Empire of Nicaea

A government in exile that worked

Drafted from Treadgold, A History of the Byzantine State and Society (1997)formation

In April 1204 a crusading army recruited to attack Egypt took Constantinople instead, sacked it for three days, and installed a Latin emperor in it. The Roman empire ceased to have a capital for fifty-seven years.

What it had instead were three governments claiming to be it. Epirus in western Greece, Trebizond on the Black Sea, and Nicaea, in the well-watered country of western Anatolia a hundred miles from the lost city.

Nicaea won because it was the only one that behaved like a state rather than a claim.

Theodore Laskaris spent his reign fighting the Latins, the Seljuks of Rum and his Greek rivals into a stable frontier. John Vatatzes, who followed him and reigned for thirty-two years, did the part that mattered. He assumed that the empire would be restored, and governed as though the revenue had to last: he promoted grain and livestock production on the imperial estates, restricted imports of Italian luxuries to keep silver in the country, kept taxes payable, and built up a treasury and an army out of a territory much smaller than the one he intended to recover. The chroniclers report him refusing to buy foreign silk and having his crown paid for from the sale of eggs off the palace farms, which is the kind of story a court tells about a ruler it is proud of and is consistent with everything else recorded about the administration.

The result was that by the 1250s the government in exile was substantially richer and better organised than the Latin regime sitting in its capital, and had taken Thrace and most of Macedonia besides. The empire was being rebuilt from Anatolia, which is where its manpower and grain had always been.

A hundred and forty men-at-arms

Drafted from Villehardouin, The Chronicle of Geoffry de Villehardouin, Marshal of Champagne and Romania, Concerning the Conquest of Constantinople (1829)formation

The Nicaean state nearly did not get a first winter. The only contemporary narrative of its earliest months was written by the marshal of the army trying to strangle it, and it is worth reading because of what it takes for granted.

In the autumn of 1204 a Latin force crossed to Asia under two knights, fortified a castle near the ruins of Cyzicus, garrisoned it, and pushed inland to take more. Theodore Laskaris collected everything he had and met them near Poimaninon on 6 December.

Villehardouin's account of the battle is four sentences and an excuse. Our people suffered greatly, he writes, and no wonder, considering the great number of the Greeks and that we had only seven score men-at-arms, not counting sergeants. But since Our Lord gives victory according to His grace and pleasure, the Greeks were beaten and scattered in spite of their numbers, and so badly that within the week much of the country came back in. Poimaninon, a strong castle; Lopadion, one of the finest cities of the country; and Polychna, a fort by a freshwater lake, were all given up, and the neighbouring provinces submitted to the pilgrims.

A hundred and forty knights took the Asian shore off the man who was building the Byzantine successor state, in his first year, and had the country surrendering inside a week.

That is where Nicaea started: not as the natural heir with the provinces behind it, but as one of three claimants with a defeated army, whose most defensible asset was that the Latins had too much else to do. Within four months of Poimaninon the emperor of Constantinople was a prisoner in Bulgaria and every knight who could be spared had been recalled to Europe, and the Asian campaign stopped. Laskaris used the interval to take the cities back, absorb the other two claimants' followings, and get himself crowned by a patriarch he had appointed.

The Latin account is also useful for what it never says. Villehardouin lists castles and men-at-arms and God's pleasure, and shows no awareness that the man he had just beaten was assembling a government — a patriarchate, a treasury, a land settlement for refugee soldiers — out of the wreck. To the marshal of Champagne this was a province being pacified at a satisfactory rate. It was the last quarter in which anyone could have thought so.

Three claimants

Drafted from Treadgold, A History of the Byzantine State and Society (1997)expansion

Three Greek states claimed to be the Roman empire after 1204.

Trebizond, on the Black Sea coast, was founded weeks before the sack by two grandsons of a previous emperor with Georgian help, which means it is not strictly a successor state at all — it was already in existence when Constantinople fell. It outlasted all of them and never seriously contended, because it was eight hundred kilometres from anywhere that mattered and its route to the west ran through everybody else's territory.

Epirus, in western Greece, was the most dangerous. Theodore Komnenos Doukas took Thessalonica from the Latins in 1224, had himself crowned emperor there, and was closer to Constantinople than Nicaea was. He was then destroyed by the Bulgarians at Klokotnitsa in 1230 and blinded, and Epirus never recovered the position.

Nicaea won, for reasons that had little to do with luck.

Anatolia

The decisive advantage was the oldest one in Roman history: western Anatolia was where the empire's manpower and grain had always been.

Theodore Laskaris established himself in the well-watered valleys behind Smyrna and Nicaea — country with cities, a taxable peasantry and a functioning agriculture, which Epirus's mountains and Trebizond's coastal strip did not have. He then did the necessary unglamorous thing: he fought the Latins, the Seljuks of Rum and his Greek rivals to a standstill and stopped, rather than marching on Constantinople with an army he could not replace.

The Seljuk frontier settled after he killed the sultan Kaykhusraw I at Antioch on the Maeander in 1211, and stayed quiet for a generation. That quiet is the whole basis of what follows: a state that does not have to defend its eastern frontier can spend everything on its western one.

The Latin empire

It is worth saying how weak the thing at the centre was.

The Latin emperors held Constantinople, a strip of Thrace and not much else. The crusade's own settlement had given three-eighths of the empire to Venice, which took the islands and the ports and had no interest in the land. The Latin regime had no revenue base, no fleet of its own, and a hinterland held by Greek and Bulgarian enemies; its emperors spent their reigns travelling in western Europe asking for money, and one of them pawned the Crown of Thorns to Venetian merchants, from whom Louis IX of France redeemed it and built the Sainte-Chapelle to house it.

By the 1240s Nicaea had taken Thrace and most of Macedonia and the Latin empire was Constantinople and its suburbs. The recovery of the city, when it came, was not a campaign. The third and fourth chapters are about what Nicaea did with the twenty years in between.

A state on a small base

Drafted from Treadgold, A History of the Byzantine State and Society (1997)peak

Nicaea lasted fifty-seven years and is better documented than most states that lasted five hundred — partly because its own officials wrote about it, and partly because Byzantine writers afterwards used it as a stick to beat the restored empire with.

Pronoia

The pronoia was a grant of the revenues of an estate to a soldier in return for service: revocable in principle, not heritable in principle, and both of those increasingly less true in practice.

The eastern Roman empire had arrived at the same arrangement six hundred years earlier with the themes, for the same reason — a state that must field cavalry and cannot pay them in coin. The striking thing is that the pronoia is not a continuation of that. The themes had dissolved well before 1204. This is an independent arrival at the same solution by a state that found itself facing the same problem again.

It keeps happening. Land granted for military service is the single most widely repeated institution in this whole period: the iqta, the timar, the pronoia, the theme, the Mamluk system, the Frankish benefice, the ilku of Middle Assyria, and the Old Babylonian ilkum that Hammurabi's code legislates in sixteen consecutive paragraphs. Some are genealogically related — the timar descends from the iqta, and the Mamluk system from both. A good many are not related to anything at all. The Ur III version predates the next one by six hundred years; the Byzantine themes and the Frankish benefice appeared independently in the same century at opposite ends of a broken Mediterranean.

A state that must put cavalry in the field and cannot pay them in coin has essentially one solution, and it has been found over again in Sumerian, Akkadian, Greek, Latin, Arabic, Persian and Turkish.

Beside the pronoia soldiers were mercenaries: Latin knights, Cuman horse archers settled on land in Anatolia and Thrace, and Turkish troops.

Vatatzes

John III Vatatzes reigned thirty-two years, and the accounts of his government read like a treatise on how to run a state that intends to be bigger later.

He put the imperial estates to grain and livestock production and sold the surplus. He restricted the import of Italian and eastern luxuries so that silver stayed in the country, and the chroniclers report him telling his son that a crown paid for out of the sale of eggs from the palace farms was worth more than one bought with foreign gold. He kept taxes at a level people actually paid. He founded hospitals and endowed monasteries and libraries. He built a treasury and an army out of a territory a fraction of the size of the one he intended to recover.

Some of that is a court's admiring story about its own ruler. The measurable part is not: by the 1250s a government in exile in western Anatolia was richer, better armed and better organised than the regime sitting in its capital, and had taken most of the European territory that regime claimed.

What it did not do

It did not go for Constantinople.

Vatatzes took Thrace and Macedonia and left the city alone. It was defended by walls his army could not have taken and by a Venetian fleet he could not have fought, so the strategy was to isolate it and wait. The strategy was correct: the city fell in 1261 to a scouting force that found the garrison away, without a siege at all.

The church in exile

Drafted from Treadgold, A History of the Byzantine State and Society (1997)peak

A Latin patriarch sat in Hagia Sophia from 1204. The Orthodox patriarch of Constantinople sat at Nicaea, elected by the bishops who had fled there and recognised by everyone in the Orthodox world who was not under Latin rule.

That is the second reason Nicaea won, and it is not the kind of asset that shows up in an account of territory or revenue.

What a patriarch is worth

The Bulgarian, Serbian and Rus churches — all of them created by Byzantine missions in earlier centuries — were canonically under Constantinople. With Constantinople in Latin hands, the question was who they were under now, and the answer everybody gave was Nicaea. Epirus, which had a rival emperor, could not produce a rival patriarch that anybody recognised; when Theodore of Epirus had himself crowned in 1224 he had to use the archbishop of Ohrid, and the canonical objection to it was immediate and general.

So Nicaea held the one asset that could not be taken by force and could not be manufactured. The Latin empire had the city, the walls and the relics. Nicaea had the church.

The price

Which is also why the union negotiations that ran throughout the exile were so difficult, and why they mattered.

Every Nicaean emperor negotiated with the papacy about reuniting the churches, because papal recognition would have removed the Latin empire's justification and possibly the crusading support that kept it alive. Every one of them failed, because what Rome wanted — acknowledgement of papal primacy, and the western position on the procession of the Holy Spirit — was the thing that would have cost them the Orthodox world that was the point of the exercise.

Vatatzes came closest, in 1250, and died before anything was signed.

The same trade was made and lost twice more afterwards, both under the restored empire at Constantinople: Michael VIII accepted union at Lyon in 1274 to stop Charles of Anjou, and was buried without a funeral by his own church; the emperor and church accepted it again at Florence in 1439 in exchange for help against the Ottomans, and the population of Constantinople rejected it.

Why it decided the contest

Because Nicaea's whole claim was that it was the Roman empire, temporarily housed elsewhere, and that the crown was the same crown — and an assertion of that kind has to be recognised by somebody to be worth anything.

The recogniser was the Orthodox church, which Nicaea held and its rivals did not. That is a strange kind of asset — not a religion propagated outward, which is the thing usually worth recording about a state and a church, but an institution held onto, whose value was entirely that other people needed its certification.

An exiled government in a provincial town won a fifty-seven-year contest for the Roman empire by keeping possession of the office that decided who the Roman emperor was. Nothing about its territory, its army or its revenue explains that.

What the restoration cost

Drafted from Treadgold, A History of the Byzantine State and Society (1997)end

In July 1261 a Nicaean commander with a small force found that the Latin garrison of Constantinople had gone off on a raid, and walked into the city. The Latin emperor took ship. Michael VIII Palaiologos — who had made himself co-emperor by pushing aside the child heir of the Laskarids, and later had him blinded — entered in August and was crowned in Hagia Sophia.

The restoration is where the Byzantine page picks the story up, and it is usually told as an ending: the empire returns, and the last two centuries are decline. The more useful reading is that the restoration cost the empire the thing that had made the restoration possible.

Constantinople was a ruin with a fraction of its former population, and it had to be repopulated, rebuilt, garrisoned and defended, and its walls and fleet paid for. The threat now came from the west — from the Latin powers who wanted the city back, and from Charles of Anjou in southern Italy who spent two decades preparing to take it — so the army and the money went west. Anatolia, which had funded the whole enterprise, was left with a frontier nobody was watching.

A frontier nobody was watching

Drafted from Treadgold, A History of the Byzantine State and Society (1997)afterlife

Within forty years that frontier had dissolved. The Turkmen bands pressing on it moved in, the Byzantine hold on western Anatolia broke, and by about 1300 the Turkish beyliks were established across it — including, on the north-western corner facing the Marmara, the one the Ottoman dynasty began with.

That is the argument for giving this fifty-seven-year state a page of its own. The empire of Nicaea is the last period in which the Roman state governed Anatolia effectively, and the decision that ended it — to spend everything on recovering the capital — is the decision that gave Anatolia to the Turks. The The Byzantine empire ends in 1453 in Constantinople. The decision that made that ending inevitable was taken here, two hundred years earlier, and was technically a success.

What Nicaea had actually built

It is worth being specific about what was given up, because the usual account of the thirteenth century treats Nicaea as a waiting room.

For fifty-seven years the Laskarid emperors governed western Anatolia as a going concern. They kept the frontier garrisoned with a settled soldiery holding land in return for service. They ran a treasury with a surplus — John Vatatzes is remembered for a state farm whose egg money paid for a crown — and they enforced a policy of buying nothing from foreign merchants that could be produced at home. They kept the patriarchate, the schools and the law courts in exile and functioning.

That was the last time a Roman government administered Anatolia in a way that worked. Everything after 1261 is a coastal empire on the wrong side of the water, holding Anatolian territory it could not garrison from a capital it could not afford.

The exchange

So the balance is unusually clean, and unusually harsh.

Nicaea traded a functioning provincial state for a ruined capital and a name. It got the name — Michael VIII entered Hagia Sophia and the Roman empire was formally restored — and within a lifetime the provinces that had paid for the whole enterprise were gone, and the empire was a city and its suburbs again, which is what the Latin empire had been.

The restoration succeeded completely and cost more than the thing it recovered was worth. That is not hindsight applied afterwards: contemporaries in Anatolia said so at the time, loudly enough that Michael's dynasty spent decades dealing with Laskarid loyalism in the provinces it had stopped defending.