Even when listening to the same recording, the fragrance in the room may change ratings of how fresh or cool a voice sounds. A 2010 study altered only the pitch of one man’s recorded voice and examined differences in impressions with and without fragrance. It reported trends toward greater perceived freshness than without fragrance for lower voices with a sweet floral fragrance and higher voices with a citrus-musk fragrance.
The paper also aimed to develop a method for measuring how fragrance affects impressions of people. However, the statistical threshold used in the results figures was the less stringent 10% level. Here, we examine the procedures and findings as an exploratory study of fragrance–voice combinations.
Looking beyond liking the fragrance itself
When developing fragrances for men’s body-care products, asking whether people like the fragrance itself is different from measuring how a man is perceived in a setting where that fragrance is present. As a method for the latter, the authors focused on voice, in which a single element can be varied more easily than in a photograph of a person.
They used a recording of one man reading the beginning of a folk tale. The original fundamental frequency was 116 Hz, and mechanical processing produced six versions at 103, 105, 112, 116, 122, and 123 Hz. The aim was to compare impressions while minimizing differences other than pitch.
The voices were evaluated by female employees in their 20s to 40s, with 22–32 participants per test. After watching an environmental video for five minutes, they recorded their current mood, listened to the six voices through headphones in random order, and rated coolness, freshness, masculinity, liking, and other impressions on line scales. Finally, they rated their own mood again.
A diffuser maintained a constant fragrance from entry to departure. Fragrance conditions were tested on separate days at the same time of day. In addition to a female examiner, a male examiner was present so that participants would be aware of a man nearby. This was an experiment using recorded voices, not an evaluation of a conversation in which the speaker actually wore the fragrance.
Voice pitch affected impressions even without fragrance
Without fragrance, the original 116 Hz voice had the highest mean liking and coolness ratings of the six versions. Masculinity ratings were relatively high at 116 Hz and below, and lower at 122 Hz and 123 Hz. It is also interesting that the lowest voice was not the most liked.
For convenience, the researchers grouped voices below 120 Hz as lower voices and those above 120 Hz as higher voices when presenting subsequent results. This boundary was based on the recording used here; the authors themselves stated that voices from several men would need to be studied before generalizing.
Fragrances were chosen through word associations and preferences among men and women
Fragrance selection was also a major part of the study. Fourteen experts evaluated 34 fragrance materials selected from literature and traditional accounts using 59 descriptive words, and the results were organized by factor analysis. Six groupings emerged: uplift and attractiveness, activity, negative mood, masculinity, coolness and clarity, and calm and fulfillment. What was measured was the imagery associated with the fragrances, not experimental verification of each psychological effect described in traditional accounts.
L-muscone, in particular, was strongly associated with impressions such as attractiveness and a flutter of excitement, and was selected as an ingredient for fragrance blends. The researchers created three blends containing it. A was a clean citrus-musk fragrance, B a gentle, sweet floral fragrance, and C a fresh, powerful aromatic-herbal fragrance.
In a separate preference survey, 40 men evaluated prototypes after using them for one week, while 35 women evaluated the scents after smelling them once. The women were instructed to imagine the fragrance coming from a man. Men’s mean preferences were highest for C, followed by A and B, with the reverse order among women. A and B also readily evoked specific images, so these two were selected for the voice tests. These results apply to the participants and product conditions in that survey, including the different ways men and women tried the fragrances.
The pattern of combinations differed between lower and higher voices
When the single fragrance material L-muscone was tested first, changes in impressions pooled across all six voices were small. However, when only the higher voices at 122 Hz and 123 Hz were considered, there was a trend toward a cooler impression than without fragrance.
With the blended fragrances, mean freshness ratings for the lower voices increased with both A and B, with a trend toward a difference for the floral-sweet fragrance B. For the higher voices, a similar trend appeared with the citrus-musk fragrance A. For the lowest voice, at 103 Hz, trends toward higher ratings of both coolness and freshness were reported with B.
How the statistics are read matters here. Although the text describes some differences as significant, the symbols in Figures 7–9 indicate p<0.1 in paired t-tests. These are not reports that the 5% threshold was met; they are better read as candidate findings emerging from comparisons across many voices and rating items.
The authors suggested that B’s gentle image may have softened the strongly masculine impression of lower voices, whereas A’s fresh image may have suited higher voices. Their explanation was that imagery associated with fragrance overlapped with evaluations of the voice. No independent manipulation was used to verify that psychological process.
The listeners’ moods were also examined
Without fragrance, irritation increased after the test, while feelings of relief, liveliness, and excitement declined. With A and B present, changes in irritation and relief were more favorable than without fragrance, and B also lessened the declines in liveliness and excitement.
These mood comparisons also used p<0.1 as the threshold in Figure 10. Moreover, B did not make the pre-to-post changes in liveliness and excitement themselves positive; rather, they declined less than in the fragrance-free condition. The authors interpreted changes in both voice impressions and listeners’ moods as suggesting a possible role for fragrance in enhancing attractiveness.
This method used one recording, female employees, and a limited set of fragrances. Randomization of the order of fragrance conditions, blinding, and correction for multiple comparisons were not specified. How far the results extend to attraction to actual people or interpersonal behavior requires separate testing.
Summary
This study proposed a method using pitch-altered recordings to measure whether fragrance affects impressions of a person. Samples were selected based on fragrance imagery and preferences. It found trends toward greater perceived freshness than without fragrance for lower voices with a floral-sweet fragrance and higher voices with a citrus-musk fragrance.
There were also differences in listeners’ moods, including less deterioration than in the fragrance-free condition. All were exploratory findings at the 10% level, but they suggest possibilities for an evaluation method that considers which voice is paired with which fragrance rather than simply scoring a fragrance in isolation.
All seven authors were affiliated with Lion Corporation at the time of the study. No separate funding or conflict-of-interest statement was found in the article.
Reference
Haga R, Suzuki K, Matsukawa H, Tanaka Y, Ichinose N, Takenaka H, Fujita S. (2010). A Study of the Effect of Fragrances on the Impression Created by Men’s Voices in the Minds of Female Subjects. Journal of the Society of Cosmetic Chemists of Japan, 44(4), 285–291. Original article
