Language is a funny thing. Most of us know what good writing looks like, but few of us can actually write good.

That’s partly because writing is so subjective; what’s funny, moving, interesting to one person is awkward, vapid, dull to another (apart from the work of Terry Pratchett which is universally fantastic.)

Sumerian is the oldest written language and was spoken in regions of ancient Mesopotamia. Clay tablets dating as far back as 3200 BCE have been found bearing Sumerian writing and, like the writing of the late great Sir Pratchett, much of the content found on the various tablets is pretty inspiring:

“A heart never created hatred; speech created hatred.” “Fate is a raging storm blowing over the land.” “A good word is a friend to numerous men.”

When the writing wasn’t being insightful it was being practical.

“Putting unwashed hands to one’s mouth is disgusting.” “Before the fire has gone out, write your exercise tablet!” (guess kids have always resisted doing their homework). “The owner of a house should reinforce the windows against burglars.”

But just to balance it out so that no-one became too self-confident or capable, the Mesopotamians also had language to remind you of your place in society.

“Your worthiness is the result of chance.” “Your role in life is unknown.” “The battle-club would not find your name – it would just find your flesh.”

You’d be forgiven for thinking that a civilisation so adept at putting ink to paper – or more accurately stylus to clay – would go hard when writing recipes.

Not quite.

Kanasu Broth

Leg of mutton is used. Prepare water add fat. Samidu; coriander; cumin; amd kanasu. Assemble all the ingredients in the cooking vessel and sprinkle with crushed garlic. Then blend into the pot suhutinnu and mint.

Recipe 23, tablet A. The Babylonian Collection at Yale University.

So why so brief? Let’s break it down.

Firstly, Ancient Mesopotamians were writing on clay tablets, not papyrus or paper. Because the letters were pushed and embedded into the clay, rather than scratched on, each tablet had to be fresh and pliable before writing. This meant that someone had to make the tablet and carve out the stylus before any writing could occur.

Secondly, the script used was cuneiform. This script (which was developed by the Sumerians), underwent a huge number of transitions during a 2000 year period (3000 BCE – 1000 BCE), developing from pictograms to glyphs to horizontal and vertical wedge shaped lines. Recipe 23 dates to about 1700 BCE, placing it in the middle of the Old Babylonian era, and uses cuneiform script but is written in the ancient language of Akkadian.

The Akkadians were another Mesopotamian civilisation who, according to linguist Guy Deutscher, developed a culturally symbiotic relationship with the Sumerians which included widespread bilingualism. By 1700 BCE, Akkadian had just about taken over Sumerian as the main spoken language of Mesopotamia, but Sumerian cuneiform script continued to be used. However, Sumerian was structurally inconsistent to Akkadian. To combat this, (and save time developing a brand new script), the Akkadians began to write out their texts phonetically using the Sumerian cuneiform symbols that most closely corresponded to the Akkadian sounds.

As if that wasn’t confusing enough, Akkadian cuneiform was also pretty wild to look at. As time went on the Akkadians altered cuneiform script into highly abstract versions of the original pictograms, some of which contained as many as 20 separate marks. Furthermore, some of the signs could be read either logographically or syllabically, making their true meaning more difficult to decipher.

This is confusing.

Basically it boils down to this: cuneiform script could be relatively time consuming to copy out and the meanings could be pretty unclear.

Also, recipe writing simply wasn’t that important to Mesopotamians, be they Sumerian or Akkadian. As Karen Radner and Eleanor Robson point out, most ancient texts dealt with numeracy and were quantitative in nature, focusing on administrative matters; wages, accounts and contract. How to make the perfect pavlova or a lasagne to wow the in-laws just wasn’t a major part of the Mesopotamian literature.

That’s not to say that Mesopotamian food was simple, far from it. In fact, you could argue that the annoyingly sparse succinct nature of the recipe suggests that Mesopotamian cooks were so adept that they knew, as standard, what phrases like “prepare water add fat” meant without needing extra hints and tips.

In addition to this the Mesopotamians enjoyed an abundance of fruit and vegetables, baked over 300 types of bread and made many types of cheese. They recorded their bounty on stelai and relief panels which show fruits like pomegranate, apricots and apples, vegetables such as radishes, lettuces and leeks, and meat including wild fowl, goat and cattle – hardly the range of inexperienced chefs.

The problem for modern historians trying to recreate these recipes, though, is that we don’t share their common knowledge. Knowing that the ancient Mesopotamians were experienced cooks who could understand the nuances of “prepare water add fat” was all well and good, but I still ended up in my kitchen boiling five kettles just in case and having a breakdown over whether to use lard, butter or olive oil.

And what the hell was samidu? Kanasu? Suhutinnu? Even Waitrose didn’t sell these ingredients (and they stock seven different types of salt and pepper!)

“Hey Google, translate ancient Akkadian”

Google translate was no help at all. It failed to detect any translation for samidu, told me that suhutinnu meant “mouthwash” and that kanasu was a “dream”. Reluctantly I accepted that I was going to have to do some proper research.

The authority on Mesopotamia was Jean Bottéro. He wrote the first book on cooking in Mesopotamia, The Oldest Cuisine in the World, and provided translations for some of the words found in the recipes.

Bottéro translated the samidu and suhutinnu as allium vegetables, like leeks or shallots. However, the food historian Laura Kelley believes that some of Bottéro’s translations are incorrect, and that rather than being a vegetable, samidu was more like semolina.

Her argument for this is based on similar words from nearby regions: The Syrians used the word semida to mean “fine meal”. The Greek word semidalis meant “the finest flour”. And “a fine flour [was] called semida in the Talmud (Pesachim 74b, Shabbat 110b, Moed Katan 28a.)” Similarly, Kelley argues that the word semida “is the Targum Yonatan translation for solet – also meaning “fine flour”.” The University of Chicago’s Assyrian Dictionary also defines semidu as semolina.

Kelley translates suhutinnu as some kind of root vegetable, but cannot be more specific. She suggests carrot, turnip or parsnip (but not an allium) based on the fact that tablets tell us nothing more than suhutinnu is “dug up”.

Bottéro gave no translation for the word kanasu (other than “a kind of edible plant”, which doesn’t really narrow it down), but Kelley believes it refers to emmer wheat flour. Emmer was one of the earliest crops domesticated in the Near East, growing naturally throughout the Fertile Crescent before domestication.

Making Kanasu Broth.

I gathered my ingredients: mutton was sourced at a local farm shop (I had to get diced rather than a full leg), emmer wheat was bought from a specialist mill, and semolina was found lurking on a back shelf in Waitrose, resentfully eyeing the more commercially successful rice pudding grains.

But how to cook it?

In his book A History of Food in 100 Recipes, William Sitwell points out that when the collected recipes are examined together, myriad cooking techniques are mentioned: “slicing, squeezing, pounding, steeping, shredding, marinating and even straining.” He suggested that despite the thousands of years separating us from the ancients, cooking techniques and food preparation hadn’t changed all that much.

So I followed my gut. Would I want to eat meat boiled in water without having been seared first? Not really; searing helps build flavour. I assumed, then, that the ancient Mesopotamians had a similar thought process.

With the meat seared and set aside I began work on the water and fat. Assuming I was making a stew rather than a broth after all, I added just 200ml of water to a pan .

The fat mentioned in the recipe was likely to have been sheep tail fat, which I couldn’t find anywhere at all at short notice, so I used olive oil. I figured that since olives had been cultivated in Mesopotamia from around 5000 years ago it wasn’t outside the realms of possibility that oil might have been used in place of animal fat when needed.

As the water and oil mixture heated, I crushed coriander and cumin seeds with a pestle and mortar and mixed the spices with semolina and emmer flour. Then the whole lot was tipped into the water and oil mixture and stirred to prevent lumps.

Once the sauce had thickened a little, I returned the browned meat to the pot. I added crushed garlic, sliced parsnips and a handful of mint and then let the whole thing cook on a low heat for just over an hour.

Final thoughts.

This meal was probably intended for the wealthy. The ingredients and tools needed (knives, caldrons, mills or grinders), to make it suggest it was probably cooked in palace or temple kitchens, rather than bog standard houses. Though literacy among Mesopotamians wasn’t just the domain of rich men, the idea that ordinary people would see the value in writing out recipes such as this just for personal use is also far fetched.

And the food itself? Delicious! Far more stew like than broth in my opinion, which I actually preferred. The mutton gave it a stronger flavour than lamb, but it wasn’t too dissimilar. The mint worked exceptionally well (who would have thought that the lamb/mint combo stretched back so far?!) and the spices were subtle enough to add depth, but not so overpowering that they drowned out the other flavours. In fact, I could probably have added half a teaspoon more than I did.

If you’d like to watch how I made this then please see the video below (and hang around til the end where I attempt a visual joke that made my husband declare he would “divorce me within 20 years” when he saw my preparations for it.)

E x

Kanasu broth

500g (1 pound) of mutton or lamb (I used diced but the original recipe refers to a leg. I give instructions below for both methods but if you are using a 1kg (2 pound) leg of lamb you will need to double the rest of the ingredients.)
200ml (6.7 fl. oz.) water
2 tablespoons of flour (any type will do)
1 tablespoon of semolina (or another tablespoon of flour if you can’t find semolina)
1 parsnip
2 cloves of garlic
1 teaspoon coriander seeds
1 teaspoon cumin seeds
2 tablespoons olive oil
Handful of mint

1. Sear the meat in a pan if using diced meat. (If using a full 1kg leg, cut holes in it and rub with olive oil and salt before roasting in an oven at 220C / 200C / gas mark 7 / 425F for about 45 minutes and then skip ahead to step 3. Remember to double the quantities for the rest of the ingredients!)
2. When the meat is browned, remove it from the pan.
3. Add the water and oil to the pan and heat.
4. Grind the coriander and cumin seeds in a pestle and mortar and then add the flour and semolina to them. Grind everything together and then add to the water and oil. Keep stirring! (If cooking a leg, skip to step 6.)
5. As the sauce thickens, add the diced meat back to the pan to allow it to cook through.
6. Chop the parsnip into chunks and add it, along with the garlic and some torn mint, to the pan.
7. Cover the pan with a lid and cook on a low heat for about an hour if using diced lamb or until the leg of lamb is cooked if using a leg. Keep checking on it regularly to stir it and add more water if it gets too thick.
8. Serve! (If using a leg of lamb allow the lamb to rest for 15 minutes or so after cooking before pouring the sauce over it.)