Journal of Chemical Ecology, VoL 7, No. 1, 1981
A C R O S S - C U L T U R A L STUDY ON THE ATTITUDE T O W A R D S P E R S O N A L ODORS
MARGRET
SCHLEIDT, I BARBARA
HOLD, x and GRAZIA
ATTILI 2
IForschungsstelle fiir Humanethologie, Max-Planck-lnstitut ffir Verhaltensphysiologie, 8131 Seewiesen, Fed. Rep. Germany ~Istituto di Psicologia, CNR, 00157 Rome, Italy (Received October 31, 1979; revised February 28, 1980) Abstract-Human axillary odor was used in testing the ability of male and female subjects to distinguish between gender and individuals. The subjects also gave a qualitative evaluation of the odors. The tests were carried out in Japan, Italy, and Germany. Of all three cultures, 80% of the participants could significantly distinguish among the odor of individuals; 50% could identify the person correctly to whom the recognized odor belonged. Discrimination between male and female odor was significantly shown by 20% of Italian, 30% of German, and 60% of Japanese subjects. The qualitative evaluation of male and female odor was the same in the three cultures: male odor was classified more unpleasant and less pleasant than female odor. Men classified their own odor more unpleasant than women did with their own. A cultural difference was found concerning partner's odor: though men classified it alike (predominantly pleasant), women differed. Japanese and Italian women classified their partner's odor predominantly unpleasant, German women predominantly pleasant. In general the Japanese subjects classified the odors less often pleasant than the Italian and German subjects did. Key Words--Axillary odor, nonverbal communication, human body odor.
INTRODUCTION D u e to t h e b e l i e f t h a t t h e sense o f s m e l l d o e s n ' t p l a y a n i m p o r t a n t r o l e in h u m a n n o n v e r b a l c o m m u n i c a t i o n , r e s e a r c h in this field has b e e n n e g l e c t e d . Although nonhuman primates can recognize conspecific individuals by odor ( H o i s t a n d L e s k , 1975, f o r Tupaia; M e r t l , 1975, f o r l e m u r s ; K a p l a n , 1977, f o r saimiris), little d a t a a r e a v a i l a b l e f o r h u m a n s . W e k n o w t h a t b l i n d a n d d e a f blind children can distinguish familiar persons from unfamiliar ones on the 19 0098-0331/81/0100-0019503.00/09 1981PlenumPublishingCorporation
20
SCHLEIDT ET AL.
basis of smell (Eibl-Eibesfeldt, 1973) and that newborns can distinguish the smell of their mothers from the smell of other women (McFarlane, 1975). Similar abilities have been observed in kindergarten age children (Montagner and Henry, 1974). In adults the capacity to distinguish individuals by their personal odor also appears to be present, at least in American and German samples (Russell, 1976; Hold and Schleidt, 1977; McBurney et ai., 1977). In these experiments axillary odor has been used, this being the main source of personal odor. Hold and Schleidt (1977) found women to be more able to recognize conspecifics by olfaction than men, perhaps reflecting their relatively lower olfactory threshold (Mattei, 1901; Elsberg, 1937; Le Magnen, 1948; Moncrieff, 1951; Koelega and K6ster, 1974). Subjects classified male odor more often unpleasant and female odor more often pleasant (Hold and Schleidt, 1977). The attitude towards one's own odor was similar: men judged their own odor mainly unpleasant, while women judged theirs mainly pleasant. This possibly reflects the fact that women have on the average smaller apocrine glands than men, and therefore produce less intense body odor (Doty, 1977; Hurley and Shelley, 1960; Shehadah and Kligman, 1963). Doty et al. (1978) demonstrated that male axillary odor is classified by subjects as more intense and stronger than female axillary odor. McBurney et al. (1977) found a positive correlation between axillary odors classified as unpleasant and a description of the likely personality of the donors with masculine as well as undesirable traits. The unpleasantness evaluation of male odor may be analogous to findings in nonverbal communication studies that men keep a greater distance and have less eye contact with each other than do women (Baxter, 1970; Exline, 1963; Mehrabian, 1972). Obviously men prefer a greater personal distance, As the qualitative evaluation of personal odor seems to parallel other social signals in nonverbal communication, the question also arose whether contact cultures differ from noncontact cultures in respect to dealing with personal odors. According to several authors (Hall, 1966; Altmann, 1977; Watson, 1970) noncontact cultures are characterized by relatively low sensory involvement, reflected by greater interpersonal distance, minimal eye contact, little touching, and a lower direct body orientation than contact cultures. The latter are to be found in southern Europe, Latin America, and the Arab world, the former in North America, northern Europe, and Asia. As it is likely that odor is a stimulus in nonverbal communication, like eye contact or direction of body axis, we undertook odor experiments in Japan (noncontact culture) and Italy (contact culture). Both were compared with earlier results from Germany (Hold and Schleidt, 1977).
ATTITUDES TOWARDS PERSONAL ODORS
21
Another question is whether there might be differences in odor perception and qualitative evaluation between ethnic groups which have apocrine glands of different size. According to Adachi (1903) and Hurley and Shelley (1960) Japanese have smaller apocrine glands and less odor than Europeans.
METHODS AND MATERIALS
Twenty-four German and 25 Italian couples took part in the investigation. In Japan we were only able to find 7 cooperative couples and therefore added 15 female and 15 male single students. The age of the subjects were: German, between 20 and 50; Italian, between 30 and 40; Japan, between 20 and 30. The German subjects had been married for 2-25 years, Italian subjects had been married more than 10 years. The experiment usually took place in groups of 10 subjects (5 couples or 5 single men and 5 single women). The married Japanese subjects were tested in one group of 4 couples and one of 3 couples. To best guarantee a constant procedure for the experiment and a comparison with the earlier German results, one or both of the German authors was present at the tests in Italy and Japan. Each participant wore the same type of cotton shirt for seven consecutive nights. Subjects were given children's soap (Penaten) to wash with, and they were asked to use no perfume and no deodorant. For the experiment the ten shirts of one group were numbered, folded, and laid out in a separate room. Each subject was asked to take the shirts and sniff them for as long and often as wished, and to complete the following questionnaire: (1) Which shirt has your own smell? (2) Which shirt has your partner's smell? (3) Which shirts smell male? (4) Which shirts smell female? (5) Which shirts smell pleasant, which indifferent, which unpleasant? About an hour later the subjects were asked to reassess the shirts, the shirts being renumbered. In all, there were three experiments with each subject. There are several factors known to influence both body odor and the ability to smell it. The olfactory threshold in women oscillates with the menstrual cycle (Le Magnen, 1952; Schneider, 1974). It is said that smoking habits might influence the ability to smell; illnesses as well as certain drugs or food intake are known to alter personal body odor. Therefore the participants were asked to complete a questionnaire about menstruation, smoking habits, illnesses, drug use, consumption of garlic. They were also asked to give information about their sleeping and washing habits. The data were statistically evaluated as follows: a particular odor was
22
SCHLEIDT
ET
AL.
said to be significantly recognized by a p e r s o n when this p e r s o n chose in at least two o f the three tests the same shirt ( b i n o m i a l test, P < 0.03). A significant difference between male a n d female o d o r w a s - - s e p a r a t e l y for each s u b j e c t - - c a l c u l a t e d with the chi-square test ( P < 0.01). In the e m o t i o n a l classification (pleasant, indifferent, unpleasant), the difference between the clusters (e.g., ' o w n o d o r ' a n d ' s u p p o s e d o w n o d o r ' as defined in F i g u r e 1) was c a l c u l a t e d with the chi-square test ( P at least < 0.05).
6 0 84
o
84
:::: ::i:: ::::
40
~ii :.:
:: i;: ::
*:.
20.
own
60
odor
-T
!:in
40
supposed
!i !ii
odor
-~ i:!
i-ld
--iili#~,
own
i
iii ::::::
9 ~~ 1 7.n6
20
|176176 9
:.-,~ .
i Giii own
odor
sup?osed
own
odor
FId. 1. Classification of own odor. Ordinate: percent responses. Each cluster of bars shows the classification of odors as pleasant = white; indifferent = stippled; unpleasant = striped. Subjects from different ethnic groups are marked: J = Japan; I = Italy; G = Germany. The three bars of each ethnic group add up to 100%. The clusters on the abscissa represent the following categories of odor: own odor = real own odor of subjects, whether correctly recognized or not; supposed own odor = odor which the subjects assume to be their own, whether they are right or wrong; $ = classification given by male subjects; 9 = classification given by female subjects.
23
ATTITUDES TOWARDS PERSONAL ODORS TABLE 1. RECOGNITION OF INDIVIDUAL ODOR a
Germany N=45
1. Subjects recognize own odor 2. Subjects recognize partner's odor 3. Subjects recognize own or partner's odor 4. Subjects mistook odor of another person as own odor 5. Subjects mistook odor of another person as partner's odor 6. Subjects could do (1) or (2) or (4) or (5)
Italy N=50
Japan(singles) N=44
Japan(couples) N = 14
%
%
%
%
31
38
25
33
30
21
49
52
29
29
34
24
30
82
88
41
57 79
~The table contains the significant (P < 0.03) recognition of each subject. RESULTS
Recognition of Individual Odor. Table 1 gives a summary on the recognition (for both sexes) of own odor, of odor of the partner, and of those cases in which the odor of a certain person was erroneously taken for the odor of the partner or of oneself. Nearly 80% of the participants (Table 1,6.) in all three cultures distinguished between the odor of individuals; 50% could identify the person correctly, at least in one case, as partner's or own odor (Table 1,3.). Only 16.5% (23 persons of 139 tested persons) failed to distinguish between any individual odor, i.e., in all three test rounds they chose different shirts. In all three cultures women were better able to identify the odors than men, which may be due to their lower olfactory threshold. This ability was not correlated with the estrogen cycle. A nearly equal percentage of subjects in the three cultures significantly mistook the odor of a certain person for the odor of the partner or for own odor (Table 1,4.5.). This means that at a subconscious level these subjects had identified a certain person by his odor. This finding fits the understanding of Cain (1978), who states that the connection between odor and language appears to be relatively weak; his subjects had difficulties in associating the appropriate label with an odorant. He writes that man is equipped to fix the identity of objects by smell at a perverbal level, and that perverbal identification is perhaps most appropriately viewed as a kind of recognition.
24
SCHLEIDT ET AL. TABLE 2. DISCRIMINATION BETWEEN MALE AND FEMALE ODOR s
Germany N = 50
Italy N = 50
Japan N=44
%
%
%
Men
20
16
50
Women Sexestogether
44 32
24 20
82 64
~The table contains significant (P < 0.01) discrimination of each subject.
Recognition of Sexual Odor. Table 2 shows how many subjects of the three ethnic groups could discriminate significantly between male and female odor. As can be seen, far more women than men could do this. According to their statements the participants seem to have estimated the stronger a n d / o r more pungent a n d / o r more unpleasant odor to be male. In Russell's (1976) study the male odors were usually classified as musky and the female odors as sweat. So there seems to be a qualitative difference between these two types of odor that may be due to different composition of hormones and other substances. Fiedler (1955) found a higher percentage of lactic acid in male sweat compared to female, and androstenon is found in axillary sweat of men (Brooksbank et al., 1974: Claus and Alsing, 1976), although not in women (Claus, personal communication), On the other hand one should not exclude the possibility that a solely quantitative difference exists between male and female odor. The statements of our subjects saying that men smell stronger than women point in this direction. Also in the studies of Doty et al. (1977, 1978), the subjects labeled strong smells more often as male and unpleasant, while less strong odors were labeled as female and pleasant. The hypothesis, that men in general smell stronger (although qualitatively men and women have the same smell) and therefore the difference between male and female odor is only a quantitative one, is further supported by the fact that there exists a large difference between individuals: in nearly every group we tested, some male odors were significantly assigned to be female by the whole group and some female odors to be male. This also corresponds to Doty's findings that given samples of only male or only female odors his subjects assigned odors to both sexes. Qualitative Evaluation of Individual Odor. The results of the qualitative evaluation of subject's own odor is summarized in Figure 1. W o m e n classified their own odor to be more pleasant than unpleasant in contrast to men who classified their own odor less pleasant and usually more often unpleasant (Significance levels of the different evaluation of own odor by men and women: German subjects, P < 0.001, Japanese, P < 0.02, Italian, P < 0.05). The "supposed own odor" (see definition in Figure 1) is classified by women as more pleasant and less unpleasant as own odor, the same is the case with
ATTITUDES
TOWARDS
PERSONAL
25
ODORS
German and Japanese men's classification. Italian men have no different classification between own and supposed own odor. One could hypothesize an underlying norm that everybody "should not smell bad." The largest number in all the men's judgement of their own and supposed own odor fell into the class "indifferent." It seems that men cannot easily decide concerning their own odor. Figure 2 gives the result of the qualitative evaluation of partner's odor. In women's classification there appears to be a difference between cultures. Contrary to the findings in Germany, Japanese and Italian women classify
80
--q, 60 i
40
:.,,
!:i:
:i:
20
partner's
odor
supposed partner's odor
80
60 i
40
:tl:: ..:~:: ,.
q
ti:!~i
20
.:rl::
JF partner's
odor
supposed partner's odor
FIG. 2. Classification of partner's odor. Bar designation and explanation of symbols as in Figure 1. The clusters on the abscissa represent the following categories of odor: partner's odor = real odor of partner, whether correctly recognized or not; supposed partner's odor = odor which the subjects assume to be their partner's odor, whether they are right or wrong.
26
SCHLEIDT ET AL.
their partner's odor more often unpleasant than pleasant (P < 0.01), The odor of the supposed partner is classified by Italian women in the same manner as by German women: predominantly pleasant. The difference in Italian women's attitude to partner's and supposed partner's odor is significant (P < 0.001). Japanese women, however, classify the odor of partner and supposed partner alike: predominantly unpleasant. In men's classification there is no significant difference between cultures. Italian and German men judge their partner's odor predominantly pleasant; Japanese men judge it pleasant and unpleasant equally often. All three groups judge the supposed partner's odor predominantly pleasant. The Japanese men show a large amount of "indifferent"; it seems not easy for them to decide. The classification of "supposed partner" compared to "partner" (with the exception of Japanese women) again brings forward a speculation about a possible underlying norm saying: "my partner smells good." Qualitative Evaluation of Sexual Odor. Figure 3 contains the classification of male odors and Figure 4 female odors. It is remarkable that the Japanese participants classify odors in all groups less often as "pleasant" than the Italian and German subjects, and in most groups more often as "unpleasant." That is to say, on the whole, they give a less positive evaluation of the odors. In all three ethnic groups men and women judge male odor different from female odor, the latter more pleasant (P < 0.001). Furthermore they classify the groups "significant male odor" and "supposed male odor" more unpleasant than "male odor." That means an odor is classified most often as unpleasant (and least often as pleasant) in two cases: when the group of tested subjects agrees that an odor smells male (significant male odor) and when the single subject's decision is that an odor smells male (supposed male odor). This parallels the findings of Doty (1978) of a positive correlation between the ratings of male and unpleasant. Female odors are classified vice versa: in both cases, by the group and by the individual, when an odor is classified as female, it is also classified most often as pleasant (and least often unpleasant). Influence of Other Factors. In analyzing the questionnaire containing the personal data of the Italian and Japanese subjects, we found, as earlier for the German sample, no influence of further factors on the results. Neither medicine intake, garlic consumption, smoking habits, women's menstrual cycle, duration of marriage or association, sleeping or washing habits, nor the age of the subjects showed any correlation to the olfactory abilities of the subjects or to their qualitative evaluation of the body odors. DISCUSSION
The results of our study show that 80% of the participants, regardless of their cultural and racial differences, could differentiate significantly between
sign. male odor
sign. male odor
~:~
supposed male odor
supposed mate odor
odor
mate odor
male
C'
FIG. 3. Classification of male odors. Bar designation and explanation of symbols as in Figure 1. The clusters of the abscissa represent the following categories of odor: male odor = real odor of males, whether correctly recognized or not; supposed male odor = odor which the participants assume to be male, whether they are right or wrong; significant male odor = odor which was classified as male by the majority of the group.
20
z.O
60
20
I-,0
60'
t~
9
9
>
> ~0
,.q
>
r
sign. female odor
I~ I G141!
li
sign. female odor
.
.
.
.
supposed female odor
supposed female odor
female
G'.,/
I I::
i odor
female odor
-i ;'1
FIG 4. Classification of female odors. Bar designation and explanation of symbols as in Figure l. The clusters on the abscissa represent the following categories of odor: female odor = real odor of females, whether correctly recognized or not; supposed female odor = odor which the participants assume to be female, whether they are right or wrong; significant female odor = odor which was classified as female by the majority of the group.
20
/.0
60
/,0
60]
>-
~7
O0
ATTITUDES TOWARDS PERSONAL ODORS
29
individual odors; 50% could correctly identify the person to whom the recognized odor belongs. Of our subjects 20-60% were able to distinguish significantly between the odor of men and women. Noteworthy is the fact that Japanese people tend to have smaller apocrine axillary glands than Europeans and presumably have less personal odor (Adachi, 1903; "the yellow race does not smell at all"). Nevertheless the same test method produced similar results in Japan as in Europe. So it seems that in Japan less intense body odors may be matched by a lower detection threshold. In the qualitative evaluation of the odors the different odor categories (male, female, own, partner) are classified nearly alike in the three ethnic groups, with one exception: judgement of partner's odor by women. Italian and Japanese women classified it predominantly unpleasant; in contrast German women classified it predominantly pleasant. Although we have to be cautious with the Japanese findings in this respect because of the small sample, it is possible that this result reflects the role women play in these different societies. Perhaps personal odor is a relevant factor in sexual attraction and is one of the variables which plays a role in the choice of a partner. Then judging the partner's odor as predominantly unpleasant could indicate a different strategy of choosing a partner compared to when a partner's odor is judged predominantly pleasant. Maybe the women's choice of a partner in Italy and Japan, in contrast to Germany, is more influenced by economic, social, and educational factors, than by sexual attraction (represented by olfactory attraction in this study). In contrast men of the three ethnic groups judge their partner's odor predominantly pleasant. Besides the observation that women's odor is in general judged in this way, one could at the same time hypothesize that men choose their partner more on the basis of sexual attraction than do women. A cultural difference is to be found between Japan, on the one hand, and Italy and Germany, on the other, concerning the overall judgement of personal odors. Men and women in Japan judge the odors in general less positively, i.e., they classify all odor categories less often as pleasant and more often as unpleasant. This might reflect a more negative attitude towards personal odors in Japanese society (Adachi, 1903). In Japan there is a strong cultural pressure to suppress obvious body odors. This attitude is also reflected by the old tradition of taking a ritual bath almost every day. If we consider Italian people as Mediterraneans typically belonging to a contact culture and Japanese people as Asians typically to a noncontact culture, we find a parallel to the different proxemic behavior in the evaluation of personal odor. Our Japanese subjects show perhaps a preference for a more distant contact--supposing odor to be a variable in proxemic behavior. Assuming that body odor plays an important role in sexual attraction, the more negative attitude towards personal odors in Japan may be
30
SCHLEIDT ET AL.
interpreted as reflecting a more negative attitude towards sexuality as compared to the situation in Italy and Germany. This hypothesis is supported by the findings of Iwawaki and Eysenck (1978) on the sexual attitudes among British and Japanese students, saying that Japan "is still very Victorian (nonpermissive) in its outlook on sex" (p. 289). Acknowledgments--We would like to thank Prof. Umemoto and Mr. Kuramitsu in Kyoto for their considerable help in finding subjects for the experiment in Japan. We are very grateful to Kayoko Kasuya, who helped us immensely with the tests, Further we thank Laura Benigni and her colleagues who helped us conduct the test in Italy. For financial support, we thank the Deutsche Forschungsgemeinschaft and the Max-Planck-GeseUschaft.
REFERENCES ADACHI, B. 1903. Der Geruch der Europiier. Globus 83: 14-15. ALTMANN,I., and VINSEL,A.M. 1977. Personal space, an analysis of E.T. Hall's proxemic framework, in J. Altmann and J. F. Wohlwill (eds.) Human Behavior and Environment. pp. 181-259 Plenum Press, New York. BAXTER, J.C. 1970. Interpersonal spacing in natural settings. Sociometry 33:444-456. BROOKSBANK,B.W.L., BROWN, R., and GUSTAFSSON,J.-A. 1974. The detection of 5a-androst-16en-3a-ol in human male axillary sweat. Experientia 30: 864-865. CA/N, W.S. 1977. Physical and cognitive limitations on olfactory processing in human beings, in D. Miiller-Schwarze and M. Mozell (eds.). Chemical Signals in Vertebrates. pp. 287-302, Plenum Press, New York. CLAUS,R., and ALSING, W. 1976, Occurrence of 5a-androst-16-en-3-one, a boar pheromone, in man and its relationship to testosterone. J. Endocrinol. 68:483-484. DUTY, R.L. 1977. A review of recent psychophysicalstudies examining the possibility of chemical communication of sex and reproductive state in humans, in D. Miiller-Schwarze and M. Mozell (eds.). Chemical Signals in Vertebrates pp. 273-286, Plenum Press, New York. DUTY, R.L., ORNDORFF,M.M., LEYDEN,J. and KLIGMAN,A. 1978. Communication of gender from human axillary odors: Relationship to perceived intensity and hedonicity. Behav. Biol. 23: 373-380. EIBL-EmESFELDT,I. 1973. Taubblind geborenes M~idehen. Ausdrucksverhalten. Homo 24:39-47. ELSBERG, C.A. 1937. Newer aspects of olfactory physiology and their diagnostic applications. Arch. Neurol. Psychialr. 37:223-236. EXLINE, R.V. 1963. Explorations in the process of person perception: Visual interaction in relation to competition, sex and need for affiliation. Z Pers. 31 : 1-20. FIEDLER, P. 1955. Der Schweiss. Editio Cantor, Aulendorf. HALL, E.T. 1966. The Hidden Dimension. Doubleday, New York. HOLD, B., and SCHLEIDT, M. 1977. The importance of human odour in nonverbal communication. Z. Tierpsychol. 43:225-238. HOLST, D. VON, and LESK, S. 1975. 0ber den Informationsgehatt des Sternaldrilsensekretes mSnnlicher und weiblicher Tupaia belangeri. J. Cutup. Physiol. 103 : 173-188. HURLEY, H.J., and SHELLEY, W.B. 1960. The Human Apocrine Sweat Gland in Health and Disease. Charles C. Thomas, Springfield, Illinois. IWAWAKI, S., and EYSENCK,H.J. 1978. Sexual attitudes among British and Japanese students. J. Psychol. 98:289-298. KAPLAN,J. 1977. Perceptual properties of attachment in surrogate-reared and mother-reared
ATTITUDES TOWARDS PERSONAL ODORS
31
monkeys, pp. 225-234, in S. Chevalier-Skolnikoff and F.E. Poirier (eds.). Primate BioSocial Development. Garland, New York. KOELEGA,H.S., and KOESTER,E.P. 1974. Some experiments on sex differences in odor perception: Odors: evaluation, utilization and control. Ann. N.Y. Acad. Sci. 237:224-246. LE MAGNEN, J. 1948. Analyse d'odeurs complexes et homologues par fatigue. C.R. Acad. Sci. 226:753-754. LE MAGNEN, J. 1952. Les phdnombnes olfactosexuels chez l'Homme. Arch. Sci. Physicol. 6: 125-160. MACFARLANE,A. 1975. Olfaction in the development of social preferences in the human neonate, pp. 103-117, in The Human Neonate in Parent-Infant Interaction. Ciba Found. Symp. 33, Amsterdam. MATTEI, L. 1901. La sensibilita nei fanciulli in rapporto al sesso ed all'eta. Arch. Psichiatr. 22:207. MCBURNEY, D.H., LEVlNE,J.M., and CAVANAUGH,P.H. 1977. Psychophysicaland social ratings of human body odor. Pets. Soe. PsyehoL Bull. 3:135-138. MEnRABIAN, A. 1972. Nonverbal Communication. Aldine, Chicago. MERTL, A.S. 1975. Discrimination of individuals by scent in a primate. Behav. Biol. 14: 505-509. MONCRIEFF, R.W. 1951. The Chemical Senses. Hill, London. MONTAGNER, H., and HENRY,J.Ch. 1974. Vers une biologic du comportement de l'enfant. Revue des Questions Scientifiques, Bruxelles. RUSSELL, M.J. 1976. Human olfactory communication. Nature 260:520-522. SChNEiDER, R.A. 1974. Newer insights into the role and modifications of olfaction in man through clinical studies. Ann. N.Y. Acad. Sci. 237:217-223. SHEHADAH,N., and KLIGMAN,A.M. 1963. The bacteria responsible for axillary odour. J. Invest. Dermatol. 41:3. WATSON, O.M. 1970. Proxemic Behaviour. Mouton, The Hague.
A C R O S S - C U L T U R A L STUDY ON THE ATTITUDE T O W A R D S P E R S O N A L ODORS
MARGRET
SCHLEIDT, I BARBARA
HOLD, x and GRAZIA
ATTILI 2
IForschungsstelle fiir Humanethologie, Max-Planck-lnstitut ffir Verhaltensphysiologie, 8131 Seewiesen, Fed. Rep. Germany ~Istituto di Psicologia, CNR, 00157 Rome, Italy (Received October 31, 1979; revised February 28, 1980) Abstract-Human axillary odor was used in testing the ability of male and female subjects to distinguish between gender and individuals. The subjects also gave a qualitative evaluation of the odors. The tests were carried out in Japan, Italy, and Germany. Of all three cultures, 80% of the participants could significantly distinguish among the odor of individuals; 50% could identify the person correctly to whom the recognized odor belonged. Discrimination between male and female odor was significantly shown by 20% of Italian, 30% of German, and 60% of Japanese subjects. The qualitative evaluation of male and female odor was the same in the three cultures: male odor was classified more unpleasant and less pleasant than female odor. Men classified their own odor more unpleasant than women did with their own. A cultural difference was found concerning partner's odor: though men classified it alike (predominantly pleasant), women differed. Japanese and Italian women classified their partner's odor predominantly unpleasant, German women predominantly pleasant. In general the Japanese subjects classified the odors less often pleasant than the Italian and German subjects did. Key Words--Axillary odor, nonverbal communication, human body odor.
INTRODUCTION D u e to t h e b e l i e f t h a t t h e sense o f s m e l l d o e s n ' t p l a y a n i m p o r t a n t r o l e in h u m a n n o n v e r b a l c o m m u n i c a t i o n , r e s e a r c h in this field has b e e n n e g l e c t e d . Although nonhuman primates can recognize conspecific individuals by odor ( H o i s t a n d L e s k , 1975, f o r Tupaia; M e r t l , 1975, f o r l e m u r s ; K a p l a n , 1977, f o r saimiris), little d a t a a r e a v a i l a b l e f o r h u m a n s . W e k n o w t h a t b l i n d a n d d e a f blind children can distinguish familiar persons from unfamiliar ones on the 19 0098-0331/81/0100-0019503.00/09 1981PlenumPublishingCorporation
20
SCHLEIDT ET AL.
basis of smell (Eibl-Eibesfeldt, 1973) and that newborns can distinguish the smell of their mothers from the smell of other women (McFarlane, 1975). Similar abilities have been observed in kindergarten age children (Montagner and Henry, 1974). In adults the capacity to distinguish individuals by their personal odor also appears to be present, at least in American and German samples (Russell, 1976; Hold and Schleidt, 1977; McBurney et ai., 1977). In these experiments axillary odor has been used, this being the main source of personal odor. Hold and Schleidt (1977) found women to be more able to recognize conspecifics by olfaction than men, perhaps reflecting their relatively lower olfactory threshold (Mattei, 1901; Elsberg, 1937; Le Magnen, 1948; Moncrieff, 1951; Koelega and K6ster, 1974). Subjects classified male odor more often unpleasant and female odor more often pleasant (Hold and Schleidt, 1977). The attitude towards one's own odor was similar: men judged their own odor mainly unpleasant, while women judged theirs mainly pleasant. This possibly reflects the fact that women have on the average smaller apocrine glands than men, and therefore produce less intense body odor (Doty, 1977; Hurley and Shelley, 1960; Shehadah and Kligman, 1963). Doty et al. (1978) demonstrated that male axillary odor is classified by subjects as more intense and stronger than female axillary odor. McBurney et al. (1977) found a positive correlation between axillary odors classified as unpleasant and a description of the likely personality of the donors with masculine as well as undesirable traits. The unpleasantness evaluation of male odor may be analogous to findings in nonverbal communication studies that men keep a greater distance and have less eye contact with each other than do women (Baxter, 1970; Exline, 1963; Mehrabian, 1972). Obviously men prefer a greater personal distance, As the qualitative evaluation of personal odor seems to parallel other social signals in nonverbal communication, the question also arose whether contact cultures differ from noncontact cultures in respect to dealing with personal odors. According to several authors (Hall, 1966; Altmann, 1977; Watson, 1970) noncontact cultures are characterized by relatively low sensory involvement, reflected by greater interpersonal distance, minimal eye contact, little touching, and a lower direct body orientation than contact cultures. The latter are to be found in southern Europe, Latin America, and the Arab world, the former in North America, northern Europe, and Asia. As it is likely that odor is a stimulus in nonverbal communication, like eye contact or direction of body axis, we undertook odor experiments in Japan (noncontact culture) and Italy (contact culture). Both were compared with earlier results from Germany (Hold and Schleidt, 1977).
ATTITUDES TOWARDS PERSONAL ODORS
21
Another question is whether there might be differences in odor perception and qualitative evaluation between ethnic groups which have apocrine glands of different size. According to Adachi (1903) and Hurley and Shelley (1960) Japanese have smaller apocrine glands and less odor than Europeans.
METHODS AND MATERIALS
Twenty-four German and 25 Italian couples took part in the investigation. In Japan we were only able to find 7 cooperative couples and therefore added 15 female and 15 male single students. The age of the subjects were: German, between 20 and 50; Italian, between 30 and 40; Japan, between 20 and 30. The German subjects had been married for 2-25 years, Italian subjects had been married more than 10 years. The experiment usually took place in groups of 10 subjects (5 couples or 5 single men and 5 single women). The married Japanese subjects were tested in one group of 4 couples and one of 3 couples. To best guarantee a constant procedure for the experiment and a comparison with the earlier German results, one or both of the German authors was present at the tests in Italy and Japan. Each participant wore the same type of cotton shirt for seven consecutive nights. Subjects were given children's soap (Penaten) to wash with, and they were asked to use no perfume and no deodorant. For the experiment the ten shirts of one group were numbered, folded, and laid out in a separate room. Each subject was asked to take the shirts and sniff them for as long and often as wished, and to complete the following questionnaire: (1) Which shirt has your own smell? (2) Which shirt has your partner's smell? (3) Which shirts smell male? (4) Which shirts smell female? (5) Which shirts smell pleasant, which indifferent, which unpleasant? About an hour later the subjects were asked to reassess the shirts, the shirts being renumbered. In all, there were three experiments with each subject. There are several factors known to influence both body odor and the ability to smell it. The olfactory threshold in women oscillates with the menstrual cycle (Le Magnen, 1952; Schneider, 1974). It is said that smoking habits might influence the ability to smell; illnesses as well as certain drugs or food intake are known to alter personal body odor. Therefore the participants were asked to complete a questionnaire about menstruation, smoking habits, illnesses, drug use, consumption of garlic. They were also asked to give information about their sleeping and washing habits. The data were statistically evaluated as follows: a particular odor was
22
SCHLEIDT
ET
AL.
said to be significantly recognized by a p e r s o n when this p e r s o n chose in at least two o f the three tests the same shirt ( b i n o m i a l test, P < 0.03). A significant difference between male a n d female o d o r w a s - - s e p a r a t e l y for each s u b j e c t - - c a l c u l a t e d with the chi-square test ( P < 0.01). In the e m o t i o n a l classification (pleasant, indifferent, unpleasant), the difference between the clusters (e.g., ' o w n o d o r ' a n d ' s u p p o s e d o w n o d o r ' as defined in F i g u r e 1) was c a l c u l a t e d with the chi-square test ( P at least < 0.05).
6 0 84
o
84
:::: ::i:: ::::
40
~ii :.:
:: i;: ::
*:.
20.
own
60
odor
-T
!:in
40
supposed
!i !ii
odor
-~ i:!
i-ld
--iili#~,
own
i
iii ::::::
9 ~~ 1 7.n6
20
|176176 9
:.-,~ .
i Giii own
odor
sup?osed
own
odor
FId. 1. Classification of own odor. Ordinate: percent responses. Each cluster of bars shows the classification of odors as pleasant = white; indifferent = stippled; unpleasant = striped. Subjects from different ethnic groups are marked: J = Japan; I = Italy; G = Germany. The three bars of each ethnic group add up to 100%. The clusters on the abscissa represent the following categories of odor: own odor = real own odor of subjects, whether correctly recognized or not; supposed own odor = odor which the subjects assume to be their own, whether they are right or wrong; $ = classification given by male subjects; 9 = classification given by female subjects.
23
ATTITUDES TOWARDS PERSONAL ODORS TABLE 1. RECOGNITION OF INDIVIDUAL ODOR a
Germany N=45
1. Subjects recognize own odor 2. Subjects recognize partner's odor 3. Subjects recognize own or partner's odor 4. Subjects mistook odor of another person as own odor 5. Subjects mistook odor of another person as partner's odor 6. Subjects could do (1) or (2) or (4) or (5)
Italy N=50
Japan(singles) N=44
Japan(couples) N = 14
%
%
%
%
31
38
25
33
30
21
49
52
29
29
34
24
30
82
88
41
57 79
~The table contains the significant (P < 0.03) recognition of each subject. RESULTS
Recognition of Individual Odor. Table 1 gives a summary on the recognition (for both sexes) of own odor, of odor of the partner, and of those cases in which the odor of a certain person was erroneously taken for the odor of the partner or of oneself. Nearly 80% of the participants (Table 1,6.) in all three cultures distinguished between the odor of individuals; 50% could identify the person correctly, at least in one case, as partner's or own odor (Table 1,3.). Only 16.5% (23 persons of 139 tested persons) failed to distinguish between any individual odor, i.e., in all three test rounds they chose different shirts. In all three cultures women were better able to identify the odors than men, which may be due to their lower olfactory threshold. This ability was not correlated with the estrogen cycle. A nearly equal percentage of subjects in the three cultures significantly mistook the odor of a certain person for the odor of the partner or for own odor (Table 1,4.5.). This means that at a subconscious level these subjects had identified a certain person by his odor. This finding fits the understanding of Cain (1978), who states that the connection between odor and language appears to be relatively weak; his subjects had difficulties in associating the appropriate label with an odorant. He writes that man is equipped to fix the identity of objects by smell at a perverbal level, and that perverbal identification is perhaps most appropriately viewed as a kind of recognition.
24
SCHLEIDT ET AL. TABLE 2. DISCRIMINATION BETWEEN MALE AND FEMALE ODOR s
Germany N = 50
Italy N = 50
Japan N=44
%
%
%
Men
20
16
50
Women Sexestogether
44 32
24 20
82 64
~The table contains significant (P < 0.01) discrimination of each subject.
Recognition of Sexual Odor. Table 2 shows how many subjects of the three ethnic groups could discriminate significantly between male and female odor. As can be seen, far more women than men could do this. According to their statements the participants seem to have estimated the stronger a n d / o r more pungent a n d / o r more unpleasant odor to be male. In Russell's (1976) study the male odors were usually classified as musky and the female odors as sweat. So there seems to be a qualitative difference between these two types of odor that may be due to different composition of hormones and other substances. Fiedler (1955) found a higher percentage of lactic acid in male sweat compared to female, and androstenon is found in axillary sweat of men (Brooksbank et al., 1974: Claus and Alsing, 1976), although not in women (Claus, personal communication), On the other hand one should not exclude the possibility that a solely quantitative difference exists between male and female odor. The statements of our subjects saying that men smell stronger than women point in this direction. Also in the studies of Doty et al. (1977, 1978), the subjects labeled strong smells more often as male and unpleasant, while less strong odors were labeled as female and pleasant. The hypothesis, that men in general smell stronger (although qualitatively men and women have the same smell) and therefore the difference between male and female odor is only a quantitative one, is further supported by the fact that there exists a large difference between individuals: in nearly every group we tested, some male odors were significantly assigned to be female by the whole group and some female odors to be male. This also corresponds to Doty's findings that given samples of only male or only female odors his subjects assigned odors to both sexes. Qualitative Evaluation of Individual Odor. The results of the qualitative evaluation of subject's own odor is summarized in Figure 1. W o m e n classified their own odor to be more pleasant than unpleasant in contrast to men who classified their own odor less pleasant and usually more often unpleasant (Significance levels of the different evaluation of own odor by men and women: German subjects, P < 0.001, Japanese, P < 0.02, Italian, P < 0.05). The "supposed own odor" (see definition in Figure 1) is classified by women as more pleasant and less unpleasant as own odor, the same is the case with
ATTITUDES
TOWARDS
PERSONAL
25
ODORS
German and Japanese men's classification. Italian men have no different classification between own and supposed own odor. One could hypothesize an underlying norm that everybody "should not smell bad." The largest number in all the men's judgement of their own and supposed own odor fell into the class "indifferent." It seems that men cannot easily decide concerning their own odor. Figure 2 gives the result of the qualitative evaluation of partner's odor. In women's classification there appears to be a difference between cultures. Contrary to the findings in Germany, Japanese and Italian women classify
80
--q, 60 i
40
:.,,
!:i:
:i:
20
partner's
odor
supposed partner's odor
80
60 i
40
:tl:: ..:~:: ,.
q
ti:!~i
20
.:rl::
JF partner's
odor
supposed partner's odor
FIG. 2. Classification of partner's odor. Bar designation and explanation of symbols as in Figure 1. The clusters on the abscissa represent the following categories of odor: partner's odor = real odor of partner, whether correctly recognized or not; supposed partner's odor = odor which the subjects assume to be their partner's odor, whether they are right or wrong.
26
SCHLEIDT ET AL.
their partner's odor more often unpleasant than pleasant (P < 0.01), The odor of the supposed partner is classified by Italian women in the same manner as by German women: predominantly pleasant. The difference in Italian women's attitude to partner's and supposed partner's odor is significant (P < 0.001). Japanese women, however, classify the odor of partner and supposed partner alike: predominantly unpleasant. In men's classification there is no significant difference between cultures. Italian and German men judge their partner's odor predominantly pleasant; Japanese men judge it pleasant and unpleasant equally often. All three groups judge the supposed partner's odor predominantly pleasant. The Japanese men show a large amount of "indifferent"; it seems not easy for them to decide. The classification of "supposed partner" compared to "partner" (with the exception of Japanese women) again brings forward a speculation about a possible underlying norm saying: "my partner smells good." Qualitative Evaluation of Sexual Odor. Figure 3 contains the classification of male odors and Figure 4 female odors. It is remarkable that the Japanese participants classify odors in all groups less often as "pleasant" than the Italian and German subjects, and in most groups more often as "unpleasant." That is to say, on the whole, they give a less positive evaluation of the odors. In all three ethnic groups men and women judge male odor different from female odor, the latter more pleasant (P < 0.001). Furthermore they classify the groups "significant male odor" and "supposed male odor" more unpleasant than "male odor." That means an odor is classified most often as unpleasant (and least often as pleasant) in two cases: when the group of tested subjects agrees that an odor smells male (significant male odor) and when the single subject's decision is that an odor smells male (supposed male odor). This parallels the findings of Doty (1978) of a positive correlation between the ratings of male and unpleasant. Female odors are classified vice versa: in both cases, by the group and by the individual, when an odor is classified as female, it is also classified most often as pleasant (and least often unpleasant). Influence of Other Factors. In analyzing the questionnaire containing the personal data of the Italian and Japanese subjects, we found, as earlier for the German sample, no influence of further factors on the results. Neither medicine intake, garlic consumption, smoking habits, women's menstrual cycle, duration of marriage or association, sleeping or washing habits, nor the age of the subjects showed any correlation to the olfactory abilities of the subjects or to their qualitative evaluation of the body odors. DISCUSSION
The results of our study show that 80% of the participants, regardless of their cultural and racial differences, could differentiate significantly between
sign. male odor
sign. male odor
~:~
supposed male odor
supposed mate odor
odor
mate odor
male
C'
FIG. 3. Classification of male odors. Bar designation and explanation of symbols as in Figure 1. The clusters of the abscissa represent the following categories of odor: male odor = real odor of males, whether correctly recognized or not; supposed male odor = odor which the participants assume to be male, whether they are right or wrong; significant male odor = odor which was classified as male by the majority of the group.
20
z.O
60
20
I-,0
60'
t~
9
9
>
> ~0
,.q
>
r
sign. female odor
I~ I G141!
li
sign. female odor
.
.
.
.
supposed female odor
supposed female odor
female
G'.,/
I I::
i odor
female odor
-i ;'1
FIG 4. Classification of female odors. Bar designation and explanation of symbols as in Figure l. The clusters on the abscissa represent the following categories of odor: female odor = real odor of females, whether correctly recognized or not; supposed female odor = odor which the participants assume to be female, whether they are right or wrong; significant female odor = odor which was classified as female by the majority of the group.
20
/.0
60
/,0
60]
>-
~7
O0
ATTITUDES TOWARDS PERSONAL ODORS
29
individual odors; 50% could correctly identify the person to whom the recognized odor belongs. Of our subjects 20-60% were able to distinguish significantly between the odor of men and women. Noteworthy is the fact that Japanese people tend to have smaller apocrine axillary glands than Europeans and presumably have less personal odor (Adachi, 1903; "the yellow race does not smell at all"). Nevertheless the same test method produced similar results in Japan as in Europe. So it seems that in Japan less intense body odors may be matched by a lower detection threshold. In the qualitative evaluation of the odors the different odor categories (male, female, own, partner) are classified nearly alike in the three ethnic groups, with one exception: judgement of partner's odor by women. Italian and Japanese women classified it predominantly unpleasant; in contrast German women classified it predominantly pleasant. Although we have to be cautious with the Japanese findings in this respect because of the small sample, it is possible that this result reflects the role women play in these different societies. Perhaps personal odor is a relevant factor in sexual attraction and is one of the variables which plays a role in the choice of a partner. Then judging the partner's odor as predominantly unpleasant could indicate a different strategy of choosing a partner compared to when a partner's odor is judged predominantly pleasant. Maybe the women's choice of a partner in Italy and Japan, in contrast to Germany, is more influenced by economic, social, and educational factors, than by sexual attraction (represented by olfactory attraction in this study). In contrast men of the three ethnic groups judge their partner's odor predominantly pleasant. Besides the observation that women's odor is in general judged in this way, one could at the same time hypothesize that men choose their partner more on the basis of sexual attraction than do women. A cultural difference is to be found between Japan, on the one hand, and Italy and Germany, on the other, concerning the overall judgement of personal odors. Men and women in Japan judge the odors in general less positively, i.e., they classify all odor categories less often as pleasant and more often as unpleasant. This might reflect a more negative attitude towards personal odors in Japanese society (Adachi, 1903). In Japan there is a strong cultural pressure to suppress obvious body odors. This attitude is also reflected by the old tradition of taking a ritual bath almost every day. If we consider Italian people as Mediterraneans typically belonging to a contact culture and Japanese people as Asians typically to a noncontact culture, we find a parallel to the different proxemic behavior in the evaluation of personal odor. Our Japanese subjects show perhaps a preference for a more distant contact--supposing odor to be a variable in proxemic behavior. Assuming that body odor plays an important role in sexual attraction, the more negative attitude towards personal odors in Japan may be
30
SCHLEIDT ET AL.
interpreted as reflecting a more negative attitude towards sexuality as compared to the situation in Italy and Germany. This hypothesis is supported by the findings of Iwawaki and Eysenck (1978) on the sexual attitudes among British and Japanese students, saying that Japan "is still very Victorian (nonpermissive) in its outlook on sex" (p. 289). Acknowledgments--We would like to thank Prof. Umemoto and Mr. Kuramitsu in Kyoto for their considerable help in finding subjects for the experiment in Japan. We are very grateful to Kayoko Kasuya, who helped us immensely with the tests, Further we thank Laura Benigni and her colleagues who helped us conduct the test in Italy. For financial support, we thank the Deutsche Forschungsgemeinschaft and the Max-Planck-GeseUschaft.
REFERENCES ADACHI, B. 1903. Der Geruch der Europiier. Globus 83: 14-15. ALTMANN,I., and VINSEL,A.M. 1977. Personal space, an analysis of E.T. Hall's proxemic framework, in J. Altmann and J. F. Wohlwill (eds.) Human Behavior and Environment. pp. 181-259 Plenum Press, New York. BAXTER, J.C. 1970. Interpersonal spacing in natural settings. Sociometry 33:444-456. BROOKSBANK,B.W.L., BROWN, R., and GUSTAFSSON,J.-A. 1974. The detection of 5a-androst-16en-3a-ol in human male axillary sweat. Experientia 30: 864-865. CA/N, W.S. 1977. Physical and cognitive limitations on olfactory processing in human beings, in D. Miiller-Schwarze and M. Mozell (eds.). Chemical Signals in Vertebrates. pp. 287-302, Plenum Press, New York. CLAUS,R., and ALSING, W. 1976, Occurrence of 5a-androst-16-en-3-one, a boar pheromone, in man and its relationship to testosterone. J. Endocrinol. 68:483-484. DUTY, R.L. 1977. A review of recent psychophysicalstudies examining the possibility of chemical communication of sex and reproductive state in humans, in D. Miiller-Schwarze and M. Mozell (eds.). Chemical Signals in Vertebrates pp. 273-286, Plenum Press, New York. DUTY, R.L., ORNDORFF,M.M., LEYDEN,J. and KLIGMAN,A. 1978. Communication of gender from human axillary odors: Relationship to perceived intensity and hedonicity. Behav. Biol. 23: 373-380. EIBL-EmESFELDT,I. 1973. Taubblind geborenes M~idehen. Ausdrucksverhalten. Homo 24:39-47. ELSBERG, C.A. 1937. Newer aspects of olfactory physiology and their diagnostic applications. Arch. Neurol. Psychialr. 37:223-236. EXLINE, R.V. 1963. Explorations in the process of person perception: Visual interaction in relation to competition, sex and need for affiliation. Z Pers. 31 : 1-20. FIEDLER, P. 1955. Der Schweiss. Editio Cantor, Aulendorf. HALL, E.T. 1966. The Hidden Dimension. Doubleday, New York. HOLD, B., and SCHLEIDT, M. 1977. The importance of human odour in nonverbal communication. Z. Tierpsychol. 43:225-238. HOLST, D. VON, and LESK, S. 1975. 0ber den Informationsgehatt des Sternaldrilsensekretes mSnnlicher und weiblicher Tupaia belangeri. J. Cutup. Physiol. 103 : 173-188. HURLEY, H.J., and SHELLEY, W.B. 1960. The Human Apocrine Sweat Gland in Health and Disease. Charles C. Thomas, Springfield, Illinois. IWAWAKI, S., and EYSENCK,H.J. 1978. Sexual attitudes among British and Japanese students. J. Psychol. 98:289-298. KAPLAN,J. 1977. Perceptual properties of attachment in surrogate-reared and mother-reared
ATTITUDES TOWARDS PERSONAL ODORS
31
monkeys, pp. 225-234, in S. Chevalier-Skolnikoff and F.E. Poirier (eds.). Primate BioSocial Development. Garland, New York. KOELEGA,H.S., and KOESTER,E.P. 1974. Some experiments on sex differences in odor perception: Odors: evaluation, utilization and control. Ann. N.Y. Acad. Sci. 237:224-246. LE MAGNEN, J. 1948. Analyse d'odeurs complexes et homologues par fatigue. C.R. Acad. Sci. 226:753-754. LE MAGNEN, J. 1952. Les phdnombnes olfactosexuels chez l'Homme. Arch. Sci. Physicol. 6: 125-160. MACFARLANE,A. 1975. Olfaction in the development of social preferences in the human neonate, pp. 103-117, in The Human Neonate in Parent-Infant Interaction. Ciba Found. Symp. 33, Amsterdam. MATTEI, L. 1901. La sensibilita nei fanciulli in rapporto al sesso ed all'eta. Arch. Psichiatr. 22:207. MCBURNEY, D.H., LEVlNE,J.M., and CAVANAUGH,P.H. 1977. Psychophysicaland social ratings of human body odor. Pets. Soe. PsyehoL Bull. 3:135-138. MEnRABIAN, A. 1972. Nonverbal Communication. Aldine, Chicago. MERTL, A.S. 1975. Discrimination of individuals by scent in a primate. Behav. Biol. 14: 505-509. MONCRIEFF, R.W. 1951. The Chemical Senses. Hill, London. MONTAGNER, H., and HENRY,J.Ch. 1974. Vers une biologic du comportement de l'enfant. Revue des Questions Scientifiques, Bruxelles. RUSSELL, M.J. 1976. Human olfactory communication. Nature 260:520-522. SChNEiDER, R.A. 1974. Newer insights into the role and modifications of olfaction in man through clinical studies. Ann. N.Y. Acad. Sci. 237:217-223. SHEHADAH,N., and KLIGMAN,A.M. 1963. The bacteria responsible for axillary odour. J. Invest. Dermatol. 41:3. WATSON, O.M. 1970. Proxemic Behaviour. Mouton, The Hague.
