Listen here for an AI podcast about this edition's feature essay:

The Machines Felt Nothing

Stinky feet.

Sour milk.

Rotten eggs.

Some smells you never forget. Others catch you off guard. Wet dog. Dirty diaper. Your stomach turns before you think.

Some bad smells are legends.

Durian fruit: smells so strong it’s banned on buses and trains in parts of Southeast Asia. It smells like paint thinner, sewage, and something oddly sweet. 

Surströmming: Sweden’s stinky canned herring. Pop the lid inside and the room clears in seconds.

Then there's the Corpse flower: a rare plant that blooms every seven to ten years. For one day, it smells like rotting meat. People line up to smell it. Some nearly faint.

Now try something:

Think of a smell you hate. Maybe it’s your boss’s cologne. A heavy perfume that overwhelmed you in a packed elevator. Close your eyes. Smell it again.

Same response? Same flinch? Same "Get me out of here!" reaction.

That's what I believed. Smell skips thought and goes straight for emotion. No filter. No warning. Your body reacts before your mind catches up.

This is perfume's oldest story. I believed it enough to put it in my book.

Then a Swiss lab handed out perfume and the sensors went quiet.

So I tore up the chapter and started over.

What replaced it was stranger.

The Test That Broke My Story

In 2015, scientists from Firmenich teamed up with researchers in Switzerland, wired people to sensors, and handed them smells.

The setup was simple:

• Heart rate
• Skin conductance
• Two facial muscles — one for frowning, one for smiling

The smells were unlabeled.  No brand. No story.

The first group smelled thirty-two odors from across the spectrum. Flowers. Fruit. Food. Animals. Sulfur. Rot.

Durian was on the list. So was rancid butter. So was the acid that makes feet smell like stinky feet.

The machines reacted.

Unpleasant smells triggered frowns. Heart rate shifted. Skin responses jumped.

The old story seemed to be right there on the screen.

Then came the second group.

Nine fine fragrances:

• Chanel N°5
• Angel
• CK One
• J’adore
• Samsara
• Trésor

The machines went quiet.

Heart rate: no difference. Skin conductance: no difference. Their frown muscles barely moved.

When asked which perfumes they preferred, people answered with confidence. They knew what they liked.

But the sensors had no clue.

I Confused Alarm with Emotion

Look back at my examples.

Rot. Sewage. Fermented fish. Sweat. A corpse.

They weren’t just bad smells. They were warnings.

I thought they proved that smell goes straight to emotion. What they really showed was simpler: the body reacts fast to possible danger.

But perfume is safe. It doesn’t need the body to scream to make us feel something.

Then I Read the Fine Print

The more carefully I read the paper, the stranger it became.

Every odor was sorted into one of two columns: pleasant or unpleasant.

The unpleasant column read like a crime scene. Body odor. Skunk. Rancid butter. Isovaleric acid—the smell of cheesy feet.

Then four familiar names stopped me.

• Melonal
• Framboisone
• Diacetyl
• Isobutyl quinoline

What? Hold on.

Those weren’t sitting in a garbage can. They were sitting in my formulas.

The Real Clue Was in the Numbers

Melonal: 50%.

Framboisone: 50%.

Diacetyl: 50%.

Isobutyl quinoline: 20%.

These weren’t formula dosages. They were lab test concentrations. No perfumer uses Melonal at 50%. We use materials like that in fractions of a percent.

Then I looked at the other end of the table.

Tiaré flower: pure.

Sulfox (a sulfur material): 0.05%.

That's a two-thousand-fold difference!

My guess: they picked these levels so both would be perceived about the same. If I’m right: the flower had to be pure to match it.

To a non-perfumer, it might sound strange. To a perfumer, it’s logical. Some ingredients need a high dose. Some work at a trace. That’s not a mistake.

Sulfox is one of those high-impact materials. At trace levels, it gives tropical fruit impression. Increase the dose, and it becomes cat-urine. That’s why it’s dosed so carefully.

That’s perfumery.

What the “Bad” Column Got Wrong

The paper labelled these materials as: unpleasant.

But I think of them differently.

Melonal (Carefree): Melon. Cucumber. Green rind. Watery. A little metallic. Often dosed well under 1%. A great modifier of heavy marine molecules like Calone.

Framboisone (Playful): Raspberry. Jam. Soft fruit. Highly edible comfort effect. But the study put it next to rancid butter. I agree. Overdose it and you get a thick, syrupy mess.

Diacetyl (Comfort): Warm melted butter. Popcorn. Creamy. Push the dose and it turns sour-milk and cheesy. Keep it low and it makes a gourmand smell like dessert. Perfect for coffee, caramel, butterscotch, and cookie scents.

Isobutyl quinoline (Rebellious): Dry leather. Green Vetiver. Earthy. Rooty. Tar. Germaine Cellier built Bandit around it in 1944. It's still one of perfumery's boldest materials. I diluted it to 1% or 10% before using it. The paper called it animal.

Then there was one more dosed at one percent.

Paracresol (Carnal): Intensely animalic. Barnyard. Horse. Ink. Medicine. Repulsive on its own. But it's an important trace material in narcissus, ylang-ylang, jasmine, or tuberose. That dirty note is not contamination.

Then there’s indole.

The Material I Refused to Use

Indole wasn’t part of the study, but it belonged in the same story.

On a blotter, it smells intensely fecal. Overwhelming. Decay. Bad breath. Mothballs.

For years, I rarely used it in my new creations.

I simply didn’t like it.

But the same bad feedback kept coming.

My white florals were pretty, but thin. Too clean. Missing something.

My colleagues were right.

So I began adding traces of indole. The flowers became fuller, warmer, more natural.

More alive.

The indole hadn’t changed. My idea of beauty had.

There are no bad materials.

Only wrong doses.

What the Paper Changed for Me

I went into this paper expecting proof that perfume impacts the body before the mind.

Instead, I learned something more useful.

A material does not carry one fixed emotional meaning. Dose matters. Context matters. Memory matters. And the person smelling it matters.

That is how something dirty can become sensual. How something strange can make a fragrance feel alive and happy.

If emotion in perfume can be shaped, it can be designed with greater care.

This paper forced me to rewrite a chapter of my upcoming book.

It may change the way you smell perfume, too.