Short Tales of Olfactory Illusions

Recently I came across a wonderful paper from the journal of Consciousness and Cognition called Olfactory Illusions: Where are they? by Richard J. Stevenson. The paper is a great review of the literature on the topic and examines the existence of olfactory illusions. The author writes in a clear, articulate style which I thinks makes exploring the ideas accessible to culinary professionals that are not used to reading scholarly journals. I recommend getting a copy and checking it out.

Richard J. Stevenson it turns out is the author of two books I’ve had in my amazon wish list for a while: The psychology of flavour and Learning to smell: Olfactory perception from neurobiology to behavior which is written with Donald A. Wilson. The amazon.com descriptions makes Wilson and Stevenson’s work sound particularly interesting.

From Learning to smell:

“Donald A. Wilson and Richard J. Stevenson address the fundamental question of how we navigate through a world of chemical encounters and provide a compelling alternative to the ‘reception-centric’ view of olfaction.”

Without having read the text, it sounds like their view tries to explain the significant role of recollection at completing aromas which is something I’ve been starting to suspect more and more over the years.

With the proliferation of modernist cooking it really comes as a surprise that there is little interest and/or no awareness of olfactory illusions. They seem like fertile ground for extraordinary creative experiences. It may help to kick things off by detailing a few of the olfactory illusions I’ve found in the wild.

One of the first illusions I had ever come across (and have never had the resources to reproduce) was an experience of eating meatloaf at staff meal with a group at restaurant Dante, probably seven or eight years ago now. I was half way through my meal and someone approached us chewing mint gum. The aroma of the chewer’s mint leaped from their mouth into mine. Frontal-olfaction became retro-olfaction and the location of the aroma changed dramatically, but that is not the end of the story.

First I inquired if it happened to anyone else at the table. Yes, to everyone, and because the experience was not harmonic we sent the chewer away (dissonance makes illusions easier to spot). No one but me seemed intrigued by the experience which is worth noting. I then went to ask chef how exactly he made the meatloaf. He confessed to adding some mint water left over from another dish (lamb?) into the meatloaf. I really feel like the mint aroma in the meatloaf was below the recognition threshold. Somehow the gum chewer awakened it.

How exactly to classify this illusion I’m not sure. I always thought it would be interesting to duplicate it in a beautiful context. Perhaps the same course could be served twice but for the second course a candle or a lamp-warmed aroma oil could be added to the table that awakened a dormant aroma in the food. These illusions can risk being inharmonic so cocktails which are often squarely in the realm of acquired tastes might be perfect for inducing such an experience.

Another illusion I had come across back in 2009 I dubbed the “Maraschino Blackberry Illusion”. I had wanted to create a cocktail garnish akin to a maraschino cherry but with another fruit. Cherries are unique because alcohol can be used to pull the aroma of the pit into the meat of the cherry creating juxtaposition. For a blackberry I would have to use another aroma source to provide the juxtaposition. I chose the spice Mace. I distilled a blackberry-mace eau-de-vie which the eventual blackberries were macerated in and brought to equilibrium with. Once at equilibrium, the eau-de-vie could be consumed alone and found harmonic but when eating a blackberry the mace aroma was intensified so significantly that you had to spit the berry out. The haptic heft of the fruit changed the threshold of perception so significantly you could not consume it. Again, in this case I suspect dissonance is important to spotting the illusion.

Years later, for a recipe in my distilling book, I re-did the illusion in a beautiful context with Fernet Branca aromatized cherries. Fernet is added to the cherry eau-de-vie at about 10% scaling, but when you eat a cherry the Fernet derived aroma qualities seem as if you are consuming 100% Fernet. The haptic heft also changes the threshold of perception of the alcohol and the the 18% or 20% alcohol cherry easily feels 40%.

Using newly developed reflux de-aeration techniques to prevent citrus juices from oxidizing, I was hoping to re-do the illusion again by placing a classic Side Car cocktail in the body of a golden raspberry (the fruit would have its color and aroma leached out as best possible by a vodka/sugar blend). The familiar cocktail and the cocktail filled raspberry (hopefully same color) would be consumed side by side to illustrate the changes. Haptic influences on the threshold of perception of alcoholic means that a generic triple-sec, which is low alcohol, might have to be used instead of the usual 40% premium stuff.

This same illusion can be seen simply when juicing fruit and was illustrated to me so many years ago by a pastry chef I worked with. Eat a strawberry and it is lively and natural. Juice the strawberry and the flavor (probably mainly the aroma) goes flat. To perk the aroma back up, now that texture has been simplified, you have to add flavor enhancers like sugar and acid to bring the fruit back to its original lively-ness. I think this impacts fruits of low sugar contents the most hence we enjoy unadulterated orange juice which has a high brix/acid ratio but not raspberry juice which has a low brix/acid ratio.

I noticed a very simple illusion when playing with Marmite. I was using Marmite to explore yeast autolysis in distillates for my book project. I was making silly stuff like Marmite aromatized rye whiskey where a commercial rye underwent simple re-distillation with a portion of Marmite. In the principle aroma categorization scheme I use, where I categorize aromas in terms of gustation, Marmite is olfactory-umami (a grounded cross modal metaphor). I smelt Marmite countless times and drank a lot of the whiskey without ever tasting Marmite alone. It always smelt olfactory-umami. I finally tasted the stuff and it quickly became olfactory-salty in a way that I could not shake for quite a while. The olfactory-umami was gone. I suspect that nutritional reward systems can have a big influence on these sudden categorical changes.

The loosest aroma illusion I have been able to generate which increases my suspicions that nutritional reward is very important to aroma illusions happened when working in my small plastic foundry. I make a few small plastic parts (champagne bottle manifold & door hardware). Anyhow, I’m drilling holes in red plastic and I start smelling cherries. When I start drilling holes in black plastic I start smelling licorice. The plastics I work with have phthalates in them which themselves produce the illusion of sweetness. The generic olfactory-sweet smell of phthalates plus color is enough to produce vivid olfactory hallucinations. The interesting context here might make for a fruitful experiment to see how many people can notice the illusion. The experience was harmonic but I still was able to notice it.

Plastics that contain Phthalates are considered dangerous to infants and young children because the illusion of nutritional value (exacerbated by harmonic olfactory illusion) makes the material more likely to be chewed on and the toxic compounds ingested. Better understanding the illusion could help make safer plastic products.

The plastic example generates an olfactory illusion from such little other stimuli and really illustrates how perception can be seen as the meeting point and synthesis of incoming sensory stimulus and outgoing recollection (completion). The meeting point is sort of like Aldous Huxley’s “doors of perception”. This door, which connects the two hallways, slides around all the time. I suspect that the presence of nutritionally valuable features makes perception recollection, and therefore completion heavy. We might be more susceptible to aroma illusions when eating, and because of the co-experience with other more salient flavor features, they are really difficult to notice. Again, the harmonic nature of many illusions may make them harder to notice.

The two congeners ethyl acetate and acetaldehyde, which are significant to most distilled spirits, I suspect work akin the the phthalates in my plastic experience. The master cognac distiller, Robert Leaute, describes these compounds as “aroma fixatives” but he does not elaborate on the term. Distillers try to keep these compounds as close to the recognition threshold as possible without going over. When above the absolute threshold but under the recognition threshold, I suspect they can turn on “nutrition mode” where perception is recollection-completion heavy and allows us to have very subjective flavor experiences with distilled spirits. All the mass-spectroscopy and chromatography data does not seem to align itself with the tasting notes. The realm of distillates and wine might be the realm of olfactory illusions. Acknowledging olfactory illusion might give us a completely different perspective by which to build spirits and wine.

If wine is the realm of illusion, it would drastically change the notions of terroir and sense of place. But I suspect the boundaries of subjectivity is narrow and can be predicted. To start, we have to understand more of how we detect contrast in olfaction. For example, we can detect contrast between blue and red with purely arbitrary linguistic categories like “blue” and “red” or use a non-linguistic frames of mind to creative alternative categories such as warm & cool which is color in terms of thermoception. We detect lots of contrast non-linguistically then translate it to named linguistic categories and then we even build layers of metaphor on top of that. And we never notice and never expand the categories (if aromas can be sweet they can also be acidic, bitter, salty, or umami). I suspect when “wet cobble stone” is used as a wine descriptor, the imbiber is detecting contrast by perceiving the aroma as a shape and the shape is akin to the surface of a wet cobble stone. I also suspect wet cobble stone can be put into the olfactory-umami category. “Acrid” is another layered metaphor. The word euphonically has an angular shape and that angular shape is a metaphor for the aroma (perhaps also to be categorized as olfactory-acid). A really fun book that illustrates a word’s euphonic convergence with its meaning is Euphonics: A poet’s dictionary of sounds.

If wine is the realm of illusion (which we should embrace), in that famous study where experienced tasters used red wine tasting terms for white wine dyed red they were perceiving normally. We may have to rethink how we abstract color in wine and whiskeys.

An illusion that I cannot believe is not better recognized is tasting certain stimulus followed by others. Such as bitter espresso after tasting something sweet (dissonance!). I call these nutritional preference comparisons and suspect that our mind often constructs reality in a way that can reinforce nutritional preference. This illusion has profound influence on food and wine interaction, but no one seems to use it beyond the adage that “a dessert wine always has to be sweeter that the dessert”. There are many layers to contrast detection in food & wine interactions but I suspect nutritional preference comparisons are a large component.

The coolest illusion I have been able to generate happens by passive training with aromatized hand sanitizers. I made a series of aromatized hand sanitizers like wormwood to teach myself to better recognize aromas and practice making distillates. They could also be used as a bizarro garnish where the application to the hands of the imbiber could add frontal olfactory top notes and push a simple drink deeper into super stimuli territory. Anyhow, I go to a grand wine tasting and two wines I’m familiar with exhibit an aroma 10x normal. Not the aroma of wormwood but one of the same olfactory-bitter category. I bet you could say the aroma would be below the recognition threshold normally. I probably tasted 30 other wines and nothing else was out of the ordinary. One of the two wines was an Anjou blanc and the other was a red “Cerasuolo” from Cos in Siciliy. The aroma I experienced was more akin to yarrow flower which is something I also have experience with, but not much relative to wormwood. I don’t know exactly how to categorize the illusion because, yes, I had significant reinforcement of a specific category but the aroma experienced in the wine was not exactly the same as wormwood. I remember Gordon Shepherd referencing a similar phenomenon in Neurogastronomy but I haven’t been able to revisit the text.

A related phenomenon happens with the aroma of dead mice. Over the years I have taken pest control at the restaurant into my own hands and developed the ability to smell a dead mouse in a trap at thresholds far below anyone else. I wonder whether all aromas have the same potential to change like this or if it is related to other phenomenons like long sensory after images which is in turn related to reward or warning systems. If the reality of an expert can be so unique it could have strange implications for spirits or wine blending. I’ve seen this hinted at in very advanced texts on whiskey making.

A most recent illusory experience came when drinking a very pale rosé that I’ve been spending a lot of time with. As I was drinking the wine, I observed a mixed bowl of nuts and raisins a co-worker was snacking on and the aroma of raisin entered the wine (the raisin itself was probably not significantly aromatic). It seems like you could never be confident enough to call out this illusion but raisin in wine for people like myself is inharmonic. Raisin is a flaw which symbolizes regret and missed opportunities. The unique category I have for raisin might have made the illusion more noticeable.

The mint gum illusion and the subtlety of the rosé-raisin illusion makes me wonder if there are parallels in wine and food interactions. Even without consuming the food, very aromatic elements from a dish perceived by frontal olfaction might be able to change contrast detection in a wine, or somehow change location from outside the head to inside the mouth juxtaposing the aroma with the wine. In the past I had categorized interactions like this as: wine pairing by “sillage” (a perfumer’s word for the scent trail). If these illusions could be better proven they could validate some recently popular ideas of pairing food & wine by similarity of aroma compounds. My suspicion is that the illusion is not generated all the time and a few instances have led sommeliers to over apply the model with no payoff.

Hopefully these illusions will help generate awareness of the phenomenon. They will either give people the confidence and a template for calling out the olfactory illusions they have experienced or inspire others to harness any illusions they can find in a beautiful context. This could be a new fun direction for modern cuisine to go.