A woman in London, a research subject, anonymized in the literature as "S.J.", who sees color every time she inhales. Coffee is a deep, shifting burgundy. Fresh-cut grass pulses in bands of electric lime and gold. The scent of her mother's kitchen, a complex layering of cardamom and ghee and warm bread, produces a visual field she describes as "amber with moving red threads, like watching embers through honey."
9 min read
S.J. is not being poetic. She is being clinical. She has olfactory-visual synesthesia, a neurological condition in which stimulation of one sensory pathway, smell, produces automatic, involuntary experiences in another, sight. When she walks through a garden, she does not merely smell the flowers. She watches them, eyes closed, as they bloom in colors that have nothing to do with petals.
She is not alone. And she is not ill. She is simply operating a version of the human brain in which the curtain between the senses has been left open.
Synesthesia prevalence and its many forms
Synesthesia, from the Greek syn (together) and aisthēsis (sensation), affects an estimated 4% of the general population according to a large-scale prevalence study by Julia Simner and colleagues at the University of Edinburgh published in Perception in 2006, though prevalence varies depending on who is counting and how strictly they define the phenomenon. The most commonly studied form is grapheme-color synesthesia, in which letters or numbers evoke specific colors: the number 5 is always green, the letter A is always red, and these associations are as involuntary and consistent as the color of the sky. Test a grapheme-color synesthete on their associations, wait a year, test again. The colors hold. They are not metaphors. They are perceptions.
Olfactory-visual synesthesia, the variety S.J. lives with, is rarer and more difficult to study, for reasons that illuminate something important about how science treats smell. Vision dominates Western neuroscience. We have standardized color charts, luminance scales, spatial frequency measures. We can describe a visual stimulus with mathematical precision. Smell resists this. There is no periodic table of odors, no wavelength that corresponds to "rose" the way 620 nanometers corresponds to "red." Olfaction is the anarchist sense, processed through the limbic system, the oldest, most emotionally entangled architecture of the brain, the same direct pathway that makes olfactory memory so emotionally vivid yet factually unreliable, and it has always made researchers slightly uncomfortable. It is too subjective, too tied to memory and feeling, too resistant to the controlled conditions that make for clean data.
This discomfort has consequences. It means that olfactory synesthesia is documented but understudied, acknowledged but poorly mapped. The handful of case studies are vivid: a man in Germany who sees geometric patterns when he smells spices, a woman in California for whom lavender is always a specific shade of periwinkle blue. The associations are consistent, automatic, and impossible to suppress. They are real in every sense that neuroscience uses the word.
But here is the stranger fact, the one that should stop us: the rest of us, the 96% without clinical synesthesia, are already halfway there.
Cross-modal perception at Oxford's research lab
In 2010, the experimental psychologist Charles Spence and his colleagues at Oxford's Crossmodal Research Laboratory published a series of studies that should have rewritten the way we think about sensory boundaries. They demonstrated that the human brain does not process senses in isolation. It processes them in conversation. Present someone with a sweet taste and they will judge a concurrent tone as higher-pitched. Show someone a red visual field and they will rate an accompanying odor as warmer. These are not synesthetic experiences (the participants were neurotypical), but they are cross-modal correspondences, deep statistical regularities in how the brain maps one sense onto another.
The implications are sweeping. The brain does not build separate models of the world for each sense and then stitch them together at the end, like a film editor syncing audio to video. It builds one model, using all available channels simultaneously, and the channels are not as separate as we pretend. There are direct neural pathways between olfactory cortex and visual cortex. There are shared processing regions in the orbitofrontal cortex where smell and taste and texture converge into the single, unified experience we call flavor. The borders between senses are not walls. They are curtains, thin, permeable, and in some brains, perpetually drawn back.
This is not new knowledge. It is new science confirming very old intuitions.
Arthur Rimbaud, in his 1871 poem Voyelles, assigned colors to vowels: A was black, E was white, I was red, O was blue, U was green. The poem has been debated for over a century. Was Rimbaud a synesthete? Was he performing a literary exercise? The answer matters less than the fact that the poem works, that readers encounter it and feel the rightness of the associations, even when they cannot explain why. The open, dark mouth-shape of "A" does feel black; the tight, bright articulation of "I" does feel red. Rimbaud was mapping cross-modal correspondences before neuroscience had a name for them.
Wassily Kandinsky painted music. His canvases were attempts to render sound in visual form, not illustrations of musical scenes, but direct translations of auditory experience into color and shape. He described a trumpet as "a high-pitched yellow" in his 1911 treatise Concerning the Spiritual in Art and believed that art should aspire to the condition of music precisely because music was already abstract, already operating in the space between defined sensory categories. Whether Kandinsky had clinical synesthesia or simply a deep sensitivity to cross-modal resonance, his work demonstrates that the space between senses is not empty. It is a creative territory, and those who inhabit it, by neurology or by training, perceive things the rest of us can only describe by analogy.
Perfumery vocabulary is already synesthetic
Now consider the vocabulary of perfumery.
A citrus note is "bright." An oud is "dark." Vanilla is "warm." Galbanum is "green." Iris is "powdery", a tactile word for a gaseous experience. A well-constructed fragrance has "depth" and "height," spatial metaphors for something that occupies no space at all. We speak of "sharp" aldehydes and "smooth" musks, of "transparent" florals and "opaque" resins. We describe certain accords as "loud" and others as "quiet." We talk about fragrances that are "round."
This is not the sloppy language of marketing. This is the working vocabulary of the trade, the words that perfumers use in the lab, that evaluators use in assessment sessions, that raw material suppliers print in their technical bulletins. It is the shared lexicon without which the profession could not function, because smell has no dedicated vocabulary of its own. Unlike color, which has red, blue, green, primary terms that refer to nothing else, smell borrows all of its language from other senses. The lexicon of absolutes, concretes, and resinoids is itself a vocabulary of borrowed analogies. It is a sense that speaks only in translation.
And the translations are not arbitrary. When Spence's team tested whether people associate lemon scent with yellow, the agreement was near-universal, cutting across cultures and languages. Cinnamon is red-brown. Mint is green or blue. These are not random pairings; they reflect deep regularities in the environment (lemons are yellow, mint leaves are green) that the brain has internalized as cross-modal expectations. But the associations go further than simple co-occurrence. People reliably rate vanilla as "warm" even in cultures where vanilla is not associated with warm beverages. They rate citrus as "bright" even in the dark. The cross-modal mapping is built into the architecture, not learned from the label.
This means that perfumery's synesthetic language is not a failure of descriptive precision. It is a success of perceptual honesty. When a perfumer calls a note "bright," she is not reaching for a metaphor because the literal word escapes her. She is reporting a genuine cross-modal perception, a correspondence between the olfactory stimulus and the visual quality of brightness that exists in the neural wiring of every human brain. She is speaking the language of cross-modal correspondence, which is the only language smell has ever had.
Training a perfumer cultivates functional synesthesia
The perfumer's training, then, can be understood as a deliberate cultivation of functional synesthesia. A student of perfumery spends years smelling raw materials, hundreds, eventually thousands of them, and building an internal library that encodes each material not just as a smell but as a complex, multisensory profile. Vetiver is more than an odor. It is dark, earthy, slightly smoky, dry on the skin, green at the top, woody at the base, with a texture like rough linen. Every one of those descriptors is borrowed from another sense. And every one of them is necessary, because a perfumer who could only say "this smells like vetiver" would be like a painter who could only say "this looks like blue." The descriptors are the tools of compositional thought. They are how a perfumer thinks through a formula, balancing warm against cool, light against dark, sharp against smooth, the way a composer balances major against minor, staccato against legato.
The parallel to music is not casual. Both perfumery and music are temporal arts: they unfold over time, with a structure that has a beginning, a development, and a resolution. Both operate on materials that are invisible and intangible. Both rely on an internal vocabulary that is partly technical, partly synesthetic, and largely incommunicable to outsiders. And both produce experiences that are felt, viscerally and immediately, in a way that precedes and often overwhelms intellectual analysis. You do not decide to be moved by a piece of music. You do not decide to be transported by a scent. The response is pre-cognitive, rooted in the oldest and least verbal parts of the brain.
What the synesthete has by nature, the perfumer develops by discipline, a discipline that smell training after Covid has now brought to millions of non-professionals. The difference is real: S.J. cannot choose to stop seeing burgundy when she smells coffee; a perfumer can discuss vetiver's darkness without literally seeing a dark visual field. But the underlying neural architecture is shared. Both are using cross-modal pathways. Both are experiencing smell as richer than a single-sense phenomenon. The synesthete's brain is simply more insistent about it.
The philosophical tradition of discrete senses
A philosophical tradition, running from Aristotle through Locke and into contemporary phenomenology, that treats the senses as discrete channels delivering separate types of information to a central processor, the mind, the soul, the homunculus sitting in the Cartesian theater. This model is intuitive. It feels right. My eyes give me color, my ears give me sound, my nose gives me smell, and somewhere behind my forehead, "I" assemble these inputs into a coherent world.
Synesthesia demolishes this model. If the senses were truly separate, synesthesia would be impossible: a wire crossed between two systems that have no business talking to each other. But synesthesia is not only possible; it is common enough to affect millions of people worldwide, and the cross-modal correspondences that undergird it are universal. The senses were never separate. The model was wrong.
What we have instead is a brain that builds a unified perceptual field from multiple overlapping and interpenetrating channels, any of which can influence any other. Smell is never just smell. It is always accompanied by associations, visual, tactile, emotional, spatial, temporal, that are not metaphorical additions to the experience but constitutive parts of it. When you smell a rose and think "red," you are not making an intellectual inference from the knowledge that roses are red flowers. You are experiencing a cross-modal resonance that is built into the way your brain processes olfactory information. The redness is part of the smell.
This is what perfumery has always known, and what synesthesia confirms. The boundaries between senses are administrative fictions, useful for organizing textbooks, useless for describing experience. A great fragrance does not merely smell good. It evokes light or darkness, warmth or cold, texture and weight and spatial depth, a full-spectrum sensory event triggered through a single channel. This is not a trick. It is not marketing. It is a fundamental property of human perception, one that a neurological minority experiences in its most extreme form and that the rest of us access every time we describe a scent as warm, or bright, or sharp, or dark.
Poets and perfumers work between the senses
The poet and the perfumer have always understood each other, even when they used different tools. Both work in the space between the senses, where a vowel can be red and a molecule can be dark and neither statement is a metaphor. Both know that the richest human experiences are cross-modal, that a sunset is more than orange but also warm and quiet and slow, that a thunderstorm is more than loud but also dark and cold and sharp. The senses are not five. They are one, variously expressed, and the proof is written in the neurology of every brain that has ever called a lemon "bright."
Rimbaud did not need an fMRI to know this. Kandinsky did not need a peer-reviewed paper. The woman in London who sees burgundy when she smells coffee did not need anyone's permission to perceive what she perceives. And every person who has ever closed their eyes, inhaled, and seen, in some interior, undeniable way, a color, a shape, a temperature, a texture that was not there but was absolutely real: they did not need to be told that the senses are connected.
They already knew. The curtain was already open.
Seven extraits at 20%, one collection. The Discovery Set carries all seven in 2 ml.
