The Flavours of Ancient Roman Cookery

To me, Ancient Roman cookery has some striking parallels with Thai cookery (sans chillies, tomatoes and squashes/pumpkins). The other surprising feature of Roman recipes to modern eyes is that salt is almost never added to anything (salt was expensive, so if it had been used this would have been mentioned). Sugar was the other ingredient that's common today, but not available in Roman times (from the fall of the Western empire it would be another 250 years before Arab cropping systems introduced sugar cane to Europe).

But back to my comparison of Roman and Thai cooking styles. Thai cookery is significantly about the blend of sweet, salt, sour and spicy in a dish. Roman cookery is similar in that the flavour profile is about the blend of salt, sour and sweet. Saltiness was provided by fish sauce, which is ubiquitous in every recipe; sweetness from honey or reduced must and sour from vinegar.

Bread and grains were the staple, with beets and taro/eddoes being the other common carbohydrate sources. Meat and fish would be primarily boiled, though it could be fried, baked (in bread ovens) or spit roasted. Most primary dishes were simply prepared and flavoured with salt, sour and sweet (though fish sauce also provides a big hit of umami) as well as herbs and black pepper. The head of the household would show off their wealth and reach in the richness of the sauces with which the food was finished. As with most ages, cookbooks were the preserve of the rich. However, Roman cookery books are different and even those, such as Cato who professed to prefer simpler fare were still writing for the Senatorial or landed classes and the foods were still largely for parties and showing off. The books themselves are a little odd to modern eyes; though they may give a list of ingredients there is no detail as to how to exact quantities or how to prepare the dish.

Most of the cooks would be Greek slaves and it's likely that any cookbook or other book would be 'coffee table' volumes. Something to show off to guests to demonstrate one's culture rather than a book from which recipes could be picked up. After all, when a book has to be copied by hand it becomes a very expensive item.

Salt and Umami: Fish Sauce

But back to the flavours of Roman food. Most people know about fish sauce, the liquamen and garum of Roman cookery. Though the use of 'fish sauce' is synonymous with Roman cookery, it was invented by the Greeks who called it γαροσ (garos); from which the Roman name garvm derives. The Greeks may originally have invented the sauce out of a need to do something constructive with the large quantities of tiny fish they caught in their nets (fish too small to be consumed directly). Over the centuries the Greeks developed a taste for this sauce, using it both as a substitute for salt and as a flavour additive in their food. It is possible that it was made from the Mediterranean fish that the Greeks called garos. Contact between the Greek and Roman worlds (and the subsequent conquest of the Greeks by the Romans) brought garos to the attention of the Roman culinary sphere.

Unfortunately there are no historical records or writings about garum/liquamen before the first century CE. At this time the writings of Pliny the Elder infor us that garum was a liquid not unlike aged mead (honey wine) in appearance and which was often mixed with wine to drink. He tells us that it was made by mixing fish, fish intestines and salt together in sealed jars until it liquefies. Writing in the same century Martial informs us that garum was made from the blood of a still-breathing mackerel. This would seem to be little more than poetic license but it does suggest that fresh, rather than rotten, fish was used. It seems that fish sauce could be made from a single fish species or a mix of different fish. Martial tells us that mackerel garum was by far the best type. He also names muria as a fish sauce made from tuna but tells us that this was considered only of secondary quality.

As well as garum and liquamen we have now encountered muria and there was a more solid paste available as well, called allec. Undoubtedly there are differences between them. However, all result from the basic process of preserving fish with salt. Of course, we are familiar with the salting of fish today, though this is generally a rapid process of semi-drying. However, if the salt and fish mixture is left for long enough the salt draws all the water from the fish. This liquid contains sufficient sugars and proteins that fermentation can occur, a process that aids in the dissolution of the remaining solid matter. Essentially this is an anaerobic process and the fermentation process itself prevents the development of bacteris so that little or no decay or putrefaction is observed. The process is even more efficient if the blood and guts of the fish are included as these contain protease enzymes that aid in the breakdown of the fish flesh. There is also sufficient salt in the final liquid that bacteria cannot grow and the final sauce is essentially sterile. The remaining solid matter sinks to the bottom of the vessel that contains the sauce and it is this that the Romans harvested as allec.

As a process, the pickling fish in salt is dependent on a number of variables in that the ratio of salt to fish will affect the resulting sauce. Some recipes also call for the addition of different herbs and spices to the initial fish and salt mixture. Food historians have spent considerable time attempting to match the various descriptions of different kinds of fish sauce to the various names that have come down to us, without much success. It may well be that there was a standard form of fish sauce that everyone used as an addition to food and then there were several more expensive variants of this which were only used for and by the richest tables. It may well be that the main factor affecting the 'quality' of the fish sauce was its pungency, which was affected both by the species of fish used and by the amount of blood and intestines added to the initial mixture.

Indeed, as Martial's writings have indicated and as Manilius I also suggests in his poem, the Astronomica a fish sauce could be made entirely from the blood and intestines of fish. Manilius I describes fish being brought whole onto the beach. These are cut up and the 'precious fluid' is saved and mixed with salt. This would suggest that all but the flesh of the fish was used at this stage. Indeed, he later says that two types of sauce are generated. One from the blood and intestines and the other from the fish's flesh. The latter form generates a deposit which he describes as a 'soft accompaniment to food'. This could only be allec and suggests that the fish-flavoured liquid above the allec might itself be used as a kind of fish sauce.

Manilius I also describes a further process whereby vast numbers of small fish are netted and poured into large wine jars called dolia. It has to be assumed that salt is added and that these jars are sealed. But was this really how the sauce was made? The problem is that though we have descriptions of garum relating to the first century CE the first actual recipe we have comes from the 10th century CE. This uses fish flesh as well as extra blood and calls the resulting sauce liquamen.

The term liquamen for fish sauce is unknown in the first century CE (when we have the first recorded references to garum) and this substance is known from Apicius' fourth century CE cookbook, De Re Coquinara. The term liquamen for fish sauce seems to arrive with Apicius and is used almost exclusively thereafter. Subsequent Medieval sources even tell us that garum and liquamen are the same thing. Many redactors of ancient recipes have suggested that garum was a solid fish paste and that liquamen was the liquid fish sauce. However, this seems unlikely as the term allec was in common usage for the former type of product. What, then, is the truth of the situation?

The Greek origin of garos is undisputed. This may originally have been a way of rendering something useful from waste fish intestines. Eventually small fish were added to the recipe which led to another form of fish sauce. With the conquest of Greece both products were imported to Rome and used extensively by slave Greek cooks. The Romans acquired a taste for these sauces and in an attempt at increasing the market the idea of making specialized sauces from individual species of fish is introduced. To maximize production the entire fish is now used in the sauce and this quality product retains the name of garum. A secondary product made from tuna (and possibly with the inclusion of other fish) is called muria. The name liquamen may have been the name given to the cheap everyday product made from small fish. This 'basic' fish sauce would be the standard ingredient used by everyone and thus would not be worthy of note for the gastronomes of the first century CE. However, as the Roman empire declined the call for the expensive single-species sauce declined and the term garum gradually fell into disuse. By the fourth century garum was no longer available and even the most expensive of palates had to use liquamen as the only product available. An alternate view was that liquamen was the sauce made from the whole fish and was used in cooking whilst garum was made from fish guts and blood and it was a condiment served at table which guests could add to their own food.

Whatever the final product is like, the process of generationg garum/liquamen was very smelly. So pungent in fact that it was declared illegal to make garum in a private house. As a result the fish sauce was produced exclusively by garum factories along the Italian coastline. In actuality salting tanks for garum/liquamen production have been found all around the southern and south-western coasts of Iberia, from the Tagus and Sado Rivers to Cadiz. One of the largest concentrations of production facilities was found at Troia at the mouth of the River Sado, in Portugal. There were also smaller concentrations in the Algarve such as those at Boca do Rio, Praia da Luz, Lagos and Quinta do Lago. In terms of quality Lusitanian garum was considered the best and most superior product.

The nature of garum is even more complicated as there were different grades. With the type serve at the master’s top table being very different from that served to the slaves (think wine). The highest grade being produced from the best whole fish. Middling grades being produced from whole small fish and the lowest grades being produced from fish guts, trimmings and blood. As classical Latin became Vulgate Latin the terms liquamen and garum came to describe the garum suitable for cookery (liquamen) and the higher grade of garum served at table.

Regardless, as garum/liquamen was produced from fish fermented in brine, it’s essentially equivalent to Thai/Vietnamese fish sauce (Nam Pla/Nuoc Nam).

This makes things much simpler for modern cooks as these sauces can be used as direct substitutes in Classical recipes.

As a final note, there were vegetarian versions of garum extant. Vegetarianism was not unknown in the Roman empire, moreover, certain rights might require someone to become vegetarian for a time. As such, Palladius in his Opus Agriculturae 3.25.12 provides a recipe for making liquamen from pears:





For a modern redaction of this recipe see this site's Liquamen ex Piris (Liquamen from Pears) recipe.

Sweetness: Honey and Boiled Must

Without general access to sugar honey was the go-to sweetener for Romans. Romans were bee-keepers as attested by Varro (116–27 BCE) who provided details of variously shaped hives made from a variety of materials in use in his day:



Varro confirms that the hives are placed in walls for protection and the beekeeper would remove lids to reach the honeycomb. Columella (ca. 60 CE) records that hives were constructed from materials that were common in different regions. Some were made of cork, fennel sticks, or woven basketry materials. Like Varro, Columella argued that pottery hives were the worst. These observations by ancient writers have been evaluated by testing the thermal properties of pottery hives, and the findings suggest that they were not as inferior as Columella claimed

However, all the techniques for bee keeping were destructive to the hive and the scale of bee-keeping was insufficient to supply all the honey required by the Roman Empire.

This practice led directly to boiling down sweet wine or must to create sweeteners (which could generally be referred to as ‘grape syrup’). One of the earliest mentions of grape syrup comes from the fifth-century BC Greek physician Hippocrates, who refers to hépsēma (ἕψημα), the Greek name for the condiment. The fifth-century BC Athenian playwright Aristophanes also makes a reference to it, as does Roman-era Greek physician Galen.

This practice led directly to boiling down sweet wine or must to create sweeteners (which could generally be referred to as ‘grape syrup’). One of the earliest mentions of grape syrup comes from the fifth-century BC Greek physician Hippocrates, who refers to hépsēma (ἕψημα), the Greek name for the condiment. The fifth-century BC Athenian playwright Aristophanes also makes a reference to it, as does Roman-era Greek physician Galen.

Grape syrup was known by different names in Ancient Roman cuisine depending on the boiling procedure. Defrutum, carenum, and sapa were reductions of must. Must (from the Latin vinum mustum, "young wine") is freshly crushed fruit juice (usually grape juice) that contains the skins, seeds, and stems of the fruit. The solid portion of the must is called pomace and typically makes up 7–23% of the total weight of the must. Making must is the first step in winemaking. Because of its high glucose content, typically between 10 and 15%, must is also used as a sweetener in a variety of cuisines. Unlike commercially sold grape juice, which is filtered and pasteurized, must is thick with particulate matter, opaque, and comes in various shades of brown and purple. One of the most commonly mentioned sweeteners in Roman recipes is caroenum; which is probably a syrup made from reduced wine or must, must being freshly pressed grape juice with the skins and stalks still in it. Another reduced wine syrup is defrutum, which appears to be thicker than caroenum and which can also be flavoured with quinces or figs, as described in De Re Coquinaria:

In Apicius 2.2.8 a couple types of defrutum are mentioned, including:



Defrutum could also have spices added to as well as other ingredients such as sweet rush (Acorus calamus), iris, and fenugreek that would make it taste sweeter.

The elder Cato, Columella, and Pliny all describe how unfermented grape juice (mustum, must) was boiled to concentrate its natural sugars:



although the terms are not always consistent. Columella identifies defrutum as



Isidore of Seville, writing in the seventh century AD, says that it is sapa that has been reduced by a third but goes on to imagine that defrutum is so-called because it has been cheated or defrauded (defrudare) (Etymologies, XX.3.15). Varro reverses Pliny's proportions altogether (quoted in Nonius Marcellus, De Conpendiosa Doctrina, XVIII.551M).

The thickened syrup was used to sweeten and preserve wine and fruit that otherwise was sour or would spoil. Cato recommends that quinces and pears be preserved in boiled must (On Agriculture, VII.3) as does Varro (On Agriculture, I.59.3). Columella insists that defrutum always be boiled with quinces or some other flavoring (XII.20.2). Apicius offers directions for preserving quinces in defrutum and honey (De Re Coquinaria, I.21) and added the rich syrup to almost a fifth of his sauces to enhance the color and flavor of almost every dish. (That color was added indicates that red, rather than white, wine was used in the reduction.) In De Agri Cultura, the earliest example of Latin prose (c.160 BC), Cato gives directions for reducing must in:



Writing in the first century AD, Columella elaborates on the process.



The author Pliny the Elder wrote the following about the several syrups made from grape juice in his work Natural History:



Not only were grape syrups knowingly used in place of honey, but apparently to make cheap, knockoff honey mixtures! For the average diner, it seems that these grape sauces were to honey as, say, artificial maple syrup is to the real deal today. The flavour and quality can be approximated, but most people agree the more expensive product is better.

A lot of re-enactors and re-creators have attempted to use grape juice to re-create caroenum, defrutum and sapa but have found the results to be disappointing at best, with little difference between the products. Often, grape juice is assumed to behave the same as must and that exact proportions (boiling caroenum down to 1/3 the original volume, defrutum down to 50% the original volume and sapa down to 2/3 the original volume). However, reading Cato carefully makes it clear that it doesn’t work like that.

I was in South Africa during lockdown and used the time to revive my brewing and wine-making skills. Grape juice is made from a blend of grapes, in North America, this tends to be native grapes, Vitis labrusca rather than Vitis vinifera. For palatability grape juices tend to be blends that are ultimately not as sweet as wine grapes (compare wine grapes to table grapes). So, if you want to make caroenum you need to source sweet wine grapes; and as Romans loved sweet wines, the sweeter the better. Grape juices, particularly North American grape juices are not your starting point.

Though how much to reduce your must by might be a good guide; reading Cato makes it apparent that taste and texture should be your guide. Caroenum should be thick and syrupy and very sweet. Defrutum should not be as thick as caroenum but should still be very sweet. Sapa will be just reduced must, boiled enough to sterilize so that it will keep. It should still be sweet, due to the starting must being sweet.

We know from Cato’s writings that Romans were not averse to adulterating their foods to make more profit or to improve sweetness. Like adding caroenum to adulterate honey. Though not universally the case, musts were often boiled down in lead vessels (the other common vessels were copper, as Cato suggests). All grape juice has some levels of acetic acid (vinegar) in it. When acetic acid domes into contact with led, particularly in the presence of heat it forms the soluble compound, lead acetate. This compound has a naturally sweet taste that would further enhance the sugary flavour of the grape syrup. So cooking in lead seemed to improve the qualities of the boiled down must. Unfortunately, as lead acetate is soluble it is readily absorbed by the body. So the Romans’ sweet tooth was slowly giving diners chronic lead poisoning.

The main culinary use of defrutum was to help preserve and sweeten wine, but it was also added to fruit and meat dishes as a sweetening and souring agent and even given to food animals such as ducks and suckling pigs to improve the taste of their flesh. Defrutum was mixed with garum to make the popular condiment oenogarum. Quince and melon were preserved in defrutum and honey through the winter, and some Roman women used defrutum or sapa as a cosmetic. Defrutum was often used as a food preservative in provisions for Roman troops.

Defrutum can be purchased commercially today

I think it’s useful to consider these sweeteners/condiments the way we would molasses today. So these would be grape molasses. As pomegranates were commonplace in Rome at the time it always intrigued me as to whether the Romans made pomegranate molasses (they were known to grow everywhere and preserved roots have been found in Pompeii).

Passum also requires a mention here. It is a sweet wine made from raisins (semi-dried grapes), originally from Carthage that was also popular in Roman cooking.

The earliest surviving instruction constitutes the only known Carthaginian recipe. It is a fragment from the Punic farming manual by Mago in its Latin translation by Decimus Junius Silanus (2nd century BC). It survives because it was summarised by Columella (1st century AD). The original Punic work is lost, but the recipe is quoted in a later Latin work, De Agricultura by Columella.

Mago gives the following instructions for excellent passum. Harvest well-ripened very early bunches of grapes; reject any mildewed or damaged grapes. Fix in the ground forked branches or stakes not over four feet apart, linking them with poles. Lay reeds across them and spread the grapes on these in the sun, covering them at night to keep dew off. When they have dried, pick the grapes, put them in a fermenting vat or jar and add the best possible must (grape juice) so that they are just covered. When the grapes have absorbed it all and have swelled in six days, put them in a basket, press them and collect the passum. Next, tread the pressed grapes, adding very fresh must made from other grapes that have been sun-dried for three days. Mix all this and put the mixed mass through the press. Put this passum secundarium into sealed vessels immediately so that it will not become too austerum. After twenty or thirty days, when fermentation has ceased, rack into other vessels, seal the lids with gypsum and cover them with skins.

Passum was produced extensively in the eastern Mediterranean through the Roman period, and its popularity is referred to by Pliny the Elder in his Natural History.

"Passum de Magon", is a modern Tunisian natural sweet wine from Kelibia in the Cap Bon region, the traditional agricultural hub of Carthage, that honours the memory of Mago and is made in this antique fashion. Passito, the modern Italian wine made in this fashion. A notable passito comes from Pantelleria, an island in the Sicily Channel not far from the site of Carthage. Vin Santo, is an Italian dessert wine made from dried grapes.

Sourness: Vinegar

The other major flavouring in Roman cookery is vinegar. It is also a product of grape/wine production but is, in terms of basic flavour, the opposite of sweet. The main vinegar for Ancient Romans being that produced from wine, though the Roman Columella described the ingredients and process for making several types of vinegar in his work Res Rustica (12.1.5, 12.1.17, 12.1.34). Vinegar typically contains from 5% to 8% acetic acid by volume. Usually, the acetic acid is produced by a double fermentation, converting simple sugars to ethanol using yeast, and ethanol to acetic acid using acetic acid bacteria.

In the Greco-Roman world, posca in ancient Rome (the everyday drink of the Roman army) was made with vinegar and he ancient Greek drink oxymel is made from a blend of vinegar and honey.

Posca was the poorest grade of Roman wine, wine that had begun to spoil and turn into vinegar. Less acidic than vinegar, it still retained some of the aromas and texture of wine and was the preferred wine for the rations of Roman soldiers due to its low alcohol levels. Posca's use as soldiers' rations was codified in the Corpus Juris Civilis and amounted to around a litre per day. Posca and was one of the most commonly available wine for the general Roman populace and probably would have been for the most part derived from red wines, since white wine grapes would have been reserved for the upper class.

This is where I started to doubt my own sanity. I knew the origins of Posca, basically vinegary wine that was given to the troops. However, all over the web people are saying with certainty that posca was a blend of vinegar and water drunk by Roman soliders. But I really could find almost no primary sources confirming this. It does make sense, carting around vinegar which you then dilute in water is much better than carting around much larger volumes of vinegary wine. What we do know is that Roman soldiers were given a vinegar ration (Vegetius, Concerning Military Matters, 3.3), and that this vinegar could be mixed with water and drunk. (Celsus, On Medicine, 2.27). Wikipedia has a page on Posca (https://en.wikipedia.org/wiki/Posca) where it’s written with certainty that it was ‘an Ancient Roman drink made by mixing wine vinegar and water’. However, there are no primary sources to verify the claim. However, they do have one source Aëtius of Amida, writing on medicine. But Aetius is a 5th or 6th century Author, so not contemporary with Classical Rome and he uses the Greek description oxykraton (which does, I must confess refer to a blend of vinegar diluted with water). So, this is not a contemporaneous account.

However, the Greek tradition medical treatise did get me thinking. Could Posca be a medicinal drink that was given to soldiers as a way to make potentially bad water safer and more potable as well as acting as a pick-me-up. Now we find references in Oxyrhynchus Papyrus 1384 (though again of the 5th century). The next is from Anthimus De Observatione Ciborum 58. But again, this is a Byzantine work from the 6th century. So though there are later texts from the Greek tradition, there are still no contemporary references. I am a little more convinced that posca as a wine and water mixture was offered to soldiers, though it almost certainly evolved from low quality vinegary wine.

However, posca as a drink really does work, think of switchels or shrubs as contemporary drinks.

Though vinegar is by far the most common souring agent it’s not the only one employed in Ancient Rome… for this we can’ entirely ignore the role of sumac. Sumac berries grow on deciduous shrubs and trees in the Rhus genus of the Anacardiaceae family, making them distant cousins to cashews and mangoes. These trees are known for their pinnate leaves, vertical clusters of red berries, and white or green flowers. Rhus coriaria (also known as Sicilian sumac, Iranian sumac, and Syrian sumac, depending on who you ask) is the variation that originated in the Mediterranean basin before spreading across southern Europe and the Middle East. It can reach up to 3m (10 feet) in height and is principally grown for its edible berries, which are dried and ground to produce powdered sumac.

In ancient Greek and Roman societies, sumac was used for dying wool, tanning leather, treating indigestion, and, of course, adding acidity to food. Its culinary uses eventually spread throughout Europe before lemons were imported from the Middle East in the second century.