Smelling Without A Smell
Stephen Pierzchajlo defended his doctoral thesis “Smelling Without A Smell: How olfactory-perceptual representations are activated by words” on October 17 at Stockholm University. In his dissertation, Pierzchajlo explores how linguistic signals (words) can influence olfactory experiences and how words can activate smell-related representations in the brain—even in the absence of actual odors. The central question is whether we can “smell with language”—that is, whether words can evoke or influence mental representations of scents.
Why study smell
The sense of smell is unique among the senses. Humans find it much harder to name and describe smells than visual or auditory impressions. This makes the sense of smell a perfect model for understanding how language and perception are connected.
How do you put smell into words?
Language and smell are weakly linked—but they do influence each other. Even though we often struggle to find words for scents, research shows that words can affect how we perceive smells and can even trigger smell-related activity in the brain without any actual odor present. A key research question in Pierzchajlos thesis is whether words “coffee” or “rose” can activate smell representations that resemble those of the corresponding smells.
What were the most interesting findings?
One of the most striking findings in the dissertation is how strongly language influences our sense of smell—and how the brain can “smell without a smell.” A key result was that words help prepare the brain’s olfactory areas even before the actual scent appears. This means that reading or hearing words like “coffee” or “rose” can cause the brain to evoke similar representations as if a real smell were present, and this preparation facilitates our perception of smells. This shows that language and smell are not entirely separate systems, as previously thought. In other words, we “smell” not just with our nose but also with our mind.
How were the studies conducted?
Study I: Measured brain responses to word–smell expectations. Participants were given words that predicted (or didn’t predict) an upcoming smell or image. Smell stimuli were delivered via an olfactometer; reaction time and brain activity were measured.
Study II: Compared smell and word similarity using experiments and AI. Participants judged how similar different smells (or smell-related words) were. The data was compared with patterns from a trained language model (LLM).
Study III: Tested smell concepts without actual odors or mental imagery. Participants rated the similarity of different smell-related words. The “Vividness of Olfactory Imagery Questionnaire” was used to measure imagination ability.
What can this research contribute going forward?
Understanding how language affects smell could have implications in fields such as food and beverage research, marketing, sensory design, and neurological rehabilitation.
Linguistic smell exercises (e.g., after smell loss) could be used to retrain the sense of smell after illnesses like COVID-19 or in neurological conditions that affect olfaction.
Finally, the research can be used to understand and influence how people perceive scents through words—for example, in product descriptions, wine tasting, perfume development, or sensory marketing.
Last updated: October 30, 2025
Source: Psykologiska institutionen