Friday, May 17, 2019

The Red Velvet Cake Effect

I recently stumbled upon a video this week where the speaker described experiments where the presence of color tricked participants into tasting flavors that weren't there. The idea seems preposterous. How can flavorless food coloring impart taste?

How Color Affects Taste, Prof. William Lidwell of University of Houston

As odd as this may seem, it must be true. I mean, I found another video on YouTube that makes the same general claim: "Soft drinks that are blue are considered to be more thirst quenching, whereas soft drinks that are pink are considered to be more sugary, even if they're not." There are not one, but two guys who are articulate, sound intelligent, and have the massive funding required to produce a YouTube video on this topic, so who am I to question their veracity?

The Taste of Color, Trace, for DNews

A little googling turned up similar results. For example, two manufacturers of color measurement devices have blog posts on the topic. Here is a quote from the blog post from Konica Minolta:

"In a study published in the Journal of Food Science, researchers found that people confused flavors when a drink did not have the appropriate color. A cherry-flavored drink manipulated to be orange in color was thought to taste like an orange drink, and a cherry drink manipulated to be green in color was thought to taste like lime."
How Color Affects Your Perception of Food

A blog post from HunterLab points to an important distinction between how color affects our expectation of taste and how color affects our perception when we taste. "[We] constantly evaluate foods based on their hue, from checking if the meat is still red to guessing an avocado is ripe when its skin becomes dark green."

The author goes on to say "Color is so powerful that [it] can override what our other senses are telling us to be true, causing us to taste sweetness that isn’t really there, experience flavors that aren’t present."
Examining the Science Behind Color Perception of Food Flavor and Quality

I'm going to invent a name for this phenomenon: The Red Velvet Cake Effect. The rich flavor of red velvet cake is is an olfactory illusion which is caused by the flavorless red food coloring. At least that's what my wife told me. Again, who am I to question?

Ok. Maybe that's not exactly true. The action of an acid (from vinegar or buttermilk) on natural cocoa imparts a red color to this cake, and this acid probably has an effect on the flavor. But pretty much all recipes for red velvet cake also include red food coloring. Why do this? The red food coloring in the cake tricks us into thinking that the cake is richer in flavor. Or maybe the cake really is richer in flavor?

These are not criticisms

Let me preface my next comments -- the comments about the aforementioned videos and blog posts. I don't mean for my comments to be criticism.

The first three of these reports are infotainment -- journalism targeted to inform and entertain. It is not expected that they go into details about how the experiments were performed. There is no talk of qualification of participants, a control group, accounting for the placebo effect, or the statistical relevance of the results. To go into such details would defeat the purpose of entertaining. Who wants to wade through boring details when all they want from the article is a factoid that they can share at the local tavern?

It's a little known fact that Cliff Clavin never actually said this.
It comes from a John the Math Guy blog post on vermillion.

Another shortcoming of the preceding infotainment  (with no criticism implied) is that none of the first three give specific reference to the experiments. Again, I am not being critical, but the articles don't offer a lot of help to the person who is a bit more than just curious. How can they find the technical paper on the topic?

Let me make this clear. I am not criticizing the first three. Infotainment is a great thing, and these are great examples of infotainment. But let's take them for what they are. Suppose you are president of the American Broccoli Growers Association, and I need to decide whether to fund research for a plant geneticist who wants to adjust the cruciferous color so as to make kids go wild about the taste of broccoli. You really need to research beyond the infotainment articles.

I commend the HunterLab article for moving from infotainment into the realm of edutainment. The article educates as well as entertains. There is a clear explanation of what experiments were performed and how they can be interpreted. And the technical papers and experts are identified so an interested reader can go out to find additional details.

But... I wonder if is there some selection bias in the choice of research papers that have been cited? Do other experiments confirm these conclusions? In true John the Math Guy style, I have done a borderline-obsessive amount of digging to get to the true flavor of this question, not colored by any preconceived notions.

Does the color of the food itself change the taste?

I followed the HunterLab link to an actual research paper from 1980. Here is what a quote from that paper:

"Results showed that color masking dramatically decreased flavor identification of fruit-flavored beverages, while atypical colors induced incorrect flavor responses that were characteristically associated with the atypical color. In addition, the color level of beverages had significant effects on their overall acceptability, acceptability of color and of flavor, as well as on flavor intensity."
Effects of Colorants and Flavorants on Identification, Perceived Flavor Intensity, and Hedonic Quality of Fruit‐Flavored Beverages and Cake

Only with a blindfold could Sherlock tell that all the glasses contained prune juice

Here is another research paper demonstrating that the taste of sweet beverages can be "flavored" by color.

"The results of the present experiment corroborate the findings of the previous experiments, demonstrating the influence of color on taste perception"
"... subjects in the present study were moved by the color stimuli to completely misjudge the flavor of the substance being tasted (calling the birch beer such things as "cherry soda" or "cream soda")."
The influence of color on the taste perception of carbonated water preparations

And yet another...
"Both the colour of the cider itself and the colour of the label significantly influenced perceived flavour and hedonic response to the ciders."
Cross-modal influence of colour from product and packaging alters perceived flavour of cider

Oh! I gotta get me some of that Hedonic Response Cider!

The only fruit-flavored beverage that I drink is wine, so I kinda don't care. The HunterLab article anticipated my predilections so they also mentioned a blog post of theirs that talked about color and wine. That second blog post summarized an experiment where researchers tried to trick professional wine tasters by adding flavorless red food coloring to white wines. The deceit was successful.

"A white wine artificially colored red with an odorless dye was olfactory described as a red wine by a panel of 54 tasters. Hence, because of the visual information, the tasters discounted the olfactory information."
The color of odors

My wife's favorite whine is "I want to go to Miami!"

It is a little known fact that olfactory means "relating to the sense of smell."

This was a bit of a disappointment to me, since I know that, all scientific research aside, I prefer red wines! I probably don't like Riesling, but I'm not sure. I've never riesled.

As you may expect, beer is also important to me. And I know that dark beer is way more better than a an icky sickly pale yellow pilsner. But if two beers differ only in color, does my eye convince my tongue and nose that the beer that is richer in color is also richer in flavor? I can't count the number of times that I have lain awake at night pondering that question!

Interestingly, experiments with beer haven't been as conclusive. The quote below is a bit complicated, but here is what I think it says: If someone is a Miller Lite kinda guy, they are more apt to be fooled by a little brown food coloring. I believe this, but only because it fits my preconceived notion that people who drink light beer are less sophisticated than I am.

"When the participants evaluated the expectations and tasting experience of the two different beers ... (pale vs. dark), after tasting, those who preferred pale beers, rated the darker beer as tasting sweeter than those who usually prefer other types of beers, such as dark ones..."
Dark vs light drinks The influence of visual appearance on the consumer's experience of beer

The Lovibond scale is used to assess beer color

One finding of the next experiment (below), is that before the mug comes to my lips, I expect that a beer with a rich brown color will be richer in flavor than a light yellow beer. Well, duh. The important part of this research is that when I actually taste the beer, my palate will not be fooled.

"Dark and pale beers were evaluated both before and after tasting. Importantly, these beers were indistinguishable in terms of their taste/flavor when tasted without any visual cues. The results indicate that the differing visual appearance of the beers led to clear differences in expected taste/flavor. However, after tasting, no differences in flavor ratings were observed, indicating that the expectations based on visual cues did not influence the actual tasting experience."
The Influence of Color on the Consumer's Experience of Beer

So, let's move on to chocolate, another place where I like to brag about my superior tastes. In my opinion, milk chocolate is for the hoi polloi. Having an affinity for dark chocolate shows that you have culture. But, at least according to this research, my false snobbery can be exposed with a little brown food coloring in the outer candy shell of an M&M. And even worse, just calling  the little rabbit pellets dark chocolate will fool me. Really? Am I that easily misled!?!?!? 

"The participants rated brown M&Ms as being significantly more chocolatey than green M&Ms and “dark chocolate”-labeled M&Ms as being significantly more chocolatey than “milk chocolate”-labeled ones."
The Influence of Color and Label Information on Flavor Perception

I like my chocolate just like my used motor oil - dark and flavorful

This sampling of research papers suggests that color can suggest taste, but that might not always be the case.

Does the color of the plate or cup influence the flavor?

I want to investigate a phrase from the cider experiments which I skipped over: "and the colour of the label". The cider house rules apparently extend beyond the color of what goes into your mouth. The effect on flavor of the color of the coffee mug seems to be fairly well researched. And the research seems to be heavily weighted toward verification of the Red Velvet Cake Effect. I have some quotes below.

"The results revealed that the colour of the cup exerted a significant influence on both pre- and post-tasting ratings for all attributes measured."
Cup colour influences consumers' expectations and experience on tasting specialty coffee

"... the coffee was rated as less sweet in the white mug as compared to the transparent and blue mugs."
Does the colour of the mug influence the taste of the coffee?

"The colour of the cup, for instance, has been shown to prime notions of sweetness (e.g., pink cup) or acidity (e.g., yellow or green cup) that may carry over to influence the tasting experience."

"Given that different styles/varieties of specialty coffee have different dominant/desirable qualities (e.g., acidity/sweetness), in the future, the design of coffee cups may need to be customized for different coffee drinking experiences (e.g., origin or roast), much as seen in the world of fine wine (with different glasses for different grape varieties)."
Assessing the influence of the coffee cup on the multisensory tasting experience

Choose your cup wisely, it will affect the taste of the coffee

Seems pretty conclusive that the Red Velvet Cake Effect extends to coffee mugs. But are pastries on plates any different from red velvet chocolate? Apparently so.

"[No] main effects of the plate colour on the evaluations of greasiness, crunchiness, creaminess, and sweetness [of pastries], as well as the hedonic value and purchase intent in stage 1 and stage 2 could be found."
Visual merchandising of pastries in foodscapes: The influence of plate colours on consumers’ flavour expectations and perceptions

But the color of the plate has a clear effect on the sweetness of strawberry mousse.

"Specifically, we investigated the influence of the color (black or white) and shape of the plate on the perception of flavor intensity, sweetness, quality, and liking for identical strawberry mousse desserts.The results demonstrated that while the color of the plate exerted a significant influence on people’s perception of the food, the shape of the plate did not. In particular, when the mousse was served from a white plate, it was perceived as significantly more intense and sweeter, and was also liked more."
Is it the plate or is it the food? Assessing the influence of the color (black or white) and shape of the plate on the perception of the food placed on it

Choose the plate wisely if the dessert is red velvet cake or strawberry mousse,
leave the ugly plates for pastries

I'm not sure what to make of all this. I am getting a bit peckish for dessert, though.

Survey papers

This limited research has led to some conflicting conclusions. I have browsed through a dozen papers and found that most (but not all) confirmed the Red Velvet Cake Effect. I readily admit that my sampling of technical journals on flavor science has not been thorough. Basically, I have only earned my associates degree in Velvet Cake from Google University. In particular, I would expect that my research would be biased in favor of papers that support the surprising results.

It is a little known fact that red food coloring might be made from Hemipterates (the order of true bugs, in the class Insecta) like the cochineal or kermes vermilio, but that Starbucks has pledged to stop using the natural dyes that are made from insects.

To get a better perspective, I will have a look at scholarly review papers.

Charles Spence (from Oxford) is a prolific author on the Red Velvet Cake Effect. (I say this despite the fact that he has not, to my knowledge, used this phrase. I am sure he will start using it when he reads this blog post.) Here are a few of his papers which provide an overview of a little bit of the research that he has reviewed. When I say "a little bit", I mean... well... the first paper has about three pages of references. The second one has 170 references. I should live long enough to read that many research papers.

Does Food Color Influence Taste and Flavor Perception in Humans
On the psychological impact of food colour
Background colour & its impact on food perception & behaviour

Spence (In his paper "Does Food Color...") makes the following statement:

"Does food coloring influence taste and flavor perception in humans? Although researchers have been investigating this important (both on a theoretical and practical level) question for more than 70 years now (see Duncker 1939; Masurovsky 1939; Moir 1936 for early research), an unequivocal answer to the question has not, as yet, been reached."

Ahhh... that explains the confusing results that I found! Then he goes on to burst my bubble.

"That, at least, would seem to be the conclusion drawn by the majority of researchers in the field."

Slicing the onion thinner

When I asked her for the color of the car, she said it was the color of an onion

Anyone who has passed an intermediate level course in "how to lie with statistics" will recognize that the answer you get depends a great deal on the precise phrasing of the question. There are a number of different but related questions that a Red Velvet Cake Effect experiment could address:

1) Does the color of the food influence my expectation of taste before I sample the food?

2) Does the color of the food influence whether I can correctly identify a given food?

3) Does the color of the food influence my assessment of the intensity of the taste of the food (the gustatory effect)?

4) Does the color of the food influence my assessment of the intensity of the flavor of the food (the olfactory effect)?

Wait... Aren't 3) and 4) the same question? Maybe to you and I taste and flavor mean the same thing, but to people who are really into the science of taste and smell, there is a distinction. Taste is the thing that we do with our tongue when we are not sticking out at people we dislike. We can distinguish only five distinct tastes with our tongue: salty, sweet, sour, bitter, and umami (which means something like savory).

I never developed much of a taste for relativity

Flavor, on the other hand, is detected in the nose. To make things even more complicated, this olfactory (smell) component might be administered orthonasally (though the nose, as when someone pretentiously sticks their nose in a wine glass to detect the nose of the wine), or it may be administered retronasally (through the mouth, like when you take a sip of the wine).

The nose of a wine is best sensed in near silhouette conditions
with clouds and mountains in the background.
Pretentiousness is accentuated by an "I forgot to shave for the past four days" beard.  

As an aside, a true oenophile (which is French for priggish wine snob) will tell you that the best way to tell the nose of a wine is to take a sip, spit it out, and then exhale through your nose. This is an example of retronasal olfaction. A true Wisconsinite like myself would be aghast at the thought of spitting out something that contains alcohol.

Now, how did Spence answer the four questions?

Spence summarized one experiment having to do with the first question. British and Taiwanese participants were asked what they expected a colored beverage to taste like. The Brits overwhelmingly (14 out of 20) expected a brown beverage to taste like cola. None of the 15 Taiwanese participants expected a cola taste - grape, mulberry, and cranberry were their choices. So, the answer to the first question is "yes", but the expectation is culturally dependent.

According to Spence, the answer to the second question is unequivocal. Color influences identification of foods.

According to Spence, the experiments that address the third question do not give consistent results. According to me, the combinations (which colors have an effect on which of the four tastes) is confusing.

According to Spence, the answer to the fourth question is clear cut. Color does flavor our sense of flavor.

Charles Spence, Food Scientist

So, I conclude that the two videos and two blog posts mentioned way back at the start of this blog post are pretty decent in their portrayal of the science behind the Red Velvet Cake Effect. They didn't quite catch all the nuances and slightly contradictory results, but we can't expect everyone to produce blog posts that are as long and boring as this one!

If you have read up to this point, I leave you with a well deserved factoid: The flavor and taste of red velvet cake is most likely due (at least in part) to the red food coloring, which might be manufactured from bugs.

Updated June 3, 2019 - In my haste to pat myself on the back for knowing that Umami has moderately recently been acknowledged as a member of the Taste Bud Band, I neglected another key member of the band, Salty Dog. I have corrected the blog accordingly. Thanks to Robin Myers for pointing this out to me.