After writing about saving a woman from a cow attack at Hadrian’s Wall, I really wanted to find a recipe I could call “Roman revenge on the cow”. Did the Romans also have problems with aggressive cows along the wall, and if so, did they ever butcher and feast on the offenders?
The answer is probably yes to both. Beef was a favorite meat for the army along Hadrian’s Wall. It was supplied regularly to the garrisons, I’m guessing in the form of live cattle, because that keeps the meat better. And where there are cows, there will always be a few that lower their horns and butt heads, and where there were Romans, there would have been an eager butcher.
I turned a 1,500-old Roman cookbook by Apicius, De re coquinaria, to find ancient beef recipes. Some of them weren’t appealing (beef hooves, anyone?), but a recipe for venison with jus caught my eye. My German translation of Apicius appears to convert the venison into beef. It’s hard to tell because it numbers the recipes differently than the Latin original, but beef would be a good substitute just in case you aren’t successful hunting deer along the wall.
But the really fun thing about this recipe is that some historians suggest it is the origin of beef bordelaise! In fact, the connection to bordelaise was the key that helped me interpret Apicius’s recipe. Julia Child, watch out, because we are going to cook up the prototype, and it ain’t French….
Challenges in recreating Roman recipes
Roman recipes can be tricky to recreate today. I had to try this recipe three times before I ended up with something that tasted mouthwatering. But this post is all about my mistakes and corrections so you don’t need to repeat them.

The first problem with Apicius is huge: He only lists the ingredients. Forget about amounts or cooking instructions. He assumed the reader would have enough cooking experience to create something scrumptious from his lists. But the ambiguity of the ancient recipes also allows us to get creative and try different approaches. Don’t be afraid to make mistakes. I sure did, as you’ll see below.
Second, historians disagree about the translations of some of Apicius’s ingredients. An incorrect translation led to my unsatisfactory first attempt and marred the flavor of the sauce.
Lastly, some ingredients might be hard to get to day. If you don’t happen to have Roman liquamen (garum, or fermented fish sauce) handy in your fridge, I can tell you what to use instead.
There is no need to cook this recipe over an open fire with ancient cooking equipment. You can try it in your own kitchen. Let’s sharpen our knives and get to work!
Ingredients and their translations
Here’s the only information Apicius gives us:
Ius in cervo – jus on venison
Piper – pepper (I used about half a teaspoon).
Ligusticum – translation uncertain, but not lovage! (I left this ingredient out).
Cepulam – onion (One small onion, finely chopped).
Origanum – oregano (or marjoram; I used about a teaspoon).
Nucleos – nuts (I’m allergic and had to leave this one out, but feel free to experiment).
Caryotas – dates (I substituted a handful of raisins).
Mel – honey (I used one tablespoon).
Liquamen – Garum (fermented fish sauce; I used a half a cup of Worcester and soy sauce combined)*
Sinape – mustard (I left this ingredient out)
Acetum – vinegar (I used a quarter cup).
Oleum – oil (I used both olive oil and butter to fry the steak).
And that’s it!
First, let’s dissect those ingredients.
A few substitutions
My German translation of Apicius suggests that the meat should be roasted. It leaves out the nuts and mustard, adds celery seeds and caraway (I added a teaspoon of each), wine (use half a cup of red wine!) and must (pressed fruit juice, especially the juice used for making wine), and substitutes raisins for the dates. I opted for each of those changes and used half a cup of pomegranate juice for the must because I know how popular pomegranates are in Italian gardens.
Liquamen, the fish sauce
Then there is the iconic liquamen, garum, or fish sauce, which was very popular in Rome. Although a surviving third-century recipe calls for salted fish to be fermented for one week with dill, coriander, fennel and other dried herbs, historians weren’t really sure how Romans made the sauce. A breakthrough occurred in 2009 when archaeologists found clay vessels containing the remains of liquamen.
Gas chromatography and an electron microscope investigation of the contents solved the riddle of the ingredients. Liquamen was composed of salted anchovies and mint, sage, thyme, oregano and other herbs. The chemical engineers who did the research then tried to recreate the recipe. After 25 days of fermentation, an amber-colored liquid developed on top of the fish and herbs. The scientists-turned-culinary-detectives sampled it. It had a rich umami flavor. Chemical analysis revealed a high level of glutamic acid, the amino acid that gives Parmesan cheese and tamari sauce their distinctive flavors.
This liquamen is now being marketed in Spain as Flor de Garum. Some top Spanish chefs, including Michelin-starred Mauro Barreiro, have endorsed it. You can now purchase the scientifically recreated liquamen online! My next culinary journey will be to order Flor de Garum and try it in some ancient recipes.
If you don’t happen to have it handy, you can substitute Worcester sauce, Asian oyster sauce, or Japanese dashi broth (often available in powdered form in Asian grocery stores. I used a combination of Worcester and soy sauce and it worked well.
Ligusticum is not lovage!
Ligusticum is a problem ingredient. It is often translated as lovage, and in fact, levisticum, the scientific name for lovage, sounds similar. Lovage has a strong celery taste and needs to be used in moderation, or else it will overpower the dish. We have lovage in our garden and I use it to make soufflés. I used it the first time I made this recipe because the translation I used said lovage. But it didn’t work. Celery seeds and lovage together were too much.

The flavor disaster drove me back to another German book on Roman cooking for further research. Grenzenlose Gaumenfreuden: Römische Küche in einer germanischen Provinz says ligusticum can’t be lovage because lovage, native to Afghanistan and Iran, hadn’t yet been introduced to Rome when Apicius wrote his book. The oldest archaeological find of lovage in Europe is from medieval Prague. There are no finds in Rome. Furthermore, Dioscurides, an ancient Greek author, describes ligusticum as having fragrant leaves that are feathered like dill or fennel leaves. Lovage has flat leaves. A seventh-century illustration of “levisticum” in the illuminated manuscript Naples Dioscurides resemble parsnip leaves. But here I have to give you a warning: Wild parsnip leaves are toxic, and there is conflicting information about whether cultivated parsnip leaves are too. Do your research and use at your own risk. I decided to drop ligusticum from my ingredient list. The recipe was still tasty.
What cut of beef?
For beef bordelaise, Julia Child recommends several cuts of beef:
Entrecôte or rib eye
Romsteck or rump steak
Faux filet or false filet
Bifteck or rumpsteak
I used the first, rib eye.
Cooking your ancient beef bordelaise
My German translation of Apicius says to roast the beef. To keep the meat both tender and Roman, I used a Römertopf, a clay pot popular in Germany that imitates a cooking technique of ancient Rome. You need to soak the clay post and its lid in water before roasting the meat in the oven, and the steam from the pot is supposed to keep the meat tender. I added both the meat and the sauce to the pot.
I’ve had good results cooking with the Römertopf, but not this time. The roast was way too tough.
My discovery that this recipe might have been the prototype for beef bordelaise was the key to reinterpreting it. I abandoned the German cookbook and turned to Julia Child. She explains that bordelaise sauce is used with pan fried steaks. Once you’ve fried your steak, you remove it and deglaze your pan with the sauce ingredients.
I tried that the second time, the result was better, but the meat was still too tough.
There are two ways to tenderize meat, with a meat mallet and by marinating, and I used both techniques the third time. After pounding the steaks with the mallet, I marinated them with Apicius’s ingredients overnight, just like I would for a sauerbraten. One needs to pat the steaks dry after removing them from the marinade – dry meat browns better. I fried the steaks about five minutes on each side until barely pink inside. Then I deglazed the pan and reduced the sauce.
This time my husband and I loved the results.
I encourage you to try it yourself and enjoy the meat of the cow that attacked you. At least we now know how beef bordelaise got started.
References
Apicus, De re coquinaria (On the art of cooking). You can read the original Latin here. The recipe is at (344) in chapter 8.
Jutta Meurers-Balke & Tünde Kaszab-Oblschewski, eds., Grenzenlose Gaumenfreuden: Römische Küche in einer germanischen Provinz (Mainz: Verlag Philipp von Zabern, 2010).
Julia Child, Louisette Bertholle, & Simone Beck, Mastering the Art of French Cooking, Vol. I (New York: Alfred A. Knopf, 2011).
Tim Cooper, Romans made us beef eaters, Guardian (August 2000).
Richard Gollmet, Das Apicius Kochbuch: Aus der Römischen Kaiserzeit (Daun, Germany: Regionala Verlag, 9th ed., 2019).
Grainger, Sally. “Towards an Authentic Roman Sauce.” 2005 Oxford Food Symposium, p. 206 ff.
Miranda Greene, Food and Diet in Roman Britain, Herefordshire Council (2004).
Taras Grescoe, Culinary Detectives Try to Recover the Formula for a Deliciously Fishy Roman Condiment, Smithsonian Magazine (November 2021).
Joseph Dommers Vehling, Apicius: Cookery and Dining in Imperial Rome (Project Gutenberg, 2009).
English Heritage, Romans: Food and Health (undated).
Legal disclaimer: We disclaim all liability for the content of websites to which our site provides links.
Three times a charm. Fascinating to be an arm-chair observer of attempts to create medival cuisine. The jus sounds delicious.
A fascinating story - and living within a few kilometres of the wall I feel I really should work on my on version after all the work you have done here
I am curious about your cooking times however - with many cuts the secret is either very fast or very slow!
Cooking in the Rumertopf I suspect you are going slowly - it should be similar to my cooking over charcoal in a ceramic oven - if doing that I would be leaving for a (very) long time 4-6 hours as I would be looking to get the meat core temp to around 90-93C
Just a thought!! Fascinating piece!!