Pinch your nose shut, take a sip of a coffee that is supposed to taste like blueberry, and the blueberry disappears. Not faint. Gone. What is left is sweetness, a little sharpness, weight on the tongue, the whole vocabulary the mouth actually owns, and none of it is blueberry. Let go of your nose and the fruit comes back inside a second, because nothing about the coffee changed between the two sips. What changed is whether a specific, narrow channel of smell was open.
That channel is called retronasal smell, and it is where almost every fruit, floral, or chocolate note on a bag of coffee actually comes from. The tongue only ever reports sweet, sour, bitter, salty, and something close to umami. Blueberry, cherry, sarsaparilla, dark chocolate: those are smells, arriving by the back door, and the only reason they feel like taste is a trick of where your brain decides to put them.
Is the Fruit You're Tasting Actually a Smell?
Mostly, yes, and coffee writing already half-says so without finishing the thought. Our own piece on what a tasting note means calls a note "a comparison," something the cup reminded a taster of, and that is accurate as far as it goes. What it does not say is which sense is doing the comparing, or that the 110 attributes in the World Coffee Research Sensory Lexicon that piece cites are mostly smell words wearing a taste word's job title.
Smell reaches your olfactory receptors two different ways, and they are treated as two separate senses by the people who study them, not two versions of one sense. Orthonasal smell is sniffing: air comes in through the front of the nose before anything is in your mouth. Retronasal smell runs the opposite direction. You swallow, warm air carrying volatile compounds from the coffee gets pushed up from the back of the throat into the same receptors, on the way out rather than the way in. Achilles Coffee Roasters lays out that split clearly, and the credit belongs there: orthonasal smell is what you get sniffing a bag before you brew it, and retronasal smell is what arrives once you have swallowed, which is where most of what a tasting note actually describes lives.
The Brazilian flavor scientist Fabiana Carvalho runs a classroom demonstration for exactly this reason, and it is the one at the top of this article. Sprudge's profile of her work puts her estimate at 75 to 95 percent of what people call taste actually coming from retronasal olfaction, and describes the test she gives students: chew something with your nose pinched shut and you get almost nothing but the tongue's handful of qualities. Release your nose and the rest floods in immediately. Nothing in the mouth changed between the two bites. The channel did, and the effect is large enough that somebody built a whole teaching exercise around it.
Why Does a Smell Feel Like It's Happening in Your Mouth?
This is the part missing from every coffee source I could find, including the two credited above, and it is worth sitting with because the answer is genuinely strange: you are smelling with the back of your throat, and your brain tells you it is your tongue.
The psychologist Paul Rozin described the split formally in 1982, with an experiment that is still cited as the founding demonstration. He trained people to recognize a set of common odors presented the normal way, through the nose, sniffing. Once they could identify those smells reliably, he introduced the identical odors retronasally, delivered near the back of the mouth instead. Accuracy fell sharply. The same molecules, reaching the same receptors by the other route, stopped being recognizable as the smells people had just learned. That is the experiment the whole retronasal/orthonasal distinction in flavor science traces back to, and it is cited that way by the researchers I was able to read directly.
The finding that matters most for a coffee bag is not just that the two routes differ in accuracy. It is where the brain decides each one is happening. Researchers studying smell and taste integration cite Rozin's conclusion directly: orthonasal odors get located out in the world, where the sniffed object actually is, while retronasal odors get located inside the mouth, so the brain mistakes retronasal olfaction for taste. I have not been able to read Rozin's original 1982 paper in Perception & Psychophysics myself. It sits behind a journal paywall I could not get past, and I would rather tell you that than cite a paper I did not open. What follows instead is sourced to a 2017 study that quotes his conclusion directly and builds an experiment on top of it, which I did read in full.
That 2017 paper, published in Scientific Reports by Kakutani and colleagues, states the mechanism as plainly as I have found it stated anywhere: "Rozin asserted that orthonasal odors tend to be localized to the external world and retronasal odors tend to be localized to the oral cavity, leading humans to mistake retronasal olfaction for 'taste.'" Your nose is doing the work. Your brain is filing the result under your tongue, because that is where the sensation seems to originate once you have swallowed, and that filing error is permanent and shared by everyone who tastes coffee, not a trick a particular person falls for and a more careful taster avoids.
Does It Matter Whether You Smell Before or After You Swallow?
According to the same study, yes, and the timing is specific rather than a general rule of thumb. Kakutani and colleagues ran an experiment giving people a sweet solution alongside a vanilla odor, varying whether the vanilla arrived orthonasally before drinking or retronasally after. Only one order changed anything measurable: "a retronasal vanilla odor after one-shot drinking enhanced a sweet taste, but an orthonasal vanilla odor before one-shot drinking did not." The paper also reports that taste intensity under the retronasal condition ran stronger than under either the odorless condition or the orthonasal one.
Smelling the coffee first, which is the ordinary advice printed on half the brewing guides in this field, primes expectation. It is not the same event as the retronasal smell that actually arrives once you swallow, and that second event is the one doing most of the flavor work. The researchers' own explanation for why only that order works is about pattern-matching against how drinking actually happens in the body: smell arriving in the sequence a real sip produces it, taste first and smell rushing up right behind it, reads as more intense than the identical odor presented out of that order. Their phrase for it is that congruency with the natural pattern of taste and odor timing, synchronized with the act of drinking, is what determines whether the odor enhances the taste at all. Coffee tasting notes describe a sensation your body is primed to register in one specific order and not the other, and nothing I found in coffee writing says so.
Why Hasn't Coffee Writing Said This Part Out Loud?
Credit where it belongs, and a concession where it is owed. Achilles Coffee Roasters correctly separates orthonasal from retronasal smell, correctly says the tongue only contributes sweetness, sourness, bitterness, saltiness and umami, and correctly says the brain links retronasal signals to memory and learned association rather than to anything physically added to the coffee. Carvalho's work, as Sprudge covers it, correctly puts the scale of the effect in front of a reader with a demonstration anyone can run at home with a piece of candy, and goes further than most coffee writing by naming environmental factors, cup shape, color, even background noise, as inputs the brain folds into the same flavor judgment.
Neither source, in what I was able to read, explains why the misattribution happens: why a retronasal smell gets filed under taste instead of under smell, rather than just noting that it does. That is Rozin's half of the story, and it is the half that turns "smell matters a lot" from a surprising statistic into an actual, named mechanism with a direction and a cause.
There is a cost to saying this plainly, and it lands on the people who write tasting notes for a living, us included. If a note is really a label for a retronasal smell that the brain has already decided to file under taste, then every note we print on a bag is making a claim about where a drinker's attention should go after the swallow, not a claim about some fruit hiding in the cup waiting to be found. We keep printing the notes anyway, because a buyer deciding between two coffees before brewing either one needs some vocabulary to compare them, and "tastes like coffee" tells nobody anything useful. What changes is how the note should be read: as a prediction about a sensation your own retronasal passage is about to produce, not a guarantee sitting in the bag independent of the person drinking it.
Our own writing has the identical gap. The piece we wrote on why coffee smells like coffee is entirely about orthonasal aroma, the compounds you can smell straight off the grounds before you ever brew them, and it says plainly that it is staying on that side of the subject on purpose. It never uses the word retronasal. Read the two pieces together and they cover different halves of the same cup, and until now neither one said so.
Does This Change How You Should Read a Tasting Note?
A little, and here is the honest version of what it changes. It does not make a note list more precise, and it does not make one roaster's five notes more trustworthy than another roaster's three. It tells you which sense actually wrote the list. When a bag says a coffee carries dried strawberry, dark chocolate, berry jam, sarsaparilla and banana milk, that description is aimed at the back of your throat, not the front of your nose, and reading the list before you brew primes a comparison your palate has not made yet, in the order the 2017 study found does not actually help.
Specialty cupping already exploits this mechanism without ever naming it. Evaluators slurp a sample hard enough to spray it across the palate and force vapor up into the retronasal passage in one motion, which is a deliberate, maximal version of the exact swallow-triggered smell this whole article is about. The Specialty Coffee Association's current Coffee Value Assessment, which we cover in our own piece on what a cupping score is measured against, splits the procedure into Physical, Descriptive, Affective and Extrinsic assessments, and the Descriptive half is substantially a retronasal exercise conducted on purpose, under conditions designed to make the effect as large and as repeatable as possible. Nothing in that SCA framework names the mechanism either, and it should have.
If you want to run Carvalho's test yourself rather than take it from a magazine profile, our Organic Peru Decaf is a fair coffee to do it with, because its published notes are specific enough to fail obviously if the mechanism is not real: dried strawberry, dark chocolate, berry jam, sarsaparilla and banana milk. Pinch your nose through the first sip and name what is left, using only sweet, sour, bitter, salty, or something close to umami. Let go on the second sip and see how much of that list reappears on its own, in roughly the order your swallow produced it. We roast it on Wednesdays, and the notes on its page are the ones we are asking you to test the mechanism against, not just take on faith.
Questions we get about tasting coffee and smell
Is taste mostly smell?
For anything beyond sweet, sour, bitter, salty, and something close to umami, yes. The tongue detects those few qualities directly. Everything more specific that a tasting note names, fruit, chocolate, floral, spice, arrives through the sense of smell, mainly through the retronasal route that opens once you swallow.
What is the difference between orthonasal and retronasal smell?
Orthonasal smell is sniffing: air carrying odor molecules enters through the front of the nose before anything is in the mouth. Retronasal smell works in the opposite direction, carrying odor molecules from something already in the mouth up through the back of the throat into the same receptors during swallowing.
Why does a smell feel like it is happening in my mouth instead of my nose?
Because of where the brain locates the two kinds of smell. Researchers describe orthonasal odors as getting located out in the world, at the object being sniffed, while retronasal odors get located inside the mouth, which is why a smell arriving through the back of the throat gets experienced as taste rather than as smell.
Does smelling a coffee before you drink it change how it tastes?
It changes your expectation, but a 2017 study found the measurable flavor-enhancing effect came specifically from odor arriving after a sip, in the same order a real swallow produces it, not from the identical odor presented beforehand through the nose alone.
Both pieces this one sits beside, the one on what a tasting note means and the one on why coffee smells like coffee, are both about where the sensation comes from. This one is about where your brain decides it is happening once it arrives. All three are honestly the same cup, looked at through a different door, and none of them was complete on its own.
