Experiment shows how different smells can influence the visual perception of emotions

Researchers at the University of São Paulo measured the speed with which volunteers recognized expressions of happiness, sadness, anger, disgust or fear. The participants were unknowingly exposed to different smells during the test. The findings are reported in PLOS ONE.

Ricardo Muniz | Agência FAPESP – Smells influence our ability to see and interpret the emotions of other people, even when we are unaware of the odor in question. This is the main finding of the master's research of Matheus Henrique Ferreira, currently a PhD candidate at the University of São Paulo's Institute of Psychology (IP-USP) in Brazil.

An article with detailed measurements of this effect is published in the journal PLOS ONE. “If I’m subjected to a pleasant smell, my perception of pleasant emotions is enhanced,” said Ferreira’s thesis advisor, Mirella Gualtieri, a professor of experimental psychology at IP-USP, with a PhD in neuroscience and behavior. “The same is true of unpleasant smells, which heighten our feelings of fear and disgust.” 

The research team began with the premise that olfactory stimuli are almost always linked to a perception of pleasantness or unpleasantness. “We don’t necessarily classify a visual scene as something we do or don’t like seeing, but it’s often the case that the only thing a person can describe the smell is whether it’s pleasant or unpleasant,” said Gualtieri, who is conducting applied research in sensory psychology.

Using this premise, the group designed an experiment to find out how being in an environment with a pleasant or unpleasant smell can affect the way a person appraises the emotions of others. Gualtieri stressed that the experiment was not novel. “The study of how we appraise emotional facial expressions is actually very old. The noteworthy aspect of our research, which few studies have featured, is that we didn’t focus on the expression of very strong emotions,” she said. “Most researchers in this field work with emotions that aren’t especially common in everyday life. Joy, sadness or anger are represented almost as stereotypes or caricatures. That’s not how emotions are typically conveyed.”

The group solicited responses to a carefully graded spectrum of emotions starting with strong facial expressions classified as 100% intensity, and morphing by 10% increments from extreme joy or sadness, for example, to neutral. Each participant was asked to say whether the face expressed happiness, sadness, anger, disgust or fear. 

“We arrived at the lowest intensity of expression needed for a person to start judging correctly the emotion it represented. We knew 100% was unnecessary, but we wanted to know what the minimum would be. We found that it was mostly between 20% and 30% of the total content of the emotion concerned,” Gualtieri explained. 

Having determined the intensity threshold required for participants to perceive these emotions, they then measured the time taken to reach a conclusion (response time). Lastly, they observed how this could be modified by the presence of pleasant and unpleasant smells.

“We showed how this effect results from all sensory modes. All five senses must interact so that human beings can adapt to their surroundings, communicate and survive. The article describes an example of this,” Gualtieri said. “The presence of a smell, whether or not I'm aware of it, will affect my visual processing and how I interpret visual stimuli as emotions.”

Individual judgments

Another novel aspect of the experiment was that each participant was allowed to decide whether smells were good or bad, rather than being required to use pre-defined categories. “Many studies of this kind use a methodology based on categories so that participants necessarily classify the smell of strawberries as pleasant and foot odor as bad. There are these ready-made labels. But we know from experience that it’s complicated, especially as far as smells are concerned, and the categories don’t always fit,” Gualtieri said. “Our analysis was based on the judgments of individual participants, on whether they found a smell pleasant or unpleasant. That was a major difference in the methodology we used, compared with the typical approach based on labels that assume a particular smell is always good or bad. This choice influenced our results significantly. We decided to conduct the entire procedure on the basis of the participants’ individual judgments of pleasantness or unpleasantness.”

The study sample comprised 20 women and 15 men. The participants did not know it was about the smell. They were told only that its purpose was to measure the speed at which they detected the emotions conveyed by facial expressions. “We didn't say anything about smells. A very small amount of a certain substance [butyric acid, smelling of rancid butter; isoamyl acetate, with a strong banana-like odor; or lemongrass scent] was placed in the foam of the headset microphone they used as they were sitting in front of the screen. The participants themselves conducted the entire experimental session to identify emotions, and we measured the success rates and response times,” Gualtieri said.

After this part was completed, the researchers explained that the purpose of the study was to find out whether the judgment of the emotions conveyed by facial expressions was affected by smells. The participants then rated each smell for pleasantness using a dial with a scale.

Patricia Renovato Tobo, Scientific Manager at Natura Inovação e Tecnologia de Produtos, a subsidiary of Natura Cosméticos, and Carla Regina Barrichello, also affiliated with Natura, are the other co-authors of the article. 

“The extent to which the hedonic valence of smells influences the emotional processing of visual stimuli had been highlighted in previous studies, but we knew several other factors could be involved. Our study showed the significant interaction between olfactory and visual stimuli, so that smells influence the identification of facial expressions and facial expressions influence the emotional response to smells,” Tobo said. 

The study was conducted under the auspices of the Applied Research Center (ARC) funded by FAPESP and Natura between 2016 and 2021, and hosted by IP-USP.

The article “Olfactory interference in the emotional processing speed of visual stimuli: The influence of facial expression intensities” is at: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0264261.