Introduction

The present study focuses on the body image of Australian recreational belly dancers. It sought to extend existing research in two main ways. First, it sought to increase knowledge about how different forms of dance are associated with body image. Previous studies conducted in the United States and United Kingdom have shown that street and modern dancers have more positive body image, while exotic dancers have poorer body image, than college student comparison groups. In this, belly dance represents an interesting dance form, as it has been conceptualized as both sexually alluring (like exotic dance) and embodying (like street dance). Second, the study sought to test a recent theoretical model, namely the embodiment model of positive body image (Menzel and Levine 2011), in the context of belly dance. This is an important general model that holds embodying experiences as core to understanding and developing positive body image, but one which has not yet been subject to empirical test. To this end, a cross-sectional design was used to compare body image and related constructs in a group of Australian recreational belly-dancers with a group of Australian college students, and to test the mediational pathways proposed in the embodiment model. Thus the present study adds to the existing scant literatures on belly dancing and on the predictors of positive body image.

Although body image is a complex and multi-faceted construct that encompasses a whole range of a person’s perceptions, thoughts, feelings, and behaviours toward their body (Cash 2004; Grogan 2008), the major focus has come to be on body shape and weight. This is likely due to current societal beauty ideals in many Western countries, including Australia, emphasizing body shape and weight. For women in Australia, like their counterparts in the United States, the societally sanctioned ideal is of a thin and toned body, while for men, the ideal is a muscular v-shaped body (Grogan 2008). For both genders, these ideals are so extreme as to be impossible for most individuals to attain by healthy means (Halliwell and Dittmar 2008; Leit et al. 2002). Hence it is not surprising that many individuals suffer some level of body dissatisfaction. Nevertheless, reviews of studies across a number of Western countries (e.g., Tiggemann 2004) have uniformly found greater body dissatisfaction for women than for men. In addition, Miller and Downey’s (1999) meta-analysis indicated that the effect size of the relationship between body dissatisfaction and self-esteem was larger in samples of women than men, indicating that appearance was more central to women’s self-concept. Indeed, body dissatisfaction is so widespread among women that weight has been conceptualized as a “normative discontent” (Rodin et al. 1985, p. 267) for United States, British, Canadian, and Australian women (Grogan 2008). Furthermore, a large-scale cross-cultural survey of body image ideals across 26 countries found few differences in women’s body dissatisfaction across urban settings in the different countries (Swami et al. 2010). The authors concluded that the thin ideal and body dissatisfaction have become international phenomena, likely due to the increasing globalization of Western media. Importantly, such body dissatisfaction is not benign. It has been shown to have negative consequences in terms of excessive dieting, disordered eating, lower self-esteem, and increased depression (for a review, see Wertheim et al. 2009), conditions which in Western countries occur more frequently among women than men (Fredrickson and Roberts 1997). Stice’s (2002) meta-analytic review particularly identified body dissatisfaction as the single most consistent and robust risk factor for eating pathology.

Nevertheless, within any one particular culture, there will likely be different subcultural norms (Tiggemann 2011a). It has been suggested that in Western countries some forms of dance may represent a subculture in which there exist particularly extreme pressures to be and to remain thin (Le Grange et al. 1994). In particular, ballet dancers have been identified as a specific group likely to suffer from negative body image. A number of studies have documented higher body image concerns and disordered eating among ballet dancers in the United States (e.g., Pierce and Daleng 1998) and Australia (e.g., Abraham 1996). Further, Tiggemann and Slater (2001) found that Australian women who had previously participated in classical ballet had greater self-surveillance and disordered eating than (non-ballet dancing) college students. In their meta-analysis of female sports participation and eating problems across Western countries, Smolak et al. (2000) concluded that, while participation in sports was generally protective of body image, dancers were at higher risk of body image and eating problems than their non-dancer counterparts. This is not surprising given that at least some dancers (e.g., ballet dancers) exist in a demanding environment likely to accentuate appearance concerns, and one in which their bodies are continually displayed and scrutinised.

However, Smolak et al. (2000) noted significant heterogeneity in their results. Although this may reflect sample differences between different studies, another possibility is that there are a number of different styles or genres of dance that differ in their aesthetic requirements, as well as in their subcultural norms. In support, Swami and Tovée (2009) found that British street-dancers (e.g., hip hop, break-dance) scored higher on body appreciation than did community and college women. Similarly, Langdon and Petracca (2010) found United States women engaged in modern dance had lower self-objectification and drive for thinness than non-dancer (college) norms. These two forms of dance demonstrate a more obvious physicality and allow a greater diversity in physique than more traditional forms of dance, such as ballet. On the other hand, Downs et al. (2006) found that United Staes exotic dancers (women who dance in adult sexualized settings, e.g., “gentlemen’s clubs”, for the pleasure or titillation of the viewer), scored higher on body surveillance and body shame and lower on relationship satisfaction than college women.

Langdon and Petracca (2010) likened belly dance to exotic dance and categorized both as appearance-focused forms of dance (in contrast to modern and street dance which were categorized as athletically-focused forms), with the clear implication that belly dance would relate negatively to body image. However, although exotic dance and belly dance share the features of being visually appealing and potentially sexualizing, they seem quite different in other important ways. In Australia, the most common form of belly dance is Middle Eastern, with some Tribal Style elements, and classes are promoted on the basis of fun and fitness (bellydanceoz.com). Accordingly, in contrast to exotic dance, belly dance is accessible to women of all shapes and sizes. If anything, a fatter and curvier body type is valued as more suitable for belly dance’s shimmies, rolls, and undulating movements (Bock and Borland 2011). As yet there exists little quantitative research on the body image of Australian or other belly dancers, but in their preliminary survey of U.S. belly dancers, Downey et al. (2010) reported that the vast majority saw belly dance as a positive influence on their feelings about their body (and none cited it as a negative influence). Consequently, in contrast to Downs et al. (2006) finding for U.S. exotic dancers, we predicted that belly-dancers would show less body dissatisfaction and more positive body image than non-belly dancing college students.

Although the complexity of body image has long been acknowledged (Pruzinsky and Cash 2002), the focus has been very much on the negative aspects of body image, particularly body dissatisfaction. Recently, however, it has been argued that such a negative focus has limited our theoretical understanding of body image, as well as treatment and prevention options (Tylka 2011). Positive body image, on the other hand, refers to love and acceptance of one’s body (including aspects inconsistent with societally-prescribed ideals) and appreciation of its uniqueness and the functions it performs (Tylka 2011). Importantly, positive body image is conceptualized as something more than just the absence of body dissatisfaction. A small but increasing body of research conducted in North America, the United Kingdom, and Australia has shown positive body image to be associated with a range of positive outcomes that go beyond body satisfaction, including intuitive eating (Augustus-Horvath and Tylka 2011; Iannantuono and Tylka 2012), self-esteem (Avalos and Tylka 2006; Swami et al. 2009), self-compassion (Wasylkiw et al. 2012), positive health behaviours (Andrew et al. 2014), and better sexual functioning (Satinsky et al. 2012).

Recently, Menzel and Levine (2011) have proposed a theoretical model of the development of positive body image which draws together the ideas put forward by Tylka (2011) and Piran (2001, 2002). They propose that participation in ‘embodying’ activities is key, because embodying experiences lead to a sense of embodiment and positive body image. Embodiment refers to the sense of ownership of the body and experiencing it as trustworthy and deserving of respect, as well as a key means of expressing competence, interpersonal relatedness, self-expression, and power (Menzel and Levine 2011). According to Piran’s empowerment-relational model (e.g., Piran 2001, 2002), it is this close and connected relationship with her body that allows a girl or woman to act effectively in the world. Embodying activities, then, are those that are situated ‘in’ the body and involve an integration or inter-connectedness of the mind and body, resulting in a sense of flow (Csikszentmihalyi 1990) and of physical (and psychological) empowerment. In this, they represent the very opposite of activities oriented toward how the body appears to others in accordance with cultural standards. This latter observer’s perspective on the physical self, with accompanying habitual surveillance of one’s external appearance, has been termed ‘self-objectification’ (Fredrickson and Roberts 1997). Self-objectification has itself been associated with a host of negative body image and other outcomes (see Tiggemann 2011b, for a recent review). Thus Menzel and Levine (2011) argue that, in addition to leading to the development of positive body image directly, embodying conditions may lead to both decreased exposure to objectifying experiences and a greater ability to cope with body image challenges without internalizing them. Thus they predict an additional indirect pathway from embodying activity to positive body image through a reduction in self-objectification.

Menzel and Levine (2011) present competitive athletics as a theoretical example of a source of embodying experiences and identify frequent states of mind-body integration, increased body awareness and responsiveness, an increased sense of physical empowerment, and an overall sense of physical competence as embodying features. We argue here that these features are largely shared by belly dance. In particular, we propose that belly dance as practised in Australia is an embodying activity, in that it is highly skilled and requires focused attention on breathing and muscles of the mid-torso, as well as strength and flexibility. Indeed, it has been argued that belly dance brings together mind, body, and spirit (Kraus 2009). Belly dance requires dancers to train their bodies to move in new ways, thus requiring concentration and communication with their body. It has been argued that getting to know and training the body helps connect it to the mind (Kraus 2009). In addition, as indicated earlier, belly dance does not subscribe to the one narrow (thin and youthful) ideal, but is accepting of different physiques and the whole range of ages (Bock and Borland 2011). In Australia it is also typically practised in a female social environment, thus enabling a sense of connectedness between women.

This conceptualization of belly dance as an embodying activity is supported by Moe’s (2012) qualitative analysis of journal narratives provided by U.S. belly dancers. A set of broad themes as to why women participate in belly dance were identified, namely, healing and well-being, spirituality, sense of community and sisterhood, and empowerment. In particular, empowerment comes as women experience enhanced strength, confidence, self-esteem, and personal liberation through belly dance. A number of participants explicitly stated that belly dance enabled them to reconnect with their bodies, to appreciate their own unique beauty and sensuality, and to transcend concern about the gaze of others. Moe (2012) concluded that belly dance is “a form of leisure that allows women a means through which to reclaim and reconnect with their physicality in empowering, self-affirming ways by supporting women’s reconciliations with their physical size and composition” (p. 225). Thus belly dance was deemed an apt context in which to test Menzel and Levine’s (2011) model. More specifically, it was predicted that belly dance (as an embodying activity) would be associated with more positive body image and less self-objectification, with the effect of the former being partially mediated by the latter.

Belly dance is also a particularly interesting form of dance because it is seen as sexually alluring in both its moves and costuming. It likely provides both a means of expressing one’s sexuality in a creative way (self as subject), as well as a means of enjoying sexualized attention (self as object). More broadly, there is no doubt that the sexualization of women and girls is pervasive in both United States and Australian culture, perhaps most obvious in media portrayals of women (American Psychological Association 2007). Although feminist theorists (e.g., Gill 2008) have argued that sexualization represents a form of oppression used to control women, more recent theorizing has suggested that some women might enjoy sexualization and engage in self-sexualizing behaviours, e.g., wearing revealing clothing or dancing provocatively, to gain power and agency from the receipt of positive male attention (Liss et al. 2011). The incidence of such self-sexualizing behaviours has been documented among U.S. college women (Nowatzki and Morry 2009; Smolak et al. 2014; Yost and McCarthy 2012), with Smolak et al. (2014) concluding that self-sexualization is a highly gendered activity that is largely confined to women. Feeling empowerment from one’s sexuality has also been framed positively as part of third-wave feminism (Attwood 2007).

However, using their recently developed scale, Liss et al. (2011) found that enjoyment of sexualized or appearance-based attention was still related to negative outcomes such as body shame and disordered eating in a predominantly college sample of United States women. In other words, there was little support for the benefits of enjoying sexualization. This presents somewhat of a conundrum in the case of belly dance. We expect that belly dancers will score higher on enjoyment of sexualization than college students, as belly dance is a performing art that entails the potential for sexualized self-expression. But we have also predicted that belly dancers will have more positive body image (and less body dissatisfaction) than college students. We tentatively predict that the context of belly dance may render these constructs as independent of each other, that is, although enjoyment of sexualization will be high in belly dancers, it will not be negatively correlated with positive body image. In contrast, on the basis of Liss et al.’s (2011) finding, we predicted a negative relationship between enjoyment of sexualisation and positive body image for the college women.

In sum, the present study sought to establish belly dance as an embodying activity and to test Menzel and Levine’s (2011) embodiment model in the context of belly dance. In order to add to the general literature on the body image of dancers, a group of belly dancers were compared with a group of (non-belly dancing) college students, the most common comparison group. Body image has been shown to be an important issue for Australian undergraduate women (Rodgers et al. 2011). Here it was predicted that belly dancers would have more positive body image than non-belly dancing college students. It was also predicted that they would score lower on self-objectification, and that this difference would mediate the difference in positive body image. A secondary aim was to investigate the newly-developed construct of enjoyment of sexualization. As belly dance is sexy and exotic in nature, belly dancers were predicted to have higher enjoyment of sexualization than the college women. In addition, in contrast to college women, it was thought that enjoyment of sexualization would be largely independent of body image for belly dancers. These aims gave rise to the following specific hypotheses:

  1. Hypothesis 1.

    Belly dancers will score higher on positive body image and lower on body dissatisfaction than college women.

  2. Hypothesis 2.

    Belly dancers will score lower on self-objectification than college women.

  3. Hypothesis 3.

    The difference between groups on positive body image will be mediated by the difference in self-objectification.

  4. Hypothesis 4.

    Belly dancers will score higher on enjoyment of sexualization than college women.

  5. Hypothesis 5.

    There will be a negative relationship between enjoyment of sexualization and positive body image, which will be greater for college women than for belly dancers.

Method

Participants

A total of 213 women participated in this study. A sample of 112 belly dancers was recruited from two belly dance schools in metropolitan Adelaide, South Australia. A sample of 101 undergraduate students who had never participated in belly dance was recruited from Flinders University, South Australia. The available demographic characteristics are presented in Table 1. The modal educational level of the belly dancers was undergraduate university. The two groups did not differ significantly in BMI, t(183) = 1.67, p > .05, with the mean BMI being in the “slightly overweight” category for both groups (Garrow and Webster 1985). Nor did the groups differ on ethnicity, X 2(1) = .16, p > .05; participants in both groups overwhelmingly identified as Caucasian/White (>90 %). The belly dancers were, however, significantly older than the college students, t(203) = 8.92, p < .001.

Table 1 Demographic characteristics for belly dancers and college students
Full size table

Measures

The questionnaire, entitled “Activities and Body Image”, contained the same body image-related measures for the belly dance and student groups, namely, positive body image, body dissatisfaction, self-objectification, and enjoyment of sexualization. Demographic information obtained for both groups included age, height, weight, and ethnicity. The belly dancers were also asked their highest level of education.

Questionnaires differed only in information pertinent to belly dance participation. The belly dance group were asked how long they had been dancing and rated some reasons for participating in belly dance on a 5-point Likert scale (1 = not at all important, 5 = extremely important). In order to support the cover story (“physical activity”) and to exclude any college women who participated in belly dance, the student group were asked to indicate whether or not they participated in sports or dance.

Positive Body Image

Positive body image was assessed by the Body Appreciation Scale developed by Avalos et al. (2005). This 13-item scale contains items addressing the appreciation, acceptance, respect, and attention given to one’s body (e.g., “I respect my body”, “Despite my flaws, I accept my body for what it is”). Responses are made on a 5-point scale (1 = never, 5 = always) and summed and averaged to produce a score ranging from 1 to 5. Avalos et al. (2005) reported unidimensionality of the scale, as well as good construct validity, 3-week test-retest reliability (r = .90), and internal reliability (α = .91–.94). In the present sample, internal reliability was similarly high for both the belly dancers (α = .92) and students (α = .91).

Body Dissatisfaction

Body dissatisfaction was measured by the Body Areas Satisfaction Subscale of the Multidimensional Body-Self Relations Questionnaire (Brown et al. 1990). Participants indicate their degree of dissatisfaction or satisfaction with nine body areas (e.g., mid torso, face, weight) using a 5-point Likert scale (1 = very dissatisfied, 5 = very satisfied). Items were reverse-scored and averaged, to produce a total body dissatisfaction score ranging from 1 to 5, with higher scores indicating greater dissatisfaction. Cash (2000) reported good reliability and validity for the Body Areas Satisfaction Scale with a 1-month test-retest reliability of .74 to .82, and internal consistency ranging from .70 to .89. In the present sample, internal reliability was within this range for both the belly dance (α = .81) and student groups (α = .78).

Self-Objectification

Self-objectification was measured by the Surveillance Subscale of the Objectified Body Consciousness Scale (McKinley and Hyde 1996). This 8-item scale addresses habitual body monitoring, the behavioural manifestation of self-objectification, and has become the most commonly used measure of self-objectification (Calogero 2011). Items (e.g., “During the day, I think about how I look many times”) are responded to on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). Scores are averaged to produce a total score ranging from 1 to 7, with higher scores representing greater self-objectification. McKinley and Hyde (1996) reported adequate construct validity and 2-week test-retest reliability (r = .79) for the measure. Internal consistency was also adequate (α = .76–.89). In the present sample, internal reliability fell within this range for both groups (belly dancers α = .85; students α = .87).

Enjoyment of Sexualization

Enjoyment of sexualization was measured by the Enjoyment of Sexualization Scale recently developed by Liss et al. (2011). This 8-item scale measures reported empowerment from feeling sexy or attractive, and enjoyment of sexualised or appearance-based attention from men. Items (e.g., “I feel complimented when men whistle at me”, “I feel empowered when I look beautiful”) are rated on a 6-point Likert scale (1 = disagree strongly, 6 = agree strongly). Scores are averaged to produce a scale ranging from 1 to 6. Liss et al. (2011) reported unidimensionality of the scale, good construct validity (moderate correlations with body surveillance, appearance-based self-worth, and participation in self-sexualizing behaviours), and adequate internal reliability (α = .85). Internal reliability was very similar in the present samples (belly dancers α = .85; students α = .87).

Procedure

Following approval by the Institutional Research Ethics Committee, and with the permission of the principals, belly dancers at two belly dance schools in metropolitan Adelaide were approached after belly dance classes. They were provided with Letters of Introduction and Information Sheets introducing the study (“Physical activity and body image”) and its voluntary and confidential nature. They either completed a hard-copy questionnaire which they returned into a secure box at the belly dance school, or an on-line questionnaire via a secure website. The undergraduate college students were recruited through an online Psychology registration system and were provided with the same information as the belly dancers. Individuals interested in participating followed the provided link to the website and completed the questionnaire on-line. Any student who indicated they had ever participated in belly dance was excluded from this sample. Students received course credit for their participation.

Results

Characteristics of the Sample

The belly dancers had been participating in belly dance for a mean of approaching 7 years (M = 6.9, SD = 4.9). Their most highly rated reason for participation was “It is fun” (M = 4.8, SD = .6), followed by “I enjoy the movements” (M = 4.6, SD = .5). The item “It makes me feel sexy” was moderately endorsed (M = 3.2, SD = 1.3). For the student group, 20.2 % reported currently participating in organized sport, 22.6 % going to the gym, 2.1 % doing ballet, and 7.4 % other forms of dance.

Table 2 provides the mean scores on the measured outcome variables. This sample of Australian college students was similar in positive body image to Avalos et al.’s (2005) validation sample of United States college students (M = 3.44). They were also similar on body dissatisfaction to Cash’s (2000) normative sample of United States women (reverse-scored M = 2.77). However, they scored a little lower on self-objectification than McKinley’s (1998) sample of United States college women (M = 4.82), but were very similar to a previous Australian college sample (M = 4.59; Tiggemann and Williams 2012). Finally, they scored considerably lower on enjoyment of sexualization than Liss et al.’s (2011) sample of United States college women (M = 3.82). An initial MANOVA indicated an overall significant difference between the two groups, R 2 = .073, F(4,208) = 5.31, p < .001.

Table 2 Mean scores (SD) for body image measures
Full size table

Group Differences in Body Image

Because of the significant age difference between groups, age was taken into account in the analyses. As the age distribution was somewhat positively skewed in both groups, a log transformation was applied (Tabachnick and Fidell 1996). Hence a series of ANCOVAs was conducted controlling for log(age). As can be seen from Table 2, the belly dancers scored significantly higher than the college students on positive body image, F(1,202) = 12.14, p < .001, ηp 2 = .06, and significantly lower on body dissatisfaction, F(1,202) = 13.33, p < .001, ηp 2 = .06, in support of Hypothesis 1. They also scored significantly lower than the college students on self-objectification, F(1,202) = 4.15, p < .05, ηp 2 = .02, thereby supporting Hypothesis 2.

As most of the previous research on body image has been conducted with college women, participants were divide into two groups on the basis of age: 18–30 years (typical college student age), and >30 years. A series of 2 × 2 ANOVAs indicated that there were no significant interactions between age and dance group on positive body image, F(1,201) = 1.08, p > .05, ηp 2 = .01, body dissatisfaction, F(1,201) = .49, p > .05, ηp 2 = .00, or self-objectification, F(1,201) = 1.70, p > .05, ηp 2 = .01. Table 2 also displays the means for the subsample of younger (college-age) women. It can be seen that the subsample means for both belly dancers and college students were very similar to the means for the whole sample.

Mediation of Positive Body Image by Self-Objectification

A series of regression analyses was conducted to examine whether the relationship between belly dance participation and positive body image was mediated by a reduction in self-objectification (Hypothesis 3). Mediation is established when the effect of the predictor variable (here dance group: student = 0, belly dance = 1) on the outcome variable (here positive body image) is less when the mediator variable (here self-objectification) is entered in the regression equation than when the predictor variable is entered on its own (Baron & Kenny, 1986). The regression analysis to predict positive body image indicated that the initial relationship with dance group (β = .26, p < .001) was substantially reduced (β = .11, p = .06) when self-objectification was included in the equation. Figure 1 provides a diagrammatic representation of this result.

Fig. 1
figure 1

Mediation pathways from belly dance group to positive body image. The number in parentheses is the regression coefficient for the direct relationship between belly dance and positive body image. * p < .001

Full size image

The bootstrapping protocol of Preacher and Hayes (2008) was then used to estimate the indirect effect of group on positive body image through the proposed mediator of self-objectification. In this approach, mediation is significant if the 95 % bias-corrected confidence interval (CI) of the indirect path does not contain zero. Results, based on 10,000 bootstrapped samples, indicated that there was a significant indirect effect of group on positive body image through self-objectification, b = .216, CI [.110, .340]. Furthermore, this represents a moderate-to-large effect size, κ2 = .163, CI [.084, .244], accounting for approximately 59 % of the total effect of group on positive body image. Thus Hypothesis 3 was supported.

The Role of Enjoyment of Sexualization

Table 2 also contains the means for each group for enjoyment of sexualization. As can be seen, in contrast to prediction, the belly dance and college student groups had almost identical levels of enjoyment of sexualization, F(1,202) = 1.25, p > .05, ηp 2 = .01. This was also the case for the subsample of younger (18–30 years) participants, F(1,114) = .03, p > .05, ηp 2 = .00. Thus Hypothesis 4 was not supported.

Table 3 displays the correlation coefficients between variables for the belly dance and student groups separately. It can be seen that the pattern of correlations was very similar for both groups. As expected, positive body image had a strong negative correlation with body dissatisfaction and self-objectification. Enjoyment of sexualization, on the other hand, was significantly associated (positively) only with self-objectification. It was not correlated with either positive body image or body dissatisfaction. In addition, in contrast to prediction, the correlations between positive body image and enjoyment of sexualisation were almost identical (and negligible) for the belly dance (r = −.04, p > .05) and student groups (r = −.05, p > .05), Fisher z = .14, p > .05. Thus Hypothesis 5 was not supported.

Table 3 Correlations between body image measures
Full size table

Discussion

The primary aim of the present study was to test the embodiment model of positive body image (Menzel and Levine 2011) within the context of belly dancing. The major findings are clear. As predicted, participation in belly dance was associated with positive body image. Further, as predicted by the embodiment model, this association was mediated by reduced self-objectification. This, to our knowledge, offers the first explicit empirical test of the predictions of the model. It is also consistent with our conceptualization of belly dance as an embodying activity. Thus the findings extend existing knowledge in two distinct ways. They offer confirmation of the predictions of embodiment theory, and they contribute to the scant literature on belly dance.

Although Menzel and Levine (2011) developed their embodiment theory from a conceptual analysis of competitive athletics, the present results empirically confirm that their theorizing extends to other contexts. To the extent that the theory provides an apt framework for belly dance, as shown, it is likely to do so for a range of other ‘embodying’ activities. Here it was found that the effects of belly dance on positive body image were mediated by reduced self-objectification. This suggests that one mechanism for the observed positive effect lies in the reduction of the relative focus on external appearance and commensurate increase in the relative focus on internal experiences and competencies of the body. Future research will need to test embodiment theory in a number of different contexts. It may well be that other more ‘mindful’ activities such as yoga or meditation, which have been shown to be associated with reduced levels of self-objectification, body dissatisfaction, and disordered eating in United States (Daubenmier 2005; Neumark-Sztainer 2014) and Australian samples (Prichard and Tiggemann 2008), may have a more direct effect on positive body image than that found here due to their greater explicit focus on internal experience.

As a set, the results support a conceptualization of belly dance, at least as practised in Australia, as an embodying activity. Belly dancers scored higher on positive body image and lower on body dissatisfaction and self-objectification than college women. In addition, the most highly endorsed motivation for belly dance participation, “it is fun”, reflects a mental and physical presence ‘in the moment’, characteristic of embodiment (Levine and Piran 2004; Piran 2001). The findings confirm in an Australian sample U.S. belly dancers’ beliefs that belly dance has a positive influence on how they feel about their bodies (Downey et al. 2010) and are consistent with Moe’s (2012) qualitative accounts of belly dance enabling women to connect or reconnect with their bodies in new and empowering ways.

The obtained positive finding for belly dance is similar to those found previously for street dance (Swami and Tovée 2009) and modern dance (Langdon and Petracca 2010). It is in marked contrast to the one previous finding for exotic dance (Downs et al. 2006), whereby United States exotic dancers had poorer body image than college students. Thus the categorization of belly dance as akin to exotic dance in this way (Langdon and Petracca 2010) does not hold. Perhaps a crucial difference between belly dance and exotic dance lies in the fact that exotic dancers perform for others in a professional or semi-professional capacity, whereas belly dance in Western countries like Australia is largely a recreational activity (Moe 2012). Thus it may be less the physical activity itself that is most important, but rather the purpose for which it is undertaken, including especially the nature of any real or implied audience. It is this observer’s perspective that is critical for the experience of self-objectification (Fredrickson and Roberts 1997). In this light it would be most informative to investigate the body image of recreational “exotic” dancers, for example, participants in the increasing number of classes in Australia offering pole dancing (Donaghue et al. 2011).

One curious and unexpected finding was that, despite belly dance being seen as an erotic and sexually alluring activity, belly dancers in fact scored no differently in their enjoyment of sexualization than did college women. This supports the earlier reasoning that belly dance is performed for the self because it makes one feel good, rather than for the actual or implied gaze of others. It may be that in belly dance sexiness is achieved through being ‘in’ the dance (and is therefore potentially achievable by all), rather than by looking or acting in a particular way, specifically conforming to an external societally-prescribed physical ideal. As suggested by Moe (2012), belly dance may allow women a (rare) safe and creative opportunity for exploring and expressing their sensual and sexual selves.

Although we predicted that relationships between body image and enjoyment of sexualization would be weaker among belly dancers than college students, this was clearly not the case. In fact, enjoyment of sexualization was not related to positive body image or body dissatisfaction in either group. This is in contrast to the finding of Liss et al. (2011) in United States female college students, but here suggests that individuals can potentially be high (or low) on both positive (or negative) body image and enjoyment of sexualization. The overall pattern of inter-correlations between variables was also somewhat curious. While enjoyment of sexualization and body image were both related to self-objectification in the predicted directions, enjoyment of sexualization was not itself related to either positive body image or body dissatisfaction at all (correlations close to zero). This perhaps suggests that there is some other (positive) component to enjoyment of sexualization that operates differently from the (negative) self-objectification component. As concluded by Yost and McCarthy (2012) in their study of United States heterosexual women kissing other women at parties, it is also possible that the same behaviour can be simultaneously both objectifying and empowering. Given that the enjoyment of sexualization concept (and associated measure) is relatively new, these results await replication. It seems likely, however, that enjoyment of sexualization will prove a more complex and dynamic construct than thought. Future research will need to disentangle this.

The set of findings has some practical implications. Importantly, they show that engaging in embodying activity (here belly dancing) is associated with positive body image, which has itself been associated with positive physical and psychological benefits in a variety of samples (Andrew et al. 2014; Avalos and Tylka 2006; Satinsky et al. 2012; Swami et al. 2009; Wasylkiw et al. 2012). Thus women and girls should be encouraged to participate in embodying physical activities, such as (non-lean) sports, athletics, hiking, or circus skills, right throughout the lifespan. Such activities will be embodying to the extent that they focus on being ‘in’, and using, the body. In addition, these activities have previously been identified as ones likely to reduce self-objectification (Tiggemann 2013) and thus, according to the logic of the embodiment model, should result in positive body image. Importantly, this recommendation represents a positive step, in contrast to most strategies used to combat negative body image, such as not reading fashion magazines or not engaging in social comparison (e.g., Tiggemann and Polivy 2010). One interesting question is whether or not there is some critical period in the lifespan that embodying activities need to be engaged in for maximum benefit. For example, such activity might be particularly crucial in adolescence, a time when many Australian girls currently drop out of organised sport, often for appearance-based reasons (Slater and Tiggemann 2010, 2011). So future research will need to test the application of the embodiment model to younger women and girls than those sampled here. A related question is what happens when women cease to engage in belly dance or another similar activity. As other Australian research indicates that former ballet dancers maintain a self-objectified view of themselves (Tiggemann and Slater 2001), it is theoretically possible that belly dancing might confer a permanently more positive body image.

As with all research, the present findings need to be considered in light of a number of limitations. First, while college students are the most common group used for comparison in this and previous studies, they likely differ from belly dancers in a number of ways besides belly dance participation. While age, BMI, and ethnicity were assessed, there are a host of other potential variables that were not measured, e.g., sexuality. More generally, research has yet to investigate positive body image among lesbian women. Of particular interest here, both groups are likely to engage in a range of other physical activities which are potentially embodying or disembodying. Future research could usefully attempt a much more detailed assessment of the activities in which participants engage. Second, the results are limited to the two particular belly dance schools in Adelaide, South Australia, and may not generalize to other belly dancers in other geographical locations. Nor do we have information on potential differences between the schools. Finally, the present study is correlational and cross-sectional in design. Thus it may be that participation in embodying activity leads (causally) to positive body image, as proposed in the embodiment model. But it is equally plausible that women with certain characteristics, in particular those who already have positive body image, self-select to engage in an activity like belly dance. Most likely the relationship is reciprocal with both processes occurring in a continuing cycle. Future experimental designs that randomly assign participants to a belly dance intervention, as well as prospective designs that follow up women from when they start belly dance over some time, are needed to come to any more definitive causal conclusion.

Despite the unanswered questions, the present study has demonstrated that belly dance is an activity associated with positive body image. It has also tested a critical prediction of the embodiment model of positive body image in the context of belly dance. In so doing, it has not only contributed to knowledge about the body image of one specific group (belly dancers), but also to a greater understanding of positive body image more broadly.