Even among ‘nice smells’, a calming scent differs somewhat from one that lifts the spirits. A smell may also evoke nostalgia or stimulate appetite. One study developed a scale allowing people to describe these feelings in detail in Japanese.
In 2025, Hirama and colleagues presented J-GEOS, a Japanese-language scale for assessing emotions elicited by odors. They asked 200 Japanese participants to smell 10 odors and analyzed responses to 25 emotional expressions. The aim was to record feelings not readily captured by pleasantness and unpleasantness alone, and to develop a tool for future comparisons with research in other languages and cultures.
- Describing feelings with 25 expressions
- Smelling for five seconds without knowing the odor’s name
- Seven factors emerged from patterns in the responses
- Even pleasant odors produced different combinations of responses
- A starting point for cultural comparisons, not uniquely Japanese emotions
- Summary
- References
Describing feelings with 25 expressions
The starting point was an odor-emotion scale previously used in languages including English, French, Chinese, and Portuguese. Two researchers independently translated English and Chinese expressions into Japanese, creating 25 Japanese items through comparison and back-translation into the original languages.
Examples include expressions meaning ‘happy’, ‘attracted’, ‘relaxed’, and ‘nostalgic’. These words describe how people feel when smelling something, not what the odor smells like. A correction to the original paper’s table of Japanese translations was published in January 2026.
Participants were aged 29–83: 100 were 65 or older, and 100 were under 65. There were 100 men and 100 women. Eligibility included being born in Japan and having spent ages 6–15 there; selection also used self-report questionnaires about everyday odor perception.
Smelling for five seconds without knowing the odor’s name
Odors were delivered using a mask-type device containing a fixed piece of fragrance-treated paper. The design reduced leakage into the surroundings. After smelling each odor for five seconds, participants removed the mask and returned it to a bag. They were not told the odor’s name.
They then rated how strongly they experienced each of the 25 feelings on a five-point scale. Three minutes were allotted for responses and a break, and all 10 odors were presented in randomized order.
The stimuli included peach-like γ-undecalactone, almond-like benzaldehyde, milky δ-decalactone, and cinnamaldehyde, a major cinnamon odorant. Earthy or musty geosmin and fecal-smelling skatole were also included, combining pleasant and unpleasant odors and food-related and non-food-related ones.
Seven factors emerged from patterns in the responses
The researchers used factor analysis to examine which items tended to receive high or low ratings together. This method organizes similarly varying responses into a smaller number of clusters.
The results were organized into seven factors: happiness–soothing–attraction, unpleasantness–anxiety, vitality, hunger–thirst, sensuality–desire, nostalgia, and spirituality. These were clusters identified from responses to 25 items, rather than seven emotions specified in the questions from the outset.
Nor did the original scale’s categories reappear unchanged. In the Japanese responses, for example, items related to calmness clustered with happiness and attraction into one large factor. The nostalgia factor also showed an association with feeling relaxed.
The authors selected seven factors based not only on statistical fit but also on interpretability. This number does not define all the types of human emotion, and the factor names describe patterns in the responses. ‘Spirituality’ likewise refers to self-ratings of that expression, not a test of special powers attributed to fragrance.
Even pleasant odors produced different combinations of responses
Comparing factor scores for individual odors revealed different profiles. For the peach-like fragrance, for example, happiness–soothing–attraction and hunger–thirst scores were high relative to the overall reference. Skatole had a high unpleasantness–anxiety score.
The milky fragrance had high vitality and hunger–thirst scores as well as a high happiness–soothing–attraction score. Citronellol, reminiscent of citrus and flowers, was distinguished by a high nostalgia score alongside happiness–soothing–attraction and other factors.
Here, ‘high’ is relative to factor scores whose mean across the analyzed data was set to zero. This was neither a before-and-after comparison showing improved mood nor a direct ranking of every fragrance against every other. The scores describe reported feelings under these conditions, not benefits to physical or mental health.
A starting point for cultural comparisons, not uniquely Japanese emotions
When the authors compared their findings with earlier studies in other countries, they found similar clusters involving pleasantness and unpleasantness, energy, and sensuality–desire. This supports the possibility of cross-cultural commonalities in odor-evoked emotion. Factors with similar names do not necessarily contain identical items, however.
The paper discusses whether memories of food and smells, the age distribution, and the particular fragrances selected may have contributed to the emergence of nostalgia and spirituality as factors. It did not directly test whether participants’ experiences generated those factors.
This study and earlier studies overseas differed in their odorants, selection methods, and procedures. Differences cannot simply be attributed to culture. Nor did the study discover emotions unique to Japanese people.
Questionnaire responses also capture only one aspect of feeling. Further validation incorporating behavior, physiological responses, and cognitive changes is needed; the scale alone cannot measure an entire experience that also includes a product’s price or appearance. The authors also note that participants were mainly people reporting good olfactory function, and that translated words may not have exactly the same meanings across languages.
Summary
J-GEOS is a tool for recording odor-evoked emotions in Japanese in greater detail than simply ‘like’ or ‘dislike’. Responses from 200 participants yielded seven factors and revealed different combinations of feelings for different odors.
The authors suggest possible applications in cross-cultural comparison and developing scented products suited to people’s everyday backgrounds. With further validation, the mask-type presentation method might also become a tool for studying individualized scent experiences. Putting feelings into words, alongside identifying what an odor resembles, is the starting point.
References
Hirama, H., Miura, T., Kanazawa, S., Imamura, Y., Kano, S., & Kobayakawa, T. (2025). What emotions are elicited by smells in Japanese people? Emotional measurement using a universal scale in Japanese. PLOS ONE, 20(5), e0323206. DOI: 10.1371/journal.pone.0323206 / Full original article. Original article: © 2025 Hirama et al., CC BY.
The PLOS One Staff (2026). Correction: What emotions are elicited by smells in Japanese people? Emotional measurement using a universal scale in Japanese. PLOS ONE, 21(1), e0340419. Correction and corrected Japanese translation table (DOI: 10.1371/journal.pone.0340419). © 2026 The PLOS One Staff, CC BY.
