-wheeled wagons generally had a carrying capacity of about 1,200 lb. Wagons, however, were rarely used before the thirteenth century because the pivoted front axle and whippletree system had not yet been developed. This created severe problems for the use of wagons even in negotiating gradual turns and partially accounts for why Roman roads were so straight.36 In addition, although harnessing techniques that improved the drawing capacity of horses were developed during the later ninth century, these were not widely diffused before the twelfth century for use in hauling because of the axle-whippletree problem mentioned above.37

Two-wheeled carts were generally preferred to wagons for both military and civilian overland transportation of heavy loads. The carts had a capacity of about 1,000 lb. and were far more maneuverable than were the wagons.38  Each cart with a full 1,000 lb. load had the capacity to haul a day's food ration for forty horses or barrels of water for about a dozen horses. Thus it should be clear that any large force of mounted troops could not rely for any great length of time upon carrying its water supply, and the hauling of grain and hay was not very much more practical.

Horse-drawn wagons and horse-drawn carts were impractical, in part, also because the animals which did the draught work had to be fed with grain which in medieval Europe was often a scarce commodity (see below). Thus oxen which subsisted without grain were generally favored for draught work.39  For example, in Anjou through the eleventh and twelfth centuries horses were not used for draught purposes.40  However, the use of oxen to haul military supplies meant that the baggage train of an army could move only about 10 miles. per day, in contrast to the 25 miles per day that horse-drawn carts could make.41

A more common means of supplying small units of mounted troops was with pack horses which, as mentioned above, could carry loads of about 200 lb. Each pack horse could carry the feed (grain and hay) required for itself and for a war horse for a period of about two days provided that water could be obtained en route. If hay also could be found easily along the route and the pack horse was required only to carry grain for itself and the war horse, then the range was extended to a maximum of ten days. Under these conditions, each additional pack horse, up to three - the number generally agreed by medieval authorities that a mounted fighting man could safely handle while on march - increased the horseman's range by two days.42  If the mounted fighting man was to remain prepared at all times during the march to engage in combat on horseback, then, without a re-supply of grain, he had an effective range of about 280 miles, i.e., 20 miles per day for fourteen days, with two days required during the march to rest the horses. If the soldier was not expected to fight on horseback, his range was increased to about 420 miles.43 The needs of the horses while on campaign make it very clear that in medieval Europe and in the crusader states exceptionally careful planning had to be undertaken in order to assure that mounted troops would be able to engage the enemy in any but the most localized defensive circumstances.

As we have seen, horses eat and drink large amounts; however, they retain relatively little of what they consume. A horse each day produces on the average 4.5 lb. of feces and .56 gallons of urine per 100 lb. of body weight. Thus a horse weighing about 1,500 lb. would produce about 65-70 lb. of feces per day in a mixture that is about 75-80% liquid and an additional 8‑8.5 gallons of urine.44  These excretions, of course, are extremely dangerous to both the animals and the men whom they serve, especially when a military force is encamped for a lengthy period of time or is besieging an enemy stronghold. Urine, while not toxic in healthy animals or humans, does in the process of evaporation produce ammonia, which is highly detrimental to horses' hooves. Feces, by contrast, are highly toxic and if not adequately dealt with can lead to health problems of epidemic proportions. At Dives-sur-Mer, where William of Normandy encamped with some 2,000 to 3,000 war horses for at least a month, a mountain of from 3,600,000 to 5,400,000 lb. of horse feces and a river of from 480,000 to 720,000 gallons of horse urine had to be, and were, disposed of safely.45

Horses have to be kept very clean and require fresh straw every day for their stalls - approximately 100 lb. of straw is needed each month. In summer, the animals must be kept cool and out of the hot sun; in winter, warm and out of draughts. They cannot be permitted to become chilled from cold rain and, in general, require a vast amount of skilled care on a daily basis.46 The chronic shortage of war horses, their high cost - for example, even in Merovingian Gaul a war horse was valued at four times a milk cow or ox, and this same ratio held in the ninth century 47 - and the effort required to train these animals (see below) generally meant that they were well treated.

As noted above, a horse that was to be kept ready to engage in combat could not be ridden much more than 20 miles on that same day. After the 20 miles limit was reached, the animal's physical ability to carry its rider into a battle of any significant duration was severely reduced. If the horse was used simply for transport, a daily distance of about 30 miles for no more than five or six consecutive days was practical and safe. After such a journey, the horse required at least a day's rest, or risk was being taken that the animal would break down.48 Simply to keep a horse in the proper shape for traveling such distances and for combat, a minimum of two hours hard training under the saddle each day was required, as well as a weekly ride of about 20 miles.49 Under extreme conditions a horse could be pushed 45 or perhaps even 50 miles, but after such a ride it generally would become useless.50

The great difficulties encountered in providing horses in medieval Europe for military or any other purposes are highlighted by the costs entailed in the support of these very expensive animals. In pre-crusade Europe, at least in much of it, seed-to-grain ratios averaged about 1:2. At this rate of production it required the true grain surplus produced by about nine agricultural workers to provide the barley, spelt, or oats needed by a war horse of about 1,300 lb. during the course of a single year. This came to about 4,370 lb. In the thirteenth century, when we can be relatively certain that on average a 1:3 seed-to-grain ratio had been achieved (i.e., a 100% increase in productive capacity), it required the true surplus produced by only two agricultural workers to provide the grain to feed a war horse of 1,300 lb. The 2.2 tons of hay and the 1,200 lb. of straw required for such a horse could be produced by the labor of an additional worker.51


As we have seen, during the Middle Ages horses were very valuable but also extremely fragile and difficult to sustain for military operations. Yet throughout western Europe, kings, dukes, counts, and others in policy–making positions insisted that in their polities there be substantial numbers of horses available for military purposes and trained men to use them. A variety of institutions and social structures either were developed anew or preserved from the Roman past in order to meet these requirements of grand strategy.52 However, despite this universal recognition of the importance of having war horses available in considerable numbers, mounted troops played only a small part in the overall military picture throughout the Middle Ages. Laying siege to fortifications - in which troops on horseback played only an auxiliary role, largely the interdiction of communication and supplies - was not only the most common type of major military operation but, from a strategic point of view, the most important.53

In contrast, large battles in the field were few during the Middle Ages and rarely decisive.54 Indeed, good generals, on balance, endeavored to avoid decisive battle with their equals because there could be only one winner and the risk of defeat was rarely worth the potential rewards of victory. Even in the few large and important battles that were fought, experienced commanders, especially when they took a defensive posture, would have the greater part of their horsemen dismount and fight on foot. A few examples of such battles - Conquereuil (992), Hastings (1066), Dorylaeum (1098), Tinchebrai (1106), Bremule (1119), Bourg Theroulde (1124), Northallerton (1138), Lincoln (1141), Crecy (1346), Poitiers (1356), and Agincourt (1415) - may suffice to make the point.55

It is also clear that when a commander underestimated the strength or resolve of a disciplined force of men fighting on foot and hurled his horsemen, inadequately supported by foot soldiers or fire power, at such a well-positioned enemy, the result was usually disaster or near-disaster for the mounted attackers. Whether we look to the early Middle Ages at Unstruct (531), Suntal (782), or to later encounters such as Lechfeld (955), Conquereuil (992), Saint Michel en l'Herm (1014), the first charge at Pontlevoy (1016), the first few charges at Hastings (1066), Legnano (1176), Courtrai (1302), Bannockburn (1314), Morgarten (1315), Crecy (1346), and Agincourt (1415), the failure of the mounted troops is evident.56  Indeed, even at the celebrated battle of Bouvines (1214), where most scholars credit the French mounted troops with winning the victory, it is clear that the Saxon infantry was superior to King Philip's own horsemen in the center of the line. The mounted charge by the French left flank against the Brabantine foot, who withdrew after a less than noteworthy resistance, is thought by many scholars to have been the decisive phase of the battle. This retreat, however, was probably due to the duke of Brabant's duplicity rather than to the vast tactical superiority of the French horse.57

When men on horseback attacked well-trained men fighting on foot in comparable numbers, or even when the mounted troops outnumbered the men on foot slightly, the latter had the advantage. Indeed, the Franks became so aware during the later ninth century that their mounted troops were no match for the Viking foot-soldiers who infested the regnum Francorum that the Annales Vedastini announced a change in strategy: Franci parant se ad resistendum, non in bello, sed munitiones construunt . . .58 (The Franks made themselves ready to resist, not in battle, rather, they built fortifications. . .).

When the armies of the regnum Francorurn did meet the Vikings in battle, moreover, able commanders such as King Arnulf showed great flexibility in tactics. In 891, Arnulf pursued a band of Vikings who took refuge in a hastily fortified camp on the river Dyle. The camp was protected by swampy ground, and the terrain was thoroughly unsuited for the deployment of men on horseback.59 The author of the Annals of Fulda describes the conditions:

The king delayed because he did not want to risk his strong army. On the other side of the river that surrounded the enemy camp was a swamp that did not give his horsemen a chance to attack. Thus he wandered here and there with his eyes, his thoughts and in seeking counsel. This counsel was necessary because the Franks were not accustomed to advance slowly into battle step by step (pedetemptim).60

Arnulf feared that the morale of his men was not up to slogging through the muck and mire slowly, step by step, under a hail of enemy missiles. Thus he told his staff that he would dismount, carry the banner, and lead his men on foot through the swamp. Arnulf's men did dismount, but his advisors impressed upon him the need for a mounted reserve so that "no warrior might fear an ambush by the enemy from behind"; thus, the king did not lead the attack, which was, nevertheless, successful.61

Although, as we have seen, the role of the medieval horseman more often than not was to dismount and fight on foot at sieges and even in set battles, there were many opportunities within the compass of sound strategy and tactics for these highly trained warriors to remain mounted and play a vital role in the course of important military operations. Thus, for example, the discussion of the battle on the river Dyle makes it clear that the use of the mounted reserve to defend against encirclement not only was a tactically sound use of horsemen but also was crucial to the morale of those who had dismounted in order to advance against the enemy on foot.

A more systematic discussion of both the training and the use of horsemen will make clear the role of the mounted element in military operations during this period. Throughout the Middle Ages, Vegetius's tract De re militari (c. 450) was of great practical importance to military commanders and policy-makers; indeed, hundreds of manuscript copies still survive not only in Latin but in a host of vernacular translations. In fact, no secular prose work of late antiquity enjoyed a popularity in the Middle Ages and even beyond that even approached that of Vegetius. Vegetian ideas and maxims permeated military planning, strategy, and tactics in medieval western Europe.62 Vegetius was known widely as the auctor or the auctoritas in military matters.63

One of the earliest extant efforts to use a selection of chapters from Vegetius's handbook for practical purposes was made by Rabanus Maurus, who during the mid-ninth century made an epitome of the De re militari and in his preface to King Lothair II indicated that he included only those things that were of importance "in modern times."64 Rabanus chose to include in this epitome the greater part of Vegetius's account dealing with the training of cavalry recruits (tyrones) to mount and dismount. Rabanus also chose to comment on this chapter when he wrote:

Wooden horses are placed during the winter under a roof and in summer in a field. The recruits at first try to mount unarmed, then they mount carrying shields and swords, and finally with very large pole weapons. And this practice was so thorough that they were forced to learn how to jump on and off their horses not only from the right but from the left and from the rear and in addition they learned to jump on and off their horses even with an unsheathed sword.65

Rabanus then adds to Vegetius's text: "Indeed, the exercise of jumping [on and off one's horse] has flourished greatly among the Frankish people."66 He also quotes "a commonly known proverb familiar to us which says `a horseman can be made in youth [in pube] but scarcely or never at all at an older age.67

This type of training, as we have seen, stood the troops of King Arnulf well in battle against the Vikings on the Dyle. Indeed, Regino of Prum, in commenting on the Frankish tactics at the Dyle, writes: "The angry king ordered his army to dismount and to do battle with the enemy on foot. This order having been given, the men jumped from their horses rapidly [citius] . . . and charged. . . ."68

The Franks also trained to fight while mounted, and early on in their history they learned the value of "war games," i.e., simulated combats for the purpose of training their troops. Nithard, a grandson of Charlemagne, witnessed a practice near Verdun in 842 which he described in the follow–ing manner:

For purposes of training, games were often arranged in the following manner. Fighting-men would be deployed in a place where they could be observed. The entire group of Saxons, Gascons, Austrasians, and Bretons were divided into two units of equal size. They charged forward from both sides and came toward each other at full speed. Then [before contact was made] one side turned its back and under the protection of their shields pretended to be trying to escape. Then those who had been engaged in a feigned retreat counter-attacked and the pursuers simulated flight. Then both kings [Louis the German and Charles the Bald] and all of the young men, raising a great yell, charged forward on their horses brandishing their spear shafts. Now one group feigned retreat and then the other. It was a spectacle worthy of being seen as much because of its nobility as because of its discipline.69

From this description it appears that the officers who orchestrated such training exercises were acutely aware of how potentially dangerous it was for men to practice these complicated maneuvers at high speed, and, thus, as a safety measure, the iron heads were removed from the horsemen's spears. The emphasis in this training, moreover, was not upon the type of mounted shock combat, so popular a picture in the chivalric literature, where horsemen with weapons couched beneath their right arms splinter their lances against each other's shields.70   Here the horsemen brandish their weapons overhand and make no contact with their adversaries.

The maneuvers described by Nithard had two basic purposes. First, the troops were training to remain in orderly formations under very strict discipline and to move as a unit. From these formations they were required to carry out repeated complicated redeployments that entailed sharp directional changes at high speed. The second major object of this training was to perfect the execution of the feigned retreat which, as we will see below, was the single most important tactic in the repertoire of the medieval mounted fighting man.

Both the feigned retreat and the emphasis upon the cohesive deployment of heavily armed mounted troops are well illustrated in the military operations carried out by King Henry I at Riade in 933. Henry set out to locate and engage a force of lightly armed Magyar horse-archers which had been raiding in the area of Mercerberg. Henry knew that his heavily armed horsemen could not force an engagement at close quarters because the Magyars, who were much more lightly armed, simply could outdistance his forces and, by remaining within bow range, pick off the slower-moving Saxons or, more probably, their valuable but unprotected horses. Thus Henry developed a plan which entailed the use of a force of lightly armed Thuringian horsemen among whom he spotted a very few heavily armed men. This force was deployed to attract the attention of the Magyars. When the numerically superior Magyars attacked, the lightly armed Germans were to wheel their horses and execute a feigned retreat. This was intended to lure the Magyars into pursuit and within range of Henry's heavily armed horsemen who, concealed by the terrain, would then attack the enemy before they could flee.71  In preparation for this encounter, Henry instructed his heavily armed horsemen in the following terms:

When you are charging forward to this initial contact with the enemy, none of you is to try to outdistance your fellows simply because you have a faster horse. Cover yourself on one side with your shield and catch the enemy's first volley of arrows on your shield. Then charge them at full speed as fast as you can so that before they have a chance to fire a second volley they may feel the wounds inflicted by your weapons.72

Maintaining a cohesive mounted formation (ordinata aequaliter acies) was well recognized in the West as crucial for the effective use of mounted troops.73 Einhard's report on the defeat of a troop of Frankish horsemen by a unit of Saxons fighting on foot in the Suntal mountains illustrates, through the description of a failed attack, what could happen when a mounted force lost its cohesion and each man attacked as an individual.74 Such mistakes seem to have been easily recognized even by court intellectuals such as Einhard, who writes:

Thus they [the leaders of the scara] decided to engage the Saxons without him [Theodoric] and took up their arms not as though they were intending to attack a prepared battle line but as if they were chasing down fugitives from behind and gathering up booty. The Saxons stood in their battle line in front of their encampment and each and every one of them [the Franks] rode at them as fast as possible. The charge was as poorly executed as the battle. Indeed, once the fighting began the attackers were surrounded by the Saxons and almost all of the Franks were killed.75

Einhard's use of the phrase "each and every one of them" (unusquisque eorum) illustrates the individual rather than the coordinated nature of the charge. It is to this failure to maintain a cohesive formation -the need of which was so fundamentally stressed by Henry at the battle of Riade - that Einhard attributes the defeat; i.e., the battle went as badly as the charge. However, Einhard does imply that it was commonly accepted practice for horsemen when pursuing fleeing enemies or when plundering to break up the battle line and operate individually.

Frontal assaults by mounted troops upon well-trained men fighting on foot, even when such attacks were well coordinated, were not likely to be successful; however, horsemen who were deployed so that they could attack on the enemy's flank could often expect better results. The battle of Adrianople in 378, where the charge of a force of mounted Goths and Alans played the key role in the destruction of a large army of Roman foot soldiers under the command of the emperor Valens, is often cited as marking the beginning of the end of the dominance of infantry in the West and the emergence of cavalry as the primary arm in warfare. It is often omitted, however, that the Alan and Gothic horse attacked by surprise and hit the Romans on the flank.76

A classic medieval example of this tactical situation occurred at the battle of Pontlevoy in the Touraine on 6 July 1016. Fulk Nerra, Count of the Angevins, in preparation for an attack on an army of foot and horse led by Count Odo II of Blois which greatly outnumbered his own, divided his own command into two parts. Fulk retained the larger part, composed of both foot and horsemen, under his own command at Pontlevoy athwart the line of advance of Odo's force which was moving to ravage Angevin possessions in the region. The remainder of the Angevin army, composed solely of horsemen, Fulk placed under the nominal command of Count  Herbert of Maine and stationed as a reserve at Bourre less than five kilometers to the south of Pontlevoy. When Odo's army arrived at Pontlevoy in the late afternoon of the sixth, tired from the day's march, it was astonished to find Fulk's army drawn up in front of it and ready to attack. With the element of surprise on his side, Fulk charged Odo's front. But after some very heavy fighting the Angevins were repelled by the hastily drawn up Blesois. Fulk was severely shaken up during the attack and unhorsed. His standard bearer, Sigebrand of Chemille, was killed. The Angevins withdrew, perhaps because the standard had fallen with Sigebrand; this was the usual signal for retreat. In any event, while the Blesois were resting after the exertion of this initial encounter, Count Herbert attacked out of the West. The late afternoon sun at his back partially shielded his movements and blinded the enemy. The Blesois' left side, taken in flank, crumbled under the impact of Herbert's charge, and just at that moment Fulk launched his second attack along Odo's front. This destroyed the Blesois resistance; the milites fled, while the slower-moving pedites were slaughtered in large numbers as the Angevin horsemen, in hot pursuit, rode them down.77

Flank attacks by one mounted unit against another were also an essential part of the tactical repertoire used by medieval horsemen. For example, at Andernach in 876 Louis the Young, the East Frankish king, defeated Charles the Bald when he used the technique of a "refused center" in order to expose the flanks of the West Frankish army to an attack by his horsemen. Louis had learned through his intelligence network of scouts and informers that Charles the Bald was going to attempt a surprise attack at night upon his camp near Andernach. Louis drew up his force in front of his camp with units of East Frankish horsemen slightly refused, or back, on either flank and a corps of lightly armed Saxons slightly advanced at the center. This formation in its initial stage was deployed to await Charles's army and resembled a slightly arched bow with the center thrust forward to attract the enemy's charge. Charles's army, somewhat surprised that the East Franks had learned of its advance, immediately charged the Saxons at the center of Louis's line. After putting up a nominal resistance, the Saxons began a feigned retreat which drew Charles's horsemen in pursuit. Now that the East Frankish center was "refused," the West Frankish horsemen were exposed to an attack on both flanks, and Louis's horsemen charged with devastating effect. Charles's army was crushed.78

The sophisticated operations that were planned in this campaign by both Louis and Charles are surely worthy of note, especially in contrast to the depiction of battles in chivalric romances, which appear to evidence a level of trickery which appears almost childish by comparison with the stratagems of real warfare. Charles's plan to launch a surprise attack, and at night, combined two of the most effective stratagems advocated by Vegetius.79  The West Frankish king early in his career had been given a copy of the De re militari by Freculphus of Lisieux, who made clear to his ruler in the dedication that books can offer practical knowledge by which he could gain an advantage at war.80  While these efforts by Charles in a losing cause are significant, the highlight of the battle is surely Louis's use of a "refused center," accomplished through a variation of the feigned retreat in order to prepare a flank attack.

The feigned retreat in various combinations, both against men fighting on foot and on horseback, was the most basic and most important tactic used by mounted troops during the Middle Ages. It was widely used also in the ancient world and usually was intended to lure troops fighting on foot from their established positions so that they would be vulnerable to an attack by mounted troops who lacked sufficient fire power to discomfit the enemy so as to force them from their emplacements. The classic example of the feigned retreat used successfully is at Hastings in 1066. Here William's archers, foot soldiers, and horsemen, after firing innumerable volleys of arrows and repeated charges, were unable to break Harold's phalanx. Thus on two separate occasions during the later stages of this ten to eleven hour battle, the Conqueror's horsemen acted as though they had been routed, broke ranks, and galloped away from the Saxon line in apparent disarray. Some elements of the great fyrd, i.e., general levies who constituted the least well-disciplined elements of Harold's force, deserted the comparative safety of their hilltop position and counterattacked in a pell-mell charge in pursuit of the retreating horsemen. But once the foot soldiers were running full tilt and were thoroughly scattered over the hillside, the horsemen turned on signal and charged the isolated individuals and small groups who were now overmatched in single combat.81 The Norman and Breton horse did not have to attack in a cohesive line, since the effectiveness of the foot soldiers, now isolated in the open field against better-armed and better-trained mounted soldiers in single combat, was limited.82  Of course, Harold's housecarls, with their massive two-handed axes, were another matter, and they gave William's men much the same treatment that the Swiss halberdmen gave the Burgundians centuries later.83

Yet another variation of the feigned retreat was used by Abul Kasim's troops against Otto II's army at Capo Colonne in 982 where the Muslims coordinated it with a hidden reserve and a flank attack. The Muslim light horse made preparations as if they would engage a force of German heavy horse in hand-to-hand combat; however, just before the armies made contact, Abul Kasim's men wheeled their mounts and began a headlong retreat. Otto's troops followed in hot pursuit. Several miles down the road, however, Abul Kasim had stationed a reserve of mounted troops hidden by the terrain, and when the Germans passed by, the Muslims struck their flank. The main body of Muslim horse then abruptly wheeled, formed up, and counterattacked. The Germans, who had broken their formation in order to carry out the pursuit - Einhard had earlier on remarked how it was normal practice for mounted troops to pursue a fleeing enemy in an individual manner - not only were surprised by the enemy reserve but were attacked on the flank and at the front at virtually the same time. Thus it should not be much of a surprise to learn that Otto's forces were all but annihilated.84  A comparison of Abul Kasim's tactics with those of Otto II's grandfather, Henry I, at Riade indicates that the coordinated use of both the hidden reserve and the feigned retreat were not only implemented against the Germans but by them as well.

It is now widely recognized by specialists in medieval military history that horsemen were deployed in a well-coordinated and highly disciplined manner when they were permitted by their commanders to remain mounted while engaging the enemy.85  The basic unit for these troops in mounted combat was called in French the conroi and usually numbered ten men. These conrois, in the West, often functioned in tactical units of three hundred men.86  The entire system seems to have been of late Roman or Byzantine origin. An early eleventh-century example is provided by the Normans who came to southern Italy only in small groups with no military organization of their own; they had only their horses and their personal equipment.87  Very shortly after the Norman arrival it is clear that they began the development of tactical units of three hundred men, which was the same size as the Byzantine tegma. These Norman units were composed of groups of ten men each and, on the whole, were easily integrated on numerous occasions into Byzantine units or used Byzantine equipment such as horse transports.88  It is also of some interest in this context that the Anglo-Norman constabularia, which was composed either of ten or twenty men, was probably a conroi.89

Thus far the emphasis, here, has been upon coordination, training, and discipline in medieval mounted warfare. This has been done, in part, to demonstrate that the effective use of horsemen depended upon the sound military principles noted above and also, in part, to highlight the gulf between the reality and chivalric literature, which focuses upon the individual warrior who is often depicted as more courageous than disciplined. However, the impression should not be left that mounted troops did not, on occasion, engage in single combats that were worthy of notice. For example, Sidonius Apollinaris, writing during the later fifth century, describes a single combat that could well have been the highlight of a medieval literary creation. He writes:

After the first, second, and third charges had been made, the upraised spear comes and pierces the man of blood. His chest was punctured and his coat of armor was penetrated twice. It was pierced even through his back and . . . the blood came pulsing through the two holes ....90

Some five centuries later Liudprand of Cremona describes a single combat in no less dramatic terms:

The Bavarian wheeling his horse, first urged him forward in a vigorous manner, then he pulled up on the reins and drew him back. In the meantime, Hubald drove his mount straight at his opponent. When the two horsemen were in striking distance of each other, the former turned his horse around in his accustomed manner and then veered him to the left and to the right in a convoluted series of circling movements that were intended to frustrate the straight on attack of Hubald. However, as the Bavarian turned his horse in retreat with the intention of wheeling suddenly and striking a frontal blow, Hubald spurred his horse straight on and drove his spear through his opponent before the latter could wheel around. Hubald's spear penetrated the Bavarian's back between the shoulder blades and entered the heart.91

Such single combats as these, however heroic they might seem, were rare, since the ebb and flow of battle permitted few opportunities for the same two men to make repeated charges against one another or to maneuver in the convoluted manner used by the Bavarian horseman described above. Nevertheless, the head-on charge of two tightly packed mounted formations at high speed could lead to the forceful clash of two identifiable individuals. The dramatic results of such an impact are powerfully expressed in D. J. A. Ross's translation from the Song of Roland:

He breaks his shield and bursts open his haubert, cuts through his breast, smashing his bones, and tears away the whole spine from his back; with his lance he casts out his soul; he thrusts it well home and causes his body to swing back and hurls him dead from his horse a full lance-length away.92

The three single combats described above indicate a variety of combat techniques. Only in the last, however, because the body of the defeated warrior has been thrown a full lance-length (plein sa hanste) over his horse's rump, is there a strong possibility - in light of the presumed force of the blow - that the victorious warrior was using the couched lance technique which is the staple of chivalric literature. Experimentation with this technique can be seen to have developed during the later eleventh century, where, for example, in the Bayeux Tapestry we can see the use of several techniques: some men are thrusting with their lances underhand; some are thrusting overhand or preparing to throw; and one is using the couched position.

The experimental nature of the couched lance position is provided with a kind of quantitative support from the Bayeux Tapestry itself, where, of thirty-five mounted figures depicted in combat with long-shafted weapons, thirty are seen to be thrusting or preparing to throw them, and only five are seen in the couched position. Indeed, Ordericus Vitalis, writing c. 1140, makes it unambiguously clear to his readers that throwing the spear from horseback is a vital military skill that must be practiced in order to be effective.93  The couched lance position was not made practical until well into the twelfth century when a combination of developments, coordinating the use of stirrups with advancements in saddle design, enabled the warrior to remain seated on his horse despite having been struck by an adversary who was moving at perhaps 25 miles per hour.94

Although stirrups became known in western Europe by no later than c. 700, they had no military impact for some two centuries or more. The diagram in figure 5.1 (Figure 5 is not included in this reproduction) shows how horsemen positioned their legs both before and after stirrups had been introduced into the West.95  However, early in the tenth century we can note that a step had been taken to make use of the stirrups in combat. Figure 5.2 illustrates this position: the rider's foot and leg are turned away from the flank of the horse; the stirrup leather cuts across the shin; the toes are pointed downward; and the leg is thrust forward against the stirrup and away from the horse. In the eleventh century, by contrast, the rider stands in the stirrups, which he uses as a platform from which to launch his attack. The technique pictured in figure 5.3 prevailed among the Normans who are depicted in the Bayeux Tapestry. Finally, in the twelfth century the horseman adopts yet another position: as we see in figure 5.4, the rider remains seated in the saddle; his legs are thrust forward against the stirrups; and his back rests against the wrap-around cantle.96

Thus in the twelfth century the rider is locked onto the horse's back in a sort of cockpit. The shock of the enemy attack is absorbed by the horseman who now, rather than being driven plein sa hanste over the rump of the horse, is driven into the cantle of the saddle which in turn is in danger of being driven off the horse's back. In order to compensate for this eventuality a double-girth system was developed by which the saddle and rider together were held on the horse's back. The front girth kept the man and saddle from going over the rump, while the rear girth prevented the unit from riding up the horse's neck in equal and opposite reaction to the initial impact. The wrap-around pommel both held the rider in the saddle during the reaction phase of the sequence and protected his groin and abdomen from enemy attack.97

The shock of being driven into the cantle of the saddle at perhaps as much as fifty miles per hour, if both horses were moving at a full gallop, was bound to have a serious effect upon the kidneys and perhaps even the spine of any horseman who took a solid blow. It is likely that in response to this problem the rigid back-plate, which had but marginal defensive value in comparison to its weight and cost, was developed. The rigid back-plate remained firm while several inches of quilted padding between the rider's back and his armor cushioned the impact. Similarly, the substantial development in helmet designs, which saw them fixed to the shoulders and thoroughly padded, cut down on the whiplash to the neck and limited the likelihood of concussion.98  By the end of the Middle Ages the participants in a joust were outfitted more like drivers in a demolition derby, and for largely the same reasons, than as horsemen who could function effectively in battle. By contrast, battle armor was considerably less elaborate and far less protective but enabled the mounted soldier to fight effectively either on horseback or on foot.99

To conclude: the popular chivalric view of the horseman as an individualistic and undisciplined dynamo who dominated medieval warfare must be greatly modified. During sieges, the most prominent type of major military engagement during the Middle Ages, troops on horseback played only minor roles dealing with logistics and communications. In large and important battles, which throughout this period were very rare and usually were avoided by able commanders, horsemen were as likely to dismount and fight on foot as they were to remain mounted and engage on horseback. When mounted units were permitted to engage the enemy on horseback, successful commanders generally used tactics such as the feigned retreat, flank attacks, and hidden reserves in order to gain the greatest advantage. Rarely did an able commander order his mounted troops unsupported by fire-power and foot soldiers to charge pell-mell at an emplaced enemy. Indeed, such frontal attacks were even more rarely carried out a second time by the same men.

The tactics that were used effectively by mounted troops, both when they fought on horseback and when they engaged on foot, required great personal discipline as well as the ability to operate effectively in groups. The requirements of medieval battle were the antithesis of the individualistic behavior heralded in the chivalric literature. The more mundane logistic aspects of medieval warfare, without which neither battles in the field nor sieges could occur, required of the military commander (a position to which noble fighting men often aspired) administrative skills and bureaucratic talents that sharply contrast with the impulsive behavior and rash courage so frequently lauded in the poetry of warfare.

How did the medieval fighting man who knew what real warfare required react to the image propagated in the literature of chivalry? Was he somewhat bemused, like the nineteenth-century cowboy who read Ned Buntline comics or the CIA agent who sees James Bond films? Was Roland's kill ratio as absurd to William the Conqueror as John Wayne's inexhaustible six-shooter was to George Patton? I am wont to believe that, then as now, on occasion imagination could suspend reality, but in general I would suspect that chivalric literature was to medieval warfare what the poetry of courtly love was to medieval sex. Honi soit qui mal y pense.

* Please note that the original text of this article, found in The Study of Chivalry contains several illustrations, which are not republished here.  Readers are advised to consult the book if possible.

End Notes

1. Medieval horseshoes usually required six nails and weighed 10-15 oz. Modern horseshoes, which are recognized to be of much higher quality and are perhaps a bit heavier than those used during the Middle Ages, last 200 to 250 mi. With these figures in mind, it is rather simple to calculate some rather mundane details. For example, when William the Conqueror was encamped at Dives-sur–Mer for a month during August and early September 1066, he had the command responsibility to assure that from two to three thousand horses were properly shod. (See below, n. 34.) Since each horse would be required to train a bare minimum of 75 mi. per week (see below, n. 32), at least one set of shoes, by conservative estimate, was required for each animal while at Dives. Thus a total of 8,000-12,000 shoes were needed for this month alone, and somewhere in the neighborhood of 50,000-75,000 horseshoe nails at about 1/2 oz. each, a total on average of at least 8 tons of iron that had been forged by skilled workers into shoes and nails. Assuming that an equal number of horses needed shoeing each day, William's herd required the services of no fewer than ten blacksmiths, each working a ten hour day every day. If we consider also the vast amount of wood or charcoal required for the forges, the hours of woodcutter-labor, the carts needed to transport the wood, and so on through the work-process, we can begin to form some idea of the great complications intrinsic to a major military operation. The systematic integration of all these elements in order to avoid a fatal "snafu" was no small task. Concerning horseshoes, see Germain Carnat, Le fer a cheval a travers l'histoire et l'archeologie (Paris, 1951), pp. 64-65, who weighs used shoes. G. Tylden, Horses and Saddlery (London, 1965), tends to be more useful.

2. The basic work is N. P. Brooks and H. E. Walker, "The Authority and Interpretation of the Bayeux Tapestry," in Proceedings of the Battle Conference on Anglo-Norman Studies (hereafter PBC), 1, 1978 (Ipswich, 1979), pp. 1-24. See also David M. Wilson, The Bayeux Tapestry: The Complete Tapestry in Color with Introduction, Description and Commentary (New York, 1985).

3. Among modern English horses, the outstanding characteristics of William's horses as depicted in the Bayeux Tapestry are found among team horses of the army service corps (thick chest and hindquarters) and Royal Horse Artillery horses (long back and thick, rounded chest). See G. Tylden, Horses and Saddlery, pl. 77, fig. 8, and pl. 76, fig. 3, respectively. The recent articles by Anne-Marie Bautier, "Contribution a l'histoire du cheval au moyen age," Bulletin philologique et historique (jusqu'a 1610), 1976 (Paris, 1978), pp. 220-49; and Robert‑Henri Bautier and Anne-Marie Bautier, "Contribution a l'histoire du cheval au moyen age: l'evelage du cheval," ibid., 1978 (Paris, 1980), pp. 9-75, provide a consider–able body of data on medieval horses, but unfortunately in the latter the authors subscribe (p. 10) to Lynn T. White, Jr.'s, now untenable argument concerning the importance of stirrups.

4. For a description of the peculiarities of the Arab horse, see the discussion by George Simpson, Horses (New York, 1961), p. 46.

5. Vegetius, Digestorum artis mulomedicinae libri IV, ed. E. Lommatzsch (Leipzig, 1903), Bk. 3, ch. 6.5, for the Huns' horses; for those of the Mongols, see D. Sinor, "Horse and Pasture in Inner Asian History," Oriens Extremus, 19 (1972), 175-79.

6. Knud Thorwildsen, The Viking Ship of Ladby (Copenhagen, 1967), pp. 7-16, concerning Viking ponies.

7. S. Kdrner, The Battle of Hastings, England, and Europe, 1035-1066 (Lund, 1964), pp. 27-271, has some useful observations on this point. See further Bernard S. Bachrach, "On the Origins of William the Conqueror's Horse Transports," Technology and Culture, 26 (1985), 306-09, for additional discussion of the value of the pictorial materials in the Bayeux Tapestry.

8. See, in general, L. Larson, The King's Household in England before the Norman Conquest (Madison, 1904), pp. 156-57.

9. Concerning the great Viking axe used by the housecarls, see: R. E. Oakeshott, The Archaeology of Weapons (New York, 1969), pp. 154, 177-78; T. Wise, 1066: Year of Destiny (London, 1979), pp. 59-60, 80-82; R. E. Oakeshott, A Knight and His Weapons (London, 1964), pp. 37-39; and J. Mann, "Arms and Armour," The Bayeux Tapestry, ed. F. M. Stenton et al. (London, 1957), p. 66.

10. Bayeux Tapestry, ed. Stenton, pls. 1, 2, 9-11, 15, 16, 20-22, 41, 46, 54-72, provides reasonable examples.

11. Concerning swords, see R. E. Oakeshott, The Sword in the Age of Chivalry (New York and Washington, 1964), pp. 25-55; and with regard to spear shafts. Oakeshott, Archaeology of Weapons, p. 96; and Wise, Year of Destiny, p. 58.

12. An easily available account of Harold is to be found in David Howarth,

1066 the Year of Conquest (New York, 1981), pp. 106-11, and p. 107 concerning his height.

13. I had the opportunity to go through the entire corpus of material at the Index of Christian Art at Princeton on three separate occasions (1964, 1971, 1972), and in 1980 and 1983, when I visited at Dumbarton Oaks, I reviewed several aspects of my research. A fair sample of what I found is discussed and presented in Bernard S. Bachrach, "Animals and Warfare in Early Medieval Europe," Settimane di Studio del Centro Italiano di Studi sull'alto Medioevo, 31 (Spoleto, 1985), pp. 707-64.

14. The basic research is cited and discussed from a methodological point of view by Donald Engels, Alexander the Great and the Logistics of the Macedonian Army (Berkeley, 1978), p. 178; additional evidence on horses from Eastern Europe that are midway in size between the smaller steppe animals and the larger Western horses is treated by S. Bokonyi, "Mecklenburg Collection, Part I: Data on Iron Age Horses of Central and Eastern Europe," Bulletin of the American School of Prehistoric Research (Peabody Museum, Boston), 25 (1968), 36, for a chart of the data, and note also the charts provided by C. Ambros; and H.-H. Muller, Fruhgeschichteliche Pferdeskelettfunde aus dem Gebiet der Tschechoslowakei (Bratislava, 1980), pp. 11-13, 23, for the charts which show the majority of stallions in the 14-hand range, with some considerably larger.

15. In honorum Hludowici, Bk. 2, lines 1116 ff. and 1126 ff. Hermold le Noir, Poeme sur Louis le Pieux et epitres au Roi Pepin, ed. and trans. E. Faral (Paris, 1932).

16. Major General Sir Frederick Smith, A Manual of Veterinary Hygiene (New York, 1906), p. 761.

17. Malcolm Vale, War and Chivalry (Athens, GA, 1981), p. 184, provides some good examples.

18. See, for example, the data developed by Tylden, Horses and Saddlery, pp. 45-46; and Engels, Alexander the Great, p. 129.

19. Mann, "Arms and Armour," pp. 56-69, has provided the basic examina–tion of the arms and armor in the Bayeux Tapestry; however, this study is badly in need of revision and expansion. See also, with regard to the weight of the armor of the period, Wise, Year of Destiny, p. 108; Claude Blair, European Armor, circa 1066 to circa 1700 (New York, 1959), p. 192.

20. The most recent and useful treatment of the Battle of Hastings is R. A. Brown, "The Battle of Hastings," PBC, 3 (1980), ed. R. A. Brown, 1-21, 97‑201. But see the critical comments by Bernard S. Bachrach, "Charlemagne's Cavalry: Myth and Reality," Military Affairs, 47 (1983), n. 41.

21. Lex Alamannorum, 62.1, 2, 3, ed. K. Lehmann, MGH, Leges, sect. 1.5.1, rev. ed. (Hannover, 1966); Lex Baiwariorum, 14.11, 12, ed. E. Von Schwind, MGH, Leges, sect. 1.5.2 (Hannover, 1926); Lex Gundobada, 4.1, ed. L. de Salis, MGH, Leges, sect. 1.2.1 (Hannover, 1892); and Lex Ribuaria, 40, 11, ed. R. Buchner, MGH, Leges, sect. 1.3 (Hannover, 1954).

22. See A. Leighton, Transport and Communication in Early Medieval Europe, A.D. 500-1000 (Devon, 1972), pp. 41-42, 163, on the basis of evidence from Domesday Book. See also J. G. Landels, Engineering in the Ancient World (Berkeley, 1978), p. 178.

23. See n. 18 above.

24. Vegetius, Mulomedicina, Bk. 2, prologus. It is perhaps of some impor–tance that these grain-to-hay ratios and the absolute necessity of the former for a healthy horse were already well known to the writers of classical Greece, as shown by J. K. Anderson, Ancient Greek Horsemanship (Berkeley, 1961), pp. 92-94.

25. U.S. War Department, Manual for Stable Sergeants: 1917 (Washington, DC, 1917), p. 64.

26. Vegetius, Mulomedicina, Bk. 2, prologus. Some useful observations are to be found in R. Lindner, "Nomadism, Horses and Huns," Past and Present, 92 (1981), 3-19.

27. See n. 6 above.

28. Victor Vitensis, Historia persecutionis Africanae Provinciae sub Geiserico et Hunirico regibus Wandalorum, ed. C. Halm, MGH, AA (Berlin, 1879), Bk. 2, ch. 12.

29. Vita Hludowici Imperatoris, Bk. 1, ch. 15, ed. G. Pertz, MGH, SS (Hannover, 1827).

30. Engels, Alexander the Great, pp. 18, 126-29. For a discussion from a snore modern perspective, see Manual for Stable Sergeants, pp. 62-63.

31. Manual for Stable Sergeants, pp. 61, 63.

32. Bachrach, "Animals and Warfare in Early Medieval Europe," pp. 717-23. Each horse required at least one exercise ride of 20 mi. per week and a minimum of 2.5 hours per day of exercise under saddle.

33. The extensive water requirements of horses have been widely recognized. See, for example, Engels, Alexander the Great, p. 127; Anderson, Ancient Greek Horsemanship, p. 95; and Manual for Stable Sergeants, p. 61.

34. William is generally agreed to have had some two to three thousand horses in combat at Hastings. See the consensus as presented by R. A. Brown, The Normans and the Norman Conquest (New York, 1968), pp. 150-51.

35. Gesta Francorum et Aliorum Hieroslimitanorum, or the Deeds of the Franks and Other Pilgrims to Jerusalem, ed. and trans. R. Hill (London, 1962), p. 22. Note Hill's trans. on p. 23.

36. Most recently John Langdon - "Horse Hauling: A Revolution in Vehicle Transport in Twelfth- and Thirteenth-Century England?" Past and Present, 103 (1984), 37-66 - has shown that in England oxen dominated as the major power for plowing through the Middle Ages; however, in the late twelfth century horses were introduced for hauling carts, and this mode became more and more popular as time passed. Wagons were rarely used before the late thirteenth century, and then they were generally pulled by oxen. Before the development in the use of the horse-cart, pack horses dominated for hauling of goods in England. The efforts by A. Leighton - "A Technological Consideration of Early Medieval Vehicles," Fifth International Conference on Economic History: Leningrad 1970 (The Hague, 1977), pp. 346-48; and "Eleventh Century Developments in Land Transport Technology," The Eleventh Century; Acta, 1 (Binghamton, 1974), 20-22 - to show that horse transport became important in the late eleventh century are not soundly based. With regard to the basic capacity of wagons and carts, see Leighton, Transport and Communication, pp. 77, 161; and Engels, Alexander the Great, pp. 14-16. Marjorie Nice Boyer, "Medieval Pivoted Axles," Technology and Culture, 1 (1960), 135, is probably correct in arguing that the whippletree-pivoted front axle system was available in the thirteenth century. The thrust of the evidence is toward the latter part of the period. Compelling pictorial and written evidence comes only in the mid-fourteenth century for the latter and later for the former.

37. Lynn T. White, Jr., Medieval Technology and Social Change (Oxford, 1962), pp. 60-62, has shown that by the late ninth century some innovations in harnessing technique were accomplished; however, the major question is one of diffusion. Thus, for example, plates in The Bayeux Tapestry, ed. Stenton, show that in England by the late eleventh century a horse is being used to pull a harrow (pl. 12) and a mule to pull a plow (pl. 1 1); however, both use the old choke collar common in the Roman Empire which cuts the hauling capacity of the animal in half. Pl. 41 shows a four-wheeled vehicle with no whippletree and flexible traces pulled by two men. It carries a wine barrel that appears to hold in the neighbor–hood of 50 gal. with a weight of about 400 lb. However, as we have seen above, n. 36, the use of oxen dominated in agricultural work and the examples in the Bayeux Tapestry mentioned above would appear to have been aberrations.

38. See the works cited above in n. 36.

39. See n. 36 and, particularly concerning the feeding of oxen, John L. Stone, "Computing Rations for Farm Animals," in Isaac Roberts, The Horse (New York, 1905), p. 364.

40. Jacques Boussard, "La vie en Anjou aux Xle and XIIe siecles," Le Moyen Age, 56 (1950), 55.

41. Bernard S. Bachrach, "The Angevin Strategy of Castle Building in th Reign of Fulk Nerra, 987-1040," The American Historical Review, 88 (1983), 54 n. 8, to which should be added Marjorie Nice Boyer, "A Day's Journey in Medieval France," Speculum, 26 (1951), 597-608.

42. Maurice's Strategicon: Handbook of Byzantine Military Strategy, trans, G. T. Dennis (Philadelphia, 1984), Bk. 1, ch. 5. Under very good conditions perhaps, well-trained men could use four horses. It cannot be assumed that Westerners handled this matter more efficiently than the Byzantines.

43. See a detailed discussion in Bachrach, "Animals and Warfare in Early Medieval Europe," p. 723.

44. Livestock Waste Facilities Handbook, ed. G. L. Bratt et al. (Ames, 1975), p. 3.

45. For the numbers of horses in William's force, see n. 34 above.

46. Concerning the Western tradition on these points, see Anderson, Ancient Greek Horsemanship, pp. 89-90, 95, 164.

47. Lex Ribuaria, 40, 11, both the early text and the Carolingian revision.

48. See n. 43 above.

49. For the background, see Anderson, Ancient Greek Horsemanship, p. 94; Vegetius, Mulomedicina, Bk. 1, ch. 56, 11-13; and for the application of the modern scientific understanding of these principles, see U.S. War Department, Cavalry Service Regulations (Washington, DC, 1914), p. 123.

50. Engels, Alexander the Great, pp. 153-58, with the scientific literature.

51. See the evidence cited in Bernard S. Bachrach, "The Cost of Castle Building," The Castle: Myth and Reality, ed. K. Reyerson and Fay Powe (Dubuque, IA, 1984), p. 61 nn. 48, 49. The discovery by Langdon ("Horse Hauling," pp. 46-58) that England in the late thirteenth century saw the beginning of the use of horse-carts should be related to these seed-to-grain ratio rises.

52. Bachrach, "Animals and Warfare in Early Medieval Europe," pp. 708-16.

53. This is widely recognized. See, for example, for the early period, Bernard S. Bachrach, Merovingian Military Organization (Minneapolis, 1972), pp. 127-28; Bachrach, "Charlemagne's Cavalry," pp. 181-87. For the pre-crusade era, see Bachrach, "The Angevin Strategy of Castle Building," pp. 533-60. For the later period, R. C. Smail, Crusading Warfare: 1097-1193 (Cambridge, 1956), p. 39, maintains that "the true end of military activity was the capture and defense of fortified places." Philippe Contamine, La guerre au Moyen Age (Paris, 1980), p. 207. observes: "Dins sa forme la plus courante, la guerre medievale etait faite d'une succession de sieges . . . ."

54. Historians often play with the notion of decisive battles. Joseph Dahmus, for example, has recently identified seven, but only four - Chalons, Hastings, Bouvines, and Crecy - occurred in Western Europe (Seven Decisive Battles of the Middle Ages [Chicago, 1983]).

55. See Bachrach, "Charlemagne's Cavalry," pp. 183-84, 186 n. 41, 187.

56. For Suntal, Pontlevoy, and Hastings, see notes below. Concerning Crecy, see Dahmus, Seven Decisive Battles, pp. 169-96. Concerning Saint Michel en l'Herm and Conquereuil, see Bernard S. Bachrach, "Toward a Reappraisal of William the Great, Duke of Aquitaine (995-1030)," Journal of Medieval History, 5 (1979), 14; and for the later battles, J. F. Verbruggen, The Art of Warfare in Western Europe During the Middle Ages from the Eighth Century to 1340, trans. W. Willard and S. C. M. Southern (Amsterdam, 1977), pp. 100-83. For the battle of Lechfeld, see Bachrach, "Animals and Warfare in Early Medieval Europe," p. 727 n. 74.

57. The best account of this battle is by Verbruggen, The Art of Warfare, pp. 220-37.

58. Annales Vedastini, an. 885, ed. B. de Simson, MGH, SRG (Hannover, Leipzig, 1919).

59. This battle is discussed in some detail by Bernard S. Bachrach, "Charles Martel, Mounted Shock Combat, The Stirrup, and Feudalism," Studies in Medieval and Renaissance History, 7 (1970), 51-53.

60. Annales Fuldenses, an. 891, ed. and trans. Reinhold Rau, Ausgewahlte Quellen zur deutschen Geschichte des Mittelalters (Berlin, 1960).

61. Annales Fuldenses, an. 891.

62. See Bernard S. Bachrach, "The Practical Use of Vegetius' De re militari during the Early Middle Ages: Some Methodological Observations," The Historian, 47 (1985), 239-55.

63. Contamine, La guerre au Moyen Age, pp. 353-56. My own research on Vegetius is being extended well beyond the article cited above in n. 62 and will concentrate upon glosses, interpolations, and fragmenta.

64. De procinctu Romanae militiae, ed. E. Dummler, in Zeitschrift fur deutsches Altertum, 15 (1872), p. 450.

65. De procinctu Romanae militiae, ch. 12.

66. De procinetu Romanae militiae, eh. 12.

67. De procinctu Romanae militiae, ch. 3.

68. Regino, Chronicon, an. 891, ed. B. Simson, MGH, SRG (Hannover, 1890).

69. Nithard, Histoire des fill de Louis le Pieux, ed. and trans. Ph. Lauer (Paris, 1926), Bk. 3, ch. 6. Cf. the translations by Lauer, pp. 111, 113; and Carolingian Chronicles, trans. B. Scholz and B. Rogers (Ann Arbor, 1970), p. 64. To translate exercitium as "exercise" is misleading as that word is generally used today in the United States. Nithard clearly means "practice" or "training." See, for example, Rabanus, De procinctu Romanae militiae, ch. 3, "exercitio et disciplina."

70. Larry D. Benson, "The Tournament in the Romances of Chretien de Troyes and L'Histoire de Guillaume Le Marechal," Chivalric Literature: Essays on relations between literature and life in the later middle ages, ed. Larry D. Benson and John Leyerle (Kalamazoo, MI, 1980), p. 8.

71. The basic sources are Widukind, Bk. 1, ch. 38, ed. H. Lohmann and P. Hirsch, MGH, SRG (Hannover, 1935); and Liudprand, Antapodosis, Bk. 2, ch. 31, ed. J. Becker, 3rd ed. MGH, SRG (Hannover, Leipzig, 1915).

72. Liudprand, Antapodosis, Bk. 2, ch. 31.

73. Verbruggen, The Art of Warfare, pp. 84-85.

74. See Bachrach, "Charlemagne's Cavalry," p. 183, concerning this battle and the likelihood that Einhard obtained his information from an eyewitness.

75. Annales qui dicuntur Einhardi, an. 782, ed. F. Kurze, MGH, SS in US (Hannover, 1895).

76. T. S. Burns, "The Battle of Adrianople, A Reconsideration," Historia, 22 (1973), 336-45.

77. Historia Sancti Florentii Salmurensis, ed. P. Marchegay and E. Mabille, in Chroniques des eglises d'Anjou (Paris, 1869), p. 274; Gesta Ambaziensium Dominorum, p. 82, and Gesta Consulum, pp. 52-53, both in Chroniques des comtes d'Anjou et des seigneurs d'Amboise, ed. L. Halphen and Rene Poupardin (Paris. 1913).

78. Annales de Saint-Bertin, an. 876, ed. Felix Grat et al. (Paris, 1964);

Regino, Chronicon, an. 876; Annales Vedastini, an. 876; and Annales Fuldenses, an. 876. J. F. Verbruggen, in "L'Art militaire dans Fempire carolingien (714-1000)," Revue belge d'histoire militaire, 23.5 (1980), 393-94, provides a satisfactory account of the battle but does not seem to appreciate that the East Franks attacked the West Franks on the flank.

79. Epitoma rei militaris, ed. Carl Lang (Leipzig, 1885), Bk. 3, chs. 10, 19, covers these points.

80. See the discussion by Rosamond McKitterick, "Charles the Bald (823-877) and his Library: The Patronage of Learning," The English Historical Review, 95 (1980), 31.

81. Brown, "The Battle of Hastings," pp. 1-21.

82. Verbruggen, The Art of Warfare, pp. 159-62.

83. Concerning the Swiss, see Charles Oman, The Art of War in the Middle Agcs, ed. J. Beeler (Ithaca, NY, 1953), pp. 77-79.

84. See the discussion by Verbruggen, "L'Art militaire dans l'empire carolingien," p. 396.

85. The major breakthrough is due to the work of Verbruggen, The Art of Warfare, pp. 72-76; and see Vale, War and Chivalry, pp. 103-04, who discusses the wide acceptance of Verbruggen's thesis. Literary scholars have begun to appreciate this, as illustrated by J. Bumke, The Concept of Knighthood in the Middle Ages, trans. W. T. H. and Erika Jackson (New York, 1982), p. 40; the author, however, is not sure what to make of the fact that "modern research stresses that there was a highly developed art of war even in the Middle Ages." See also p. 206 n. 77.

86. In Christian sources of a non-technical nature, the number 300 can be of potential trouble because of the penchant of chroniclers to be influenced by symbolism. In the numerical-symbolic system called gematria, 300 equals the Greek letter tau, which itself is the letter that stands for the cross of Christianity. Therefore when a force of 300 men is said to support a hero in battle, it may be the intention of the chronicler to suggest that he was enjoying the protection of the cross. Concerning gematria, see Vincent Hopper, Medieval Number Symbolism (New York, 1938), pp. 75-76, concerning the number 300. With regard to units of 300 men, see nn. 87, 88 below.

87. E. Joranson, "The Inception of the Career of the Normans in Italy - Legend and History," Speculum, 23 (1948), 353-96, examines all of the relevant sources, and, although his aim is not to discuss military organization, a careful study of the account he treats enables us to understand that the Normans brought little of an institutional nature with them. They did, of course, come with their horses and weapons. See pp. 358, 371, but cf. 359-60, 365, where the topoi of pilgrim-legend distort the picture of Norman practice. Concerning Byzantine military organization, see A. Toynbee, Constantine Porphyrogenitus and his World (London, 1973), pp. 287-88.

88. Amatus of Monte Cassino, Ystoire de li Normant, ed. V. de Bartholomaeis, Fonti per la Storia d'ltalia, Scrittori (Rome, 1935), Bk. 2, chs. 8, 18; Annales Romani, ed. G. H. Pertz, MGH, SS (Hannover, 1844), 5, 471; Geoffrey Malaterra, Historia Sicula, ed. E. Pontieri, Rerum italicarum scriptores, new ed. (Bologna, 1928), Bk. 2, chs. 8, 10, 17, provide evidence for the use of 300 man Norman units and for the integration of Norman units of 300 into Byzantine operations. There also may be reason to believe that the Normans in Normandy used 300 man units in the tenth century, although the evidence is problematic. On William Longsword's following of 300, see Verbruggen, The Art of Warfare, p. 68. D. P. Waley, "Combined Operations in Sicily, 1060-78," Papers of the British School at Rome, 22 (1954), 118-19, discusses the Norman use of Byzantine transports. Now see Bachrach, "William the Conqueror's Horse Transports," pp. 513, and 527 n. 50.

89. C. Warren Hollister, The Military Organization of Norman England (Oxford, 1965), pp. 31-42, reviews the evidence on the constahularia. More recently, Emily Zack Tabuteau, "Definitions of Feudal Military Obligations in Eleventh-Century Normandy," in On the Laws and Customs of England: Essays in Honor of Samuel E. Thorne, ed. M. S. Arnold et al. (Chapel Hill, 1982), 18-59; and Marjorie Chibnall, "Military Service in Normandy before 1066," PBC, 5 (1982), 65-67.

90. Panegyricus Avitis, lines 289-92 in Epistolae et Carmina, ed. C. Luetjohann, MGH, AA (Berlin, 1887), 8.

91. Liudprand, Antapodosis, Bk. 1, ch. 21. Is the Bavarian riding a la jineta, i.e., with short stirrups, to obtain the mobility described by Liudprand?

92. "Plein Sa Hanste," Medium Aevum, 20 (1951), 10.

93. Brown, "The Battle of Hastings," p. 12, discusses Ordericus's observation and finds it compelling.

94. Bachrach, "Animals and Warfare in Early Medieval Europe," pp. 747-48. See also the discussion of the mechanisms of combat in the article by Rosemary Ascherl in the present volume.

95. Concerning the introduction of stirrups, see Bernard S. Bachrach, "A Picture of Avar-Frankish Warfare from a Carolingian Psalter of the Early Ninth Century in Light of the Strategicon," Archivum Asiae Medii Aevi, 4 (1986), 5-27.

96. Bachrach, "Animals and Warfare in Early Medieval Europe," pp. 737-48.

97. Bachrach, "Animals and Warfare in Early Medieval Europe," pp. 747-48; Victoria Cirlot, "Techniques guerrieres en Catalogne feodale: le maniement de la lance," Cahiers de civilisation medievale, 28 (1985), 35-43, arrives at similar conclusions for Catalonia from Catalan materials.

98. Concerning the great jousting helms and the smaller and lighter field equipment, see Vale, War and Chivalry, pp. 106-11.

99. See, for example, John Keegan, The Face of Battle (New York, 1976), pp. 97-102, for an account of the French horsemen who had dismounted to fight on foot.

This article was first published in The Study of Chivalry: Resources and Approaches, edited by Howell Chickering and Thomas H. Seiler (Kalamazoo: Medieval Institute Publications, 1988).  We thank Bernard S. Bachrach and the Medieval Institute at Western Michigan University for their permission to republish this article.  Interested readers can learn more information about this volume and purchase it at the website for Medieval Institute Publications.