Back in 2018, at the very beginning of my work on Your Tasting Brain, I reached out to some experts in various aspects of the chemical senses and the things we use them for. One of them was Mary Svoboda, a director of R&D for a flavor company. Her insights helped me shape the character of the book.
RM: How did you come to this field?
I came to this work through a culinary background. I guess I come from a flavor family. My mom can describe a meal from 30 years ago with amazing detail of the way it tasted.
RM: How do you organize the universe of aroma relative to the things you work with? Is it by chemical group?
MS: Yes, thinking about it, I guess it is.
Aroma molecules are mostly based on carbon chains of different lengths with various side chains. I think a lot about those side chains, which determine if it the molecule is an ester, an alcohol, an aldehyde or whatever. Just changing that side chain changes the personality dramatically. Aldehydes, known for being green and fatty, always stick out in a formula. There’s a series of ‘goaty’ chemicals: Caproic acid is really, strongly goaty; ethyl caproate is fatty/fruity; and as an aldehyde, it’s called hexanal, which is green, but also goaty.
Some flavorists have a kind of synesthesia, where they see aromas as color. A deep berry fragrance, methyl anthranilate might be thought of as a deep purple color; strawberry furanone, a sort of cotton candy aroma, is thought of as brown. Benzyl acetate, a fruity-floral lavender chemical, is visualized as more pale, less vivid in color.
RM: Can you talk about flavor creation?
MS: You’re really trying to recreate the color, the shape, the form of the flavor experience. Think about your internal model of the flavor—let’s say strawberry. We’re going to create the form of the strawberry. As a client, you might ask me for a true, accurate version of a strawberry, but honestly, you don’t want real strawberry! That’s not what’s right in a food or drink product. It’s always going to be a specific interpretation, always a bit of an abstraction. That’s the art of it.
You know, Picasso did that painting, “Nude Descending a Staircase.” He wasn’t interested in showing us a nude. He was showing a motion. Flavor chemists are creating a language, trying to tell a story, but it’s abstract—never reality.
When I was a young flavorist, I was given a project: a peach cobbler. I worked on it a little every day for weeks, trying all kinds of different formulas. I don’t think I ever got it right, but that was OK. I was learning. These days, time is compressed. Clients are in a hurry for everything. There’s no time to fail. To build up your sense memory, you have to try a lot of different things.
RM: The difference between “almost” and “there it is.”
MS: It’s amazing. Just a few parts per billion can completely change how something smells. It’s like putting some anchovies into a dish you’re cooking. You can’t really call them out, but they make a big change. Just a few ppb of undecatriene, a green, fatty compound, can really make citrus flavors pop, and the same is true for a fishy-smelling compound called dimethyl anthranilate. Raspberry has some fruity components, but also dimethyl sulfide (DMS). If you put fresh raspberries in a paper bag and come back and smell it after a few minutes, it will smell just like creamed corn [the signature of DMS]. But add a little DMS to your raspberry flavor and it comes to life. If you take a basic raspberry flavor and add a compound called ionone, now you’ve moved into a kind of icy blue raspberry flavor.
Like chefs do, we to develop the ability to find the “hole” in the product: what’s missing? And also to tell what’s sticking out, like something doesn’t belong or is just is too much.
RM: Unexpected flavor notes can either bring everything into focus or take it off the rails, right?
MS: Perfumers have an idea that it’s not just about the pleasant smells, but just a little of something unpleasant adds a lot of depth and emotion. When you put just a little methyl mercaptan—that chemical they put into propane to make it smell—into something butter-flavored, it really makes it come alive. A tiny amount of indole [a chemical that smells like feces, but is also a vital component of jasmine] in a raspberry flavor makes it smell much more natural.
There’s this analytical approach, where you run things through the CG [gas chromatograph, a machine that separates and graphs volatile molecules], and then you just put those together but it doesn’t always work like that. We were duplicating a honey flavor, and when added all the identified compounds, the mix smelled just like genet, a floral ingredient with rose, citrus and tobacco notes. We had to figure out what wasn’t showing up on the GC and add those back to make it smell like honey.
RM: I’m pretty much blind to indole.
MS: Yes, that’s not that rare. We all have those blind spots.
RM: Do you have any hints or tips learned the hard way about any aspect of tasting for people who may be just starting their journey.
MS: Flavor chemistry is terrifying for most people. Most untrained people have about 40 to 60 vocabulary terms for flavor. For some people it’s only 15.
So the challenge is to create in your own head, your own flavor vocabulary, but for professionals we need to speak in a common sensory language. When we start training people, we encourage them to just say whatever strikes them, so for a particular compound: “I think this is green.” That’s close, but not precise. It’s actually a woody aroma, so we guide them through the shared vocabulary.
Today, flavor chemistry is all about the chemistry, and there are engineering types who see an analysis of a flavor and think that it should all be cut and dried. If they add all the chemicals they see on that chromatograph trace, they’ll have the flavor. But it doesn’t work like that, They’re missing the beauty.
And we like people to have that life experience and be familiar with a lot of flavors in the real world. We had a trainee who thought margarine tasted like butter. His idea of whipped cream was Cool-Whip. When I gave him real whipped cream, he was amazed: “This doesn’t taste like whipped cream. This has too much flavor.”
For me, the fun and interesting part is that you have to integrate both the left and right halves of the brain.
RM: The science is advancing rapidly, it seems. Is there anything new in the science that has gotten you really excited and helped do the work you do?
MS: There’s a lot of interesting work looking non-volatile (not normally aromatic) compounds, and how they contribute to overall flavor. There are a lot of non-aromatic things important in flavoring systems. Stevia is a plant you can grow in your backyard and if you chew on a leaf it will taste sweet, with no bitterness. They’ve identified a chemical called reb-A [rebaudioside-A], that’s the sweet component, and they’ve been separating that out and using it as a naturally-derived, non-caloric sweetener.
But there’s a problem. reb-A has a bitter, drying, nasty aftertaste. There are other reb types: B, C, and so on. So they’ve found that while these don’t have so much sweetness, but they’re very good at masking that bitterness, so they’re going back and recovering the part they used to throw away and extracting these other reb compounds and blending them in, and the flavor is much better.
All these innovations in non-volatiles are helpful because food processing is so extreme these days. People want a protein shake to have the same nutritional content as pea soup—these products have a lot of pea protein in them—but they want it to taste like a chocolate milkshake. These vegetable proteins and amino acids have a lot of bitterness and unpleasant aftertastes, so we need the tools to do that. But even though you can make a believable illusion of a milkshake, it’s impossible to make a pea shake taste like a really great milkshake, to make it taste perfect. We struggle with that.
We like to demo our products at trade shows; there is so much interest in these very highly processed foods, which, no matter what, will never be perfect products, will never be as delicious as real food. So we have to show products that are good as present science will allow.
RM: How is the flavor world changing?
MS: I think one big thing is sustainability. You have a shortage of really good raw materials. There’s a plant in South Africa called buchu, that has a fuzzy peach note and also some mintiness, which gives peach and cassis flavors a real lift. But there’s been a terrible drought. People don’t even have drinking water. So they’re telling us now they’ve had a little rain, so there may be some buchu, but the price will be going up, of course. At some point, people may have to get ready to reformulate their peach flavors if the situation doesn’t change.
Citrus is in trouble. There’s this problem called “greening,” that causes citrus to fall off the trees before it’s ripe. You see these trucks leaving the farms that used to be full of orange fruits, and now they’re a mix of orange and green. The juice isn’t as sweet, and there’s not as much of it. Juice producers are adding back distilled oils to make the flavor acceptable. It’s caused by a bacteria spread by an insect. The only cure is to burn down the groves to keep it from spreading. It’s happening everywhere: Florida, Mexico, Brazil.
So to satisfy people’s thirst for flavor, the future may just have to be artificial molecules. There is just not enough natural flavoring material in the world to meet demand. And it certainly won’t be about organics, which further limits yield, consistency and quality.