The caliphs' kitchens

Al-Warraq’s title means the copyist or the bookseller, and the book reads like the work of someone who handled a lot of them. He names his sources as he goes. The most famous is Ibrahim ibn al-Mahdi, born in 779, the son of the caliph al-Mahdi and half-brother of Harun al-Rashid. Ibrahim was a celebrated singer, and in 817 a faction in Baghdad proclaimed him caliph against al-Ma’mun. The reign lasted about two years. He went into hiding, was caught, was pardoned, and spent the rest of his life as a court musician and, by repute, the best cook in the family.

His own cookbook is lost. Dozens of its recipes survive inside al-Warraq, which makes a failed caliph one of the earliest named cookery writers in Arabic.

He was not unusual. Al-Warraq cites a run of earlier collections compiled by courtiers, physicians and men of letters, and he gives recipes by name to members of the ruling house. Kushajim, a tenth-century poet at the Hamdanid court in Aleppo who was also known as a cook, contributes two dozen poems about food that al-Warraq sets beside the recipes they describe.

The poetry was not decoration. Al-Mas’udi, writing his history in the 940s, describes an evening at the court of the caliph al-Mustakfi when the guests were told to recite verses about dishes. As each poem finished, the caliph had the dish it described brought to the table. A cookbook in that world was a book of taste in both senses, and it belonged on the same shelf as the anthologies.

The recipes are precise about ingredients and vague about quantities, which is normal for medieval cookery. What they describe is a cuisine of sour and sweet together.

Sikbaj is the dish everyone mentions. It is meat stewed in vinegar with sweet things, a Persian name and probably a Persian origin, and it was eaten hot or cold for days. Zirbaj is a relative, chicken or meat in vinegar sweetened with sugar and thickened with almonds, scented with rose water. Madira was meat cooked in soured milk. Harisa was wheat and meat cooked together overnight and beaten into a smooth paste. Tharid, bread broken into broth, was the old Arabian dish, and a hadith has the Prophet compare his wife Aisha to other women as tharid to other food.

The strangest category is the judhaba. A pudding of bread or rice, sweetened, was set in a pan under a chicken or a joint roasting in the tannur oven, so that it caught the dripping fat and juice. Then the bird and the pudding were served together.

Sugar was everywhere, in savoury dishes as much as in sweets, along with honey, dates and date syrup. Lawzinaj was almond paste rolled in pastry thin enough to see through, soaked in syrup. Spice use was heavy and specific: saffron, cinnamon, galangal, mastic, cumin and coriander, with musk and camphor for scent.

Then there was murri. Barley dough was left to rot in the summer heat for forty days, then pounded with salt and water and left again, until it became a dark salty liquid. Charles Perry, who has done more than anyone to reconstruct these recipes, compares the taste to soy sauce. The process sounds alarming and was clearly considered worth the trouble, since murri appears in cookbook after cookbook for four hundred years and then drops out of the record.

Where did this cuisine come from? The conquerors who came out of Arabia in the 630s ate dates, milk, bread and meat when they could get it. Two centuries later their descendants in Baghdad were eating sikbaj in houses a short ride from the ruins of the Sasanian capital at Ctesiphon.

The vocabulary tells part of the story. A large share of the dish names in al-Warraq are Persian, and so are many of the words for kitchen equipment. The caliphs built their court on the Sasanian model, staffed it in part with Persian secretaries, and took over Persian ideas of what a king’s table should look like. Maxime Rodinson, who opened the modern study of these books in 1949, traced much of it to the Sasanian court, and later work has mostly confirmed him. The Arab part was the bread, the dates and the old dishes like tharid, kept on and honoured.

The other ingredient was medicine. The Greek physicians translated into Arabic in the ninth century, Galen above all, taught that every food was hot or cold, moist or dry, and that diet was the first line of treatment. Al-Warraq’s early chapters set out the properties of foods, and his last ones give advice on digestion, on table manners and on washing the hands after eating. A cook was expected to know why vinegar suited a fatty meat. The same translation movement that put Galen into Arabic is described on the page on the translators.

Washing mattered more than the recipes let on. The best-known food story from the Thousand and One Nights turns on it, and the passage is quoted below the page. A merchant who marries a girl from the household of Harun’s wife Zubayda eats a dish of zirbaj at the wedding and forgets to wash his hands. When she smells it on him she has his thumbs cut off.

Behind the table there was a long argument about the fields. In 1974 the economic historian Andrew Watson proposed an Arab agricultural revolution. In his account a list of crops from India and further east, among them rice, sugar cane, hard wheat, sorghum, cotton, citrus fruits, aubergines and spinach, moved west through the caliphate between 700 and 1100, carried by a single market and a single law. Summer irrigation let farmers grow a second crop in the hot months when Mediterranean fields had stood empty. He set it out in full in 1983, in Agricultural Innovation in the Early Islamic World.

The idea spread fast, and it is still repeated in most general books. Specialists have been harder on it. Michael Decker showed in 2009 that several of Watson’s crops were grown in the late Roman Mediterranean before Islam, rice and hard wheat among them, and that others stayed minor for centuries. The archaeobotany from Spain and Sicily shows new crops arriving, but slowly and unevenly.

What survives of Watson’s case is smaller and still real. Sugar cane, citrus and aubergines did spread west in the caliphal centuries. Al-Andalus did gain new crops and new irrigation. The cookbooks are part of the evidence, because the ingredients they take for granted in Baghdad turn up a few centuries later in the cookbooks of Cordoba and Seville.

Sugar had a darker side in Iraq itself. The salt-crusted marshes of the south were cleared by gangs of enslaved East Africans, and their revolt in 869 is told on the page on the Zanj.

Nobody in a porter’s household cooked from al-Warraq. The book assumes a kitchen with a tannur, a hearth, several cooks and an unlimited budget for saffron. What it shows indirectly is how much of a city’s food came through the market.

The hisba manuals, written for the market inspector, the muhtasib, are the best guide to that. Al-Shayzari’s manual from twelfth-century Syria has separate chapters on the bakers, the harisa makers, the cooks, the roasters of meat, the sellers of fried fish and the sausage makers, each with the frauds they were known for and how the inspector should catch them. A city that needed those rules was a city that bought its food ready-made, and Paulina Lewicka has shown that in medieval Cairo most households did exactly that. Many had no kitchen at all.

Ordinary meals were bread with whatever went with it. Bread is the measure of everything in the legal and medical sources. Lentils, chickpeas, beans, onions, cheese, dates, pickles and salted fish filled out the plate, with meat on feast days. Al-Jahiz’s Book of Misers, written in ninth-century Basra, is a comedy about middle-class hosts who count every loaf their guests take, and it is full of exactly this food.

The Quran forbids khamr, and the jurists spent two centuries arguing over what counted. Grape wine was clearly out. Nabidh, a drink of dates or raisins steeped in water, was another matter, and the Hanafi school allowed some kinds that had fermented a little, provided the drinker stayed sober. The other schools did not.

Practice was looser than either. Several caliphs drank openly, and the court poet Abu Nuwas built a career on wine poems. The monasteries round Baghdad sold wine to Muslim visitors, and a whole genre of books about Christian monasteries is mostly about drinking in their gardens. Taverns were usually run by Christians or Jews, who were permitted to make it. The Maqamat illustration of 1237 shows one, with the jars stacked against the wall and a man drawing off a jug.

The cookbooks keep their distance. Al-Warraq gives recipes for nabidh and for sweet drinks of sugar and fruit, sharab, from the Arabic root that gave English both sherbet and syrup, and he says almost nothing about wine. The table he describes was a dry one in theory and a mixed one in fact.

What outlasted the caliphate was the food. Al-Baghdadi’s cookbook of 1226, written on the eve of the Mongol sack, still cooks sikbaj, harisa and judhaba. The Spanish word escabeche, for fish or meat preserved in vinegar, is usually traced back to sikbaj, and harees is still made across the Gulf and in Hyderabad during Ramadan, pounded overnight as the tenth-century cookbook describes.