Abstract
An impressive work of neurophilosophy, Ann-Sophia Barwich’s rigorously researched and historically tethered Smellosophy is interwoven with historical developments and insightful interviews. Alongside Stuart Firestein’s research, Barwich tracks in the empirically minded spirit of Patricia Churchland, unspooling what is undoubtedly one of the most comprehensive and accessible studies of olfaction. Advancing through history, philosophy, psychology, and neuroscience in integrated fashion while plucking insights from wine sommeliers and perfumists—whose trained pattern-recognition provides a fascinating study in odor constancy—this book will delight nonspecialists and specialists working in olfaction alike, particularly scientists who working on human sensory perception and those who are interested in both recent developments in olfaction research, as well as olfactory research’s history. The central queries echoing through Smellosophy include: Does it make sense to model olfaction via a mapping of odors onto neural structures? Ought we conceive of smells as mental objects, given that smell sits at the “border of conscious and unconscious perception” (10)? Does smell communicate conceptual content? And, most critically, ought we to consider perception by way of an “interpretive” paradigm, contra the stimulus-response model?
According to Barwich, to understand how the brain knows what really happens outside the nose we have to start from receptor behavior—the brain recognizes smells via pattern recognition, not combinatorial coding and topographic mapping (195). And when it comes to combinatorial coding of odors at the receptor sheet and signal distribution, iterative evaluation facilitates selectivity and learning processes, which result in observational refinement. What this does not explain is how meaning gets attached to odors; while signal distribution may seem random, its measure is far from arbitrary—in olfaction, the relative timing of different neurons carries information (259). Perceptual representation in olfaction takes up brain activity as a processual framework, where this framework is about informational content. Brain activity herein is self-organized and selective, functioning as a measure of the world. Barwich underscores that, with olfaction, the representational expression of physical structures breaks down—olfactory perception does not mirror the world, “interpreting/measuring” it instead.
This interpretation, in generating perceptual content, must be modeled via the biological mechanisms that code the stimulus and determine the patterns of recognition. However, olfactory recognition involves both the physical stimulus and learned behavior (114–115). Barwich’s resounding message is to temper the chauvinism of inherently visuocentric approaches when it comes to how we think about perception, its functions, and its conditions. And while a vast majority of scientists working on human perception, including vision researchers, will likely agree with Barwich that perception is an active process, with learned behavior and interpretation being key, Barwich’s message is valuable for those working in the philosophy of perception, where old terminology and the mapping approach continues to be the dominant paradigm. Barwich details how smells turn into objects of cognition—that is, when we associate their perceptual content with semantic objects. Notably, smells are an interpretation of physical information in the context of continuous physiological and cognitive operations. Consequently, the same stimulus can have various interpretations and get processed into different odor images.
One particularly fascinating anecdote Barwich retells is perfumer Christophe Laudamiel’s distribution of sulfurol strips during a public symposium. Switching between slides of milk to ham, audience members’ olfactory perception shifted, tracking the projected image. Barwich thus shows that perception of smells is not merely a sensory abstraction from physical features, but an active interpretation under mental and biological measures. Odorants can have multiple interpretations, but not randomly so—their disposition is grounded in the chemical features of the stimulus, while their perception is based on an individual’s receptor repertoire.
Barwich brilliantly dovetails psychology with neurophysiology, detailing how olfaction operates by markedly distinct principles of neural representation in comparison with vision, audition, and somatosensation—it is highly variable due to its receptor genetics and does not operate by additive stimulus coding nor employs stereotypic, topographical stimulus representation. Barwich shows that olfaction stands in contrast with the notion of stimulus mapping, which occupied the center of sensory neuroscience during the 20th century and continues to dominate representationalism discussions in the philosophy of perception. When it comes to “mapping talk,” veridical perception is taken as an accurate mental representation of physical properties, particularly by those philosophers who espouse so-called “(strong) representationalism” such as Michael Tye; however, this obfuscates what perception truly is: associative learning, observational refinement, and context-sensitive decision-making (310). Olfaction demonstrates that what is common among perceptual effects—including conceptual imagery—hinges on the development of neural architecture more than stimulus topology.
Barwich uses olfaction to motivate an alternative conception of perception as a skill, built on stratified levels and centered on the measure of information in decision-making rather than stability. As Barwich notes, “How sensory information is processed into perceptual categories and cognitive objects, including what kind of classificatory activity a sensory system affords, is the very thing in question” (305). Olfaction offers an opportunity to arrive at a comprehensive theory of the senses and clarify the premises by way of which we model the neural basis of information processing. Visuocentric theories often pursue the idea that perception is about the stable representation of objects (despite visual processing being strongly centered on motion detection). This prioritizes how one obtains constant perception on the basis of consistently diverging sensation. Olfaction challenges key philosophical assumptions in theories of perception—that individual variation in perceptual experience is incompatible with the objectivity of perception, that perceptual constancy is the chief function of sensory processing, and that neural topography is the brain’s fundamental organizing principle.
Barwich highlights that both philosophical and scientific debates pose stability as an expression of objectivity to “demarcate the difference between appearance and reality in sensory perception" (303–304). Specifically, Barwich cites the late James Gibson’s work and its influence on David Marr—and while contemporary perception research, privy to contextual and learned behavioral influences, may be no longer track in outdated paradigms, the philosophy of mind and perception continues to be plagued by nonempirical approaches, hampered by old terminology (e.g., qualia) and mapping perception on to perspective-invariant perceptual objects. The stereotypic topography of the visual system might bolster this view—according to it, we can map the world onto the brain, with the brain resembling a passive platform upon which mind and world correlate. Olfaction does not comport with this correspondence model and, as Barwich’s details by tracing olfactory research’s history, stereotypical representation and chemotopy are inadequate for the olfactory bulb—in olfaction, glomerulus formation fundamentally depends on receptor genetics and developmental mechanisms (229). Smell is a sense which faces an unpredictable stimulus, both in its behavioral context and physical localizability. The chemical environment is always in flux and olfaction is a strongly evaluative and enactive sense, with perception and judgment always inherently entangled.
Olfaction, serving as a “measuring device/instrument,” offers a dynamic model of the brain. Barwich, partial to current theories of the predictive brain that integrate action-perception coupling—as well as Helmholtz’s prescient work on illusions as unconscious inference and Walter Freedman’s nonlinear dynamics model—notes that the brain “weighs smell to answer how, what, and when to choose” (305). Afferent connectivity in the piriform and temporal coding—primacy coding at the receptor level and population coding at the neural level—present cellular expressions of the two main principles in recent theories of the “predictive brain” (250). These theories motivate that our brains learn stimulus regularities, measuring the environment via prediction, where perceptual content matches current input with prior experience. Barwich convincingly shows that our brains’ high plasticity and contextual encoding make olfaction an excellent model to analyze how the computational principles of anticipation (top-down) and error correction (bottom-up) link to cellular mechanisms at the periphery/receptor level and in the brain (260).
Smells are thus understood as an interpretation of ratios and an evaluation of signal combinations/magnitudes, embodying a representation of qualitative analogs. Barwich’s book is not just about olfaction but perception tout court—the olfactory brain demonstrates that perception is not primarily about stable object recognition and identification, but the observational refinement of experience to flexibly evaluate the qualitative differences in varying decision-making contexts. In centering the brain-as-measuring machine, olfaction and odor situations’ informational content is variable, contingent upon associations formed between specific ratios/combinations of inputs and the expected value of their interactions. Seeing perceptual generalization as action-relative and memory based, the olfactory brain motivates understanding perception via interaction of the perceiving organism with its environment, centering “odor situations” instead of “odor objects.” Smellosophy is not simply about olfaction—the book is undoubtedly valuable for its historical and empirical insights, and will greatly interest those Perception readers who, like Barwich, study the mechanisms underlying human perception; furthermore, the book might even encourage those working in the philosophy of perception to take up empirical philosophy and catch up to the novel paradigms and problems of ongoing perceptual research.
