PPC 105 (April 2016)

Ancient Fish Sauce in a Modern WorldSally Grainger

I was delighted recently to receive a parcel containing little bottles of fish sauce which Tom Jaine had been sent by a lady in the United States. The internet blurb claimed that the sauces therein were a serious attempt at reproducing ancient garum.
I was of course intrigued and impressed in equal measure as they were remarkably good sauces. Whether they were an accurate reproduction of garum remained to be seen. Tom Jaine has asked me to write a review, more of which anon, but first I thought I would consider the tradition of making fish sauce in the Mediterranean today and look particularly at those sauces which either claim a link to the ancient past or which we associate with ancient Roman sauces: namely Colatura di alici from the Bay of Naples and pissalate from the south of France, particularly that made in Antibes and Nice. I will also take a look at a Japanese squid viscera sauce called ishiri which has some similarity to the black and bloody ancient garum.
As many of you will know I have been researching ancient fish sauces over many years and I am particularly concerned with the precise nature and texture of these products: how they might appear in use in the kitchen and at table; how this might be different from their appearance in transit in amphorae; and, crucially, how they appear in production. This knowledge will allow archaeologists to recognize the residues of the sauces in the archaeological record, which in turn will hopefully allow a greater understanding of the economic system behind their trade. I am nothing if not obsessive about these issues, particularly as there is still no clear agreement about the ancient sauces among scholars of archaeology and history and this lack of clarity is mirrored in the way in which modern interpretations of garum and liquamen are understood and reproduced.
A simple definition of terms and products is necessary – I admit, these definitions are mine. Ancient fish sauce came in three varieties:
1) Garum / garon melan  / haimation in Greek: a ‘black and bloody garum, a blood and viscera sauce – crucially no fish meat is present in the recipe. The process is an enclosed lactic acid bacterial (LAB) fermentation and enzyme action with an indeterminate amount of salt over about three months. It was either made from tuna or mackerel viscera and blood or blood and a miscellaneous variety of whatever viscera was available, which was probably considered less exclusive. The small-scale recipe for garum / haimation in the Geoponica tells us to seal the viscera, salt and blood in a relatively small ceramic vessel. The vessel is eventually pierced and the sauce flows out while leaving the undesirable semi-decomposed visceral material behind. There is no indication of extra heat, though a vessel placed in full sun would allow the contents to increase its temperature a little. My own attempts to reproduce this sauce were inadequate for a number of reasons, not least the difficulty of harvesting blood from mackerel before it coagulates, as this involves bleeding the fish while very fresh – essentially, still alive. We have fish bone evidence for the bleeding of tuna through the gills, as this leaves a tiny bone residue which has been found in the bottom of a jar in Roman Jordan. My mackerel garum did display distinct qualities such as an absence of the fishy taste one expects from nuc nam or nam pla and a distinct iron favour characteristic of other umami-rich dishes made with blood such as black pudding and jugged hare. Black and bloody garum was not used in cooking but at the end of the process, at table, or in special dips known as oenogarum which were blended at table.
2) Liquamen / garon in Greek: a whole-fish sauce made with relatively small fish such as anchovy, sardine, mullet, or bream and additional salt with an open or closed fermentation process for a period of about three months. This sauce resembles south-east Asian sauces in every respect: the fish dissolve by enzyme action derived from the viscera, in summer air temperatures of c. 25–35°C. A natural brine is generated by the fish and when the sauce is made open to the sun, this always has to be topped up with extra brine thanks to evaporation. Modern fish sauce does not utilize bacterial fermentation but this is probably due to the high sodium content. Ancient sauces use a good deal less salt and may therefore involve LAB fermentation. Liquamen resembles nuc nam and nam pla in terms of colour, as experiments have produced shades of amber to dark brown, while Pliny describes it as looking like aged honey wine (HN 31.95). A quick method of making liquamen recorded in the recipes involved boiling fresh fish and straining the liquor but it is very unlikely that the two methods were combined during the initial production process. Much later, the residue of liquamen known as allec may have been washed with hot brine to extract a second sauce at the place of sale and / or utilization. The ancient sauce was used in cooking just as are the modern varieties, though it is clear that a luxury liquamen made from mackerel could also be used as a table sauce for oenogarum, which may be the origin of the recurrent ancient and modern confusion of garum and liquamen – as we will see, this confusion is even more complex than simple semantics. The residue of liquamen looked like a fish paste resembling piperatum, the anchovy paste. Ancient allec could also be marketed as a desirable product but was certainly a bone-free fish paste, and allec was also made with bone-free seafood such as sea urchins, oysters and mullet livers. When made just from viscera, garum did not generate a fish-paste allec and this is a fundamental distinction when attempting to identify the products in archaeology, through either residue analysis or, particularly, amphora shape.
3) Muria / halmer: the brine or pickle in which cleaned salted fish was stored in during transit. It was considerably saltier than the other two and might also come in various qualities. The absence of viscera and therefore enzymes means that there was limited transfer of nutrition to the sauce and this would result in a paler colour, though a brine derived from long-term storage of specific species such as mackerel and tuna could result in a stronger taste and darker colour. We do find reference to labels suggesting that some forms of salted fish were aged for 3–4 years in brine and it is unlikely that the fish itself was consumed after such a time but the brine may have developed flavours that could be considered more desirable than that derived from a relatively short-term storage.2 It is apparent that the Greeks valued tuna brine as an élite product during the Hellenistic period and blended dips with it similar to oenogarum. The Romans continued this practice.
To sum up: we have a very dark (it is called by Galen ‘black garum’), iron-rich, nutritious table condiment made from viscera and blood of mackerel and tuna with salt in a long fermentation; a fish sauce in shades of brown used as a general cooking sauce, made from the whole small fish, with a fishy and salty taste; and a fish brine that was naturally pale in colour, subtle in flavour (save that it was very salty) and weak in nutrition, though long-term storage may have improved all of these characteristics.