#156 - Smell and Memory

Matt Walker (0:02)

Hi, everyone, it's Matt and welcome back to the podcast. Today, we'll look at how one particular sense reaches into your past in a way the others don't. We'll start with a glass jar without a label that sits on the back shelf of a spice rack. We'll open it, smell what's inside, and immediately recognize it. But then when you try to name it, that's when you'll likely struggle. That's because it's incredibly hard to identify familiar scent, even if your sense of smell is intact, and even if you can name objects you see without any trouble. And not only do you struggle with the name of a familiar scent, but often a memory comes to mind instead, frequently a long forgotten one. In psychology, this effect is known as the Proust phenomenon. Researchers who set out to test it noted that support for the idea had mostly relied on anecdotal sources and a later review of the field described scientific interest in it as sporadic. So let's ask our question. Why does a scent go back further than a photo? The first clue is that one sensory system bypasses a stage in which all the others participate. Imagine the brain is a hotel and all guests must pass through the lobby before entering the hallways. Sighted people walk through the lobby as do those who hear and feel.

Smell has a secret back entrance with a staircase that leads to a hallway outside a storage room containing your memories. It may seem like a strange image, but the physical layout of the brain matches it surprisingly well. For the senses of sight, hearing, and touch, the lobby is the thalamus. It is the main source of sensory information for the brain's primary sensory cortices. Smell bypasses this lobby.

The sensory neurons in the nose transmit signals to the olfactory bulb, which sends them directly to the primary olfactory cortex.

This region includes the piriform cortex, part of the amygdala, and the lateral entorhinal cortex. Within one or two steps, the information arriving in your nose has reached regions involved in processing emotion, in processing reward and decision making, and in processing memories of the events in your life. When subjects are placed in an MRI scanner and instructed to simply lie still, we can examine which areas exhibit synchronized changes during rest. One study conducted this analysis for all four senses simultaneously within the same subjects. Smell, touch, vision, and hearing. In each case, the primary sensory area was contrasted with the hippocampus. This was a small sample of 25 healthy participants. During rest, the primary olfactory cortex showed greater connectivity with hippocampal circuits than the other sensory cortices did. The authors offer an evolutionary explanation, and suggest is their word for it. During the expansion of the primate brain, the hippocampus' functional networks shifted away from primary sensory areas and towards association areas, a shift observed in vision, hearing and touch. The olfactory cortex appears to have remained largely unchanged across mammalian evolution, and the findings imply that this connection with the hippocampus may have persisted, while the other modalities underwent structural relocation. Back in the hotel, there was a remodeling in which every other department relocated to the upper floors, but the smell department retained its original workspace near the archives.

When the olfactory cortex encounters a scent, it constructs an object from it. It extracts the physical and chemical properties of the molecules and converts them into an object, complete with associations and with a pleasantness or unpleasantness attached. If you scan people with fMRI as they sniff a range of odors, the pattern of activity in the posterior piriform cortex corresponds to the way they rate each odor. Two odors that generate similar patterns will be rated by people as similar. Two odors that generate dissimilar patterns will be rated as more different. The same relationship doesn't appear in the anterior piriform cortex, the amygdala or the orbitofrontal cortex.

It's in a specific region of the brain, in short, where a smell becomes a recognizable thing. This makes the blockage, when you can't name an object but remember it, we sometimes experience all the more mystifying. If we think of the pathway from odor to verbalization as a conveyor belt with three checkpoints, we can see why this happens. One checkpoint is for perceiving, one connects it to meaning and vocabulary, and one for the verbalization of the word itself. When you combine the behavioral data with the imaging results, the losses at each of the three checkpoints add up. And by the third checkpoint, the signal has thinned out. The scientists who designed this model titled their paper, The Muted Sense. But losing the label doesn't just cost us the name of an object. A name is like a draw handle. Without it, the draw is still there, but you cannot draw it open when you need to. When you ask subjects to call up a smell in their imagination, the odors that are hardest to name are those they find hardest to imagine. When you train subjects to name a set of odors, then ask them to imagine these smells. Their ratings show that they find many of those odors easier to imagine than the subjects who simply smelled the odors or were only told the names of the odors or received no training at all. The authors are careful to point out that labeling might explain only some of the difficulty in recalling an odor. The draw matters all the same. Which brings me to the one experiment I'd like you to keep in mind. Imagine that a memory is a room with two entrances. You can enter through the door marked word or the door marked smell.