Why are smells so evocative?

It’s our most ancient sense – one we share with bacteria which can detect whiffs of ammonia emitted by hostile microbes. Smell is also our most mysterious sense. Scents have a legendary ability to unearth our earliest memories. American author Michael Chabon alludes to this when he recalls opening a box of comic books: “I was struck by a rush of memories, a sense of my childhood self that seemed to be contained in there.” Why is it so?

Scent-evoked memories tend to be pure and emotional. This is connected to how the memories are stored. Most of the action takes place in an ancient part of the brain – the limbic system. Odour receptors in the nose, like tiny antennae, send messages to the brain’s olfactory bulb, located above the roof of the nasal cavity, for processing. The olfactory bulb sends its information directly to the amygdala – the memory bank for emotional experience. There is no extra processing en route, as there is for our other senses, so smell memories link to emotional memories in a raw state. From there, the interwoven memories of smells and emotions are sent to the hippocampus for long-term filing. This fits with Chabon’s anecdote. He found the scent of his old comics pulled a woven thread of memory, which included an early emotional experience.

Human noses can distinguish more than a million, and maybe

even a trillion smells

It’s not surprising that emotion and scent are strongly linked. Scent is crucial for our survival. If food smells good we want to eat it; if it smells bad it’s probably poisonous. 

What’s less understood is how smell works. 

When you sniff, airborne molecules called odorants are drawn up to the roof of the nasal cavity which is lined with 40 million receptors – each attached to a neuron that wires into the olfactory bulb directly above. There are some 400 different types of receptors, each responding to one or more odorants. According to a 2014 paper in Science, human noses can distinguish more than a million, and maybe even a trillion smells. 

Some odorant molecules create a scent on their own; for instance, isoamyl acetate elicits the smell of banana. More often, odours are created by combinations. For example, a tomato odour results from 16 odorants in a specific ratio, even though the tomato gives off 400 separate odorants.

When the odorant triggers its receptor, it sends a signal along its attached neuron. But we don’t know how the triggering takes place. We also can’t predict from the chemical structure of the odorants what smell they will produce. So if a chemist made a new molecular structure in the lab, we would be clueless as to its smell. This is quite different from vision or hearing – we can predict the colour or sound of a stimulus from the composition of its light or sound waves. To further complicate matters, the odorant concentration can have a huge effect. At dilute concentrations, 4-mercapto-4-methylpentan-2-one smells like blackcurrant. Raise the concentration 30 fold and it smells like cat urine.

There are two theories of how odorants trigger smell: the standard “shape” theory and the controversial “vibration” model.

The shape theory, originally proposed by Scottish scientist Robert Wighton Moncrieff in 1949, holds that odorants trigger receptors based on their shape, rather like a key in a lock. Your brain then pieces together the specific combination of activated receptors to interpret the smell. 

With 400 different types of receptors, the combinations can easily account for a trillion different scents. But the shape theory is flawed. For instance molecules with different shapes can have similar smells while those with similar shapes can smell differently. 

That has left room for the vibration model, championed by biophysicist Luca Turin since the 1990s, though the original idea dates back to the 1920s. It’s a variation of the lock and key idea but the idea here is that the odorant “key” works by vibrating at a specific frequency that is in harmony with that of its receptor. Allowing for some imaginative licence, an odorant molecule unlocks its receptor by singing the right note. 

Turin (who admittedly thought the theory was a little wacky when he first heard it) tested it by seeing if human subjects could differentiate between two odorants with the same shape but different vibrations. Replacing a single hydrogen atom with deuterium (a heavier cousin with different vibrations) he showed humans could indeed discern between some normal and deuterated musks. But others have been unable to reproduce this result when repeating the experiment with other odorants including acetophenone, a fragrance ingredient in soaps and perfumes – and one study, published in March, suggested there’s no evidence for the vibrational theory at all.  

So we do know why smells are connected so strongly to memories and emotions, but we don’t know how our sense of smell works. If you manage to work it out, you’re probably in the running for a Nobel Prize.