Smell and memory

Why can a passing smell drag back a childhood scene more sharply than a photograph of it ever does?

You walk past a doorway, catch something — wax polish, a particular soap, wet chalk — and a childhood afternoon arrives whole, before you have identified the smell or even decided to remember anything. Look at a photograph of the same afternoon and you get recognition, information, mild warmth. Something about the smell skips the queue. But what queue, exactly, and why does one sense get to jump it?

Notice first what is odd about the experience itself, because the clue is in the ordering. With the photograph you recognise, then feel. With the smell you feel, then — often clumsily, sometimes never — work out what you are smelling. Most people are strikingly bad at naming odours they know perfectly well: the memory is right there and the word is not. So whatever route the smell takes seems to arrive at feeling earlier than it arrives at identification.

That inverted order is a strong hint about wiring. Ask how the other senses reach you. Light hits the retina, and the signal is passed to a relay station deep in the brain — the thalamus — and only then forwarded to the visual cortex, where an image is assembled and recognised. Sound takes the same shape of route. The thalamus is the great switchboard: sensory traffic goes through it, gets gated and filtered by attention, and emerges at a cortical map. Emotional colouring is applied afterwards, once the thing has been worked out.

Smell is built differently. Receptor neurons in the nose — a few hundred distinct types — run their fibres straight up into the olfactory bulb, sitting just under the front of the brain. And from the bulb, the output projects directly to the piriform cortex and to the amygdala, and on into the entorhinal cortex, which is the doorway to the hippocampus. Piriform is olfactory cortex. Amygdala is the structure that tags experience with emotional significance. Hippocampus binds an episode into a retrievable scene.

Read that list again and the phenomenon almost falls out of it. The olfactory pathway lands, in very few steps, in exactly the structures that handle emotion and episodic memory — structures the other senses reach later and by a longer road. The smell is not stronger than the photograph. It is closer.

Why would evolution build it that way? Because olfaction is old. It is the sense by which an animal decides whether something is food, kin, mate or rot, and those are decisions where the appropriate output is an immediate pull or recoil, not a considered description. A pathway optimised for approach-or-avoid has no reason to route through a switchboard built for deliberate attention.

Two corrections, and the piece is not honest without them.

The first concerns the anatomy. You will very often read that smell is "the only sense that bypasses the thalamus", and that is an overstatement. Olfaction does have a thalamic pathway — through the mediodorsal thalamus — and it appears to matter for discriminating odours, for attention to them, and possibly for conscious smell perception. The defensible claim is narrower and still striking: olfaction reaches limbic and memory structures without an obligatory thalamic relay first, whereas vision and hearing do not. Direct enough to matter. Not the clean exception it is usually sold as.

The second concerns the memory itself. Does the smell actually bring back a better memory? The evidence says no. What odour-cued autobiographical memories reliably are is more emotional, more strongly accompanied by the sense of being transported back — and notably older: cued by odour, these memories cluster in the first decade of life, whereas the same people cued by words or pictures produce memories peaking in adolescence and early adulthood. That age shift is a real and repeated finding. Greater accuracy or detail is not: the experiments do not support it, and there is no reason they should, since odour memories are reconstructed like any other.

So the feeling of vividness is genuine, but it is a report about emotional intensity and about how far back you have been thrown, not about fidelity. A plausible contributing factor — suggested rather than settled — is that we rarely rehearse or narrate our smells, so those memories are retrieved less often, from further back, and with less accumulated overwriting from retelling.

• Why naming a familiar odour is so much harder than recognising it.
• What the age skew of odour-cued memories implies about how early childhood is stored.

Every term the collection defines is gathered in the glossary.