This item is only available as the following downloads:
i WHAT'S WITH THE FLOWER BOYS? COMPARING KOREAN AND AMERICAN PERCEPTIONS OF FEMININITY AND MASCULINITY BY REBECCA ALICE FURLOW A Thesis Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the re quirements for the degree Bachelor of Arts Under the sponsorship of Heidi E. Harley Sarasota, Florida May, 2012
ii Dedication Page This is dedicated to all those who share the love for Korean culture, a nd to all of the incredible, talented hardworking people in the Korean entertainment industry. Especially the flower boys.
iii Acknowledgements I would first like to express my most sincere appreciation for Dr. Harley and all of her contributions and efforts in helping me achieve academic success. You have always given me exactly the right kinds of understanding, support, encouragement, and motiva tion throughout your time as my mentor. Thank you for holding my hand through the thesis process, for reassuring me of my competence in my times of doubt, for always having my back, and for being such a wonderful person in general. I would also like to t hank the other members of my thesis committee: Dr. Barton and Dr. Graham. Dr Barton, you were also my mentor for a short period during my time at New College. Y our high expectations for me were ultimately necessa ry in pushing me to work harder to achieve m y academic goals Thank you for caring about my future Dr. Graham, thank you for agreeing to take on my thesis despite all of the confusion and the obstacles thrown in your way Also, to Dr. Cooper : thank you for spending so much of your own time helping me sort out numbers. Without your help, I would have been lost. I would a dditionally like to thank all of my other professors who have ever taken me under their wing during my time here. Of course, I owe more than I could ever give back to my family. To my parents, Tina and Bill Furlow, who have always supported and provided for me to the absolute best of their abilities, t hank you for your continuing support throughout my time here, not to mention my life To Nana, thank you for always caring for me so much, and for continuing to check up on me even when I couldn't seem to find the time to return the favor. I would like to thank all of my friends, whose presence has me ant the world to
iv me. T o Jeremy Blackowiak thank you for your companionship and supp ort throughout the thesis process and for making me challenge my ways of thinking T o Winn Haslam thank you for ultimately convincing me to pursue the topic I love. To JP Paiva and Mackenzie Pawliger, though you may not realize it, you have been a tremendous source of support in my recent times of instability, and have kept me standing on my feet when I may have otherwise fallen. Thank you for accepting me as your bro. To Gwen Luffman, thank you for he lping me keep my head up through the final stretch. To others not mentioned by name here, you know who you are, and you know how grateful I am I would finally like to thank my translator Seongeun Park and my backtranslator (who will remain anonymous) for their hard work once again, I could not have done this without you! Lastly I would like to thank all of the anonymous people of the Internet, both in the U.S. and in South Korea, who participated in my study. Thank you for taking the time to help me de spite not even knowing who I am.
v Table Of Contents Dedication .......................................................................................................................... ii Acknowledgements .............................................. .............................................................iii Table of Contents ................................................................................................................v Abstract .. .............................................. .............................................. ......................... .....vii i Introduction .........................................................................................................................1 East and West: Differences Across Culture 5 Differences Across Gender ...12 Parallels in Culture and Gender ....14 Perceptual Implications: Femininity and Masculinity ..18 The Current Study .20 Methods ............................................................................................... ..............................23 Results ................................................................................................. .............. ................28 Discussion............................................................................................ ..............................33 References.......................................................................................... .. ..............................43 Appendix A ......................................................................................... ..............................51 Appendix B ........................................................................... .............. ..............................53 Appendix C ......................................................................................... ..............................54 Appendix D .............................................................................................................. .........55
vi List of Tables Table 1 : Image Ratings ....................................................................... ..................... .........30 Table 2 : Trait Ratings ......... .................................................................. ............................31 Table 3 : Acceptance Ratings .... .......................................................... ....................... .......33
vii List of Figures Figure 1: Androgyny Scale ... 47 Figure 2 : Image Ratings 48 Figur e 3 : Trait Ratings .. 49 Figur e 4 : Acceptance Ratings ... 50
viii WHAT'S WITH THE FLOWER BOYS? COMPARING KOREAN AND AMERICAN PERCEPTIONS OF FEMININITY AND MASCULINITY Rebecca Furlow New College of Florida, 2012 ABSTRACT South Koreans and (U.S.) Americans promote social norms (e.g., "flower boys" and muscular men) that suggest the cultures have divergent views of femininity and masculinity. In a formal investigation of these differences, 26 Americans and 13 Koreans completed a survey in which they rated the femininity and mascu linity of potentially androgynous individuals and personality traits, and indicated their feelings toward hypothetical cross gendered expression. Koreans rated individuals as being more masculine than did Americans, especially when rating ma les Koreans al so rated "playfulnes s" as being a predominantly masculine trait, whereas Americans rated the same trait as being predominantly feminine. When controlling for subjective differences in perceptions of femininity and masculinity, Americans had a more acceptin g attitude than did Koreans toward both a masculine female and a femini ne male, and both cultures were more accepting of the masculine female than of the feminine male. The results suggest that Korean androgyny is more feminine and less masculine than is A merican androgyny, and that this discrepancy is influenced by views of playfulness. _________________________ Dr. Heidi Harley Division of Psychology
1 What's With the Flower Boys? Comparing Korean and American Perceptions of Femininity and Masculinity A fter spending any amount of time on the Internet as an average American absorbing media related to Sout h Korean pop culture, it would become apparent that there are numerous di screpancies between (U.S.) American and (South) Korean cultures. One noticeable difference is that the men in Korean media often seem to be, in some way or another, more effeminate than the media's presentation of American men One would unavoidably encounter the concept of the idolized kkotminam or "f lower boy (literally, flower beautiful man [Turnbull, 2010]), i.e. Korean teenage to adult males who might be best described to an American audience through the following depiction taken from Eat Your Kimchi a popular Canadian video blog dedicated to Korean pop culture: Flower boys are pretty boys, VERY pretty boys. They're usually skinny, have gorgeous hair and have feminine features. They're not beefy or muscly [sic] but are quite lean. Fashion is extremely important to the m and they dress really really well, often with tight fitting clothing. They have exceptionally smooth skin, quite possibly from usage of BB Cream and sometimes they might wear gray or blue contacts [sic] lenses, along with circle lenses. They're ridiculou sly and unfairly pretty looking. (Martina & Simon, 2011) Though the prominence of this particular male beauty ideal in Korean culture has yet to be empirically examined, many online sites and blogs attest to its widespread desirability, declaring, "S outh K orean males are proud to b e called flower men' ( Ardery, 2012 ) This "flower boy" ideal poses in stark contrast to the most typical
2 American ideal of male beauty, which generally endorses a more muscular (and therefore more masculine) than average body t ype ( Barlett, Vowels, & Saucier, 2008 ). T his apparent cultural discrepancy in male beauty ideals would lead an American to ask: To what can the existence of flower boys' be attributed ? Do Koreans see them as less effeminate than Americans do? Do they see that level of femininity as fairly neutral? Or rather, do Koreans really see "flower boys" as being just as effeminate as we do, but actually prefer admittedly more feminine men ? I s it a bit of both?" To date, there are no systematically collected data i ndicating that Korean men are more effeminate than American men; that the neutral Korean male beauty ideal is objectively more feminine than the neutral American male beauty ideal; that what Koreans might perceive to be fairly neutral, or even masculine qualities in males Americans might perceive to be much more feminine; or that when controlling for potential cultural differences in subjective perceptions of male femininity Koreans would have a generally more positive or accepting attitude toward what they would perceive to be a high level of femininity than would Americans in reaction to what they would perceive to be an eq ually high level of femininity Despite the lack of empirical research on this subject, these ideas are still regarded in varying combinations as truths, even by scholars. One sociologist in particular, who has been cited in publications such as TIME Magazine and The Washington Post has created an entire website containing hundreds of pages of articles dedicated solely to "Korean s ociology through advertising, gender, and pop culture," in which he frequently cites and analyzes different aspects of the "flower boy" phenomenon and the feminization of men in Korean media, as can be seen from the titles of articles such as "The
3 effeminacy of male beauty in Korea," and "Why size matters: Feminine representations of men in Korean adverti sing" (Turnbull, 2008; 2010). In the absence of scientific data, he relies on his own observations, as well as the observations of others, of appar ent patterns in Korean media, and his and others' thoughts on how these patterns differ from those seen in Western media. Having been raised in the UK, his observations and lines of reasoning are inevitably informed by a Western mindset, unavoidably limiti ng his analyses to those allowed by views through this specific cultural lens For this reason, s cholars such as these would benefit greatly from more objective, empirical data in this area. While consuming Korean media, one may also possibly notice a rela tive absence of "tomboys," or females who are fairly masculine in appearance and mannerisms, as compared to females in American pop culture (and American culture in general), though this difference may not be quite as striking as the apparent cultural disc repancy in male expression s of femininity. Overall, Korean women may actually seem to more closely adhere to more traditional female gender role expectations, being portrayed as more passive and submissive than are women in American culture Jung & Lee (20 09) found this pattern to hold true for magazine advertisements through a content analysis of magazines from the two cultures They also found that women in American advertisements were more often portrayed in a way that emphasized their corporeal sexualit y. Similarly, Nam, Lee, & Hwang (2011) found that in adolescent Korean girls' magazines, Korean females were more often portrayed with more innocent or "cute" facial expressions ( such as smiling pouting or making childlike faces) than were
4 W estern female s who were more often shown in a way that emphasized their bodies as the main focus Academics have attributed this cultural discrepancy in portrayals of Korean and American women to a Korean ideal of beauty particularly femal e beauty as one that more strongly emphasizes innocence and purity ( Chen g, 2005; Jung & Lee, 2009; Nam et al. 2011 ). According to these findings, the A merican ideal of female beauty would be relatively more sexual ized a decidedly non childlike, and therefore less innocent c haracteristic than the Korean ideal. Because on the other side of innocence lies maturity, and maturity is unavoidably accompanied, even necessitated, by a certain autonomy and independence, it would seem that Americans favor a more mature, and therefore more indep endent quality in females, which would be consistent with the typically Western idealization of the "independent woman". As the quality of independence can arguably be tied to some extent to assertiveness (to be indepen dent, one must assert one' s independence), which is generally perceived to be a more masculine than feminine personality trait (Costa, Terracciano, & McCrae, 2001; Feingold, 1994) it could be proposed that the American female beauty ideal is more masculine than the more feminized Korean female beauty ideal. This discrepancy, when combined with the apparent feminization of the Korean male beauty ideal as compared to the relativ ely masculine American ideal, suggest s that Korean and American people may actually have divergent percepti ons of what it means to be "feminine" or "masculine," or perhaps different value judgments about these qualities However, again, empirical research has yet to break into this area, and therefore the validity of th ese ideas has gone unchallenged and remain s to be tested.
5 These predominantly informal observations are the true origin s of the current study's hypotheses H owever, the empirical literature on culture and gender differences in self construal and other aspects of personhood can lend support for th e plausibility of cultural differences in the perceptions and judgments of femininity and masculinity and therefore further justification for academic exploration in this area East and West: Differences Across Culture Collectively, members of Western and Eastern cultures in particular, people from the U S and from certain East Asian countries often differ to varying magnitudes in numerous psychological aspects, including attitudes, motivations, priorities, character traits, and self construals (Ch ang, Sanna, & Yang, 2003; Christopher, D'Souza, Peraza, & Dhaliwal, 2010; Kashima et al., 1995; Kim, Pan, & Park, 1998; Kim, Sohn, & Choi, 2011; Markus & Kitayama, 1991; Singelis & Brown, 1995 [as cited in Kashima et al., 1995]). In cross cultural psycholo gical research, arguably the most prominent, widely studied of these aspects is self construal, or self concept: the way in which individuals define, view and describe themselves. The main dimension of self construal is generally thought to encompass a spe ctrum ranging from interdependence to independence, in which more interdependent self construals more strongly emphasize relationships between the self and others, idealizing harmonious interdependence, and more independent self construals more strongly em phasize the separation between the self and others as defined by the individual's unique inner attributes, seeking to maintain independence (Markus & Kitayama, 1991). In cross cultural research, the terms "independent" and "interdependent" in reference to self construal are frequently used interchangeably with "individualist" and "collectivist," respectively; However the latter
6 set of terms (individualist, collectivist) is arguably more suited to discussing the general self construals of members of a cultu re, whereas the former set (independent, interdependent) is arguably more suited for use in reference to self construals at a more individual level (Cross, Hardin, & Gercek Swing, 2011). As with independent self construal, individualist self construal is c haracterized by a stronger focus on what separates the individual from the rest of society; similarly, as with interdependent self construal, collectivist self construal is characterized by a stronger focus on the individual's relationships to others and o ne's role within society (Cross et al., 2011; Kashima et al., 1995). Other terms associated with the interdependent collectivist end of this spectrum include ensembled communal and relational whereas the other, independent individualist end is ass ociated with terms such as autonomous agentic and separate (Kashima et al., 1995). Whatever the terms used, t hese fundamental differences in self concepts inevitably influence other areas of cognition in social as well as nonsocial thinking: those with more interdependent self construals are more aware of relevant "others" within a given context (Markus & Kitayama, 1991). Because of this elevated awareness, an interdependent self construal means higher susceptibility to the influence of the reaction s of others in the expression and experience of emotion and motivation, quite possibly leading to the restriction of the expression of certain emotions (such as anger) thought to promote separation between individuals. Additionally, more interdependent sel f construals should coincide with more other serving than self serving motives as compared to more independent self construals (Markus & Kitayama, 1991). According to Cross et al. (2011), individuals possess both independent and interdependent self constr uals, but cultural context typically promotes the development
7 of one or the other self construal more strongly ." Based on examinations of previous literature outlining general differences between Western and Eastern perspectives of personhood, Markus & Ki tayama (1991) conceptualized the classification of cultures according to self construal tendencies, following observations that generally, members of Eastern cultures tend to define themselves more in terms of their relationship s to others that is, to ho ld more interdependent, or collectivist, self con struals whereas members of W estern cultures tend to define themselves more in terms of how they are s eparate from others in society that is, to hold more independent, or individualist, self construals. T his I ndependent Western vs. Interdependent Eastern distinction has been supported by and reflected in existing research overall some of which directly demonstrates cultural differences in self construal (Christopher, D'Souza, Peraza, & Dhaliwal 2010; Kashima et al., 1995; Singelis & Brown, 1995 [as cited in Kashima et al., 1995]), and some of which implies these differences by demonstrating differences in socially related thoughts, motivations, and priorities (Kim, Pan, & Park, 1998; Kim, Sohn, & Choi, 2011). Further support for this particular cultural distinction can be gathered from research in disciplines outside of psychology, such as economics: Long predating Markus & Kitayama's (1991) proposal of cultural differences in views of the self, Hall (1976; as cited in Kim et al., 1998) proposed a comparable structure involving cultural differences in the importance placed on context in social interactions, wherein he named more typica lly Eastern cultures (such as those of East Asia) as "high context", and more typically Western cultures (such as the U.S. and those of certain European countries) as "low context". These classifications serve to outline the ways in which people in a given culture generally relate to one another, particularly in the areas of social bonds,
8 responsibility, commitment, social harmony, and communication (Kim et al., 1998). As the terms suggest, highe r context cultures tend to emphasize the importance of context ("the other") more strongly in these areas of interaction, whereas lower context cultures tend to be more individualized, exhibiting less concern for context. For high context cultures, this means deeper interpersonal involvement, stricter control of inne r feelings, emphasis on conformity, higher commitment due to a stronger dependence on close connections and relationships, and, as a result, a stronger social hierarchy. In contrast, low context cultures emphasize less involvement of individu als with other s, more explicit and nonpersonal communication, and less of a social hierarchy. These qualities, outlined through a business oriented perspective seem to match up quite well w ith the qualities of Independent Western and I nterdependent Eastern cultures as described by psychologists Markus & Kitayama (1991): high context characteristics, as depicted by Kim et al., 1998 (i.e., emphasis on the importance of context ["the other"], deeper interpersonal involvement, stricter control of inner feelings, and higher importance given to conformit y and social hierarchy), mirror Markus & Kitayama's depictions of cognitive effects and patterns associated with an interdependent view of the self (i.e., higher awareness of relevance of "others" within a given context, higher value assigned to relationships between individuals, suppression of certain emotions thought to promote separation between individuals, and more other serving than self serving motives ). One particular set of empirical findings incorporates both sets of ideas. Kim, Sohn, & Choi (2011) had 240 Korean and 349 Ame rican college students complete surveys examining the use of, and motivations behind the use of, social networking sites. In their regular use of the sites, Koreans were more motivated by social fac tors ("seeking
9 social support"), while Americans were more strongly motivated by more internal, individualized motives ("seeking entertainment"). Additionally, Koreans had smaller, yet more closely involved "friend" networks than did Americans, whose relat ively much larger "friend" networks consisted of much shallower friendships overall. This discrepancy can be explained by differences in social priorities and relationship building values between the cultures: Maintaining social networks requires time and effort; therefore, the larger someone's network, the less amount of these limited resources that person can afford to spend managing each relationship. The refore, the smaller networks more typical of Koreans are more conducive to closer, deeper, relationsh ips and more meaningful interactions than the larger networks more typical of Americans, in which, generally, relationships and interactions can only exist at a more casual and shallow level. Not having as many "friends" generally allows one to be more dee ply involved in existing relationships. Collectivist cultures (such as that of South Korea), whose conceptions of individuality are defined more in terms of the relatedness of people to one another than the unique inner qualities that separate them, tend t o place higher value on forming stronger, longer lasting relationships than the more short lived, relatively weaker relationships that tend to result from the more individualistic schools of thought characteristically fostered by individualist Western cult ures. As a whole, the findings of Kim et al. (2011) suggest that Koreans are more relational, socially concerned, and collectively interdependent than are Americans, who are more inwardly concerned and more collectively independent. Because Markus & Kita yama's (1991) proposed qualities of independence and interdependence so cleanly match with Hall's (1976) qualities of low contextualization
10 and high c ontextualization, respectively the two sets of ideas can be comfortably collapsed into one somewhat broad er, though potentially more clearly or fully defined paradigm, wherein the higher context, more interdependent mindset characteristic of certain Eastern cultures can be collectively re defined as an "outward focus," and the lower context, more independent mindset characteristic of certain Western cultures can be collectively re defined as an "inward focus." This Western/Eastern distinction is supported by more recent empirical research. Christopher et al. (2010) had 360 Americans (a group highly representative of "Western culture") and 367 Thais (a group that is arguably to some substantial extent representative of "Eastern culture") complete a survey incorporating the Self Construal Scale (SCS), an instrument developed i n previous studies to measure independent and interdependent self construal. Both the English and Thai versions of the survey had been previously shown to have adequate internal consistency and validity. The scale consisted of two factors (independent and interdependent) with 15 items per factor. Responses to the items were on a seven point scale ranging from "strongly disagree" to "strongly agree". As predicted, Americans scored higher on independent items than did Thais, and Thais scored higher on interde pendent items than did Americans thus implying that t he "Western" Americans had more independent self construals than the "Eastern" Thais, who held more interdependent self construals. Kashima et al. (1995) found similar differences in self construal be tween cultures commonly regarded as "Eastern" (Japanese and South Korean) and "Wes tern" (American and Australian). Five hundred ten "Eastern ers" and 292 "Westerners" completed a Collectivism scale, which was created using 28 items from Yamaguchi's Collecti vism
11 scale (1994; as cited in Kashima et al., 1995) that generally described hypothetical situations in which there was a conflict between personal and (friendship) group goals. Some items favored the personal goal, while others favored the group goal. Par ticipants rated each item on a 5 point scale according to how well they believed it described them. Because collectivism is the tendency to place group goals above personal goals in importance, ratings that favored group goals over personal goals were indi cative of more collectivist self construals. Survey responses revealed a tendency of Easterners to have more collective self construals than Westerners, who tended to be more individualistic overall. Additionally, Kashima et al. (1995) noted that Asian Haw aiians' scores on a measure of self construal were more indicative of collectivism than were those of Euro Hawaiians, whose scores were more strongly typical of individualism (Singelis & Brown, 1995; as cited in Kashima et al., 1995). Furthermore, though H all's (1976) theory rests perhaps a bit too comfortably on predominantly anecdotal evidence, empirical marketing research has si nce supported his claims. In an attempt to develop an instrument to investigate Hall's (1976) proposed dimensions of the major d efining aspects of low and high context cultures (i.e., social orientation, communication, commitment, and confrontation avoidance), Kim et al. (1998) created a compilation of items adapted from preexisting scales assessing similar or related concepts in psychology, such as social conformity (authoritarianism) need for cognition, and VALS (values). The researchers gave this measure to 96 American, 96 Chinese, and 50 Korean business management students to complete. Responses were analyzed in a way that made higher scores more indicative of higher context culture. As expected, the Chinese and Korean groups both scored significantly higher overall than
12 did the American g roup that is, they scored in a way more strongly indicative of high context cultures than did Americans, whose scores were more consistent with low context cultures (though the researchers did note that due to the nature of the instrument, the magnitude of these differences could not be determined). Differences Across Gender In light of the literature addressed thus fa r, it can be posited that there are indeed cultural discrepancies in self construal. However, culture is not necessarily the only factor a cross which self construal varies in some way: Self construal discrepancies have also been observed to exist across gender Specifically, studies conducted primarily in the U.S., but also across several other cultures, suggest that w omen tend to define the mselves more in terms of their relatedness to others, placing more value on their interpersonal relationships than do men, who tend to define themselves more in a way that emphasizes the separateness of the self from others as created and defined by charac teristics unique to the individual ( Clancy & Dollinger, 1993; Josephs, Markus, & Tafarodi, 1992 ; Kashima et al., 1995 ) As noted by Kashima et al. (1995), the claims of theorists that women are more relational than men (Gilligan, 1982; Miller, 1986; both as cited in Kashima et al., 1995) do not go entirely unsupported. Fairly recent empirical findings have been consistent with these assertions (Clancy & Dollinger, 1993; Josephs et al., 1992 ). In an investigation of gender differences in sources of self est eem, Josephs et al. (1992 ) had 47 women and 43 men report their feelings about their own abilities in different areas, the importance they placed on being skillful in those areas and how they thought they compared to others who were skilled in the same areas in which they perceived themselves as being highly
13 skilled Men with higher self esteem perceived themselves as being uniquely superior in their abilities while men with lower self esteem did not seem to think they were particularly more skilled tha n others. W omen did not follow this self esteem pattern, instead yielding results most similar to men with low self esteem across the board. Additionally, 35 women and 30 men completed a word recall test that included words both in association with and not in association with important people in participants' lives Women with higher self esteem showed better recall for information tied to important others than not tied to important others, whereas women with low self esteem showed no difference in recall b etween the two categories; men with both high and low self esteem scored most similarly with women with low self esteem. These results suggest that for men, self esteem is more typically determined by individuation, and the perceived desirability of distin guishing personal characteristics, whereas for women, self esteem is more strongly reliant on the quality of connections and relationships to others more simply state d, self esteem is derived from separation and independence for men, and connection and i nterdependence for wo men. In further support of a self construa l difference across gender, Clancy & Dollinger (1993) asked 142 women and 59 men to create a personal photo essay out of any 12 autobiographical pictures of any kind they felt would best illust rate who they w ere as they viewed themselves. W omen included more pictures of themselves in social situations than did men, who chose to include more photos in which they were alone. Considering the knowledge that participants chose pictures of subjects th at they thought best demonstrated their views of themselves, thereby creating visual representations of their self construals, the more social focus of women and the more individualized focus
14 of men support the assertion that women are more relational in t heir self construals than men, whose self construals tend to put relatively more emphasis on separation. When viewed collectively with the results of Josephs et al. (1992), these findings lend support to the already common contention that generally, women, as compared to men, are more socially minded and place higher value on interdependent connection. This gender difference in self construal tendencies appears to parallel the self construal discrepancies found between Eastern and Western cultures, where f emales, with more relational, socially focused self construals as compared to those of males, fall generally on the same inwardly focused end of a self construal spectrum as do Easterners when compared to Westerners, whereas males, with more separatist, ou twardly focused self definitions as compared to those of females, fall generally on the same side of the spectrum as Westerners Parallels in Gender and Culture In addition to this apparent gender culture parallel in self construal, research has demonstrated comparable differences in other areas some of which seem to be only distantly (if at all) related to self construal, but many of which can be easily tied back into self construal through the grouping of social ly related aspects into "independent" and "interdependent" qualities. For example, Koreans (the Eastern culture of interest in the current study) tend to have higher negative affect in general than do Americans (the Western culture of interest) ( Chang, Sanna, & Yang, 2003), and as the parallel would dictate, women tend to have higher negative affect than do men ( Costa, Zonderman, McCrae, & Coroni Huntley, 1987 [as cited in Thomsen et al., 2005]; Fujita, Diener, & Sandvik, 1991; Thomsen et al, 2005; Vesg a Lpez et al., 2008). In a similar, more
15 specific vein, Asian Americans report higher anxiety and depression than do Euro Americans (Okazaki, 1997), just as women tend to report higher anxiety and sadness than do men (Costa et al., 1987 [as cited in Thoms en et al., 2005]; Thomsen et al., 2005; Vesga Lpez et al., 2008). Additionally, various studies and meta analyses of gender differences in personality traits have fairly consistently shown that males tend to be more assertive than females, whereas females tend to be more trusting, nurturing, and agreeable (Costa, Terracciano, & McCrae, 2001; Feingold, 1994; Maccoby & Jacklin, 1974 [as cited in Feingold, 1994]). When more closely examining these personality traits and their definitions in a socially oriente d light, a fairly clear distinction arises wherein the more typically "male" and more typically "female" qualities can be grouped together as more conducive to "independence" and "interdependence", respectively. The more predominantly male c haracteristic o f assertiveness is more likely to aid in the separation between individuals than in forming cl oser, more interdependent bonds and is therefore much more conducive to independence than to interdependence, whereas the more predominantly female traits of tru st, nurtu rance, and agreeableness all social in nature all involve the reliance of people on o ne another, and as a result, facilitate or are required for interdependence. This classification of males and females as more on the independent and interdepe ndent ends of a spectrum (respectively), stemming from empirical findings of personality trait differences across gender, serves as additional support for the previously outlined gender culture parallel. Kashima et al. (1995) recogniz ed this apparent parallel, and tak ing issue with its broadness particularly as a result of the existing literature's ostensible ambiguity in definitions of individualism and collectivism proposed that culture and gender
16 differences in self construal may actually be cha racterized by different sets of psychological dimensions. Specifically, they professed that discrepancies can be found along three main dimensions: the individualistic, which encompasses the independent, autonomous, "agentic," separate aspect of the self; the collective, which concerns the relationship between the individual and the collective (society); and the relational, which addresses the relationships between the individual and other individuals, describing the degree to which the self is construed in a way that emphasizes relatedness with other selves. The researchers argue that gender differences in self construal are most prominent in the relational dimension, whereas cultural differences are the most pronounced in the individualist and collectivist dimensions that is, gender differences are more strongly defined by perceived emotional relatedness to others, whereas cultural differences are more strongly defined by perceptions of self agency. To test this proposal, the researchers conducted a surv ey of 510 peo ple from Japan and South Korea and 292 people from Australia and the mainland United States cultures w idely regarded as collectivist and individualist, respectively. The survey consisted of four scales related to the proposed three main dimen sions of self construal: the Collectivism scale, the Kanjin shugi scale, the Allocentrism scale, and the Friendship Questi onnaire. The Collectivism scale, as previously detailed, was the primary measure for collectivism in self construal. The Kanjin shugi (translated as "between people ism" or contextualism) scale, developed by Hamaguchi (1987), was used to assess self construal along the relational dimension, as the content o f the scale generally emphasized the emotional relatedness of the self with others Participants responded to items on this section, which were in the form of self statements about feelings toward several specific
17 close others, via a 5 point scale ranging from "strongly disagree" to "strongly agree". The Allocentrism scale was used to m easure collectivism at a more individual (as opposed to cultural) level, and included items relating to an individual's relationships with different groups of people (such as friends and family). Items on this scale were in the form of allocentric (other f ocused) or ideocentric (self focused) statements about the self, to which participants responded "true" or "false". For the Friendship Questionnaire, participants identified their five best friends, and how many of these friends were friends with each othe r this was used as a measure of friendship group cohesiveness, which has been used to measure allocentrism in the past (according to Triandis , closer knit groups are more characteristic of collectivism). A factor analysis of the Collectivism scal e yielded three factors: Collectivism, which was defined positively as positive feelings toward the friendship group and negatively as the desire to leave the group; Agency, seen as tapping into individualism, and defined positively by an emphasis on indep endence of action and opinion, and negatively by emphasizing conformity to the group; and Assertiveness, also seen as tapping into the individualist dimension, as it was defined by verbal expression of opinion. Factor analysis of the Kanjin shugi scale rev ealed a single factor: Relatedness, or the relational dimension of self construal. Participant responses were examined collectively in terms of these factors. As would be expected, in terms of scores on items within specific factors, the collective and rel ational dimensions correlated positively with each other, and negatively with the Agency factor (though not the Assertiveness factor) of individualism. Allocentrism and cohesiveness correlated positively with each other and with collectivism and relatednes s, and negatively with individualism in the Agency
18 factor. Along the collective dimension of self construal, differences were present across culture, but not across gender: As expected, the Japanese and Korean ("collectivist") samples scored higher than th e American and Austr alian ("individualist") samples Expected cultural differences were also found in the Agency and Assertiveness dimensions (sub dimensions of individualism), with Americans and Australians scoring much higher in individualism overall tha n Asians S omewhat surprisingly, no significant gender differences w ere found in these dimensions. However, a long the Relational dimension, both cultural and gender effects were found: Women were generally more relational than men across cultures, and most notably, Korean men and women scored relatively high in the relational dimension as compared to men and women from the other cultu res. When these Korean results were analyzed alongside the results from Japanese participants as a collective "Eastern" group against t he "Western" group's results, this pattern disappeared yielding no significant differences across culture However, because the Eastern culture of focus within the current study is that of Korea and not Japan (which is actually closer to America n culture than to Korean culture in some social aspects [Park, Killen, Crystal, & Watanabe, 2003]), the Japanese results can be disregarded, revealing a cultural discrepancy comparable to the gender discrepancy in the relational dimension. Therefore, after a narrowing of cultural focus from "East" versus "West" to the more specific and currently relevant "Koreans" versus "Americans t hese observations further contribute to the plausibility of the gender culture parallel along the inward to outward focus sp ectrum.
19 Perceptual Implications: Femininity and Masculinity Th e importance of th e apparent gender culture overlap, where characteristics seem to be somewhat shared between women and Koreans and between men and Americans via more typically outward and inward foci, respectively is that it potentially points to cultural d ifferences in perceptions of femininity and masculinity Consider an analogy: I magine a person from Alaska and a person from Ecuador. The person from Alas ka is constantly exposed to temperatures that, as compared to the rest of the world, are generally relatively cold, whereas the person from Ecuador is constantly exposed to temperatures that are generally relatively hot. However, because of the constant ex posure to these fairly extreme temperatures, the temperatures become more normalized to the respective people: the person from Alaska will have a more perceptually normalized temperature range towards the colder end of the objective temperature spectrum, w hereas the person from Ecuador will have a more perceptually normalized temperature range towards the hotter end of the objective temperature spectrum. As a result, the person from Alaska would be more sensitive to objectively hotter temperatures and the p erson from Ecuador would be more sensitive to objectively colder temperatures, as these extremes would be far away from their normalized centers. This would mean that if these two people were to meet in someplace such as South Carolina at the same point in time the person from Alaska, with a lower temperature normalization, might perceive the weather to be fairly hot, while the person from Ecuador, with a higher temperature normalization, might perceive the weather to be fairly cold. Just as temperatures t oward the colder end of a spectrum are normalized to the person from Alaska, resulting in higher sensitivity to heat, and temperatures toward the hotter end of the spectrum are more normalized to the
20 person from Ecuador, resulting in higher sensitivity to cold, a similar structure could be proposed for perceptions of masculinity and femi ninity across cultures. Because (as the literature suggests) qualities more prevalent in Korean culture than in American culture tend to be shared more strongly with women, or femininity, than with men, or masculinity, these coincidentally more characteristically "feminine" qualities would presumably be more normalized to Koreans Similarly, the coincidentally more characteristically "masculine" qualities that tend to be more prominent within members of American culture would presumably be more normalized to Americans Therefore, Koreans due to their potential relatively higher exposure to and plausibly resulting normalization of more "feminine" qualities, could be more sensi tive to ma sculinity than would be Americans and would therefore judge a ny given quality as more masculine than wo uld an American. T he same could be pr oposed of Americans and masculinity: that Americans, possibly having relatively higher exposure to and plausibly resulting normalization of more "masculine" qualities, may be more sensitive to femininity than are Koreans, and would therefore judge a ny given quality as more feminine. Current Study Korean beauty ideals seem to be objectively more effeminate than American beauty ideals. Perhaps in response to this discrepancy, there are numerous claims throughout various media that femininity in general seems to be much more desirable, prevalent and accepted in Korean c ulture than in American culture. However, there is no empirical research that directly tests these claims A review of differences across culture and gender mainly in terms of self construal, suggests differences across people fr om W estern and Eastern cultures.
21 Westerners tend to hold more independent self concepts and Easterners tend to hold mor e interdependent self concepts ( Christopher et al., 2010; Kashima et al., 1995; Kim et al. 1998; Kim et al. 2011 ; Markus & Kitayama, 1991; Singelis & Brown, 1995 [as cited in Kashima et al., 1995] ) Specifically, Koreans, as members of Eastern society, tend to define themselves more in terms of their relationships to other individuals and the roles they play in the lives of others, whereas A mericans, as members of Western society, place more emphasis on their own unique qualities that separate them from other individuals. In addition, Koreans are more socially motivated at least during certain online activities whereas Americans have more internal, individualized motives (Kim et al., 2011). There also exist fundamental differences along the same veins across people of female and male genders. Compared to males, females tend to define themselves more in terms of their relatedness to others, and to give more importance to interpersonal relationships to have a more "outward" focus whereas males tend to define themselves more in a way that emphasizes the separateness of the self from others via the indiv idual's unique characteristics to have a more "inward" focus ( Clancy & Dollinger, 1993; Josephs et al. 1992 ; Kashima et al., 1995 ) T hese discrepancies across culture and gender seem to parallel each other in their ability to be placed along an inward focus to outward focus spectrum with women (femininity) and Eastern culture (Koreans) lying toward the outward focus end, and men (masculinity) and Western culture (Americans) lying toward the inward focus end. Therefore, Koreans may be more sensitive to masculinity than are Americans whereas Americans may be more sensitive to femininity than Koreans. In this case one
22 should predict that given an image of a fairly androgynous individual, and asked to rate the femininity and masculinity of this person, Koreans would rate the person as more masculine than would Americans, who would ra te the person as more feminine. Furthermore, when given several personality traits and asked to rate them in terms of femininity and masculinity, one could predict that if there are cultural differences in perceptions of what defines femininit y and masculinity Koreans and Americans may differ in how they would rate these traits. Significant differences in particular traits may help to explain with more detail exactly which qualities sway the opinions of mem bers of the two cultures into perceiving a given target as more or less feminine or masculine. O ne might expect Koreans to rate what the literature suggests are typically thought of as more fem ale" qualities as to some degree less feminine than would Americans; similarly, that Koreans may rate more typically "male" qualities as somewhat less masculine than would Americans. In particular, the previously mentioned Korean emphasis on innocence in ideal s of beauty in general (Cheng, 2005; Jung & Lee, 2009; Nam et al., 2011) might encourage the expressions of child like behaviors, or playfulness, in attempts to be more desirable, regardless of gender This would mean that playfulness would be seen as a pos itive expression f or both males and females. As playfulness seems to be thought of as a more uniquely feminine trait in American culture, one might expect a large discrepancy in ratings of this particular trait specifically, that Koreans would rate it as being much more masculine and less feminine than would Americans. An additional possibility is that when presented with a hypothetical relatively masculine female and a hypothetical relatively feminine male in a way that would control for cultural differ ences in subjective perceptions of femininity and masculinity, Koreans
23 might hold a more accepting attitude toward the feminine male than would Americans, whereas Americans might hold a more accepting attitude toward the masculine female than would Koreans Furthermore, it would reasonably follow that Koreans would be more accepting toward the feminine male than they would be toward the masculine female, whereas Americans would be more accepting toward the masculine female than they would be toward the femi nine male. These issues were addressed by having Korean and American participants complete an online survey in which they rated the femininity and mascul inity of different individuals and personality traits, and indicated their feelings toward a hypotheti cal masculine female and feminine male Method Participants In total, 99 respondents elect ed to participate in the survey: 33 responses to the Korean version and 66 responses to the English version. Data from 13 Koreans (eight males, four females, one oth er) and 26 Americans (10 males, 15 females, one unspecified) ages 18 40 were analyzed. Eleven participants of other nationalities completed the survey; however, as this study is only focused o n differences in the perceptions of Koreans and Americans, their data were omitted from analysis. Additionally, data from 46 participants (19 from the Korean version and 27 from the English version) were excluded from analysis due to their nondisclosure of nationality, or their admittance in the demographics section to being under 18 years of age. Nationality/ethnicity classification was determined by participants' agreements with definitional statements: "Korean" was defined as "One or both of my parents are
24 Korean, and I have lived in Korea for most of my life." Korean American was defined as "One or both of my parents are Korean, but I have lived in the U.S. for most of my life (or for about as long as I lived in Korea); or both my parents are American, but I have lived in Korea for most of my life." "American" was defined as "Neither of my parents are Korean, and I have lived in the U.S. for most of my life; or one or both of my parents are American, but I have not lived in the U.S. for most of my life." "Other" was defined as "Neither of my parents are Korean or American, and I have not lived in the U.S. or South Korea for most of my life." American participants were recruited via posts to "Reddit," an international online forum. Posts were made to academically related subsections of the site, and included a s mall amount of information about the current study as well as a link to the survey. Korean participan ts were recruited via posts to the Korean online forum "todayhumor.com," which was chosen based on suggestions, given via an online forum and a social netw orking site, from individuals with knowl edge of Korean Internet culture who agreed that the site was a sufficient Korean equivalent to Reddit Posts were made to a thought related subsection of the site, and were translations of the English posts. Materia ls and Procedure The self administered online survey was originally written in English. It was translated into Korean by a bilingual individual (who received payment for her services), and then back into English by a different bilingual individual (who wa s offered, but refused payment for her services), to assure the retention of critical wording and connotations, and to assure the comparability of English responses to Korean responses.
25 The English backtranslatio n was reviewed and judged to be sufficientl y comparable to the English version in all areas of concern. The survey was prefaced with an initial "Informed Consent" segment outlining the purpose, content, risks and benefits of the study as well as other pertinent information, and providing responden ts with the option to consent or not consent to participate in the study. To minimize potential respons e biases, participants were informed that the study aimed to investigate perceptions of femininity and masculinity, but not that it was cross cultural. The main body of the survey was composed of four sections, the general content and approximate lengths of which were provided to participants in the initial description of the survey's content and again before each section. Section 1 consisted of eight dif ferent images, each followed by a set of questions assessing respondents' perceptions of the femininity and masculinity of the subjects in the images. Section 2 asked respondents about their perceptions of the femininity and masculinity of six different pe rsonality traits. Section 3 separately presented a hypothetical masculine female and feminine male, and asked participants about thei r judgments of the two people. Section 4 had participants provide their demographic information. Comment fields were provid ed throughout the survey to give respondents the opportunity to share any additional relevant thoughts or explanations they may have had during their participation. The eight images in Section 1 each pictured one of eight different, fairl y androgynous indi viduals (see A ppendix A for these images ). The subjects of the photos were male and female Korean and American musical artists (two Korean females, two Korean males, two American females, and two American males). Images in which
26 subjects were more androgy n ous were chosen over images in which subjects were more strongly feminine or masculine so as to minimize any potential ceiling or floor effects in ratings of masculinity or femininity. Because the issue of interest in this study was the incidence of Koreans' and Americans' perceptual differences, rather than absolute magnitudes of the masculinity and femininity of the subjects in the images, the exact quantifications of the androgyny of the images' subjects were not of concern; therefore, in choosing images for the survey, informal opinion polls were sufficient for roughly assessing the magnitude a nd judging the appropriateness of the androgyny of potential image subjects. These polls mainly consisted of informally asking peers chosen through convenience sampling to provide their opinions about the levels of androgyny exhibited by subjects in potent ial images. Additionally, femininity and masculinity ratings for the image subjects given in a small pilot study by participants once again chosen via convenience sampling were assessed and judged to indicate sufficient androgyny Each picture in Section 1 was followed by two sets of question items: femininity ratings and masculinity ratings. These items required participants to rate the femininity and masculinity of the pictured individual in three aspects of expression: Appearance (facial features and bo dy), Presentation (posture and facial expression), and Styling (clothing and hair). Responses to these ite ms were on 5 point scales in which 1 = Not feminine/masculine at all, 2 = Somewhat feminine/masculine, 3 = Fairly feminine/masculine, 4 = Very feminin e/masculine, and 5 = Extremely feminine/masculine. Femininity and masculinity were measured on separate scales in accordance with Bem's (1974) findings that participants perceive these measures as independent from one another. For the sake of consistency, femininity and masculinity
27 rating sets varied in order of appearance on the survey according to the gender of the corresponding pictured individual: the femininity set appeared before the masculinity set for female image subjects, and the masculinity set a ppeared before the femininity set for male image subjects. See Appendix B for example items from this section. Section 2 presented six personality traits of which all but one (Playfulness) have been found to be more prominent in males than in females, or vice versa: Assertiveness, Playfulness, Confidence, Shyness, Cautiousness, and Neuroticism (Costa et al., 2001; Feingold, 1994; Maccoby & Jacklin, 1974 [as cited in Feingold, 1994]) D espite not bein g a particular quality of interest in past literature relating to personality dif ferences across gender, Playfulness was included in consideration of the previously discussed relative emphasis on innocence in Korean ideals of beauty, and the proposed impli cations of this cultural difference in terms of perceptions of childlike expressions (i.e. playfulness) Using the 5 point femininity and masculinity scales from Section 1, participants assigned femininity and masculinity ratings to each of the six traits. See Appendix C for example items from this section. Section 3 presented two separate descriptions in each of which respondents were asked to imagine themselves in a hypothetical situation where they saw a very androgynous person whose gender they could n ot discern. In the first description, a friend eventually told respondents that the stranger was female, and in the second description, the same friend eventually told respondents that the stranger was male. Immediately following each of the two descriptio ns was a set of rating items, each consisting of six self statements concerning respondents' fe elings toward the hypothetical stranger 's levels of femininity or masculinity as related to their gender ( levels of masculinity for the
28 female, and levels of femininity for the male). Responses to statements were on a 7 point scale ranging from "Strongly disagree" to "Strongly agree". The set measured three aspects of attitude: Self Comfort, Social Acceptability, and Cultural "Okayness". Self Comfort referred t o participants' levels of comfort with the femininity or masculinity exhibited by the hypothetical individual (i.e., "I would feel comfortable seeing a female that masculine"). Social Acceptability referred to how socially acceptable participants viewed th e individual's levels of masculinity or femininity to be within the context of the participants' cultures (i.e., "I feel like it's socially acceptable for a male to be that feminine"). Cultural "Okayness" referred to how "okay" participants thought other p eople in their country would be with the individual's levels of femininity or masculinity (i.e., "I feel like most people from my country would be okay with a female being that masculine"). Statements were also either positive ("I would feel comfortable") or negative ("I would feel uncomfortable"), yielding two statements per aspect of attitude. See Appendix D for example items from this section. In Section 4, participants reported their gender, age, and nationality/ethnicity. Detailed definitions were pr ovided on the nationality/ethnicity item for Korean, Korean American, American, and Other, and an extra comment field was provided for participants who were unsure of their classification to allow them to explain their individual living situations. Upon co mpletion of the survey, participants were again t hanked for their participation.
29 Results Section 1: Image Ratings Androgyny ratings were created from femininity and masculinity ratings by subtracting each masculinity rating from its corresponding femin inity rating, so that androgyny ratings ranged from 4 to +4, where more negative ratings indicated more masculinity overall, more positive ratings indicated more femininity overall, and ratings of 0 indicated complete androgyny (Figure 1) Each participan t's androgyny ratings for Appearance, Presentation, and Styling for each image were then averaged into single overall expression ratings, yielding one expression androgyny rating per image per participant. As expected, Koreans rated images overall (i.e., of both male and female subjects) as significantly more on the masculine side of androgyny ( M = 0.6667, SD = 0.7061) than did Americans, who actually rated images overall as slightly more on the feminine side of androgyny ( M = 0.0540, SD = 0.4208), t (37) = 4.00, p = 0.0003. For female subject image s, the difference was less pronounced ; a one tailed t test in the hypothesized direction show ed that Koreans rate d images of females as significantly more on the masculine side of androgyny ( M = 0.3397, SD = 0.8789) than did Americans, who again rated the same images as somewhat more on the feminine side ( M = 0.0929, SD = 0.6580), t (37) = 1.73, p = 0.0461 (Figure 1) The two tailed t test did not yield a statistically significant difference, t (37) = 1.73, p = 0.0922. For male subject image s, the difference was large; Koreans rated male images as much more on the masculine side of androgyny ( M = 0.9936, SD = 0.7426) than did Americans, who again rated the same
30 images as slightly more on the feminine side ( M = 0.0150, SD = 0.6636), t (37) = 4.30, p = 0.0001 (two tailed test ; Figure 2 ) Table 1 Mean (SD) Androgyny Ratings for Subject Images Group ______________________________________________________ Subject Image Korean s Americans Female 0.3397 (0.8789) 0.0929 (0.6580) Male 0.9936 (0.7426) 0.0150 (0.6636) Both 0.6667 (0.7061) 0.0540 (0.4208) = p < 0.05 Section 2: Trait Ratings For traits, androgyny ratings were again created by subtracting masculinity ratings from their co rresponding femininity ratings to form single androgyny ratings for each trait. Differences in Korean and American participants' androgyny ratings of Assertiveness closely approached, but did not quite achieve statistical significance at the = 0.05 leve l; Americans rated the trait as slightly more on the masculine side of androgyny ( M = 2.0769, SD = 1.1974) than did Koreans, who also rated the trait as more masculine overall, though not to quite as great an extent ( M = 1.3077, SD = 1.1821), t (37) = 1. 90, p = 0.0654. However, as expected, Koreans rated Playfulness as far more
31 on the masculine side of androgyny ( M = 1.2308, SD = 1.4233) than did Americans, who rated the trait as about as equally far on the feminine side of androgyny ( M = 1.1600, SD = 1. 0677), t (37) = 5.84, p < 0.0001. Analysis of the other four traits (Confidence, Shyness, Cautiousness, and Neuroticism), yielded no further significant differences across nationality (Figure 3 ) Table 2 Mean (SD) Androgyny Ratings for Traits Group ______________________________________________________ Trait Korean s Americans Assertiveness 1.3077 (1.1821) 2.0769 (1.1974) Cautiousness 1.0000 (1.6833) 1.0800 (1.0770) Confidence 0.9231 (1.0377) 1.3077 ( 1.1232 ) Neuroticism 0.2308 (2.1274) 0.8462 (1.2551) Playfulness* 1.2308 (1.4233) 1.1600 (1.0677) Shyness 2.0000 (2.1213) 1.4615 (1.1741) = p < 0.05 Section 3: Attitude Ratings In order to analyze acceptance, indicated by levels of agreement with statements concern ing feelings toward the hypothetical individuals, r esponses to negative versions of statements were reversed and averaged with their positive counterparts, resulting in three
32 adjusted attitude ratings per hypothetical individual per participant (Self Comfo rt, Social Acceptability, and Cultural "Okayness"). Ratings of these three aspects of attitude were then further averaged together to form an overall Acceptance rating. Note that the reliability and validity of these measures as a true indicator of "accept ance" has not yet been assessed, and should therefore be viewed as preliminary As expected, a matched pairs t test of differences between ratings for the hypothetical masculine female and feminine male showed that Americans were significantly more accep ting of the masculine female ( M = 4.9487, SD = 0.9065) than they were of the feminine male ( M = 4.4231 SD = 0.8784 ), t (25) = 2.86, p = 0.0085. However, contrary to expectations, Koreans followed a similar trend, also being significantly more accepting of the masculine female ( M = 4.1923, SD = 0.8046) than they were of the feminine male ( M = 3.5000, SD = 0.9329) though to a slightly le sser extent, t (12) = 2.58, p = 0.0243. As predicted, Americans were more accepting of the masculine female ( M = 4.9487, SD = 0.9065) than were Koreans ( M = 4.1923, SD = 0.8046), t (37) = 2.55, p = 0.0152. However, opposite to expectation, Americans were mor e accepting of the feminine male ( M = 4.4231, SD = 0.8784) than were Koreans ( M = 3.5000, SD = 0.9329), t (37) = 3.03, p = 0.0044. Koreans ( M = 0.6923, SD = 0.9691) and Americans ( M = 0.5256, SD = 0.9377) did not differ significantly in differences in acce ptance levels of the masculi ne female and the feminine male. T he two groups were both more accepting of the masculine female than they were of the feminine male, by about the same amount, t (37) = 0.52, p = 0.6078 (Figure 4 )
33 Table 3 Mean (SD) Acceptance Scores for Hypothetical People Group ______________________________________________________ Person Korean s* Americans* Masculine female* 4.1923 (0.8046) 4.9487 (0.9065) Feminine male* 3.5000 (0.9329) 4.4231 (0.8784) = p < 0.05 Discussion A s expected, the visible charac teristics of other individuals that Americans judge as being fairly androgynous or somewhat feminine, Koreans judge as being much more masculine Apparently Koreans are much more perceptually sensitive to m asculinity than are Americans, especially when assessing qualities as exhibited by men. Additionally, Koreans regard "playfulness" as being a predominantly masculine trait, whereas Americans regard it as being predominantly feminine However, contrary to p redictions, Americans are actually more accep ting of what they perceive to be highly effeminate males than are Koreans of males they perceive to be equally as effeminate As predicted, Americans are also more accepting than are Koreans of high levels of m asculinity exhibited by females. In general, Americans have a higher acceptance (as compared to Koreans) of cross gendered expression overall.
34 Image Ratings Koreans rated image subjects overall as being significantly more masculine than did Americans a finding which is consistent with the proposal that Koreans are more perceptually sensitive to masculinity than are Americans, who are in turn more perceptually sensitive to femininity. As previously outlined, this proposal was derived from the reasoning that Koreans, due to their apparent general tendency to possess qualities and traits shared more strongly with femininity than with masculinity along the proposed "outward focus to inward focus" spectrum, could therefore in the most simplifi ed terms be des cribed as being objectively more feminine than are Americans, whose more typically possessed qualities are shared more strongly with masculinity along the same spectrum, thus warranting the description of being more objectively masculine. As a consequence of Koreans' relatively high er expression of more feminine qualities as compared to Americans' relati vely higher expression of more masculine qualities, Koreans would unavoidably experience higher exposure to more objectively feminine qualities than those e xperienced at equivalently high exposure by Americans, and would thus normalize these objectively higher levels of femininity as being basically neutral, androgynous centers of a shifted subjective masculinity to femininity scale. Accordingly, Americans wo uld experience higher exposure to more objectively masculine qualities and would thus normalize these objectively higher levels of masculinity as being basically neutral, androgynous centers of a shifted subjective masculinity to femininity scale. These sh ifted centers of normality would yield a discrepancy in the two populations' judgments of levels of femininity and masculinity
35 exhibited by a given target, as the two cultures would be assessing these levels according to their own subjective scales, which would in definition differ from one another. Applying a more specific example to this model, because the neutral point on the subjective Korean masculinity to femininity scale would be shifted more toward femininity as compared to the neutral point on the subjective American masculinity to femininity scale (which would be shifted relatively more toward masculinity), an objectively completely androgynous person (if such a person can exist) would fall somewhere on the masculine side of Koreans' subjective ne utral point and would thus be judged by Koreans as being more on the masculine side of androgyny; similarly, the same androgynous person would fall somewhere on the feminine side of Americans' subjective neutral point and would thus be judged by Americans as being more on the feminine side of androgyny. Because the results from the image judgments were consistent with this pattern, they provide support for the plausibility of the outlined model, which at its core rests on the theoretical "outward focus to i nward focus" spectrum paradigm constructed within the original empirically based argument according to findings from past research relating to both culture and gender. Markus & Kitayama's (1991) classification of "Eastern" and "Western" cultures along two opposite ends of a self construal spectrum ranging from Interdependent to Independent set the foundation for the connection to parallel structures. "Eastern" and "Western" cultures appear to adhere to this general trend in mindset, wherein Easterners (spec ifically Koreans) tend to be more socially oriented and outwardly focused, placing more importance on context in a given situation than do Westerners (specifically Americans), who tend to be more concerned with separateness
36 and inwardly focused, placing re latively less importance on context in a given situation (Chang et al., 2003; Christopher et al., 2010; Kashima et al., 1995; Kim et al., 1998; Kim et al., 2011; Markus & Kitayama, 1991; Singelis & Brown, 1995 [as cited in Kashima et al., 1995]). Gender d ifferences appear to follow this trend in similar areas as well, with men and women sharing a great deal of overlap with Americans and Koreans respectively on the inward to outward focus spectrum: Men, like Americans, tend toward a more separatist mindset, most highly valuing what distinguishes the self from the other, whereas women, like Koreans, tend toward a more social mindset, most highly valuing what connects the self to others ( Clancy & Dollinger, 1993; Josephs et al. 1992 ; Kashima et al., 1995 ). G iven these parallels in cultural and gender differences along the spectrum, it was possible to construct the current line of logic that yielded the aforementioned predictions about perceptual scale shifts, which were ultimately supported by the data from i mage ratings analyzed here More detailed analysis of these data showed the cultural difference in ratings of male image subjects to be much greater than the cultural difference in ratings of female image subjects This trend was also proposed in the orig inal argument as a possible reflection of the seemingly much larger cultural discrepancy in ideals of male beauty than in ideals of female beauty: The "flower boy" aesthetic, which Westerners seem to perceive as being highly effeminate (Ardery, 2012; Marti na & Simon, 2011; Turnbull, 2008; Turnbull, 2010) appears to pose in sharp contrast to the American male beauty ideal, which tends to endorse a more muscular than average b ody type (Barlett et al.,
37 2008). However, cultural differences in female beauty idea ls seem arguably less pronounced, potentially being defined predominantly by a cultural discrepancy in the perceived desirability of innocence (Cheng, 2005; Jung & Lee, 2009; Nam et al., 2011). Trait Ratings The latter contention regarding the potential role of innocence in cultural discrepancies in beauty ideals could arguably be closely tied to the finding that for Koreans, playfulness is a predominantly masculine trait, whereas for Americans, it is predomin antly feminine. The relationship between playfulness and innocence is arguably mainly defined by a childlike quality: "Playfulness" is defined as "f ull of fun and high spirits; frolicsome or sportive (Playful, 2009), whereas "innocence" is defined as "u nc orrupted by evil, malice, or wrongdoing; sinless (Innocent, 2009). Playfulness implies an element of innocence, as it is light hearted and free from the presence of malice though it also seems to include an additional, more energetic element. Assuming a substantial overlap of these two qualities, the cultural discrepancy found in the perception of playfulness is arguably consistent with the assertion that Koreans regard innocence as a positive trait to possess regardless of gender. Because of this appa rent role of innocence in the cultural discrepancy in the definitions of masculinity and femininity, it would have been wise to include "Innocence" as one of the target traits to be rated in the survey, and it would be a useful addition to future expansion s of the current research. Attitude Ratings The somewhat unexpected trend in attitude ratings in which Americans were more (rather than less) accepting of the feminine male than were Koreans, may be
38 explained by the image judgments: B ecause what Koreans would judge to be completely androgynous would be objectively more feminine and less masculine than what Americans would judge to be completely androgynous, it would follow that when asking members of the two cultures to create a mental image of a person exhibiting a set objective amount of femininity or masculinity, the images conjured by Koreans would feature individuals who ar e substantially more effeminate than the ind ividuals imagined by Americans Thus, the two different cultures would ultimately be judging images that would be quite objectively different in levels of femininity and masculinity. Because of this difference, it would not be accurate to report that Americans' higher acceptance than Koreans' of feminine males implies that when viewing the same image of a highly masculine female or highly feminine male, Americans would have more accepting attitudes toward the high levels of masculinity or femininity exhibited by the person than would Koreans. T he cultural differences in subjective perceptio ns of femininity and masculinity as found in image judgments suggests that there is no way for Koreans and Americans to be shown the same image of a person and to interpret the person as expressing the same subjective levels of masculinity and feminini ty; therefore, it is not possible to make claims about Koreans' attitude ratings in the current study as compared to Americans' That is, it would not make sense to ask to what extent the seemingly exaggerated effeminacy of the "flower boy" aesthetic is attributable to Koreans possessing a more accepting attitude than that of Americans toward high levels of femininity in males, since according to these r esults, Koreans' subjective classifications of what is feminine versus what is masculine dictate that the individuals Americans regard as exhibiting high levels of femininity are actually much more masculine in their
39 presentation than Americans would judge them to be for Koreans This finding then nullifies the need to question the role of attitudes toward cross gendered or androgynous presentation as a potential contributor to the apparent cultural discrepancy in beauty ideals, because the neutral androgyn y point differs across the two cultures Health and Testosterone The results of the current study serve as fairly solid support for the assertion that Koreans and Americans differ in perceptions of femininity and masculinity. According to the presented lo gic, this discrepancy is attributable to differences in the prevalence of more feminine versus more masculine traits in either culture that is, Koreans are more perceptually sensitive to masculinity due to a relatively higher expression of and therefore exposure to more typically "feminine" traits, whereas Americans are more perceptually sensitive to femininity due to a relatively higher expression of and therefore exposure to more typically "masculine" traits. As for why the two cultures exhibit these t ende ncies to express more strongly feminine or masculine traits, the current argument points to the relationship between qualities of these traits and tendencies in certain cognitive aspects that is, cultural discrepancies in the femininity and masculini ty of subjectively normative presentation are attributed to significant overlaps of traits more typical of femininity or masculinity with traits belonging to more interdependent or independent self construals, more other serving or self servi ng motives, an d generally more outwardly focused or inwardly focused mindsets, respectively. But beyond the cultural discrepancy in tendencies toward, for example, more interdependent or independent self construals, one might be curious as to what could be the driving f orce or the mechanism responsible for the original cultural
40 discrepancy. Why exactly might the two cultures be different in this way? One could speculate that t he answer to this question might possibly be related to culturally specific levels of health an d testosterone, as well as other factors in attraction. As has been reiterated throughout the current study, cultures most typically subscribing to more interdependent mindsets, as opposed to more independent mindsets, place more value on interpersonal rel ationships and social harmony (Markus & Kitayama, 1991) These particular values would imply that people within interdependent cultures would find the most desirable traits for a potential partner to possess to be those most conducive to social harmony. A ccording to Perrett et al. (1998), when attempting to ascertain qualities possessed by men through facial cues alone, women tend to associate increased levels of masculinity in facial features with decreased levels of warmth, emotionality, honesty, coopera tiveness, and quality as a parent all of which could easily be argued to be conducive to social harmony. On this finding alone, it would seem that all women seeking male partners would find men with the least masculine facial features to be the most attr active. However, common sense would dictate that this is not actually a universal truth, and that there must be other factors involved in women's judgments about the attractiveness of male faces. Furthermore, as Korean ideals of male beauty seem to be rela tively more effeminate than American ideals, and there must be similar kinds of discrepancies across other cultures throughout the world, it is likely that women's judgments of men's attractiveness would vary across cultures as a function of culture specif ic levels of a potentially wide array of variables.
41 One of these variables is health, specifically the overall health of a culture: In addition to socially oriented personality characteristics, women can also associate varying levels of health with varyin g levels of masculinity when judging men's faces. Specifically, women see increased levels of masculinity as an indicator of better health. Therefore, in cultures of particularly good general health (as measured by mortality rates according to the World He alth Organization), where the health of potential male partners is therefore not of very significant concern, women can afford to place much less importance on facial indications of good health (increased masculinity), ultimately finding less masculine fac es (which again are positively associated with socially conducive personality traits) to be more attractive (DeBruine, Jones, Crawford, Welling, & Little, 2010). Though South Korea was not one of the countries sampled in DeBruine et al.'s study, its mortal ity rates are consistent with the observed countries said to have better national health as indicated by their lower mortality rates. Because these countries with comparable levels of health to Korea seem to hold relatively more effeminate than average ide als of male beauty, one could assume that Korea could plausibly fit into this trend, possessing comparably effeminized ideals of male beauty a proposal that is consistent with the argument of the current study. In short, given these patterns of women's c ulturally variant perceptions of the attractiveness of levels of masculinity in men's faces as influenced by national health and by levels of importance placed on socially conducive qualities in potential partners one could speculate that within an interd ependent society, if a high level of national health is maintained, women seeking male partners would continue to hold a preference for relatively less masculine facial features in men. This trait would then be genetically
42 perpetuated, leading to a higher instance of the trait; that is to say, lower levels of masculinity (and thus higher levels of femininity) would become more prevalent in the culture, which, according to the current study's original logic, would lead to a more effeminately shifted perceptu al normalization of characteristics. Again, this trend is consistent with the findings of the current study. As perceived levels of masculinity in physical traits are determined largely by levels of testosterone (DeBruine e t al., 2010), it would follow tha t relatively less masculine men would possess relatively lower levels of testosterone. Given that according to the current argument, Korean men are outwardly more masculine than American men, in future research, it would be useful to empirically assess and compare testosterone levels across the two cultures in order to test this assumption from a different perspective, therefore potentially lending it stronger support. Of course, this study would only address masculinity in terms of men's morphological feat ures (e.g., facial characteristics, hair density) versus other areas of gender expression.
43 References Ardery, J. (2012, January 25). Korea's lovely kkotminam [Web log post]. Retrieved from http://www.humanflowerproject.com/index.php/weblog/comments/koreas_lovely _kkotminam/ Barlett, C. P., Vowels, C. L., & Saucier, D. A. (2008). Meta analyses of the effects of media images on men's body image concerns. Journal Of Social And Clinical Psycholog y, 27(3), 279 310. doi:10.1521/jscp.2008.27.3.279 Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology 42(2), 155 162. doi:10.1037/h0036215 Cheng, S. (2005). Popularising purity: Gender, sexuality and nationalism in HIV/AIDS prevention for South Korean youths Asia Pacific Viewpoint, 46 (1), 7 20. Chang, E. C., Sanna, L. J., & Yang, K. (2003). Optimism, pessimism, affectivity, and psychological adjustment in US and Korea: A test of a mediation model. Personality and Individual Differences,34 (7), 1195 1208. doi:10.1016/S0191 8869(02)00109 5 Christopher, M. S., D'Souza, J. B., Peraza, J., & Dhaliwal, S. (2010). A test of the personality culture clash hypothesis among college s tudents in an individualistic and collectivistic culture. International Journal of Culture and Mental Health 3(2), 107 116. doi:10.1080/17542863.2010.491707 Clancy, S. M., & Dollinger, S. J. (1993). Photographic depictions of the self: Gender and age diff erences in social connectedness. Sex Roles 29(7 8), 477 495. doi:10.1007/BF00289322
44 Costa, P. r., Terracciano, A., & McCrae, R. R. (2001). Gender differences in personality traits across cultures: Robust and surprising findings. Journal of Personality and Social Psychology 81(2), 322 331. doi:10.1037/0022 35188.8.131.522 DeBruine, L. M., Jones, B. C., Crawford, J. R., Welling, L. L. M., & Little, A. C. (2010). The health of a nation predicts their mate preferences: Cross cultural variation in women's preferences for masculinized male faces. Proceedings of the Royal Society of London B, 277 (1692): 2405 2410. doi:10.1098/rspb.2009.2184 Feing old, A. (1994). Gender differences in personality: A meta analysis. Psychological Bulletin, 116 (3), 429 456. Fujita, F., Diener, E., & Sandvik, E. (1991). Gender differences in negative affect and well being: The case for emotional intensity. Journal of P ersonality and Social Psychology 61(3), 427 434. doi:10.1037/0022 35184.108.40.2067 Innocent. (2009). In The American Heritage Dictionary of the English Language Retrieved from http://www.thefreedictionary.com/innocent Josephs, R. A., Markus, H. R., & Tafaro di, R. W. (1992). Gender and self esteem. Journal of Personality and Social Psychology, 63 391 402. Jung, J., & Lee, Y. (2009). Cross cultural examination of women's fashion and beauty magazine advertisements in the United States and South Korea. Clothing & Textiles Research Journal 27 (4), 274 286. doi:10.1177/0887302X08327087 Kashima, Y., Yamaguchi, S., Kim, U., Choi, S., Gelfand, M. J., & Yuki, M. (1995). Culture, gender, and self: A perspective from individualism collectivism research. Journal of Personality and Social Psychology 69(5), 925 937.
45 doi:10.1037/0022 35220.127.116.115 Kim, D., Pan, Y., & Park, H. (1998). High versus low context culture: A comparison of Chinese, Korean, and American cultures. Psychology & Marketing 15(6), 507 521. doi:10. 1002/(SICI)1520 6793(199809)15:6<507::AID MAR2>3.0.CO;2 A Kim, Y., Sohn, D., & Choi, S. M. (2011). Cultural difference in motivations for using social network sites: A comparative study of American and Korean college students. Computers in Human Behavior, 27 (1), 365 372. doi:10.1016/j.chb.2010.08.015 Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98 (2), 224 253. Martina & Simon. (2011, June 30). TLDR shiny suits and fl ower boys [Web log post]. Retrieved from http://www.eatyourkimchi.com/cool korean trends/ Nam, K., Lee, G., & Hwang, J. (2011). Gender stereotypes depicted by Western and Korean advertising models in Korean adolescent girls' magazines [Abstract] Sex Role s 64 (3 4), 223 237. doi:10.1007/s11199 010 9878 z Okazaki, (1997). Sources of ethnic differences between Asian American and White American college students on measures of depression and social anxiety [Abstract]. Journal of Abnormal Psychology, 106 (1), 52 60. Park, Y., Killen, M., Crystal, D. S., & Watanabe, K. (2003). Korean, Japanese, and US students' judgments about peer exclusion: Evidence for diversity. International Journal of Behavioral Development 27(6), 555 565. doi:10.1080/01650250344000217
46 Pe rrett, D. I., Lee, K. J., Penton Voak, I., Rowland, D., Yoshikawa, S., Burt, D. M., Henzik, S. P., Castles, D. L., & Akamatsu, S. (1998). Effects of sexual dimorphism on facial attractiveness. Nature, 394 (6696), 884 887. doi: 10.1038/29772 Playful. (2009). In The American Heritage Dictionary of the English Language Retrieved from http://www.thefreedictionary.com/playful Thomsen, D., Mehlsen, M., Viidik, A., Sommerlund, B., & Zachariae, R. (2005). Age and gender differences in negative affect -Is there a ro le for emotion regulation?. Personality and Individual Differences, 38 (8), 1935 1946. doi:10.1016/j.paid.2004.12.001 Turnbull, J. (2008, November 3). Why size matters: Feminine representations of men in Korean advertising [Web log post]. Retriev ed from htt p://thegrandnarrative.com/2008/11/03/why size matters feminine representations of men in korean advertising/ Turnbull, J. (2010, December 9). The effeminacy of male beauty in Korea [Web log post]. Retrieved from http://thegrandnarrative.com/2010/12/09/kkot minam male beauty korea/ Vesga Lpez, O., Schneier, F. R., Wang, S., Heimberg, R. G., Liu, S., Hasin, D. S., & Blanco, C. (2008). Gender differences in generalized anxiety disorder: Results from the national epidemiologic survey on alcohol and related cond itions (NESARC). Journal of Clinical Psychiatry 69(10), 1606 1616. doi:10.4088/JCP.v69n1011
47 Figure 1 Androgyny rating scale, in which 4 represents the highest level of masculinity, 4 represents the highest level of femininity, and 0 represents complete androgyny.
48 Figure 2 Korean and American Androgyny ratings of images by image subject gender. -1.5 -1 -0.5 0 0.5 Female Images Male Images Androgyny Rating Image Subject Gender Image Ratings Koreans Americans
49 Figure 3 Korean and American androgyny ratings of traits. -3 -2 -1 0 1 2 3 Assertive Confident Playful Shy Cautious Neurotic Androgyny Rating Trait Trait Ratings Koreans Americans
50 Figure 4 Korean and American acceptance levels toward the hypothetical masculine female and feminine male. 0 2 4 6 Masculine Female Feminine Male Acceptance Level Hypothetical Person Acceptance Ratings Koreans Americans
51 Appendix A Korean female subject images Korean male subject images
52 American female subject images American male subject images
53 Appendix B Section 1 Example Please answer the following questions as applied to the female subject pictured below exactly as she appears in this particular image. 5. Please rate the person in the above image based on your perceptions of her femininity in the following aspects: 6. Please rate the person in the above image based on your perceptions of her masculinity in the following aspects:
54 Appendix C Section 2 Example 26. Please rate the following 6 personality traits in terms of how characteristically feminine you perceive them to be. When making your judgments, you should consider the extent to which a stronger presence of each characteristic in an individual would cause that person to be perceived as more feminine. You should also consider how typical you think each characteristic would be of an average highly feminine individual.
55 Appendix D Section 3 Example 29. Imagine that you and your friend are walking a round somewhere near where you live. You see another person walking around in the same area. You look at the stranger's face, hair, clothes, posture, and movement. You decide that the stranger is not particularly attractive or unattractive, but you can't r eally tell whether the person is male or female. Your friend, who has met this person before, tells you that the person is actually FEMALE. Please indicate how much you agree with the following statements about your feelings toward the highly masculine fe male stranger.