Will It Garum?
This blog has only existed for three weeks, but my post on hacking dietary supplements to create garum in one one-thousandth of the time as traditional garum has certainly made the most noise.
I’m excited to explore this culinary technique further, and it seems many of you are as well. To facilitate innovation in this space, I’m going to update this post regularly with “quick-garum” test results and learnings from myself and others. Feel free to comment with your findings!
Garum is an umami-rich sauce made from fermented proteins. You're probably familiar with fish sauce, a type of garum.
Unfortunately, garum typically takes at least 4 months to make. Alternative approaches pioneered by noma (i.e. using pancreatic hydrolysis via whole pork pancreas) are much faster, but they are practically infeasible for home cooks due to the inaccessibility of whole pancreas and the impracticality of producing small batches with whole organs.
Last week, however, I discovered a method for using easy-to-purchase dietary supplements to speed up garum production to an insane 3 hours. This totally democratizes garum production, and I’m pumped about it.
One advantage of this approach is rapid iteration. Since last week’s post, I’ve tried 11 different garum recipes. And I’ve already learned a great deal. In this post, which I’ll update regularly, I’ll share my findings as I hone this new technique. I invite others to test and iterate as well, and I’ll post those findings here, too.
Disclaimer: I’m not a chemist. I’ve been reading scientific literature to support this research, and I am following standard food safety protocols with my tests. If you’re a chemist and you see that I’ve misunderstood a chemical process, please let me know!
Substrate selection
The defining flavor of garum is umami. Umami is produced in garum through the breakdown of proteins by protease enzymes. Protease enzymes work with water to deconstruct complex proteins into their component parts. These component parts are easier for tastebuds to understand, so they are often more delicious. One of the resulting molecules of this process is glutamate, the source of umami.
As such, when choosing a substrate for quick-garum, you should choose a substrate with adequate protein. Merely breaking down food with enzymes is not garum. Otherwise, lactose-free milk (which is produced with lactase that breaks down the lactose in milk into simple sugars) would be garum! Similarly, my tests with breaking down pineapples (which are relatively low in protein) did not produce flavors that were reminiscent of umami or garum.
Thankfully, many substrates with high protein content exist — including vegan ones. You could try:
Meats, poultry, fish, and eggs
Dairy products (e.g. whey, milk, yogurt, buttermilk)
Seeds and nuts
Legumes and beans
Bee pollen
High protein fruits (e.g. guava)
With a traditional garum, you should avoid fats. Thats because both animal intestines and koji (the sources of protease enzymes in most garums) also contain lipase (enzymes that break down fats). When exposed to fats, lipase produce simple fatty acids that can become rancid and taste bitter or “soapy”. My pistachio garum test, for example, was far too bitter to be enjoyable. Pistachios are high in fat.
But unlike traditional garum, there are enzyme concoctions for sale that have protease but do not have lipase. If you can acquire such enzymes, it might be possible to produce garum with a fatty substrate and then separate the fat in the fridge after processing. Otherwise, just look for substrate without fat (e.g. for pistachio garum, I will try de-fatted nut flour next).
Since quick-garum lacks the long-term maillard reaction that adds flavor to traditional garum, you should consider searing or otherwise browning your substrate before you cook. This would not be possible with a classic fish sauce because the searing process could kill the natural enzymes in the fish’s intestines.
ENZYME selection
Enzymatic fermentation is not like the lacto-fermentation you’re probably used to. Unlike lactobacillus bacteria, protease enzymes do not reproduce. So, to break down your protein entirely, you’ll need to add the right dose of protease. There’s really no use with “backslopping” a new garum with a finished ferment.
In my dosing experiments with powdered protease (e.g. pork pancreas supplement), I start by testing 1% of the weight of the protein. Nik Skoufis has found that you can get away with 0.5% if you cook for twice as long. And if the protein has not adequately broken down after a few hours, you can always add more of the protease until you determine the right dose.
When choosing an enzymatic supplement to use for this garum technique, keep in mind that not all protease enzymes are the same. In fact, the various types of protease differ in their ability to produce either umami-rich or bitter flavors. The protease in pork pancreas and koji lean more towards umami, but the protease in pineapples (i.e. bromelain) lean more towards bitterness.
I’m satisfied with my pork pancreas supplements, but I understand that many people have dietary restrictions that prevent them from using pork. In my vegan garum tests with a plant enzyme supplement which contained both bromelain and papain protease, the results were too bitter to be enjoyable. (Update: I’ve also tested industrial protease derived from koji built specifically for maximum umami.)
Each protease enzyme also has an ideal temperature and pH level. Thankfully, these are well-documented. In general, however, these enzymes seem to prefer neutral pH, which might make combining quick-garum production with lacto-fermentation difficult. Futher, the ideal temperatures are typically lower than would safe to cook meats for a long period of time (i.e. lower than 140f). Thankfully, you can always make up for a less-than-ideal enzymatic environment by simply adding more enzymes (provided the enzymes exhibit some activity at your chosen temperature and pH level). 3 hours with my pork pancreas supplements at 140f liquifies proteins entirely.
When choosing a supplement (and especially when choosing a pancreatic supplement), look for supplements that are not heat-treated. Too much heat will kill the enzymes before you even use them. Mine are freeze-dried. I also prefer to find supplements that come in easy-to-pull-apart capsules.
Lastly, to ensure that the enzyme is properly distributed, you’ll likely need to agitate your garum during cooking multiple times. I take the bag out and shake it around about every 30 minutes.
Water ratio selection
Protease enzymes require water to do their thing. Water is also going to increase your yield (at the expense of a more diluted product). I tend to experiment with water ratios based on my assessment of the water percentage of the substrate. For example:
Fresh meats are very moist, so I use 50% of the meat’s weight in water.
Seared meats are less moist, so I prefer using 70% of the meat’s weight in water.
Bee pollen is not particularly moist, so I used 100% of the bee pollen’s weight in water.
Chicatana ants are dried for preservation (and they are expensive so I wanted higher yield), so I used 200% of the ants’ weight in water.
After blending the substrate with the enzymes and water, the result should look like a loose puree or porridge, think at least as loose as apple sauce.
Summary of findings
I’m going to keep track of all of my garum projects in this Google Sheet. Please feel free to comment in this blog post if you’ve stumbled upon useful findings as well. Provided the tests seem rigorous enough, I’ll add them to this Google Sheet.
And with that, I wish you best of luck! Happy testing.
