Friday, July 20, 2007

Psychoanalysis (More Detailed)

The application of psychoanalysis to cinema is by no means new. In particular, the productions of Hollywood (the 'dream factory") were amenable to psychoanalytic interpretation, displaying, as they did, the familiar repertoire of Freudian motifs. Such readings, however, tended towards reductionism in that the ostensible meaning of the film (comparable to the manifest content of the dream) was displaced by the hidden, Freudian meaning (equivalent to the dream's latent content), which tended towards a certain sameness. The loss of par­ticularity and difference effectively discredited psychoanalysis as a crit­ical method. However, its reintroduction into film studies was on die very different grounds of the need for a theory of the relations of the subject to discourse, which is exactly what Jacques Lacan's reworking of Freud appeared to offer. Although we shall show at the end of this chapter that a reading of Lacan as making good a deficiency in Marxism and semiotics was more problematic than was assumed, we shall for the moment simply sketch those aspects of his work that were taken up during the 1970s, most notably his account of the development of the subject.

The child is born into the experience of lack, what Lacan terms the manque a etre (the ‘want to be'); and the subject's subsequent history consists of a series of attempts to figure and overcome this lack, a project that is doomed to failure. Though the form and experi­ence of lack may alter, the basic reality of its persists and defies re­presentation. In retrospect - and for Lacan this history, like all his­tories of the subject, his own theory included, can only be retrospec­tive - the child interprets the prior union with the mother as anterior to lack, a condition where it was everything and lacked nothing.

Throughout its life the child will attempt to recapture this imagined entirety in a search for that which will overcome the lack, the missing component Lacan terms I'objet petit a and whose most obvious proto­type is the breast. This stands as a representation, no more than that, of what is ultimately unrepresentable, in that the object that could overcome the lack is non-existent. As compensation for the continual failure to re-establish unity, the child will console itself with imaginary solutions, notably in idealised images of itself as complete.

In Lacan's account of the child's development there are three deter­mining moments: the mirror phase (the acquisition of a sense of self), the fort-da game (the accession to language), and the Oedipus complex (the submission to the laws of society). We shall summarise each of these in turn.

The mirror phase occurs between the ages of six and eighteen months and is the child's primary means of establishing the difference between itself and the world. On seeing itself in a mirror, or more complexly through identification with the body of another, usually the mother, the child responds jubilantly to this image whose com­pleteness and unity contrast with its own experienced disunity and lack of motor control, and assumes 'that's me'.

In certain respects the child gains from the assumption. It facilitates an awareness of the body as localised and separate from the environ­ment, which is a prerequisite for coordinated physical activity; and on the basis of this newly acquired awareness of boundaries the child is then able to develop a sense of its own separate identity, without which there can be no social interaction. However, these gains are offset by losses, in essence those of misrecognition, alienation and division. Because the self-image represents a state of maturation not yet achieved and a degree of completeness and perfection never to be attained, the image is a narcissistic self-idealisation or, as Lacan puts it, 'a mirage', designed to parry die lack in being and 'to preserve the subject's precarious pleasure from an impossible and non-com­pliant real'.1 As well as misrecognition the identification involves alie­nation, in that it is typically sanctioned from another, from elsewhere. The announcement 'That's me' (though not yet in so many words) is verified by an adult, again, usually the mother, holding the child up before the mirror. The child thus identifies with her perception of it (or more accurately what it imagines she wants it to be). In saying 'That's me' it is saying 'I am another5. As Juliet Mitchell points out, Lacan considers that the subject 'can only conceptualise itself when it is mirrored back to itself from the position of another's desire'.2 That is, the child is divided from the moment it forms a self-conception. In finding compensation for the manque a etre in the fictive unity offered by the mirror, % encounters yet further divi­sion, which overlays and complicates the original lack, an irreducible gap between the reality of the child's being and the idealised image it assumes as its own.

The mirror phase is usually conceived as emblematic, or indeed the founding moment, of the so-called imaginary, one of the three constitutive orders of subjectivity, of which the other two are the symbolic and the real. Our reading of Lacan, however, would draw all three into the domain of the mirror phase, as indeed into the other formative phases of the subject. The imaginary comprises the repertoire of images that the subject invokes to annul the originary gap, and is present in the mirror phase as the image of the other with which the child identifies and which masks its division. The symbolic, it will be recalled, comprises the Other of laws, rules, codes and prohibitions, to which the child must submit in order to enter society, and is present in the mirror phase because the identity avail­able to the child comes from elsewhere, from another subject — in the above example, the mother - whose desire is always already con­ditioned by it. The Other is present here as it is everywhere, assigning the child a place even before it is born. Finally, the real is defined negatively as that which the imaginary seeks to image and the sym­bolic seeks to symbolise. But it necessarily eludes all such attempts, remaining outside imagination and symbolisation, while retaining an effectivity. An example given by Lacan is the trauma of the separation from the mother, which is present in the mirror phase both as the manque a etre prompting the narcissistic idealisation of self and as the gap between this idealised image and the subject. The interpenetration of the three orders is worth stressing because they have sometimes been thought of as separable, a misunderstanding that has had adverse consequences for die attempt to theorise the relation between film and spectator.

Like the mirror phase, the fort-da game brings the child both gains and losses. It was originally named and described by Freud in 1915, having watched his grandson Ernst throw a cotton reel out of sight then retrieve it by means of an attached thread, while accompanying the two actions with the sounds 'o' and 'a' respectively. Freud hypothesised that the reel symbolised the child's mother and that the game was a way of coming to terms with her absence. In throwing the reel away the child moved from a state of helplessness (I am abandoned by my mother) to one of agency or even mastery (I aban­don my mother). Lacan's subsequent reading of the game de-emphasised the supposed bid for mastery and instead proposed that it represented the child's accession to language. The emphasis came to be placed on the child's invocation of a symbol to stand in for what was missing and through which the mother's comings and goings could be represented. It was her absence that prompted the adoption of the reel as signifier, which stood in a metaphoric relation to mother and child, present only by virtue of the absence of what it represented. Like Hegel, Lacan conceived of the word as the mur­derer of the thing: no representation is ever adequate to what it claims to represent. As we shall see, this is crucially the case with the subject's self-representation. The fort-da game is a language system in micro­cosm, in that the signifiers 'fort’ and 'da' are defined relationally, each by what it is not. The subject, therefore, is caught up in a pre-existing language whose terms are organised diacritically and not by any relation to the real.

The gains for the child are those of entry into society, with all the concomitant possibilities for cooperation and communication. But these gains exact a toll of yet further division and experience of lack. For instance, although the child can now articulate its needs as demands, like calling for 'juice' to assuage its thirst, there is always a surplus in demand that amounts to a request to the Other, here usually the mother, to make good the manque a etre. Typically figured as the demand for unconditional love, it is bound to remain unsatis­fied because no such love exists - hence Lacan's comment that loving is giving what one does not have. The discrepancy between the satis­faction of the need and the unsatisfied demand for love is the condi­tion for desire, the unfulfillable search for the eternally lost object (objet petit a). The entry into language and the discovery of lack in the Other therefore precipitates the child into the constitutionally unsatisfiable state of desire.

In a further Sense, too, the entry into language is the birth of desire. Because the laws of society are inscribed within language, entry into the symbolic order entails that the child submits to its pre-given place and role, while that which is not consonant with such a social identity is consigned to the unconscious. When the child accepts this identity, as it must, its desires and the terms in which they are figured are determined by the Other, by the laws of society. Desire is always destined to pass through 'the defile of the signifier’; that is, 'man's desire is the desire of the Other’.3 Henceforth the subject's relation­ships are, at their simplest, triangular: not subject and object, but subject, Other and object (where the terms of both subject and object are given by the Other). 'Everything emerges from the structure of the signifier’.4

The relationship of subject, Other and object is perhaps best approached from the side of the subject, and since this aspect of Lacan's theory has been a major focus of attention by film theorists we shall look at it in some detail. As will have become apparent already, Lacan's subject is nothing except by and through language. The subject is such only by virtue of a self-conception, but this neces­sarily involves misrecognition. The subject's self-representation occurs either by a name or by the first person singular pronoun, in both cases by a signifier taking its meaning from other signifiers. As with the mirror phase, this necessarily involves alienation and divi­sion, because here too there is an acceptance of an identity determined elsewhere, in this case from within the Other, the symbolic order, language. That is, the subject can only appear if represented in the Other, while simultaneously and consequently all that is repressed as heterogeneous to the identity given through any signifier means that representation is always inadequate to the subject's being. As Stephen Heath put it, the subject is at once 'represented and excluded, becoming some one by its constitution as less-than-one'.5 The uncon­scious is the field of exclusion entailed by representation. Therefore in representing itself by a signifier the subject appears only to disap­pear. Because of the inescapable coexistence of the unconscious with language, the subject is as inescapably destined to division in its attempts at self-representation.

The necessary division of the subject can be articulated as the dif­ference between the subject of the enunciation and the subject of the enounced. The T who speaks is always in excess of the T who is spoken of. In other words, the first person singular pronoun 'de­signates the-subject of the enunciation, but it does not signify it'.6 That is, by virtue of the unconscious and the fact of being in process, the subject eludes all attempts to pin it down in language: When the subject appears somewhere as meaning, he is manifested elsewhere as "fading", as disappearance'.7

Lacan proposed that the 'no win' situation of the subject could be explicated by the so-called 'vel’ of alienation'. He likened the ‘or’ to that offered by the highwayman to his victim in the phrase Tour money or your life': choose money and you lose both, choose life and you lose your money, ending up with a life deprived of some­thing. Lacan's version of a Venn diagram similarly offers a choice between meaning and being.

Non –






Choose being, and the subject disappears into non-meaning; choose meaning and 'meaning is only left curtailed of the part of the non-meaning which is, strictly speaking, what constitutes, in the realisa­tion of the subject, the unconscious'.8 The general implication is that there is no way you are going to keep everything, and the attempt to do so will result in your losing both. Better, as Roger Thornhill was advised by his mother in North by Northwest, to 'just pay the two dollars'.

Thus the subject comes into being at the cost of division, only to experience a renewed sense of lack, of manque a etre; this it seeks to fend off with an idealised self-image as a unity, which however is always subverted by the process of the subject in the symbolic, by the challenge of the unconscious to any identity.

The third major constitutive moment of infancy is in the Oedipus complex, when the child encounters sexual difference. As we have seen, the human infant is entirely helpless and dependent on the mother, a potentially terrifying state. To ameliorate its condition the child takes refuge in the fiction that it is as indispensable to the mother as she is to it. The child imagines itself to be what she lacks, and therefore desires: what Lacan terms the phallus. 'The phallus signifies what the mother on her own has not got. It indicates lack . . . [and] at the same time it stands for what makes up that lack.'9 It is precisely and only a signifier, not an existing object, which represents the neces­sarily absent object of desire. Lacan insisted that it was not to be equated with the penis (for neither men nor women possess it). Through identification with the phallus the child imagines it will complete the mother and be itself completed, recovering the lost unity it has come to believe preceded its own lack and which, if regained, will annul it. Such a state of union with the mother is figured as an Edenic condition of plenitude and the absence of all lack. However, the child's solution to its problem is exposed as illusory when it dis­covers that it is not and cannot be the phallus. For there is a third person, the father, whom the mother desires and who the child pre­sumes possesses the phallus. The Oedipal scenario turns on this inter­vention of the father, the moment of castration when the Law of the Father (‘Thou shalt not desire what was my desire') places an interdiction on the child's desire to be what the mother desires.

The father, then, is the third term, breaking the mother-child dyad and rupturing its imaginary wholeness. Henceforth the child must exchange its earlier identification with the phallus for the identity assigned by the Name of the Father, and the phallus now figures the fact of sexual difference, with the father perceived as having it and the mother as lacking it. Given this sexual division between those representing lack and those having what would seemingly make it up, the child has to come down on one side or the other: it cannot be both, it cannot be neither, and it has no choice which one it is.

Confronted by this divide the boy assumes the masculine identity assigned to him by the symbolic. But as we have seen in our discussion of the accession to language, entry into the symbolic always produces a renewed sense of lack, recapitulating the original manque a etre. In an attempt to compensate for this the boy indentifies with an idealised figure, namely the father as the supposed possessor of the phallus. This move, though, is ultimately doomed to failure, as in taking his place within the symbolic the boy comes to understand 'that there is desire, or lack, in the place of the Other, that there is no ultimate certainty or truth, and that the status of the phallus is a fraud.'10 Thus castration is the moment when the Other (O) is recog­nised as the barred Other (Ǿ). Whereas in the mother-child dyad the Other had been fantasised as the place where demands are met, it is now revealed that there is nowhere demand can be fulfilled and desire satisfied. Rather than a secure destination the child has reached only a precarious sexual identity, one always liable to challenge from the unconscious, constantly fading, decentred and divided.

The girl's Oedipal trajectory is harder to specify, its difficulty com­pounded by what Jane Gallop has referred to as 'the contagion…from subject matter to theoretical description.11 Freud's own prob­lems here are well known, with his talk of 'the dark continent of female sexuality" and the notable absences in his account of its development. His basic conception was that the girl's trajectory is the same as the boy's until the discovery of the mother's castration, when she is faced by three options: either to give up on sexuality altogether because she cannot compete, or to seek to acquire the phallus herself, or to take her father as love object. Only through the last of these options does she enter the female Oedipus. Though the father apparently has the phallus, under the pressure of social taboos the girl must renounce him and seek substitute love objects who also appear to have the phallus, at the same time as identifying with her mother who has not got it. However, this identification with the supposedly castrated mother has proved difficult to explain. The Lacanian psychoanalyst Catherine Millot has put the problem thus: for the girl 'there is no ideal feminine identification possible other than the phallic woman; but this is precisely a pre-Oedipal identification.'12 Hence, 'the recognition of castration. . . leaves no possibility. . . of a straightforward post-Oedipal identification with the woman.'13

Nor are matters any easier in Lacan's version. The broad outline of his account, at least as expounded by Jacqueline Rose in her intro­duction to Feminine Sexuality, is clear. The subjectivity of women, like that of men, is constituted within language and the symbolic; their sexual identity is enjoined on them by the law, is therefore not pre-given but legislated; this identity, as for men, is taken up with reference to the phallus, in this case a matter of not possessing it. It is also clear that the woman's relation to this assigned identity is even more troubled than is the man's to his assigned identity in that she is constituted as a subject within a symbolic order where women are treated as objects. Not the least aspect of this objectification is the role women play in the fantasies of men who, divided in the symbolic, use them to represent their own problems in relation to desire, to object petit a. Women figure both as the representation of lack, in that the 'lack inherent in being human. . . is projected onto women', and as that which can make good the lack, as the woman (though of course, as Lacan insists, the woman does not exist).14 On the one hand she is constituted as what men are not, as lacking, as 'not-all', and on the other as the terminus to desire, the site Jouisance, the everything.

Lacan argued that the woman's difficulties in the symbolic are such that she is excluded, which has been on occasion interpreted as meaning that women lack a voice. For example, Ann Rosalind Jones has written: 'Lacanian theory reserves the "I" position for men. Women…occupy a negative position in language.'15 But such read­ings have been contested. Thus Millot reiterates Lacan's argument that there is nothing missing in the real. It is not a question of women being denied access to the symbolic, of being deprived of speech, but of a lack arising from the symbolic itself, of which they are a part. Jacqueline Rose too stresses that women's exclusion is by not from language. The 'not1 of the 'not-all' derives from women being defined against men, as the exception to the phallic rule. What is unclear in either interpretation is why women take up their assigned identities and the consequences this process has for their sexuality. The absence within psychoanalysis of a satisfactory explanation of feminine sexuality was, as we shall see, severely to restrict its value in understanding the exchange between film and women spectators.

The critical question emerging from this selective account of the con­struction of the subject is how the relations of the subject to his or her discourse are to be conceived. The particular problem Lacan faced was how to represent the situation of a subject that is at once con­stituted and constituting. Lacan wished to say both that the signifier is anterior to the subject (or that the subject is an effect of language) and that the subject is anterior to the signifier (so that language represents the subject). The problem lay in overcoming the explicit contradiction of maintaining both. The answer was the graphe complet, Lacan's figuration of the subject in process.16

For our purposes a full explication of the graph is unnecessary, it being sufficient to note that it is organised around two vectors: the vector of speech, which is the subject's signifying chain; and the vector of the drive, which is the search for satisfaction. Both of them pass through the Other (the symbolic), and the graph illustrates the effect of that passage on the constitution of the subject.

The vector of speech, the lower of the two left-to-right arrows, is a different version of the graph already discussed in the previous chapter. The vector passes through the Other (A/l'Autre), language's synchronic dimension, the place where every signifier is stored and where the subject is constituted. But for there to be meaning, the signifying chain must be punctuated by the barred subject $ (barred because the subject is never fully represented in speech, i.e., there is always the unconscious), who is thereby constitutive as well as con­stituted. The subject calls a momentary halt to the slide of the signifiers at the punctuation point s(A). Belonging to the diachronic not to the synchronic, this is the moment when meaning emerges in the retroactive contextualising of signifiers and the anticipatory construction of those to come. In this moment something of what the subject desires is at once expressed and repressed. The barred subject $ is represented at the point s(A.), but only inadequately. The subject's appearance in the field of the Other occurs only through a stand-in, a representative, and therefore at the cost of division. The subject is thus 'constituted only by subtracting himself, because by punctuating his or her discourse to give it sense s/he can only appear after the fact, by which time s/he is in the process of becoming some­thing else.17 The appropriate tense for this situation is 'the future anterior of what I shall have been for what I am in the process of becoming'.18 This fading of the subject at s(A) prompts him or her to seek compensation in an idealised image, i(a), which will fend off the lack. Such a specular image, whose prototype is the image in the mirror, produces a misrecognition of the self, m, the moi, thereby effecting the junction of the imaginary and the symbolic, otherwise known as suture.

The upper part of the graph duplicates the lower, with the arrow here representing the vector of the drive. In seeking satisfaction, the drive addresses a demand to the Other, one that is made at the site ($ ◊ D), which like A below it is also a place. It is where the subject seeks to overcome the lack in being by demanding unconditional love, and because in so doing s/he discovers the impossibility of what s/he demands it is therefore the site of castration. The various symbols of the formula expand as follows. $, again, is the barred subject, eclipsed and fading, and D is the demand. The lozenge, ◊, has two meanings, each dependent on an interpretation of its graphemic origins. If read as the conjunction of the mathematical symbols < (less than) and > (greater than), it signifies the impossibility of the demand being granted. If, on the other hand, it is taken as the symbol used by silversmiths to guarantee authenticity, then it stands for the uniqueness of each particular individual's demand. The correspon­dence between upper and lower parts of the graph continues with S(Λ), which like s(A) is a moment, not a place. It occurs when the subject enters the symbolic and discovers that the Other is lacking. Finally, castration gives rises to desire, d, on the upper right of the graph, which loops over the top in the direction of fantasy, ($ ◊ a), where $ is the barred subject, a is the objet petit a, and ◊ is the screen onto which the subject projects his or her own uniquely fantasised objet petit a..

Broadly speaking, two phases may be distinguished in the use made of psychoanalysis, each of which emphasised one aspect of the Lacan-ian model. In the first, structuralist, phase, centred on the work of Baudry, Metz and Mulvey, the emphasis was on the constitution of the ego in the mirror stage (on graphe complet, the portion

m i(a)).

In the second, post-structuralist, phase, associated for instance with the work of Cowie and Rose, the emphasis was on desire and fantasy on the graph

($ ◊ a d).

Another way of distinguishing the two phases would be to say that the first was concerned with sameness and the second with difference.

In her article 'Visual Pleasure and Narrative Cinema', Laura Mulvey concentrated on the more classically Freudian features of Lacan's theory, notably identification, voyeurism and fetishism. In view of its impact, especially among feminist writers on cinema, we shall follow her argument through in some detail.

The argument opens with the suggestion that the pleasure offered by mainstream Hollywood-type cinema, its 'fascination', depends on pre-existing psychological patterns at work within the spectator. Such pleasure is indissociable from dominant cinema's capacity to articulate patriarchal ideology around sexual difference, or more precisely, its capacity to negotiate the contradictions inherent in this ideology. By offering a kind of satisfaction to the alienated subject of patriarchy, cinema ensures its own commercial success. Principal among pleasures offered are those of identification, where the spectator narcissistically identifies with an idealised figure on screen, typically a male hero whose actions determine the narrative, in a process that recapitulates the discovery of the image of oneself in the mirror phase; and scopophilia, or pleasure in looking, through which the spectator indulges in a more socially acceptable form of peeping tom-ism whereby the other, typically a woman, is turned into an object of fantasy, so giving the voyeur a position of control and mastery. In respect of this latter pleasure, however, the security and mastery of the spectator seated in the solitary intimacy of darkness is in fact deeply problematic. For the image of the woman also brings with it what the spectator's look would disavow, the fact of sexual differ­ence itself and the concomitant threat of castration. In order to allay the anxiety so engendered, the Hollywood film tends to respond with two basic strategies. One of these is to deepen the already present voyeuristic aspect of cinema, coupling it with sadism, so that the difference figured by women is investigated by the constant re-enact­ment of the discovery of the lack. Such texts, of which films noirs are outstanding examples, attempt to master the anxiety both through their typical investigation of the woman and by their narrative punish­ment of her. The other strategy is to turn the woman into a fetish object, thereby containing the threat of difference through disavowal. Instead of lacking, women are represented as being complete and perfect. Here, parts or the whole of the female body are endowed with a value that compensates for the lack elsewhere, typified by the overinvestment in Monroe's breasts, Grable's legs, Hayworth's shoulders. Though this fetishisation is found throughout Hollywood, and is indeed an integral part of the star system, it finds its apotheosis in the films of von Sternberg, where the female body is no longer relayed through the looks of the male characters in the fiction but is offered directly to the spectator as a highly stylised and fragmented perfect product.

The avowed purpose of Mulvey's analysis was destruction. By examining dominant cinema's pleasures and exposing their phallo-centrism she hoped to create the space for an alternative cinema speak­ing 'a new language of desire', involving a totally different regime of spectating.19 There is, towards the end of the article, a shift in approach that brings it close to the Althusserian orthodoxy of inter­pellation. Implicit in her discussion of voyeurism and fetishism, the conception of the spectator being positioned by the text in such a way as to be blind to other perspectives becomes explicit around the organisation of looks within cinema. (Since we shall return to this topic in chapter 5, we shall do no more here than note its presence in Mulvey's article.) Interpellation, in providing an identity and a perspective on reality, bears on that aspect of the graphe complet which involves the identifications of the mirror phase. And this aspect, as we shall see, proved to be central to the theories of the apparatus associated with Jean-Louis Baudry and Christian Metz.

In developing their ideas the theorists of the apparatus were in­debted to earlier studies of perspective in Renaissance and post-Renaissance painting. Such studies, for the most part from within a Marxist framework, had emphasised the importance of form over content in determining the ideological effectivity of art. It was not simply that the paintings of the Renaissance began to speak the lan­guage of bourgeois ideology through their adoption of greater realism, more secular subject matter and an increasingly marked indi­vidualism of style, but also, and more importantly, the system of perspective based on a convergence towards a vanishing point in the picture indicated that there was a single, unique point in the imaginary space outside it from which its content was perceived. In other words, perspective gave the spectator an omniscient unitary place from which to view what was depicted, thus reinforcing the bourgeois notion of the subject as a free unique individual.

Jean-Louis Baudry's article 'Ideological Effects of the Basic Cinematographical Apparatus' was an early attempt to think through the implications of perspective for cinema. Because the camera lens is modelled on the same optical principles that underlie the perspectival system of Renaissance painting, cinema ensures that the spec­tator is established as the active centre and producer of meaning. The events, people, landscapes and objects of the film, its fictional reality, are always and necessarily seen from a fixed point in its imagin­ary space, one that is occupied by the spectator. Thus visually positioned, the spectator is blinded to the work of the film, its frame-by-frame construction of what passes for reality. And just as the suc­cession of those frames is effaced in favour of a continuous vision, so too is the film's ideological operation. Far from disrupting the unitary, perspectivally-defined position of the spectator's vision, thereby revealing both their and the spectator's contingency, the suc­cession of images in fact augments the spectator's imaginary dominance. The movements of the camera, the refraining of the shots, the cuts from one image to another are indeed comparable to the operations of an eye 'no longer fettered by the body, by the laws of matter and time' and are not, as might have been expected, a threat to that transcendental supremacy.20 The movability of the camera, with the resulting multiplicity of perspectives, in fact provides the most favourable circumstances for 'the manifestation of the trans­cendental subject'.21 Further, the various cinematic devices of framing, movement and editing are perceived by the spectator as acts of syn­thesis and constitution, and hence serve as evidence for the existence of a synthesising, constitutive, that is, transcendental, subject. Those very textual operations that would appear to put in question the spec­tator's self-identification as transcendental actually work to confirm this misrecognition. Although the subject's (mis) recognitions and (mis) perceptions are in reality a function of the text, the textual strategies in play convince him or her of the opposite. In sum, Baudry argued, the spectator is constituted by the meanings of the text but believes him or herself to be their author. This is precisely an account of the process of interpellation, where the subject appears to be the source of the meanings of which he or she is an effect.

In a later article, The Apparatus' (1975), Baudry moved away from a reliance on the Lacan of the mirror phase towards a more classically Freudian model to explain the ideological effects of the cinematic apparatus and in particular the impression of reality it created. Once again he challenged the view that this was a con­sequence of cinema's uniquely mimetic power, arguing here that this was less significant than its capacity to institute a mode of subjectivity analogous to the state of dreaming. At once enabled by and enabling of sleep, the dream, according to Freud, involves a state of regression comparable to the beginning of psychic life, where perception and representation are not differentiated. Thoughts are transformed into images, word presentations are transposed into thing presentations, repressed desires find expression and satisfaction in hallucinated images. That is, the fantasies of the dream wish appear as reality, indeed as 'more that real', for unlike waking perceptions the represent­ations of the dream impose themselves on and submerge the subject. The desire to recreate this state of regression is, Baudry maintained, 'inherent in our psychical structure' and has in the course of history given rise to a number of art forms, like painting and opera, although these are mere 'dry runs' for cinema, failed attempts to achieve its unique capacity to correspond to the dream state.22 Nowhere is this more effectively brought about than in the darkness of the auditorium with the spectator immobile and passive, gazing at moving images. The apparatus, projecting images onto a screen, mimes a form of archaic satisfaction, returning the spectator to a time when the se­paration between the subject's body and the world was ill-defined. It is this archaic identification rather than those secondary ones of the mirror phase that is fundamental to the desire for cinema and that explains the spectator's attachment to the images. The peculiar impression of reality engendered by cinema derives from the subjec­tivity constructed by it rather than the content or formal organisation of the film texts themselves. Only by acknowledging the force of the unconscious in the subject was it possible to account both for the desire for cinema and its reality effect.

Despite Baudry’s second thoughts it was the first of these articles that exercised the greater influence, not least on Christian Metz. In his seminal article 'The Imaginary Signifier', Metz returned the Lacanian concept of the mirror phase to the centre of the theoretical stage as he addressed the question of what, in the light of psychoanalysis, can be understood of cinema's specific characteristics in relation to the spectating subject. At the outset, he distinguished two 'machines' operating within the cinematic institution: one being cinema as indus­try, making commodities whose sale as tickets provides the return on the original investment; the other being the spectators' psyches, experiencing film as the pleasurable 'good object" and hence wanting more of the same. The economy of the former, the circulation of money, is interdependent with the economy of the latter, the circu­lation of pleasure. Metz's major theoretical concern was with the second of these two machines.

His analysis opens with a definition of the cinematic signifier as distinctively different from those of the other major art forms. In the first place, cinema operates over a much wider perceptual range than most other arts, bringing in sound, vision and the perception of movement in an ordered time sequence. And secondly, in compari­son with those arts, such as theatre, opera and other spectacles, that do involve as rich a perceptual register, cinema does not offer percep­tions belonging to the same time and space as the audience but images and recordings of what is absent. Instead of the presence of real players, sets and props, the screen presents its audience with a world that is physically absent, 'in effigy…in a primordial elsewhere'.23 Whether or not what is perceived is functional, the actual unfolding of it on screen is - which is why Metz can write, 'every film is a fiction film'. The distinctive feature of cinema, therefore, is this necessarily imaginary, absent quality, quite apart from whatever imaginary world it may happen to represent. The cinematic signifier is itself imaginary. Hence cinema uniquely involves its audience in a play of absence and presence, whose prototype is the imaginary completeness of the absent image of the child in the mirror. Screen images are 'made present in the mode of absence', and it is this com­bination of the presence of a rich perceptual field and the absence of what it conveys that essentially defines cinema.24

With this much established, the analysis then proceeds to enquire what modes of subjectivity are implied by the characteristic play of absence and presence. Paralleling Mulvey, but differing in certain crucial respects, Metz maintained that three basic processes were operating: identification, voyeurism, and the related phenomena of disavowal and fetishism. Of these three, his proposals around iden­tification proved to be the most influential.

In order to function at all in social life, Metz argued, people must have a sense of identity; hence the process of identification must accompany all social practices. The question then is, ‘With what…does the spectator identify during the projection of the film?'25 Given that cinema has already been established as a technique of the imagin­ary and in so far as the screen bears a certain resemblance to a mirror, the obvious answer would be that the spectator identifies with an image of him or herself in a manner analogous to the mirror phase. This, however, is ruled out on two grounds. One is that the screen, unlike the mirror, does not image the spectator's body. And the other is that films only make sense to subjects who have passed through the mirror phase, entered language and accepted the laws of the sym­bolic. In other words, the imaginary of cinema presupposes the symbolic. (It is here, incidentally, that Metz claims he departs most markedly from Baudry.)26 An alternative answer to the question would be that the spectator identifies with a fictional character or with a star (the answer given by Laura Mulvey). While accepting that such identification does exist and provides one of the pleasures of cinema, Metz argued that it is in fact secondary, requiring an act of recognition by an already constituted identity for it to occur. If this is so, it is necessary to discover the source of this prior, primary identification that enables the spectator to recognise his like on screen.27

The answer to this further question is that 'the spectator identifies with himself, with himself as a pure act of perception'.28 The reasoning runs as follows: conscious always that he is in the cinema, in the presence of something only imaginary and hence, regardless of what happens on screen, unthreatening, the spectator is aware, firstly of himself as absent from the screen, placed outside it in a position of all-seeing mastery; and secondly, of the condition of films being per­ceived, namely that he exists there in the auditorium as the seeing, hearing subject without which the film would have no point or even existence. This dual knowledge on the part of the spectator permits him to identify with his own act of seeing and hearing, as a pure instance of perception, 'as a kind of transcendental subject anterior to every there is'.29 The point has been made by subsequent commen­tators in possibly more accessible terms. Thus, Stephen Heath writes that the apparatus of look and identification institutes the spectator in 'the totalising security of looking at looking3.30 And John Ellis states that identification with the cinematic apparatus 'involves the fantasy of self as a pure perceiving being.'31

Such identification is powerfully reinforced - and here Metz's debt to Baudry becomes more pronounced - through the spectator's con­comitant identification with the camera, whose monocular perspectival regime inscribes the place from which vision acquires a godlike omniscience. The spectator is the camera at once actively training its gaze upon objects and passively receiving the imprint of its percep­tions. The homology between apparatus and spectator extends to the cone of projection of light onto the screen, which parallels an ideology of vision as just such a searchlight beam illuminating the field of the subject's intentionality. So profound is the homology between apparatus and the activity of perception that the two are taken as identical and indissociable — on the side of activity: camera eye, the projector, the cone of visual intention; on the side of passivity: sensitised emulsion, screen, retina. Yet for all the omniscience and omni- vision deriving from the identification, indeed because of them, the subject is nonetheless caught up in alienation and misrecognition, both 'transcendental yet radically deluded'.32

In his discussion of voyeurism and fetishism Metz covered some of the same ground as Mulvey, but with a very different theoretical concern. Whereas Mulvey had seen them as the basis for understand­ing the relationship between the spectatorial gaze and representations of women, Metz was concerned to show that their structures duplicated the cinematic machine of spectating, and hence are deeply impli­cated in it. To oversimplify, one might say that Mulvey's concern was with the cinematic signified and Metz's was with the cinematic signifier.

The two sexual drives on which cinema relies are scopophilia, the desire to see, and what Lacan referred to as the invocatory drive, the desire to hear. What distinguishes these drives from others is their even greater dependence on lack, not only in that they foster a fruitless search for the lost object that can never be recaptured, but also in that they must maintain a necessary' distance from their respective objects. While the drives associated with orality and anality seek a degree of fusion between their source and their aim, those of vision and hearing require the maintenance of the gap between the body and the object. This holding at a distance is indeed a diagnostic feature of the major arts such as painting, sculpture, music, and theatre, but what further distinguishes cinema in respect of the scopic drive is not so much the distance as the absence of the object. In theatre and opera the co-presence of the spectator and performer ensures that there is a presumed complicity between the two, so that the spectator's voyeurism is matched by the performer's exhibitionism, making them 'the two protagonists of an authentic perverse couple'.33 In cinema, on the other hand, the performer is present when the spectator is not (during shooting) and the spectator is present when the performer is not (during projection). The reciprocating acknowledgement of the existence of the other is far less pronounced, even absent, as, for example, in the convention of the actor never looking directly at camera and so confronting the spectator with his own voyeuristic gaze. This shame-faced, unacknowledged, non-consensual looking of the cinema spectator at that which 'lets itself be seen without pre­senting itself to be seen' places cinema in a direct line of descent from the primal scene, the unwitnessed witnessing of the parents' copulation.34 Various features of the cinematic institution contribute to this affinity. The spectator sits in darkness before a lit screen, making for an 'inevitable keyhole effect'; then, though a member of an audience, he nonetheless remains essentially solitary; the actors necessarily remain in ignorance of the spectator; and finally, the film unfolds in a place that is simultaneously close and yet definitely in­accessible, all of which makes the experience of voyeurism in the cinema one of transgression.35

The pertinence of psychoanalytic concepts to the elucidation of the relation between subjectivity and the cinematic signifier is most marked, according to Metz, in disavowal and fetishism. Like the cinematic signifier itself, these turn on the play of absence and pre­sence, and, while not going as far as to posit a simple equation between the cinematic situation and fetishism, Metz points to their shared features. The structure of disavowal can be conceived in terms of discrepant knowledge and belief. The child's discovery of sexual difference institutes an anxiety in the face of the threat emanating from the lack in the Other, figured as the absent maternal phallus. To ward off the threat the child disavows difference and the lack in the mother, resulting in the characteristic formula encompassing the contradiction: 'I know very well but all the same. . .' If, subsequently, some object is elected to mask the lack disavowed, an object that simultaneously denies that anything is absent but whose presence acknowledges that it is, then disavowal has taken the specific form of fetishism.

Metz argued that something very comparable to this takes place around the cinematic institution. The spectator knows very well that what he is watching is a fiction, but all the same he maintains the belief, indeed his pleasure is dependent on the belief, that it is not. Cinema is thus founded on a regime of spectating at once knowing one thing and believing its opposite, which, as we have seen, is pre­cisely the structure of disavowal. Indeed, with its rich sensorial presence and objective absence, the cinematic image 're-plays the game of castration: "to be or not to be", death, anxiety'.3*' Like the fetish, which 'disavows a lack and in doing affirms it without wishing to'37 the cinematic apparatus itself is 'a kind of substitute for the penis'.38 Or rather, 'it is not exactly a substitute for the penis, but for the absence of the penis', both affirming the presence of what is absent and emphasising the fact of that absence.39 Cinema's technical achievement is to make what is absent so forcefully present that the spectator almost, but never completely, forgets that it is absent. Unless this awareness of absence is sustained there cannot be an appreciation of what is made present. The spectator's enchantment depends, as in classical fetishism, on a simultaneous awareness of what is present and what is absent. Just as the fetish completes the female body and disguises its lack, so the technical accomplishment of the cinematic apparatus perfects the imaginary signifier and masks the absence on which it turns. The fetishist gains pleasure from the object that stands both for the woman's lack and her lack of lack; the cinephile gains pleasure from the never-quite-closed gap between imaginary presence and real absence. So for Metz there are effectively two levels of fetishism: one where the apparatus is the fetish, the other where the image and its meaning (as for Mulvey) become the fetish.

Thus Metz answered his original questions: How does cinema per­petuate itself? How does cinema produce pleasure and the desire to return for more? Cinema is the imaginary signifier, that is, it involves a process of signification turning on an absence that it seeks to fill but never finally does. In the gap between presence and absence a lack constantly reappears, and it is this lack that renews desire, so guaranteeing the perpetuation of cinema as institution. Although his solutions were not received without some dissent - John Ellis, for example doubted the transgressive quality of film spectating pointing to the authorisation conferred on voyeurism by the present of others in the cinema — and despite an undeniable current of phallocentrism running through his analyses (to which we shall return), Metz's work had a tremendous impact on film studies.

Central to the thinking of both Baudry and Metz was the theme of the misrecognition devolving on cinematic signification, a theme that was taken up again in the debate around suture. This very complex concept refers both to the relation of the subject to his or her dis­course, and to the junction of the imaginary and symbolic thereby entailed. In speaking, or in enunciating a text, the subject is divided, but defends itself against this division by a pseudo-identification in which it imagines itself a unity. As a concomitant of every act of signification, suture in some form or other accompanies all linguistic and social practices. However, when the concept was taken up by film theory it was considerably simplified. With film theory's early emphasis on cinema as a discourse organised around absence and lack, subsequently inflected through an Althusserian terminology, the moment of pseudo-identification was understood as an instance of misrecognition. Suture, therefore, was held to be an effect only of certain texts, or rather of certain textual practices, which (along the lines of the post-1968 typology around political functioning) were those that alienated and deceived. It was only in the latter stages of the debate that the complexities of the concept were duly acknow­ledged.

The first important theorist of suture was Jean Pierre Oudart, who advanced a description of its operation within cinema in a series of articles in Cahiers. He proposed that Lacan's notion of the subject suturing the lack opened up by enunciation with an imaginary entirety fitted the logic of cinema spectating well. His argument ran as follows. The spectator's initial response to the cinematic image is one of jubi­lation, not unlike that of the child in front of the mirror. The image offers an imaginary plenitude, 'a pure expanse of jouissance’, in which the spectator is caught up in a fascination with the unreal.40 Such dyadic bliss is ill-founded and short-lived, for in the cinema, as elsewhere, there is no imaginary without the symbolic. The first intimation of crisis is the discovery of the frame, the terminus of the image that reveals the absent space out of frame and induces anxious questions in the spectator's mind. The image is no longer innocently there; it is there for someone. From this certain questions arise (Who is this missing spectator whose point of view this is? And who is ordering and framing the image?), questions that threaten to expose film as signifying practice, as a constructed and enunciated operation. What annuls the threat is the system of shot/reverse shot, by which a second shot shows the first to have been the field of vision of a character within the fiction. In this way the Absent One turns out to be a particular character whose point of view is disclosed, and the threatening absence is reappropriated within the film. By introducing a character to take the place of the Absent One, the system of shot/ reverse shot sutures the rupture in the initial relation of image to spectator and envelops cinematic discourse within the imaginary.

As conceived by Oudart, suture is the tragedy inherent in cinematic discourse, entailing as it does the loss of the totality of the image and hence of spectatorial pleasure. Such a conception is oriented towards an evaluation of films according to whether they expose the specifically tragic nature of cinematic language, with Oudart citing The Trial of Joan of Arc as an example of a film that does do so and Au Hasard, Balthazar as an example of one that does not. In the case of films that do expose their discursive processes, that do move from the imaginary towards the symbolic, the spectator is no longer positioned in an illusory relation to the text but is actively involved in a process of reading that reveals the film's textuality. In this way the truth of cinema is allowed to unfold and reveal itself.

With Daniel Dayan, the emphasis shifted from suture as the tragedy inherent in cinematic discourse to the ideological operation of a par­ticular mode of discourse.41 In taking up the concept from an Althusserian standpoint he represented the play of absence and presence in ideological terms, as a particular mode of interpellation or filmic address constituting the individual as subject. Along with other post-1968 theorists, he held that a film's ideological functioning was less a partisan depiction of the world than a mode of enunciation that masked the ideological origin of its discourse. Like Baudry and Oudart, Dayan understood the cinematic image as the equivalent of classical painting organised by perspective, with the spectator consti­tuted as a subject in a position of imaginary dominance by the specular effect of the image's spatial organisation. Unaware of the codes posi­tioning him or her, the thus-constituted subject is denied the know­ledge that the representations of the film are the product of a semiotic system. What is threatened by the potential exposure of this, through framing and so on, is the film's successful ideological operation. In order to sustain it, various strategies have been developed, primarily the shot/reverse shot system, which, by locating the origin of the image in the diegesis rather than in the process of representation, is able to render the working of the film's codes invisible. Consequently 'the spectator…absorbs an ideological effect without being aware of it.42

Dayan's account of suture has been criticised on various grounds. Barry Salt, whose exhaustive analyses of the textual procedures of classic cinema have overturned many received ideas, has demonstrated that shot/reverse shot comprises only some thirty to forty per cent of the total cuts in Hollywood narrative from the 1930s onwards.43 Contra Dayan, who maintained that other forms of shot were unusual, Salt pointed out that for most films the majority of shots were not within the shot/reverse shot format, and moreover that films such as Birth of a Nation, in which only three out of over 2000 cuts employ the reverse angle, work powerfully on their audiences. And if the device was so effective why, Salt demanded, was it not pushed to extremes (say, seventy percent or more of the cuts) in all commer­cial films rather than just a few.

William Rothman questioned the dominance of the two-shot sequ­ence (shot/reverse shot) and suggested instead that for Hollywood the norm is a three-shot sequence: first the character looking, then what is seen, then the character again.44 For example in the Bodega Bay scene of The Birds an initial shot of Melanie looking prompts the question, what is she looking at? This is answered by shot two, showing what she is looking at, namely, the Brenner house. Then follows shot three, showing Melanie's reaction to what she has seen. The significant point of this is that the Absent One has no role here at all: the question 'Whose point of view is this?' simply does not arise. Spectators know a point of view shot when they see one, and indeed know that all shots are produced by the cinematic apparatus. The capacity to read a film and to pass judgement on the veracity of the representations is not therefore determined by the suppression of codes at all - if it were, cinema would be an ahistorical institution endowed with the power to deceive in all conjunctures. Hence the process as described by Dayan, functioning on behalf of ideology, is itself a fiction.

A further contribution to the debate was made by Stephen Heath, who, while accepting many of the criticisms made by Salt and Rothman, still held the concept of suture to be important for an understanding of cinema as discourse producing a subject address. In contrast to linguistic utterances, film images bear few, if any, of the marks of their enunciation, instanced by the relative difficulty of contradicting them. Images make it much harder to quarrel with the ideological representations they offer. The concept of suture is there­fore valuable in its emphasis on the cinematic image as an utterance and in making it clear that the apparent completeness of the image is only illusory, that it requires for its completion a subject of enun­ciation. But at the same time that subject is never finally and fully represented there, because the subject is always fading. This play of incompleteness and completeness is what suture can help to specify: meaning and subjectivity come into being together in the endless process that is the subject's emergence into the symbolic.

But in arguing for the concept's pertinence Heath was not uncritical of his predecessors in the debate. He took issue with the emphasis on shot/reverse shot, pointing out that, as the juncture between the imaginary and the symbolic, suture is present in every enunciation: all texts suture, though they do so differently. In Chantal Akerman's film News from Home there is no instance of shot/reverse shot, so according to the Oudart and Dayan conception it should therefore be unsutured, but the spectator is nevertheless 'included and moved…in a structure and a rhythm of lack and absence'.45 More generally he criticised Oudart and Dayan for their transformation of a purely descriptive concept into a means of evaluation, hence into the basis for a typology of films. Their mistake was to suppose that certain films were less implicated in the imaginary than others — for Oudart, the work of Godard and Bresson, for example. Since the imaginary and the symbolic are always co-present, this shift blocks thinking about the relation of spectator to film. Dayan's linkage of suture with interpellation also came in for criticism. As with all Althusserian readings, Heath said, there was a tendency to emphasise the imaginary at the expense of the symbolic and unity at the expense of division. Moreover, whereas interpellation conceives of the subject as produced, psychoanalysis makes it at once production and product: film does not position the subject but performs it, just as there is a 'permanent performance of the subject in language itself 46

The most important theoretical work of the 1970s deriving from Lacan centred on what Metz termed 'the social regulation of the spectator's metapsychology’.47 In conceiving of cinema as an institution with both technological and psychological components, the theory offered explanations of how, variously, the subject acquired an imaginary unity, an impression of reality was created, and the institution repro­duced itself in promoting the desire to return for more. The strength of the 'hegemonic and totalising model' of the cinematic apparatus developed by Baudry and by Metz was, however, also a source of potential weakness, as two main lines of criticism made apparent.48

In the first place, because the substantive features of cinema dis­cussed by Baudry and Metz are found, if not in every film, at least in certain broad categories, the variability of spectator response remains unaccounted for. The fact that different cinema goers react differently to the same film implies that some essential determinants of a film's reception are being neglected by the theory. Secondly, the theory in effect foreclosed on the possibility of transforming cinema, as was argued by Constance Penley. She pointed out that if the effects of the apparatus were total and irresistible then there could be no form of cinema that could subvert its power.49 Since such forms of oppositional cinema were held by many theorists, notably by feminist critics and filmmakers, to be not only desirable but possible, then there must be some places where the control of the cinematic institution was not total. Both lines of criticism were to lead towards the psychoanalytic concept of fantasy.

It is, of course, something of a commonplace that a parallel may be drawn between the condition of the spectator in watching a film and the condition of the dreamer, daydreamer or fantasist. As John Ellis puts it, 'Images and sounds are received in a state where the normal judging functions of the ego are suspended to some degree (near to sleep), so that what is seen is not subject to the usual expec­tations of plausibility that we apply to everyday life.'50 There is, moreover, similarity between filmic form and content and fantasy itself, so much so that, for instance, Laplanche and Pontalis can describe fantasy in terms of a cinematic metaphor when they write-off it as 'the mise-en-scene of desire.'51 After all, what is Holly­wood, with its stars, its happy endings, its interminable elaboration of Eros and Thanatos, but fantasy? However, with this acknow­ledged, there is the risk of simply repeating the reductionist interpre­tations offered by the earliest application of psychoanalysis to cinema, the invariably successful hunt for sexual symbolism in apparently innocent texts. The reintroduction of the notion of fantasy might, in other words, turn out to be a limited and regressive move.

That such was not the case may be attributed to two significant differences between its more recent and its earlier application. The first of these was that fantasy was conceived not as originating in the mind of the director but as operating in the exchange between the film and the spectator. The concept, then, was not intended to replace the imbrication of text and spectator with a content analysis, but was a means, precisely a means, of elaborating the relationship between text and spectator. In enunciating a fantasy that drew support from the text, the spectator was at once constituting and constituted.

The second major difference was that instead of fantasy being considered as straightforward wish-fulfilment, it was acknowledged to be a more complex compromise formation in which the repressed ideas were given expression, but only in a distorted form, dictated by the repressing agency. As with the classic symptom, then, enun­ciated fantasy contained both the unconscious wish and the defence against it. As defined by Laplanche and Pontalis, fantasy is an 'imagin­ary scene in which the subject is a protagonist, representing the ful­filment of a wish (in the last analysis, an unconscious wish) in a manner that is distorted to a greater or lesser extent by defensive processes'.52 Fantasy never articulates desire alone but always desire and the law. And even more complexly, it may express conflicting desire and the law in a single ensemble.

An example of fantasy as compromise formation in the analysis of a film is Elizabeth Cowie's reading of Now Voyager. In it the fantasy played out is that of the phallic mother. Having displaced her own mother, Charlotte becomes a surrogate mother to her lover's child by taking her in and caring for her, while at the same time denying herself sexual relations with him and leaving his unhappy marriage intact. In this way her initial transgressions, her enactment of homosexual and aggressive impulses, are indirectly punished and thereby legitimated by the film. By not marrying her lover and yet retaining a part of him in his daughter she at once becomes the phallic mother and abides by patriarchal law. Desire and its prohibition are thereby both articulated through the film. Cowie's analysis is not that of Charlotte as analysand, which would merely have a sophisti­cated yet still reductive content analysis, but is instead one of the film itself. The fantasy is the film's, not Charlotte's; it is an effect of its narration and therefore available to the spectator alone, the 'place in which all the terms of the fantasy come to rest'.53

Thus the concept of fantasy at once continued and developed the earlier application of psychoanalysis to film. The continuation lay in the attention to the spectator as the subject of the enunciation, the development in the greater complexity accorded to this subject. For in fantasy the spectator engages in multiple identifications and in its filmic scenarios may identify with several figures simultaneously, women and men, winners and losers, heroes and villains, the active and the passive, This conception of the subject as occupying con­tradictory positions and thus articulating conflict within the psyche, is distinctly different both from traditional notions of identification with the star of the same gender and from the Althusserian model in which the spectator is fixed in position by an assumption of a unified self-image.

The psychoanalytic text most frequently referred to here is Freud's paper 'A child is being beaten', concerning a masturbatory fantasy reported by some of his patients. The fantasy had three phases, which differed according to the gender of the patient. In the first phase, the child being beaten was not the patient and the beater was an adult of indeterminate identity. Freud interpreted this adult as being the father and surmised that the fantasy demonstrated the father's wished-for love by having him beat a rival sibling. In the second phase it is the fantasist who is being beaten by the father. For the female patient this represents a compromise between her desire for the father and her guilt about that desire, the sadism of the first phase thus being transformed into masochism. For the male it represents both a masochistic attitude towards the father and an identification with the feminine position, thereby expressing a wish to take the father as love object. The third phase, though resembling the first, continues to act as the agency of the second phase, that of the passive desire for the father. But the male represses his homosexual desire by identifying with the beating adult not the beaten child, so adopting the active masculine position; and simultaneously he adopts a passive position towards the desired phallic mother through the variation 'I am being beaten by my mother'.54 The female also seeks to avoid incestuous attachment to the father in this phase, but does so by replacing the father with some other adult such as a teacher and by adopting the position of an onlooker, 'a spectator of the event which takes the place of the sexual act’.55

Freud's analysis clearly demonstrates that fantasy entails multiple points of identification and places of enunciation. An example of the way such complexity can bear on filmic reading is given by Laura Mulvey's study of the Western, where, she proposes, narrative closure typically takes one of two forms, either a marriage, resolving the Oedipus complex and integrating the hero into the symbolic order, or non-marriage, 'a nostalgic celebration of phallic, narcissistic omnipotence'.56 The tension implicit in this alternative often results in there being two heroes, as in The Man Who Shot Liberty Valance, in which Tom (John Wayne) embodies primitive phallic power, but whose defeat of Vallance goes unrecognised and who loses the woman, and in which Ranse (James Stewart), 'the upholder of the law as a symbolic system', is misrecognised as the victor and marries the woman.57 With this fantasy scenario the spectator is able to iden­tify simultaneously with Tom and Ranse, with Ranse mourning Tom, and with the woman marrying Ranse but loving Tom. Thus the spec­tator's desire in all its complexity is given expression and becomes 'pleasured'. But the pleasure so gained is not to be conceived in terms of the traditional wish-fulfilment model.

This last point was something particularly stressed by those most influenced by Lacan, for whom fantasies involve not satisfiable needs but unsatisfiable desires. Because, for Lacan, desire is in pursuit of an eternally lost object, it is more accurate to say that fantasy sustains rather than satisfies desire, that it is the staging or mise en scene of desire rather than its fulfilment. As Lacan put it, 'the fantasy is the support of desire, it is not the object that is the support of the desire';58 and Cowie, 'the fantasy depends not on particular objects, but on their setting out; and the pleasure of fantasy lies in the setting out, not in the having of the objects'.59 Desire is therefore perpetuated through ever more elaborate signifying ensembles, one of which is of course narrative, where the spectator both desires and does not desire resolution. When resolution occurs the lost object figured by the narrative will be achieved only to fall once again into loss by the very fact of that achievement. So long as the narrative delays the desired moment, leaving open the question of how and when it will occur, while leading inexorably up to it, then pleasure is the outcome.

An example given by Elizabeth Cowie is the film Reckless Moment in which the central characters Lucia and Donnelly circulate through a number of positions.60 Lucia is variously Donnelly's lover and mother and, in the absence of her husband, father of the household; Donnelly, for his part, is variously lover, son, father and mother. At the end of the film the tensions produced by this sliding between positions are not resolved, only halted; in any case what matters is not so much any would-be resolution, but the succession of figures, equivalences and exchanges put into play by the narrative. Similarly Elisabeth Lyon's analysis of India Song as fantasy establishes that the spectator is variously and simultaneously 'the 'I/ego of the camera -the beggarwoman - Ann-Marie Stretter - the Vice-consul - death'.61 Taking up Lacan's formula $ ◊ a (where $ is the barred subject, a the objet petit a), she describes fantasy as the relationship of the subject to the non-existent object of desire, figured in the film by the key image of the naked breast of Ann-Marie Stretter. The lozenge in the formula stands for the third element, the Other, which always separates subject and object, and here represents the interchangeabil-ity of positions inside and outside the fantasy, the positions of par­ticipant and observer. As in Reckless Moment there is a circulation that never reaches a resolution: fantasy as the staging of desire can never provide an answer to the question of desire, but can only re-pose it.

More generally, the formula $ ◊ a. conveniently summarises the several ways in which the concept of fantasy has reoriented the appli­cation of psychoanalysis to film. The emphasis is now on objet petit a, desire and the symbolic rather than on the imaginary, concomitantly acknowledging that for the subject there is nowhere outside the sym­bolic order and no Other of the Other. Whereas, then, the earlier metapsychologists argued for the sameness of the effects of the apparatus on spectators, the concept of fantasy entails difference, not simply pertaining to a male/female dichotomy, but in recognition of the uniquely determined complexity of the psychic economy of each spectator. Fantasy means diversity of response to the same film, with each spectator enunciating their own economy of desire through it. Finally, if among the earlier theorists there was a tendency to invoke the unconscious only to ignore it, fantasy insists on it. Fantasy's sub­ject is barred, mobile, fading, present only 'in a de-subjectivised form ... in the very syntax of the sequence in question'.62

Psychoanalysis was introduced into film theory as a supplement to historical materialism and semiotics. The fact that it has not only remained but has moved to a position of centrality might perhaps seern surprising given the problems attendant on it. However, as we have already shown, psychoanalysis in its Lacanian mode proved to be remarkably consonant with post-structuralism, whose underlying precepts came to dominate film theory after the mid-1970s. Neverthe­less, despite this alliance, there were a number of outstanding unre­solved problems, which we shall discuss in the remainder of this chapter.

The first of these concerned the specific ways in which psychoanalysis had been applied to the study of film. The objection that the use of psychoanalysis in film studies ran counter both to the classic theory and to the more recent Lacanian reworking of it was not simply a matter of defending doctrinal purity. Rather, it intended to signal that any improper use of psychoanalysis would inevitably store up problems for whatever theoretical project it was informing.

Metz's proposition that cinema, more than any other art, involves its audience in the imaginary was challenged along such lines by a number of critics, including most notably Jacqueline Rose and Con­stance Penley. His claim that the cinematic image places it within the register of the imaginary is, they pointed out, fundamentally at odds with Lacan's thought, for no image has meaning in itself, given directly through vision, but only acquires it within a particular cul­tural order. Just as for the child the idealised self-image of the mirror phase is given by the mother's look, that is, by the Other, so too is the imaginary always informed by the symbolic. This failure to acknowledge the importance of the Other has significant reper­cussions for Metz's central thesis, that of the spectator's identification with his own act of perception, leading to a delusory omniscience. According to Constance Penley, Metz's line of argument that the apparatus always installs the spectator as an all-perceiving subject, confuses its aim with its effects, which is particularly apparent when one considers the subject in relation to vision and desire. For in look­ing, as in speaking, the subject is divided: 'vision always takes place in the field of the Other's vision and desire'.63 In treating the imaginary as a state of plenitude antecedent to accession to the symbolic, a state to which the spectator as an effect of the cinematic image regresses, Metz misses the workings of desire in the exchange between spectator and film. What is necessarily a triangular relationship is misconceived as dual. A number of consequences accrue to Penley's avowedly Lacanian respecification of these scopic relations. In the first place, the subject of vision is itself an object of representation, because what determines the subject is a look that is outside - for Lacan, in the field of the visible: 'I am looked at, that is to say, I am a picture'.64 Secondly, there is no such thing as a purely perceptual look, it is always a matter of it being conditioned by the desire of the Other. Because of the structure of desire there can be no question of the look achieving the satisfaction of seeing what it wants to see, but only perpetual deferral down the metonymies of narrative. Finally, the place the subject occupies, more seen than seeing, is itself, by virtue of its implication with the unconscious, unstable and dispersed. All in all, these consequences conspire to thwart the apparatus's aim of constructing a transcendental and secure subject identifying with itself in an act of pure perception. Faced with these criticisms Metz had no defence but to confess 'I am not a Lacanian'.65

The second problem, that of the supposed ahistoricality of psychoanalysis, may be dealt with quite briefly. Lovell's objection that psychoanalysis is an 'a-historical theory of the constitution of the subject and its entry into language and culture' can hardly be sustained when one considers that each individual is formed within a unique family configuration that is itself an effect of a wider histor­ical matrix.66 Psychoanalysis does not, and does not need to, offer a theory of social and familial change; its concern is to chart their effects through the oral histories recounted on the couch. Where Lovell, however, does have a point is that it has proved difficult to theorise the two together, to find a means of integrating psychoanalysis and theories of social change, whether Marxist or otherwise. These dif­ficulties emerged in concrete form, as we have seen already in chapter 2, when attempts to theorise the historicality of the subject in relation to reading texts remained largely unsuccessful.

The third problem, that of feminine sexuality, is by far the most pressing of psychoanalysis's lacunae. Freud's difficulties here are well known, not just in relation to his contentious privileging of the penis in the concept of penis envy, for his failure satisfactorily to theorise the female Oedipus leaves us with no overall explanation of the con­struction of female subjectivity. While modern feminist advocates of psychoanalysis have tended to turn to Lacanian theory as the place of resolution of these difficulties, it is still far from settled that Lacan has made any significant advance on Freud. At its simplest, the prob­lem is that if in patriarchal culture women are seen as lacking why should anyone assume a feminine identity.

For some feminists any problems here are outweighed by Lacan's central emphasis on the symbolic in the construction of subjectivity, which disposes equally of reductionist notions of biological deter­minism and of mystical notions of feminine essence. For such as Juliet Mitchell, Jacqueline Rose, and Ellie Ragland-Sullivan, Lacan's theories offer an explanation both for the construction of subjectivity under patriarchy and resistance to it. Their value lies in their 'exposure of the inevitable alliance between "feminine essence" and the natural, the given, or precisely what is outside the range of political action and thus not amenable to change'. 67 The insistence that the 'subject is not constructed from sexuality, [but] sexuality is constructed in, the history of the subject’ marks therefore a complete break with any idea that anatomy has to be destiny.68

Or does it? Is there not one almost axiomatic concept through which anatomy, despite all protestations to the contrary, returns? The phallus, according to Lacan, is not the penis. Possessed by neither men nor women, belonging to the symbolic order and not nature, taking its value like all signifiers from its relation to other signifiers, the phallus signifies the lack indissociable from entry into culture. As such, it permits sexuality to be conceived 'as an arbitrary identity that is imposed on the subject, as a law. . . legislated rather than autonomously assumed', hence provisional, often inappropriate, and potentially open to change.69 Not everyone, though, is convinced of the phallus's radically non-biological status. For example, Jane Gallop, in many respects sympathetic to Lacan, has detected in his and his followers' work an 'endless repetition of failed efforts to distinguish phallus and penis clearly3.70 Whatever else it means, the phallus also always stands for the penis, a confusion that is symptomatic of the impossibility of conceiving a non-phallic masculinity at this historical moment. This problem, along with the more general one around the construction of female subjectivity, was reflected in the uneven develop­ment of psychoanalytically informed film theory. Although by the end of the 1970s psychoanalysis had contributed to a persuasive account of the exchange between a film and the male spectator, no comparable account existed for the female spectator. According to Laura Mulvey's widely accepted analysis of the voyeuristic and fetishistic structures organising the male gaze, the woman was what was looked at, not the one who looked. The place of enunciation, the place of the look was for the majority of films that of the male. As Mary Ann Doane has noted, 'historically there has always been a certain imbrication of the cinematic image and the representation of the woman'71 and cinema has 'articulated its stories through a con­flation of its central axis of seeing/being seen with the opposition male/female'.72 Debarred, except as objects, from the characteristically masculine scopic regime of voyeurism and fetishism, women have a very different relation to the image from that of men. The reason for this difference, Doane proposed, is 'the overwhelming presence to itself of the female body3, a theme of self-proximity that has been elaborated by Luce Irigaray, Helene Cixious and Michele Mon-trelay.73 As a consequence woman are unable to establish the distance from the image that is the condition for voyeuristic pleasure and control, narrativising the Other as female image, and instead remain in a relation to it of identificatory narcissism. Women are similarly barred from fetishistic structures. As Doane has written elsewhere, what can fetishism 'have to do with the female spectator for whom castration cannot pose a threat since she has nothing to lose?'74 In its place the woman's relation to the image is one of'over-identifica­tion', one that is founded on the absence of a distance between seeing and knowing. Such a position is finally untenable precisely because it fails to confer on the spectator the distance needed to read the image adequately.

It follows from the above that female spectators have two principal options: either the assumption of a masculine position; or the assump­tion of a passive or masochistic position through identification with a female character. We shall consider each of these in turn.

The principal theorist of the masculinisation of the female gaze has been Laura Mulvey, who returned to and expanded on her earlier position in response to criticisms. According to D. N. Rodowick the only place in Mulvey's scenario for the female subject was as a negativity defining castration.75 Linda Williams similarly objected that the concentration on the male look at the woman left no place for women's own pleasure in looking.76 Mulvey responded by arguing that Hollywood masculinised its female spectators in offering them male points of view and male identifications. Indeed, there was nothing specific to cinema in this, for across a whole range of folk and mass culture the grammar of narrative 'places the reader, listener or spectator with the hero'.77 Through such identifications women can enjoy the freedom and control typically given to the hero by narratives. The hypothesis could be supported by Freud's own later writings on women, which propose that, because feminity is gained through repressing the masculine tendencies of the phallic phase, netjrosis in women is often to be explained in terms of the irruption of this repressed material. What Hollywood offers the female spec­tator is a socially sanctioned access route to her repressed masculinity. And there is a reinforcement of this process of masculinisation through the parallel, already noted by Freud, between the ego and the hero of any narrative. The phrase 'Nothing can happen to me', so succinctly expressive of the ego's sense of superiority and invulnera­bility, finds its narrative correlate in the typical hero's passage through the text. In identifying with such a figure and confirming the ego's fantasies, the female spectator is habituated to transsexual identifica­tions. Though the gain is that of the reactivation of the fantasy of 'action', which a proper femininity represses, it is at the cost of a certain uneasiness at violating patriarchal precepts.

Mulvey's version of the female spectator, however, commanded less widespread assent than her theorisation of the male spectator. One alternative version was that provided by Mary Ann Doane, at least over the scopic field sustained by so-called women's films. In analysing a corpus of films addressed to women (e.g. Rebecca., Suspi­cion, Gaslight, The Two Mrs Carrolls, Caught, Possessed, Secret Beyond the Door — tides that in themselves are revealing), she found that the spectatorial options were not limited to either a narcissistic identifi­cation with the woman as spectacle or a transsexual identification with the male hero. Instead, these films summoned up an interactive process between the text and the spectator that could be best comprehended in terms of the third stage of the fantasy ‘A child is being beaten', where the woman/girl no longer figures as a participant in the scenario but as a spectator. Such films turn on 'masochistic fantasy instead instead of sexuality’. 78 In them women are de-eroticised, functioning not as spectacle to be looked at but as protagonists in masochistic scenarios.

Of the two forms these tend to take, one has the woman as the agent of the gaze investigating a secret whose solution entails an act of aggression against her (paradigmatically, the locked room where her husband is planning to murder her), as if 'the woman's exercise of an active investigating gaze can only be simultaneous with her own victimisation'.79 In the other the woman is afflicted by an illness, the object of the medical rather than the erotic gaze, and constituted within a medical discourse that seeks to tell her story through interpre­tation of her symptoms. Often blind or mute, she must wait for a man to disclose her truth through medical or psychological discourse. In both kinds of scenario the effect is to desexualise the woman's body, and concomitantly to address the female spectator in such a way that she 'loses not only her sexual identity in the context of the scenario but her very access to sexuality3, so recapitulating the de-eroticised, specular stage of the female version of 'A child is being beaten'.80

These considerations, however, were part of a larger debate around the work of the theorists of the apparatus, notably Baudry and Metz. While it was generally agreed that they had performed a valuable service in shifting the emphasis from the reproduction of objects to the production of subjects, serious doubts existed about the implica­tion of their work for the female spectator. So inseparable was the machinery of image reproduction and projection from the psycho-perceptual machinery of scopophilia, identification and fetishism, that the place of women in such a system must be deeply problematic. In their account there were two levels of exclusion: from representa­tion and from spectating. Both of these conspired to make 'the very idea of a feminist filmmaking practice seem an impossibility" because 'the simple gesture of directing a camera towards a woman has become the equivalent to a terrorist act'.81 Given this situation, one strategy, exemplified by the films of Peter Gidal, has been systemat­ically to exclude all images of woman on the grounds that they could only partake of the dominant system of meanings. Yet the very fact that this strategy compounds the exclusion of women makes all the more evident the impasse film theory had reached.

In consequence women theorists became increasingly critical of the concept of the cinematic machine. Drawing on Freud's remark that all complicated machinery and apparatus occurring in dreams stands for the male genitals, Constance Penley suggested that the apparatus as conceived by Baudry and Metz was a 'bachelor machine' with a characteristic 'bacheloresque emphasis on homogeneity and closure'.82 Their very mode of theorising effectively closed off ques­tions of sexual difference. For Joan Copjec, the apparatus thus con­ceived was a machine to defend against the alienation and division experienced in the symbolic and hence was a denial of the sexual difference inscribed in the symbolic. Any conception of the apparatus that meets the spectator's demand and fixes him in a secure and unified identity runs counter to the whole drift of Lacan's thinking, and therefore represses those aspects of it most important for women. The only role for women, given the existing theories, was to adopt strategies of subversion: Mary Ann Doane proposed one such strategy at the level of representation: the masquerade. In flaunting herself, in producing herself as an excess of femininity, the woman can reveal that 'it is femininity itself which is constructed as a mask - as the decorative layer which conceals a non-identity1, thus challenging the iconographic patterns that function as a support for the male gaze.83 At the same time, by opening up a distance from the female image the masquerade allows it to become controllable, readable, producible by the woman.

More promising, by and large, than subversion was the prospect of finding an alternative theory of the institution of cinema. What was needed was a mode of theorising that would retain the radical implications of Lacan's notion of the complex constitution of both subject and object through discourse, but would avoid the phallocen-trisrr implicit in Lacan's thinking. Just such an ideologically accept­able, de-phallicised recasting of the relation of subjectivity and dis­course was to be found in the work of Foucault.

By adopting Foucault's conception of power, whereby discourses produce domains of objects and modes of subjection, Copjec and Doane advanced a theory of femininity as constructed within discur­sive practices. Thus, Doane writes: 'Femininity is produced very pre­cisely as a position within a network of power relations.'84 And Cop­jec: 'Patriarchy can only be an effect of a particular arrangement of competing discourse, not an expressive totality which guarantees its own self-interest.'85 For Doane, referring back to her discussion of fetishism and voyeurism, the distinctive closeness and 'presence-to-itself of femininity is not, as some have supposed, an expression of some essence but is rather the outcome of the place women are cul­turally assigned. And for Copjec the consequent need was for an analysis of how the multiformity of sexual difference and subject position is related to particular discursive formations and practices.

The work of Foucault was also to provide support for those who believed that psychoanalysis, despite various attempts to feminise it, was indissociably complicit with phallocentrism. Like any other dis­course, Foucault said, psychoanalysis was grounded in nothing beyond historical contingency, and therefore could not be judged in terms of any supposed truth, but must stand or fall on the basis of its effects. Its theory and practice amounted to a discourse on sex that was necessarily implicated in power relations. As to whether it had done anything to question or redefine power, his conclusion was that it had not. But there were possibilities for resistance to the subjecting alliance of discourse and power, and in his less guarded moments he suggested that any theorisation of this must start from the body. The suggestion was taken up by feminist film theorists, who sought to re-theorise the relation of the body and discourse in such a way as to do justice to feminine specificity while avoiding essentialism. A conception of the body was necessary, said Doane, 'in order to formulate the woman's different relation to speech, to language'.

One aspect of this reassertion of the feminine body, indeed of the reintroduction of a concept of the body into psychoanalysis, was the adoption by Doane and Copjec of the notion of anaclisis. This term had been used by Freud to designate the way the infant's sexual drives trench on its ego or self-preservation instincts, an obvious example being provided by the oral phase, when the erotic pleasure of sucking the breast is associated with satisfaction of the need for nourishment. Only later does the detachment of sexuality from the bodily function occur, after the child has come to want the secondary pleasure inde­pendently of the original somatic need. Taking up this idea, Copjec suggested that the deviation of the drive from the instinct was caused by the introjection of 'a scene of satisfaction into the subject5.87 Although the body is not the cause of the psyche it nonetheless has a role in structuring it. Likewise Doane argued for the irreducibility of sexuality to bodily function, but also for the latter acting as a support for the former.

Copjec and Doane also both referred to ideas elaborated by women working within or proximate to a Lacanian framework that had been treated with suspicion by the anti-essentialists. Julia Kristeva's linking of discourse with its pre-linquistic somatic precursors (this and other aspects of Kristeva's work are discussed in chapter 7) and, of greater immediate relevance, Luce Irigaray’s respecification of the feminine were called on by Doane in support of her position. For Irigaray, because of the phallomorphic tendencies of all existing theories of subjectivity, there was the need for a break with them, but in a non-essentialist direction. Starting from the specificity of the female body, in particular the multiplicity of its erotogenic zones and the nature of female genitals, whose lips constitute a constant and unforbiddable source of mutual embrace and self-touch, Irigaray proposed that female sexuality was plural, non-unifiable and could not 'be subsumed under the concept of subject’.88 And mirroring this sexuality was an equally non-masculine relation to language, at once polyvalent, plural and free of the restrictive insistence on identity of patriarchal law.

While many feminists considered Irigaray to have lapsed into the very essentialism she sought to avoid, Doane maintained this was altogether too dismissive. It was possible, she held, by using the notion of the body as a support or 'prop', to rethink the relationship between the female body and signifying processes in such a way as 'to define or construct a feminine specificity (not essence)' and 'to provide the woman with an autonomous symbolic representation'.89 There could be no question of some natural feminine body directly finding expression in a transparent medium. For one thing, the ideological complicity of 'the natural' was such as to rule it out; for another, the body was always written, coded, a function of discourse, with its sexuality implicated in language. Through these qualifications to Irigarays position the hope was that a middle course could be steered, avoiding equally a reductive essentialism and an anti-essen-tialism where the body either vanished or was simply a derivative of discourse.

Doubts about this strategy were expressed by Constance Penley. It was, she felt, bound to reproduce the difficulties associated with the essentialist position, where identity and difference are established before they can be adequately questioned. In particular she doubted the value of the concept of anaclisis for these purposes, in that any attempt, however indirect, to derive gendered sexuality from the body endangered the uncompromising insistence of psychoanalysis that sexuality is an arbitrary identity imposed by convention. The law of sexual division requires that everyone take up a position in relation to the phallus, a non-biological entity. Hence, the concept of anaclisis effaced 'the difficulty of femininity as a sexual position or category in relation to the symbolic'.90 A far more effective counter to the maleness of the cinematic apparatus than the reintroduction of the female body was through the concept of fantasy, which offered a way of preserving and accounting for sexual difference without pre­determining what any given individual's sexual identity should be. How any one spectator relates to a filmic narrative depends on their unique pattern of desire, with the only fixity coming from the formal masculine and feminine positions as defined by the fantasy. By giving an explanation variously of the spectator's desire of the image, the fantasmatic relation to that image, including a belief in its reality, and his or her multiple and changing identifications, the theory of fantasy could retain a notion of the cinematic institution while 'con­structively dismantling the bachelor machines of film theory’.91

There is, finally, the most fundamental problem of all: the fact that psychoanalysis is founded on the discovery of the unconscious. Since the implications of this are so far-reaching we shall leave discuss­ion of it until our concluding remarks, when its significance for the entire film theoretical project will be easier to gauge.