Playing God (or Carving Venus): Food Product Design

When writing my article on terpene removal a search for an author I quoted led me to this interesting 1998 article, The Sweet Taste of Success, published by Food Product Design. I have bunch of masters program text books on food science for food product designers and some of the ideas from industrial food scientists range from insightful & interesting to startling & creepy. They sometimes pen justifications for using artificial ingredients they call nature equivalent and rationalize them as more friendly to ecosystems than growing natural ingredients. They are known for not liking to waste anything so they take every fatty scrap and invent snacks for children (the road to hell is paved with good intentions).

But there is also great ideas to be found and I’m only high lighting this article because when I started collecting vermouth literature so many years ago, I was looking for unique language that flavor professionals used to discuss the very complex things they were constructing. Did they have language the flavor layman didn’t have and did that help them achieve so much? Sadly, I didn’t find anything too unique and I started creating my own language using ideas from aesthetics, sensory science, cognitive linguistics, metaphor theory, and category theory.

Here goes, lets highlight some passages.

Before becoming a food scientist, I couldn’t understand why my homemade yellow cake and freshly squeezed lemonade didn’t pack the full flavor of grocery-store products. It was only after touring my first flavor-manufacturing facility did I understand why my creations paled next to commercially prepared foods.

Oh god, what an introduction. What author Lisa Kobs is getting at is how commercial food manufacturers use every trick in the book to create a supernormal stimuli.

Flavor chemists have access to thousands of flavor compounds capable of accentuating the subtle nuances of sweet goods. The literature tends to focus more on the application of flavor to savory, rather than sweet, food products. But with a basic understanding of how to properly use flavoring ingredients, the food scientist can create the right flavor system for sweet applications.

This implies fragmenting something into a series of categories and manipulating them independently until you can create a seductive experience that exploits all of our reward mechanisms.

The four most common processing methods – Bourbon, Mexican, Tahitian and Java Indonesian – vary in the length of time beans are grown before picking; duration of drying; and the drying method used, which can include sun-roasting and fire-curing.

This differentiation of vanilla beans is new to me and very interesting. She describes vanilla as the chief way to enhance sweets but personally its a flavor I’ve rebelled against, often seeming too plebian and ordinary.

An aroma profile common to all vanillas is described as sharply acidic with slightly bitter back notes and a pronounced pungency.

In this statement note that she is describing olfaction in terms of gustation which is the first layer of my aroma categorization schema. I had also never seen vanilla referred to as acidic before.

However, vanillas have characteristic flavors and aromas based on their country of origin. Bourbon-processed vanilla beans, grown mostly in Madagascar and the Comoro Islands, produce a high-vanillin-content vanilla described as rich, smooth, rummy and full-bodied. Mexican vanilla beans have a lower vanillin content and the vanilla lacks the body associated with the bourbon beans. Its flavor profile has been described as sharp, slightly pungent, woody, resinous, sweet and spicy. Tahitian vanilla is distinctively sweet, very fragrant and perfume-like, with coumarinic flavor and heliotropine notes. Java vanilla beans, from Indonesia, produce a vanilla described as deep, full-bodied, harsh, smoky and phenolic.

Awesome descriptive language and differentiation here. She uses varying categories to describe each of the beans even using two iconic object comparisons for the Tahitian beans.

Ethyl vanillin is a chemically processed flavor made from the coal-tar derivative, guaiacol. It has an intense, vanilla-like odor, and has a more powerful flavor than vanillin. It can feature a harsh “chemical” character when used at too high a level. A number of other, less well-known components delivering a vanilla flavor include: veratraldehyde, which is herbaceous and warm; heliotropine, which is sweet, spicy and floral; anisyl acetate, which is powdery and floral; and vanitrope, which has a warm, spicy medicinal sweetness.

Coal-tar, who would have thought? I’m not afraid of that kind of thing but it is surprising. Here we see a “chemical” descriptor among many other categories. Powdery is a surprising one and the paper Understanding the Underlying Dimensions in Perfumers’ Odor Perception Space as a Basis for Developing Meaningful Odor Maps helps correlate such descriptors to others that are better known.

The category of sweet, brown flavors includes those flavors having the connotations of roasted, burnt or caramelized flavor systems, according to Carol Pollock, director, sweet and beverage flavor creations, Wild Flavors, Inc., Cincinnati. They can be extracted from botanicals and supplemented with other natural and artificial flavors, or they can be created by a reaction process. Flavors within this category include brown sugar, graham cracker, malt, honey, maple, molasses, caramel, butterscotch, coffee and chocolate.

Here she uses the term category which may seem insignificant but believe me its significant.

Flavor profiles for the base notes in many sweet brown flavors are similar. St. John’s bread, an extract of the carob plant, forms the base note for many brown flavors. Brown sugar gets its distinctive flavor from a thin coating of molasses on the granulated sucrose. Butterscotch flavor is made from heating butter, sugar, fat and salt. Lipase activity from the butter, caramelization from heated sugars, and Maillard reactions from the sugar and protein generate this flavor. Many of caramel’s flavor notes can be found in butterscotch, but with a twist. Botanical extracts that make up the sweet browns include black hawthorne, fenugreek, yerba mate and lovage. Brown flavors tend to contain more backnotes and mouthfeel rather than aromatics, and many of them have actual extracts of the ingredient in them, such as coffee or chocolate.

I love the idea in here of yerba mate. Flavor formulators love to surprise and here is an example of it in action. Yerba mate is a fragment or sub category of a larger category like sweet-brown so it fits because it fills its category role but it turns heads because its different and that is relatively more extraordinary. A pattern is found and put to use with a fun variation.

Honey. Honey is considered a sweetener, but one with a characteristic flavor. A complex flavor results from the sugars, acids, tannins, and volatile and nonvolatile components within it.

This is one reason why I specify non-aromatic when I use white sugar. It eludes to variations that could provide aromas such as using honey which is more than just aroma but rather flavor.

Using honey at high levels also can be quite expensive. The solution may be a honey flavor. The flavor chemist can engineer an excellent artificial honey flavor, and a blend of honey and other sweeteners boosted with a honey flavor would provide the desired flavor characteristics at a lower cost without the accompanying texture problems. Often a mixture of real honey and honey flavor can taste more like honey than actual honey does.

Lets quote that last sentance again:

Often a mixture of real honey and honey flavor can taste more like honey than actual honey does.

Text book supernormal stimuli: where there is a response tendency we create an exaggerated response tendency. Boom! Don’t let flimsy symbolic constraints like being natural get in your way…

Maple syrup. Maple syrup, the sap of black maple and sugar maple trees, is another sweetener containing a characterizing sweet brown flavor. The sap is concentrated through an evaporative process, which thickens it and intensifies the flavor. Syrup right out of the tree is mostly sucrose. Evaporation produces some glucose and fructose upon inversion at a low pH. One group of flavoring components comes from the ligneous materials from the sap, but a second group is formed by the caramelization of sugars.

A really interesting way to sum up maple. I didn’t know it started as sucrose.

Maple flavors have been developed by the extraction of botanicals, such as fenugreek and lovage, or chemical compounds, such as cyclotene and methyl cyclopentenone. It’s important to distinguish real maple flavor from maple syrup flavor. Processed, artificially flavored maple syrups have become almost a standard of maple flavor, while a true maple flavor has a completely different character.

Really interesting ideas on how to elaborate maple. And then the ubiquitousness of the artificial version has superseded the natural version? Interesting.

Chocolate flavors typically contain actual chocolate, or extracts and distillates from the cocoa beans. Artificial chocolate is difficult to make without any real chocolate extractive components because of the complexity of the flavor, according to Gary Reineccius, professor in food science, department of food science and nutrition, University of Minnesota, Minneapolis. “It’s very difficult to make a totally natural chocolate flavor, because the chemicals comprising chocolate flavor aren’t available in natural form, and the flavorist won’t even get close to a mediocre natural chocolate flavor by putting together pure chemicals without adding chocolate products.”

Its amazing how chocolate can elude forgery. Is the word forgery appropriate?

Vanilla and vanillin are commonly added to enhance the flavor of chocolate. They also are the primary source of flavor in white chocolate, which is a blend of cocoa butter, sugar and milk. Another developer’s trick to increase the perception of chocolate flavor is to darken the food matrix. The deep brown color of a chocolate cake will send connotations of rich chocolate flavor to the consumer’s mind before it is ever tasted.

Perception is the meeting point of incoming sensation and outgoing recollection. He color primes your recollections before you even taste. #phenomenology.

Aside from the adjective “coffee-flavored,” it can be called acidic, full-bodied, mellow, mocha, soft, nutty, rich, smooth, acidic, spicy, smoky, winey, heavy, chocolate, bright and earthy.

She goes from one upper level object comparison to other lower level object comparisons, sensations, and grounded metaphors where one sense in described in terms of another. Separating sensations like acidic from grounded metaphors like heavy is not always easy. In another context without much cluing, acidic could also be a grounded metaphor.

Coffee flavors have been developed by profiling the extractives of the native beans for their flavor, and then analyzing these chemicals and their composition. Reineccius explains that a compound called furfurylmercaptan can help the developer create coffee flavor without using coffee. Since this flavor isn’t available naturally, it must be labeled as artificial. It’s impossible to make a natural coffee flavor without starting with some coffee, as there are no other naturally occurring substances that capture this flavor. “Making coffee flavors challenges the flavor chemist because different levels of oils exist in the beans themselves,” Pollock explains. “In addition, different amounts of oils can be extracted, and coffee contains many reactive ingredients. Coffee flavor is temperature-dependent; freshly brewed coffee loses its impact within a minute of brewing.”

Adding furfurylmercaptan to coffee to stretch it would fit the intention of creating a supernormal stimuli. Interestingly its not to be more seductive but to be more economical. Like chocolate, coffee might be very symbolically significant to our culture because it resists forgery. #mythologies

Caramel. Applying heat to sucrose above its melting point catalyzes the reaction of caramelization. Sugar breakdown products create a mixture of aldehydes and ketones and, most importantly, furanones. These can be characterized as caramel-like, sweet, fruity, butterscotch, nutty or burnt, and are the backbone of the caramel flavor. “The decomposition of sucrose by heat is a challenge in a plant situation because it is difficult to control the reaction,” Pollock says. “It’s much easier to simulate caramel flavors by using compounded flavors.” Maltol, ethyl maltol and cyclotene are components commonly found in caramel flavors. Caramel candy’s flavor comes from heating and concentrating sugar and milk, so simulated caramel flavorings often are enhanced by added dairy notes. Caramelization occurs in baking and cereal manufacturing, and the product base can be enhanced by adding caramel-type flavors.

Wow, the inputs seem so cheap, but because its difficult to control the reaction at the large scale formulators often go artificial.

Fresh-fruit flavor can be achieved by blending juice with aromatics recovered from the rest of the fruit. Natural and synthetic flavors can be added to juice to boost flavor and reduce expense.

Good advice, press and then distill. This is very important for liqueur manufacturing. And then synthetic flavors make it go turbonormal stimulating.

Concentration via vacuum distillation separates solid matter from the aromatic substances. These can be partially recovered and added to the concentrate, but the finished product still will be deficient in top notes. Freeze concentration uses no heat, so the finished product’s profile is closer to real fresh fruit.

I tried to turn freeze concentration into a trend yeas ago because it is so cheap and easy on the small scale but no one bit.

Citrus fruits are made into essential oils because much of the characteristic odor is found in the peel’s oil. Citrus oils have a high percentage of terpenoid hydrocarbons. These carry smaller levels of oxygenated compounds such as alcohol, aldehydes, ketones and esters. These are responsible for the characteristic odor and flavor. The terpenes contribute an odor/flavor of their own, and a citrus oil with the terpenes removed will be flatter-tasting and lack freshness. Terpenes are typically removed because they will oxidize, resulting in lower flavor quality.

This is why I found this document. Interesting sensory descriptors of terpenes.

To develop a fruit flavor, flavor chemists start with what nature starts with: amyl, butyl and ethyl esters, organic acids, aldehydes, alcohols, ketones and lactones. These build, characterize and enhance fruit flavor. Some chemicals instantly conjure the image of the fruit they are meant to depict, such as amyl butyrate with its banana-like scent. Others, such as ethyl acetate, will suggest an overall unidentifiable fruit note that will enhance and round out the flavor. Green, fresh, earthy, overripe, cooked and floral notes all can be added for complexity.

Playing God. What a great rationalization in the beginning.

Organic acids occur naturally in fruits, giving them their distinguishing flavor and bite. The same flavor will deliver differently depending on the acid used to enhance it. While citric and malic are very close to each other chemically, their profile and sharpness in the mouth vary considerably, and each individual acid will enhance fruit differently. Citric acid enhances cherry and strawberry flavors, Pollock explains, and malic works with apple and pear. Blends of malic with tartaric are great for raspberry as the tartaric has a slight metallic aftertaste that fits with the seediness of a berry. The goal is stimulating other areas on the tongue. A subliminal amount of acidity, not specifically tart, can work well to add a different dimension. Phosphoric acid at less than 100 ppm, or acetic acid used at a level at which the scent isn’t noticed, are other atypical ways of using acidity.

This is great stuff and the descriptors are spatial. One problem with spatial descriptors like sharpness is that they are hard to make scaler with any concensus on meaning. I proposed to overcome that by using hypertext controls.

Grape typically has been associated with the use of malic and tartaric acids, according to Jim Lewis, director, flavor applications, Bush Boake Allen, Montvale, NJ. Today, citric acid is often used to enhance grape flavor, and many people have become accustomed to the different flavor that results. Because of this, some will perceive an off-note to grape enhanced by tartaric or malic acids.

We have been so warped by the works of flavor formulators that the artificial has become the norm and the natural seems off. #JorisKarlHysman #AgainstNature

Another option is using a nut flavor. “True and characteristic nut flavors can be developed from synthetic ingredients that not only convey a nutty characteristic,” Pollock explains, “but can simulate the specific nut, such as a filbert, hazelnut, cashew or pecan.” Many nuts contain allergens, so a great need exists for flavors that aren’t nut-based. Using only natural flavors restricts the flavor chemist’s compound options. A nutty character can be developed, but it won’t possess the unique nuances of the individual variety that can be found in the artificial flavors. Since these natural flavors require the use of actual nut extractives, it’s not easy to develop an all-natural flavor that is allergen-free.

Giving us allergies by saving us from allergies. Here the main category nutty is broken down into sub categories which are object comparisons.

Lets requote this:

A nutty character can be developed, but it won’t possess the unique nuances of the individual variety that can be found in the artificial flavors.

This refers to using natural non nut ingredients to synthesize the character of nuts. Kobs claims only artificial ingredients can push natural non nut ingredients into believable nut territory. I personally like artistic constraint and don’t feel the need to have nut named stuff when no nuts are present. This is a semiology issue, they are forcing a symbol on a sensation.

Spices. What would pumpkin pie be without the spiciness of cinnamon, ginger and cloves? Spices are defined as natural vegetable products used for flavoring, seasoning and imparting aroma to foods. Small quantities of spices add dimension to a food product, and their connotations of naturalness appeal to the consumer. However, spices vary in strength and flavor profile; their flavor is often less evenly distributed within the food matrix; they can represent a microbiological hazard; and they lose flavor strength upon storage. Occasionally, a large spice volume can make the food matrix muddied or speckled and bitter-tasting.

Connotations of naturalness… so what something symbolizes is important. #semiology

Often, an essential oil or extracted oleoresin is preferred. Essential oils help control flavor strength and character. They are microbe- and enzyme-free, and are stable under good storage. One drawback of the essential oil is that it only represents a portion of the total available flavor in a spice. The volatile oil of ginger won’t provide any of the pungent qualities because these qualities come from non-volatile components. Oleoresins contain the volatile and nonvolatile compounds from the spices, so their flavor is more characteristic of the spice. Oleoresins are thick, viscous liquids, making them difficult to incorporate into the food matrix evenly. They also are very concentrated, so weighing errors are dramatic.

A very interesting differentiation between an essential oil (only the volatile part) and an oleoresin (volatile and involatile). This fragmentary thinking is so much more important than people think.

Spices also may be found in the form of essences, emulsions and encapsulates, and plated onto sugar. Often, a blend of forms represents the perfect solution. In a cinnamon roll application, cinnamon essential oil will provide the flavor strength, while a dusting of ground cinnamon will give a quality, homemade appearance.

Homemade appearance. We’ve jumped from sensations to what something symbolizes.

Maltol and ethyl maltol can improve overall flavor, potentiate sweetness, increase the sensation of creaminess, mask bitterness and suppress an acid bite or burn. Marketed under the name VeltolÆ by Cultor Food Science, Ardsley, NY, these ingredients have a mild flavor and sweet caramel-like odor. While both compounds must be labeled as artificial flavors, the product line also includes product enhancers that can be labeled as natural flavor.

Potentiate sweetness here might be what I call olfactory-sweetness.

Licorice extracts, derived from the roots of the licorice plant Glycyrrhiza glabra, also possess flavor-potentiating properties.

More potentiating. What I’d love to know is if its an industry term or the authors personal term.

Going beyond the obvious can lead the developer into flavor areas that might sound unlikely, but the results speak for themselves. There’s no reason why a grape flavor can’t be enhanced by a less recognizable flavor such as melon or plum, which provides roundness and depth. Fantasy flavors, or flavors with no real characterizing base flavor, can come from all sorts of unlikely blends and can be great fun to the creative flavorist.

This is really great and it elludes to the power of the grotesque to be attractive and extraordinary.

“What the developer is doing is adding interesting notes,” says Reineccius, “and even though the product is sweet, the flavors don’t necessarily have to be. Odd items can contribute interesting notes – there’s really no limit. Garlic oil works nice in butterscotch because it provides a warm feeling, and chocolate often has been enhanced with low levels of fermented soy-based flavors.” Using 300 ppm of monosodium glutamate in maple syrup will help open up taste buds, and make the flavor come alive through this very viscous product, Pollock says.

Collage creative linkage. A plane is a fragment of the architecture of space -Hans Hoffman.

When 20 new flavors come in, it’s tempting to open the bottle, take a sniff, and make a decision. But flavors shouldn’t be screened in their pure state, as many of the notes will appear unbalanced or even unpalatable. The best screening method is trying a flavor in its final application. With a cake, bake a plain batter containing the flavors and evaluate to determine how they interact with other ingredients and heat. With time lines as short as they often are, and 30 flavors staring at you from the shelf, this may be unfeasible. The next best thing is to dilute the flavors in water, comparing them for quality, character and impact. Just as a sprinkle of sugar will tone down the bitterness of a slice of cinnamon toast, sweeteners make flavors come alive. This phenomenon is apparent when screening flavors. Diluting an almond extract in plain water will produce a slightly bitter and unpleasant liquid that would appear to contribute very little to the finished product. Adding sugar will accentuate its rich and fruity notes and bring out flavor more realistically. Many of the components of sweet flavors don’t have a very pleasant flavor on their own, so it’s important to screen sweet flavors with sweetened water. It also takes a great deal of imagination to recognize the capacity within a flavor.

This parallels my idea of making a series of sketches to get familiar with flavor fragments when making products like amaros or aromatized wines.

The way sweeteners interact with flavors and deliver to the human olfactory system is quite complex and almost totally unpredictable. When flavoring based on sweetness concentration, mildly sweetened products require the use of less flavor as the flavor comes through more clearly. At very high levels, sweetness becomes intense and begins masking the overall flavor. As a result, higher flavor levels are required.

When sweetness masks the overall flavor, I’ve called this cloying. Sweetness can be a aroma enhance to a point then it is an aroma distractor. Enhancement could be defined as lowering the threshold of perception.

The best method for developing products with balanced flavor is learning to speak the language of the flavorist, and to have them involved at the conceptual get-go. Don’t be afraid to answer their questions truthfully. The flavorist isn’t trying to steal your concept. Instead, he needs this information to provide the best product possible for a given application. How many hours, dollars and pounds of ingredients have been lost because a flavor didn’t act as predicted? Granted, there’s no guarantee changes won’t occur, but at least you’ll rest easier knowing you did everything possible to prevent it.

Does the flavorist actually have a language like aesthetic sensory language? or is she talking about business language and logistics of developing a formula?

It’s important for every food scientist to learn the language of flavor, because within every flavor category, a subset of many characterizing flavor descriptors exists. A fruity strawberry can be very unripe and green, very ripe, seedy tasting, or cooked so as to resemble preserves. It’s not enough to say one is seeking a chocolate flavor, because the terms tobacco, barny, fruity, musty, milky, woody, oily, green, hay-like and floral all have been used to characterize chocolate flavor. Telling the flavorist one is looking for a vanilla that is creamy, custardy, spicy, smoky, floral, caramellic, baby-powdery or fatty will save time by reducing the number of samples that need to be submitted and screened, resulting in shortened development time. Discussion can be promoted and expectations clarified by using food-item terminology, such as fruit punch, cough syrup, vanilla wafers or even brand names like Captain CrunchÆ cereal and Juicy FruitÆ gum.

So they think the have a language…

Developers and flavorists must have this list of vocabulary words, and agree on what flavor is being perceived. If one person describes a flavor as “hay-like” and the other person describes the same flavor as “barny,” then there should be a common word agreed upon so everyone knows this particular flavor will be described as such. This is not as easy at it might appear, as each individual has his own sensory strengths and abilities to communicate their reactions.

Agreeance is what I called Endoxa in my analysis of wine descriptors.

Granted this article is from 1998 and a lot has happened since in the industry, but it seems like there is tons of room to advance. The skills and ideas of the industrial flavor formulator are relevant to the cocktail creator or the micro distiller formulating new non traditional products.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

search previous next tag category expand menu location phone mail time cart zoom edit close