I. Film Studies and Melodrama

That cinema has suffered a loss of autonomy over the past twenty years is hardly in doubt. As movie theatres have become showcases for story and spectacle values equally exploitable elsewhere in the distribution/exhibition circuit, the films, too, seem to have lost their textual closure in favour of other aggregate states - material and discursive – of the commodity cinema. Control of production having passed to multinational commercial interests and media conglomerates, a movie undergoes several metamorphoses, from record to paperback, from object of an advertising tie-in to topic of a TV programme or university exam, as it makes its way through the audio-visual and print environment.

Yet, during the same period the cinema also gained a new specificity as a theoretical object. Semiology, formalism, psychoanalysis were mobilised in order to come to grips with the cinema's status as a distinct mode of signification. In turn, literary criticism, photography, art history, women's studies and cultural studies now show the influence film studies has had, for instance, on reformulating problems of realism, revitalising the study of narrative and narration, and posing the question of subjectivity and sexual difference.

Whether this intensity of the cinema's theorisation already contains a keen knowledge of an irrevocable historical loss is a moot point, but the paradox is worth stating – if only because the most interesting work done in film studies today seems hyperconscious of the need to engage with arguments emerging from already constituted debates and, at the same time, generate from these issues insights that either are not specific to film, or help to transcend the horizon (historical, theoretical) set by the cinema as traditionally understood.

The study of melodrama is a case in point. Retrospectively, its function has been to act as a relay for a series of displacements within the critical discourses on film, opening a space for negotiating a number of changes: not so much in film history or film form, but in the shifting relevance of the Hollywood cinema to critical theory and cultural practice. Initially, in the promotion of Sirk and Minnelli over Hitchcock, Ford and Hawks, the term melodrama wanted to make the auteur theory readable for questions of ideology and gender, while building on mise-en-scène criticism for a more socially informed history of Hollywood style.1 Melodrama also helped to think the opposition author versus genre (terms held together by a production-oriented logic of self-expression and ideological self-reproduction) from the other side, as it were, that of an audience-oriented model of the efficacy of the Hollywood text, and its articulation of the peculiar economy inherent in the viewing situation.2

Subsequently, similar displacements allowed questions of identification, of distance and affect, irony and pathos, of transgression and excess, of gendered spectatorship, the circulation of knowledge and narrational authority, domesticity and consumption, of cinematic duration, televisual flow and narrative closure to be convincingly argued around melodrama and the melodramatic imagination. Nowhere did the notion of the film text as a network of intersubjective relations linking spectator and screen, and therefore available for a symptomatic reading in the manner of a psychic event, seem more plausible and productive than in the melodramas and woman's films of the 1930s to the '50s.

If some commentators detected in these extensions of melodrama conceptual muddles and critical solecisms, complaining about the lack of precision and intoning a 'requiem for a phantom genre'3, the term has clearly survived its various rewritings, and now connotes a distinct 'mode' among Western systems of representation.4 It also seems to work as a genre within the different taxonomies of periodisation, subgenre and cycle developed for the Hollywood output and its general mode of production. Thus, Mary Ann Doane's The Desire to Desire5 takes 'The Woman's Film of the 1940s' as a sufficiently distinct group for a historical and textual reading of several cycles which over the years have also been discussed as melodramas, weepies, or belonging to that other 'phantom genre', film noir.6

II. Melodrama: the Floating Signifier of Feminist Film Theory?

One wonders, though, whether melodrama would have enjoyed quite such critical prestige if it did not figure – in film studies and cultural theory – as a floating signifier, a kind of shifter, or at least a marker staking out a theoretical and even political terrain for which melodrama – at times metonymically standing for the Hollywood cinema as a whole – is itself merely a symptom: the terrain is of course that of sexual difference, the outer envelope, as it were, of the past decade's debates.

Doane is very aware of this conjuncture: her articles in Screen, enclitic, Wide Angle, Cinetracts, Poetics Today as well as in Re-vision, The Cinematic Apparatus and Cinema and Language rarely address the question of melodrama as such, but discussing irony and pathos, narrative authority and the representation of women, the concept of masquerade, subject-object relations and identification, she maps out a similar critical space with respect to both gender and the cinema's enunciative apparatus familiar from the debates about melodrama.

The Desire to Desire expands and reworks the themes of many of the essays. It reads its four sub-groups of the woman's film (the female patient films, the maternal melodramas, the stories of impossible love and the paranoia films) across a double axis of interrogation. The first is historical: 'Hollywood women's films of the 1940s document a crisis in subjectivity around the figure of the woman - although it is not always clear whose subjectivity is at stake' (p 4). The second set of questions, though, is theoretical and meta-critical:

[M]y interest in the films is primarily inspired by certain issues and theoretical blockages in contemporary feminist work. The insistence of their address and the forcefulness of their tropes make the woman's films of the 1940s an appropriate textual field for the investigation of issues surrounding the concepts of subjectivity and spectatorship and the ability or inability of feminist theorists to align these concepts with sexual specificity. (p 5)

These programmatic statements come from the opening chapter, as concisely and densely argued a piece of writing as anything one may hope to come across in either film studies or feminist theory. The Desire to Desire is an immensely intelligent book, but it is also a difficult one, at times almost as difficult to read as it must have been to write. The difficulties stem from the scope it sets itself, or rather, from Doane's awareness of arguing simultaneously from within and without several already very complex and contested debates, while firmly keeping to her historical and generic brief: re-reading the woman's film of the 1940s. By the same token, it is not an easy book to review, if one wants to do justice to the critical perspectives and deconstructive projects making up the conjunction of feminist theory/film theory which the book wants to address and advance.

While Metz, Baudry, Bellour and Heath had largely concentrated on questions of identification, subject construction and the subject effect as the conditions of the classical Hollywood cinema's legibility and textual coherence, feminist film theory, it will be remembered, began with the realisation that the cinema was pre-eminently a system of representation implicated and complicit in the production of sexual difference. In the words of Jacqueline Rose: 'In this debate (about sexuality and representation), the cinematic image is taken as both the model of and term for a process of representation through which sexual difference is constructed and maintained.'7 At least since Laura Mulvey's article, published in 19758, much of film theory addressing the task of deconstructing the terms of this complicity has focused on the possibility or impossibility of female spectatorship.

The crux that emerged from 'Visual Pleasure and Narrative Cinema' is too well-known to need repeating here in detail. But briefly, if Mulvey's argument is correct about the functioning of narrative and spectacle in classical cinema then the subject addressed by the Hollywood text is gendered and male by virtue not only of the dominant forms of visual pleasure – voyeurism and fetishism, traditionally analysed as male perversions – but also by the close formal convergence of narrative progress (the desire to see and to know) with the filmic process itself and the working of the cinematic apparatus. As Doane puts it, reviewing the debate: 'With respect to a narrativization of the woman, the apparatus strains; but the transformation of the woman into spectacle is easy. Through her forced affinity with the iconic, imagistic aspects of cinema, the woman is constituted as a resistance or impedance to narrativization' (pp 5-6). The conditions of female spectatorship in narrative cinema are therefore either submission to the regime of the gaze, the perversion of a perversion so to speak, or they involve taking pleasure in the subject effects of a sexual identity not her own. 'Confronted with the classical Hollywood text with a male address, the female spectator has basically two modes of entry: a narcissistic identification with the female figure as spectacle, and a "transvestite" identification with the active male hero in his mastery' (p 9). The very concept of female spectatorship thus names a theoretical impossibility, and in practice produces 'a mixed sexual body, ... a hermaphrodite.' While the institution cinema has always courted women as an audience, its textual system works to deprive the woman 'of a gaze, ... of subjectivity and repeatedly [transforms her] into the object of a masculine scopophilic desire' (p 2).

Throughout the opening chapter, The Desire to Desire succinctly and lucidly restates the terms of these impossibilities. Yet Doane's intervention is not another summary coming after so many commentaries and revisions of Mulvey's theses.9 Nor is it exactly a matter of testing them against concrete examples, even if the time frame and generic focus at first suggest it. Rather, her move is, as it were, to exacerbate the issue of female spectatorship by choosing movies explicitly made to address women, at a moment in history when women's presence in the labour force gave their presence on the screen a different ideological urgency. This choice has at least four consequences for Doane's readings.

1. It entails an explicit refutation of a nostalgic or revisionist project: 'There is an extremely strong temptation to find in these films a viable alternative to the unrelenting objectification and oppression of the figure of the woman in mainstream Hollywood cinema. … Yet, the woman's film does not provide us with an access to a pure and authentic female subjectivity, much as we might like it to do' (p 4).

2. The focus on films with a female addressee allows Doane to operate a sort of reductionism on the multiple subject positions10 suggested to cope with Mulvey's theoretical crux of the narrative cinema's male gaze overriding representations of women, even where they are depicted as active or where the fiction privileges their point of view. According to Doane, in the woman's film the textual system and the mode of address work in complex ways to de-eroticise the look, despecularise and desexualise the female body, and substitute masochistic fantasy for sexuality:

Because the woman's film purportedly directs itself to a female audience, because it pretends to offer the female spectator an identity other than that of the active male hero, it deflects energy away from the second ‘transvestite' option. … But since the woman's film reduces the specularizable nature of the female body, this first option of a narcissistic identification is problematized as well. In a patriarchal society, to desexualize the female body is ultimately to deny its very existence. (p 19)

3. By addressing the question of female spectatorship within the terms of masochistic fantasy, Doane is able to offer a theoretically coherent account of the mobility of identification involved in the cinematic spectacle but also begin to historicise the dominance given to specularity and vision in recent theories of cinema, subjectivity and desire. If masochism requires spectatorship as a position or role within the fantasy scenario, it nonetheless, according to Freud, functions for the woman by assigning spectatorship in a context where looking is already desexualised and 'disembodied'. Here 'A child is being beaten' is the fantasy that takes on paradigmatic significance.

4. Doane's choice of the woman's film of the 1940s raises the question of the commodity and consumption, in its relation to the cinema as an institution, to femininity as historically constructed, and to subjectivity as constituted by desire.

III. The Desire to Desire: A Meta-Theoretical Treatise

What makes these various moves less straightforward than I have presented them here is the meta-critical or meta-theoretical dimension of The Desire to Desire. Doane's critique of classical cinema regarding the representation of female subjectivity runs parallel with a critique of psychoanalysis. 'Reading Freud is often as strangely compelling as watching a woman's film – both entail the simultaneously pleasurable and unpleasurable effect of recognition/misrecognition of one's own cultural positioning' (p 21). This observation defines the overall dialectic of her study, namely to trace the coincidence of cinematic scenarios and psychoanalytic scenarios of female subjectivity: both 'are symptomatic of a more generalizable cultural repression of the feminine' (p 21). In this respect the critical object of her book is neither melodrama nor the woman's film, but 'the cinema', and by extension, those theorisations of the cinema which have relied on Lacanian theory and psychoanalytic semiotics.

Consequently, traversing the book is an internal argument with feminist film theory, and the extent to which its proponents11 rely on or redefine the Lacanian model. By narrowing her focus to the woman's films, and their flawed, problematic and highly ideological construction of a female subjectivity, Doane may be able to bring extra-textual determinants to bear on the issue, and also to simplify the paradox of women taking spectatonal pleasure in films entirely aligned with male subjectivity. Yet choosing films with a female addressee only makes her more conscious of the theoretical difficulties of relying on psychoanalysis for her key concepts. She constantly reminds the reader that the problem is not cinematic spectatorship alone, nor is the aim the assertion of some illusory feminist identity. Instead, across the woman's film what needs investigating is 'the representation of female subjectivity or its failure in a variety of discourses – film, psychoanalysis, literature, law' (p 9).

Doane's own displacement of the psychoanalytic paradigm appears initially to take the direction of Foucault's critical positions about the complicity of psychoanalysis with other nineteenth century discourses (echoing Stephen Heath's points about the novelistic and cinema both supporting specifically historic constructions of subjectivity). The Desire to Desire, in this perspective, is not 'the definitive psychoanalytical account of the repression of woman in Hollywood cinema', that Tania Modleski praises in the blurb, but a demonstration of the woman's film's (and by extension, of the Hollywood cinema's) capacity to produce feminine subjectivity. At once rewriting the woman's body (destroying it as image, rearticulating it as a site of symptoms and illnesses) and reformulating sexual desire as masochistic fantasy in scenarios of persecution, suffering and self-sacrifice, the woman's film does not repress woman, but supports, in often contradictory ways, an overdetermined production of the feminine.

IV. Female Spectatorship: A Gaze without a Body?

Chapters on 'The Medical Discourse' (Cat People, Possessed, The Locket, Beyond the Forest), on 'Pathos and the Maternal' (Stella Dallas, To Each His Own, The Reckless Moment), on 'The Love Story' (Humoresque, Deception, Letter from an Unknown Woman, Love Letters), on 'Paranoia and the Specular' (Secret Beyond the Door, The Spiral Staircase, Suspicion), systematically and in detail investigate the textual and narrative form this production takes. Finally, Doane reads Caught and Rebecca as representations of the feminine caught in and defined by the 'machines of projection'. Although Doane herself is at times ambiguous on this point ('cinema and psychoanalysis partake in the cultural repression of the feminine') it seems that what is at stake (and what must have determined her choice of the woman's film) are not so much the classical cinema's strategies of exclusion (as typified by the films discussed, for instance, by Mulvey), but figurations of the feminine across a number of culturally coded discourses, including that of the cinematic apparatus itself. Here Doane's object is to describe how a male cinema imagines the functioning of female subjectivity, across a mise-en-scène of (psychic) scenarios of female spectatorship, as if imagining the woman as spectator not only constructed a viable mode of address for both sexes, but by the same token neutralised the power of that spectatorship within the diegesis (in terms of its potential for agency and control) by ensuring that the woman's ability to look is cancelled by her body becoming the problem (the site of disorders). As Doane puts it: 'a bodyless woman cannot see' (p 20).

Doane's close textual analyses are organised around a number of key notions and fields, for example, not only the body, but time (repetition, waiting as the typical temporality of the woman's film, structurally related to agonising choices and all-or-nothing affectivity as the signs of a troubled relationship to narrative and agency); pathos (aligned with mis-timings, deferral, the hope for a reversibility of time, and the gap between knowledge and power). Such concepts, situated at a level where they can have a purchase on the individual texts as well as on the cycle and subgroup as a whole, are clearly also chosen for their relevance to larger debates about narrative, femininity, and systems of representation. In particular, they emphasise the instability of subject positions entailed by the blocking of object choice and its conversion into identification with the object, typical of the narcissistic scenarios offered to the female characters in the woman's film.

These individual readings deserve (and hopefully get) detailed commentary from reviewers more competent or at least more directly engaged in the issues of feminist theory. In this respect, Doane seems careful to inscribe two kinds of readers, or rather, she comes to her object from two directions at once: the woman's film seen as symptom of a generalisable truth about the cultural and social discourses concerned with situating, placing and thus keeping in its place, certain definitions of female subjectivity. Conversely, the woman's film also provides textual examples of an insufficient theorisation of spectatorship in the cinema. For Doane's arguments raise the question (a question which the notion of visual pleasure may have answered too categorically) central to film semiotics: 'to understand how films are understood'12, in the sense of how films construct a space of intelligibility as well as pleasure, but also how subject positions not experienced as pleasurable can be made consumable.

Doane follows psychoanalytic film theory in seeing cinematic spectatorship as a privileged figuration of human subjectivity, the mise-en-scène of the subject's quest for coherence across a set of identifications in which the other as image plays a decisive role. In the terms of Mulvey's argument (or Lacan's mirror phase), fetishism and an over-investment in the visual ensure both distance and identity. However, Doane recognises that even where this coherence-effect becomes problematic, as it does in the woman's film, which creates highly unstable and often unpleasurable subject positions, the spectator is captivated, fascinated, and bound to the representations.

The woman's film, therefore, stands for a question that has in the past been underlying discussions of melodrama generally: namely, why are subject positions emotionally moving which, as it were, dramatise the impossibility of even the fiction of specular coherence and dwell on incomplete entry into the symbolic order? In other words, the Mulvey paradox could be turned round: why does the woman's film appeal, beyond its historical and generic addressee, to both sexes?

V. Stolen Gestures

There is no reason why Doane's study should take up the issue in this form, but I think it is possible to use The Desire to Desire to continue a line of inquiry also implicit in Steve Neale's essay on the melodrama.13 Two directions offer themselves: one is to read Doane's analyses as suggesting, despite her warning against seeing the woman's film as radical or transgressive, the possibility of re-negotiating the subject's (imaginary) relation to the symbolic. What appears in one respect as the oppressively obvious construction of an all too familiar femininity (signalled by the destabilisations of the image, the confusion of subject-object divisions, and the alignment of sight and hearing with crises of perceptual mastery) might be interpreted as valorising less rigid or fixated forms of imaginary investment than the object choices typical of 'male' genres. Doane wants to chart a difficult path between Mulvey (who, within the logic of her argument about a textually constructed spectator, was asking for a cinema practice that would forego pleasure and thus captivation by the textual play of identification and positionality), and writers like Janice Radway14, who pin their hopes on the possibility of actual, historical readers/spectators appropriating the texts for their own 'more positive and more empowering meanings' (p 180). The reason why the woman's film is appealing to women, according to Doane, is that the genre (like other 'popular' texts of mass culture) selects and in effect 'steals' women's gestures and desires. Stylising them, and returning them to the spectator in a more perfect form, the films create the effect of recognition and of estrangement simultaneously: a rhetorical strategy as aesthetically powerful as it is ideologically problematic. Doane, following Barthes, advocates a 'double mimesis' or mimicry, as a way of devaluing the currency of the woman's film's notions of femininity. 'What is needed is a means of making these gestures and poses fantastic, literally incredible" (p 180).

VI. Desire and the Commodity Form

Whether this appeal is addressed to the textual or the historical spectator is not made altogether clear. However, as Doane herself realises in her opening chapter, such a renegotiation of subject positions requires a more detailed analysis of the function of the commodity in the economy of the narcissistic subject addressed by the cinema. Thus, when Doane speaks about 'The Commodity Form in/of the Cinema', she remarks that commodity fetishism should more properly be called a form of narcissism, meaning thereby that the indifferentiation between subject and object associated with narcissism – the confusion between having and being – has traditionally defined the relation of the commodity to the consumer, especially the female consumer.

Yet in a society where the management of this confusion has become an industry and a prized personal asset called style, giving the illusion of mastering the symbolic order by playing with its signs, strategies of mimicry are politically highly problematic. The woman's film – prone to parodying the feminine – might come to be seen as the genre where the spectator (female or male) becomes, if not 'feminised', then adept at investing objects with the signifiers of the self mirrored by or attributed to the commodity. The ambiguity then transfers itself onto the consumer, and her/his way of using the commodity as sign, or rather, as the sign of a sign, since the commodity is, like myth in Barthes, a species of 'stolen language'.

The other way of answering my question of what it is that draws men to melodrama may also require a switching between the textually constructed spectator and the 'historical' spectator. Whatever difficulty the woman's film has in situating and thereby containing female subjectivity, the fact that the genre thematises the woman's desire gives it a fascination not that different from what can be found in certain forms of pornography: the woman's film, however indirectly, satisfies the desire to see the woman's desire. By dramatising the difficulties of her having access to her own desire, the genre adds its own sadism to the masochistic 'scenarios of waiting, giving, sacrificing and mourning' (p 180).

However, there may be yet another way of reading the situation. The spectator in the cinema is someone who is lacking, a lack which makes her/him not just an addressee but a desiring subject. All desire is grounded in absence, or as Doane puts it 'desire is a form of disengagement – from need, from the referent, from the object'. If male desire ideally involves the perception and maintenance of spatial distance (which makes voyeurism such a perfect form of desire) then captivation in the cinema is for the male spectator desire deferred.

For Doane the impossibility of the female spectator having direct access to desire in the woman's film 'produces perturbations and contradictions within the narrative economy, ... ideological stress points ... [which] can hopefully be activated as a kind of level to facilitate the production of a desiring subjectivity for the woman - in another cinematic practice' (p 13). But has Doane made a case why such a 'desiring subjectivity for the woman' should have the cinema as its site of production, since, as she herself points out, the cinema by its very nature as a signifying system (due to the heterogeneity of its signifying materials; image, dialogue, music) is incapable of producing anything other than the fiction of a coherent I, even assuming that desire, in the form of being subject to the symbolic order, is desirable?

If we follow Lacan, desire is always a relational term, it is generated in the gradient emerging from perceived difference. This would suggest that the 'desire to desire' is in fact a kind of double negative, and grants the female spectator, or for that matter any spectator caught up in the signifying process of the woman's film, a special sort of intensity, a radicalism of desire: it is not desire denied, but desire doubled.

In fact, it might even be possible to argue that if the goal is access to desire, then the desire to desire is the more desirable desire. Woman's 'mediated' access to desire would then be her passage to a more intensified pleasure in the cinema: she would enjoy her pleasure, as it were, not doubled but 'squared', according to an asymptotic relationship between access to desire and intensity of affect: the more difficult or mediated the subject's access to desire, the greater the subject's jouissance. This, at any rate, might be desire according to Lacan Surrealist, as against Lacan Saussurised.

If, on the other hand, the woman lacks 'lack', her femininity would indeed consist of not being subject to the symbolic order, founded as this is on castration anxiety and lack. She would not need desire in order to negotiate separation and loss, in which case, going to the cinema would be a kind of luxury, an excursion into the land of lack, desire and difference. Feminism and film theory would have to part company.

I began by mentioning the paradox of the cinema's intense theorisation of spectatorship at the precise moment of this spectatorship's historical disappearance. If I end on a slightly heretical note, it is because the television spectator – a viewer, or in Sandy Flitterman's phrase, 'a new social subject: part viewer, part consumer'15 – is, I think, no longer a spectator driven by lack. Her/his access to desire, even more clearly mediated by the commodity, will have to be theorised differently. As television positions the viewer not in the field of vision, but generates its hold (such as it has) on subjectivity out of the multiplicity of its voices and modes of address, the heterogeneity of its images and views, so intelligibility and interpretation, social preconstruction rather than textual construction and imaginary coherence will dominate the question of spectatorship. From the television armchair, and not only for the male spectator, then, the desire to desire of the woman's film may soon be that most desirable of places: a lost paradise.