Every medieval historian, whatever his race, invariably indulges in wild and picturesque exaggeration whenever he has to estimate numbers that cannot easily be counted. It is therefore impossible for us to-day to establish the actual size of the Crusading armies. When Fulcher of Chartres and Albert of Aix tell us that the fighting men of the First Crusade numbered 600,000, while Ekkehard gives 300,000 and Raymond of Aguilers a modest 100,000, or when Anna Comnena declares that Godfrey of Lorraine brought with him 10,000 knights and 70,000 infantrymen, it is clear that the figures are only meant to denote a very large number indeed. But when they are dealing with smaller numbers the chroniclers need not be entirely distrusted, though they like to give a round figure that can only be approximate. From their evidence we can make certain deductions.
The proportion of non-combatants in the armies cannot be estimated. It was certainly high. A large number of knights brought their ladies with them. Raymond of Toulouse was accompanied by his wife, and Baldwin of Boulogne by his wife and children. Bohemond had at least one sister with him. We know the names of several ladies that took part in Robert of Normandy’s expedition; and occasionally other ladies appear in the story. All these ladies brought attendants; and there was certainly a large number of humbler women, respectable and the reverse, with the army. We continually hear of male non-combatants, such as Peter Bartholomew and his employer. The clergy with the army was numerous. But it is probable that most of the male non-combatants would be pressed into service in times of danger. The proportion of permanent non-combatants, women, old men and children, cannot have been more than a quarter of the whole force.
It is probable also that the rate of mortality was particularly high amongst these non-combatants, especially the old men and children. Among the combatants the infantry must have died off from disease and hardships in greater proportion than the knights and ladies, who were better tended and better able to buy food. In battle the cavalry played a more exposed role than the infantry and therefore suffered as heavily.
The proportion of cavalry to infantry seems to have been about one to seven when every possible combatant was enrolled into the latter. Anna’s estimate of the relative strength in Godfrey’s forces, though her figures should be divided at least by ten, is probably correct. At the battle of Ascalon, when every available man in Palestine was employed, there were 1200 cavalry and 9000 infantry, a proportion of one to seven and a half. At the siege of Jerusalem there were, according to Raymond of Aguilers, 1200 to 1300 knights out of an army of 12,000; which, however, included Genoese and English engineers and marines. The term ‘knights’ must be used to mean armed horsemen, and not in any chivalrous sense; while many of the infantrymen were not fully armed. The archers and pikemen were probably only a fairly small proportion of the whole.
Of the individual armies it is almost certain that Raymond’s was the largest; but we have only one indication of its size. When he heard at Coxon the false rumour that the Turks had evacuated Antioch he sent a cavalry force of 500, including some of his leading knights, to occupy the city. The number 500 occurs with suspicious frequency; but it may well have been considered the proper unit for a large raid or expedition of this type. It is unlikely that Raymond would have spared half his cavalry strength at this stage. If we accept this figure of 500 as approximately correct, his whole cavalry strength must have been 1200 or more, and his total force about 10,000, apart from old men, women and children.
The Chronicle of Lucca tells us that Bohemond went to the East with 500 knights. Anna Comnena notes that he did not have a particularly large army; so this figure may well be correct. He allowed Tancred 100 knights and 200 infantrymen for his Cilician expedition, though he sent another 300 infantrymen after him. These numbers fit together reasonably.
The only indication that we have of the proportionate size of the other armies is given by Raymond’s action at Rugia, when he attempted to bribe his rivals to accept his leadership. He offered Godfrey and Robert of Normandy each 10,000 sous, Robert of Flanders 6000, and Tancred 5000, and lesser sums to the lesser chiefs. The sums must have been fixed in relation to the strength that each prince could now supply, though Tancred was probably offered a disproportionately high sum in order to detach him and as many Normans as possible from Bohemond.
Our only evidence for the size of Godfrey’s army, apart from Anna’s fantastic figure, is provided by his willingness to spare 500 cavalry and 2000 infantry to his brother Baldwin for his Cilician expedition. It is most unlikely that he would have parted with more than half his cavalry strength, even though he intended this force to rejoin him before reaching Antioch. It is tempting to assume that Raymond’s offer at Rugia was made on the basis of ten sous for each head of cavalry. If at the same time we divide Anna’s figures by ten, we may credit Godfrey with some 1000 cavalry and 7000 infantry at the time of his arrival at Constantinople. He must have suffered considerable losses before the date of the conference at Rugia, quite apart from the knights that accompanied Baldwin to Edessa; but he had been joined by survivors from Peter the Hermit’s Crusade and the abortive German Crusades, as well as by some of Guynemer’s marines; who, as their master was a Boulonnais, would naturally associate themselves with the Count of Boulogne and his brothers.
Robert of Normandy ranked equal to Godfrey at Rugia. If Godfrey commanded 1000 head of cavalry, he must have been equally strong. A century later Normandy was obliged to provide its duke with slightly under 600 knights. For the Crusade Robert might well have been able to raise a rather larger number of horsemen, perhaps 650. He was joined by soldiers from Brittany and across the Channel; which may have given him another 100 or 150 horsemen. Moreover, after the return to Europe of Stephen of Blois and Hugh of Vermandois, he had assumed command of such of their forces as remained behind. Stephen, whose territories were not large but were rich, may have provided 250 or 300 horsemen. Hugh probably did not bring with him many more than 100. In all Robert may well have had close on 1000 under his command at the time of Rugia.
On the same basis Robert of Flanders must be credited with 600 cavalrymen, some of whom came from the territory of his neighbour, the Count of Hainault. Robert legally owed his liege, the king of France, only twenty fully armed knights; but in 1103 he offered in a treaty to provide Henry I of England with 1000 cavalrymen. He could therefore easily raise 600 for the Crusade.
Bohemond’s force of 500 cavalrymen, mentioned by the Chronicle of Lucca, fits in with these figures. If we assume that the armies of the lesser lords are to be counted in with the greater armies, and that the sums offered to them by Raymond at Rugia were purely personal, we reach a total for the whole expedition of roughly 4200 to 4500 cavalry and 30,000 infantry, including civilians that could be pressed into service. The letter written by Daimbert to the Pope numbers the army of Crusaders at 5000 cavalry and 15,000 infantry. By the latter armed combatants alone were probably included. The former figure is a permissible exaggeration from 4000.
This seems a small enough army. Yet when we come to the figures given by the chroniclers for individual battles, the numbers are smaller still. At the battle of the Lake of Antioch, when, we are told, all the available knights were used, there were only 700 of them. But many of the knights were sick at the time; and it appears from a letter of Anselm of Ribemont that the real shortage was of horses. He estimates that only about 700 were available for use at the time of the siege of Antioch, so many had perished from hunger and from cold. He declares that there was no shortage of men. Moreover, on this occasion, it is probable that Raymond’s cavalry remained with him to guard the camp. The raiding expedition led by Bohemond and Robert of Flanders the following month was said to have been composed of 2000 cavalry and 15,000 infantry; and this definitely excluded Raymond’s army. But again, only 1200 or 1300 cavalry were present at the siege of Jerusalem, and a little over 10,000 infantry; and the strength of the army at Ascalon was very similar. Though many soldiers had died or been killed and many had returned home, it is impossible that the strength of the army should have declined by two-thirds between the time of the conference at Rugia and the siege of Jerusalem.
We can only therefore repeat that any estimate must be taken with reserve. I believe that the whole army at the time that it left Constantinople roughly reached the total that I have suggested above. In the course of the next two years it was very much reduced; and at Rugia Raymond was using an out-of-date and highly optimistic calculation on which to base his offers. The comparatively small figures given in the chronicles of Baldwin’s exploits can, I think, be accepted as roughly accurate.
The size of Peter the Hermit’s original expedition is equally impossible to calculate. The figure of 40,000 given by Albert of Aix is clearly excessive; but his followers may have numbered as many as 20,000. Of these non-combatants formed the vast majority.
For purposes of comparison it may be noted that the whole Byzantine army in the ninth century has been calculated to have numbered 120,000. The loss of the Anatolian provinces must have resulted in a reduction of available forces by the end of the eleventh century; but Alexius could probably dispose of about 70,000 men, most of which were needed to garrison his far-flung frontiers; while a large proportion was probably disbanded every winter for purposes of economy. It is improbable that the largest army led into battle by the Byzantines at this period numbered more than 20,000 men, well equipped and well trained. It is impossible to estimate the size of the Moslem armies. Kerbogha’s army probably numbered about 30,000; but no actual evidence exists. It was able to undertake a more effective blockade of Antioch than the Crusader army could manage. The Egyptian army at Ascalon was certainly larger than the Crusaders’; but its actual size can only be guessed. It is doubtful if the Turkish army at Dorylaeum was as large as the Crusaders. The Turks relied on their sudden attack and their mobility to compensate for any disadvantage in numbers.