Male homosociality and the buddy movie

There have been various accounts of the allegorical significances of nostalgic narratives in revisionist westerns for North American society during the 1970s which are not relevant to my present argument. What is most relevant to an analysis of representations of sex-gender in this cycle of films is the idea that the increasing popularity of buddy films in the 1970s and 1980s may have reflected more general male anxieties. The buddy movie represented an intensification of the (phallic) nostalgic mythologisation of the figure of the untrammelled western outlaw as anti-hero of the times (Wright, 1975; Films, 1983). The visual and narrative ordering of the buddy film marginalised women, or excluded them entirely. This exclusion of 'the feminine' was motivated by the need to deny the destabilising effects of feminism and to shore up the institution of homosociality itself.

[The buddy film] reconfirmed white male self-identity within radically shifting social and political frames; these men without women defined themselves in opposition to the (absented) other. (Fuchs, 1993: 196)

Mulvey's (1975) argument that the male buddy movie circumvented problems of verisimilitude produced in the classic relay of the male gaze through that of male protagonist onto the female was purely textual. For Mulvey, this form addressed a fundamental problem of the homosociality of the scopic relay itself:

The presence of woman is an indispensable element of spectacle in normal narrative film, yet her visual presence tends to [...] freeze the flow of action in moments of erotic contemplation [...] (A recent tendency in narrative film has been to dispense with this problem altogether; hence the development of what Molly Haskell has called the "buddy movie," in which the active homosexual eroticism of the central male figures can carry the story without distraction.) (Mulvey, 1975: 62)

Mulvey's tangential remark did not develop the theme but, in terms of Neale's (1983) revision of her general argument, homoeroticism would inevitably be invoked by the scopic organisation of cinematic representations of the exclusively male bond. Thus "the exclusion of women compelled overt condemnation of implicit and even explicit homoeroticism" (Fuchs, 1993: 196).

Such homoeroticising effects may not only have been due to the absence or marginalisation of women in the buddy film and resultant disturbance in the relay. There may also be a feminising tendency in representation of the corporatising effect of male comradeship. That is, collectivity always tends towards feminisation precisely because it signifies invasion of individual boundaries, cancels differences between men, and effects a transgression of boundaries between self and other, inside and outside, "legitimate and illicit" (Fuchs, 1993: 194). In militaristic comradeship, this effect is enhanced because the presence of death and mutilation is also signified (Neale, 1983; Theweleit, 1989). Militaristic authority and the intense bonding process which incorporates and supersedes the individual male into a corporate military identity were identified as the homosocial processes most potently haunted by the spectre of homoeroticism (Waugh, 1977; Daly, 1978; Lewis, 1983; Theweleit, 1989). In such environments, the male bond is argued to develop fascistic undertones when a militaristic, violent, elite group is forged in a flight from "the feminine" (Waugh, 1977; Lewis, 1983; Theweleit 1989: 319-320).

The pervasive anxiety about homosexuality which haunts this era might, then, be seen as an effect of the defensive intensification of male homosociality in response to feminism. (It is important to emphasise here that it is not homosexual activity or identity which is under discussion, but phantasmic heterosexual male anxieties. Although Neale's (1983) analysis indicated that homoerotic subtexts in action genres may be subversively re-appropriated by gay men, this is obviously not their direct mode of address or preferred reading.) In his critique of An Officer and a Gentleman (1982), Lewis (1983) linked the buddy-film not only with militaristic nationalism but explicitly with anti-feminism (13) and misogyny:

[D]iscipline shapes rather well-defined 'individuals' who are not only tenacious (in their ability to survive and succeed) but also quite willing to accept (paradoxically) that incredible sublimation of selfhood for the good of the male group [...] women here remain at the periphery of concerns and action [...] women become the real object(s) of his scorn and blame and externalised rage [...] Their very existence threatens the male group. (Lewis, 1983: 13)

Rosenthal's description of male bonding in buddy films also linked the homoerotic subtexts of buddy films to misogyny:

[R]elationships with women are always secondary, [...] the homoerotic tensions that structure these films tend toward an anti-social misogyny, which reduced love to a fascistic military solidarity. The outlaw code — the band of brothers living honestly outside of civil law can easily be transformed into the fascist band of brothers living above the law and enforcing it [...] the films are homophobic as well as misogynist. (Rosenthal, 1978: 1)

Rosenthal (1978) saw The Wild Bunch (1969) (in its corporatising representation of the male bond) as prototypical of the buddy movie, but most critics claimed Butch Cassidy and the Sundance Kid (1969) as the archetype. This latter is, more typically, organised around a paired-bond between two men rather than a militaristic or quasi-militaristic group of men.

Wood (1986) did not think that the more typical 'paired' buddy films should be interpreted mainly as a reaction to feminism, however (230). If women carry the symbolic function of marriage and society in Western narratives, Wood saw the homoeroticism of the buddy movie as breaking through the repressive dominance of the heterosexual family.

[I]f women can be dispensed with so easily, a great deal else goes with them, including the central supports of and justification for the dominant ideology: marriage, family, home. (Wood, 1986: 227)

Such representations appealed to what Wood saw (somewhat optimistically) as the heterosexual male's "unconscious but immensely powerful need to validate love relationships between men" (Wood, 1986: 230). Of course, in terms of the model of male bonding being used by the other theorists of the buddy movie, this would still cite homosociality as the guarantor of heterosexual masculinity. Wood did, however, note the almost hysterical level of denial in these films. The homoerotic implications of such intimacy (and of the relay of the phallic gaze onto the male) are most usually displaced by projecting homosexuality outside the central male bond as the property of another, despised or comic, character.

The overt homosexual (invariably either clown or villain) has the function of a disclaimer — our boys are not like that. The presence of women in the films seems often to have the same function: they merely guarantee the heroes' heterosexuality. (Wood, 1986: 229)

The male buddy movie originally occurred most commonly in the western genre, which is associated with adolescent male thrills. It is now most common in the detective genre, which arguably took over the western's thematic and adolescent audience when the latter's popularity lapsed (Films, 1983). It would seem that the western's generic function of underpinning a mythos of society and nation, and thus of reflecting and controlling anxieties about race relations, also shifted to this new genre. The detective-buddy film also began to articulate white anxieties concerning shifting racial boundaries. Anxieties about race and sex-gender boundaries were frequently conflated.

The Lethal Weapon series (1987–1992) focused critical debate on a cycle of interracial buddy films in the 1980s. Whilst Guerrero (1993) saw it as one of the "least-worst" of the interracial buddy sub-genre (244-5), Modleski saw Lethal Weapon as "a heavily condensed mixture of racism, misogyny, homoeroticism, and heterosexual panic" (Modleski, 1991: 141). Jones (1993) saw Glover's "normally situated" Black character as "sanitised beyond recognition. Robbed of all traces of ethnicity or real sexuality . . ." (252-3). Fuchs (1993) and Tasker (1993) both also noted that the Lethal Weapon series displaced homosexual anxieties evoked by representation of the homosocial bond onto a racially defined boundary of 'otherness.'

The secrecy of masculine intimacy and vulnerability is sustained in these films by the 'marriage' of racial others, such that the transgressiveness of black-white difference displaces homosexual anxiety. (Fuchs (1993: 195)

The Lethal Weapon films manage to avoid any homoerotic inflections of the buddy pairing by reinscribing difference within the terms of race [...] Both men are further defined in the film [Lethal Weapon 2] against Leo Geetz (Joe Pesci), a witness they are assigned to protect. Leo's comic role functions to produce a figure who is both feminised and trivialised by his hysterical speech. (Tasker, 1993: 46-7)

Ames saw a revisionist imperialist agenda as more central. In his analysis of the interracial detective-buddy film, he related the significance of this new sub-genre to Fiedler's (1949) comments on the Western 'dark girl'/buddy and concluded:

This [American] archetypal relationship is manifested today most prominently in the black-white "buddy films," [...] Black and white protagonists are paired, not in a virgin wilderness, but in its metaphorical and corrupted contemporary counterpart, the urban jungle of violent crime [...] [in a mythic reversal] the white man is clearly the savage equipped for survival, while the black man has become a highly civilized figure who has lost touch with his savage masculinity [...] As in the traditional myth, a homoerotic or homosocial bond emerges, uniting black and white man against the civilized repressions represented by women and the law [...] [expressing] a powerful mythic longing to reverse or purge historical guilt. (Ames, 1992: 52-3)

Fuchs noted a similar sublimating effect in Heart Condition (1990):

[T]he plot device that transplants upscale lawyer Stone's heart to racist cop Mooney's chest [...] points out spiralling ideological and cultural tensions that structure the generic buddy film [...] First, the men's literally shared heart figures a masculine hegemony which appears to subsume such tensions, an all-male unit transcending race and class distinctions to produce stable self-identity. (Fuchs, 1993: 194)

As in the traditional Western, the possibility of substituting a female for the 'dark companion' was exploited in films as diverse as the genre spoof Streets of Fire (1984) and Lethal Weapon 3 (1992):

Lethal Weapon 3 actually proves the rule that these heroes cannot be allowed a central erotic relationship with a woman because it distracts from the primary love of the male companions [...] Lorna Cole is essentially [...] a female man. She is a permissible sexual companion for Riggs because she combines macho physical intensity with a female body - thus the eroticism latent in the scenes of the two male heroes can find acceptable expression [...] (Ames, 1992: 55)

The female protagonism of pro-feminist buddy-films such as Thelma and Louise might, then, be seen as effecting a similar mode of displacement — but in the replacement of both buddies by female protagonists, a subtextual homoeroticism is feminised (lesbianised).

The male buddy film, then, seemed implicated in an effort to re-establish and simultaneously conceal a threatened exclusionary white, male bond through a drastic denial of differences of both race and sex — often through the substitution of one set of signs for another. Simultaneously, in a fetishistic manoeuvre, the male spectator is able to deploy such differences to a denial of the homoeroticism of the male homosocial bond which he seeks to verify. This is possible through the inversion of the dominant tendency to hyper-virilisation of black masculinity on the one hand, and the 'inverted' virilisation of the white woman on the other. The two main 'buddy movie' forms (corporate or paired) show similar tendencies. Both evade the problematic of the cinematic representation of women. Both display a more general fear of the 'castrating' effect of closeness with women as well as a homophobic anxiety about the feminising effects of closeness with men. Both nevertheless evoke a homoerotic intensity to the male (or female) bond. Both are linked with revisionist agendas. However, whilst the former is usually overtly nationalistic and militaristic, and concerned with the maturing of the individual (his incorporation to dominant forms of masculinity) through suffering, the latter aligns itself with the phallic-nostalgic thrills of a fantasy wilderness — be it urban or 'natural.' Both forms have also developed pro-feminist analogues.

New Women's Cinema and the buddy movie

There was already a long tradition of setting novels and films in all-female contexts such as schools, prisons or other institutions, for a range of allegorical or titillating purposes; but a pro-feminist discourse of 'female-bonding' was carried over to the New Women's Cinema during a peak in feminist critical activity in the 1970s). The feminist buddy movies of the 1970s which are analysed in the next chapter can thus be seen not only as a product of a feminist discourse of female-identification, but also as a reversal of the male-buddy genre which was popular at that time. Feminist accounts of Thelma and Louise cite this generic origin (Rapping, 1991: 30; Dowell, 1991: 28). Given the radical disjuncture in the meanings ascribed to male and female same-gender bonding, however, a wide gulf must be expected between the preoccupations of the male buddy movies and the cinematic exploration of female bonding and female identity. The interest of pro-feminist filmmakers in representing women's relationships in the mainstream was also seen as an appropriative or incorporative mode of "digesting the women's movement" (Vicinus, 1978; 1).

Feminist reception of what Kuhn (1982) called "New Women's Cinema" and Haskell refers to as "neo-women's films" or "quasi-women's lib pictures," was very mixed. In terms of a feminist construct of positive images, these films generally registered as negative in the 'by-men-for-men' category. Haskell argued that women's "rebellion is defensive, and her victory often pyrrhic" (Haskell, 1974: 336). Much feminist criticism took a cynical view of changes in the representation of women. For Haskell (1974), New Women's Cinema was, in any case, a mere flash in the pan:

The trickle of feminist inspired movies of the mid-seventies - A Woman Under the Influence, Alice Doesn't Live Here Anymore, and An Unmarried Woman - had led us to anticipate, if not a revolution, at least a gaggle of films that would chart our evolution as emerging feminists [...] Instead, women virtually disappeared from the screen [...] for over a decade [...] as the multinational executives [...] of the "New Hollywood" panted after the youth market [...] (Haskell, 1987: 276)

Kuhn, however, disagreed that the phenomenon of pro-feminist cinema could be treated as a simple reflection of a social preoccupation with feminist agendas; particularly as it coexisted with a spate of exceptionally violently anti-women films. Rather, Kuhn argued that there had been a segmenting of Hollywood audience, with these films being aimed at a specifically feminist 'sub-audience' (Kuhn, 1982: 125). Kuhn turns out to have had the more accurate reading, historically.

There was also a divergence between heterosexual feminist and lesbian-feminist attitudes towards representation of the intra-female bond. Neither heterosexual nor lesbian feminists responded well in the 1970s to the sexualised exploitation of female intimacy in male-directed female buddy films (Vicinus, 1978). Heterosexual feminist filmmakers accordingly struggled to effect denials of any (homo)eroticisation of the relay which might be evoked as women look at other women. By the 1980s, however, a generally deconstructive attitude towards the hetero-institutionalised structure of the cinematic gaze seems to have facilitated lesbian pleasures in pro-feminist mainstream buddy films such as Times Square (1981) and Thelma and Louise (1991). This may have been partly a factor of an inevitable diversity of lesbian response at any given time; and partly due to a paradigm-shift in lesbian attitudes towards popular cinema which seemed to occur during the 1980s. However, it may also be because the later pro-feminist buddy films incorporated iconography and narrative codes already associated with the tomboys of 1950s and 1960s westerns and thus already had some 'built in' appeal for lesbians.

Heterosexual feminist critiques of the 1970s buddy film were all too often informed by negative attitudes towards both male and female homosexuality. Haskell was angered by the marginalisation of women by a male-homosocial focus in mainstream film — the British allegedly being the worst offenders:

[T]he English cinema — swinging-mod and kitchen-sink — established a certain cynical, derisive, and crypto-homosexual attitude toward men and women [...] The homosexual theme finally becomes explicit in Sidney J Fury's The Leather Boys, where poor Rita Tushingham is enough to drive her young husband from their marriage bed onto the back seat of another fellow's motorcycle. (Haskell, 1974: 326)

Indeed, she seemed to feel that it was male homosexuality, rather than the homosociality of heterosexual men, which was responsible for the exclusion of women from the representational order of "New Hollywood Cinema" (Haskell, 1987: 358). This seemed a popular radical-feminist view throughout the late 1970s and 1980s. In any case, lesbians do not fare any better in Haskell's critique. After disapproving of the sensationalism with which The Killing of Sister George (1968) tackled what she described as "unpleasant truths" (337), she went on to lambast "militant lesbians" for "looking for clues to lesbianism" in the texts and biographies of their authors:

The vision of the virginal woman as some kind of freak (a vision no less perverted than the concept of her as sacred) is purveyed not just by conventional heterosexual society, but by militant lesbians as well. Village Voice writer Jill Johnston ransacks the lives of deceased artists looking for clues to lesbianism and is concerned when their lesbian potential has gone unfulfilled; [...] an over-simplification that does great injustice to the female sensibility, to isolate orgasmic sexual fulfilment as the supreme and only form of sexual experience. In this, Johnston and other chauvinistic lesbians merely adopt the imperialistic characteristics of male sexuality that they presumably abhor, and use absolutist labels that exclude a wide range of affective feelings and behaviour. (Haskell, 1987: 339-340)

An anxiety that female autonomy or intimacy might provide clues to lesbianism also seems to pervade the heterosexual feminist approach to the buddy film. Whilst male-directed films were criticised by heterosexual feminists for lesbian-eroticising women's relationships; lesbians seemed to find little that is erotic about them (Vicinus, 1978; Bailin, 1979).

Vicinus (1978) saw Julia as a project of containment of the feminist 'threat' which exploits a subtextual lesbian-eroticism whilst suppressing any politicised signs of lesbianism. The film focuses primarily on the adult, heterosexual Lillian and her relationship with Hamnet, locating her love for Julia in their adolescence. Julia's rebellious, independent and highly politicised existence is marginalised, and Julia's death resolves the narrative for the heterosexual character, Lillian.

[F]emale buddy films] [...] are about female relationships which are based in the past, before each pair of women has begun "real" life. . . films about male friendship depend upon an outlaw code, placing them outside of, and in opposition to, civil law and domestic life. But the outlaw code has always been ambiguous - while life outside of the civil order may remain an extended adolescence [...] it may develop into an alternative Civitas of its own [...] I can't think of a corresponding example, in popular culture, where a female outlaw society would be allowed to become strong enough to constitute an alternative civil order [...] female friendships are perhaps most easily portrayed as occurring in adolescence. (Rosenthal, 1978: 19)

Both Rosenthal and Bailin noted that, as with the 'phallic' pro-activity of the tomboy, bonding between women is confined to the period of adolescence and is 'always already' displaced by family responsibilities. There were also formal differences between female and male buddy narratives. Rosenthal described how, in the male buddy film, the male bond is represented through the intimacy of sharing everyday life, resolving conflict and tension, laughing off undignified errors, as well as pursuing personal aspirations; and, in the end, sharing death.

[...] [M]ale-male relationships are constituted of endless joking and bickering - infinite adjustments to be made between inflated egos, a mock antisocial marriage in the face of danger and death [...] (Rosenthal, 1978: 1)

Cinematically, the impression of intimacy between two men is built up through the frequent "appearance of characters together in a frame or the strong connective cutting between them" (Rosenthal, 1978: 1). By contrast, Julia represents little of a process of bonding. Rosenthal noted that the two women barely meet during the course of the film and are rarely in frame together. The girls' relationship is frozen in time; an "oasis" in memory for the adult Lillian. This past relationship is represented as pure and ennobling - one cannot imagine the girls rough-and-tumbling (like tomboys), making fools of themselves, learning, and moving on. Though the girls do face danger, this, again, is for an ennobling cause — represented almost as spiritual. There is no sense of the achievements, risks, petty scuffles, and power-struggles that make up everyday life and relationships.

Bailin (1979) agreed that, in spite of their apparently self-conscious feminist intentions, female buddy films did not present an entirely positive image of female friendship. The female buddy film of New Women's Cinema did not displace or marginalise heterosexual bonding as the male buddy film tends to do. Rather, a feminine protagonist, firmly situated within a present primary heterosexual bond, reconstructs the memory of a bond which is positioned as pre-heterosexual. Nevertheless, these films-texts did reveal anxieties about the potential homoeroticisation of the women's relationship. As in the male-buddy film, these were displaced onto a character external to the buddy relationship:

Fred Zinnemann [...] seems unable to admit to the possibility of close friendship between women (or men, or men and women, for that matter) without a sexual component; [...] Yet this very attraction appears to be denied in the melodramatic scene of Lillian turning the table onto the smirking Sammy for hinting at a sexual relationship between her and Julia [...] the connection is clearly made between lesbian love and Sammy's incest with his sister. The vileness of Sammy confirms the vileness of female sexuality. (Vicinus, 1978: 6)

For Bailin, the connection made between female sexuality and Sammy's incest confirms the 'vileness' of the intra-female bond as much as it confirms the characters' (hetero)sexuality.

Lesbian-eroticism or woman-identification quotients were often still lower in feminist-directed buddy films which, self-consciously rejecting commercialising sexual sensationalism, tended to work still harder at the denial of any lesbian subtext. Girlfriends (1978) was both woman-directed and an overtly feminist film. It was seen by its female director and star as a transparent window onto a feminine world: as expressing the 'truth' of female identity. The identities of director, protagonist, feminine object of the protagonist's "desire," and of the (heterosexual) feminist spectator, were all collapsed together:

[F]or the first time she [Mayron] was playing a part that was close to herself, so it felt as if she "didn't have to do anything". . . Now most of her friends tell her it's the first 'women's picture' where they can really identify with the heroine; she feels the same way and so will many other women in the audience. (Kinder, 1978: 46)

and:

Weill claims that the film "is autobiographical only in the sense that I know the material — the milieu, not the events. I feel that I've been Susan. I've been Anne." (Kinder, interviewing Weill, 1978: 46)

Weill's stated intention was to show, realistically, the complexity of a female bond seen as comparable with a marriage. Fischer described the way in which the women's intimacy is set up in the film's early scenes:

[W]e see Susan early one morning photographing Anne as she sleeps. In the next scene, we watch Anne read her poem to Susan, soliciting an opinion. Clearly, the women share a spirit of creativity [...] As Anne reads her work, Susan sits on the toilet, demonstrating the mundane intimacy of the roommates' lives. (Fischer, 1989: 234)

From a lesbian point of view, however, female bonding is quickly resubordinated to the primacy of heterosexuality. Bailin noted that, as with Julia, there are few cinematic codes for intimacy between the women. Anne and Susan have few scenes together. There is little eye-contact, touching, framing of the girls together, or connective cutting. The film focuses more on what separates the women than that which bonds them. The women's relationship is overdetermined by their heterosexual relationships — their only hug celebrates the success of their separate heterosexual bondings (Bailin, 1979: 3). This can, of course, be seen partly as an effect of its realist aesthetic since the reality for heterosexual women is that a primary focus on heterosexual bonds with men characteristically stresses, and supersedes, relationships between heterosexual women. Girlfriends' narrative concerns Susan's, apparently doomed, efforts to change this tendency.

In common with Julia, Girlfriends deploys the strategy of including a marginal homosexual character as a way of externalising and denying any homoerotic content to the buddy relationship. The film was thus criticised for marginalising and pathologising lesbianism:

Ciel's relationship to Susan is, I feel, indicative of how liberal feminism sees lesbianism. Susan isn't freaked or repulsed by Ciel's sexual overtures; she simply says no. Straight feminism accepts lesbianism but doesn't see it as significantly different or representing a real alternative. Nor do straight feminists acknowledge the effect of gay life on heterofeminist realities. (Bailin, 1979: 3)

Weill, on the contrary, saw herself as representing lesbianism as a positive choice for feminists; but her articulation of this intention reveals a level of denial. The reference is to a "spectrum" (continuum) of women's relationships which, here, effectively resubordinates female friendship to heterosexual relationships.

When Susan is so upset with Anne's marriage, one can wonder whether she's in love with her [...] The film is about the whole range of experiences a young woman can go through. A love relationship with another woman is one of those real possibilities, but one that Susan doesn't choose. Friendship and love are levels of intensity on a spectrum. (Kinder, 1978: 46)

These pro-feminist films might, on the other hand, be argued to have represented a more positive departure in mainstream film practice not only in terms of their content, but also in form. Kuhn argued that New Women's Cinema brought a self-reflexive approach to the gendered representational regime of Hollywood cinema:

The rise of the second wave of feminism, according to this explanation, brought about a backlash effect: the threat posed by the liberated woman was actually contained in films, often by a literal containment, at the level of story, of female protagonists [...] At the same time, however, since the middle 1970s [...] a number of Hollywood films have been made which may be read as indicating an opposite trend. In these films, [...] Narratives [...] are frequently organised around the process of a woman's self-discovery and growing independence (Kuhn, 1982: 125).

Kuhn based her analysis on the same two films Girlfriends (1977) and Julia (1977), noting that both resist classic narrative closure. She further argued that, particularly in the case of Girlfriends (which was feminist-directed), classic closure would probably be impossible:

[R]esolution in Girlfriends might be brought about by the establishment of love relationships for Anne and Susan: either with each other, or with new partners. Although the first option would fit in well with the structural demands of classic narrative, as well as with the powerful Hollywood "romance" model, its content is excluded by rules, conscious and unconscious, currently governing representations of homosexuality in dominant cinema. But at the same time the second option [...] is also ruled out, in this case by the demands on the narrative set up through the characterisations of the two women: it would simply not be plausible [...] perhaps because it would undermine the 'buddy' structure that governs the organisation of the narrative. (Kuhn, 1982: 127)

But Kuhn's remarks can also be applied to the male buddy movie:

[I]n all these [buddy] films the emotional center, the emotional charge, is in the male/male relationship, which is patently what the films are about. Obvious, of course: yet the fact stands in direct opposition to the usual account of the Classical Hollywood text in terms of the happy ending in heterosexual union, promising the continuance of the nuclear family. (Wood, 1986: 228)

In the male buddy genre, 'the woman' is usually marginal (pure signifier of the heterosexuality of the male subjects) as in Pat Garrett and Billy the Kid (1973), or shared or exchanged homosocially as in Butch Cassidy and the Sundance Kid (1969). The male buddy movie resists the classic closure of marriage and so the homoeroticism of the bond cannot be thus repressed. The only remaining re-normalising resolution is death.

Whilst many lesbian critics expressed pleasure that a more positive image of female independence was being put forward, these films clearly had little appeal for lesbian audiences and were criticised for their formal denials of lesbian eroticism. The female-buddy films did not, either, represent any form of politicised, or even prioritised, "woman-identification" which might constitute any alternative order of lesbian discourse (Rosenthal, 1978; Bailin, 1979). Despite such explicit denials of lesbianism, however, it cannot be denied that the production of subtextual lesbian eroticism cannot be entirely controlled:

Susan's dominance in the film is already clearly established in the opening. In the first sequence we see Anne asleep in the morning light as we hear a strange clicking that awakens her and is revealed to be the sound of Susan's camera. Anne is immediately introduced as the passive object while Susan actively controls the point of view. (Kinder, 1978: 46)

This effectively sets up a relay through which a female spectator is estranged from the (male) position as controller of the image, looking onto the passive feminine. Anne is actually asleep at the time (cannot return the look), intensifying the voyeuristic effect. Somewhat incidentally, this scene also evokes the auditory hallucination of a "clicking" sound referred to in the paranoid fantasy described by Freud (1915) in which a female patient claimed to have been watched and photographed having sex with her lover (in Freud's analysis disguising a homoerotically charged primal fantasy of watching/being watched by the mother). Thus, it doubly raises implications of lesbian eroticism through inversion of gendered positions in the relay, and through the evocation of a polymorphous fantasy scenario. Such analysis remains, however, caught within the dualistic structuration of gender in post-structuralist theorisations of film-text. However, it is important to open out such closed systems if a lesbian appropriation of the text (in addition to a subtextual eroticism inherent to the text itself) is to be figured.

The Color Purple (1985) was the only Hollywood buddy film based on the work of a black feminist writer and with a largely black cast. It was, however, white-produced and directed, and was denounced by male-dominated black organisations in America. The film was seen as fuelling images of black men as violent and animalistic and as divisive in the black community. Particularly resented by men was a perceived "authentication" of racist attitudes towards black men by the participation of Walker as "representative" of black women (Butler, 1991: 63-65). In spite of this, the film was often popular with black women who greeted it as a relatively positive development in popular film. Butler suggested that differences in reception between black men and women may be intrinsic to popular Hollywood representations of black history:

[T]he black male spectator, recognizing the self in the black male image upon the screen, rejects that self as unrealistic, evil, and resists the film's discourse completely. The black female spectator, on the other hand, recognizes herself in the sympathetic black female upon the screen and accepts that self [...] [she] casts a 'blind eye' at the film's brutal depiction of the black man. (Butler, 1991: 65)

Some black female spectators, indeed, responded in a way which recalls white heterosexual women's accounts of Girlfriends: by collapsing text back into history, arguing that black women did suffer such abuse and this should be openly addressed:

One black woman who had seen the film was quoted in the New York Times as saying that she knew many Celies when she was growing up [...] (Bobo, 1988: 44)

Black women's ability to access pleasures in this text may have been partly "authorised" by Alice Walker's participation, but was also effected by mobilising oppositional reading strategies located in African-American history and culture. Bobo refers black women's textually transformative activity to the materiality of black women's discourses:

Paradoxically, the film The Color Purple was constructed according to mainstream values; the meanings embedded in the film are ones that are deeply ingrained in this culture. The struggle to resist the pull of the film, and to extract progressive meanings, is the same struggle needed to resist domination and oppression in everyday life. This battle is not a new one for Black women. Their cultural competency (the repertoire of discursive strategies brought to bear on interpreting a text) stems from growing up Black and female in a society which places little value on their situation. (Bobo, 1993: 285)

Whilst this film did seem to have an appeal for black women, it might be argued that The Color Purple almost entirely suppressed the lesbian presentation of the relationship between Celie and Shug in Walker's original novel. Nataf's suggestion that lesbian readings of Ghost (Nataf, 1995: 71) might have been intertextually informed by reference to The Color Purple does indicate, however, that lesbians might effect a contextualised interpretation drawing not only on Walker's personal identification but on black lesbian discourses more generally.

In spite of the polarisation of political, cultural, and academic accounts of lesbian spectatorial practices, I hope to draw out the ways in which lesbians deploy feminist discourses alongside other modes of interpretation in order to effect an oppositional reading of popular films. Analysis of these modes might be thought to split along a camp axis of cross-gendered signifiers and a feminist axis of female-bonding. I hope to show, instead, that second-wave feminist practices of lesbian spectatorship appear to have absorbed those which I have previously discussed in relation to the figure of the tomboy into the construct of a transpersonal (discursively effected) bond between lesbians. Furthermore, I want to draw out the ways in which the discourses of dominant cinema themselves absorb counter-discourses and thus are also constantly transformed through struggle. As feminist discourse and practice increasingly impacted upon mainstream cinema, in the context of the wider impact of feminist contestation on dominant discourses, lines between inclusion/exclusion in any case became increasingly blurred. In short, I intend to analyse popular cinema in terms of its own significatory system (IMR) but I shall not treat the IMR as a closed system, but as a contested field of discourse within western cultures.