Start with the strangest fact about your own body. Smelling something is the only sense where the outside world makes physical contact with your nervous system. A molecule comes off a bottle, floats up your nose, and binds to a receptor on a nerve cell that is, in effect, part of your brain reaching out into the air.
The receptors, and a Nobel Prize
In 2004 Richard Axel and Linda Buck shared the Nobel Prize in Physiology or Medicine for discovering odorant receptors and how the olfactory system is organised.
The work was published in 1991, and they did it on rat olfactory tissue. What they found was a gene family of around 1,000 genes, described by the Nobel committee's own announcement as roughly three per cent of our genes, each giving rise to a different receptor type. That is an enormous share of an animal's genetic instructions spent on one sense.
Humans have fewer, because a number of those genes were lost over evolutionary time and no longer work. The figure commonly quoted for functional human olfactory receptors is around 400, though treat the precise number as approximate.
The important part is not the count. It is that no single receptor corresponds to a single smell. A molecule activates a pattern across many receptors, and your brain reads the pattern. It is closer to a chord than a note: many receptors firing together, not one receptor for one smell.
The equipment replaces itself
The cell that molecule landed on is a neuron, sitting in the roof of your nose, exposed to whatever you breathe: cold air, smoke, viruses, cleaning products. Nothing else in your nervous system is that vulnerable.
So it does something nothing else in your nervous system does. Olfactory receptor neurons die and are replaced continuously, with an average life span of about 30 to 60 days, regenerated from basal stem cells in the epithelium underneath. Over a year you cycle through several complete sets.
Smell loss after a virus is often recoverable, because the tissue can rebuild. That is the mechanism behind the smell-training programme in our guide to training your nose. The population of receptors you are smelling with today is also not quite the one you had two months ago.
Why smell hooks memory harder than anything else
This is the one piece of fragrance folklore that turns out to be true, and there is a structural reason for it.
Every other sense routes through the thalamus first, which acts as a relay station before information reaches the rest of the brain. Smell does not. The olfactory bulb connects directly to the amygdala and the hippocampus, which handle emotional response and memory formation.
So a smell arrives at the emotion and memory systems before you have consciously identified it. That is why the recognition often comes second: you feel something, then work out what you are smelling.
There is an evolutionary reason for the wiring. The limbic system, which governs emotion and memory, developed out of olfactory tissue. Smell did not get plugged into the emotional brain. The emotional brain grew out of smell.
The phenomenon gets called the Proust effect, after the passage where a taste triggers a flood of childhood memory. It is a real, documented pattern: scent-triggered memories tend to be more vivid and more emotionally loaded than those triggered by sight or sound.
How many smells can you tell apart? Nobody knows.
This is where we have to be straight with you about the state of the science, because the confident numbers you will read are all doing more work than they can support.
For decades the standard claim was that humans can distinguish about 10,000 odours. That figure was never well founded.
In 2014 a paper in Science by Bushdid and colleagues estimated that humans can discriminate more than one trillion olfactory stimuli, and the number went everywhere.
Then it got taken apart. Markus Meister at Caltech pointed out that the conclusion rested on 260 comparisons of pairs of smells, and argued the extrapolation was wrong by astronomical factors. He showed the same data could be read as consistent with humans discriminating as few as ten odours, which is not a serious claim about human ability but is a devastating one about the method.
Richard Gerkin and Jason Castro, publishing in the same journal as Meister's critique, showed the mathematical framework was fragile enough that small changes in the experiment produced almost any answer. With about a hundred more subjects it would have concluded humans can discriminate every possible odour. With a more conservative analysis it would have returned a figure close to the pre-2014 estimate.
That paper's title is the right answer, and we will borrow it: the number is still unknown.
If you see a fragrance article state a figure confidently, it has not read the argument.
Specific anosmia, and why people argue online
Anosmia is the inability to smell. It can be total, from injury, illness or age. Specific anosmia is different and much more common: you smell normally, but you are blind to one particular molecule or family of molecules.
Our materials guide gives the clearest example. Roughly one in five people is mildly anosmic to certain isomers of Iso E Super, one of the most widely used materials in modern perfumery. Musks are the other famous case, and there are many more.
And in one case the gene is known, which turns this from a curiosity into demonstrated mechanism. In 2007 a team including Andreas Keller, Hiroaki Matsunami and Leslie Vosshall published work in Nature on a single receptor, OR7D4. In the lab it responded to androstenone and androstadienone and to almost nothing else, staying silent against a panel of 64 other odours. Then they genotyped people. Subjects carrying particular OR7D4 variants were less sensitive to those two molecules and found them less unpleasant, and the variation in the gene accounted for a meaningful share of the variation in how intense and how pleasant people rated them.
This is the clearest single piece of evidence for it. Two people smell the same molecule and report different things because the difference is in a gene, not because one person imagined it, suggested it, or has a more sophisticated nose than the other.
One person in five cannot properly perceive a material that is doing structural work in an enormous number of fragrances. To them, a fragrance built on it smells thinner, shorter and quieter than it does to everyone else.
This resolves an argument you have certainly seen. Somebody says a fragrance is a beast that lasts twelve hours. Somebody else says it vanished in twenty minutes and must be fake or reformulated. Both are describing their own noses accurately, and neither is lying or wrong.
It also means you cannot fully trust a review, including ours, and you cannot fully trust a recommendation from a friend. The apparatus differs between people at the hardware level.
Our guide to the perfumers has the extreme case: Jean Carles went anosmic and kept composing, building Ma Griffe and Miss Dior from a mental model of what materials do, with his son smelling the trials.
Adaptation, which is not the same thing
Your nose stops registering a constant smell within minutes. This is olfactory adaptation, it is normal, and it is not your fragrance disappearing.
You stop smelling your own fragrance long before other people do. The urge to reapply because "it wore off" is usually your own adaptation, and the result is that everyone around you is getting considerably more than you think. Our guide to wearing fragrance covers the etiquette of that.
You cannot evaluate many fragrances in one session. Our guide to testing perfume properly is entirely about this, and it is the practical guide that follows directly from this one.
Taste is mostly smell, and the number everyone quotes is made up
You have read that most of taste is really smell, often given as a specific percentage in the high double digits. It is the kind of fact that sounds settled.
The mechanism is real. Chewing pushes volatile compounds up the back of the throat into the nasal cavity, which is retronasal olfaction, and it is why food goes flat when you have a cold and why a gourmand fragrance can read as edible when nothing about it is.
The specific percentage is not real. Charles Spence went looking for the source of the commonly quoted figure and found that the study usually credited with it does not support it, and that much of the confusion comes from people using "taste" to mean gustation in one sentence and flavour in the next.
We are including this because it is the same failure as the trillion odours, one section up, and the same failure as the freezing-point myth in our storage and care guide. A number gets stated, gets copied, and acquires authority from repetition alone. This field generates those constantly, and that pattern matters more than any individual figure above.
What to do with all this
1. When a review and your nose disagree, believe your nose. Not as self-affirmation. As a claim about hardware, per the OR7D4 finding.
2. Before you decide a fragrance has no projection, ask somebody else. You adapted to it in minutes and they did not. This one costs nothing and prevents the most common over-application mistake in the hobby.
3. Treat any confident number about smell as unverified until you have chased it. Ten thousand odours, one trillion odours, a fixed percentage of taste. All three fell apart the same way once someone checked the source.
4. If you have lost some sense of smell, do not assume it is permanent. The tissue regenerates. Our training-your-nose guide has the protocol, and it is the one piece of advice in this manual that is properly clinical rather than merely useful.
So when a page on our site tells you a fragrance is beautiful, or loud, or dull, understand that we are reporting what a particular set of receptors did with it. You are allowed to disagree, and you may be right. Our guide to fragrance myths takes on more of this territory; this is the one place where the correct answer really is that it depends on you.
Part of Nose to Bottle, our complete fragrance manual. Read it in order here.