by Dr. Arielle Johnson
Dr. Arielle Johnson is the Science Director at Noma Projects. With a PhD in Flavor Chemistry, Arielle was a key member of the research and development team at noma and a part of establishing its original fermentation lab. Before rejoining the team at Noma Projects earlier this year, she has been across many exciting projects including her work at the MIT Media Lab.
“Traditional” XO sauce is a relatively recent invention, originating out of the Hong Kong restaurant scene of the 1980’s (which restaurant has the most legitimate claim over it is still debated over). True to its namesake (high-grade XO cognac), it’s a study in plush and warm richness. Rather than grape alcohol and oak, XO sauce constructs its richness from umami-abundant dried scallop, shrimp, and Jinhua ham, plus aromatic and warm chile pepper and garlic.
Vegan XO sauce is to XO sauce as XO sauce is to XO Cognac. In other words: a layered reinterpretation of a certain vibe, rather than an attempt at a direct flavor copy.
Each component brings something unique to completing the picture, where specific well-informed choices in selection and processing subtly but decisively push it in a certain direction. Let’s use it as a case study in understanding flavor knowledge in practice.
Wild Pepper
Piper borbonense, or in Malagasy, voatsiperifery, is a wild member of the genus Piper. “Regular” black pepper, Piper nigrum, is the most famous member of the genus, but only one of many used as a spice. While black pepper dominates its relatives today in Western cuisine, this wasn’t always the case: the culinary canon of spice-loving Medieval and Renaissance Europe also made heavy use of long pepper and cubeb pepper, and their relatively low profile today probably has less to do with anything lacking in their flavor and more to do with 17th-century colonial trade wars.
So what’s in a Piper? Pretty much all of them contain the spicy molecule piperine, but if you’re only looking for spiciness from pepper, you’re missing out on a kaleidoscope of delicate and complex flavors coming out of their aroma profiles. Black pepper and long pepper both go hard on woody and resinous flavors, with long pepper balanced towards deeper, warm-clove-y base notes and black pepper expressing more woody middle and top notes, enriched with citrusy limonene. Choosing wild borbonense, you get a much smaller dose of pungent spiciness, and a way more top-loaded aromascape, with huge notes of citrus, cardamom, heady pine resin, green wood, and a spiced quality with elements of saffron, dill, and nutmeg.
Pumpkin-bushi
Getting katsuobushi from raw fish muscle to ribonucleotide-laden partner to kombu (the marriage makes Dashi) is an exercise in patiently layering, transforming, and infusing flavor, through stages of boiling, drying, smoking, and mold fermentation. Executed traditionally, they’re an icon. Taken abstractly, they’re an opportunity to create something deliciously new.
When I first met Yusuke Sezaki, he had me climbing into the bucket of a forklift to get a better view (and better camera angle) of the layout of his family-owned Katsuobushi facility, Kaneshichi, within five minutes of arriving. Then, he showed me a ring he had made by cutting and polishing the translucent, ruby-pink inner core of a piece of katsuobushi until it looked like a gemstone.
From the outside it could look like irreverence or eccentricity, but what it really reveals is the skill of looking at a situation to deeply understand its nature and function—then taking whatever practical step you need to get the most out of it or express a different side of it. Yusuke-san thinks like really good R&D or test kitchen chefs think: let’s work hard to understand it, let’s not wait to try it out, let’s ruthlessly edit whatever turn out to be the hack ideas (however cool they sound) from the actually good ones.
After years of friendship and collaboration with Yusuke-san, with the knowledge that he was open to experimentation, we suggested applying his craft to other ingredients than bonito. He has since tested dozens, maybe hundreds, of ingredients for how they respond to the katsuobushi-making process. Think: persimmons, flowers, octopus, deer antlers. With the concept in mind of making a meatless reinterpretation of katsuobushi, he’s gotten stunning results with (relatively) humble pumpkins. Pragmatically skipping the mold fermentation step—it didn’t add anything flavor-wise, regardless of how cool it might look on Instagram—halved, seeded, guava-shaped pumpkins lose water, concentrate their flavor, and absorb just the right dose of smoke (assertive, but not harsh or overpowering) until they emerge in their final form as a richly flavored, leathery, essentially indefinitely shelf-stable grateable condiment, and the flavor heart-note of Vegan XO Sauce.
Dried Tomato
While most ingredients have just one or two of the five tastes in their profile (sour and sweet oranges, or salty and sour kosher pickles) tomatoes in their naked state check the boxes on four of them: sweet, sour, bitter (focused in the seeds) and umami. Sprinkle salt on a tomato slice and you’ll bring it over the finish line with 5 for 5.
Flavor is a dance between taste and aroma, and tomato’s umami sets the tone for how we read its aroma profile as a vegetable rather than a fruit. There’s familiarly vegetal grassy, metallic, and potato-y notes to it. But others take it in a decidedly fruity, dessert-y direction: geranium, citrus, caramel, coconut, roasted apple. Without the umami (which is anomalously strong in tomatoes compared to other fruits) a tomato would have a flavor somewhere in the ballpark of “weird strawberry”. With umami, it’s grounded as a definitively savory ingredient, (if one of the more flamboyant ones).
Tomato umami is also multidimensional. It contains both glutamate, which tastes umami, and ribonucleotides, which don’t taste like much but hugely intensify the umami taste of any glutamate around. It’s also distributed in a gradient, where the jelly around the seeds has about 3X the levels of glutamate and ribonucleotides as are in the flesh of the tomato. (We can thank chef Heston Blumenthal’s keen perception for that data point).
Drying a tomato totally changes its compositional balance and what it does in a dish—it goes from about 95% water to 15%, and takes on a jammy, chewy, leathery texture instead of a fresh and wet one because of this change. It can take a place in pastes and other recipes it would wash out in its raw form. It becomes even more of a taste powerhouse, since glutamate, acids, and sugars stick around during drying—and having removed flavorless water, you’ve concentrated them and their umami, tangy, and sweet effects.
Rose Oil
A rose is a rose is a rose, but like “black pepper”, there’s an innate elasticity to “rose” as a flavor. “Rose” comes in shades of sweetly citrusy-waxy rose; honeyed, more cocoa/wine-y rose; and rich, woody, earthy, roasted-fruit pipe-tobacco rose—each from a different kind of molecule, each expressed in different intensities depending on the chemical makeup of the rose variety in question.
Unlike the wild pepper used in its whole state, the rose in Vegan XO Sauce goes through a process of mediation: it’s extracted into oil, first, then the spent petals are separated from the now-infused liquid.
An infusion is never really an exact carbon copy of what it’s infused with—think how different coffee beans smell compared to their infused form in brewed coffee. Exactly how it differs—which side of its flavor is expressed—depends mostly on what ingredient you’re infusing into.
Specifically: is it, chemically, more like water or more like oil?
Hydrophilic (water or water-like) ingredients disproportionately pull out and get infused with hydrophilic molecules—in the realm of flavor, this means taste molecules like salt, sugars, amino acids, etc. Oleophilic (oil or oil-like) ingredients disproportionately pull out and get infused with oleophilic, hydrophobic molecules, which covers smell molecules. So a tealike infusion of something will express more of its tastes, and a lighter profile of its aromas, than an oil infusion will, which will have stronger aromas and very weak tastes.
That’s the simple version. The more interesting one is that the realm of “oleophilic aroma molecules” is not homogenous, but a spectrum from “pretty oleophilic and somewhat hydrophilic” to “extremely oleophilic and not at all hydrophilic”.
In other words, no smell molecules are as mixable with water as a salt molecule is, but some are way more mixable than others. They’re like a metal band with enough pop-y elements to have crossover hits.
With rose, specifically, the molecules responsible for the honeyed, wine-y rose quality are about 10 times more hydrophilic than the sweet citrusy-waxy or rich pipe-tobacco ones, so an infusion into water (literally, rosewater) has a much stronger, soft-honey-wine quality than an oil infusion. Rose essential oil, which is what floats to the top of rosewater as insoluble oil droplets while you’re distilling it, has those water-tolerant smell molecules depleted, expressing mostly a sweet-citrusy-waxy/rich-pipe-tobacco aroma from the molecules that remain. Infused rose oil has a more equitable distribution of all of them.
Don’t think of it as one version being necessarily better than the others—it’s more like having three shades of rose instead of one and understanding them all well enough to get just the effect you want.
Kanzuri
If you’re a fermentation enthusiast, you may spend time thinking about what makes a particular style a style: Light and dark misos can taste incredibly different and still be called “miso”, and differences in seasoning, water content, and technique take “fermented cabbage” in completely different directions as sauerkraut, paocai, and kimchi.
I really like thinking about how ratios of ingredients can be style-defining. Imagine a triangle where each corner has a name: Koji, Chile, and Yuzu. Any point on the triangle could represent a different type of salted, lactic fermentation.
If you’re directly on the “koji” corner, you’re making shio-koji, which takes koji, salt, and water and ferments them together to make a paste that can season other ingredients with creamy, tangy, salty flavors, or transform them with enzymes, making starchy things sweeter and protein-rich things tenderer and more umami.
If you’re sitting on the “chile” corner, you’re making fermented chiles—which in the end could look like a thin hot sauce, a thick sambal-like paste, or whole pickled peppers.
On the “Yuzu” corner, you’re making something I’m not sure commonly exists, which would be a salt-fermented yuzu that would look a lot like North African preserved lemons, or Vietnamese chanh muoi, pickled limes.
Stray from the corners and things start to get really interesting.
Hang out halfway between chile and yuzu and you’re making yuzukosho, a deliciously citrusy-aromatic, sharply spicy fermented yuzu-chile paste. Make a move to the middle of the triangle and you’re in the realm of Kanzuri: koji, chile, yuzu. It’s a little sweet and creamy, it’s a little earthy, it’s got tempered spiciness, and a fruity expression starting with fruit-leathery chile base notes and ending with sparkling yuzu topnotes. Of course, if you think only in triangles you miss out on some real-life details. Before their fermentation three-way, the salted togarashi chiles in Kanzuri get a cold snow bath called yuki-sarashi, which is said to leach out some of this initial salt dose and temper their spiciness.
Flavor Notes
Maximum Umami?
by Dr. Arielle Johnson
Dr. Arielle Johnson is the Science Director at Noma Projects. With a PhD in Flavor Chemistry, Arielle was a key member of the research and development team at noma and a part of establishing its original fermentation lab. Before rejoining the team at Noma Projects earlier this year, she has been across many exciting projects including her work at the MIT Media Lab.
The other day we got talking about “what is the most umami you could put into something?”
The dumbest, most brute-force answer to that question is to think like a chemist: umami is the taste of the molecule glutamate, so, what’s the most glutamate you could cram into something? If that “something” is water, the answer is about 650 grams per liter or 65% by volume. If you were stupid enough to do that, you could probably consume enough of it to get close to the lethal dose of glutamate (as measured in rats, about 16 grams per kilograms of bodyweight—as a benchmark, the lethal dose of salt is about 3 grams per kilogram). So, while it’s nice to know a theoretical upper limit to the concept of Maximum Umami, the practical answer is going to be significantly lower.
As a taste, umami is part chemistry and part sensation. Part of it is material you can measure, and part of it happens inside your brain. The link between the two is the umami taste receptor, which grabs glutamate molecules and trips the signal to the brain that it is experiencing the corresponding amount of the feeling of umami. Taste and smell receptors usually have a saturation point: they’re like a bunch of catcher’s mitts sticking out of the surface of your tongue and nose, waiting to catch the correctly-shaped “balls”. If you start throwing more than they can catch at them, they’re going to send the same “high” signal regardless of how much more than maximum capacity they’re at.
So you could reasonably approach this question like a neurobiologist, and ask what the maximum amount of glutamate the umami receptor can handle is. Scientists have done some good research on the dose-response curve of the umami receptor, but since we’ve only known about it for around 25 years, they’ve been mostly spending their time on more useful versions of this question, and not the “poke-it-with-a-stick” approach of trying to deliberately overwhelm the receptor. For isolated cells growing in a petri dish, the number looks to be around 160 mg per 100g, but since we have them over our whole tongue, the answer for most people is almost certainly quite a few times higher than that.
We (and I presume, you also) care about umami not because we’re particularly excited about the glutamate molecule or the umami receptor as isolated phenomena—we want to know because we like delicious food, and want to understand how it works. In that context, useful questions look less like “what’s the highest limit of solubility for glutamate?” or “How much can the receptor take?” and more like “What kind of foods have a lot of umami from glutamate?” and “where does that glutamate come from?” and “can we make more of it” and “what molecules or flavors can enhance umami indirectly?”
Approaching the question like a flavor chemist, I’d want to know how much glutamate tends to be present in food, and what foods or ingredients have the most. This is a question that there is a lot of data to answer, if you know where to look. (two good places that have pulled lots of literature together into single sources: The Umami Information Center website, or Ole Mouritsen’s book Umami). The short answer is: kelp (konbu, specifically), nori, aged cheeses, fermented fish, fermented soy, and dried mushrooms tend to have the highest levels of glutamate, potentially above 1000 mg/100g and up to 3000+ for the most umami-rich kombu.
Since cooking doesn’t just mean assembling ingredients, the biochemist in me wants to know if there’s anything we can do to increase the amount of free glutamate in an ingredient that feels lacking in umami. Looking at the list we just saw, there’s a lot of protein-rich ingredients (milk, fish, soy) that have been fermented or aged. The process in common? Breaking down those proteins over time, enzymatically. Proteins are made from a grab bag of 20 amino acids, including glutamate, but many proteins have a disproportionately high amount of glutamate woven in, about 20% of their weight. If you unleash protease enzymes—one of the best sources of which is the mold Aspergillus oryzae and the koji you can make from it—onto these proteins, they’ll eventually chew them up entirely and free all that glutamate to be tasted as umami. The theoretical upper limit for typical protein-rich ingredients is about 2600-3300 mg free glutamate per 100 grams (which, if you’ve been paying attention, is pretty close to those upper-level umami ingredients—meaning fermentation, given enough time, is one of the most effective ways to umamify protein ingredients.
Thinking about maximum umami like a sensory scientist, I might take it as a given that the ingredients I’m using have a certain amount of glutamate in them. And maybe even that I’d freed as much glutamate as possible from as many proteins as possible. Is that the end of the story?
Certainly not! Not least because glutamate isn’t the only factor in umami. Our sense of flavor is full of amplifications, enhancements, and synergy through suggestion, that makes eating as interesting an experience as it is.
Some things can augment umami directly. Food molecules called ribonucleotides enhance the umami intensity of glutamate by interfering with the umami taste receptor. They can’t activate it on their own (so they don’t taste particularly umami) but, they keep any molecule of glutamate clamped into the receptor for way longer than normal, which is like keeping your finger on someone’s doorbell. One press, but it keeps ringing and ringing. Same thing with the umami receptor and ribonucleotides—they can make an umami taste signal up to eight times more intense.
The supreme queen of secret umami enhancement is the absolutely ribonucleotide-packed ingredient Katsuobushi, the cooked, smoked, dried, and fungally-fermented loins of bonito that are an essential ingredient in many types of dashi, and hold a second title as the hardest food ingredient in the world. Other strong sources are fresh fish and meat, dried mushrooms (or fresh shiitake or matsutake), asparagus, nori, and tomatoes.
Since our sense of flavor is basically about figuring out how to feed ourselves, get the most out of that food, and avoid poisoning ourselves, there’s a lot of mutual signal enhancement going on behind the scenes. With flavor, never underestimate the power of suggestion. Sweet things will literally taste sweeter if they have aromas we associate with sweetness, like fruity or caramel-y ones. Same thing with umami: we associate funky, aged, cheesy, fermented, seaweedy aromas so much with umami, they amp up whatever umami is there in a big way. Unlike ribonucleotides, not so much of this happens at the umami receptor, and more of it in the places in the brain that package all of the separate flavor signals together into one.
Kokumi, a sensation of thick and rich mouthfeel and lingering richness, is neither a taste nor an aroma—we sense it with CaSR receptors, which are usually used for sensing calcium ions. This weird not-a-taste can make glutamate-water solutions taste around 50% more umami-intense in lab settings. Kokumi comes from small molecules called peptides, especially Glutathione, a 3-amino-acid long chain of glutamate, cysteine, and glycine.
Some good places to find kokumi peptides include beef, chicken, foie gras, scallops, tomato juice, brewed alcohol, soy sauce, garlic, onions, and long-aged cheeses like Gouda. Like free glutamate, it’s often, but not exclusively, associated with protein-rich aged and fermented ingredients—places where proteins have had time to break down into smaller pieces.
Just throwing together the most umami-rich or umami-enhacing ingredients and calling it a day may guarantee you’ve made an umami bomb, but it doesn’t guarantee that what you’ve made is as delicious as possible. Finding the umami sweet spot is less like layering every strategy on top of each other, and more like my grandmother’s rule (lifted from Coco Chanel) for dressing with good taste: when you leave the house, look in the mirror—then remove one accessory.
Cooking is a craft and an art because it requires you to pay attention, work with what you taste rather than how something is “supposed” to taste, and adapt to your ingredients using your judgement and intuition. On the other hand, thinking about umami like a chef doesn’t have to mean rejecting data and going with vibes only. It means using numbers and/or science for as long as they’re helpful—but trusting your palate as the final judge. Practically, it’s having that knowledge in your back pocket—as numbers, or even just as general trends—and using it to make decisions as you go.
Ways to get more umami, maximum or otherwise (to use as your palate and judement see fit).
- Use glutamate-rich ingredients like kombu, fish sauce, fermented soy, cheeses, dried mushrooms, dried tomatoes, or autolysed yeast.
- Rule of thumb: aged proteins, seaweeds, plants
- some mid-range (but still pretty glutamate-heavy) ingredients: walnuts, green tea, dry-cured ham, sake, fresh tomatoes, fresh molluscs like squid or scallop, corn, peas, potato, garlic, cabbage.
- Ferment protein-rich ingredients with protease-rich microbes like Aspergillus oryzae (in koji) and some lactic acid bacteria or blue cheese molds.
- Turn up the signal from naturally-occurring glutamate with ribonucleotide-rich ingredients like katsuobushi, fresh fish or meats, and for the non-meat eaters, mushrooms (especially dried), tomatoes, asparagus, or nori.
- Use kokumi-rich ingredients to create a sensation of richness that amplifies umami: garlic, onions, tomato, soy sauce, brewed alcohol (like beer or wine), foie gras, chicken, or beef.
- Enhance umami with other flavors: saltiness, sweetness, funky/fermented, cheesy, seaweed-y, meaty.
- Glutamate is really soluble in water and pretty poorly soluble in fat: infuse glutamate-rich ingredients in water (or wine, sake, juice, something water-based) for maximum umami transfer
- Glutamate is non-volatile: you can boil down or reduce a glutamate-rich liquid to remove water and concentrate the umami, if you don’t mind that it changes the aroma (that change might be delicious)
Flavor Notes
Travel Letter from Kyoto
by Kevin Jeung.
Jeung is Chef of Research and Production for both noma and Noma Projects, spending most of his time in the Fermentation Lab. Jeung spent three months in and around Kyoto as part of the noma Kyoto pop-up where the team researched new and exciting ingredients and production methods.