Your DNA contains roughly 1,000 genes dedicated to smell. Over 600 of them are broken. Dead code, accumulated mutations rendering them nonfunctional. In mice, only about 20% of olfactory receptor genes have gone silent. In humans, the number is over 60%. We carry more broken smell genes than working ones.
The leading hypothesis for why: as our primate ancestors developed complex trichromatic color vision, the survival pressure that kept olfactory genes functional relaxed. We could see ripe fruit instead of smelling it. We could spot predators instead of scenting them. Over millions of years, the unused genes quietly decayed. Humans traded smell for sight, and the evidence of that trade is written in every cell of your body.
What remains, those roughly 400 working receptors, constitutes one of the least understood sensory systems in biology. The mechanism by which your nose detects an odor is the subject of a genuine, unresolved scientific dispute. And the things smell can do, triggering memories more vivid than any other sense, detecting disease before medical instruments can, influencing who you find attractive without your knowledge, make the sense we are losing one of the most interesting ones we still have.
How Your Nose Actually Works
In 1991, Richard Axel at Columbia University and Linda Buck published a paper in the journal Cell that transformed the field. They discovered that roughly 3% of the entire human genome was devoted to olfactory receptor genes. This was an enormous genetic investment in a single sense. They received the Nobel Prize in Physiology or Medicine in 2004.
Here is how the system works. Each olfactory receptor neuron in your nasal epithelium (a small patch of tissue high in the nasal cavity) expresses exactly one type of receptor protein. When an odor molecule drifts in and binds to a matching receptor, it triggers a nerve impulse. But most odors are not a single molecule. Coffee, for example, contains over 800 volatile compounds. Each compound activates a different combination of receptors, creating a unique activation pattern across the ~400 receptor types. This is the combinatorial code: like a barcode, where the pattern of activated and non-activated receptors identifies the smell.
The system is elegant. But there is something even more remarkable about it.
Smell is the only sense that bypasses the thalamus. Every other sensory input, vision, hearing, touch, taste, passes through the thalamus first, the brain’s central relay station, before being sent to the relevant processing area. Olfactory signals skip this step entirely. They project directly to the amygdala (emotion processing), the hippocampus (memory formation), and the piriform cortex (primary olfactory cortex). No other sense has this kind of direct line to the emotional and memory centers of the brain. This is not a metaphor. It is wiring.
One more thing about the numbers. In 2014, a team at Rockefeller University led by Andreas Keller published a study claiming humans could distinguish over one trillion distinct odors. The finding made global headlines. The following year, Gerkin and Castro at Arizona State University published a rebuttal in eLife showing that the mathematical framework was deeply flawed. Different assumptions plugged into the same model produced estimates ranging from 5,000 to 10 to the power of 29. The original study had tested subjects on only 148 odor pairs and extrapolated from there.
The honest answer, as of today: nobody knows how many distinct smells the human nose can detect. The number is genuinely unknown.
Two Thousand Years of Wrong
Before modern chemistry existed, smell was not just a sense. It was a theory of disease.
In ancient Egypt, priests burned three different incenses daily in the temples: frankincense at dawn, myrrh at midday, and kyphi (a compound incense made from honey, wine, raisins, resins, and imported spices) at dusk. Kyphi was also consumed as medicine, believed to cleanse the body and bring restful sleep with vivid dreams. The Ebers Papyrus (c. 1550 BCE) contains dozens of formulations using aromatic substances for healing. For the connection between ancient aromatics and healing traditions, see our article on the magical properties of frankincense and myrrh.
The Greeks formalized this into a system. Hippocrates in the 5th century BCE and Galen in the 2nd century CE developed the miasma theory: disease was caused by “bad air” identifiable by foul smell. In this framework, you did not just detect sickness by smell. The smell WAS the sickness. This is not a subtle distinction. For over two thousand years, the dominant medical theory in the Western world held that poisonous vapors, identifiable by their odor, were the direct cause of epidemic disease. The word “malaria” is Medieval Italian for “bad air.”
The famous plague doctor mask, the one with the long beak, is often associated with the Black Death of the 14th century. It wasn’t. The beaked mask was first documented in 1619, designed by French physician Charles de Lorme during a plague outbreak in Paris. The beak was stuffed with dried flowers, herbs, camphor, vinegar-soaked sponge, juniper berries, ambergris, cloves, myrrh, and storax. De Lorme shaped the beak specifically to give inhaled air enough time to be “purified” by the aromatics before reaching the nose.
The mask did nothing against plague bacteria. But here is the twist: some of the herbs it contained, particularly camphor and certain aromatic compounds, may have incidentally repelled fleas. Fleas were the actual vectors carrying Yersinia pestis, the plague bacterium. The mask was designed according to a wrong theory and it may have worked for a completely different reason than anyone understood at the time.
The miasma theory was not abandoned until after the 1880s, when germ theory finally replaced it. Two thousand years of medical practice, built on the idea that smell was the mechanism of contagion, dissolved in a generation.
The Quantum Nose
The mainstream explanation for how olfactory receptors identify a molecule is the shape theory: the molecule fits the receptor like a key in a lock. Its three-dimensional shape triggers the signal. This is clean, intuitive, and incomplete.
The problem: there are molecules with nearly identical shapes that smell completely different. And there are molecules with very different shapes that smell the same. If shape were the whole story, this shouldn’t happen.
In 1996, biophysicist Luca Turin published a paper in Chemical Senses proposing an alternative: the vibration theory of olfaction. He suggested that olfactory receptors detect molecular vibrations through a quantum mechanical process called inelastic electron tunneling. In his model, the molecule must first fit the receptor binding site (shape still matters), but then its vibrational frequency determines what signal is sent. He called it a “swipe card” rather than a “lock and key.” The card has to be the right size, but it also has to carry the right information.
The key test: replace hydrogen atoms in a molecule with deuterium. Deuterium has the same number of electrons and nearly the same shape as hydrogen, but twice the mass. This changes the molecule’s vibrational frequency without changing its shape. If the vibration theory is correct, deuterated molecules should smell different despite being structurally identical.
In 2013, Turin’s team published results in PLOS ONE showing that humans could distinguish deuterated musks from regular musks in double-blind tests. Fruit flies showed the same ability in separate experiments.
The rebuttal came from Leslie Vosshall’s lab at Rockefeller University in 2015. When they tested highly purified musk isotopomers against nine different olfactory receptors in cell culture, they found no difference in activation. They concluded the vibrational theory was “implausible.”
Turin’s supporters responded that testing isolated receptors in a dish may not replicate what happens in a living nose. The tunneling mechanism might require the intact cellular environment.
This is where the debate stands. The shape theory cannot explain all the data. The vibration theory cannot survive the receptor-level tests. Some researchers have proposed that both mechanisms operate simultaneously. Nobody has settled it. The basic question of how your nose identifies a molecule remains genuinely open.
Why a Song Can’t Make You Cry Like a Smell Can
Marcel Proust described it in 1913. He dipped a madeleine in tea and was flooded by memories of his aunt’s house in Combray. The experience was so vivid, so emotionally overwhelming, that it became the most famous passage in 20th-century French literature. Neuroscience has since confirmed that Proust was describing something real, and the explanation goes back to that thalamic bypass.
Rachel Herz, a psychologist at Brown University, put the Proust phenomenon in an fMRI machine. When subjects recalled memories triggered by personally significant odors versus visual or auditory cues, the fMRI scans showed significantly greater activation in the amygdala and hippocampal regions. The emotional intensity was measurably higher for smell-triggered memories. And those memories showed a distinct pattern: they tend to come from the first decade of life, a “bump” that other senses don’t produce as strongly.
The explanation is the wiring. Because olfactory signals skip the thalamus and go directly to the brain’s emotion and memory centers, smell-triggered memories arrive with their emotional charge intact. Other senses go through an additional processing step that dilutes the raw emotional impact.
But here is the part that surprised researchers most.
The Jahai people, hunter-gatherers of the Malay Peninsula, have approximately 12 distinct abstract words for smell. Not descriptions by reference (“smells like lemon”) but true abstract categories. The word ltpit, for example, covers the smell of various flowers, ripe fruit, perfume, soap, and bearcat. The Jahai find naming odors exactly as easy as naming colors.
English speakers, by contrast, are terrible at it. Lab tests show remarkably low agreement when English speakers try to name the same smell. We resort to source descriptions because we lack the vocabulary. And the kicker: this appears to be entirely cultural, not biological. The Semaq Beri, another hunter-gatherer group in the same region, show the same rich olfactory vocabulary. The non-hunter-gatherer Semelai nearby do not. It is the lifestyle, not the genetics, that determines whether a culture develops smell words.
Industrialized humans did not lose the nose. We lost the words.
The Diagnostic Nose
In 2012, a retired Scottish nurse named Joy Milne mentioned to researchers at the University of Edinburgh that she had noticed her husband Les developing a new, musky odor years before he was diagnosed with Parkinson’s disease. She had hereditary hyperosmia, an unusually heightened sense of smell. The researchers, intrigued, tested her.
They gave her twelve T-shirts, six worn by Parkinson’s patients and six by healthy controls. She correctly identified all six patients. She also flagged one of the control subjects as positive.
That person was diagnosed with Parkinson’s eight months later. Milne had smelled the disease before any medical instrument or clinical examination could detect it.
The odor, researchers at the University of Manchester later determined, comes from changes in sebum, the oily substance produced by skin glands. Altered cell metabolism in Parkinson’s patients produces a distinct signature of volatile organic compounds (VOCs) detectable in skin swabs. The team developed a diagnostic test based on this discovery, achieving 95% accuracy under laboratory conditions.
Dogs, with roughly 300 million olfactory receptors compared to our 6 million, take this further. Trained medical detection dogs can identify cancer in blood samples with approximately 97% accuracy. Published detection rates include breast cancer at 88% sensitivity and 98% specificity from breath, colorectal cancer at 91% sensitivity and 99% specificity, and ovarian cancer at 97% sensitivity and 99% specificity. Dogs have also been trained to detect malaria, diabetes, and epileptic seizures before they occur.
Ancient physicians used smell diagnostically for millennia, even though their theoretical framework (miasma) was wrong. Modern science is circling back to the same practice with a different explanation. The nose knew something. It just took 2,000 years to understand what.
The COVID Catastrophe
In 2020, smell became global news for a terrible reason. Approximately 62% of COVID-19 patients experienced anosmia (loss of smell), and in 11.8% of cases it was the first symptom, sometimes the only symptom. With cumulative infections exceeding 250 million worldwide, an estimated 10 million people developed lasting smell loss.
What COVID-19 taught neuroscience was, in a grim way, invaluable. Olfactory sensory neurons are the only neurons in the mammalian nervous system that naturally regenerate and project to specific brain targets. They are directly exposed to the environment (they contact inhaled air), which is why they need to regenerate. The virus damaged not just the neurons themselves but also the supporting sustentacular cells in the olfactory epithelium.
Most people recovered. But many developed something worse than absence: parosmia, distorted smell perception. Coffee smelling like sewage. Onions smelling like rotting meat. The onset typically came about 2.5 months after infection, during the regeneration phase. The mechanism: as damaged olfactory neurons regrew, they sometimes reconnected to the wrong glomeruli in the olfactory bulb. The wiring was restored, but to the wrong targets. The brain received signals, but wrong ones.
The standard treatment protocol was developed in 2009 by Dr. Thomas Hummel at the University of Dresden, years before COVID made it urgent. Four essential oils: rose, lemon, clove, and eucalyptus. These four were chosen based on Hans Henning’s 1916 “odor prism” categorization of primary odor qualities: floral, fruity, spicy, resinous. The protocol is simple: sniff each oil for 10-20 seconds, twice daily, for at least 12 weeks. It works like physical therapy for the nose. Repeated stimulation strengthens the olfactory pathway. For more on essential oils and their properties, see our article on essential oils.
About 7% of affected patients remain anosmic after 12 months. For millions of people, the pandemic permanently altered their relationship with one of their oldest senses.
The Secret Nose
Your nose does things you do not know about.
In 1995, biologist Claus Wedekind at the University of Bern conducted what became known as the sweaty T-shirt experiment. Forty-four men wore T-shirts for two days. Forty-nine women then rated the odors. The result: women consistently preferred the smell of men whose MHC (major histocompatibility complex) genes were most different from their own. MHC diversity in offspring produces stronger, more versatile immune systems. The nose was selecting for immunological compatibility without the conscious mind knowing it.
The finding that drew the most attention: women on oral contraceptives showed reversed preferences. They favored MHC-similar men, the opposite of the natural pattern. Women described MHC-dissimilar scents as reminiscent of current or past boyfriends. MHC-similar scents reminded them of their fathers or brothers. The implications for partner selection remain debated, but the data is clear.
Whether this constitutes “pheromone” signaling is an open question. Most mammals have a vomeronasal organ (VNO) specifically for detecting pheromones. Humans have a physical remnant of the VNO, but no active sensory neurons have been found in it, no nerve connections to the brain have been identified, and the key VNO genes are pseudogenes. The hardware appears to be dead. But the behavioral effect is documented. Whether humans accomplish this through pheromone signaling or through standard olfaction of body chemistry is unknown.
And the nose has deeper secrets still. Cadaverine and putrescine, chemicals produced by bacterial decomposition of dead tissue, trigger an innate avoidance response that researchers estimate evolved approximately 420 million years ago. The response is mediated by trace amine-associated receptors (TAARs), a receptor system that triggers behavioral responses without learning. No one teaches you to recoil from the smell of death. The wiring is older than terrestrial vertebrates.
You also smell in stereo, though you cannot report it. A 2020 study published in PNAS confirmed that humans use their two nostrils independently, sampling odor concentrations on each side and using the difference for spatial navigation. Block one nostril and your ability to locate an odor source degrades. But when asked, subjects cannot consciously identify which nostril detects a stronger signal. Stereo olfaction operates entirely below awareness.
What We Don’t Know
The vocabulary problem is not just a linguistic curiosity. It points at something deeper. Industrialized cultures have systematically deprioritized smell. We have hundreds of words for colors, dozens for textures, extensive musical terminology. For smell, we have almost nothing. We describe odors by pointing at their sources. “It smells like cinnamon.” “It smells like rain.” No abstract categories. The Jahai have them. We don’t. And the difference correlates with lifestyle, not biology. When the hunting-gathering way of life disappears, the smell words disappear with it.
The shape-versus-vibration debate remains open. Shape theory cannot explain all the anomalies. Vibration theory cannot survive receptor-level testing. Both camps have published in major journals. Neither has won.
Olfactory adaptation remains poorly understood in its extremes. You stop smelling your own house within minutes, but you never stop seeing your furniture. Both are constant stimuli. Both are processed by the brain. Why does the olfactory system shut down so aggressively? The peripheral mechanism involves calcium feedback loops that literally desensitize the receptor cell. The central mechanism involves the hippocampus dampening signals from the piriform cortex. The functional purpose is clear: filter out the constant so you can detect the novel. But why smell does this so much more than vision or hearing has no complete explanation.
And then there is the graveyard itself. Six hundred dead genes. What could we smell if they still worked? We know they accumulated mutations faster in humans than in any other primate, and that the acceleration correlates with the development of trichromatic vision. But correlation is not mechanism. And we have no way to reconstruct what those lost receptors once detected. The information is gone. We can only know that the sense we have now is a fraction of what it was.
The science of smell is not a solved field. It is a field where the most basic question, how does the nose identify a molecule, does not have a confirmed answer. Where the number of distinguishable odors is unknown. Where a retired nurse in Scotland can outperform a hospital’s diagnostic equipment. Where a chemical that triggers panic in humans has been doing so for 420 million years, through an unbroken chain of receptor inheritance that predates the dinosaurs.
What remains of this ancient sense is more interesting than most people realize. And we are still losing it.
Sources
Key scientific papers:
- Richard Axel and Linda Buck, “A novel multigene family may encode odorant receptors,” Cell 65 (1991)
- Andreas Keller et al., “An olfactory demultiplexer,” Science 343 (2014) [trillion smells claim]
- Richard Gerkin and Jason Castro, “The number of olfactory stimuli that humans can discriminate is still unknown,” eLife 4 (2015) [rebuttal]
- Luca Turin, “A spectroscopic mechanism for primary olfactory reception,” Chemical Senses 21 (1996)
- Turin et al., “Plausibility of the vibrational theory of olfaction,” PNAS 112 (2015)
- Block et al., “Implausibility of the vibrational theory of olfaction,” PNAS 112 (2015) [Vosshall rebuttal]
- Claus Wedekind et al., “MHC-dependent mate preferences in humans,” Proceedings of the Royal Society B 260 (1995)
- Rachel Herz, “Testing the Proustian Hypothesis,” Chemical Senses (2004)
- Asifa Majid and Niclas Burenhult, “Odors are expressible in language,” Cognition 130 (2014) [Jahai smell vocabulary]
- Thomas Hummel et al., “Effects of olfactory training in patients with olfactory loss,” Laryngoscope 119 (2009)
- Nobel Prize in Physiology or Medicine 2004
Historical sources:
- Plutarch, De Iside et Osiride [on Egyptian kyphi practices]
- Ebers Papyrus (c. 1550 BCE)
- Hans Henning, Der Geruch (1916) [odor prism classification]
Medical and diagnostic:
- Trivedi et al., “Discovery of Volatile Biomarkers of Parkinson’s Disease from Sebum,” ACS Central Science 5 (2019) [Joy Milne research]
- Hackner et al., “Canine olfaction in cancer diagnosis,” Journal of Breath Research (2022)
- Parosmia and COVID-19 olfactory dysfunction reviews in Frontiers in Neural Circuits (2025)
Archaeological evidence:
- Plague doctor costume, first documented Charles de Lorme, 1619
- Kyphi incense recipes from Egyptian temple inscriptions
Related articles on this site:
