A particular terror in losing a sense one had never given a thought to. Not blindness, which the imagination rehearses. Not deafness, which literature has ennobled. But anosmia, the total absence of smell, which, until recently, most people could not even name.
11 min
In the spring of 2020, millions of people woke up in a world emptied of its invisible architecture. Coffee no longer had warmth. Garlic no longer had bite. Their own children smelled of nothing. The loss was not painful the way a broken bone is painful. It was ontological. The world was still there, visually intact, but it had been emptied of a dimension no one had bothered to tell them was holding everything together.
What happened next is one of the strangest episodes in the history of medicine. A rehabilitation protocol developed in a German university lab, obscure, unglamorous, studied by perhaps two dozen researchers worldwide, became a global phenomenon overnight. Millions of people, desperate and largely abandoned by their doctors, began sitting at their kitchen tables twice a day, holding small pots of essential oil under their noses, sniffing with the deliberate concentration of someone learning to walk again.
They were doing smell training. And in doing so, they stumbled accidentally onto one of the most striking demonstrations of neuroplasticity modern science has produced.
The protocol has a name and a father. Thomas Hummel, professor of otorhinolaryngology at the Technische Universität Dresden, had been studying olfactory disorders since the early 1990s, a period when telling colleagues you studied smell was roughly equivalent to announcing you collected bottle caps. The olfactory system was, in the hierarchy of neuroscience, a blind spot. Vision had the cortex. Hearing had cochlear implants. Smell had anecdote and resignation.
Hummel did not resign himself. He had noticed something the clinical literature had largely ignored: some patients with post-viral anosmia recovered their sense of smell over time, and those who reported actively seeking out smells during recovery seemed to recover more fully. The observation was informal. It was also, as it turned out, the seed of everything.
By 2009, Hummel and his team had formalized the intuition into a protocol. Four essential oils, rose, eucalyptus, lemon, and clove, were selected not arbitrarily but according to a classification system proposed by German psychologist Hans Henning in his 1916 book Der Geruch. Henning had described olfactory perception as organized along a geometric prism with six vertices: floral, fruity, resinous, spicy, putrid, and burnt. Hummel’s four oils were chosen to represent four of these six categories. Rose for floral. Lemon for fruity. Eucalyptus for resinous. Clove for spicy. The putrid and burnt categories were excluded for obvious reasons of domestic harmony.
The instructions were disarmingly simple. Twice a day, morning and evening, the patient opened each pot in sequence and inhaled gently for ten to fifteen seconds, concentrating on the smell, or on the memory of the smell, if the smell itself was absent. The minimum duration was twelve weeks. Improvement, when it came, often continued for months afterward.
The first randomized controlled trial, published by Hummel’s group in 2009 in The Laryngoscope, showed a statistically significant improvement in olfactory function relative to controls. Later studies replicated the result. A 2017 systematic review by Sorokowska and colleagues in Rhinology confirmed it. By the standards of olfactory medicine, a field where therapeutic nihilism was the default posture, the results were notable. Here was an intervention that cost almost nothing, had no side effects, and produced measurable structural and functional changes in the nervous system.
Almost no one noticed.
To understand why smell training works, one must understand something unusual about the olfactory system: it is the only sensory system in the human body that continuously regenerates its primary neurons throughout adult life.
The olfactory epithelium, a postage-stamp-sized patch of tissue high in the nasal cavity, contains roughly six million olfactory receptor neurons, each genetically unique to the individual. Each neuron expresses a single type of odor receptor on its surface, drawn from a repertoire of roughly four hundred functional receptor genes, as mapped by the Nobel Prize-winning work of Linda Buck and Richard Axel, published in Cell in 1991. When an airborne molecule binds to one of these receptors, the neuron fires. The signal travels down the neuron’s axon, through tiny perforations in the cribriform plate, a sieve-shaped bone at the base of the skull, and into the olfactory bulb, the brain’s first relay station for smell.
Here is the crucial detail: olfactory receptor neurons live only thirty to sixty days. They are born from a population of basal stem cells in the epithelium, mature, extend their axons through the cribriform plate, form synaptic connections in the olfactory bulb, function for a few weeks, and die. The cycle never stops. You are literally rebuilding your sense of smell every month.
This constant regeneration is the system’s great strength and its great vulnerability. Under normal conditions, newly born neurons follow chemical guidance signals to find their correct targets in the olfactory bulb. Neurons expressing the same receptor type converge on the same glomerulus, a spherical cluster of synapses, creating a precise spatial map. The map is continuously rewritten, but because the guidance signals are stable, each new generation of neurons recreates the same topography. The result is seamless. You never notice the renovation, because the floor plan stays the same.
When a virus damages the olfactory epithelium, which is exactly what SARS-CoV-2 does, along with influenza, rhinoviruses, and others, the regeneration process can go wrong. The stem cells still divide. New neurons still emerge. But the guidance signals can be disrupted. Newly born neurons, like commuters in a city where every street sign has been removed, extend their axons into the olfactory bulb and connect to the wrong glomeruli. A neuron that should plug into the glomerulus coding for rose instead arrives at the one coding for sulfur. The subjective experience of this miswiring is parosmia, that terrifying condition in which familiar smells become distorted, usually into something nauseating. Coffee smells like sewage. Chocolate smells like gasoline. Your partner’s skin smells like burnt rubber.
Smell training intervenes at precisely this point. By repeatedly presenting the same four odors, and crucially, by asking the patient to deliberately direct attention to each one, to recall its true character from memory even when current perception is distorted or absent, the protocol appears to provide a form of guided neuroplasticity. Repeated stimulation encourages regenerating neurons to find their correct glomerular targets. The attentional component may increase top-down neural feedback that helps strengthen correct connections and prune incorrect ones. Over weeks and months, the spatial map in the olfactory bulb is gradually restored.
The mechanism is not fully understood. No one has performed serial biopsies of human olfactory epithelium during smell training, for obvious reasons. But converging evidence from functional brain imaging, psychophysical testing, and animal models is compelling. Smell training accelerates and directs natural recovery.
One word buried in the protocol deserves more scrutiny than it usually receives. That word is attention.
Hummel’s instructions do not say: expose yourself to four smells twice a day. They say: focus on each smell. Concentrate. Try to recall what it should smell like. The distinction is not incidental. Multiple studies have shown that passive exposure to odors, ambient room scenting, for instance, produces significantly less improvement than the same exposure paired with deliberate, focused attention. The act of trying to smell, of directing consciousness toward the olfactory signal, appears to be pharmacologically active in a way that mere proximity to molecules is not.
This is a deeply strange result if you think of smell as a passive sense, which most people do. We tend to imagine olfaction as something that happens to us: a smell passes by, we register it, end of story. But neuroscience tells a different story. Olfactory perception is a construction, assembled in real time from the interplay of bottom-up sensory signals and top-down expectations, memories, and attentional states. When you concentrate on a smell, you are not simply receiving more of it. You are changing the neural calculation that turns a chemical signal into a percept.
Philosopher Alva Noë argued, in his 2004 book Action in Perception, that perception is not something organisms undergo but something they perform. Smell training is perhaps the most literal embodiment of this thesis in clinical medicine. The patient is not a passive recipient of a treatment administered by someone else. The patient is the treatment. Their attention is the active ingredient.
This is also, incidentally, why smell training is so difficult. Not physically: opening a pot and sniffing requires no special equipment or ability. But attentionally. Sustaining focused olfactory attention for even fifteen seconds is genuinely hard work for most people. The mind wanders. The visual system, accustomed to its dominance, reasserts itself. The temptation to go through the motions, to hold the pot under the nose while thinking of something else, is overwhelming. And going through the motions does not work nearly as well.
The post-Covid explosion of smell training revealed a vast, unmet hunger for olfactory education. Before 2020, the only people who deliberately trained their sense of smell were professionals: perfumers, flavorists, sommeliers, oenologists, tea tasters, and a handful of sensory scientists. These disciplines had always recognized that olfactory acuity is not a fixed trait but a skill, developed through years of systematic practice. A trainee perfumer spends months learning to identify raw materials blindfolded, not because their nose is anatomically different from anyone else’s, but because they have built, through repetition and attention, a cognitive architecture for discriminating and categorizing olfactory information.
Covid’s anosmia crisis democratized this knowledge. Suddenly, ordinary people were learning things perfumery students learn in their first year: that smell requires active engagement; that naming a smell helps you perceive it; that the same molecule can smell different depending on concentration, context, and expectation; that olfactory memory is more durable and more emotionally charged than visual or auditory memory; that the nose adapts quickly and needs rest between exposures; that on some days your sense of smell is sharper than on others, for reasons that remain largely mysterious.
A cottage industry materialized overnight. Smell training kits, small boxes containing the four canonical oils, appeared on Amazon, Etsy, and in pharmacies across Europe. Advocacy groups like AbScent in the UK, founded by smell-loss campaigner Chrissi Kelly, became lifelines for hundreds of thousands of people. Facebook groups swelled into mutual-aid communities where members tracked their progress in obsessive, moving detail. The medical establishment, which had never taken olfactory complaints particularly seriously (there is no olfactory equivalent of an audiologist, no insurance code for smell rehabilitation), was belatedly forced to pay attention.
Some converts to smell training went further. Having recovered their sense of smell, they did not stop training. They found that deliberate olfactory practice had sharpened their perception beyond their pre-illness baseline. They could detect subtleties they had never noticed before. Their vocabulary for describing smells had expanded. They had become, modestly but genuinely, more present to the olfactory world.
This may be the most interesting implication of the smell training story, and the least discussed in the medical literature. If directed olfactory attention can repair a damaged sense of smell, what can it do for an intact one?
The answer, based on decades of sensory science evidence, is: a great deal. Smell training studies in healthy subjects have shown improvements in odor discrimination, identification, and sensitivity. The gains are not enormous, and they require sustained effort. But they are real. The human nose is not a fixed instrument with immutable specifications. It behaves more like a muscle, or, more precisely, a neural network whose discriminating power increases with structured input and feedback.
The broader principle applies to every sensory modality but is most dramatic in olfaction, because smell is the sense we neglect most. We live in a culture that is visually saturated and olfactorily impoverished. We have a hundred words for colors and almost none for smells, a poverty that the synesthetes who see scents as colors navigate more fluently than the rest of us. We can describe the exact shade of blue in a painting but struggle to articulate the difference between two white wines. This is not a limitation of the nose. It is a limitation of attention.
Smell training, whether practiced by an anosmic patient with four pots of essential oil, or by a perfumer with a thousand raw materials, or by anyone who simply pauses to notice what the air actually smells like, is fundamentally an exercise in reversing this neglect. It is the practice of paying attention to information that was always there, arriving at the same neurons, triggering the same molecular cascades, but processed at the periphery of consciousness rather than at its center.
One final irony deserves note. Despite all the scientific validation of Hummel’s protocol, despite the randomized trials, the imaging studies, and the neurobiological models, the central mechanism of smell training is something humans have done for millennia without giving it this name. The incense rituals of ancient temples. The spice markets of the medieval trade routes. The garden walks prescribed against melancholy in eighteenth-century medicine. The sommelier who swirls a glass and inhales with eyes closed. All of these are, in their own way, structured practices of olfactory attention.
Thomas Hummel did not invent smell training. He measured it. He formalized it. He proved it. And in doing so, he gave a name and a mechanism to what the human olfactory system had always been waiting for: the simple, radical act of being asked to pay attention to what it was telling us.
It took a pandemic to make us listen.
Seven extraits at 20%, one collection. The Discovery Set brings all seven together in 2 ml.
