An Ottoman counted the houses of Iași and the people of Brașov. How much did he exaggerate?

When

4 December 1659, the day named in the letter

What he counted

20,000 houses, 11 monasteries, 2,060 shops, 1,060 cellars

Checked against

Marco Bandini, 1646: 15,000 houses and 11 monasteries

The letter that brought him to Moldavia carried a date: the twentieth day of Rebiülevvel 1070, that is 4 December 1659. It came from Ștefăniță, son of Vasile Lupu, whom the Porte had just made prince. Iași was held by Constantin Șerban, put on the throne with the help of George II Rákóczi; the Ottoman text calls him Burunsuz, “the noseless”, and the Romanian chronicles call him Cârnul.

The battle was fought in snow on the plain outside Iași, the enemy's wagon-fort drawn up across the town's lake at Galata. Constantin Șerban fled, and Evliya stayed on for forty days – “in forty days we came to know their ways”, he writes.

His count: some twenty thousand houses, not one roofed with tile or lead, masonry rare, seven hundred streets floored with planks, 2,060 shops, 1,060 cellars with taverns above them, six merchants' inns, and not a single fountain – water came in barrels on carts.

Some thirteen years earlier the Catholic bishop Marco Bandini had written about the same town, in Latin: fifteen thousand houses, sixty churches in all, eleven monasteries in the town and near it. On houses, Evliya adds a third. On monasteries he writes the same number: eleven.

What he describes around the princely court checks out as well. The lake below the palace, he says, was made by Vasile Lupu in the year 1045 of the Hijra, 1635–36, by damming a stream between two hills, and its lease brought in forty thousand gold pieces. The bath built on its shore was, he writes, brighter than any bath in all Frankish lands. Archdeacon Paul of Aleppo, who stayed in Iași in 1653, saw both the lake behind the palace and the bath “built by Vasili the Voivode, on the plan of the Turkish baths, with cupolas and an abundance of marble”. Lupu himself Evliya places exactly: a prisoner in Yedikule, in Istanbul, where he died in March 1661.

When

January–February 1660

What he counted

500 captives ransomed at Bucharest; 17,000 dead, “so they relate”

Checked against

Paul of Aleppo, 1657: 6,000 houses and 40 churches in Bucharest

From Moldavia the army crossed into Wallachia. He gives the reason himself: Mihnea III, the prince the Porte had installed two years before, had turned against it and sent word to his men to kill the Muslim merchants and soldiers in the towns.

The figure he gives for those days is the largest in the chapter – seventeen thousand killed in three days, in Bucharest, Târgoviște and the other towns – and it is the one he twice holds at arm's length: “so they relate”, and, in Arabic, “the responsibility rests with the teller”. No other source gives a number at all.

The march through the country he writes as an inventory of burnt villages and captives taken: three hundred here, a thousand there, two thousand along the Argeș. On the way back, he says, the people of Târgoviște brought gifts and ransomed their captives by the hundred, and in Bucharest the inhabitants “beginning to come down from the mountains” paid for five hundred.

One of his sentences about Târgoviște can be checked from the other side. Evliya writes that the town had only just begun to be rebuilt. Archdeacon Paul of Aleppo, who had passed through two years earlier, recorded why: in Lent 1658 the Pasha had ordered Târgoviște to be set on fire, “walls and all”.

Bucharest he does not count. The only count of the town from those years is Paul of Aleppo's: it is “said to have contained for many years about six thousand houses”, with “forty churches and convents”.

When

June 1660, on the way to Oradea

What he counted

1,200 houses in the fortress, 1,500 in the town, four baths, four mosques, 400 shops

Checked against

Nobody else counted the same thing in those years.

In June 1660 he set out north from Belgrade to catch the siege of Oradea. At Timișoara, an Ottoman town for a hundred years by then, he writes the most detailed description of the whole journey.

The fortress, he says, is built of oak rather than stone: two rows of palisade packed with lime, cement and gypsum, fifty feet thick, because stone would have to be carted from far away and brick would sink into the marsh of the Timiș. Inside he counts twelve hundred houses, all shingled, four mosques, seven schools, three inns, four baths and four hundred shops. The streets are laid with planks, there is no well at all, and the water comes from the river through filters.

The difference from the Christian towns is in the list itself. In Iași he had counted cellars and monasteries; here he counts baths and mosques. The same categories do not appear on both sides of the border, so the two lists do not compare directly.

For Timișoara in 1660 there is no other count to set beside his. His figures stand alone: nothing contradicts them and nothing supports them.

When

27 August 1660, the day of the surrender

What he counted

73,000 captives, as the Tatars told it

Checked against

The garrison numbered 850 men and marched out with 300 wounded

He reached Oradea in midsummer and stayed to the end. The fortress was surrounded on 14 July and surrendered on 27 August, after forty-four days.

His account is that of a man standing in camp rather than over a map: the rumours going round the besiegers, the quarrel between the commanders, the woman who showed where the moat's sluice was, the Hungarian who slipped out of the fortress with news of a mine. Then the surrender: of the seven commanders in the fortress, some come out unarmed first, as hostages, to ask for a week's truce, and at the end the whole garrison marches out.

The figure he gives for captives he does not claim as his own: he heard it from the Tatars, “our warriors hold seventy-three thousand captives”. On the other side, the garrison that held the fortress for forty-four days numbered 850 men at the start and 300 wounded at the end.

What he watched himself agrees with what is known elsewhere. In front of the fortress stood the gilded bronze statues of the kings of Hungary, among them the mounted Saint Ladislaus cast in 1390 by the brothers Márton and György of Cluj. Evliya describes an agha riding at them with a sword and the statues being cut to pieces and loaded onto carts with the bells. The statues were indeed destroyed then, in August 1660. What he adds are the eyes of night-shining stone and the nails of twenty-carat diamond.

When

August 1661

What he counted

A wall 4,000 paces round, 40 towers, 4 gates, 11 monasteries

Checked against

Nobody else counted the same thing in those years.

In the summer of 1661 Ali Pasha's army entered Transylvania to remove János Kemény and install Mihály Apafi. Evliya rode with it, and the campaign let him see the towns inside the Carpathian arc one after another: Hațeg, Orăștie, Alba Iulia, Turda.

At Cluj the fortress at first refused to submit: inside, he writes, were three thousand of the emperor's soldiers. Once the suburb had been attacked and burned, the flags of submission went up on the bastions and the captain came out with gifts. Evliya writes one sentence that is pure curiosity: “I too, poor man, went in and looked about.”

He paced the wall – four thousand paces round – and counted forty towers, four gates and eleven monasteries with tin-covered spires, each with a gilded cross on top. Inside, the villagers of the whole district had crowded in for fear of the army, “so that hardly a foot's width of space was left”.

For the buildings of Cluj in 1661 there is no independent count. What can be said is the order of magnitude of the town: the first modern census, in 1850, found 17,465 people there.

When

October–November 1661

What he counted

100,000 people in the fortress, 70 churches, 2,000 shops, 17 inns

Checked against

The 1850 census: 21,782 inhabitants

He was sent to Brașov with four thousand soldiers to collect the tribute. The Saxons met them outside the town on horseback, silver staves in hand, and said exactly how long the hospitality would last: “for ten days you are our guests, and on the eleventh we let you go”. In the meantime, Evliya walked the town.

His count here is the largest of the whole journey. Ten thousand rooms inside the walls, forty thousand in the town outside them “so the judges said”, two thousand shops, seventeen inns, seventy churches of which twenty-two belong to monasteries, each holding two hundred priests. And at the end, the judges' reckoning: a hundred thousand people inside.

He explains why there were so many: for fear of the Ottoman army, the country around had shut itself inside the walls. The explanation is plausible; the figure is not. The first modern census, in 1850, counted 21,782 people in the whole of Brașov with its suburbs – after nearly two centuries of growth.

The monasteries are a different kind of problem. Brașov had gone over to Lutheranism in the 1540s under Johannes Honterus, and the monastic houses were dissolved then. Twenty-two monasteries had nowhere to stand in the town of 1661, and the four thousand four hundred priests he puts in them would be a fifth of everyone the census found in the whole town two centuries later.

When

November 1661

What he counted

74 towers, 70 churches, 70 inns, more than 100,000 souls

Checked against

The 1850 census: 12,765 inhabitants

From Brașov the road ran to Sibiu. Evliya walked right round the fortress in an hour and a half, and the next day went inside with Ali Pasha's steward.

He counted seventy-four towers, eleven brick bastions, five gates, seventy churches, seventy inns and forty schools. For people he gives two figures in the same description: first six thousand soldiers and sixty thousand men, then, a few lines later, “more than a hundred thousand souls”. The 1850 census found 12,765.

The reason he lingers on the walls is military: he describes the marshes around them, says which side the fortress is weak on, and adds a sentence that leaves nothing to interpretation – should the Ottomans ever besiege Sibiu, a road can be made through the marsh.

He also supplies the proof that the walls held. Two years earlier George II Rákóczi had shut Ákos Barcsay inside them and kept the town under siege, Evliya writes, for a full seven months. The Hungarian archives show the same span: the siege began at the end of 1659 and Rákóczi was still attacking the town in May 1660. Here his figure falls inside what the documents give.