In 1147 a Frankish attempt to take the town of Bosra ended in complete failure. The army retreated under the most difficult circumstances but, at a time when morale was at its lowest, spirits were raised by the exploits of one particularly brash cavalryman. William of Tyre tells us that ‘a certain knight’ broke ranks (contrary to their instructions) and charged towards the enemy. He struck a Muslim emir with a spear and then followed this up by running him through with his sword. He came back to the crusaders’ lines accompanied by loud cheers, but facing disciplinary action for having disobeyed orders.
The moral of the story, from William’s point of view, was to explain how the knight’s success was sufficient for him to get away with infringing army discipline, and how he avoided punishment ‘rather because of the result, than because it was right’. Far more surprising for a modern reader, however, is the line of defence with which the soldier escaped censure: he was Turkic, and could not understand Frankish commands.
Just as most aspects of Frankish warfare, from grand strategy down to the minutiae of tactical manoeuvres, were deceptively different from European practice, so too were the troops with whom that warfare was conducted. What we perhaps think of as archetypically ‘European’ armies were actually very different on the ground.
When looking at different kinds of troops, it is important to resist the temptation to be overly rigid or definitive. These were desperate times, with manpower in short supply. There have been lengthy debates, for example, about whether a ‘sergeant’ was a mounted soldier or an infantryman. In practice, we can find instances of both, as both were needed. The way men were used was largely dictated by need, availability and skills, rather than by a manual. If cavalry were in short supply, or if the army was expected to be on campaign for a long period and over extended distances, any sergeant who had a horse, or who could be provided with one, would doubtless have gone mounted. If fighting nearer home, on the other hand, a sergeant with a crossbow and expertise in how to use it would have been deployed on foot.
Mercenaries had begun to specialise depending on their preferences and expertise, and the availability of equipment. Some, for instance, were equipped and trained to ride in the second rank of the charge, supporting the knights. Others would have specialised in the use of bow or spear. Similarly, there has been a lot of discussion about the ethnicity of the Turcopole recruits in Frankish armies, but it is hard to imagine the circumstances in which a man with the skills to make a good mounted archer would be turned away, regardless of his ethnic origins or denominational preferences.
The kinds of troops involved would also inevitably depend on the type of actions envisaged and the men available at the time. There was no uniform, ‘one size fits all’, Frankish army. A long-range patrol in the desert areas south of Kerak or Darum, for instance, would have had a higher than usual proportion of cavalry, particularly lighter cavalry, and would have included local Bedouin guides and auxiliaries, some of whom would probably have been riding on camels. Any infantry on the patrol would perhaps have been mounted on mules or other lower-value baggage animals. A frontier garrison might have been a small but relatively professional group, probably with a disproportionate number of crossbowmen. Similarly, a force that had to move fast to relieve a castle or go to the aid of another crusader state would have had more mounted men and a high proportion of ‘regular’ and ‘professional’ troops in it (though both terms need to be used in a less demanding, medieval way, rather than in a modern context).
At the opposite end of the spectrum, in the case of a local incursion all able-bodied men might be called up for emergency defence, at least until non-combatants could be taken to a place of safety. This scratch defence force would be reinforced or relieved as soon as possible by regional troops led by the local nobility or the nearest troops from the garrisons of the military orders. And if a full-scale invasion was under way, as was increasingly the case in the 1170s and 1180s, the size of the army would increase, but with inevitable compromises with regard to the expertise and training of those who were mobilised.
It is important to look at different categories of troops within the armies of the crusader states, but we need to be continually aware that any definitions we might try to impose would have been far more flexible in practice, and subject to major variation.
Knights were the celebrities of Frankish society and the main battle tanks of the army. They shared some of the characteristics of both. Small in number and extremely high profile, they could win a battle in minutes, even when facing seemingly overwhelming odds. But there were other similarities too. Like celebrities, they could be fragile and fractious. And like main battle tanks, they were few in number and strangely brittle: vulnerable if they were not carefully supported or if their attack was launched under adverse conditions.
The knight was an expensive but highly trained and adaptable weapons system. They could be deployed on horseback or on foot, using a lance in the charge, a sword or mace in close combat, or perhaps, less commonly, wielding a crossbow or bow during a siege. The knight was versatile and an excellent all-round warrior but he could deliver most value for money in the shock of impact. When dismounted as an armoured infantryman, they became a solid and reliable fixed point for an army. But there were many other infantrymen who were almost as good in that role, and far less expensive. Where the knight excelled, where he was unique, was in the charge. It was in the charge that he could, if everything went well, win the battle.
Luckily for Saladin and the others who faced them, there were never many knights available at any given time. They were formidable but expensive. A primarily rural economy with limited economic surplus could never support as many knights as were needed to defend the Holy Land. Every effort was made to force fief-holders to fulfil military obligations even when they were personally unable to do so. Widows had to pay for mercenaries. The disabled and the elderly had to hand over their horses and armour to someone who could do the job for them.
Over time, the strain started to show. In 1150, for instance, when Baldwin III needed to go to the aid of Edessa, many of his noblemen in the south of the kingdom refused a series of personal summonses from the king to join the muster. This was partly because of tensions between the king and his mother, Queen Melisende, but also a reflection of the resentment engendered by the remorseless personal demands of chronic warfare across the entire region. The unremitting nature of crusading warfare had a disproportionate impact on the knights, whose small numbers and vital battlefield role meant that they were called upon to campaign and fight with appalling frequency.
We have a fascinating list of knightly service owed to the king of Jerusalem by the leading fief-holders. The ‘servitia debita’ lists of John of Ibelin were written around 1265 but internal details suggest that they set out the state of affairs in the Latin Kingdom of Jerusalem at some time between 1183 and 1186. The lists are beguiling. On one level they seem to be a full review of the knights available to the king, divided into each of their feudal contingents. They also seem to be relatively accurate, and on one level they are. But that is misleading: they are perhaps better viewed as a legal document, setting out the rights and obligations of the nobility, rather than as a muster sheet for the military.
The list shows that the range of knightly service due to the king varied enormously, depending on the size and economic value of the fief. At the top end of the scale the largest of the lordships, the major baronies, such as Sidon, Galilee and the County of Jaffa and Ascalon, each owed the service of 100 knights. At the opposite end were the fifty-eight fief-holders who owed the service of just one knight: unless they were too old or too sick, they would probably have served in person.
If everyone fulfilled their obligations the total number of knights was around 675. We do not have comparable lists for the other crusader states, but the Principality of Antioch seems to have had a broadly similar level of knightly service. When Roger of Antioch went into battle at Ager Sanguinis in 1119, without waiting for help from the other Frankish armies, his knightly contingent was said to be some 700 strong. The County of Tripoli was always smaller and had a more limited number of knights. The County of Edessa was similarly smaller in scale, and even this contingent was dramatically reduced after the loss of much of its lands in 1144 (with the county being completely lost by 1150).
On the basis of these legal obligations, the combined total of the feudal knightly contingents available across all four of the crusader states amounted to just 2,000 men: approximately 700 each from Jerusalem and Antioch, with Edessa and Tripoli perhaps having access to another 300 men each. These figures are not necessarily a minimum: there were always bound to be some knights who were too ill to attend a muster, some who were too old, or who were prisoners of war.
In practice, however, the lists almost certainly underestimate the number of knights available. On the few occasions when we come across details about the composition of feudal contingents in the field, the numbers are invariably different. In 1158, for instance, the future King Amalric arranged for fifteen men to act as witnesses for a legal document. Ten of these were described as vassals, who would naturally have been included on a list such as that compiled by John of Ibelin. A further five knights, however, were described as retainers (‘stipendarii’) who would not have been. Similarly, in 1232 Balian, son of John of Beirut, was reported as riding into battle with a unit of five knights. Of these, only two knights (Philip of Novara and Raymond of Fiace) were his vassals. The others were household retainers who had been knighted by Balian (Robert of Maumeni and Eudes of La Fierte) and a highly regarded mercenary (Peter of Montolif). In both instances, where we have a list of knights, there are significant numbers present who were not direct vassals (and who therefore would not appear on any traditional service lists). Additional knights were clearly being used to bolster the size of feudal contingents by those who could afford to do so.
The requirements of feudal service were extremely demanding. A knight could be asked to fight for a full year if the homeland had been invaded, in what was effectively full-time conscription, or for four months if required to serve beyond its borders or when going to the aid of other crusader states. Feudal service meant that a knight was expected to serve until he was 60, though over the age of 40 he could provide another knight to deputise in his place. In an age of low life expectancy, in which warfare was brutal and dependent on muscle and endurance, the military elite were continually expected to justify their status. Being a knight in the Frankish east was never a sinecure. This unprecedented level of commitment, with all the hardships and danger that went with it, was inevitably unpopular from time to time. There were occasional undercurrents of unrest, normally after several consecutive years of distant campaigning. Yet outright refusals to serve were surprisingly rare. Ultimately, the interests of the men defending the Holy Land were aligned: their continued survival always depended on effective military action, and there was little point complaining about it.
Although, as we shall see, Frankish ‘crusader’ armies were different from their European counterparts in terms of how they functioned, they were very similar in terms of equipment. Indeed, as far as we know, much of the armaments used in the crusader states continued to be manufactured in Europe and, for most practical purposes, knightly armour was largely unchanged during most of the twelfth century.
The effectiveness of their armour and equipment was reflected in the sheer difficulty of trying to stop them. Anna Comnena, the Byzantine historian and princess, reflected the views of her father and the Byzantine military establishment when she wrote that the Franks’ ‘cuirasses and coats of mail made them almost, if not entirely invulnerable . . . [Frankish] armour consists of a tunic interwoven with iron rings linked one with another; the iron is of good quality, capable of resisting an arrow and giving good protection to the soldier’s body. This armour is supplemented by a shield . . .’ The quality of their protection meant that even in battle against very heavily armed opponents, casualties among the men could be surprisingly light. Writing of the battle of Brémule in 1119, Orderic Vitalis tells us that ‘in the battle of the two kings, in which about 900 knights were engaged, only three were killed. They were all clad in mail and spared each other on both sides, out of fear of God and fellowship in arms.’
The Rule of the Templars (the twelfth-century ‘best practice’ guide to life in the order) makes it clear that knights of the period were extremely well equipped. The brothers had several layers of protection, a range of weapons for contact with the enemy and good horses to carry them. Each was a medieval armoury in microcosm. Rule 138 shows that their knights were equipped almost to the point of overkill, stating that:
[each] knight brother of the convent should have three horses and one squire and a fourth horse and a second squire, if he has them, are at the discretion of the Master [of the Templars], and they should have . . . a hauberk [i.e. a coat of mail], iron hose, a helmet or chapeau de fer, a sword, a shield, a lance, a Turkish mace, a surcoat, arming jacket [a padded jerkin worn under the hauberk], mail shoes, and three knives.
The head and face were always particularly vulnerable and received a great deal of attention. At the beginning of our period the helmets worn by the knights and many of the better armed infantry would be very familiar to us from the Bayeux tapestry and from a mass of other artistic representations. These conical helmets generally had a distinctive nasal bar to give some facial protection without detracting too much from visibility. They protected the head from enemy archery and most horizontal and downward sword thrusts. That very familiarity is a function of the universality of the design among heavily armoured European troops of the time, giving a deceptive sense of uniformity in a world without mass production techniques. These helmets were worn over a mail coif, looking rather like an open-faced ski mask, which gave additional protection to the neck, face and throat. The coif could either be attached to the hauberk or worn as a separate item. A leather cap was worn underneath the coif and helmet, for comfort and as yet another layer of protection.
This traditional conical helmet and nasal guard remained in use throughout the period but by the middle of the twelfth century we find other types emerging alongside it, as a number of round-topped or cylindrical helmet designs were gradually introduced. Most significantly, increasing efforts were made to protect the face, particularly important in light of the high volume of projectiles that were inevitably directed at the front ranks of any charge. Over time we find that the nasal guards became bigger, covering parts of the face as well as the nose. By 1170 we also see a more radical approach to this same problem. Face guards began to appear on helmets, sometimes as an integral feature, and sometimes added on.
The other core element of defensive equipment was the hauberk, the main body armour worn by a knight. It was an essential piece of equipment but it did not come cheap. A full suit of mail took a minimum of 140 hours of highly skilled labour to construct, in addition to the cost of the metal, itself hugely expensive in a pre-industrial society. The mail shirt was built up by adding row after row of rings, with the rivet heads facing outwards to avoid damage to the padded leather jerkin worn underneath. A mail hauberk and jerkin may not have made the wearer invulnerable, but they did provide excellent protection.
The main objective was, of course, to prevent penetration by enemy projectiles or weapons, and it did so by trying to distribute the impact of any blow across as wide an area as possible. The larger the number of rings (and hence the higher the cost of the hauberk), the greater the extent to which the force could be dissipated. The other key variable was the weight and flexibility of the armour. Mail was a good way of ensuring that its wearer retained mobility, but if the rivets were made stronger and thicker, there would be an inevitable decrease in flexibility. The hauberk would reach down towards the knight’s knees and would be split front and back to allow more comfortable riding. Most would have weighed around 25lbs and, with the weight distributed evenly across the shoulders, would not have presented too much of a problem for a fit soldier. But there was always a trade-off between cost, protection and mobility.
At the upper end of the social scale, and for those with bigger budgets, there were additional refinements. By the time of the First Crusade, knightly hauberks commonly extended down to the wrists, but by the end of the century they often also included mailed gloves or mittens (also known as ‘mufflers’) for the hands. Mail leggings (‘chausses’) also seem to have become increasingly common among knights in the second half of the twelfth century. This was not just a fashion statement. There were very practical reasons why one would want to cover any exposed limbs. The bodies of the doomed garrison at Jacob’s Ford, destroyed by Saladin’s forces in 1179, show a range of extremely unpleasant wounds, many of which were associated, not surprisingly, with the parts of the anatomy that were traditionally less well protected by armour. The surviving skeletal remains were found in the back of a burnt-out building within the castle, where some of the Templar soldiers gathered to make a final stand. As well as the usual arrow wounds that one would expect to find when facing Turcoman nomads, the bodies showed signs of traumatic wounds to the limbs. One sword cut was so devastating that the Frankish soldier on the receiving end of it had his arm completely severed at the elbow. Another soldier had his lower jaw split into two, again, at a point where a nasal helmet would have offered less protection to its wearer. Defensive armour was always a wise investment if you could afford it.
The large kite shield provided yet another layer of protection, particularly along the front and left side of the charging knight. It gave a good degree of cover for the left leg and the effectiveness of the shield may go at least part of the way towards explaining why mail leggings were not universally worn in the early part of our period. The shields were made of a leather covering on top of a wooden backing, reinforced by a metal boss in the middle and metal strips around the edges of the shield. The kite shield was also curved along its primary axis and partially wrapped around the soldier carrying it, giving him even greater protection. This remained the dominant knightly shield design throughout the twelfth century, although the curved top of the shield became less pronounced as the century progressed.
Ironically, the most striking difference between a knight arriving at Jerusalem in 1099 and a knight fighting at Hattin in 1187 is the one we know least about. From the mid-twelfth century onwards we find knights wearing long, often sleeveless, fabric garments over their hauberks. This fashion was so popular that it must have brought benefits to those who adopted it. But there is still some speculation about what those benefits might have been. A partial explanation is the weather. A mail hauberk is never easy to wear but is significantly less comfortable under the desert sun or in a rainstorm. The surcoat may have originated in the Frankish armies of the Holy Land, as they copied the example of their Muslim opponents and adopted a loose-fitting and light-coloured covering for their body armour. This would have had the effect of simultaneously deflecting the sun’s rays and creating a breeze between the surcoat and their body armour. Similarly, for knights based in western Europe, a surcoat treated with animal fats may have provided a welcome element of waterproofing.
There were other, less tangible but nonetheless helpful benefits as well, however. The introduction of visors and full-faced helmets in the second half of the twelfth century led to greater protection for the face, particularly against enemy archery, but there was always a compromise between facial protection and visibility. In a mêlée a visor slit could easily be knocked to one side, making an already impaired visibility even worse and, with both hands fully occupied, it was very difficult to make adjustments. The surcoat allowed individualised patterns and bold colours to be incorporated in a knight’s appearance, creating a unique ‘branding’ and heightening visibility. This, combined with distinctive shield designs, aided visibility and unit cohesion, perhaps by increasing recognition of where one needed to form up or, in conjunction with standards, where one needed to regroup. It saved lives by encouraging opponents to pause and consider the possibility of ransom, rather than just cutting the throat of a fallen knight. Perhaps even more importantly for this ever competitive social elite, the surcoat increased visibility among one’s peer group: the point of great deeds of valour was having someone to see them and talk about them. Reputation among the knights was everything.
As we have seen, the knights were equipped with a range of weapons. The lance was the preferred weapon for the initial charge, approximately 4m long and ideally made of ash wood. These would often shatter on impact with the enemy, and as the mêlée slowed, lances would be dropped and weapons such as swords or maces drawn instead. Sword design was not subject to meteoric change in this period. Most were generally straight, double-edged weapons, broadswords which were very similar to those used by the Vikings hundreds of years earlier. They weighed about 3lbs or a little less, and were well suited for the slashing or hacking actions of a cavalry mêlée.
The lance typified the style of fighting for which the knightly class were best known. They also trained with a variety of other weapons, however, and the skills of the hunt were often transferable to military activity. Knights, competitive in all things, could also have a reputation for archery as well as the more traditional jousting skills that we usually associate with them. On the Third Crusade, for instance, we are told that ‘the Norman William du Bois, who was a very skilled archer, untiringly fired arrows and darts, this way and that, routing . . . [the enemy]’.
Similarly, while on campaign in the north against Nur al-Din in 1150, one of the most senior crusader knights, Humphrey of Toron, the constable of Jerusalem, had a bow to hand, with which he and presumably some of his men chased off a unit of Turkic light cavalry. According to William of Tyre, ‘Humphrey, the constable, armed with his bow, was pursuing the retreating infidels a little apart from the army when a soldier from the enemy’s ranks approached him’. Humphrey, who had linguistic as well as archery skills, could speak Turkic. He engaged in some banter with the enemy cavalryman and the encounter ended peacefully. Clearly even some of the more senior Frankish knights were capable of operating as mounted archers and, given that they were living in a part of the world where mounted archery was a central part of enemy military practice, it is perhaps not surprising that this should be so. It is also interesting that William of Tyre mentions this purely in passing, as an anecdote that speaks well of Humphrey’s bravery and skills but with no obvious expectation that his readers would find the idea of a Turkic-speaking horse archer Frankish lord outlandish or entirely unexpected.
Archery skills would also have been useful on garrison duty, one of the less glamorous but entirely typical roles which a knight would have been called on to fulfil. Evidence is slight but there is an interesting detail in the summary of court-martial proceedings taken against a Templar brother knight. He was found guilty of ‘losing’ two weapons that he had been entrusted with, while travelling from the castle of Blanchegarde down to Kerak, to the east of the Dead Sea. The suspicion must have been that he had sold the weapons, and he was eventually expelled from the order. It is interesting to see, however, that although one of his weapons was a sword (predictably enough), the other was a longbow. It seems likely that many knights, with little else to occupy them for lengthy periods of time, would have wanted to be able to pick up a longbow or crossbow in the event of a siege, helping the rest of the garrison to keep enemy miners and engineers at bay for as long as possible.
The most important piece of equipment used by a knight, however, as well as the most vulnerable, was his horse. Training was important, both on an individual level and as military units. On a personal basis, it started very young and it was thought to be vital that good riding skills were already in place before puberty. Ongoing training was important, for the horse as much as the rider, and the knightly pursuit of the hunt as a form of relaxation was a very practical way of ensuring that riding skills were continually honed by both.
The knight’s horse was a stallion, but smaller than one might imagine, perhaps no more than 15 hands high (i.e. some 5ft high at the shoulder). Horse armour later became more prevalent but during the twelfth century almost all Frankish horses were unarmoured and the vulnerability of their horses dictated much of the battle tactics of the period. For a nomadic light cavalryman, kept at a distance by the threat of Frankish infantry archers or crossbowmen, it was hard to kill or wound a Frankish knight. Only a very lucky shot could bring him down and, while under fire oneself, this was never an attractive option. The knight’s horse was a different matter, however. This was a much bigger target and much less well protected. Without his horse, the Frankish knight was transformed into a slow-moving and easily tired heavy infantryman, almost entirely incapable of offensive action against light cavalry. An altogether easier prospect.
So, while the strength of a Frankish knight was well known to their opponents, whether Byzantine Christians, Turkic warlords or Egyptian viziers, their vulnerability was also widely recognised: an obvious tactic was to try to wound or kill as many of their horses as possible, before they could even form up to launch one of their renowned charges. Anna Comnena, writing of a confrontation between Bohemond and her father’s troops, shows the explicit emphasis the Byzantines placed on aiming at mounts rather than riders. Her father ensured that the Byzantine cavalry:
were issued with a plentiful supply of arrows and told not to hold back in their use; but they were to shoot at the horses rather than the [Franks] . . . Shooting the riders, therefore, would in his opinion be pointless and quite crazy . . . the emperor, who had plenty of experience of [Frankish] armour and our arrows, ordered them not to worry about the men but rather to attack the horses.
If the Byzantines were aware of the need to focus on the crusaders’ horses, the Turkic light cavalry that formed the backbone of most of the Muslim armies operating in Syria in this period were even better motivated to do so. They had large numbers of archers and the capacity to lay down dense showers of arrows. Although their arrows did not have the penetrative power to inflict grievous casualties on the Frankish heavy troops, whether infantry or knights, the horses were extremely vulnerable. A Muslim source, Abu Shama, analysed the situation in a way which was uncannily similar to Anna Comnena’s. Writing about the battle of Hattin, he commented that a ‘Frankish knight, as long as his horse was in good condition, could not be knocked down. Covered with mail from head to foot, which made him look like a block of iron, the most violent blows make no impression on him. But once his horse was killed, the knight was thrown and taken prisoner. Consequently, though we counted them by the thousand, there were no horses amongst the spoils, whereas the knights were unhurt.’ It is significant that although there were huge numbers of horses in the Frankish army (perhaps some 10,000 animals), none had emerged unscathed from the arrow storms unleashed by the Muslim archers.
Archery was a great leveller. In a matter of seconds, a couple of cheap arrows, fired by a poorly armed man on a pony, could turn the Frankish knight, the most expensive piece of military kit on the battlefield, into a slow-moving and disorientated infantryman. The knights, and their unstoppable charge, were at the core of how a crusader army needed to perform. Until the moment was right, however, a delicate balancing act was always required to simultaneously protect the horses and stop the enemy overrunning the flanks and rear.