SIX

We Are the Anti-Pandas

THE BEST THING ABOUT BEING A PARENT OF YOUNG children is that sometimes they forget that you’re there. In their minds, the back seat of the car is a private space where they can talk to their friends without fear of parental eavesdropping.

NATALIE (age eight): What’s your favorite food? Pickles are mine! I love pickles!

JACOB (her twin brother): Pickles are disgusting, Natalie. My favorite food is French toast.

NATALIE: I like French toast too, but not as much as pickles. Pickles are the best!

SARAH (Natalie’s friend): French toast is gross. It smells like eggy farts! I like mangoes.

JACOB: Yuck! Mangoes are slimy!

NATALIE: No they aren’t. You’re stupid, Jacob.

JACOB: No, you’re the stupid one for liking stupid, salty pickles!

You might imagine that this line of conversation could only be sustained for so long. After all, there’s no moral point here. People just like different foods. And you’re never going to convince someone to change their liking or disliking of a particular food through argument or insult. Still, a back-seat conversation about food preferences could easily go on for thirty minutes, all the way home from school, regularly punctuated with delighted howls of disgust.

Many years later, when I found myself looking for a match on OkCupid, I was surprised to see that, seeking to convey her unique qualities, nearly every woman on the site spent quite a few words on her food preferences.¹ I understand why people turn to food as a topic when trying to paint a picture of themselves as individuals. Everyone has a distinct set of food preferences and they are easy to articulate. Nonetheless I remember thinking, “Hey, CharmCitySweetie! You like spicy food and hoppy beer but despise mayonnaise, mustard, and runny eggs. OK. But really, so what? Are we truly doomed as a couple if you like Stilton and I prefer cheddar?” As it is inclined to do, my mind began to wander and I started to imagine dating websites for other critters:

HOTGIANTPANDA4U: I like to eat bamboo shoots. Nothing else, really. Just bamboo shoots.

SICHUANPANDAMONIUMGAL: Same here. You wanna eat bamboo shoots and chill?

When it comes to food, humans are the anti-pandas. Giant pandas inhabit a small ecological niche—the cloud forests of southwest China—and they eat only bamboo. Humans have spread all over the globe, from the polar regions to the tropics, and so, as a species, we have come to eat lots of different plants and animals. We have succeeded by being food generalists. As a species, we can’t be overly predetermined when it comes to food. We must adapt to local availability through learning, giving rise to lots of individual variation. This is the main reason why food preferences are considered such a mark of human individuality.

We can see this idea come out in everyday language, where the phrase “to have good taste” means to have enviable likes and dislikes, including things that never go in the mouth—clothes, music, books, or what have you. Taste has come to mean individual preference considered generally and not just related to food. This meaning is not just a peculiarity of English. In Spanish, for example, the verb gustar, which means “to like,” derives from the Latin root gustare, meaning “to taste,” which also gives us the English words gustatory (relating to the sense of taste) and gusto (with an appetite; enthusiasm).

To understand how the food preferences of individuals become so varied, we’ll need to explore the neurobiological basis of flavor sensation. In everyday speech, when we say that something tastes good, we’re not just referring to the five basic tastes detected by sensors on the tongue (sweet, salty, sour, bitter, and umami), but rather to a unified sense of flavor that blends smell, taste, and touch and refers the sensation to the mouth. When we say “taste” we mean “flavor,” and we tend to use the two terms interchangeably.² Here, to avoid confusion, I’ll use the word taste in the narrow sense to mean only the five basic tastes, and I’ll use the word flavor to mean the blended multisensory experience produced by putting food in the mouth.

Most people who go to a doctor complaining that they have lost their sense of taste—which can result from head trauma, side effects of a drug, an infection, or a few other reasons—have actually lost their sense of smell.³ If you test them by dropping salty or sour or sweet solutions directly onto their tongue, they perceive these sensations normally, proving that their sense of taste is intact. When you think about it for a moment, this is extraordinary. You would never go to the doctor saying that you couldn’t hear and be told that, actually, you were blind. You would never report that you couldn’t feel your legs and be told that, really, you were deaf—that you had just confused one sense for another. The fact that deficits in smell are often attributed to the sense of taste underscores the unusual degree to which these senses are combined in our experience.

I know it’s obvious, but it bears mentioning anyway. By the time you are tasting something, you’ve probably already investigated its odor through sniffing and have decided to put it in your mouth. The taste sensors on your tongue are there to help you make the further decision: Should I ingest this item or spit it out? In some cases, this is a life-and-death decision. Not surprisingly, the need to make it is found in nearly all animals, and so taste sensors are evolutionarily ancient, probably originating about five hundred million years ago. The modern sea anemone is similar to some of the very earliest animals with neurons. It has no real mouth to speak of, and no brain, yet it can detect bitter substances that have entered its simple closed-end digestive system and then, using its rudimentary nervous system, send commands to contract the relevant muscles and barf it up. In a way, this ancient behavior is the deep evolutionary origin of our cross-cultural human “yuck face,” which includes a tongue thrust to eject unwanted food from the mouth.⁴

In humans, the taste-sensing cells are organized into about ten thousand taste buds, clustered into visible bumps called papillae, which are spread across the tongue. The taste buds also reside in the soft palate and the portion of the upper airway called the pharynx, but are not clustered into papillae in these smooth tissues. Each bud is a basket-shaped cluster of fifty to one hundred cells with a tiny pore at the top, and each individual cell within the taste bud is dedicated to one of the five basic tastes.⁵ In this way, there are not separate taste buds for, say, sour or bitter, but rather each taste bud has separate detectors for each of the five basic tastes. Crucially, the electrical signals produced when the individual cells are activated by food or drink appear to be kept mostly separate as they are conveyed to the brain.⁶ The sensors on the cell surface that bind to taste molecules are proteins, and so their expression is directed by genes. At present, biologists have identified twenty-five human bitter sensors, two sweet sensors, one salt sensor, one sour sensor, and two umami sensors.⁷

Each of these sensors has a particular job in terms of evaluating food and making the decision to swallow or spit it out. Most bitter-tasting chemicals are made by plants and many, like the caffeine in coffee beans or the isothiocyanates in broccoli, are toxins, designed to protect the plants from bacterial or fungal infections or from predators, mostly insects. To a certain degree, when we taste bitter plants, we’re eavesdropping on a conversation that has little to do with us: we’re the bystanders in an ongoing chemical war between plants and insects.

Other bitter compounds are produced by bacteria. As a result, for most animals, bitter taste indicates either plant toxins or bacterial infection, and so will trigger food rejection.⁸ This is an inborn trait. A newborn baby will reject bitter foods with a tongue-thrusting yuck face the very first time they are encountered. No learning is required. Sour taste is also mostly aversive. A little bit of sour can be nice, but strong sour tastes tend to indicate either fermentation, as in sour milk, or hard-to-digest foods, like unripe fruit. Like bitter taste, newborns arrive with sour taste aversion built in.

Sweet taste is the opposite. We are born already finding sweet taste pleasant. If sugar is placed on a latex nipple, a baby will suck longer and harder than if the nipple is coated with plain water. Sweet taste comes from both natural sugars and, to a lesser degree, from the sugars that are released in the mouth as foods containing carbohydrates are chewed and partially broken down by the enzymes in saliva. Throughout human evolution, it has been mostly adaptive for humans to enjoy and consume sweet and carbohydrate-laden, calorie-dense foods, including breast milk. So it makes sense for sweet taste to be hardwired for pleasure.

Umami is primarily the flavor of the amino acid L-glutamate, which is found in many foods with a “meaty” taste, including beef broth, some fish, mushrooms, parmesan cheese, and tomatoes, as well as many fermented products like soy sauce, miso, and fish sauce. Breast milk also has a lot of umami flavor—about as much as beef broth—and this is likely to be at least part of the reason that we are born with a built-in liking for umami.

Salty taste is slightly more complicated. We are born finding salty foods pleasant, but only up to a point. Both babies and adults enjoy salty foods, but the experience of them becomes unpleasant at very high concentrations. This makes biological sense. We need to keep the sodium concentration in our bodies within a fairly narrow range. Either insufficient dietary salt or too much salt can cause trouble in several organ systems, including the nervous system. This problem of optimal salt consumption appears to have been solved in part by having two different populations of taste receptor cells, one that is activated only by low salt and is connected to brain regions that evoke pleasure, and another that is activated only by much higher salt concentrations and is linked to brain regions that produce aversion. At present, the low-salt sensor, called ENaC, has been identified but the molecular identity of the high-salt sensor remains a mystery.⁹

Here’s a question: Why are there at least twenty-five human sensors dedicated to bitter taste but only one for sour taste? It’s as if nature is trying to tell us something with the math. The reason is that all sour taste comes from the same simple chemical, H⁺, also known as a free proton. A lemon may have a different flavor than vinegar because of additional chemicals in each food, most of which are detected by smelling, but the thing that makes both vinegar and lemons sour are free protons (molecules that donate free protons are called acids). When sour taste receptor cells are built to detect free protons and nothing else, they don’t need many types of sensor to do it.¹⁰ Likewise, all salty taste comes from sodium ions, almost all umami taste from L-glutamate (and a few structurally similar molecules like L-aspartate), and all sweet taste from a small number of sugar molecules (fructose, glucose, sucrose, etc.) that share a similar chemical structure. So sweet, salty, and umami taste can also function well with one or two receptors each.

Bitter taste, on the other hand, can originate from thousands of different bitter chemicals that are not structurally related. As a result, there’s one sensor called T2R38 that appears to be particularly well suited to detect the bitter chemicals produced by certain classes of bacteria, as well as bitter chemicals, called glucosinolates, found in cruciferous vegetables like broccoli and brussels sprouts.¹¹ Another bitter sensor, T2R1, is one of several tuned to detect, among other things, chemicals called isohumulones that give hop flowers their bitter taste, to the delight of IPA beer drinkers everywhere (including CharmCitySweetie). Some bitter sensors respond to a broad variety of chemicals and others to a single compound. But the overall theme is clear: we need many bitter receptors in order to detect a large and chemically diverse range of bitter substances that we should avoid.

When a particular taste sensor is no longer required in the life of a species due to a change in diet, then the gene that encodes that sensor can accumulate mutations; eventually, some of those mutations will break the gene entirely such that no functional protein is produced. These broken genes are called “pseudogenes.” If we look at certain carnivores—like house cats, lions, tigers, vampire bats, and western clawed frogs—they are unable to sense sweet taste; their sweet sensor gene, T1R2, has become a pseudogene, shot full of mutational holes. On the other end of the dietary scale, some herbivores, like the giant panda, do not encounter umami in their strict bamboo diet, and so have lost their ability to taste it. Again, we can see the giant panda umami sensor pseudogene T1R1 lying in the genome like a rusted-out car, up on the blocks.

Perhaps the strangest case of lost taste sensation is found in whales and dolphins, who evolved from a plant-eating terrestrial ancestor about fifty million years ago, becoming carnivorous aquatic mammals. These marine mammals have not only lost their ability to sense sweet taste but they have also lost sour, bitter, and umami.¹² At first look, this is surprising, as whales and dolphins might still need bitter and umami sensors to avoid toxins and enjoy flesh. One hypothesis to explain this limited gustatory repertoire is that because dolphins and whales swallow their prey whole, they do not need to make a decision about which foods get ingested and which are spat out. When everything goes down the hatch, there’s no need for taste sensors to inform such a choice. Interestingly, not all marine mammals have lost their taste sensors. Manatees, which are plant eaters, have kept their functional sweet and bitter sensors, reinforcing the idea that the retention of basic tastes across species is determined by diet.

IN ORDER TO MAKE sweet taste intrinsically pleasant and bitter taste intrinsically unpleasant to newborns, a particular wiring pattern must be laid down. The sweet and bitter taste cells each send their electrical signals to dedicated neurons in a region called the taste ganglia. From the taste ganglia, they course into the brain, passing through three processing stations before contacting neurons in the insular cortex, the region of the brain responsible for identifying tastes.¹³ Importantly, throughout this taste pathway from the tongue to the insular cortex, the bitter- and sweet-evoked electrical signals are kept mostly separate. Experiments in mice indicate that axons conveying bitter taste information will make synapses and activate neurons in one patch of insular cortex, while those carrying sweet signals will activate a separate but adjacent patch. This pattern, where different types of sensory information are strictly segregated, is called labeled-line signaling. Artificial electrical activation of the sweet taste cortex will make mice behave as if they are experiencing a sweet taste and lick a water spout over and over. Likewise, artificial electrical activation of the bitter taste cortex will mimic a bitter taste and suppress licking behavior.

However, the pleasant or unpleasant quality of taste is not produced in the insular cortex itself. That requires a further continuation of the labeled lines. The sweet cortex sends axons to activate neurons in the anterior basolateral amygdala, while the bitter cortex mostly projects to the nearby central amygdala. Sure enough, artificial activation of sweet cortex nerve terminals in the anterior basolateral amygdala produces a pleasant sensation in laboratory mice, while activation of bitter cortex nerve terminals in the central amygdala produces an unpleasant sensation. When the neurons of the amygdala are prevented from firing, then the mice can still distinguish sweet from bitter (which requires the insular cortex), but the tastes no longer evoke pleasant or unpleasant responses—sweet and bitter are rendered emotionally neutral.¹⁴ If molecular genetic tricks are used to cross-wire the taste systems in mice so that the bitter taste cells send their information along the sweet pathway and vice versa, then sweet tastes will be perceived as both bitter and unpleasant and bitter tastes will be perceived as both sweet and pleasant.¹⁵ The system is modular, with separate stations in the brain for taste identification and taste-evoked emotional responses. When brain researchers speak of a “hardwired behavioral response” in newborns, this is typically metaphorical language. Most often, we don’t really have the neural wiring diagram to explain the innate behavior. But here, we do.

IF HUMAN TASTE RESPONSES are really so hardwired, then why don’t we all like and dislike the same foods? One reason is that we all carry individual variations in the genes that encode our taste sensors. While almost no one is completely ageusic (insensitive to taste), careful testing in the laboratory reveals that there is considerable variation in individual response to pure taste-activating chemicals applied directly to the tongue (the experiments are done this way to minimize activation of the sense of smell). For example, testing with L-glutamate droplets in a population of French and American adults revealed that about 10 percent can only sense umami taste weakly, and about 3 percent cannot sense it at all.¹⁶ Sure enough, when we look into the DNA, there are tiny variants in the human umami sensor genes (T1R1 and T1R3) that confer greater or weaker sensitivity to L-glutamate. These probably underlie the individual differences in umami sensitivity.¹⁷ Similar stories have been found for genetic variation in one of the two genes that encodes the sweet sensors.¹⁸ In the case of bitter sensors, where there are at least twenty-five different ones, genetic variation in a specific bitter receptor can give rise to individual variation in sensitivity to the particular chemicals that activate it, but that variation does not generalize to all bitter substances. For example, if you carry a bitter sensor variant that makes you extra sensitive to the bitter taste of caffeine, you will also be extra sensitive to quassia bark extract, which activates the same receptors, but you will not necessarily be extra sensitive to sinigrin, a bitter chemical from black mustard seeds, which activates a different group of receptors.¹⁹

To this point, we’ve talked about genetic variation in taste sensors that is specific to one of the five basic tastes. In addition to that, there’s another, more general type of heritable variation: the number of taste-bud-bearing fungiform papillae on the surface of the tongue. People with lots of papillae are more sensitive to bitter taste, particularly that produced by an artificial chemical called PROP.²⁰ They have been called supertasters by researcher Linda Bartoshuk, and they represent about 25 percent of the population. Conversely, another 25 percent of people cannot taste PROP at all and so are called nontasters. While nontasters can taste some other bitter chemicals, their overall sensitivity to a range of bitter chemicals is reduced. The remaining 50 percent of people, simply called tasters, are in the middle: to them, PROP tastes bitter but not extremely so.²¹ Supertasters also have a somewhat higher sensitivity to sweet, salty, and umami tastes, as well as oral non-taste stimuli like the burn of chili peppers or alcohol. There is a corresponding reduction in the sensitivity of nontasters to these diverse mouth sensations.²²

I dislike these terms because they are laden with positive and negative associations. Who wouldn’t want to be a supertaster? It sounds so cool, like Superman. On the other hand, being labeled a nontaster seems like an insult—the ultimate bland milquetoast. In fact, most supertasters tend to be picky eaters who shun strong flavors to avoid overstimulation. Often, they are particularly averse to vegetables, which are the source of many bitter chemicals. Enthusiasts of strong flavors, who typically enjoy a broader range of foods, are much more likely to be tasters or nontasters. But even this is not guaranteed, as there can also be an interaction with personality traits. A minority of supertasters who are strongly novelty seeking and risk-taking in all aspects of their lives tend to enjoy strong-tasting foods, even if they find them overwhelming.

All of the genes we’ve discussed so far have been expressed in the taste receptor cells, but they are almost certainly not the whole story. It’s likely that genetic variation in the later stages of the taste-processing pathways can also influence taste experience. In that vein, it’s worthwhile to reemphasize that taste identification and emotional taste reactions are processed separately. Although there is no evidence of this yet, we can easily imagine that genetic variation in the taste-activated neurons of the amygdala might give rise to changes in the perceived pleasantness or unpleasantness of a particular taste sensation, without changing one’s sensitivity to it or ability to identify it.

In addition to genetic variation, there are changes in taste perception that depend on stage of life. For example, when people are given a set of twelve-ounce glasses with increasing concentrations of sugar dissolved in water and asked to identify the one with the ideal level of sweetness (charmingly called the bliss point), most adults will pick a glass with about ten teaspoons added. It’s troubling to me that this sugar concentration is even slightly sweeter than most soft drinks. Children have an even sweeter tooth—they choose an average of eleven teaspoons per glass of water.²³ For babies, it’s not really possible to make a sugar solution that’s too sweet for them. They will happily lick thick sugar syrups that even an eight-year-old would reject. At present, we don’t know if these developmental changes in sweetness perception only represent changes in the sensitivity to sweet taste or also involve the perceived pleasantness of sweet taste. The change could be in the tongue or in the brain or both.

The opposite age trajectory is found for bitter tastes: babies are the most intolerant, followed by children, followed by adults. This is generally the case for all five of the basic tastes: as we age, our sensitivity is gradually reduced. On average, women rate bitter tastes as stronger than men, and they experience a temporary increase in bitter sensitivity in the first trimester of pregnancy.²⁴ It has been suggested that this transient bitter aversion serves to reduce the chance of maternal toxin ingestion during the critical early stages of pregnancy, but this idea, while plausible, is just speculation.

We are born predisposed to like sweet, umami, and mild salt tastes but to dislike bitter and sour tastes. Yet many adults, and even some children, enjoy bitter and sour foods like broccoli, coffee, sour candy, and yogurt. While genetic variation and age-related changes contribute to individual food preferences, they are not the whole story. When fraternal and identical twins reared apart were asked to fill out a self-report of their diet, the results indicated that only about 30 percent of the variation in adult food preference was heritable. Surprisingly, the remaining variation had almost no contribution from the shared environment, suggesting that by adulthood the learned portion of our food preferences mostly occurs outside of the family.²⁵ Indeed, other studies have shown that even identical twins raised together develop some divergent food preferences by the time they become adults.²⁶

Beyond genetic and life-stage influences on taste sensation, there are two main contributors to individual food preference. The first is learning. As we gradually try new foods, we develop associations, good and bad. That cup of coffee is bitter, but it gives me a nice little buzz. With time, I can come to enjoy its flavor. That yogurt is a bit sour, but I like the feel of it in my mouth. Perhaps it can be added to my list of acceptable foods as well. We are all embedded in social groups with complex ideas about foods and eating, and it is our lifelong job as omnivorous food generalists to figure out what we like to eat and to do so in a social context.

The second major factor in individual food preference is the role of the non-taste aspects of blended flavor sensation: mostly smell, but also sight, hearing, and touch. For example, experimental subjects chewing potato chips rate their freshness and appeal based in part on the mouthfeel and sounds produced as the chip shatters. These sorts of multisensory influences are not limited to the food itself. The laboratory of Charles Spence at the University of Oxford has performed many studies showing that our perceptions of food flavors can be affected by all sorts of factors: the sounds of the restaurant, the color and size of the plate or bowl, the weight of the dining implements, etc. In the case of potato chips, the researchers found that the crinkly rattling sound of the chip bag contributed to the perceived crunchiness of the chip itself. In another study, they showed that white yogurt was perceived as slightly sweeter when eaten with a white spoon compared to a black spoon.²⁷ These effects tend to be small, on the order of 15 percent, but they are significant and underscore that flavor is truly a multisensory experience and that individual food preferences could potentially be influenced by variation in hearing, sight, or touch sensations.

MOST OF OUR EXPERIENCE of flavor comes from the sense of smell. You may not think either Coke or 7UP has much of an aroma, but without smell, they are indistinguishable sweet fizzy water. A slice of steak is merely salty chewy stuff with a bit of umami mixed in. Lemonade is just sweet and sour water. Not surprisingly, people who have lost their sense of smell (a condition called anosmia) take little pleasure in food and often struggle to eat enough to maintain a healthy weight. In addition, they often report difficulty sleeping, cognitive disruption, and loss of motivation and feelings of social connectedness. Most importantly, anosmia sufferers have a significantly elevated risk of depression and suicidal thoughts. Anosmia is a serious, life-threatening condition that goes well beyond the evaluation and enjoyment of food, as we also rely on smell for social and sexual cues, detecting danger, learning, and even navigation.

In relation to food, the sense of smell supports three main decisions. The first is: Where can I find food? The second: Should I put this food item in my mouth? The third, also informed by the sense of taste: Should I swallow this item that’s in my mouth or spit it out? Considering these questions is key if we are to develop an understanding of the human sense of smell and how it varies among individuals. The first and second decisions involve evaluating odor molecules originally located outside of the body. These odorants are inhaled through the nostrils to reach a patch of about twenty million specialized olfactory receptor neurons, located in the upper wall of the nasal passages. This sniffing route is called orthonasal olfaction (figure 13).

However, when food is in the mouth, odor molecules released from it are carried by exhaled breath through a passage at the rear of the roof of the mouth called the nasopharynx and so reach the patch of olfactory receptor neurons through the back door. This smelling through exhalation is called retronasal olfaction and is only found in a subset of mammals, including primates and dogs. Importantly, sniffed and exhaled odorants do not reach the array of olfactory receptor neurons in exactly the same pattern and concentration. This means that if you sniff something, the odor you experience is somewhat different than if you hold it in your mouth and exhale. This is probably why some foods, like certain ripe cheeses, can have an offensive odor when sniffed but have a fine flavor (to which odor is a major contributor) once they are in your mouth.²⁸

Let’s consider the smell of a ripe tomato. The tomato is composed of thousands of different molecules. Of those, only about 450 will be sufficiently small and volatile to be released into the air, where we can measure and identify them using a machine called a gas chromatograph. Of those that travel on the air, only sixteen will bind to the specialized odor receptors on the olfactory receptor neurons with sufficient affinity to reach the threshold of perception by humans and so contribute to the blended odor we call tomato smell.”²⁹ Even then, it’s likely that not all sixteen different odor molecules are necessary to evoke a tomatoey sensation. One could probably make a very convincing artificial tomato smell from a subset of those chemicals. In fact, chemical companies do this all the time. For example, roses emit hundreds of volatile molecules, but just one, phenylethyl alcohol, is sufficient to convey a convincing rose smell. In fact, when asked to sniff two vials, one containing natural rose essence and the other containing only phenylethyl alcohol, most people in modern societies, having been exposed throughout life to artificial rose scent in products like hand soap, will misidentify the pure chemical as the natural product.

FIGURE 13. Odors can take two different paths to the olfactory receptor neurons. The orthonasal route is engaged by breathing in and samples the external world, while the retronasal route is engaged by breathing out and carries food odors from the mouth to the olfactory receptor neurons via the nasopharynx, the airway at the rear of the roof of the mouth. The olfactory receptor neurons pass their electrical signals through local circuits in the olfactory bulb before these signals branch to reach five different brain regions, each concerned with a different aspect of odor processing. This figure has been adapted from Shepherd, G. M. and Rowe, T. B. (2015). Role of ortho-retronasal olfaction in mammalian cortical evolution. Journal of Comparative Neurology, 524, 471–495, with permission of the publisher, John Wiley and Sons. © 2019 Joan M. K. Tycko.

The olfactory receptor neurons are clustered together in a yellowish patch of mucus-covered tissue in the upper wall of the nasal cavity. There are about twenty million of these cells, and each one expresses a single one of the four hundred or so types of olfactory receptor. A particular smelly chemical, like the aforementioned rosy odorant phenylethyl alcohol, will activate many different types of olfactory receptor—perhaps ten to forty of the four hundred possible types. A different pure odorant molecule, like eugenol, which smells like cloves, will activate a different group of olfactory receptors, a few of which might overlap with that of phenylethyl alcohol. Of course, natural smells, like freshly mown grass or wood smoke, are composed of many different odorant molecules at various concentrations, so the pattern of olfactory receptor activation will be even more complex. The main point here is that there is rarely a single receptor for a single odor, even if that odor is from a single pure chemical odorant. If you carry mutations in a single odorant receptor, this usually won’t cause you to become supersensitive or insensitive to a single odor, but will more likely have complex effects on your perception of a range of odors.

Like the receptors for a particular taste sensation, such as sweet, which are distributed across the surface of the tongue, the olfactory receptor neurons that express a particular odorant receptor are not clustered together in one patch, but rather are spread across the array of olfactory receptor neurons. However, the information-conveying axons from all of these dispersed receptor neurons converge on one spot in the next processing stage, the olfactory bulb, a specialized part of the brain. This means that different spots in the olfactory bulb where axons converge (called glomeruli) correspond to different types of odorant receptor.³⁰ Furthermore, at least in mice and rats, the dorsal portion of the olfactory bulb seems to convey signals that result in innate avoidance responses, like the odors of carrion or fox urine, or innate attractive responses, like those evoked by the mouse sex pheromone darcin. From the olfactory bulb, smell information is carried by axons that split to send the information to five different brain regions (figure 13), including the piriform cortex, which is responsible for odor recognition, and the cortical amygdala, which is required to attach positive or negative emotional valence to intrinsically attractive or aversive odors.

I hate to burden you with neuroanatomical details, but there’s one subtle point here that’s really important for understanding our experience of odors. When the axons from the dorsal olfactory bulb travel to the cortical amygdala, they cluster together, meaning that activation of an odorant receptor can activate both a particular patch of adjacent neurons in the olfactory bulb and a particular patch of adjacent neurons in the cortical amygdala. This is exactly the labeled-line pattern of connection one would expect for innately aversive or attractive odors. However, when the axons from the other parts of the olfactory bulb travel to the piriform cortex, they do not terminate in patches. Rather, they distribute their signals widely across the piriform cortex.³¹ This means that a single smell-receptive neuron in the piriform cortex receives information from many, many different types of olfactory receptor, like a giant switchboard.

At first glance, this arrangement seems wasteful. Why go to all of the trouble to cluster information from all of the olfactory receptor neurons together in the olfactory bulb, only to then scatter it pell-mell across the piriform cortex? The likely answer is that the piriform cortex is an olfactory learning machine, where neurons are tuned by the pattern of inputs they have received from the experience of smelling the world. The odor signals to the cortical amygdala are hardwired to produce invariant responses, but the piriform cortex is the proverbial blank slate, waiting to be molded by life experience.

THE POPULAR NOTION THAT humans are inferior to most other mammals in the smell department simply isn’t true. There are several factors that determine olfactory ability of a species, including the number of olfactory receptor neurons (about 20 million for us but about 220 million for a bloodhound) and possibly the number of different olfactory receptor proteins (about four hundred in humans, one thousand in dogs, and nine hundred in mice).

While there are some odors that other animals can detect that we cannot smell at all, we’re surprisingly good at detecting most odors that originate from plants, bacteria, and fungi. When Matthias Laska of Linköping University in Sweden compared the sensitivity of many species to a panel of different pure odor molecules, he found that, on average, humans were generally more sensitive than many species thought to have a refined sense of smell, including rabbits, pigs, mice, and rats. Dogs, however, kick our human butts, often detecting odors at concentrations over one-million-fold lower than us. When the ability to tell two simple odors apart was tested, humans were in the middle of the pack: worse than dogs, mice, and Asian elephants, but similar to squirrel monkeys and fur seals.³²

A bloodhound, with its superior ability to detect and discriminate among faint odors, is far better than a human at tracking animals by scent. The other advantage dogs have is their nose location, which makes it easy for them to sample the ground for odors. In one of the more amusing experiments reported in recent years, intrepid smell researcher Noam Sobel and his colleagues showed that college students, fitted with blindfolds, earplugs, and thick gloves, could do a decent job of tracking a scent trail of chocolate extract laid through the grass of an athletic field. All it took was for the students to swallow their pride, get down on all fours, and sniff. The students were pretty good even the first time out and, with repeated practice, could learn to track faster and with greater accuracy.³³

WHEN WE THINK ABOUT the olfactory abilities of different animals, it’s useful to consider the various problems that they have to solve through smell and how these have changed through evolutionary time. Dolphins appear to have no sense of smell at all.³⁴ This isn’t just a result of living in water, however, as salmon, which navigate by smell to find their natal river at spawning time,³⁵ arguably have the most sensitive smeller of any animal. Mice can sense odors in mouse urine that are undetectable to us but that convey social information to them. Even the very same odorant chemical can have different innate meanings to different animals: the predator odor 2-phenylethylamine is understandably aversive to mice, but it serves as a sex pheromone in tigers. Carrion odors, like the evocatively named molecules putrescine and cadaverine, are aversive to mice but attractive to scavengers like vultures. Each species has a different set of decisions to make based on olfactory information; their sense of smell, from nose to brain, has evolved to support those decisions.

Humans have about four hundred genes encoding functional olfactory receptors and about six hundred more nonfunctional olfactory receptor pseudogenes. Presumably, these olfactory pseudogenes detected odors that were important in the lives of our ancestors, but which are no longer important today. If we compare olfactory receptor genes across species, from the numbers I just mentioned, we can calculate that humans have about 60 percent pseudogenes. Our primate relatives that share three-color vision with us—like chimpanzees, gorillas, and rhesus macaques—have about 30 percent pseudogenes, whereas primates with two-color vision like squirrel monkeys and marmosets have only about 18 percent pseudogenes. This observation has led evolutionary biologist Yoav Gilad and his colleagues to suggest that the development of three-color vision relaxed the evolutionary selective pressure on the sense of smell, allowing the loss of more olfactory receptor genes.³⁶ For example, one can imagine a scenario in which improved three-color vision could aid in finding ripe fruit that previously had to be sniffed out with a broader range of odor receptors.

ON AVERAGE, HUMANS ARE good at detecting faint odors, and we’re reasonably good at telling two odors apart, but when it comes to identifying even familiar odors by name, we leave a lot to be desired. Imagine if I were to sneak into your home and raid your fridge and bathroom for familiar objects—foods, beverages, cosmetics, medicines—that emit odors. Then, I’d blindfold you and wave the objects under your nose, allowing you to get a good strong sniff. What fraction of the objects do you think you could correctly identify? The answer, from laboratory experiments, is between 20 and 50 percent, with better performance for young adults and a gradual decline with age. These rates of identification fall further if the odors are not necessarily familiar ones.³⁷ For comparison, if I were to perform a similar task asking you to name familiar objects by sight, you’d likely get nearly 100 percent correct. Vision can reliably trigger the memories for object names, but smell is much less able to do so.³⁸

Our mediocre ability to identify odors is likely related to our poor ability to describe them. In most languages, there are only source-based descriptors for various smells: a whisky’s bouquet might be described as smoky or peaty. A wine might be redolent of pear, tropical fruit, tobacco, or barnyard aromas. The important point is that all of these odor descriptors refer to a particular source—an object or a process. In English, as in most languages, we do not have abstract terms for odors the way we do for color. A tomato, a firetruck, and a stop sign are all red, and the name of the color makes no explicit reference to the objects that share that property. “Red” is an abstract descriptor, while “smells like a banana” is source based.³⁹ If we described colors the way we do odors, we’d describe the American flag as having alternating cloud and cherry-colored stripes with cloud-colored stars on a rectangular field of dark sky.

Ethnographers have uncovered examples that challenge the idea that our ability to recognize and name odors is intrinsically limited by brain structure and function.⁴⁰ The Serer-Ndut people of Senegal have five abstract odor descriptors. Of these, pirik is the smell of bean pods, tomatoes, and various spiritual beings, while he is the smell of raw onions, peanuts, limes, and the Srer-Ndut themselves.⁴¹ Nomadic hunter-gatherers living on the Malay peninsula, who speak Maniq or Jahai, have about fifteen abstract terms for odors. For example, the Jahai word itpt can be used to describe odors as distinct as soap, bearcats (Arctictis binturong), durian fruit, and certain flowers.⁴² The Maniq word mi is used to describe the smell of animal bones, mushrooms, snakes, and human sweat.⁴³ Perhaps it’s not surprising that Jahai speakers perform better than English speakers at naming both familiar and unfamiliar odors.⁴⁴ My own reading of the literature leads me to believe that identifying objects by smell and by vision are fundamentally different processes in the brain. Nonetheless, within the constraints of the smell-identification system, intense experience evaluating odors, like that found in populations of hunter-gatherers, can push the boundaries of olfactory naming.

Can we train ourselves to be more like the Jahai? Bianca Bosker was a plucky technology reporter living in New York City with no special affinity for or knowledge of wine or food when she set herself the task of passing the famously difficult Court of Master Sommeliers certification exam within eighteen months. Impressively, she succeeded. She recounts her struggles to rapidly become a wine expert in her hilarious and informative book Cork Dork. Bosker writes:

I liked wine the same way I liked Tibetan hand puppetry or theoretical particle physics, which is to say I had no idea what was going on but was content to smile and nod. It seemed like one of those things that took way more effort than it was worth to understand.… I was captivated by these people who had honed the kind of sensory acuity I’d thus far assumed belonged exclusively to bomb-sniffing German shepherds.

In order to progress from a wine novice to a certified sommelier, she had to learn a lot of facts about wineries, grapes, and food pairing, as well as how to recommend and serve wine with aplomb. Most importantly, to pass the exam she had to learn to identify two mystery wines, one red and one white. To accomplish this feat, she had to hone her wine senses by drinking a ton of wine and concentrating hard to isolate the various smell, taste, mouthfeel, and sight components. She learned to attend to many individual sensations, ranging from the color of the wine in the glass, to the degree of burn from the alcohol, to a faint odor of violets that might emanate from an Oregon pinot noir. Most importantly, she had to learn to put words to her wine-drinking experience, and because she’s an English speaker, not a Jahai speaker, her words were mostly source based: “Notes of tropical fruit and green grass” and so on.⁴⁵ In a sense, she had to train for eighteen months to get some of the olfactory fluency that a Jahai person would have already acquired by age twelve.

It’s to be expected that trained wine experts, as well as perfumers and other odor professionals, can perform much better than the average person on recognizing individual familiar odors. However, when familiar odors are blended in a mixture, even the world’s most highly trained noses struggle to identify more than three or four components, which is not that much better than untrained people.⁴⁶ There appears to be a hard limit on identifying odors in a mixture than can’t be extended even with the most intensive training. This makes you wonder what wine experts are reporting when they give tasting notes with lists of ten or more smells.

When experts are evaluating wine, they are using every available sense. What’s surprising to me is the degree to which vision can overshadow taste and smell in these evaluations. In one experiment, a panel of wine experts was asked to evaluate two glasses of wine, one a white (a 1996 Bordeaux containing sauvignon blanc and Sémillon grapes) and the other consisting of the very same white wine with an odorless, tasteless organic red dye added to make it appear red. When the panel was asked to describe the flavor of the white wine, they mostly used typical white wine descriptors like grapefruit, pear, and floral bloom. However, when evaluating the white wine dyed red, they almost completely switched to typical red wine descriptors like tobacco, cherry, and pepper.⁴⁷ The point here is not to belittle wine experts, but to point out an important aspect of smell: in the real world, olfaction is most often used in combination with other senses, and those other senses can strongly influence the perception of odors.

There’s something about the sense of smell that lends itself to imagination and, hence, deception. Consider this devious experiment, reported by E. E. Slosson of the University of Wyoming in 1899:⁴⁸

I had prepared a bottle filled with distilled water carefully wrapped in cotton and packed in a box. After some other experiments I stated that I wished to see how rapidly an odor would be diffused through the air, and requested that as soon as anyone perceived the odor he should raise his hand. I then unpacked the bottle in the front of the hall, poured the water over the cotton, holding my head away during the operation and started a stop-watch. While awaiting results I explained that I was quite sure that no one in the audience had ever smelled the chemical compound which I had poured out, and expressed the hope that, while they might find the odor strong and peculiar, it would not be too disagreeable to anyone. In fifteen seconds, most of those in the front row had raised their hands, and in forty seconds the ‘odor’ had spread to the back of the hall, keeping a pretty regular ‘wave front’ as it passed on. About three-fourths of the audience claimed to perceive the smell, the obstinate minority including more men than the average of the whole. More would probably have succumbed to the suggestion, but at the end of a minute I was obliged to stop the experiment, for some on the front seats were being unpleasantly affected and were about to leave the room.

Lest we think that audiences in 1899 were more susceptible to olfactory suggestion, we can turn to a more recent experiment by psychologist Michael O’Mahony. He arranged for a television show broadcast in the area of Manchester, United Kingdom, to induce olfactory hallucinations in its viewers.⁴⁹ At the end of a show about the sense of taste and smell, viewers were told that it was possible to transmit smell by sound. They were shown a bogus apparatus consisting of a two-foot-tall cone called the “taste trap,” which had contained a commonly known odorous substance for twenty-three hours, and which was connected to a suitably technical looking nest of cables and electronic equipment with flashing lights. It was falsely explained to viewers that smells were characterized by the frequencies of vibration of the molecules of a substance, and that the vibrations of the molecules causing the odor in the taste trap were being picked up by sensors. A sound of exactly the same frequency as the odor would then be played and broadcast. The brains of the listeners would recognize these frequencies as smell frequencies, which would cause them to experience an odor. Viewers were asked to phone or write to the television company saying whether they had smelled anything or not, and if so what they had smelled. They were particularly urged to write should they not experience any smell at all. Because it was a late-evening show, viewers were told that the smell transmitted would be something they would not normally smell in their house but rather an outdoorsy, country smell. At this point, the studio audience laughed because they guessed it might be manure, so it was clarified that the smell would be a pleasant country smell, not manure. Afterward, 130 people contacted the TV station to report smells. The most common smells were hay and grass, but onions, cabbage, and potatoes also made the list of phantom odors.⁵⁰

It’s not just imaginary odors where our perception is malleable. We’re deeply influenced by suggestion, context, personal history, and, of course, the evil forces of advertising. If you live in the United States or Europe, then you are likely to have been told by personal care corporations or so-called aromatherapists that the smell of lavender is relaxing and the smell of neroli (an extract from blossoms of the bitter orange tree) is stimulating. This is true, but only if you believe it already. There’s nothing intrinsically stimulating about neroli or relaxing about lavender. In one experiment, C. Estelle Campenni and her coworkers exposed college students individually to either lavender essence or neroli essence. The odor was not identified, but in some cases the subjects were told that it was known to be stimulating and, in others, relaxing. Sure enough, students told that lavender was stimulating reported feeling stimulated and had an increased heart rate, while the opposite effects were found when they were told that lavender was relaxing. Of course, the same thing was true of neroli. The odor didn’t matter. Only the suggestion.⁵¹

The power of verbal suggestion on olfactory experience was also tested by psychologists Rachel Herz and Julia von Clef, who presented test subjects with ambiguous odors with labels that carried either positive or negative associations. One of these odors was a mixture of isovaleric acid and butyric acid labeled either “parmesan cheese” or “vomit.” Not surprisingly, participants rated the parmesan cheese odor as significantly more pleasant than the vomit odor, even though they were the very same substance. In fact, 83 percent of the subjects were convinced that they had actually smelled two different odors.⁵² While all the senses are subject to manipulations by learning, expectation, and context, smell perception seems to be particularly malleable.

UNLIKE MICE, HUMANS ARE born with few innate emotional responses to odors. This is a good strategy for a wide-ranging omnivore who must learn to consume a variety of smelly foods. Newborns arrive with an inborn aversion to the rotten-fish odor trimethylamine and the spoiled-meat odors putrescine and cadaverine,⁵³ but an attraction to the secretions of the Montgomery’s glands on the breasts of nursing mothers. Presumably, these odors activate the dorsal olfactory bulb to cortical amygdala pathway. But these innate odor responses are the exception, not the rule. Beyond these few limited examples of innate odor responses, our like or dislike of odors is mostly a matter of learning in a social context.

You might think that poop smells are intrinsically unpleasant, and indeed most adults around the world shun fecal odors, but babies happily play with their own waste. They must be taught that fecal odors are disgusting. This teaching is culturally specific. Notably, several groups in Africa, including the Maasai of Kenya and the Mwila of Angola, mix cow dung with other ingredients like butter to make a hair treatment. It probably helps that cows, being herbivores, consume a diet low in cysteine, an essential amino acid that is broken down during digestion to form the intrinsically aversive smelling chemical hydrogen sulfide. Meat eaters consume much more cysteine in their diet, and so their farts and poop are much more redolent of hydrogen sulfide. For this reason, there’s probably nowhere in the world where people smear cat poop on themselves.

But what about pepper spray or raw onions or ammonia-laden smelling salts? Aren’t those substances universally aversive, even to newborns? The answer is yes, but the reason is because they contain volatile chemicals that are not odors at all. Rather, they activate a special part of the touch sense. In addition to olfactory receptors, the nasal cavity has free nerve endings that are specialized for detecting certain irritating chemicals, like capsaicin (the heat-mimicking chemical of chili peppers, which binds a receptor called TRPV1) and menthol (the cold-mimicking chemical of mint, TRPM8), as well as the warm-mimicking chemicals of horseradish, onions, garlic, and ginger (TRPA1).⁵⁴ Ammonia, found in smelling salts and hydrogen sulfide, as well as rotten eggs and carnivore poop, can also activate TRPA1 receptors.

While these chemical-sensing nerve endings are found in the nose, they are also in the mouth, skin, eyes, and cells that line the airways. This is why, for example, you can feel warmth from chili pepper extract or cooling from crushed mint, even when they are smeared on the skin of your arm. While smell information is conveyed to the brain through the olfactory nerve, the chemical touch sense from the face is carried by a completely different pathway, the trigeminal nerve, and is ultimately processed by different regions of the brain. Strong activation of these chemical touch senses is intrinsically aversive, which is why newborns have an innate, unlearned aversion to chili peppers, raw onions, and ammonia-based smelling salts.

WHEN LESLIE VOSSHALL AND her colleagues at Rockefeller University tested 391 adults of varying backgrounds from New York City on odor perception tasks, there were some interesting trends that emerged.⁵⁵ Smell acuity was gradually reduced with age. On average, women had somewhat lower thresholds for detecting faint odors than men.⁵⁶ Not surprisingly, smokers tended to have reduced olfactory acuity. While you might expect that blind people would have developed a more acute sense of smell by means of sensory compensation, that did not seem to be the case.⁵⁷ Overall, individual variability in overall olfactory acuity was the norm. Some people are just better smellers than others, and some people have even lost their sense of smell entirely. However, when you ask people to rate their own sensitivity to smells by merely reflecting on their life experience, most rate themselves as above average.⁵⁸

While there is variation between individuals in overall sensory acuity, there is even more variation in the responses to individual odors. For a given odorant, like 2-ethylfenchol, which has an earthy, mossy smell, certain people might be able to detect it at concentrations one hundred times lower than others, while another group might fail to detect it at all. These differences in sensitivity to particular odors are found widely across individuals. We each live in somewhat different olfactory worlds. My strawberry is not your strawberry and my Gouda cheese is not your Gouda cheese.

When scientists sequenced the genes that encode the four hundred or so functional olfactory receptors in many people, they found that, compared with other parts of the genome, these genes are unusually susceptible to large functional changes—they are more likely to be rendered completely nonfunctional or duplicated in their entirety. In addition, they found an unusually large number of subtler variants in olfactory receptor genes: changes in the DNA that might alter a single amino acid in an olfactory receptor protein and thereby make it more or less sensitive to a given odor. One such study estimated that, between any two unrelated people, on average there would be functional differences in 30 percent of their odorant receptors. That’s a lot of variation, and it could account for a significant share of individual odor perception.

The first example of a link between human genetic variation and odor perception involved the chemical androstenone, a metabolite of testosterone, which is perceived by different people as offensive (sweaty, urine-like), pleasant (floral), or odorless. Hiroaki Matsunami and his colleagues found that subtle single-nucleotide variants of a particular odorant receptor called OR7D4 could predict individual sensitivity to androstenone.⁵⁹ Since then, subtle genetic variation (single-nucleotide mutations) in other genes coding for odorant receptors have been linked to individual perception of odors. These include the smells of isovaleric acid (cheesy, sweaty), beta-ionone (floral), cis-3-hexen-1-ol (fresh-cut grass), guaiacol (smoky), and several others.⁶⁰ While at present we only have a handful of these examples, it’s likely that many more will emerge as this topic is investigated further.

One of the best-known individual differences in odor perception involves the smell that some people detect in their urine after eating asparagus. For some, like Marcel Proust, the odor was pleasant. He effused that eating asparagus spears “transform[ed] my humble chamber pot into a bower of aromatic perfume.” Benjamin Franklin could also detect an odor in his urine after eating it, but wrote that “a few stems of asparagus eaten shall give our urine a disagreeable odor.” While Proust and Franklin disagreed about the pleasantness of asparagus-pee odor, clearly both could smell it. Others are unable to detect the smelly methanethiol and S-methyl thioesters that are produced in urine from asparagus consumption. The rate of asparagus-pee anosmia has been reported to vary across different populations. One recent report that sampled adults of European American ancestry found that about 58 percent of men and 61 percent of women failed to detect asparagus odor in their urine. Because most people don’t go around smelling each other’s urine, this failure could be a metabolic inability to produce the odorants, an inability to detect them, or both. This question was addressed by some rather unsavory studies showing that people who cannot detect the odor in their own urine also cannot detect it in the urine of known producers, supporting the selective anosmia hypothesis. When the genomes of this population were scanned, three different single-nucleotide mutations were statistically associated with asparagus-pee anosmia, indicating a potential genetic contribution to this trait.⁶¹

CHARLES WYSOCKI, A RESEARCHER at the Monell Chemical Senses Center in Philadelphia, had assumed that he was one of the 30 percent of people who couldn’t smell androstenone. He could weigh it out in the lab and load it into vials for others to sniff, but it never had any kind of an odor to him. But after a few months of working with it, he noticed a new smell in the lab, and sure enough it was the androstenone. Repeated exposure had changed him from a non-smeller to a smeller. Intrigued, he found twenty people who were non-smellers (but who had normal sensitivity to other odors) and had them take a whiff of an androstenone vial three times a day for six weeks. Of those twenty non-smellers, ten became smellers within a week or two. Presumably, the ten non-smellers who didn’t improve had broken versions of the key androstenone receptor OR7D4, so they had no possibility of becoming more sensitive. Importantly, the ten androstenone non-smellers who did improve did not have an overall decrease in odor thresholds, as their ability to detect two other smelly molecules, amyl acetate and pyridine, was not changed.⁶² This finding has since been replicated by other researchers, and it has been shown that it’s not just that non-smellers of androstenone can become smellers, but that people who can smell it weakly can have their sensitivity increased further with repeated exposure.⁶³

One possible explanation for this sensitization effect is that intermittent exposure somehow makes the olfactory receptor neurons more responsive to the exposed odor, causing them to send stronger electrical signals on to the odor-evaluating parts of the brain. Another possibility, which is not mutually exclusive with the first, is that the odor-detecting circuits, particularly those in the piriform cortex of the brain, are plastic and change by repeated exposure, so as to efficiently extract the relevant odor-evoked electrical signals coming from the nose. When a tiny electrode is threaded up the nostril to record electrical activity of the olfactory receptor neurons of non-smellers, the androstenone-evoked signals gradually increase with repeated androstenone exposure in those who become smellers, indicating changes in the nose itself.⁶⁴ A recent experiment in mice suggests how this might happen. Repeated, intermittent exposure to a particular odorant changed the expression pattern of certain olfactory receptors in the olfactory receptor neurons, possibly rendering the nose more sensitive to that particular smelly chemical in the future.⁶⁵

In a different experiment, androstenone non-smellers were repeatedly exposed to androstenone in one nostril only (using a nose plug/air blower arrangement to carefully restrict exposure to a single nostril both orthonasally and retronasally). After three weeks of exposure, sensitivity to androstenone was found with separate application to either the exposed or nonexposed nostril.⁶⁶ This finding suggests that the changes occur in the brain, where information from the two nostrils is combined.⁶⁷ This is consistent with a number of brain-scanning experiments showing changes in the electrical activity of odor-processing regions of the brain after odor-related learning.⁶⁸

When Bianca Bosker developed her wine expertise through repeated careful tasting, is it possible that such training increased the sensitivity of her nose to faint wine-related odors? Perhaps there are odorants in wine like androstenone, for which one can increase one’s sensitivity with repeated, intermittent exposure. While there is still much work to be done on this question, the early indications are negative. Wine experts (and other odor experts like perfumers) have an increased ability to put names to familiar odors, but their sensitivity to faint odors, even those commonly found in wine, does not appear to be different than the average person off the street.⁶⁹

LEARNING ABOUT FLAVORS STARTS in the womb. The world’s expert on this topic is Julie Mennella, who has shown that the substances a mother consumes while pregnant, from foods to cigarettes, will influence her baby’s flavor preferences in early life. Odor and taste molecules can pass from the maternal circulation into the amniotic fluid and can be smelled and tasted by the developing fetus during pregnancy. Mennella and her coworkers reported that when pregnant women consumed carrots, anise, or garlic during pregnancy, this increased the acceptance of these flavors when their children were reexposed during infancy or early childhood. The caveats to this finding are that it doesn’t necessarily work for every food eaten during pregnancy and that it’s not yet clear if this fetal exposure produces lasting effects that influence food choices later in life.⁷⁰

Learning about odors and flavors is a lifelong pursuit for humans. We are not so influenced by the foods of early life that we cannot change our preferences in adolescence or adulthood. We continuously learn to associate particular odors with tastes. This shared experience even makes its way into our language. Most people from the United States will say that vanilla, strawberry, or mint odors smell sweet. On the face of it, this doesn’t make sense. Sweet is a taste, not a smell. A substance cannot smell sweet any more than something can sound red. If we break it down, when we say that something smells sweet, what we mean is that, through our experience, we have come to associate that smell with sweet taste. In the case of strawberries, they are naturally sweet when ripe. In the case of caramel, vanilla, and mint, most Americans have experienced these odors in sweet foods like cookies or gum or sweetened drinks. There’s nothing intrinsically sweet about these odors—we’ve just learned to associate them with sweet taste. As a counterexample, in Vietnam, where caramel and mint are used primarily in savory dishes, their odors are not typically described as sweet.⁷¹

These effects of odor-taste associations can also be studied in the lab. When Richard Stevenson and his colleagues paired vanilla or caramel odor with a sugar solution, it was rated by test subjects as sweeter than the same sugar solution alone, in a population accustomed to vanilla- and caramel-flavored sweets (Australian college students). Similarly, those “sweet” odors reduced the perception of sourness when added to a sour citric acid solution. After novel odors (like water chestnut) were repeatedly paired with sugar solution, they were rated as smelling “sweeter” than they were before the pairing.⁷² These experiments reinforce the idea that we are constantly learning (and unlearning) associations between odors and tastes.

We also form a strong association when we eat something and then feel ill afterward. Everyone has one of these stories. For me, I was put off lasagna for about twenty years after a childhood bout of gastrointestinal distress following a family dinner at an Italian restaurant. Obviously, learning strong food aversions is adaptive—if something is likely to have made you ill, you don’t want to eat it again, lest it be infected or poisonous.

Compared to other animals, humans are unusually adaptable when it comes to foods, even those that cause a degree of pain. Through learning and deep cultural influences, humans can come to enjoy a wide variety of foods, even those that produce mild pain like chili peppers or raw onions or the ammonia-laden fermented shark dish from Iceland, called hákarl. By comparison, it’s nearly impossible to train your dog or cat to enjoy chili peppers (please don’t try this at home). Even rats, which have a famously varied diet, cannot be trained to like mildly painful foods like chili peppers or wasabi. But humans, as the ultimate food generalists who can live and eat in nearly any location on earth, can learn to overcome our innate aversions to these chemical irritants, as well as to sour and bitter foods and foods with odors that otherwise might indicate dangerous bacterial infection (stinky cheeses, beer, miso, and sauerkraut).

Our individual food preferences are deeply influenced by culture, which, these days, includes advertising. Ethnographers have shown that, around the world, there are specific foods that are a mark of inclusion in nearly every culture. They also often serve as a xenophobic mark of exclusion: “We eat pigs but those other people in the next valley eat fish and it makes them stink.” Our individuality, when it comes to food preferences, is not unbounded, but rather is molded and constrained by cultural ideas that influence taste and odor learning.

These cultural ideas are not limited to food odors, and they can be quite specific. One might imagine that the United States and the United Kingdom share a lot of cultural similarities, but there are some notable differences in ideas about odors. One involves the odor of wintergreen (methyl salicylate), which, in a sample of Americans published in 1978, was ranked the most pleasant of twenty-four odors tested.⁷³ This ranking is in stark opposition to a 1966 survey in the United Kingdom, in which wintergreen was ranked as one of the most unpleasant odors.⁷⁴ While there are a few other examples like this, most odors are ranked similarly in the two countries; people in both countries tend to like jasmine but dislike pyridine, which has a stale, fishy smell. The divergent responses to wintergreen do not arise from genetic differences in the odorant receptors between these populations. Rather, it’s because of associative learning. In the United States, wintergreen is used in candies and gum. In the United Kingdom (at least in 1966), it was almost entirely used in medicines that are rubbed on the skin for pain relief. The pure sensory experience of wintergreen odor is the same for both groups, but the learned associations, and hence the emotional responses, are completely different.

Cultural ideas about odors can change, and this change is not just an invention of our present trend-chasing society. Pliny the Elder, writing in Rome during the first century CE, opined: “The iris perfume of Corinth was extremely popular for a long time, but afterwards that of Cyzicus. Then vine flower scent made in Cyprus was preferred but afterwards that from Adramyttium, and scent of marjoram made in Cos, but afterwards quince blossom unguent.” Being on trend with your Roman perfume was no easy task. Interestingly, these favored perfumes were the same for men and women, a practice that would mostly continue in Europe for centuries. For example, George IV, who ruled England from 1820 to 1830, first encountered a scent on a visiting princess at a royal ball that he later adopted as his own favorite. Fifty years later, styles had changed, and sweet floral blends were deemed exclusively feminine, while men adopted more woodsy scents.⁷⁵ Although perfume companies might tell you otherwise, there’s nothing intrinsically womanly about the smell of flowers. It’s merely a cultural construction of the present moment, aided by an unusually malleable human olfactory system.