Skip to content Skip to navigation

Full citation: 

Penrose, Walter. 2001. “Hidden in History: Female Homoeroticism and Women of a ‘Third Nature’ in the South Asian Past” in Journal of the History of Sexuality 10:3-39.

Contents summary: 

[Note: I have some reservations about this article because it feels very much like a western outsider using primarily western/translated sources to try to say big-picture things about gender and sexuality in South Asia.  There is a fair amount of speculative language (“such women could have...”) and conflation of historic evidence from wildly disparate times and places whose primary common theme is “not part of western Christian culture.” Take it for what it’s worth.]

Penrose looks at a variety of evidence to see if he can reconstruct a picture of variant gender and sexuality roles for women in South Asia (primarily the Indian subcontinent) in the pre-colonial past. In modern India, colonial attitudes (to some extent, reinforced by the interests of Hindu nationalism) have suppressed traditional roles falling outside a gender binary. Traditional roles such as hijras have been re-framed as falling into western concepts of transgender rather than being seen as distinct and independent from male or female.

Modern Indian society has no role parallel to hijras for women (or to be more precise, for persons assigned female at birth “AFAB”) and modern Indian women who identify as lesbians come into conflict with strongly patriarchal cultural imperatives toward marriage. But Penrose traces historic remnants and suggestions of “third gender” and gender-variant roles for Indian women in the past, starting with anthropological data from isolated traditional societies, references in ancient Sanskrit texts, and later historical documents.

Penrose begins by reviewing a theoretical framework for studying third/fourth-gender roles in indigenous North American societies which can be defined by the following characteristics: economic specialization (that is, specific economic roles apart from those associated with men or women), special ritual/religious functions, and gender difference or same-sex relations associated with non-procreation. Within Native American societies there are a variety of gender role systems that may include either a third gender or both a third and fourth gender, depending on whether AMAB and AFAB non-binary people are treated as a single category or two different categories. When considering third/fourth gender categories worldwide, it becomes clear that the development of variant gender categories is not related to the degree of gender difference within the society as both highly patriarchal and more gender-egalitarian societies may have them. Further, non-binary gender roles do not automatically correlate with non-heterosexual activity, as variant genders may include the renunciation of sexual activity.

The first category under consideration for India, and the one with the strongest surviving tradition, is the hijra, most typically involving AMAB (assigned male at birth) people who typically undergo castration, but can also include intersex people or non-menstruating women. Hijras typically dress in female-coded clothing or a mixture of male and female garments and engage in special ceremonial roles associated with births and weddings, as well as in prostitution (with men). Social attitudes towards hijras have changed under colonial influence and hijras themselves may be redefining their understanding of their identity. Historically, the role was an accepted part of society in part because it was viewed as a group identity rather than an individual one, and in part because it was viewed as being part of individual spiritual evolution within the cycle of reincarnation.

When looking for parallel or equivalent female (or AFAB) roles, one encounters the strong pressure on women to marry and the negative social implications of refusing or failing to do so. But there are remnants of established roles for non-married women in some isolated traditional societies, such as the sadhins among the Gaddhi people in the Himalayan foothills. The root of the word means “a holy person” but unlike the masculine sadhu it doesn’t identify someone with religious obligations but rather a “sworn celibate woman.” Such women renounce marriage but continue to live with their families, retain female names, and are referred to with feminine language. They may socialize as women but may also on occasion socialize as men. The renunciation must occur at puberty, in contrast to other female ascetics who may change their status later in life after marriage or widowhood. A sadhu who later changes her mind and has sex with men is stigmatized.

The more common form of Hindu female aseticism involves a renunciation of the material world performed later in life, which does not follow the usual understanding of a separate gender role.

There are records in the 1890s of a category of Hindu women called basivis who would be given male privileges and be allowed to wear male clothing in order to bury their parents and pass on the family name, which by definition also involved being permitted to have sexual relations and bear children. (It isn’t clear whether this would be a life-long role or a temporary one.)

The sadhins don’t meet the three-part framework for a “third gender” established above as they take on male economic roles (rather than a specialized separate economic position), do not perform ritual functions, and are not associated with same-sex activity.

A different context is associated with a group of “transvestite” men and women in southern India associated with devotion to the goddess Yellamma who change gender presentation in terms of clothing and are considered to have changed sex (rather than being a distinct and separate gender category), and are known respectively as jagappa (for those AMAB) and jagamma (for those AFAB). Membership in this group is considered to be divinely mandated and is signaled by a set of conventional physical symptoms. On accepting membership, a jagamma will typically change to wearing male garments, but they do not typically adopt male names, use male pronouns, or take on male professions. Both jagammas and jagappas participate in public ritual functions associated with membership.

Leaving the possible modern examples, Penrose looks for deep historic roots for third-gender functions, starting with physical artifacts from the Indus Valley civilization and then looking for potential connections with Mesopotamian and Sumerian traditions from the Bronze Age. [Note: In my opinion, this is reaching a bit and smacks of “all ancient civilizations can be equated with each other. Penrose suggests similarities between hijras and the galli associated with various mother goddess cults in the Mediterranean region. Like I say, I think he’s stretching here.]

Early legendary material from India includes divinely mediated change of sex, in contexts where the change is motivated by misogynist or heteronormative imperatives. A queen who had given birth to seven daughters and no sons was told her next child must be a son. The resulting daughter, Amba, was raised as a boy. When she married, her true sex was discovered but she was able to become physically male by exchanging sex with a supernatural creature. In various stories, a princess is disguised as a prince and changes sex (by entering a body of water) in order to marry another woman. [Note: if I were reaching widely across cultures and time periods for thematic connections, I might note the motif of the bath in the context of magical sex-change in the story of Yde and Olive.]

Sanskrit religious and medical texts provide ambiguous and confusing hints at possible third-gender roles for women. The confusion is increased when western translations of the works impose ill-fitting categories on them via vocabulary choice. References to a tritiya prakriti (third nature) “one with neither masculine or feminine nature” have been interpreted as applying only to AMAB people who may take on female clothing or may undergo castration, but other interpretations suggest it includes cross-gender roles for both sexes. As the term appears in the context of sexual texts (the Kama Sutra), the term may be concerned specifically with roles associated with sexual activity. Within this context, the female term purushupini for a “third nature” may be connected with terms discussing “virile behavior in women” or women who take a “masculine role” in sex (e.g., penetration of either male and female partners). But much of the ambiguity comes from the focus in the text on male sexual activity and appropriate sexual partners for men, thus the possible implications of female same-sex relations are not explored.

The Kama Sutra does explicitly discuss female same-sex activity as a situational behavior when men are not available (e.g., in sex-segregated environments), but this doesn’t invoke a distinct gender category.

Another social category with sexual implications is sanvahika “women who do arduous work, women who carry burdens” but it isn’t clear that this is a gender category rather than an economic/class one.

The sexual role of svairini (who can take on a penetrative role) is problematic in translated works, which generally try to shoehorn it into meaning “lesbian”. Contextual examples seem to imply a meaning more like “a woman who operates sexually outside the normative female role.” For example the svairini is listed as a type of prostitute (or, at least, a type of woman with whom it’s permitted to engage in certain sex acts). But elsewhere it’s noted “Svairini are independent women who frequent their own kind or others.” (But does the “own kind/others” distinction mean “other svairini / non-svairini women” or “women/men”?) In another passage, “The svairini is one who refuses a husband and has relations in her own home or in other houses,” which could imply simple marriage resistance while rejecting celibacy. And elsewhere the svairini is specifically described as engaging in sex with women, but specifically in a penetrative role. So is the svairini a gender role or a sexual one?

Sanskrit medical literature touches on categories of variant women who may be infertile or who lack sexual desire (treated as functionally equivalent) due to actions by the parents during conception or before birth. But these are not consistent with a concept of a distinct gender role.

Penrose follows this with a long discussion of various gender-linked occupations that contradict traditional gender roles, such as female bodyguards and warriors, or “wandering nuns” who were not bound by traditional restrictions of female movement and association. The various passages on female guards/warriors are fascinating, but fall more in the category of occupation than gender category, despite the masculine coding of the underlying activities. There is, however, a passing reference to an Indian tradition of a (legendary) Amazon society known as Strirajya whose members engaged in same-sex erotics.

Women entering various ascetic religions traditions often left behind feminine-coded behaviors, e.g., by cutting their hair and no longer being considered sexual objects (though not necessarily renouncing sex).

The next section of the article discusses same-sex relations within homosocial environments such as harems and other sex-segregated palace arrangements. These institutions might recognize (and in some cases try to regulate) sex between women and might have institutionalized systems for women to educate each other in sexual techniques (either for their own satisfaction or for the benefit of their husbands). This might include some of the women or their female attendants dressing in male garments as part of sexual relations. (Which is the tenuous connection with the theme of the article, i.e., variant gender roles.) Examples are brought in from similar social arrangements in the Mediterranean Islamic world (of questionable relevance). This section includes a lot of speculative language regarding sex between women in harems that I’m skipping over.

Under British colonial rule, many of the social structures around variant gender roles were deliberately eliminated and traditions disrupted to the point where modern Indians often consider female same-sex relationships to be a Western import. This suppression on western “moral” grounds has, to some extent, been continued in the interests of Hindu nationalist identity, emphasizing hyper-masculine identity for men and an image of self-sacrifice and chastity for women.

But there is enough evidence to support an understanding of a variety of “separate gender” roles for AFAB people in the past, some with religious functions, some with economic basis, and some associated with specific sexual interests. The degree to which these roles were self-chosen may have varied. While the concept of a “third nature” was viewed, in some circumstances, as an inherent trait, there is no clear unifying single model, nor do the Indian examples clearly align with “third gender” roles in other traditional societies worldwide. The Indian subcontinent was settled by successive waves of invasion or migration across a long period and the disparate “third gender” roles may represent the remnants of many different traditions (and tend to be geographically localized). Even before the effects of colonialism, cultural influences that resulted in the decline of Buddhism and those associated with the Muslim invasions altered the nature and understanding of third-gender roles.