The same face photograph can feel a little less pleasant when an unpleasant odor is present. A 2018 study investigated whether not only odor type, but also the timing of seeing a face, changes evaluations and brain responses.
Faces were rated less pleasant in the unpleasant-odor condition, and some EEG responses were larger when faces appeared during the unpleasant odor than afterward. What we smell, and when it overlaps with visual information, can both matter when considering interactions between the senses.
Combining odor type with the timing of face presentation
The study directly compared how the interval between odor and face presentation related to pleasantness evaluations of faces and brain activity after seeing them. It was not simply a test of whether wearing a pleasant fragrance improves someone’s impression.
Twenty-nine people participated. Behavioral data from 28 were analyzed after one person withdrew, while EEG analyses included 26 after two further participants were excluded because of excessive noise. Participants were aged 18 to 31, and their sense of smell was checked beforehand.
The photographs showed 90 male and female faces without strong emotional expressions. Prior ratings had also confirmed broadly neutral pleasantness. There were three odor conditions: pleasant jasmine, unpleasant methyl mercaptan resembling rotten cabbage, and clean air as a comparison.
Conditions producing similar perceived intensities for the two odors were selected using a previous experiment. Ratings in this study likewise showed no significant differences in intensity or familiarity, while jasmine was clearly rated more pleasant.
Comparing faces seen during and after an odor
Odors were delivered for three seconds, synchronized with the beginning of inhalation. Face photographs were shown for just 0.3 seconds at one of two timings.
In one condition, the face appeared two seconds after odor onset, while the odor was still being delivered. In the other, it appeared one second after the three-second odor presentation ended.
The paper calls the first condition simultaneous presentation, but the odor and face did not begin at exactly the same instant. Rather, the two stimuli overlapped in time.
Each participant saw every face twice, paired with the same odor. What differed was the timing of face presentation. Face-odor assignments were counterbalanced across participants, and presentation order was also controlled.
After seeing a face, participants rated its pleasantness or unpleasantness on a scale from 0 to 100 and also reported the odor’s intensity on that trial. EEG was recorded continuously with a 128-channel system.
Faces were less pleasant in the unpleasant-odor condition
Combining the two timings, faces were rated significantly less pleasant with methyl mercaptan than with clean air or jasmine. There was no significant difference between jasmine and clean air.
Thus, the main finding in this experiment was that an unpleasant odor lowered face evaluations, rather than a pleasant fragrance raising them.
Mean ratings were 47.94 when faces appeared during the unpleasant odor and 49.39 when they appeared afterward. With clean air, the corresponding means were 54.11 and 53.76. The direction of change differed by odor condition.
The analysis also found an interaction between odor type and timing. However, the difference between the two timings within the unpleasant-odor condition alone did not reach statistical significance. That comparison alone cannot establish that faces were reliably rated lower when seen during the unpleasant odor.
Differences in EEG around 0.5 to 0.7 seconds after seeing a face
EEG signals from many trials were averaged relative to face onset. The resulting event-related potentials provide a way to examine when responses change after a stimulus.
First, in an early time range around 0.17 seconds after face onset, responses were larger when faces appeared during odor delivery than afterward when odor conditions were combined. This was a timing effect, not an effect established as specific to unpleasant odor.
At later times, an interaction between odor type and timing emerged. Two responses centered around 0.52 and 0.71 seconds after face onset showed larger amplitude changes when faces appeared during the unpleasant odor than after it. These EEG timing differences were significant.
These time ranges correspond to relatively late responses known as the LPP, which is associated with processes including emotional information processing. Jasmine and clean air showed no significant timing differences in the same two time ranges.
Information beyond the face itself contributes to evaluation
The authors suggest that when an unpleasant odor overlaps with a face, the unpleasant surrounding context may influence face processing more strongly. This could relate to mechanisms directing attention toward situations involving unpleasant odors.
However, EEG amplitude is not a direct numerical display of “how much someone dislikes” a face. It may also reflect attention to the face or the integration of multiple senses. An adaptive function in avoiding danger is likewise an interpretation, not a behavior tested in this experiment.
Participants reported pleasantness or unpleasantness on seeing face photographs. The study did not assess a person’s trustworthiness, personality, or attractiveness, nor did it demonstrate changes in real-world interpersonal relationships.
The experiment also involved only two odors and brief face presentations. EEG analyses searched across many electrodes and time points to identify differences, making replication with other stimuli and populations important. Even so, the findings show why presentation order and intervals need to be controlled when studying relationships between smell and vision.
The study received support from a UK ESRC CASE studentship and additional support from Unilever. Two company employees were coauthors, and the company’s involvement in the research was described as limited.
Summary
Neutral face photographs were rated less pleasant in an unpleasant-odor condition. In addition, EEG responses around 0.5 to 0.7 seconds after face onset were larger when faces appeared during the unpleasant odor than after it ended.
The timing difference in face ratings within the unpleasant-odor condition was not significant on its own. It is therefore important to distinguish subjective evaluations from brain responses. Understanding how odor contributes to impressions requires attention not only to its pleasantness, but also to when it meets other information.
References
Cook S, et al. Simultaneous odour-face presentation strengthens hedonic evaluations and event-related potential responses influenced by unpleasant odour. Neuroscience Letters. 2018;672:22–27. doi:10.1016/j.neulet.2018.02.032
