A perfume is an illusion, and the illusion is honest

The centre of the book is a refusal. Ellena does not reproduce the smell of tea, of flour, of figs; he says reproduction is not creation. Instead he turns odors into signs: green tea becomes a sign for Japan, flour a sign for skin, green mango a sign for Egypt. The sign is built from a handful of materials, and the listener’s memory does the rest. The plausible, he argues, convinces more deeply than the true.

Underneath sits his strange arithmetic: put two materials together and each stays audible while a third smell appears. One plus one makes three. Odors do not follow each other like notes in a melody; they coexist, argue, and form meaning. That is why he calls composing writing, and why his formulas read like sentences rather than pyramids.

The cabinet of illusions · tap a smell to see how he builds it

Choosing is the creative act

Over twenty years he cut his collection from a thousand raw materials to fewer than two hundred: one third natural, two thirds synthetic, no ready-made bases. His last ten launches, he notes, used no more than a hundred and thirty materials between them. The beauty of a perfume, he insists, comes not from the sum of its ingredients’ qualities but from the harmony between them.

The discipline extends to amounts. He changes a proportion only by at least a factor of two, because timid adjustments teach nothing; middling quantities, he warns, produce middling ideas. And he distrusts the famous pyramid: it describes evaporation, not construction. Bergamot lives six hours on a blotter, rose alcohols twenty, musks for days; a perfume is a balance of intensities, not a stack of floors.

He writes formulas the way writers write, far from the lab

The working image of the book: a man with a sheet of paper, thirty lines by six columns, writing material names from olfactive memory, nowhere near a smelling strip. An assistant weighs the sketch to the milligram; he smells the result the way a novelist reads back a draft. He keeps an annual notebook of pretexts, ideas, quotations and sketched formulas, works by elimination, keeps every trial for months, and defends idle time, since boredom is where his ideas arrive.

Two desk habits are teachable outright. The forty-eight-hour blotter: leave a dipped strip for two days and smell what refuses to leave; only then do you know your base. And the association chain: he lets one smell suggest another until a theme appears, trusting that the most distant associations make the most interesting perfumes.

One chain, as he tells it

The drills that built the nose

The book opens its training chapter with his grandmother in the rose fields asking whether he could smell the chafer beetle among the flowers. The drills that followed are all reproducible. Study smells in sets of ten, at least an hour apart, at five per cent on blotters, in a room that smells of nothing. Sort material names on cards into families from memory, then re-sort blind and score yourself. Keep an odor notebook in two columns, the material and what it evokes, so angelica sits beside iris roots. Play the two-player game he tested with a university: matched sets of numbered vials, one player describing, the other finding.

Then the apprenticeship of copying: rebuild the historic bases and the era-defining perfumes exactly, as in every artistic discipline, before you argue with them. The vocabulary is borrowed from touch, hard, soft, velvety, dry, silky, because for a perfumer the word jasmine names an odor, not a flower. The perfumer, he says, invents the object of his own science. And the arithmetic of the school road is stated plainly: each school admits around twenty from hundreds, and across all schools fewer than a dozen students a cohort become perfumers.

The body knows before you do

The physiology chapter earns its place at the bench. The smelling surface is two to five square centimetres per nostril, six to ten million receptors, renewed every thirty to forty days, wired almost directly into the brain’s memory and emotion circuits, which is why a smell can ambush you with a year you had filed away. A separate system entirely, the trigeminal nerve, reads mint as cold and ammonia as sharp: feeling, not smelling.

He also gives the trade’s ruler: a material’s presence is the ratio of how much of it reaches the air against how little the nose needs. Vanillin barely evaporates yet is detected at almost nothing, so it lasts and lasts; banana ester floods the air and burns out in a minute. Of the ten thousand molecules the industry has made, a working specialist can name about one thousand blind; the amateur distinguishes three or four smells in a perfume, the professional ten times finer. Training, not talent, is the difference he defends.

The nightingale, synthetic

Modern perfumery, the book argues, was born of chemistry, and the synthetics were a liberation: a release from the obligation to copy nature. In 1900 science knew eight components of rose; by the century’s end, more than four hundred. Over thirty per cent of the synthetics he uses exist nowhere in nature, and that is precisely their value: they are words no flower ever said.

His disillusionments are as instructive as his loves. Headspace analysis, which photographs the air around a living flower, converted him instantly and then lost him: a soulless photograph, he decided, a simulacrum, which in twenty years yielded the industry fewer than a dozen usable molecules. The gas chromatograph he treats with respect and suspicion at once: he used it to study the masters he admired, and the industry uses it to copy everyone.

The synthetic century, material by material

The brief, the cursor, and the paradox of the good

The marketing chapters are the book’s bitterest and most useful. A traditional brief arrives as brand history, positioning, mood images, a price per kilo of concentrate and a deadline. Since the seventies, demand marketing has added trend books, focus groups and the cursor method, tick-boxes sliding between feminine and woody and long-lasting, which he says has distanced perfumers from the judgement of their own senses. The result is his paradox of the good: these perfumes are technically excellent, identifiable, and therefore incapable of surprise. Two hundred and fifty launches hit France in a single year of the book.

Against this he sets two working alternatives. The niche model: the perfume speaks for itself, name first, no market tests, reputation by word of mouth. And his own arrangement at Hermès, in-house since 2004: one trusted composer, the Hermessences as protected free time, the garden series built each from one sign, fig leaf, green mango, ginger and rain. The market is not his benchmark, he writes; competition has no place in creation, because the work is not to do better, it is to be unlike.

Time gets its own chapter, and its saddest page: the old houses matured a concentrate for months in tanks; the automated industry compounds in under an hour, so the maceration, he jokes darkly, now happens on the store shelf. His answer is the evolving perfume, a composition that changes honestly as it evaporates, with moments of emptiness and fullness. A department-store customer gives a fragrance seven minutes; a boutique customer, thirty. He composes for the person willing to wait a day.

What it costs, how it is made, and why it cannot be owned

The industrial chapters are a gift to any small house. Only a handful of names, Chanel, Hermès, Guerlain, LVMH, Patou, still compound their own concentrate; everyone else buys it from the flavour giants, five of which hold sixty per cent of a sixteen-billion-euro market. Production runs: pilot batch, a few days of maturation so the materials settle into each other, a week to a month of maceration in alcohol, chilling, filtration, nitrogen. Concentrations, in his figures: cologne two to four per cent, eau de toilette five to twenty, eau de parfum ten to twenty, extract fifteen to thirty-five. And the yields behind the prices: a kilo of absolute costs four tonnes of tuberose flowers, or six hundred kilos of jasmine.

Then the strangest fact in perfumery: the formula, the whole art, cannot really be protected. A patent would publish it and expire; trademark law has refused to register a smell; the French courts said perfume was a work of the mind in 1999 and again in 2006, and the highest court answered the same year that it was merely the application of a skill. So the industry keeps its secrets the old way, by keeping them, while the chromatograph quietly reads everyone’s diary. Secrecy, speed and taste are the only locks that hold.

The register of perfumes and their creators

The book closes by naming names: the perfumers behind the bottles the century knows by heart. A selection from his appendix, kept here as the house crib sheet.