The project of articulating a lesbian position within a feminist critique of film theory has been dominated by Mulvey's (1975) dualistic account of the gendered organisation of cinematic space:
In a world ordered by sexual imbalance, pleasure in looking has been split between active/male and passive/female. The determining male gaze projects its phantasy on to the female figure which is styled accordingly. In their traditional exhibitionist role women are simultaneously looked at and displayed, with their appearance coded for strong visual and erotic impact so that they can be said to connote to-be-looked-at-ness. (Mulvey, 1975: 62 – emphasis hers.)
Various efforts were made to trouble the binarism of this model and to figure an active female spectatorship within its parameters. Among these were Mulvey's own Afterthoughts on transvestite female identifications (1981) and Doane's (1982) masquerade theory which argued that a covert feminine activity could be read in to the text as a troubling of the institutional mode of representation (IMR) in the noir and western genres in particular. Copjec (1982) redeployed Laplanche's (1976) model of anaclisis to figure pre-discursive (pre-Oedipal) 'feminine drives' which might trouble the emphasis on (phallic) narcissism in accounts of cinematic identification. Dyer (1982) and Neale (1983) also addressed the possible objectification of the male body as object of phallic desire in action genres.
By the end of the 1980s there was a period of reassessment. In her (1987) overview, Merck noted that Mulvey (1981) had, herself, found her female spectator in "the awkward position afforded by the binary structure of the 1975 essay" (5). The possibilities for exit afforded variously by Kristevan semiotics, the theorisation of a pre-Oedipal bond with the mother or of Bakhtin's concept of carnival as ritualised transgression had also been explored. Merck identified a range of specifically feminist approaches to the Lacanian problematic of sexual difference including an Irigarayan celebration of an oppositional construct of feminine difference; a multiplication of categories of difference in order to displace the oppressive hierarchy of a single, over-arching dualism (for example by drawing attention to differences between women); and a Derridean criticism of the foundations of such dualism itself in western metaphysics.
The contradictions which feminist approaches to the film-studies model have produced were summed up by Penley:
In examining the model or metaphor of the cinematic apparatus, the most useful and successful feminist approaches have been those that take film theory on its own terms — semiology, psychoanalysis, textual analysis — while questioning the capacity of each to elide the difficulties specific to feminine sexuality, if not gendered subjectivity tout court. Another approach to the same apparatus question in its relation to ideas like the Imaginary, identification, and repetition, would be to reject, out of hand, all the work produced by film theory on the grounds of its manifest exclusion of the woman; and then strike out along the well-worn dissident paths of a reductive biologism, sociologism, or mysticism of the feminine, resurrecting once again the expressiveness of the women's body or 'women's experience' ... (Penley, 1989: 60)
The limits of the feminist-psychoanalytic model of 'sexual difference' seemed to have been reached by the close of the 1980s. This exhaustion was perhaps an element in a shift, at the end of the 1980s, away from a mode of enquiry focusing on the category of femininity (with its problematic drift towards the naturalisation of 'the woman' noted above by Penley) and towards the category of masculinity.
Silverman (1988) deployed Deleuze's (1971) analysis of Freud's (1924) account of the masochistic fantasy scenario to undermine the purported fixity of masculine subjective dominance. If the resignification of femininity as a positive category could not be effected, then the controlling dominance of the masculine gaze might be troubled instead. Rather than being perceived as universal 'law,' masculinity might be decentred or pluralised and differences of race, or sexual orientation, be opened out.
Why should feminists discuss masculinity now? Why should it be given any priority over other pressing issues? ... [I]t is urgent that feminists examine the cultural construction of masculinity, as well as the psychoanalytic structure of male sexuality, in order to avoid reducing either to monolithic and universal categories. (Penley and Willis, 1988: 4-5)
In this context, Clover (1992) challenged de Lauretis' (1984) argument that, in the mythological structure of the horror genre, there are two basic and fixed subject positions: that of a (phallic) mobile, heroic being who penetrates closed spaces and that of an immobile (feminine) being who personifies the damp dark space constituting that which is to be overcome. Clover counter-argued that gender formations in horror and science fiction genres recall an older 'single-sex' construction of gender in which identifications were more plastic and reversible. The scopic regime of the IMR is effectively disrupted in the horror genre since the phallically-controlling (voyeuristic) gaze of the killer is genre-coded as psychologically out of control and as always-already doomed to annihilation. The phallic gaze is thus disempowered and the male spectator re-aligned with the reactive (feminine) moment of looking — suspended in horrorful anticipation of events which he ultimately cannot control.
Feminist re-conceptualisations based on a de-subjectivising Freudian account of the fantasy scenario, or on Laplanche and Pontalis' (1986) re-reading of this account, also appeared to open out possibilities for moving beyond the binary terms of sexual difference which had foreclosed discussion of lesbian spectatorship. For lesbians, the limits of a discourse of sexual difference are marked not only by the subordination of the feminine sign to the masculine, but also by the hetero-gendered structure of desire which inscribes homosexuality as a pathological disavowal of sexual difference. The desiring gaze must be aligned with an active/ male appropriation of a passive/ female spectacle and lesbian desire can thus figure only as active/ male (inverted).
Studlar (1991) and de Lauretis (1991) constructed desubjectivising (non-identificatory) models of lesbian spectatorship in order to retain a psychoanalytic paradigm of gender and yet avoid its inversionary effects. Whilst Studlar (1991) focused on the lesbian-iconic representation of Dietrich in popular film, for de Lauretis (1991), a lesbian interpretation is possible only when film production and spectatorship both take place in a lesbian context — and thus text and context can be treated as continuous. The specificity of lesbianism is constituted under the sign of desire. In other words, the lesbian spectator can only be discerned in the production of a specifically lesbian desire through the workings of the specifically lesbian text read in a specifically lesbian context. Lesbian desire thus reads not as inversionary but as an interpretative pact among women. However, the power to produce lesbian meanings seems vested solely in the textual producers — the consumer seems to be dragged along "by the scruff of the text".
Although de Lauretis (1993) stated that "redefining the conditions of vision, as well as the modes of representing, cannot be predicated on a single, undivided vision ... (whether as 'lesbians' or 'women' or 'people of color' ...)" (152), this recognition seems difficult to implement in terms of a purely textual practice. In the discussion following her presentation of her (1991) analysis of She Must Be Seeing Things, it was pointed out that de Lauretis had ignored racial specificities in the presentation of its protagonists (264). De Lauretis replied that "... I don't think the film allows one to deal with it beyond locating it as a problem ..." (264 – emphasis mine). She added that for this reason her work had focused on the (non-racially-specific) question of fantasy (264). In the (1993) text, de Lauretis again addressed She Must Be Seeing Things, but restated her view that "[t]he originality of [the film] is in its representing the question of lesbian desire ... through the feminist-lesbian debates on sex-radical imagery as a political issue of representation, as well as in real life." (153 – emphasis hers). Although she had noted the specificity of the character Agnes' Latina identification, effectively it still failed to 'signify'. That is, the critic may only deal with issues as the text allows.
In terms of this argument, it is difficult to avoid the conclusion that the lesbian text produced in a white context continues to compel the silence of its racialised 'others' — both as performers and as spectators. Thus, whilst lesbian film-theorists have begun to argue that lesbianism cannot be treated as an undifferentiated category, those models which remain within the textual field of film-theory (which absorbs 'the spectator' to the text) might still be argued to reproduce an exclusionary effect. Furthermore, the constitution of lesbianism under the sign of desire might also be seen as universalising. Despite reference to extra-film-textual lesbian debates (over sexuality) and, whilst 'the lesbian' spectator is no longer characterised merely as an ideal construct of instabilities inherent to the popular film-text and its apparatus, there still seems no way to address issues or identifications which are not addressed by the text itself.
Lesbian identifications might, however, actually resist address through such exclusively textual approaches as that offered by psychoanalytic film theory. Mayne (1991) argued that all spectatorship is inherently contradictory and questioned whether textual analysis of the disruptive potential of lesbianism to the scopic relay should be the primary focus of lesbian criticism. The textual theorist can only point up incoherences in the internal workings of the text and try to 'work' the exposure of its illusions of monolithic power and control. This necessarily effects a largely negative critical practice. "A pessimism about any alternative pleasures has characterized many feminist discussions of the cinema" (Mayne 1993: 70).
De Lauretis (1991), whilst celebrating marginal lesbian performativity, was nevertheless highly critical of the identificatory modes of spectatorship offered by popular representations of lesbianism in mainstream lesbian films such as Desert Hearts. She dismissed male-directed films such as Black Widow or Personal Best, which have been popular with many lesbians (Ellsworth, 1986; Traub, 1991), as "outright obnoxious" (de Lauretis, 1991: 257). A celebratory attitude towards lesbian film which nevertheless articulates a rejection of lesbian readings of the popular text might seem to be challenged by the evidence that many lesbians have generated a culture of pleasure in popular cinema (see, for example, regular reviews of popular film in Diva magazine; Merck 1980; Becker et al 1985; Ellsworth 1986; Case, 1988/9; Traub 1991; Weiss 1992; Tasker 1994; Heller 1994; Hamer and Budge (Eds) 1994; Wilton (Ed) 1995b; Olson 1994, 1997; Patton 1995; Nataf 1995; Halberstam 1995; Modleski 1995/6 — to name but a few). Clearly, these lesbians must have found some way of disentangling themselves from preferred readings or identificatory modes offered by the dominant text.
The film-studies model has tended to rule out any meshing of its conceptual, speculative, textual analysis with historical, sociological or anecdotal accounts of spectatorship. The former disallows the latter since if subjective-identificatory practice is a product of the structure of the text, then there can be no reference to any identificatory processes external to the text: or at least no way to relate them to a specifically textual mode of analysis. Textual theorists also objected that reference to the specificities of extra-textual spectators' identifications (such as lesbianism) tended to posit a mis-placed concreteness (or essence) to such identifications. Most historically based analyses of film, therefore, "tend to be done independently of the kind of apparatus concerns that characterize 1970s film theory" (Mayne 1993: 63). Textual practice is not, however, itself immune from a tendency to reification:
Indeed, some feminist work which takes Mulvey's essay as an inspiration puts forth just as rigidly a master plot of men looking at women from which any possibility of the female subject has been evacuated. (Mayne 1993: 52)
American lesbian critics such as Becker et al (1985) and Ellsworth (1986), who had been less confined by the European (and especially British) mode of textual film criticism, had already taken informal account of anecdotal evidence from lesbian (bums-on-seats rather than theoretically inferred) spectators. More recently, there has also been some general movement in British film studies towards the historicisation of the film text. In feminist work, this has often been effected through conceptualisations of intertextuality: "to refer less to the discursive and self-referential quality of all signification, and more to the ways in which film addresses its viewers across a wide range of texts." (Mayne 1993: 64). This inter-textual approach refers to a wider film culture of popular magazines and advertising as well as to variously-focused accounts of the historical contexts of film production, exhibition, or reception.
While much of the work on spectatorship done in the name of history is extremely critical of that theory, textual analysis has not been rejected but rather revised. For a common point of agreement in studies of intertextuality, exhibition, the cinematic public sphere, and reception is the need not to reject textual analysis, but rather to expand its parameters beyond the individual film text. Textual analysis thus becomes attentive to the intersecting and sometimes contradictory ways in which different forms of address function across different textual registers. (Mayne 1993: 68)
Feminist film theory has not proven satisfactory to lesbian purposes and could also be argued to have reached its limits more generally. In spite of various challenges, however, the psychoanalytic model of gender in film studies has remained crucial to the intelligibility of feminist and lesbian film criticism and practice. At the same time, popular cinema has itself been exhibiting an ever-increasingly self-reflexive mode of address. Awareness of the possibilities for commercial exploitation presented by a diversity of spectatorial practices has been encoded in the diversification of the forms and address of the popular film-text itself. Constructs originally developed within film theory and subsequently transmitted into popular culture by filmmakers who had been through a process of higher education are, by now, very often crucial to the popular intelligibility of dominant texts. The self-motivated femme fatale now lives to fight another day in commercially successful 'post-feminist' films such as A Rage in Harlem or The Last Seduction. In a more limited opening out of address, 'positive image' black male heroes or white female heroines tokenistically save a 'multicultural' world in blockbusters such as Terminator II, or Independence Day.
Idealised 'postmodern,' 'multicultural,' or tokenistic modes of incorporative address may, however, not only be a-historical but actually misrepresent material conditions. For economically marginal communities, and especially for women, these may not have improved or may actually have worsened during the last two decades. Discrepancies between a 'post modern,' or 'multicultural,' representational order (in all its forms of presentation, not just in relation to cinema) and present-day conditions of economic marginality have given rise to some comment, particularly in the UK:
... [T]he lesson of our time is that the distinctive logics of racism can rapidly transport ... neighbours from benevolence and intimacy into a non-negotiable mutual loathing that denies the premises on which the idea of a multi-cultural utopia has so far been constructed. (Gilroy 1993a: 25)
[The political climate] ... has seen a fierce move to the right in the western world in the last 15 years, the balance of forces having shifted firmly against the social and political gains made by women, the black community, and the working class during the 1970s and early 80s ... Poststructuralism has played a major part in the backlash . . . (leading) to a situation whereby any materialist analysis is viewed with suspicion. It has become extremely unfashionable for anyone to claim that material oppression, injustice and inequality exist in the world as structures ... (Wingfield 1994: 21-24)
[A post-modern discourse] is gender depoliticised, sanitised and something difficult to associate with sexual violence, economic inequality, women dying from backstreet abortions. It is gender reinvented as play for those who see themselves far removed from the nitty gritty of women's oppression. (Jeffreys, 1994: 98)
This [culturally-focused] writing suggests the far distance between contemporary consumer culture and the world of long hours, unrewarding work, drudgery and brutal exploitation ... (McRobbie 1997: 76)
Whilst it is clearly crucial that some means of historicising accounts of gendering in cinematic processes be effected, Mayne suggested that it would hardly improve matters to replace the reified dominance of the male gaze by any new "empiricist orthodoxy;" since "history" can also acquire the status of "unquestionable evidence" (1993: 64). A Foucauldian model of discourse, on the other hand, does not posit any area of human consciousness, fantasy, or impulse as a priori to the production of signs. Foucauldian method, which holds textuality (or discourse) in tension with a contingent account of material conditions, may thus offer a solution.
Foucauldian approaches do not currently offer any developed conceptualisation of the specificities of cinema as medium or apparatus, however. Feminist film theory, it would seem, continues to provide the only developed language in which to specify the particular ordering of gender articulated by the cinematic apparatus. Psychoanalytic accounts of representation are obviously not, on the face of it, compatible with Foucauldian (anti-psychoanalytic and post-gendered) perspectives. But in Foucauldian terms, all conceptualisations must recognise the necessary incoherences of any kind of account of 'reality.' By re-articulating models of the apparatus made available by the work of feminist film-theoreticians into a discursive mode, it becomes possible to approach the cinematic process not as the hopelessly monolithic process of the reproduction of women's subordination, but as a site of contestation and transformation.
This is necessary if a lesbian 'history of the present' is to be constructed as an effective map of continued resistance.
Without some kind of extra-textual or inter-textual contextualisation it is difficult to avoid not only the negative positioning of the lesbian spectator by heterosexualising feminist accounts of the scopic relay; but also the occlusion of the specificities of lesbian oppositional reading practices themselves. Black feminists working within a film-studies paradigm (who had also been marginalised by the dualistic construct of gender in film-theory) seemed to be moving in similar directions. Bobo (1988), for example, cited black women's deployment of a "cultural competency" which she defined as "the repertoire of discursive strategies brought to bear on interpreting a text" which had developed through the processes of black history and culture (Bobo, 1993: 285). These politicised cultural competencies enabled black female spectators to recontextualise, or resignify popular films such as The Color Purple which had, elsewhere, been rejected by black critics as negative portrayals of black communities. Whilst it might legitimately be taken for granted that lesbians do, in fact, also deploy such (sub)cultural competencies in reading the popular text, any account of lesbian spectatorship which does not explore the discursive construction of these lesbianisms tends to depoliticise. Lesbianism has too often been articulated in film-theory as signifying an alternative content of desire rather than as effecting a politicised resistance. Interdisciplinary lesbian work on popular culture has more recently begun to take account of lesbian diversity, however:
While tracing this history of the exclusion of images of lesbians and gay men from film has been an important liberationist project, it is important to recognize that lesbians and gay men have in no way been coherent categories. (Patton, 1995: 22)
How may we adequately define a lesbian film, a lesbian film-maker, a lesbian spectator ? Of course, unless we resort to essentialism we are forced to conclude that these things are contingent ... (Wilton, 1995c: 4)
Women of the baby-boom generation ... believed that they could construct a collective sense of what it meant to be a lesbian, and also develop representation of that collective identity. Today's emergent generation, much more aware of the limitations of identity politics, seemingly does not. (Stein, 1994: 26)
In referring conceptualisations from the textuality of feminist film theory to a lesbian (sub)cultural context one is effectively re-referred to the problem of defining 'the lesbian' in the first place. It seems inevitable that, whilst formally acknowledging diversity, lesbian cultural critique has thus almost surreptitiously referred its otherwise inarticulable speculations either to the specificities of the writer's own lesbian 'knowledges' or to an amorphous construct such as lesbian 'cultural competence.' The citation of lesbian 'memory' is not, however, inconsistent with Foucauldian practice — which informs much of the more recent lesbian work (as well as my own). Such work does not seek to establish historical 'facts' by empirical methods, but to draw out, or render more transparent, the cultural conceptualisations which are productive in every-day lesbian strategies and identifications:
In the second volume of his last major work, The History of Sexuality, (Foucault) describes the object of the studies there as being "to learn to what extent the effort to think one's own history can free thought from what it silently thinks, and so enable it to think differently." (Hamilton 1996: 140; citing Foucault, 1992: II, 9)
Any critical concepts, nevertheless, always need to be situated by what Foucault calls "the historical conditions which motivate our conceptualisation ... (and) the type of reality with which we are dealing" (Foucault 1982: 209). But to Foucault, historical narrative does not constitute a 'factual' or 'authentic' account of the past or the present. It is no more or less unreliable than any other.
History has no meaning though this is not to say that it is absurd or incoherent. On the contrary, it is intelligible ... but this is in accordance with the intelligibility of struggles, of strategies and tactics. (Foucault, 1980: 115)
Foucault advocates using the historical sense to "construct a countermemory — a transformation of history into a totally different form of time" (Foucault 1984: 93) ... Superficially, this 'countermemory' sounds like an exercise in redemption, a remembering of all the voices conventionally dismissed by history ... But Foucault believes that a catastrophic unconnectedness is exactly the authentic state of affairs, past and present. (Hamilton 1996: 138-9)
Although working basically within a film-studies paradigm, Traub (1991) nevertheless rejected any universalising categorisation of 'the lesbian' constituted under the sign of desire.
Whatever a 'lesbian' 'is' is constantly negotiated — a matter of conflicting and contradictory investments and agendas, desires and wills. (Traub, 1991: 305)
Traub's interpretation of Black Widow (1986) implicated a (quasi-historicised) lesbian spectator in the institutionalised organisation of desire in the popular film-text by requiring her to supply 'lesbian knowledges' in excess of those offered by the film's narrative but which were opened out by its incoherences and necessary in order to effect a (lesbian) reading. 'The lesbian' nevertheless remained more or less a prisoner of the naturalising effects of the cinematic process:
'Lesbian' appropriation of the 'gaze' comes only at the price of acquiescence to a system of sexual (gender and erotic) regularization that reproduces dominant taxonomies of sexual (gender and erotic) difference. (Traub, 1991: 308)
... [T]he majority of 'lesbians' I know who saw the film defied spectator conventions, constructing Alex as an erotic object to suit their desires. At the same time, their construction enacts precisely the conflation of gender and eroticism endorsed by the film and by which the dominant ideology homogenizes 'lesbian' desire as 'narcissistic': identification with Alex leads to desire for Alex. (Traub, 1991: 308)
Butler's (1993) Foucauldian re-reading of Lacan's primary account of narcissism, however, offers an opportunity to reposition the Lacanian account of the cinematic process itself within a historicising frame. Lacan's account of the 'mirror phase' (and his pupil, Miller's (1977-8), account of 'suture') are crucial to the film-studies model of the (gendered) identifications offered by the visual relay. If Lacan's originary model of the acquisition of the bodily imago upon which the ego is articulated can be treated, as Butler has done, as an account of the materialisation of "a sedimented history of imaginary relations" (Butler, 1993: 74); then there seems no reason why the psychoanalytic account of the replication in cinematic processes of identification and desire may not be treated as similarly contingent.
Psychoanalytic film theory need not be approached, as it generally has been, as describing the processes by which the cinematic apparatus reproduces the (culturally) gendered organisation of (essentialised) 'drives.' Models of identificatory practice might, instead, be treated as accounts of discursively effected institutionalised relations which are repetitively cited in the (re)production (or refusal) of the gendering of spectatorship. The contingency of cinematic narrative coding has, indeed, long been recognised:
Narrative and image in film are never entirely reducible to one another, if only because the demands of the classic narrative could in fact be met by a range of conventions of cinematic narration, of which the classic system is but one. Conventions, by their nature, are subject to change. Even if the classic narrative retains its dominance as a structure, its basic requirements could conceivably be met by cinematic codes different from those of classic cinema. (Cook, 1985: 215 – emphasis hers.)
The gendered organisation of 'the gaze' might be read as subject to exactly the same productive incoherences and contestations as any discursive construct or cultural code.
[T]he theoretical material on 'the gaze' also fails frequently to distinguish between the look (associated with the eye) and the gaze (associated with the phallus). (Evans and Gamman, 1995: 16)
The (white, male) 'look' might strive to replicate the seamless dominance of the 'gaze' but since the 'gaze' is, in 'reality,' merely a cultural ideal, appropriations of it can never achieve a 'perfect' dominance — that is, no matter how he tries, the male individual can never entirely exclude or annihilate his own 'lack of fit' or the resistances of 'others' to his acts of dominance. Women and lesbians, on the other hand, would not only be debarred from such performances of dominance, but would not necessarily have any interest in attempting to replicate the 'gaze' as dominance. On the contrary, they might very well be motivated, instead, to resistance.
The conventions which render visual-spatial organisation intelligible are, arguably, produced within the same discursive formations as the conventions of any culturally intelligible representation. This is increasingly being recognised in feminist film studies:
We would also argue that the 'cultural competence' of the lesbian spectator (and lack of such competence in other viewers) may influence the way representations are viewed and understood by some women. Using Foucault's model of discourse we would argue against any essentialist model of the lesbian gaze and instead suggest that lesbian viewers may bring certain subcultural experiences and knowledge to the reading of specific texts. (Evans and Gamman, 1995: 35)
There is often a danger, however, for the dominant text to become entirely relativised.
But does the queer gaze always reconstitute the visual text as queer? Or do some images encourage polymorphous identifications more than others? As we argued ... context is important, but the text also is a structuring discourse. (Evans and Gamman, 1995: 45)
It is important to retain a sense of mainstream modes of representation as effecting dominance, and of lesbian reading as representing a challenge — rather than simply as 'diversity' with minimal political valence. If adequate attention is not paid to the processes of popular cinema as a mode of dominance, there can be a tendency to idealise constructs such as 'contradiction' in feminist discourse or 'transgression' in queer discourse — to the point at which they appear as ends in themselves:
For what, after all, is transgression? Transgression has a slightly Sadeian ring, almost as if a notion of blasphemy were lurking somewhere about. In secular terms, what is implied is a flouting of the rules, or a rule, behaviour antagonistic to what is established, the opposite, a radical challenge to what is prescribed. Yet, just as the only true blasphemer is the individual who really believes in God, so transgression depends on, and may even reinforce, conventional understandings of what it is that is to be transgressed. (Wilson, 1993: 109)
My own analysis attempts a critical reading project in which close readings of film-texts are effected by citing lesbian contexts which are, in turn, delineated by a discursive treatment of lesbians' own competing political, cultural, and historical narratives. I want, however, also to try to articulate the productive effects of lesbianisms' and feminisms' own discursive conflicts and struggles — between lesbians themselves as well as between feminisms, lesbianisms, and the dominant formations of popular cinema. That is, in order to historicise the text, as well as the spectator, I have set out to model lesbian (re)readings of dominant cinema by reference to the same discursive struggles which organise the formations of feminism and lesbianism as modes of resistance, and which thus make it possible also narrate a (dominant) context for the films themselves (which may, increasingly, be reflexive not only to cultural contestation and change but to critical discourses in film theory itself), but without being bound over to its modes of domination. That is, it becomes possible to articulate the relationship of lesbian subjectivities to dominant cinema as a politicised process. My own mode of analysis thus takes into account 'preferred' readings offered by popular films.
The intelligibility of popular film obviously depends on readers' familiarity with the dominant cultural codes which it deploys and which its readers re-deploy. In Foucauldian terms, these conflicts are ordered by complex and culturally produced motivations in the (re)production or contestation of power relations. Any spectators who can read conventional cinematic codes, including marginalised spectators, thus have a range of compliant or oppositional reading strategies made available to them. That is, in Foucauldian terms, coercive practices must inevitably refer themselves (metonymically) to the possibility of motivation — either to participate or to resist (for if we were robotically servile coercion would merely signify as excess). Those who are motivated to claim dominance (a will to access power or court its approval) might thus attempt an appropriative or compliant reading of popular texts; whilst those motivated to resist might attempt a disruption or articulate a downright refusal. It therefore seems important to articulate lesbian motivations produced in discursive struggle rather than any fixed or universal identity or interpretative strategy.
I do not treat lesbian identity as a meaningful essence in itself but rather as motivation to resist the processes of women's subordination which finds a variety of modes of articulation. Motivations to resistance may engage, as well or instead, other discourses such as race, sexuality, or class. Furthermore, modes of resistance change over time. Resistance is, above all, a motivated (politicising) activity, and not a passively inevitable product of (lesbian) 'nature.'
I would like to suggest another way to go further towards a new economy of power relations, a way which is more empirical, more directly related to our present situation, and which implies more relations between theory and practice. It consists of taking the forms of resistance against different forms of power as a starting point. To use another metaphor, it consists of using this resistance as a chemical catalyst so as to bring to light power relations, locate their position, find out their point of application and the methods used. Rather than analyzing power from the point of view of its internal rationality, it consists of analyzing power relations through the antagonism of strategies. (Foucault, 1982: 211)
This model of resistance has often been subjected to a binaristic interpretation in Anglo-American literatures — that is, the axis of enquiry articulates a dialectical relation between 'dominant' and 'abject.' Foucault expressly rejected the pessimism of such articulations in which resistance appears as "only a reaction or rebound, forming with respect to the basic domination an underside that is in the end always passive, doomed to perpetual defeat" (Foucault, 1976: 96). A micropolitical analysis must, on the contrary, take account of the production of resistance as not only incoherent but also active.
Lesbian discursive (sub)formations can be discerned as much through the manifestation of lesbian struggles over the meaning of lesbianism as through the struggles of lesbians against 'the dominant'— an equally incoherent process. A discursive method offers a means to posit oppositional discursive strategies which work their reversals within the cultural sphere of the dominant but which come, gradually, to constitute relatively autonomous 'sub-domains,' or formations (shared modes of language use and reference), which are capable of sustaining materialised (apparently substantive) counter-identifications. Whilst any illusion of coherence produced as an effect of such identificatory practices cannot be taken as a lesbian 'authenticity,' they can provide a point for the re-transmission of cultural codes. In other words, they enable a diversity of lesbians to read mainstream culture from lesbian perspectives whilst, from a theoretical point of view, offering no universalising 'truth' of lesbianism.
The incoherence of the discursive productivity of power posited in Foucault's work can, therefore, be used to account for a lesbian ability to effect re-readings of dominant cultural products which are incoherent with preferred readings — and, indeed, with each other — yet intelligible as transformative modes of resistance. In effect, the relation between spectator and text is a dynamic one — the spectator is not the passive victim of textuality (as in psychoanalytic film theory) but then nor is the text a passive object which is entirely relativised by the specificities of the gaze (as in much queer theory). Each represents an attempt to control the production of meanings in the discursive field of popular cinema. I will thus be addressing questions of lesbian identity to a diversity of lesbian discourses, as articulated in politics, literatures, anecdotal accounts and media.
This project, however, is in no way intended to constitute an exhaustive representation of all formations of lesbian culture, identity, or "community" (an unwieldy and impossible task) — nor, indeed, a comprehensive history of cinematic practice. Instead, I have tried to develop a discursive analysis of feminist engagement with a popular genre and to distinguish some basic modes of lesbian spectatorship. The lesbian-contextualising material which precedes my discussions of lesbian readings of specific film texts is intended to sustain lesbian-ising readings of cinematic texts not by empirical reference to extant communities of lesbians, but by reference to the conflicting "agendas, desires, and wills," which have mobilised lesbian practices. Although I have effected a schematic distinction between lesbian modes of reading cinema in a roughly chronological development, clearly lesbian practices do not confine themselves neatly to period. Metaphors and practices such as 'woman-identification' or crossdressing may persist and yet transform over time. Furthermore, oppositional strategies do not always take the form of counter-identifications and, increasingly throughout this decade, they do not do so. Many lesbians now see a proliferation of 'perversity' or 'irrationality' as the cutting edge of change, expanding what might be admitted to a 'pluralistic democracy' (Butler 1990).
Lesbian representation has, indeed, recently achieved a degree of marketability in the mainstream representational order which has, itself, fragmented. Hollywood now markets popular films which represent lesbianism as attractive, glamorous, and desirable. Producers of mainstream cinema are now able to recirculate lesbian subcultural codings; and the lesbian 'knowledges' brought to decoding are culled from a bricolage of lesbian cultural-historical practices. However, this very degree of incorporation and fragmentation problematises conceptualisations of 'the lesbian' not only as the unified figure of identity politics but also as a queer figure constituted through exclusion from dominant modes of representation.
At the same time, I have addressed what I see as some of the limitations of Foucauldian analyses of lesbian cultures, particularly in the field of queer discourse. I have therefore included a critique of Foucault's occlusion of gender in his formulation of the discourse of 'sexuality' as well as an evaluation of lesbian debates surrounding the influence of Foucault on queer discourse.
I have referred my own assumptions and conceptualisations of lesbian reading strategies to the context of a more general body of American, European, and British lesbian political debates, histories, literatures, and media through which key struggles over lesbian identity, sexuality, and representation have been articulated. In dealing with the relevance of specific film-texts to lesbian sub-cultures, I have referred to lesbian and gay overviews of popular cinema in which particular popular films are mentioned again and again as being of interest to lesbians (see, among others, Russo, 1981; Dickens, 1982; Sulter, 1985; Becker et al, 1985; Whitaker, 1985; Boffin and Fraser (Eds), 1991; Weiss, 1991; Weiss, 1992; Bell-Metereau, 1993; Gever et al, 1993; Johnson, 1993; Hart, 1994; Olson, 1994; Hamer and Budge (Eds) 1994; Arthurs, 1995; Wilton (Ed), 1995b; Evans and Gamman, 1995; Williams, 1996).
Finally, and in conformity with a Foucauldian method, I will bring my own authorship into discourse. I write as a white, educationally privileged, lesbian. I am trained in philosophy, literary/cultural and film critical theory: specifically in the critical application of theoretical models and in close reading. My academic context is dominated by Foucauldian queer theory as it is constituted in the cultural studies discipline in which I am currently based. My own lesbian 'counter-memory' is drawn from personal involvement in London lesbian culture and politics, as well as my work in reviewing, researching, teaching, curating, and programming lesbian film. In these contexts, I am grateful for all the illuminating insights which I have received over the years from lesbians as activists, filmmakers, curators, journalists, academics, students and spectators which has been so influential in shaping my work.