The mammal olfactory system is astonishingly capable, able to detect an uncountable number of odorous molecules. It does this by using close to four hundred different olfactory receptors, most of them responsive to more than just a single chemical. This comes at a certain cost: extraordinarily complex response patterns that could overwhelm the brain’s processing capabilities. One important and fascinating question in olfactory research is how we make meaning of smells while avoiding data overload.
In many of our senses, input telegraphs to an anatomical map of sorts. It’s often possible to follow the mapping of stimuli through several layers of neurons as they are processed, then handed off to various brain regions. In olfaction, input from the receptor cells are sorted into the olfactory bulb go through glomeruli at the input end of the olfactory bulb, to some degree approximating the how receptor types are positioned in the nasal epithelium. At this point, their organization still resembles a map.
In most mammals, the organization of olfactory receptor input to the olfactory bulb’s sorting stations, or glomeruli, seems mostly based on solubility in water (hydrophilicity), which at least partially tracks innate like/dislike of particular chemical classes. This results in a pleasantness gradient across the olfactory bulb’s input side. It’s a little imprecise in humans, but at the water-insoluble (hydrophobic) end are esters and other molecules common in food, for which we obviously have some affinity. The arrangement progresses through grassy-to-fatty aldehydes, spicy phenols, cooked or burnt pyrazines and others.
At the hydrophilic far end, you’re in very dark territory: chemicals like amines possessing strongly negative, even repellent characters. Encoding of like/dislike starts right at the receptor level, especially for dangerous predator odors and the stench of death and rot, which in most mammals are hardwired to drive instinctive action. Humans have receptors capable of detecting these “behavioral” odor chemicals, and these cluster together in the olfactory epithelium, roughly telegraphing up into the olfactory bulb as well.
By the time aroma information exits olfactory bulb just two synapses from the nose, all bets are off. By the time neural signals reach the first region beyond—the primary olfactory cortex (piriform cortex)—the pattern of responses to chemistry disappears and is replaced by something utterly different. The brain simply does not have the computational capacity to manipulate the massive amounts of data the receptor responses contain. Once it uses the patterns to recognize odors, there’s no further use for it, so there’s no reason to preserve this bulky data.
So that dispatches the question about an odor map in the cortex. There isn’t one. But what is in its place? The full picture is a little unclear, but we know what kind of information it contains. Every odorous experience leaves the olfactory bulb tagged with everything from reward and like/dislike (hedonic) valuations to category, context and suitability for tasks and other behavior. We know the generalities, but we can’t read the specifics in the brain traffic. There may not even be a universal code, as it seems like each individual and every brain region extracts just the information it needs to do its job. We don’t know how the brain makes this dramatic translation, but it obviously involves learning of various kinds, since the new coding is about the odor’s relationship to us. At this point, this may be one of the deeper mysteries in olfaction.
From a taster’s point of view, one important effect of this process is that we experience odors configurally. The brain’s enforcement of efficiency means individual features become subsumed into a nearly impenetrable whole. This is, in fact the superpower of olfaction—to take a set of diverse chemical features from an endlessly complex environment and process them into a unified representation.
Access to the components of a smell, if they were available as they are in other senses, could theoretically improve our ability to conjure up vocabulary terms. Vision, for example, is processed in stages, each of which feeds information about lines, planes, shapes and more to particular brain regions. This enables features to be recognized even as the representation is still being processed, providing clues to identity and ultimately naming. Odors have no primitives such as the subcomponents of visual objects known as “geons”— cylinders, rectangles, cones and others.
Instead, odor signals gushing into the language network by just the third synapse up from the nose. This a raw, indeterminate format that defies translation and honestly, can feel a little scary at times. With no features, smells are not easily manipulated in the mind, one reason why it takes a lot of learning and experience to be able to do so successfully.
—————————————————————————————
Citations:
Florence Kermen, “Topographical Representation of Odor Hedonics in the Olfactory Bulb,” Nature Neuroscience 19, no. 7 (2016): 876–78, https://doi.org/10.1038/nn.4317.
Hadas Lapid and Thomas Hummel, “Recording Odor-Evoked Response Potentials at the Human Olfactory Epithelium,” Chemical Senses 38, no. 1 (2012): 3–17, https://doi.org/10.1093/chemse/bjs073.
Anat Arzi, Noam Sobel, “Olfactory perception as a compass for olfactory neural maps.” Trends in Cognitive Sciences, (2011): https://doi.org/10.1016/j.tics.2011.09.007.
Artin Arshamian, Patricia Manko and Asifa Majid, “Limitations in odour simulation may originate from differential sensory embodiment,” Philosopical Transactions of the Royal Society B 375, no. 1800 (2020): , https://doi.org/10.1098/rstb.2019.0273
Jonas K. Olofsson and Jay A. Gottfried, “The Muted Sense: Neurocognitive Limitations of Olfactory Language,” Trends in Cognitive Sciences 19, no. 6 (2015): 314–21, https://doi.org/10.1016/j.tics.2015.04.007.