Cinema and lesbian visibility

Specifically lesbian significances to the crossdressed figure (rather than the significance of crossdressing for gay men or for heterosexual women) has been elided in the field of queer discourse as well as in heterosexual feminism. Queer discourse shows a tendency towards a homogenising analysis of the crossdressed figure which does not take account of gender as a field of discursive power in which 'male' is the dominant, or substantive, category. Instead, an undifferentiated construct of 'normalisation' is treated as the dominant discursive formation. 'Heterosexuality' and 'gender' are treated more or less as interchangeable terms although the discourses of sex-gender very obviously continue to differentiate a dominant (male) category and a subordinate (female) category in the production and normalisation of cultural dominance. At the same time, feminist film studies has tended to treat both 'male' and 'female' as categorisations which exclusively constitute (white, middle-class) heterosexual relations. Either way, the specificities of lesbianism are ignored. Before analysing specific film texts, I will therefore attempt to outline the contestations surrounding crossdressing in specifically lesbian literatures, cultures, and theorisations — both within and independently of the contexts of cinema or film studies.

Lesbian feminism has tended to figure crossdressing as an identificatory practice, the significance of which remains fixed in a static (a-historically figured) structure of power relations. I will argue, on the contrary, that signifiers such as 'crossdressing' do not refer themselves to static forms or constructs but reference relatively fluid and mobile discursive strategies. The shifting significances of such tropes register transformations and proliferations in lesbian discourses which are produced in diverse struggles to challenge or maintain specific modes of dominance. In this way, I hope to delineate (very broadly) a discursive context for lesbian practices of spectatorship between the emergence of the discourses of first- and second-wave feminism.

There is a very limited amount of material specifically concerned with female-to-male crossdressing in film on which to draw. Those texts which do address crossdressing in cinema focus mainly on male-to-female transvestism as 'camp' — a theatrical mode of performativity. Where female-to-male crossdressing is theorised, it is treated as a similar phenomenon which operates performatively through denaturalisation effect (Kuhn, 1985; Straayer, 1992; Garber, 1992). On the other hand, commentaries such as Bell-Metereau's (1993), which do treat crossdressing as gender-specific, tend to leave lesbianism in the background. For example, whilst Bell-Metereau does briefly refer to lesbian undertones in a scene from Sylvia Scarlett (1935) in which the crossdressed Sylvia is kissed on the mouth by another woman, overall she treats the significance of the film's crossdressed protagonism in these far more general feminist terms:

Sylvia's character offers a delicate balance of ethical principles and a spirit of fun, a combination all too rare in film heroines of later decades. (Bell-Metereau, 1993: 113)

In order to avoid the overdetermination of my own account of the crossdressed cinematic figure by models of 'camp' or 'performativity,' I shall not focus on the much-discussed lesbian-appeal of theatrical crossdressing in films such as Sylvia Scarlett, or Morocco. The limited anecdotal material which is available tends to suggest that this was not the only mode — and probably not the defining mode — of lesbian appropriation of crossdressed imagery. I shall therefore begin by addressing specifically lesbian theorisations of crossdressing, 'mannishness' and 'butch' identifications, and relate these to the genesis of the crossdressed tomboy figure in mainstream popular literature. I will then look briefly at lesbian discussion of the emergence of 1940s and 1950s lesbian subcultures and of the relationship of these subcultures to popular literature and cinema.

Crossdressing in a feminist frame

The modern association of female-to-male crossdressing with lesbianism is sufficiently ubiquitous in western culture to be received as a necessary one. The connection is, however, of relatively recent date. The significance of the 'maiden' in male clothing across a range of cultures and time-frames is far too extensive even to summarise here. However, prior to the 19th Century in Europe, female-to-male crossdressing seems rather to have been associated with sexual laxity and female rebellion than with lesbianism or, indeed, with any particular psychological state of being. Of course, prior to the 'medicalisation' process, women loved women and engaged in (what would now be defined as) lesbian acts and crossdressing was considered sometimes to accompany inter-female acts of 'sodomy' (a catch-all term for illicit forms of penetrative intercourse) but not to be in any way coextensive, or even necessarily always connected, with such acts (Faderman, 1981; Rose, 1984; Smith-Rosenberg, 1885; Friedli, 1987; Trumbach, 1991: 121-123; Howard, 1993).

An association between crossdressing and lesbianism began generally to be made at the end of the 19th Century (Foucault, 1976; Faderman, 1981 and 1991; Rubin, 1984; Jeffreys, 1985: 102-126; Weeks, 1985). Foucault (1976) argued that there had been a shift away from a conceptualisation of (hetero)sexuality which had seen illicit sexual acts as proceeding from the soul's imponderable tendency to evil and towards a scientific discourse "which spoke, not of sin and salvation, but of bodies and life processes" (Foucault, 1976: 64). Darwinian biologism was a profound influence on late-19th-Century and 20th-Century scientific thinking about sexuality (Weeks, 1985: 67). The publication of Krafft-Ebing's (1886) sexological study had paved the way for an explosion of sexological literature (Weeks, 1985: 67). Science and medicine began to define a "homosexual personality." Ellis' influential study (1901) delineated 'inversion' as a congenital or organic condition — although not always "physically visible" in lesbian women (Ellis, 1901: 255-258). In Freud's work, 'inversion' figures as inversion of (sex) object choice (Freud, 1905: 46-52). In Ellis, it figures rather as an inversion of gender-identity. Most usually these modes of 'inversion' were elided together (as they were in European films of the 1920s and 1930s — see below). Through the processes of these biologistic and medicalising discourses, the lesbian body (rather than the soul) came to represent the site of lesbian 'difference.'

This naturalising representational order was frequently resented by lesbians (and gays); not least because of its extensive deployment to underpin oppressive practices in the institutions of medicine, law, and family. 'Third sex' and 'homophile' organisations in Europe from the 1920s to the 1950s attempted to reverse the pathologising tendency of the 'inversion' model by reversing the implications of its biologism. That is, they posited homosexuality as a blameless and harmless product of nature which was needlessly persecuted by the unenlightened (Dyer, 1990: 74-75; Wolff, 1986; Weeks, 1985: 118-119). The extent to which the dominant construct of the 'mannish' lesbian actually reflected, conditioned, or produced, lesbians' own practices has been a matter of heated debate in feminism (Shockley, 1979; Faderman, 1981; 314-371; 1991: 37-187 and 159-187; Newton, 1984; Davis and Kennedy, 1986; Nestle, 1987; Whitlock, 1987; Jeffreys, 1987; Bogus, 1992; Weiss, 1992: 30-50).

Faderman's (1981) groundbreaking analysis of representations of lesbianism in literature represented lesbianism, instead, as a cultural construct formed through politicised conflict and thus as subject to historical transformation. Faderman's study associated constructs of lesbianism with broader discursive struggles over the boundaries of female identity and activity. Faderman (1981) and van Casselaer (1986) drew out connections between the success of first-wave feminism -- in expanding the sphere of women's activities into "male" professional and political fields -- and the codification in medical discourses of lesbianism as a form of "phallic" competition with men. The adoption of "masculinising" clothing (that is, clothing suitable for physical activity) by women thus cited not an organically "perverse" organisation of desires; but a political struggle between feminism and masculinist reaction.

Faderman's (1981) argument was that, in contrast with the turbulent early-modern period, the sexual passivity attributed to femininity had become so naturalised by the 18th Century that intimate relationships between women were seen as entirely unthreatening. The subsequent shift towards an increasingly virulent pathologisation of lesbianism in the early 20th Century was a direct response to Suffragism:

[W]hen the Conciliation Bill, which would have given a million women the vote, was vetoed in November 1910, the suffragists decided to start throwing bricks and make complacency impossible. Antifeminists must have considered that whatever would scare women away from the movement was fair play. And to associate feminism with what the experts had called morbid would surely scare women away. (Faderman, 1985: 337)

A connection between crossdressing and lesbian sexuality was thus made directly by sociologists and sexologists themselves (van Casselaer 1986: 10), with the motive of halting the entry of women to the public sphere. Sexologists argued that "masculine" activities required "transvestite" (utilitarian) clothing, and encouraged women to emulate a male virility which they could not achieve. Thus, feminist women merely lost their femininity and became "de-sexualised" (van Casselaer: 1986: 10-16). Lesbians such as Radclyffe Hall, who adopted and publicised a crossdressed, mannish, persona, were thus often interpreted by 1980s radical lesbian feminism as colluding with anti-feminist attempts to frighten and pressure women out of 'masculine' professions and back into 'feminine' passivity (Faderman, 1981: 316-326) or with a will to appropriate male dominance to themselves. Faderman (1981) argued that many contemporaneous lesbian figures had rejected this pathologising model outright, including Natalie Barney, a friend of Radclyffe Hall's, who saw lesbianism as "an expression of her feminism" (Faderman, 1981: 372). Faderman later, however, described Barney as "reflecting the sexologists' influence on her conception of her own homosexuality" (Faderman, 1991: 58).

Faderman (1991) also noted that mannishness was often associated by sexologists most particularly with working-class women. This association was originally formed partly as a result of the class prejudices of medical observers who already considered working class women to be less feminine or "refined." Sexologists were also reluctant to "see a deviant sexuality" in women of their own class (Faderman, 1991: 41). It was also engendered by the selection of environments in which women were available for intrusive observation by the sexologists.

The females that the earliest sexologists ... defined as sexual inverts were often a captive population in prisons and insane asylums, daughters of the poor. (Faderman, 1991: 41)

In any case, the passivity enforced in middle-class women would have made survival far more difficult in the relatively challenging environments in which poorer women were obliged to survive (Faderman, 1991: 41). If feminine passivity was actually produced by restrictive clothing and restraints on gesture, social behaviour and sexuality; then refusal of these restraints could be seen, instead, as facilitating the attainment of economic and social autonomy for women. A second-generation of "New Women" wanted not only economic opportunity, but also freedom from the restrictions of femininity (Newton, 1984: 564). Public acknowledgement of 'inversion' could facilitate resistance to marriage, and to the economic dependence and passivity associated with 'femininity' (Faderman, 1991: 58).

Jeffreys, on the other hand, saw any advantage acquired by adopting virilising codes as an appropriation of male power at the expense of more feminised women:

An advantage must have been the superior status. Amidst all the hatred directed at lesbians by the straight world, butch lesbians were at least able to feel superior to someone, the femme. (Jeffreys, 1985: 74)

Both second-wave feminism and gay liberation tended to perceive crossdressing and role-playing as forms of internalised self-hatred left over from the bad-old-days before Women's Liberation and Stonewall (Jeffreys, 1985: 80). For lesbian feminists, crossdressing also articulated as an alignment with gay men:

Lesbian-feminists protested that the butch image was created by males so that "the female homosexual was groomed to appear as a burlesque of licentious, slightly cretinous, ersatz men" and that some lesbians had accepted that image because they had been saturated with it and believed it was the only way to feel authentic ... Lesbian-feminists were sometimes revolted at signs of what they regarded as excessive sexuality among a few lesbians and took a moralistic ... stance ... They would tolerate nothing that resembled the raw sexuality of male eroticism. (Faderman, 1991: 232, quoting Susan Helenius, "Returning the Dykes to the Dutch" in Everywoman, 2:10, 1971)

According to most lesbian accounts, however, lesbian crossdressing practices have always differed from the "camp" presentation associated with male homosexuality (Faderman, 1991; Davis and Kennedy, 1986; Nestle, 1987; Penelope, 1986). Even within the cross-fertilising queer arena, lesbian crossdressing appeared to retain laws of its own (Davy, 1994: 130-149). These specificities may be seen as produced by the inequalities subsisting between men and women. Furthermore, there are diachronic contextual specificities. The mapping of feminist critique over the bar cultures of the 1950s can be perceived as de-historicising. Penelope, for example, described her own transition from a butch identity in the context of the embattled bar-cultures of the 1950s to a lesbian-feminist consciousness -- the possibility for which had not opened out until the 1970s (1986: 14-17). Jeffreys also acknowledged that "a (feminist) critique was not available in the bar scene of the 1950s" (1987: 90). Furthermore, Faderman (1991) noted synchronic diversities in specificities of race and class identifications.

The 'butch' stereotype is also racially inflected for British women of African-Caribbean descent. Black women are already constructed in white culture as less feminine:

The belief that Black men are better in bed is mirrored in the lesbian community ... The idea that Black women are dominating, insatiable animals and undesirable can stay with us throughout adulthood. We are often pushed into these roles by both the white heterosexual and lesbian and gay communities ... (Mason-John, 1995: 80)

Asian women, on the contrary, have conventionally been constructed in orientalist representation as the hyper-feminised objects of white male desires (Karimjee, 1991). However, Asian lesbianism has also been negatively associated with mannishness in orientalist literatures:

The Moslem Harem is a great school for this "Lesbian" (which I call Atossan) love; these tribades are mostly known by peculiarities of form and features, hairy cheeks and upper lips, gruff voices, hircine odour and the large projecting clitoris with erectile powers. (Karimjee, 1991: 32; quoting Burton's Thousand Nights cited by Rana Kabbani in Europe's Myths of the Orient, 1988: 53)

Whilst many rejected the stereotype, some black lesbians recorded specifically black lesbian butch identities or crossdressing practices. Many of the women of the Harlem Renaissance who were recorded as having bisexual lifestyles crossdressed. They were referred to as 'bulldaggers — a term which Bogus linked specifically to popular cinema (1992: 34). Moraga, a Latina lesbian, described her own specific experience of roleplay and its relation to feminism:

... I feel a lot of pain around the fact that it has been difficult for me to conceive of myself as thoroughly female in that sexual way. So retaining my "butchness" is not exactly my desired goal. (Moraga, interviewed in Hollibaugh and Moraga, 1983: 400)

On the other hand, in the video Framing Lesbian Fashion (1992), a woman of Asian descent who posed for Gill Posener's image in the notorious On Our Backs calendar (1989) (which does not credit her name) stated that she specifically wanted to phallicise her self-image to counter the racist hyper-feminisation of Asian women.

Although there is some commentary on the contradictory and ambivalent significances of "mannish" coding to lesbians of colour, the majority of black lesbians in the 1980s situated themselves in a more globalised network of political concerns and expressed little interest in what many black lesbian theorists saw as essentially a white concern:

I cannot align myself with anything that is not about the liberation of Black and Third World women and this does not speak to my needs ... I do think it's a white woman's issue. (Sims, 1982: 99)

By the mid-1980s, opposition to radical-lesbian-feminist constructs had emerged within white lesbian discourse. Whilst the feminist project had been extremely successful in alleviating the stigmatisation of lesbianism, this was thought by some groups of lesbians to have been achieved at the cost of largely eliminating desire or pleasure, as well as history, from the lesbian agenda (Moraga and Hollibaugh, 1983; Hollibaugh, 1984; Nestle, 1987). Papers given at the Barnard Conference (The Scholar and Feminist IX Conference, "Towards a Politics of Sexuality," in 1982) coalesced discussion of lesbian sexuality — including practices of role-playing — re-signified as a liberatory sexual freedom posited against a perceived "anti-sex" and victimising bias in radical feminism:

... [L]ooking at the danger and damage done us is only a part of coming to terms with sex. We should also begin to look at sexuality itself and at what we mean by words like desire, passion, craving, need. (Hollibaugh, 1984: 402)

Radical feminist lesbians, however, saw such arguments as collaborating with an anti-feminist backlash:

The context in which lesbians ... are promoting role-playing is that of a massive and coordinated onslaught upon feminism. (Jeffreys, 1987: 90)

The decade nevertheless saw some rehabilitation of the archetypal figure of the mannish lesbian and the butch bar-dyke. Lesbian-feminists contributors to the US lesbian-feminist academic journal Lesbian Ethics, in dialogue with British lesbian-feminists such as Jeffreys, radically re-signified the butch figure from within the lesbian-feminist field of discourse. Strega (1985) criticised a "superficial romanticising of butches [sic] 'erotic power'" (78) in libertarian lesbian discourse (although she did not mention Nestle (1987) specifically, this would appear to have been her primary reference). Retaining a radical feminist position (see, for example, Daly's exposition, 1978: 1-9) that "femininity" can be separated from "femaleness" (the former being a product of patriarchal culture and the latter of "nature"), Strega argued that heterosexual feminism confused the terms, colluding with heterosexist models of "femininity:"

Feminism has remained extremely heterosexist and Lesbophobic, and so it supports the male ideals of feminine values. That is why most ex-het feminist Lesbians think that butches are in a role, but that femmes are not. Like with other privileges, femme-ness is considered the norm. And of course it's those with the privilege who have the power to define what the norm is. (Strega, 1985: 78)

Butch-ness was thus re-figured as resistant to the feminine role rather than appropriative of the male role. The butch thus represented the 'true' (ontological) female:

If something in her [the butch] cannot and will not accept the artificiality of femininity, the trappings and mannerisms of the feminine role, everyone around her will begin telling her she's not really a girl ... When she continues to resist, and as her Lesbian identity becomes obvious, she is labelled a deviant, a freak of nature, a man-in-a-woman's body . . . The pressure is meant to humiliate and bully her into accepting femininity. (Strega, 1985: 75)

Ruston et al extended this reversal to a discursive de-naturalisation of heterosexual gender:

It is men and het women who truly play roles. Their roles are so much a part of dominant male culture that they are taken for granted and considered to be natural. (Ruston et al, 1986: 27 -- emphasis theirs)

Penelope's re-conceptualisation, which seems to owe much to Wittig (1981), removed the butch entirely from "the masculine" -- which for Strega, Ruston, et al, had figured as a pre-cultural space:

English, like any language, describes the version of reality preferred by the majority of speakers in a society, and forces us to accept the reality agreed upon by those speakers. The United States is a heteropatriarchal society, a culture that assumes that heterosexuality is "natural", that male dominance is "natural", and that female subordination is "natural", and English provides its speakers with ways of expressing these assumptions as though they were incontrovertible facts. The words masculine and feminine exist only because they express concepts essential to the maintenance of H[etero]P[atriarchal] reality. But the existence and continual use of these words doesn't mean that they denote "real" or actual things. (Penelope, 1986: 60 &mdahs; emphasis hers)

Use of the term 'real' man, or woman, thus effectively posits an 'unreal' man or woman — and it is in this 'unreality' that the figure of the butch locates. Queer space opens out as this field of linguistic exclusion translates into social exclusion. The use of terms such as 'mannish' lesbian or 'effeminate' homosexual operate to reconfirm the linguistic dichotomy which produces men, women, and queers in the first place. Penelope also noted that the gradation of lesbian behaviours in the 1950s bar-scene, of which 'ki-ki' was the median term, disappeared in the process of the feminist transformation of lesbian culture; reflecting, perhaps, the paradoxical extremity of feminist re-polarisation of gendered identity.

Whilst Penelope's insistence on the productive effect of patriarchal language constitutes a discursive, or anti-essentialist, position, she nevertheless retained a sub-textual extra-heterosexual 'reality:

If we're real, then their conceptual framework is flawed ... I suggest we toss out HP semantics along with MASCULINE and the Lesbian labels derived from HP, BUTCH/FEMME ... Most importantly, we must stand our ground outside of HP reality, occupy it with determination, and resist efforts to assimilate us and dilute the radical force of our perceptions. (Penelope, 1986: 79 — emphasis hers)

Penelope was, probably quite deliberately, unclear as to the status of this 'ground' outside of HP reality. It first appears as a reverse-discursive space opened out as an articulation of the exclusions effected by heteropatriarchal language. However, if lesbianism occupies a reverse-discursive space, we can only work-over and not "toss out" patriarchal definitions. On the other hand, Penelope's account of language as culturally productive and her reverse-discursive formulation of lesbian occupation of the "irreal" posited by the terms 'real man' and 'real woman' seems to deny any substantive position. It becomes, effectively, a lesbian-feminist double-reversal which re-frames the dominant as 'unreal' and the lesbian as "real."

Jeffreys (1994) acknowledged that this kind of approach politicised the issue of roleplaying, but maintained that such arguments did not escape from essentialism.

They call upon all lesbians to simply relinquish the privileges of passing and give up femininity so that "butches" would no longer suffer for their visibility. This is a more dynamic lesbian positive solution [than that of Nestle, 1987] ... But their use of role-playing vocabulary in situations in which it hardly seems appropriate does undermine the important political points they are making ... It rigidifies butch-femme categories and does not allow for change. (Jeffreys, 1994: 92)

Jeffreys does seem, however, basically to agree with Strega, Ruston, and Penelope's constructions of lesbians' relationship to the patriarchal category of femininity. Jeffreys further contended, and I would strongly agree with her, that the rejection of femininity has been a more communal lesbian construct, and has not been confined to those adopting a butch style

There has been a historical tradition of lesbians rejecting femininity in different ways and to different extents but the rejection of femininity has been, I would suggest, a common theme. (Jeffreys, 1994: 93-4)

This is not, however, fundamentally in contradiction with the Lesbian Ethics position which did not support roleplaying but merely argued that a femme identification articulates a less radical deconstruction of the patriarchal category of femininity than that achieved by a butch identity in the urban bar settings of the 1950s. Penelope, furthermore, specifically recommended that a lesbian-feminist identification with 'femaleness' (as opposed to 'femininity') can more profitably be approached by abandoning lesbian roleplaying altogether. In any case, unless a lesbian femme adopts an excessively (hetero)feminine personal stylisation, heterosexist culture interprets lesbians (and feminists) willy-nilly as masculinised — that is, as signifying 'aggression' (Nestle, 1986: 102; Hart, 1994). The Lesbian Ethics contributors seem merely to have wished to emphasise that lesbian feminism does not constitute a celebration of (an effectively heterosexual) femininity, but a rejection of it. In other words, it seems to be an argument better addressed to a tendency in 'cultural feminism' to naturalise patriarchal constructs of femininity (see my discussion of radical feminism, Ch.3) rather than to libertarian lesbianism.

Queer theory's model of sexuality seemed to offer a discursive means to supersede the problematic of heterosexually gendered subject positions altogether. Crossdressing could be figured as a deconstructive 'play' with heterosexual gender-signifiers effecting a de-subjectivised lesbian desire (Case, 1988/9; Butler, 1990; de Lauretis, 1991). Radical-feminism conflicted with this queer agenda; partly because of a continued commitment to identity politics but mainly because radical feminism retained the view that a male perspective over-determines any and all representation of desire. A 'true' lesbian desire would, therefore, be unrepresentable by any available means:

We must remember that homosexual (lesbian feminist) desire will not be recognised as "sex." We do not even possess suitable words to describe it. The course of eroticising equality and mutuality has received no prizes from male supremacy ... (Jeffreys, 1990: 315).

Jeffreys, indeed, saw the postmodern project, which was primarily associated with libertarian lesbianism and queer theory, as little more than the queer articulation of a more general reactionary attack on radical feminism.

Feminism as it has been generally understood has been declared impossible. Post-modern theory has been enlisted to support the sexual libertarian and specifically sadomasochistic project. (Jeffreys, 1994: 101)

These issues will be explored in more detail in later chapters dealing both with feminism and with queer discourse, however. I am primarily concerned here with a model of lesbianism as a rejection of (hetero)feminising patriarchal practices. Although avowedly constructionist, this model can be argued, nevertheless, to tend towards the universalisation of the lesbian figure. As Faderman noted (1991) mannishness is very differently coded in different cultural contexts because gender articulates differently in racialised contexts. Shockley (1979) argued that a mannish lesbian stereotype derived from white literature was deployed to shore up African-American cultural myths that lesbianism is acquired from white culture. The sociological myth of the 'black matriarchy' represented black women as "towers" of limitless strength (hooks, 1982: 75-85; Smith, 1983: xxvi), placing additional pressures on black lesbians. Black men often accepted it as a reality which they perceived as further disempowering oppressed black men. Whilst some black feminists felt empowered by it (hooks, 1982: 80), others saw it as occluding the heterosexist oppression of black women (Shockley, 1979; Combahee River Collective, 1981; Smith, 1983).

Black lesbians could also be mythologised as sharing in the black man's "legendary sexual prowess" (Shockley, 1983: 85). The universalising model of lesbianism as a resistance to (white) feminisation becomes still more problematic.

The butch-femme debate in white lesbian culture was primarily a politicised conflict about appropriate strategies for resisting the production of white heterosexual femininity as passive and powerless (this model also seems also to have some resonances for Asian lesbians). (White) lesbian feminisms often saw the valorisation of 'the feminine' as a radical challenge to inversionist models of mannishness. Other lesbians, white and of colour, have re-deployed phallic imagery to a project of signifying lesbian (female) autonomy. Queer lesbians' fascination with the figure of the butch was seen by radical feminists as, once again, colluding with anti-feminist strategy. The political valences of crossdressing practices seem, however, contingent to specificities of context, both diachronically and synchronically.

Conflict and compromise in popular culture

In spite of the sexologists' efforts, the boundaries of femininity continued to expand in the early part of this century. As part of the freedoms associated with the 'New Woman,' Edwardian girls began to be permitted to take part in sport and athletics. Given the association of utilitarian clothing with (lesbian) sexual pathology, however, they were expected to wear 'normal,' feminine clothing with its heavy garments, restrictive stays, and suspenders. Hence, although genteel girls participated in athletic activities in the pages of middle-class Edwardian schoolgirls' magazines, they were not permitted practical clothing (Cadogan and Craig, 1976: 78-79).

An amazonian figure first appeared in the very un-genteel, and male-addressed, form of western 'dime novels' in the latter half of the 19th Century (Smith, 1950: 112) and in Hollywood westerns in the 1920s. Perhaps it was only at the frontiers of society that such risqué representation of women was possible. Smith argued that it was also by coding the heroine as Native American that a greater range of female activity could be represented:

One method of transforming the heroine from the merely passive sexual object ... was to introduce a supposed Indian girl able to ride and shoot who later proves to be an upper-class white girl captured long ago by the Indians. (Smith, 1950: 112)

The dime novels later began to feature crossdressed heroines as a "much more promising means of effecting a real development in the heroine" (Smith, 1950: 112). The earliest crossdressed heroine appeared in the late 1860s (113). Dime-novel crossdressed heroines were armed and often actually fought.

Cadogan and Craig associated the appearance of the crossdressed figure in female-addressed mainstream culture with the entry of women into the professions in greater numbers during the 1920s and 1930s (304). Mainstream representations of this new kind of career girl popularised a new kind of adolescent heroine — the tomboy. The terms in which this figure is here articulated recall the lesbian-feminist model of a refusal, or transformation, of femininity:

[T]he two professions which, in the 1930s, lent themselves most readily to an up-to-date, popularized interpretation were those of girl reporter and girl detective — interestingly enough, these carry with them implications of initiative, mental alertness, deductive ability, physical mobility, courage, technical skill and personal ambition — all qualities which had for so long been out of bounds to women ... (Cadogan and Craig, 1976: 304)

In literature aimed at girls and young women, however, the tomboy's image was "hedged around with compromise" (304). It was in popular cinema, in films such as His Girl Friday, that the tomboy really hit her stride:

The image of the 1930s girl reporter belongs indubitably to the cinema, where it was redefined, enlivened, shown in accelerated motion and embodied in Rosalind Russell or Joan Crawford. (Cadogan and Craig, 1976: 304)

The tomboy figure might thus be seen as a trope which effected some sort of compromise in the struggle between conservative constructs of femininity and an increasingly popularised feminist discourse of "emancipation."

It is interesting to note that this figure maps easily over a Freudian psychoanalytic model of a gender-undifferentiated infantile libido which becomes pathological (lesbian) in women only if retained into sexual maturity (see Freud, 1905; 1925; 1933; and Mulvey, 1981). Love and Abbot (1985) noted the relative acceptability of tomboyish behaviour in pre-adolescent girls (or at least in young, unmarried girls) which would not be permissible in sexually mature women. Pleasure in 'masculine' pursuits would, in a sexually mature woman, inevitably be associated with lesbianism.

Girls going through a tomboy period are not imitating boys as much as they are experiencing a fuller range of activities that will be permitted until puberty brings an end to their freedom ... she [the lesbian] learns that she is "not a woman" if she persists in being competitive or aggressive and independent. (Abbott and Love, 1985: 21-23)

Olson also recalled that the tomboy figure seemed to have no adult analogue:

On screen, tomboys were socially acceptable. As a young butch dyke coming out in 1986, I looked for their grown-up counterparts. I couldn't find anything. (Olson, 1994: 58)

The tomboy tale did, however, spawn an adult escapist version in Britain where there was an existing literary and theatrical tradition of female crossdressing from which to draw a relative respectability. There was also the aristocratic class which the British middle-class traditionally viewed as disgracefully decadent and onto which 'perversity' could easily be projected and distanced. Romantic fictions eroticised tomboy fantasies by the addition of a male desire-object. This male object of desire is often almost as adrogynously presented as the tomboy heroine herself, however — and thus often seems rather to (homo)eroticise the inversionary undertones to the tomboy persona than guarantee her heterosexuality. It would not seem, therefore, that the figure of the pro-active woman can be entirely successfully re-stabilised within heterosexual-feminine parameters. The following extract is from These Old Shades, one of the most popular British romantic novels of its era (reprinted 19 times between 1928 and 1935), aimed primarily at young, middle-class women. This heroine is a teenaged girl passing as a page-boy who falls in love with a foppish and sinister aristocrat (whose presentation strikingly recalls male homosexual stereotyping of the time) in a French, ancien régime period setting. As though the 18th-century weren't quite distance enough, a geographic displacement also seems necessary to open out a safe space for an articulation of resistance, or at least ambivalence, towards the feminisation process.

This is the tomboy's reaction on being obliged to wear feminine clothing for the first time:

"Bah!" said Leonie, and gathering up her skirts, walked carefully across the room. "Certainly I shall burst," she sighed ... "I cannot wear women's clothes. It is as though I were in a cage." (Heyer, 1926: 119)

This unimpeachably-heterosexually-addressed novel does not appear to have had any particular following amongst lesbians, but it was enormously popular with female readers more generally. Another of the same author's popular romances featured a heroine who also passes as a man (also in a period setting) and includes her participation in a sword-fight.

Daphne DuMaurier was another extremely popular writer of romantic novels in the 1930s and 40s. The following extract is from Frenchman's Creek, one of her period romances. The heroine here is a mature wife and mother who has an illicit affair with a freewheeling, sensitive, and artistic lover — the titular Frenchman (again, the British notion of France as already somewhat sexually suspect provides some cultural distance). Again, her crossdressing denotes both rebellion against the limitations of the female role and a fantasy of proactivity:

"If you lend me a pair of your breeches I could climb better", she said... The blood was running from the cut on her chin, and she was soaked to the waist, but she did not care ... (Du Maurier, 1941: 130-2)

This heroine begins the novel roistering with 'roaring boys' in Restoration England — a signifier of female rebellion from the early-modern period (Rose, 1984). Possibly due to gossip regarding the author's personal sexual orientation, sub-textual homoeroticism floats much closer to the surface and Frenchman's Creek, along with du Maurier's still more perverse anti-heroine Rebecca, did appeal to lesbians. The 1970s TV adaptation of Rebecca was included in the British National Film Theatre/Museum of the Moving Image's Out of the Archives III lesbian TV season (screened in July 1994, and reviewed in Diva, #2, 1994, p.42). The film version of Frenchman's Creek, made in 1944, is also specifically referred to by one of Whitaker's (1985) lesbian interviewees:

I really like Joan Fontaine's 1940s film Frenchman's Creek ... She runs away to sea with the pirate, kills a man who tries to rape her, and wears trousers with blousey shirts ... She's in drag, looking like an adolescent boy. (Interviewee quoted by Whitaker, 1985: 109)

The fantasy of sharing a sensitive lover's outlaw adventures in the persona of an androgynous cabin boy might evoke a lesbian-eroticism, but the narrative ends in heterosexual adjustment to dutiful marriage and motherhood. For heterosexual women, the innocent tomboy, unlike the more alarming mannish figure, could function as a comfortably nostalgic fantasy figure upon which the frustrations and conflicts of the feminine role might find a fantasy resolution. Crossdressed adventure can thus signify a temporary flight from the oppressive patriarchal imposition of femininity. However, it also opens out to a lesbian interpretation — as a more permanently real-isable refusal of the restrictions of a heterosexual feminine role.

Crossdressing in lesbian subcultures

As a result of the medicalisation process, lesbianism had been available as an articulable identification prior to the 1940s but conditions had not existed for the formation of substantial lesbian communities. Faderman saw the context of the two World Wars, and especially the aftermath of World War II, as crucially productive in the formation of visible, lesbian, urban-subcultures:

World War II in particular brought great numbers of females of all classes into a society of women where they were able not only to expand friendships but to learn to appreciate other females as serious, self-sufficient human beings ... not long after Freud, The Well of Loneliness, and the term "lesbian" became household words ... [women's] lives were much more open to lesbian possibilities than they could have been earlier. Since World War II also brought large numbers of women to big cities, ... finally relatively large lesbian communities could be created. (Faderman, 1991: 121)

Furthermore, the practical clothing which had been coded as virilising by sexologists during the opening decades of this century (Faderman, 1981 314-331; and 1991: 37-61) had became more acceptable to the wider culture, partly as a result of World War II. Until the 1940s, it had constituted a scandal for a woman to wear trousers in public:

"Innocent bystanders gasped in amazement to see ... Greta Garbo striding swiftly along Hollywood Boulevard dressed in men's clothes." (Faderman, 1991: 125, quoting Harrison Carrol's commentary, "Miss Dietrich Defends Use of Pants," in World Telegram, January 17, 1932, on a newspaper headline quoted by Mercedes de Acosta in Here Lies the Heart, 1968; reprinted Arno Press, 1975, NY, p.229)

In the 1940s, however, women had been obliged to wear trousers in the munitions factories and in other wartime jobs. Thus, not only did World War II create the economic conditions for the emergence of modern urban lesbian subcultures, it also influenced the stylisation of lesbian identity:

In addition to the changing attitudes about what constituted morality, the war also contributed to an easier formation and development of a distinctive lesbian "style" because it made pants acceptable garb for women ... The lesbian who loathed dresses felt much freer to wear pants out of doors than she had in the prewar years. Pants soon became costume and a symbol that allowed women who defined themselves as lesbians to identify one another. (Faderman, 1991: 125-6)

During the 1940s, the lesbian dress code "announced them as lesbians to their neighbours, to strangers on the streets, and of course to all those who entered the bars" (Davis and Kennedy, 1986: 8).

In the fifties, with the increased visibility of the established gay community ... the street dyke emerged. (Davis and Kennedy, 1986: 8)

Most commentators noted the "prominence of butch-fem roles" in these subcultures (Davis and Kennedy, 1986: 13; citing Martin and Lyon, 1972; Lorde, 1979; D'Emilio, 1983; Nestle, 1987). Although this visual coding of lesbianism is now more usually associated specifically with the figure of the butch, anecdotal evidence tends to suggest that "at times femmes dressed similarly to their butch lovers (Nestle, 1986: 102); and that "you can't tell butch-fem by people's dress. You couldn't even really tell in the fifties" (13). That is, to the heterosexual world, both butch and femme undifferentiatedly registered as 'dykes.'

Davis and Kennedy argued, however, that this lesbian-coded environment was also constitutive of lesbian sexual identity and that it was lesbian sexuality, as much as a confrontation with dominant culture, which was organised by signifiers of role-playing:

It is the nature of this community that it created public space for lesbians and gay men, while at the same time it organized sexuality and emotional relationships. (Davis and Kennedy, 1986: 9)

Sexual autonomy for women nevertheless still depended crucially on economic autonomy from men and, at least to some extent, could only be imagined in terms of male privilege. It is, perhaps, difficult to make an effective distinction between sexual and social or economic autonomy:

The rebel lifestyle, in which these women as lesbians demanded some of the social privileges and customs ordinarily reserved for men, ... challenged social orthodoxies about sexuality in the 1950s and 60s ... (Faderman, 1991: 164)

Some of Davis and Kennedy's respondents did, however, see mannishness as confrontational and as having laid the foundations for gay liberation rather than as contributing to the stigmatisation of lesbians:

Things back then were horrible, and I think that because I fought like a man to survive I made it somehow easier for the kids coming out today. (Davis and Kennedy, 1986: 8)

By the 1950s, more formalised lesbian organisations had developed, leading to a public articulation of dissent among lesbians over the conceptualisation and stylisation of lesbian identification.

The decade following the war that expanded the potential of lesbian life-styles did see the formation of the first lesbian organization in America, Daughters of Bilitis (DOB), which was originally founded as a private social group to give middle-class lesbians an alternative to the gay bar scene. (Faderman, 1991: 148-9)

DOB did not long remain a social group, but politicised around demanding lesbian rights and "improving the lesbian image" (Faderman, 1991: 149 quoting a personal interview with Del Martin and Phyllis Lyon in San Francisco, August 14, 1987). "Image improvement" was seen in terms of aligning lesbian style with heterosexual norms in style and dress and an avoidance of those confrontational codes which proclaimed the existence of urban lesbian subcultures.

Some middle-class lesbians complained that it was butches and their femmes who made lesbians outcasts. One of the earliest issues of The Ladder proclaimed: "The kids in fly front pants and with butch haircuts and mannish manner are the worst publicity that we can get." (Faderman, 1991: 180, quoting The Ladder (1958) 3:1, p.30)

Whilst all agreed that the visual stylisation of lesbianism led to confrontation with the dominant order, there were differences in strategy. Most commentators appear to have linked these differences with class position. On the one hand, Nestle (1987) saw the visibility of the bar cultures as a working class "flag of rebellion" (100):

The butch-femme couple embarrassed other Lesbians (and still does) because they made Lesbians culturally visible, a terrifying act for the 1950s. (Nestle, 1987: 101)

Whilst on the other:

[The Ladder's readership] believed that unpopular forms of overt self-expression such as wearing masculine garb led not only to danger for lesbians, but also to further alienation from the parent culture, which was especially painful during a time when the middle-class lesbian culture was still in a relatively inchoate form. (Faderman, 1991: 180)

Jeffreys (1985) argued that, on the contrary, the association of role-playing with working-class lesbians was "not well substantiated and could lead to the creation of false stereotypes of the working class butch or femme" (Jeffreys, 1985: 78). Faderman, however, held that whilst there were role-playing lesbians among the middle-classes this was uncommon, and conflict over the issue between blue-collar bar culture and white-collar assimilationist cultures was documented:

Statistical studies of lesbian couples during the period also concluded that middle- and upper-middle class lesbians preferred to blend in with heterosexual society in terms of their styles ... [and that] "role is more enforced [among lesbians] in the blue collar and lower white collar classes". (Faderman, 1991: 181; quoting Suzanne Prosin, "The Concept of the Lesbian: A Minority in Reverse" in The Ladder, July, 6:10, pp.5-22)

At the same time, the political environment for more confrontational forms of female rebellion was becoming increasingly harsh. The popularisation of Freudianism in the United States during the 1950s had focused public anxiety concerning female emancipation and homosexuality (Faderman, 1991: 130-134). The upheavals of the Wartime period had led to a compensatory conservatism in American society.

The "homosexual" became a particular target of persecution in America. He or she presented an uncomfortable challenge to the mood that longed for obedience to an illusion of uncomplicated "morality". (Faderman, 1991: 140)

During the era of the McCarthy 'witch-hunts' (the investigations of the House Committee on Un-American Activities), homosexuals were associated with subversive threats to the United States and were harassed and persecuted (Faderman, 1991: 141-148; Rubin, 1984: 270). As a result, lesbian sub-cultures went underground. "Lesbians often felt they could not trust close acquaintances with knowledge of their personal lives, even if they suspected those acquaintances might also be lesbian" (Faderman, 1991: 148). For lesbians (and gays), any relation to popular culture would have to be effected through the clandestine transmission of those codes which had come to constitute the visibility of their sub-cultures.

Re-reading popular culture

In the period following World War II, an expansion in publishing diversified popular literature considerably and even produced a specifically lesbian form (Uszkurat, 1993: 26-7). Although often rejected by more literary and academic lesbians, who focused rather on 'high' literature, lesbian 'pulp' novels such as Anne Bannon's Beebo Brinker series remained popular with lesbians in the new urban communities and were regularly reviewed in The Ladder (27). Lesbians identified with Anne Bannon's butch protagonist and accepted such novels as part of lesbian reality (44). Indeed, many looked to them for information about lesbianism (44).

By contrast with popular literature, the cinema of the 1950s has often been characterised in film-studies criticism as reflecting a period of rather seamless moral consensus in US society. This is partly an effect of a critical focus on direct censorship (in the form of the Production Code Administration or formed in 1934) as the primary mode of controlling cinematic representation:

Once the Code's provisions had entered cinematic currency, in fact, they informed both the style and the content of the genres themselves ... Indeed, as long as the industry remained vertically integrated, the power of the PCA remained virtually unchallenged. (Cook, 1985: 8)

Biskind (1983), however, saw the representational field of 1950s Hollywood as far more complex:

Indeed, perhaps the most striking thing about the films of the fifties is that they reflected not one but several warring ideologies, so that it is possible to speak of radical (left- and right-wing) as well as mainstream films. Moreover, they waged this combat ... without explicit political allusions. Films of the fifties, in short, pitted different ways of being and acting against each other. While, say, sensitive men like Montgomery Clift, Tony Perkins, and James Dean were quivering and quaking their way through films like A Place in the Sun, Fear Strikes Out, and East of Eden, John Wayne was still gunning down bad guys in films like Rio Bravo. (Biskind, 1984: 4 -- my emphasis)

In other words, films of the 1950s may frequently be regarded as highly polemical — despite offering a preferred reading of happy conformity and consensus. Homosexual anxieties were among those most often subtextually rehearsed in popular film:

The fear of homosexuality emerged in several films of the late 1950s ... the pressure to conform, to hide any secret sensitivity out of fear of the word queer, was a popular subtext. (Russo, 1981: 108)

Hollywood also reflected a conservative association of homosexuality and lesbianism with "red" subversion:

Several films of the fifties make an interesting set of connections between feminine clothing and capitalism versus masculine clothing and Communism ... At the same time that hardness in women is identified with Communism, softness is men becomes the correlative symptom. (Bell-Metereau, 1983: 101)

Although representation of homosexuality was expressly forbidden by the PCA "in every statute" (Russo, 1981: 30), there were many ways for lesbians and gays to get around this. Gay men involved in Hollywood production found ways to encode subtexts in the films they made, including the use of an underground gay language:

In late 1933 the Hays Office sent out memos to several studios announcing that pansy was now a forbidden word. A few months later Raoul Walsh's Sailor's Luck opened and, in a bathhouse scene, James Dunn pointed at a lisping attendant and said to his seafaring buddies, "Hey, fellas, etgay the ansypay!" To which the ansypay replied by saying, "Hi, sailors!" and wiggling five fingers. (Russo, 1981: 40)

Indeed, lesbian or gay reading of popular film is often argued to require the involvement of gays in the production of the text in order to effect or read appropriate recognition codes, but a lesbian presence could be inferred from stylistic markers associated with extant subcultures included by the film's producers:

The burden of proof for a lesbian analysis frequently depends upon the interpretation of style. (Becker et al, 1985: 301)

This kind of requirement would leave lesbians with considerably fewer options than gay men, however, since there were few identifiable lesbians working in Hollywood film production: only one director (Arzner) and two stars (Dietrich and Garbo) seemed to be widely implicated in lesbian gossip. However, gay audiences were clearly also capable of appropriating cinematic imagery without pointers to gay involvement in production. Babuscio described the analogising of gay knowledges to popular modes of cinematic performance and representation. For example:

The experience of passing is often productive of a gay sensibility. It can, and often does, lead to a heightened awareness and appreciation for disguise, impersonation ... and the distinctions to be made between instinctive and theatrical behaviour. The experience of passing would appear to explain the enthusiasm of so many in our community for certain stars whose performances are highly charged with exaggerated (usually sexual) role-playing. (Babuscio, 1984: 45)

It is less clear what would constitute a specifically lesbian aesthetic unless, perhaps, crossdressing as signifier of female autonomy is cited — and Arzner, Dietrich and Garbo are all associated with crossdressing.

In the field of lesbian literary studies, although work has also generally been dominated by a search for an codifiable lesbian aesthetic in lesbian-authored high literary texts, it has also been argued that lesbian readers do re-signify popular texts. Lesbians "resist 'heterotexts' by privately rewriting and thus appropriating them as lesbian texts" (Zimmerman, 1993: 139). Anecdotal evidence such as that gathered by Whitaker (1985) further suggests that lesbians used similar strategies to re-appropriate the moving image. Critics such as Russo (1981: 170) and Weiss (1992: 64), as well as some unknown lesbians referred to by Sheldon (1984: 16), saw even such pathologising representation of role-playing as that in The Killing of Sister George (1968) as recognisable and re-appropriable by lesbian subcultures.

[Some lesbians] enjoyed some of George's outrageous anti-establishment behaviour (particularly the incident with the nuns) ... or spoke of the character played by Coral Brown as a turn-on (Sheldon, 1984: 16)

A majority of lesbians probably do balk at such alarming images, however (Sheldon, 1984: 12; Florence, 1993: 131). The cost of such lesbian visibility in terms of social exclusion, or even internalised self-hatred, was heavy. The popularity of the compromise figure of the tomboy with many lesbians may thus be partly due to the overwhelmingly stereotypical and tragic representation of lesbianism as such in popular films up to the 1960s, as well as to the invisibility of lesbianism in popular films of the PCA period. Those lesbians who did not simply reject popular forms wholesale inevitably sought ways to access their pleasures by identifying a rebellious and gender-ambiguous female image; but one which did not carry the heavy stigma of pathological inversion. The tomboy seems to have appealed to heterosexual women by offering a nostalgic flight from the restraints of adult femininity, but with a reassuringly normalising resolution. This figure may, however, also have provided a gender-unstable, homoerotic, yet non-stigmatised point of entry for a clandestine lesbian pleasure in popular cinema.

This is not to say that all lesbians read subtextually. The assimilationist attitude overdetermined The Ladder's film reviews, for example. Britain's roughly equivalent political organisation, the Minority Rights Group, and its publication, Arena 3, seems to have taken up a broadly similar stance on discretion, assimilation and the avoidance of sub-cultural lesbian stylisation or "stereotypes" (Groocock, 1994: 92-98). Weiss (1992: 66) stated that attitudes towards popular culture in The Ladder's reviews were generally negative. But it would not seem that its reviews were by any means consistent in relation either to popular culture or to crossdressing. It was often sympathetic to butch characters in lesbian pulp fiction, such as Bannon's novels (Zimmerman, 1990: 9; Grier, 1976: 203; Uszkurat, 1993: 26). Film reviews were by no means overwhelmingly negative either. The following reviews of Club des Femmes (which originally appeared in The Ladder's forerunner, Vice Versa) and Therese and Isabel (negative attitudes towards which latter probably reflected) rather nicely sum up assimilationist attitudes:

The role of Alice [in Club des Femmes] was portrayed with admirable good taste. One could not help but admire the character of the quiet, self-contained lesbian who bore such great affection for her girlfriend, yet forebore ... to enter her girl friend's room to keep her company in the evenings. Possessed of great beauty, both spiritual and physical, [the role of] Alice was handled in a sane, intelligent manner rather than furnishing the usual subject for harmful propaganda or mere sensationalism. (Lisa Ben, 1960, The Ladder, April, 4:7, p.8-9)

On the other hand the European art-house/soft-porn aesthetic of Therese and Isabel did not go down well at all:

A movie has been made from Violette Leduc's Therese and Isabel. It is worse than the book, ... see it if you must, but you have been warned. (The Ladder, Dec-Jan, 1968-1969, reprinted in Grier, 1976: 121)

Another important result of The Ladder's humanistic and aspirational approach was that it tended to deal with texts pretty much at face value. That is, its reviewers either approved or disapproved of preferred readings offered by dominant film texts, but did not recognise any potential for resistant reading. For a film to be of interest to lesbians, therefore, it had to be lesbian-authored, or at least to deal overtly with the theme of lesbianism. Setting aside a post-modern critique of authorship and authenticity, there was, in any case, virtually no lesbian-produced cinema which could be seen as in any way culturally analogous to lesbian-authored literatures. Maidens in Uniform (1931) has sometimes been offered as an early candidate, since the scriptwriter identified as lesbian, but even so the evidence does not support a view of lesbian control over its collective production (Dyer, 1990: 30-32). There were, however, a number of films made in Europe between the 1920s and the 1950s which treated lesbian themes relatively openly and sympathetically. Besides Maidens, these included Club des Femmes (1936), and Olivia (1950). They represented an intra-feminine desire between conventionally-feminine-styled women in professionalised, all-women, environments. As such, they were generally acceptable to assimilationists (Ermayne, for example, described Club des Femmes as "The best of its kind made to date" in The Ladder, 4:7, 1960: 5 — but it is not clear whether this film was viewed in the censored form exhibited in the US, from which "all intimations of lesbianism were cut." (Faderman, 1991: 103 quoting Time, 25 Oct, 1937, pp.26-28).) The Haunting (1963) with its elegantly stiff-upper-lipped lesbian heroine was also recommended by The Ladder (Jan 1964: 23).

Such conventionally-styled and tastefully restrained representations were few and far between, however. It would not, in any case, seem at all evident that a clear dichotomy between cinematic representation of lesbianism as intra-feminine desire or as identification with masculinised female figures can actually be drawn in practice. As Weiss (1992) pointed out, representations of same-sex desire in same-sex environments (which were more acceptable to assimilationist and feminist lesbians) usually also made iconic reference to inversion (visually signified by crossdressing). Furthermore, films in which women wore trousers frequently also represented an attraction (usually based on mis-identification) between two women. The two modes were often effectively difficult to distinguish:

[There was during the 1920s and 1930s] an inconsistency or confusion of visual codes . . . lesbianism pertained primarily to desire ("object choice"), but also surfaced as gender inversion. These images of crossdressing and gender inversion on the one hand ... and a feminine-identification involving a female object choice, on the other, were concurrent images in the cinema, and the former never entirely gave way to the latter. (Weiss, 1992: 17)

The wearing of trousers by female characters had seemed to provide a suitably visual means (for the visual medium of cinema) of encoding a sub-textual working-through of heterosexual spectators' conflicts and anxieties concerning the masculinisation of women during World War II:

The popularity of the tomboy reached its peak in the years after the Second World War. A number of these films showed women participating in the masculine realm, often dressed in appropriate male attire, but reduced the threat of such women by placing them under the tutelage or control of a man. (Bell-Metereau, 1993: 96)

If it was by the wearing of trousers that many urban lesbian sub-cultures effected mutual recognition (by deploying a subcultural knowledge), as Faderman argued (1991: 125-6, quoted above), then it seems logical to assume that similar visual codes might also be deployed to effect clandestine lesbian readings of popular film.

Whitaker's interviewees (1985) tended to confirm that lesbians, like their gay counterparts, could read lesbian subtexts in(to) popular film texts. However, whilst Babuscio referred gay recognition codes to iconic Hollywood divas such as Bette Davis and Mae West, lesbian commentators appear to have cited a different style altogether. Sheldon's remarks on the type of films which were popular with lesbians of the 1930s and 1940s closely parallel Faderman's account of the origins of lesbian bar culture during the postwar period:

[L]esbians' interest in the cinema seems to be oriented towards those exceptional films made in Hollywood during the late 1930s and the 40s, when the needs of patriarchy/capitalism to make war and money demanded that women be oriented away from home-making and into industry to replace men sent away as cannon-fodder. These films often has as central characters strong and resilient women ... (Sheldon, 1984: 17)

In the 1950s, the tomboy seemed to offer lesbians a similar form of recognition:

More recently, the cow-girl films [Annie Get Your Gun (1950), Calamity Jane (1953), etc] ... enjoy some popularity. (Sheldon, 1984: 17)

Fantasy identification with androgynous, or even male, characters has been widely noted as a strategy used by lesbians to engage with Hollywood texts in the 1920s and 1930s. Crossdressing comedies and stylised dramas such as Sylvia Scarlett, Queen Christina or Morocco are referred to again and again in what little literature exists (Russo, 1981: 14-15; Koch, 1982: 51-3; Sheldon, 1984: 17; Studlar, 1991: 4; Weiss, 1992: 30-50; Straayer, 1992: 38-39; Olson, 1994: 66). In relation to the 1950s, the western tomboy figure, and even the cowboy, also come up surprisingly often (Sheldon, 1984: 17; Whitaker, 1985: 107, 116; Bell-Metereau, 1993: Olson, 1994: 58; 80-95; Modleski, 1995/6). There is also more recent evidence of lesbian interest in western iconography refracted through a lesbian craze for Country and Western music in the late 1980s linked to the popularity of kd lang [sic]. Western iconography has been identified as a significant factor in the lesbian appeal of films such as Thelma and Louise, or Salmonberries (Johnson, 1993; Ainley and Cooper, 1994; Tasker, 1994: 182; Thynne, 1995).

The dominance of discourses of mannishness during the 1950s, combined with the structural imperatives of the visual relay identified by feminist film theory, might very well have been expected to promote a lesbian form of male identification in popular film. The anecdotal evidence does not seem to suggest, however, that lesbians identified fully with male characters any more than they did with negative stereotypes of lesbianism. Lesbian sub-textual strategies do not necessarily imply a sutured identification with male or masculinised characters or spectatorial positions. That is, they do not imply a mimetic identification. On the contrary, it seems that lesbians attempted to insert themselves into the cinematic visual relay with a marked degree of self-consciousness about the 'lack of fit.' Indeed, this often seemed sufficient to invoke an impression of "watching the effect" of the IMR upon the (lesbian) self as the lesbian spectator (Whitaker, 1985: 110). It has to be remembered, though, that Whitaker's interviews are anecdotal and do not constitute formal "sociological or psychological studies" (Whitaker, 1985: 106) — the interviewees, although all college educated women (106), were not necessarily familiar with the precise usage of terms such as 'identification' in film theory and "some of the women use 'identify with' in a way that could be interchangeable with 'love'" (107). Some of the interviewees did, however, clearly mean quite specifically to articulate identification with male or masculinised characters — and the western genre is specified surprisingly often in this context:

ROMAINE: To the extent that I identified with film characters, it would be with Robin Hood, male characters, cowboys ... (in Whitaker, 1985: 107)

EA: ... I always identified with cowboys in the westerns . . . (Ibid: 110)

ROMAINE: I identified with cowboys too, and paid no attention to the sex differences ... I identified with male characters even after adolescence, and maybe still do ... (Ibid: 110)

LULU: From westerns I incorporated into myself a strong silent image. Enough that I'd like to get rid of some of it. (113)

J A MARQUIS: I've thought about it. I was identifying with men. At age eight, nine or ten and on, when it came to Elvis Presley, I didn't scream, or drool ... I wanted to be Elvis Presley. (Ibid: 113)

These little tomboys empowered me to think of myself as a hero. They were strong and smart like the movie cowboys and gangsters I emulated . . . (Olson, 1994: 58)

Again, these lesbians seem to be articulating identification with this figure in terms of an idealised fantasy of proactivity offered by the image rather than in terms of desire for the woman as 'other'.

The western Johnny Guitar is frequently cited as a camp classic which "features a butch Joan Crawford and an even more butch Mercedes McCambridge in a series of confrontations that keep present-day gay audiences howling" (Russo, 1981: 103; see also Bell-Metereau, 1993: 87; Olson, 1994: 58). Bogus (1994) also recalled identifying with cowboy characters in black-cast films (made for the new urban black audiences in the 1920s and 1930s, copying Hollywood themes and presentations, with white-backed production, but all-black casts). The word signifying a "mannish" black women in the US in the 1920s and 1930s was "bulldagger." Bogus (1994) associated this term with Bill Pickett, a real-life black cowboy who became a legendary rodeo performer before starring in early black-cast Westerns.

Once we had seen that movie, we talked about Bill Pickett, and we talked about his exotic life, and we said "bulldogger" whenever we could. (Bogus, 1994: 34-35)

Identification with male characters is, however, articulated far more equivocally by some, and rejected outright by others:

DAGMAR: I did try hard to like male characters because I thought I was supposed to, but I found most of them nebbishes. (Whitaker, 1985: 109)

ANNA MARIA: I always worried about it ... I was always really pushed to resolve in some way what just would not fit, and so I found it hard to identify with male characters. I had to change them in some way. (Whitaker, 1985: 110)

FANNIE: I was very confused. I knew I wasn't an Indian, Superman or a cowboy. I was Lois Lane and Wonder Woman a little. (Whitaker, 1985: 112)

One of the seven (out of nine) of Whitaker's interviewees who did articulate her identification with male characters pithily explained this preference:

ROMAINE: ... Women characters were all boring ... (Whitaker, 1985: 107)

It is particularly interesting to note (given the emphasis of lesbian historians such as Nestle (1987) and Davis and Kennedy (1986) on butch-femme roleplaying as fundamentally sexual) that, among Whitaker's interviewees, those who most strongly articulated desire for the feminine star also articulated considerable ambivalence towards masculinised identifications:

You're attracted by these women and yet you don't fit in ... At times I'd identify with a character. Other times I'd float outside the situation, sort of watching the effect this attractive woman was having on me ... or sometimes I'd be Katherine Hepburn. And I might be sort of behind Spencer Tracy but I wouldn't be Spencer Tracy ! ... (Interviewee in Whitaker, 1985: 108-110)

None of the interviewees articulated identification with male stars in terms of this as offering a position from which to desire 'the woman' but solely in terms of their pleasure in the fantasy of autonomy and untrammelled activity.

ROMAINE: The cowboy archetype meant being strong and independent, not needing anybody, and moving around a lot. Being free and living with nature were especially important. As a child, independence was something I needed to feel. I had to feel free of society -- an attitude I now see as alienated in many ways. (Ibid, 1985: 110)

Bogus also made it clear that it was the representation of skill, independence, and capability in the character of the "bulldogger" which appealed to her black lesbian contemporaries (Bogus, 1994: 34-35). She does not mention sexual desire in this context. A black lesbian interviewee of Whitaker's, on the other hand, reported finding cowboys too "clean cut" (116). She articulated a preference, instead, for outlaw figures and for "smut." Again, there is a dislocation of sexual desire from specifically gendered identificatory practices. She seemed quite indifferent to the gender of the star, as long as they signified as rebels and "outsiders:"

Gene Autry was the dullest. I would identify with the cowboy who got away with the money and the girl, or who died in the street ... [In Anna Lucasta] Eartha Kitt ... ran away from her family and did what she wanted to do. I can relate to that as far as being lesbian ... (Ibid: 116)

The following anecdotal evidence from Whitaker may throw further light on the rather fragmentary processes by which lesbians seem to have engaged with popular film and, like Bogus' commentary, it also identifies cinema as having had a function in forming lesbian identity. Rather than referencing fixed identifications, there appears to be a process of trying out gendered roles to find a workable compromise with dominant representations that more or less 'fits':

JUDY: It wasn't film that told you it was okay to love women.

ANNA MARIA: No, but it did give me the context in which to play out all these ideas, fantasizing romantic encounters and playing one of many roles, always switching around when it got too hot in one seat. I eventually worked out something acceptable to me. (Whitaker, 1985: 110)

Anecdotal lesbian accounts of film spectatorship seem to concur rather strikingly with theoretical feminist accounts of the significances of lesbian deployments of masculinising iconography which are outlined at the start of this chapter. The constant emphasis on the physical activity of the crossdressed icon also seems to effect a spontaneous echo to Mulvey's (1981) theorisation which posited the instability of gendered identifications in Hollywood film, by reference, precisely, to the compromise figure of the tomboy. The tomboy icon always seems to refer itself to a struggle against the restrictions of femininity rather than directly to masculinity as a mimesis. It evokes female autonomy, skill and motor pleasures (which have, of course, an erotic component) which dominant discourses make available — and even enforce as the presentation of masculinity — but which conflict with a conventional presentation of femininity within the heterosexual order. It thus enables some refusal of patriarchal modes of feminine performance without binding lesbians over to the category of masculinity.