County of Edessa

The knight who went east

Drafted from Barber, The Crusader States (2012)formation

In the autumn of 1097 the armies of the First Crusade were grinding south through Anatolia towards Antioch, and at Marash one of the leaders left them. Baldwin of Boulogne, younger brother of Godfrey of Bouillon, took a small force east instead — away from Jerusalem, away from the campaign, across the Euphrates into a country of Armenian towns that no crusading vow had mentioned.

He was invited. The Armenian lords of the region had been living between the Seljuk amirs of Mosul and Aleppo for a generation and were looking for a military patron, and T'oros, the ruler of Edessa, went furthest: in February 1098 he adopted Baldwin as his son and heir in a public ceremony. Within weeks T'oros had been dragged out of the citadel and killed by his own townspeople, and Baldwin was in possession.

Whether Baldwin arranged the killing is disputed, and cannot be settled from the sources.

What the episode is not in dispute about is chronology. Edessa was taken in March 1098. Antioch did not fall until June. Jerusalem was not stormed until July 1099. The first Crusader state was founded before the crusade reached its objective, a hundred and eighty kilometres in the wrong direction, by a man who never went to Jerusalem with the army at all.

He went later, and as king. Godfrey died in 1100 and Baldwin rode south to take the crown as Baldwin I, handing Edessa to his cousin Baldwin of Bourcq — who in 1118 did exactly the same thing and became Baldwin II. The two men who built the kingdom of Jerusalem into a state both learned the work here first.

Nothing passed from the county to the kingdom except those two men, which is a fact about careers rather than about states. It is still the most consequential thing Edessa produced: the two rulers who turned a captured city and a coastal strip into a functioning monarchy both served their apprenticeship on the Euphrates frontier, governing Armenians.

Ruling a country that was already Christian

Drafted from MacEvitt, The Middle Ages Series (2008)expansion

The standing image of the Crusader states is a Frankish garrison sitting on a Muslim population. For Edessa that is close to backwards.

The county's people were overwhelmingly Eastern Christian — Armenian and Syriac Orthodox, with Greek and Muslim minorities — and the Latin element was a few hundred knights and their households in a country of hundreds of thousands. MacEvitt's argument, for which Edessa is the central case, is that what developed there was neither the systematic persecution of the older accounts nor anything resembling tolerance as a principle. It was an unarticulated working accommodation: local lords kept their lands, Syriac and Armenian bishops kept their sees and their disputes, Frankish and Armenian families intermarried at the top, and nobody wrote down a theory of what was going on because writing it down would have required deciding questions everyone was better off leaving open.

The county was administered in Latin, Armenian and Syriac at once, and not as a courtesy. The charters were Latin. The men who collected the revenue, held the smaller castles and knew where the boundaries ran were mostly not.

The army says the rest. Fiefs held of the count, in the ordinary Frankish way, over a Latin fighting strength counted in the hundreds throughout the county's fifty-two years. Antioch, Tripoli and Jerusalem were all thin. Edessa was thin and also landlocked, which is the difference that killed it: it was the one Crusader state that no fleet could reach, four hundred kilometres from the sea, dependent for reinforcement on a road through Antioch that an enemy holding Aleppo could cut.

Nobody has published a figure for how much ground it covered, and the figure would mislead if anybody did. The county was large, and it was empty of the thing that mattered, which was Franks.

A castle for a castle

Drafted from al-Qalanisi, The Damascus Chronicle of the Crusades (1932)peak

The county at its height is not visible in a battle. It is visible in the fact that the atabeg of Damascus received a letter from the lord of Tell Bashir proposing a swap of fortresses and a mutual non-aggression pact, and had to decide what to do about it.

The proposal

Joscelin of Courtenay had fallen out with his own maternal uncle, Baldwin the Little, who held Edessa, and had gone over to the king of Jerusalem, who gave him Tiberias as a fief. The two of them then worked out a plan.

Joscelin would write to Zahir al-Din, atabeg of Damascus, offering friendship and urging the establishment of peaceful relations. He would hand over the castle of Thamanin near Jabal Amila, and receive in exchange the castle of Habis in the Sawad, together with half the Sawad itself. And he would guarantee, on the king of Jerusalem's behalf, that there would be no further attacks on any province of Damascus — provided Damascus attacked no Frankish province in return.

That is a completely ordinary piece of twelfth-century diplomacy: a territorial exchange, a defined frontier, and a bilateral truce, negotiated between a Frankish count and a Turkish atabeg, with a Frankish king's ratification promised in advance.

Nothing in it treats the other side as an anomaly. Ibn al-Qalanisi records the terms the way he records terms between Muslim rulers, because that is what they were.

The refusal

Zahir al-Din rejected the proposals — and the reason was not that the Franks were untrustworthy.

He set out from Damascus with his askar to meet the amir Mawdud and join him in the holy war. They met at Marj Salamiya, agreed to attack Baldwin, and went together. The atabeg took the whole Damascene army with contingents from Homs, Hama and Rafaniya, encamped at Qadas on the Feast of Sacrifice, and marched by Ain al-Jar and the Biqa to Banyas.

So the two things were live at the same moment, in the same council: a negotiated frontier with the Franks, and a coalition to destroy them. Damascus chose the second that year. It would choose the first in other years, and did.

What peace looked like when it held

The chronicle shows the alternative working a decade later, on the northern frontier, in a passage that is easy to read past.

An armistice was concluded between the lord of Aleppo and the Franks, on condition that the castle of Atharib be surrendered to the Franks. It was handed over and they took possession. Then: friendly relations were maintained on this footing, the provinces prospered on both sides, and the roads became safe.

A castle changes hands, the frontier is agreed, trade moves, and a Damascene chronicler writes it up as good news for everybody.

That is the county of Edessa at its peak — not a crusader outpost besieged in a hostile sea, but one of half a dozen small powers in northern Syria that fought each other, allied with each other, exchanged castles, and occasionally kept the roads open. Joscelin's own career runs on those terms: he quarrels with his uncle, takes service with a king, negotiates with Damascus, raids when it suits him, and is eventually taken prisoner at Saruj by Balak ibn Ortuq along with his cousin and a number of his leaders.

Being captured by a neighbour is a normal risk for a regional lord. It is a very different condition from the one the county ended in.

Kharput

Drafted from al-Qalanisi, The Damascus Chronicle of the Crusades (1932)contraction

The recurring problem of the county of Edessa was that its rulers kept being captured.

It was a frontier state with a Latin fighting strength in the hundreds, holding a long border against Turkish lords who raided constantly and who had learned that the most valuable thing on the far side of it was a Frankish commander. Ransoms were enormous. So the counts led from the front, out of necessity, and were taken prisoner at a rate no other crusader state matched.

Two in six months

In the autumn of 1122 the Artuqid amir Balak marched against the Franks of Edessa, fell on them at Saruj and defeated them, taking prisoner their commander Joscelin, together with his cousin Galeran and a number of their leaders.

The following April, Baldwin — count of Edessa before he became king of Jerusalem — moved on Aleppo to deal with Balak, who was besieging a castle. The armies met near Manzara. The king was defeated and captured, along with a number of his principal knights, and Balak put him in a dungeon in the castle of Kharput, together with Joscelin and the rest.

So within six months the same man had taken the count of Edessa and the king of Jerusalem and had them both in the same fortress, four hundred kilometres from anywhere either of them governed.

The escape

What happened next is one of the strangest episodes of the period, and Ibn al-Qalanisi's text at this point is damaged, which is a pity because he had the story.

What survives is this: the Frankish prisoners in Kharput devised a stratagem among themselves, seized the castle, and got out. One of them escaped and could not be recaptured. On the same day the bishop of al-Bara, imprisoned there too, got away from his confinement.

Then Balak marched on Kharput with his troops and besieged his own citadel until he had taken it back from the Franks who had surprised it, and put a garrison in it with instructions to be more careful.

The prisoners who did not get away were still prisoners, in a fortress they had briefly captured.

What it cost

Rulers of Edessa spent years of their reigns in captivity, and the county spent those years being governed by whoever was left and paying for whoever was not. Ransoms went east, hostages went with them, and the frontier castles that required a lord's attention did not get it.

That is the ordinary condition of the county for most of its existence, and the reason it was the crusader state that could least afford a bad year. When one finally came, in 1144, there was no reserve of anything: no army beyond the garrison, no fleet that could reach it, and no ruler in the city.

Four weeks at Christmas

Drafted from Barber, The Crusader States (2012)end

Joscelin II had moved his residence west of the Euphrates to Turbessel and was there in November 1144 when Zengi, atabeg of Mosul and Aleppo, appeared before Edessa with an army. The city had walls and no field force. The count raised what he could and did not arrive. Antioch, with whose prince Joscelin was quarrelling, sent nothing. Jerusalem sent a relief column that got there too late.

Edessa fell on 24 December 1144, after four weeks.

The county was finished when the city fell — a landlocked state named for its capital, with the capital gone — but it took six more years to stop existing. Joscelin held the fragment west of the Euphrates until 1150, when he was captured; his wife sold what was left to the Byzantines, and Nur al-Din took it off them within months. The same dynasty did both ends of it.

What 1144 changed is larger than the county. It was the first demonstration that a Frankish state in the Levant could be destroyed, forty-five years after the first one was founded, and it converted the counter-crusade from a series of raids into a programme with a proof behind it. Zangi and Nur al-Din are usually identified as the point at which the Muslim response to the crusades acquired an ideology. Edessa is where it acquired the evidence.

In Europe the news produced the Second Crusade, which arrived in 1148, decided not to attempt Edessa at all, besieged Damascus for four days, and went home. Damascus was at that moment the one Muslim power in Syria still hostile to the Zangids. The crusade raised to avenge the loss of Edessa succeeded chiefly in pushing Damascus into Nur al-Din's hands, and ended the independent Burid emirate there.

The sap

Drafted from al-Qalanisi, The Damascus Chronicle of the Crusades (1932)afterlife

The Damascene chronicler's account of how Edessa fell is a technical one, and that is what makes it useful: he is interested in the engineering.

The besiegers drove a mine under the wall, shoring the tunnel as they went with timbers and special appliances. When it was finished, all that remained was to set fire to the shoring. Zangi went down into the sap himself to see what had been done, expressed his admiration, and gave the order.

The fire caught the timbers, the timbers went, and the wall came down with them. The Muslims forced their way in over the breach with heavy losses on both sides, and so many of the Franks and Armenians were killed and wounded that the rest abandoned the town. It was taken by the sword in the early forenoon of Saturday 23 December 1144.

What the victor did next

The troops sacked it — killing, capturing and looting — until, Qalanisi says, their hands were filled with money, furnishings, animals and captives enough to gladden their hearts.

Then Zangi gave orders to stop the killing and the plundering, and set about rebuilding what had been destroyed and repairing what was damaged. He detailed a picked force to govern and defend the city, and reassured the inhabitants with promises of good treatment and of justice for all of them, near and far.

That sequence is not a contradiction and should not be softened into one. A storming was expected to be followed by a sack; a ruler who intended to keep a city stopped the sack when he judged it had gone far enough and started paying for roofs. Both halves are the same policy, and the second half is why Edessa was still a functioning city under his successors.

Everything else fell by itself

He marched on to Saruj, which the Franks had already abandoned, and took it. After that, Qalanisi notes, there was not a place or a stronghold he passed and camped in front of that was not immediately handed over.

Only al-Bira held — a castle of the utmost impregnability, he says — and Zangi sat down in front of it and cut off everything and everyone trying to reach it.

That is the measure of what 1144 did. The county of Edessa had held its ground for forty-six years against everything sent at it; and once the capital was gone, the rest of it surrendered on sight, in the order in which the army happened to arrive.

The proof

The Frankish states in Syria had been founded on the assumption that they could not be dislodged, an assumption every ruler in the region had been testing for half a century without success.

The mine under the wall of Edessa ended that. Everything afterwards — the programme of Nur al-Din, the reconquest under Saladin, and, eventually, the Mamluk sultans working down the coast city by city — proceeds from a night in December 1144 when somebody set fire to the props and stood back to watch.