Montgisard Was Not a Miracle. Saladin's Army Had Stopped Being an Army.

The version you have seen is a thumbnail: five hundred knights against twenty-six thousand, one afternoon, the sixteen-year-old leper king in a silver mask riding out of Ascalon into the sunset. Almost every number in that sentence is doing something other than counting, and the mask is from a film. What actually happened on 25 November 1177 is better than the thumbnail, and it is better for an uncomfortable reason: the real battle teaches you something you can use, and the miracle version teaches you nothing at all.

Here is the claim. Nobody destroyed twenty-six thousand men that afternoon, because twenty-six thousand men were not there. Saladin arrived in the coastal plain with a large army and then, over the days that followed, stopped having one. An army is not a number. An army is a number plus the ability to put that number in one place at one time, and by the time Baldwin's column came over the rise, Saladin had the number and had lost the other half. The famous ratio is not a measure of courage. It is a measure of how completely a force can dissolve while remaining, on paper, enormous.

Two counting systems, multiplied together

Start with where the numbers come from, because that is where the trick lives.

The principal source is William of Tyre, archbishop, chancellor of the kingdom, and — this matters — Baldwin's own childhood tutor, the man who first noticed that the boy could not feel his right hand.¹ He gives the Frankish knights as 375 and puts Saladin's force at 26,000.² Those two figures, sitting in one paragraph, produce the ratio that has been travelling ever since.

But they are not two measurements of the same kind of thing. The 375 counts knights. It does not count the several thousand infantry who marched out with them, and it does not count the turcopoles — the locally recruited light cavalry that every Frankish field force in the Levant depended on and that no chronicler bothered to enumerate.³ It does not count squires, some of whom fought. It may or may not include mounted sergeants; the sources are not consistent enough to tell.⁴ Historians of crusader warfare treat the Frankish field force that day as something in the range of several thousand men in total, not four hundred.

Why the omission? Not deception. The chronicles count knights because, in the society producing the chronicles, only knights counted. The arithmetic in the manuscript is a social hierarchy wearing the costume of a headcount. The men who died holding the line were present at the battle and absent from the sum.

Only the knights were counted, because only the knights counted.

Now the other side of the fraction. The 26,000 is the top of a range, not a consensus. Other contemporary accounts give figures in the order of twelve thousand Turkish and nine thousand Arab troops; across the surviving sources, estimates of Saladin's strength run from roughly eight thousand to twenty-six thousand depending on who is counting and what they are counting.⁵ Modern editors have suspected at least one of the surviving numbers of being a scribal corruption — a figure that reads twenty thousand and was very likely ten thousand before a copyist's hand slipped.⁶ Medieval numerals are fragile in transmission in a way medieval prose is not. A single stroke changes an order of magnitude, and there is no surrounding context to catch the error the way context catches a misspelled word.

So the celebrated ratio is manufactured by taking the most generous available estimate of one army and multiplying it against a count of the other that deliberately leaves most of it out. Do the same operation honestly — total force against total force, using mid-range figures — and you get something closer to one against three or four. Still bad odds. Still the underdog. Not a miracle.

Then popular retelling adds its own inflation on top. The 375 in the chronicle becomes 500 in the video title, because 500 is rounder and sounds more like a last stand. Nobody decided to lie. The number drifted upward through a thousand retellings the way numbers do once their job stops being accuracy and starts being effect.

What Saladin actually had at Montgisard

Here is the part the thumbnail cannot accommodate, because it is not heroic. It is administrative.

Saladin came north out of Egypt expecting no serious opposition in the field. Baldwin was a boy with a disease; the kingdom's best troops were understood to be committed elsewhere; the road looked open. So he did what a commander does when the road looks open. He let the army spread out and eat. Detachments went off to raid and forage across Ramla, Lydda and Arsuf — a wide arc of countryside, with a long way between the pieces.⁷ The baggage train, which is to say the food, the water, the spare mounts and the tents, had bogged down at a crossing behind the main body.⁸ And crucially, nobody was watching Ascalon. Saladin did not post scouts on the fortress that contained the king, because the king was not considered a threat worth the pickets.⁹

Consider what that combination means in practice, before any charge happens. Communication in a twelfth-century army moves at the speed of a rider. A detachment fifteen miles away that has not been told to come back is not a reserve. It is a rumour. There is no way to recall it, form it up, and feed it into a fight that will be decided inside an hour. The men are alive, they are loyal, they are counted in every total anyone will ever write down — and they might as well be in Cairo.

An army is a number plus the ability to put that number in one place. Saladin had the number.

Baldwin, meanwhile, did the one thing available to him. He broke out of Ascalon with everything he had, joined by Odo of St Amand and eighty Templar knights who had come out of Gaza, and moved fast enough to make contact before the situation could correct itself.¹⁰ That is the whole of it. The Frankish force was concentrated and the Ayyubid force was not, and at the point where the two touched, the local ratio — the only ratio that decides anything — may well have favoured the Franks for the few minutes that mattered.

The heavy cavalry charge, delivered against troops who are not formed to receive it, is close to unanswerable in this period. It is not magic and it is not faith. It is several hundred armoured men and horses arriving as one mass against a line that has not had time to become a line. Every Frankish commander in the Levant knew this. The problem was almost never how to win the charge. The problem was always getting the enemy to be standing still and disorganised when you delivered it, which a competent opponent would never permit. Saladin permitted it, once, by underestimating a sixteen-year-old.

The rout that followed was severe, and the retreat was worse than the battle. The chronicle tradition has Saladin escaping on a fast camel and describes his surviving troops harried and picked apart on the desert road back to Egypt — flagged here as the tradition rather than as a secure figure.¹¹ What is not in dispute is that Muslim writers treated the defeat as a genuine humiliation, one that Saladin's reputation did not fully shed until 1187.¹² Montgisard was a real, unambiguous, embarrassing defeat for the most capable commander of the age.

The same lesson, ten years later, running backwards

If the argument is right — that Montgisard was decided by concentration and not by numbers — then it should be testable somewhere else. The test exists. It is called Hattin.

In July 1187 the Kingdom of Jerusalem put into the field the largest army it had ever assembled: something in the range of eighteen to twenty thousand men, including roughly twelve hundred knights, paid for in part with English money and gathered by stripping the garrisons out of the castles and towns to do it.¹³ Saladin brought more, but not overwhelmingly more. By the arithmetic of the thumbnail this should have been the best chance the Franks ever had.

They were annihilated. Not because they were outnumbered in any decisive way, but because they marched across waterless ground in July under continuous harassment from horse archers, and arrived at the fight already unable to use what they had. Twenty thousand men dying of thirst on a hillside are not twenty thousand men. They are a number that has been separated from its ability to act — which is precisely the condition Saladin had been in at Montgisard, arrived at by a different road.

He had learned the lesson. He spent ten years making sure that when the decisive battle came, the enemy's mass would be the thing that killed them: too big to move fast, too big to water, too big to feed off the country it was crossing. Then he took Jerusalem, and the kingdom that had won the miracle at Montgisard lost everything, because it had never been the miracle doing the work.

Why the story keeps growing

Three things inflate this battle, and they are worth separating, because they are three different failures.

The first is the interested witness. William of Tyre is our best source and he is not a neutral one. He taught Baldwin, served his government, and wrote the history that would fix his reputation. That does not make him a liar; by the standards of medieval historiography he is unusually careful, and modern scholarship treats his chronicle as a serious historical document rather than a panegyric.¹⁴ But when a boy's tutor writes down the size of the army his former pupil defeated, the number he selects from a plausible range is not going to be the low one. This is the same problem as the Rosetta Stone and the stele of Hammurabi. The surviving record is not a neutral deposit left by the past. It is a thing somebody made on purpose, and the purpose is usually still legible in it if you look.

The second is the mask. Baldwin IV almost certainly never wore one. The silver mask is an invention of Ridley Scott's Kingdom of Heaven in 2005, and it has since colonised the popular image so completely that people are surprised to learn it is not attested anywhere in the sources.¹⁵ The medical assessment of his condition — a case that appears to have begun in the polyneuritic form and progressed to lepromatous leprosy after puberty — suggests that in 1177 he was not conspicuously disfigured, and that what the disease had taken from him by then was chiefly the use of his right hand and arm.¹⁶ He rode. He commanded. He was obeyed. The genuinely remarkable fact about Baldwin is not that he hid a ruined face behind silver. It is that a visibly ill king, carrying a disease with enormous social stigma attached to it, remained an effective and respected monarch for eleven years — which tells you something real and surprising about that kingdom, and which is a duller, truer, better story than the mask.

The third is the format. A video title is a promise about the ratio between what you know now and what you will know in eleven minutes, and the ratio has to be steep or nobody clicks. "500 versus 26,000" is a load-bearing structural element of the video, not a claim about the past. This is not a complaint about the internet. Chroniclers rounded numbers up for effect eight hundred years before the recommendation algorithm existed. It is a note that the pressure has run in the same direction the whole way down, cumulatively, for eight centuries, and that nobody at any single point in that chain had to be dishonest for the number to end up where it is.

The thing that was actually admirable

None of this argues the courage away, because it was real, and it was of a specific kind that deserves naming.

Baldwin's decision at Ascalon was not the decision to be brave. It was the decision about when. He was shut in a fortress with a force far too small to face Saladin's army in the field, and the correct textbook move — the move nobody would ever have criticised him for — was to stay behind the walls and let the raiders burn the countryside, because that is what a garrison does when it is outnumbered. Sitting still was safe, defensible, and would have cost him nothing but land.

He went out anyway, and he went out at the one narrow moment when going out was not suicide: after the enemy had dispersed to forage, before it could reassemble. Getting that timing right requires knowing something the numbers do not tell you — that the twenty-six thousand were not twenty-six thousand today, and would be again tomorrow. Then it requires betting the kingdom on your own reading of that, at sixteen, with a hand you cannot feel.

That is a judgment about the state of an opponent, executed at speed, with no margin for being wrong. It is the most transferable kind of skill there is, and it is completely invisible in the miracle framing, because miracles do not have timing. If God wins the battle, it does not matter when you leave Ascalon.

Which is the real objection to the thumbnail. Not that it exaggerates — everything exaggerates. That the exaggeration removes the only part of the story anyone could learn anything from. Five hundred against twenty-six thousand is a story about divine favour, or about the innate superiority of one kind of man over another, and both of those readings have been put to ugly uses. Several thousand concentrated men against a much larger force that had come apart is a story about paying attention, moving before the window closes, and understanding that mass you cannot assemble is not mass.

Saladin lost at Montgisard because he counted his army instead of locating it. Ten years later he made the Franks make the same mistake, on a hill with no water, and he took Jerusalem for it. The lesson was on the table both times. It is still on the table, and it has nothing to do with a mask.