FOUR

The Perversion of I/Eye
in Un Chant d’amour

The indexical pull of metaphor in Genet’s novels parallels the indexical focus of the image in Genet’s film, Un Chant d’amour. The cinematic image is both an icon and an index: it produces a perfect copy of its object, and points both to the former presence of the object before the camera and to the place of the spectator’s look behind the camera eye. While dominant, narrative cinema produces films which mask the work of reference and subject-address in film discourse in order to assure the spectator’s immersion in the plenitude of the visual field, Genet’s film draws attention to the spectator’s implication in the events on screen.

Since cinematic representation is not simply a signifier of something but a signifier for someone, the question of cinema is inseparable from the question of performance, a process which determines the subject’s implication in social and ideological discourses. As Stephen Heath has said,

Heath identifies film performance with the subject’s psychological implication in narrative, a system of codes for ensuring closure between the “I” of the enunciation and the “I” of the enounced in the imaginary-symbolic realm. Heath, following the example of Christian Metz in “The Imaginary Signifier,”² situates the spectating subject in the framework of Lacanian psychoanalysis. In Lacan, the subject of the “mirror phase” is originally divided into a subject which speaks and a subject which is “spoken” in language.³ The imaginary integration of these two faces of the subject occurs by means of the subject’s participation in discourse, a symbolic medium in which the subject’s projections are held and returned in the form of a coherent image. Cinema is a type of social institution in which the subject’s desire is staged; narrative is a socially determined structure which regulates and perpetuates that staging. Heath says,

In Heath, “classic narrative film” performs the relation between representation and the subject in cinema, while non-narrative films or filmic figures which disturb narrative closure are not, in the first instance, a function of film performance. Though Heath suggests at the end of his discussion that certain avant-garde practices, rather than destroy performance simply perform “differently,” he tends to divide cinema into classical narrative performance and its “other,” namely, transgressive, avant-garde practice.

By defining performance as the act of being something for someone else, a function of the subject’s moment by moment construction in discourse, we manage to transcend the boundaries separating dominant cinema from its others. The point of the previous chapter was to show that performance is inseparable from the movement of figures in discourse, rather than a kind of master code structuring specific styles or modes of representation to the exclusion of others. Likewise, all kinds of films engage the subject’s desire to identify with the fiction, but certain film practices, including Genet’s, place in question the terms of the contract binding the spectating subject to the voice of authority in dominant film discourse. It is in this framework that my analysis of performance in Genet’s novels leads directly to questions of film performance raised by his only original film, Un Chant d’amour.

Jean Genet has been associated with the cinema in one way or another since the 1950s. His film projects include directorial collaboration on scripts he did not write, film adaptations of his plays, and an original screenplay he did not direct. Genet directed Nico Papatakis’s film Les Abysses in 1963. Joseph Strick filmed Genet’s play The Balcony in 1963, and Christopher Miles filmed The Maids in 1973. Genet wrote the script for Tony Richardson’s Mademoiselle in 1966, and his poem, “Le Condamné à mort,” is the basis of a film by Albert-André Lheureux, entitled, “Possession du condamné” (1967).

Un Chant d’amour, a silent, black and white film made in sixteen millimeter and lasting thirty-five minutes, is singular in that it is neither an adaptation nor a collaboration. Though Jean Cocteau assisted Genet with the technical aspect of the film, as Jean Mitry says, Un Chant d’amour is “entièrement réalisé par Jean Genet.”⁵ Moreover, unlike the other films in which Genet was engaged, this film transcribes into cinematic terms Genet’s ideal of textual performance along the lines of the index, the sign function which traces relations of reference and subject-address in discourse.

Un Chant d’amour has been generally avoided in scholarly studies of Genet, perhaps because of the pornographic aspects of the film, which caused it to be banned in France and the United States for some twenty years. The nature of the film medium and the disposition of the spectator in front of the film might explain why scholars discuss Genet’s quasipornographic literature and not his film. Film images present characters “in the flesh.” Therefore the explicit representation of sex organs on screen is more susceptible to the label “pornographic” than the same representation in literature. This very statement about the difference between cinema and literature was used to uphold the court decision to ban Genet’s film from screening in Berkeley, California, in 1966:

Moreover, the potential for spectator identification is greater in cinema than in literature or theater, because of the difference between the disposition of the spectating subject of cinema and the reading subject or the theater spectator.⁷ The film spectator is isolated, passive, his attention more or less captive in the darkened theater. The reader can emerge from the book; the theater spectator can recognize the difference between the on-stage fiction and off-stage reality by merely looking beyond the sets to the spaces back-stage. Also, as Metz points out in “The Imaginary Signifier,” it is the very absence of the reality traced on the film image which invites the spectator’s projections into the film, thereby determining him or her as the locus of cinematic discourse, a voyeur constructing the film according to his or her own desire.⁸

Ironically, the little critical acclaim given Un Chant d’amour in the United States, primarily by American avant-garde filmmakers and critics, commits a type of censorship by idealizing the work on esthetic grounds. Justifying the pornographic aspects of the film in the name of high art, critics such as Jonas Mekas and Anaïs Nin ignore the perversely ironic nature of Genet’s eroticism, which victimizes rather than gratifies the spectator.⁹

There is a movement in Genet’s film between the explicit representation of sexual perversions, including fetishism, onanism, and voyeurism, and the displaced representation of the same in fairly commonplace symbols, such as a flower being plucked or a pistol being shoved into a prisoner’s mouth by a guard. As we discovered in the novels, Genet’s original contribution to erotic literature consists in neither the explicit nor the symbolic representation of sexual perversions, but in a perversion of the subject’s performance of dominant discourse. Genet eroticizes the cinematic signifier, as it were, engaging directly the question of the spectator’s desire of and in the film. In Un Chant d’amour he frustrates the spectator’s voyeurism by perverting normal relations between the subject of the enounced and the subject of the filmic enunciation.

Un Chant d’amour can be divided into five segments: an opening scene which is to be read in symbolic terms and which prefigures the rest of the film; two more or less realistic scenes in a prison depicting the actions of the main characters, two prisoners and a guard; a departure from the prison setting in fantasy sequences; and a closing which recapitulates the various motifs of the film.

In the opening a prison guard walks forward, looking up at something off-screen. Cut to a medium close-up of two barred (prison) windows. Out of the window on the left, a hand attached to a virile arm swings a garland of flowers to another hand attached to a virile arm hanging out of a window on the right. After several attempts the hand on screen right fails to catch the garland.

It should be noted that the garland is an important symbol in Genet’s novel Miracle de la rose. As I discussed in chapter 3, the transformation of Harcamone’s chains into a garland of white roses is a precondition of the narrator’s symbolic intercourse with him. One is also reminded of the systematic representation of bodies by arms and hands in the novel, Pompes funèbres. In Un Chant d’amour, as in the novel, the isolation of arms and hands from the body symbolizes the imaginary fragmentation of the body in castration anxiety, a theme which links homosexuality to figures of desire in Genet’s work. These metonymies also invite the spectator to imagine what is not directly represented on screen, the absent bodies, thereby projecting his or her imaginary into the film.

The symbolism of the opening sequence is developed throughout the film in the themes of alienation and frustrated desire. The prison walls, penetrated by virile arms attempting union by means of arms, hands, and the garland of flowers, constitute paradigmatic elements which are later replaced by straws passed through holes in the wall, smoke passed through the straws, and a flower passed between the mouths of two lovers. All symbolize the unfulfilled desire for homosexual possession.

Most of the second segment consists of cuts from a well-lighted cell with a fair prisoner in a white T-shirt, to a darker cell with a swarthy, hairy prisoner in a dark uniform. The fair man is dancing slowly, caressing a tattoo of a woman on his upper arm with one hand and his crotch with the other. The man in the dark cell attempts to contact his neighbor. He knocks on the wall, then penetrates a small hole in the wall with a straw from his mattress and blows cigarette smoke into the fair man’s cell. The latter ignores these signals. The swarthy man, angry and frustrated, seals up the wall with bread paste and masturbates against the wall. The fair man next door falls onto his own cot in orgasm. Meanwhile the guard from the opening shot has entered the prison: he peeps in on the dark man’s climax, then spies on the fair man.

In the third segment of the film, the guard walks to other cells and spies through peep-holes in the doors at various half-naked prisoners masturbating. In these shots the prisoners gaze directly into the camera, evidently taking pleasure in being placed on exhibit. In the first shot, a white man dressed only in a striped T-shirt smiles into the camera. In the next cell, a black man in white pants masturbates in rhythm to a calypso he performs for the camera. He finally collapses on his cot in a climax of pleasure. Their gazes into the camera are intercut with close-up shots of the guard’s eye penetrating the peep-holes of the cell doors and long-shots of the guard, obviously excited, walking from one cell door to the next in the hallway.

The guard, returning to the two original men, peeps in on the fair man caressing his own foot and on his swarthy neighbor. The fair man, tired and frustrated by his solitary eroticism, knocks on the wall separating him from the swarthy man. The latter responds by caressing the wall. The fair man takes a straw from the mattress and penetrates the hole in the wall, inviting the other man to pass him smoke from his cigarette. A series of shots alternating between the two cells represents the two men finally making contact: the fair man inhales the smoke exhaled through the straw by the swarthy man.

The guard, meanwhile, has been spying on their “intercourse,” visibly aroused and grabbing his crotch. The swarthy man, observing the guard’s look, signals a warning to his neighbor to stop their communication. Deprived of his scopic gratification, the guard acts out his sadism by entering the cell of the swarthy man and beating him into submission. This scene is intercut with two sequences depicting the inner visions of the guard and his victim.

It is worth noting that the motif of the voyeur peeping through keyholes of doors along a hallway figures prominently in Cocteau’s Le Sang d’un poète. This motif is significantly transformed in Genet’s film, and does not put into question the authorship of Un Chant d’amour.

The guard’s scopic relation to the prisoners defines a perversion of the erotic drive toward union or continuity, since the scopic drive demands a physical distance between the voyeur and the object of desire. Freud says that voyeurism not only displaces the sexual drive from the genitals to the eyes, but is accompanied by sadism, since the object of desire is subjugated by the controlling look of the subject.¹⁰ The prisoners are trapped in their desire, captives of the guard’s look. Solitude characterizes all of the perversions shown in the film. However, none of the other perversions, such as the fetishistic substitution of the sex organs by feet, straws, and flowers, or the onanism of the prisoners, contains the element of sadism implicit in the guard’s voyeurism.

Where the other perversions are substitutes for sexual contact—means of transcending the distance separating the prisoners—voyeurism is a drive based on the distancing (the “garde,” or psychic defense) of the voyeur from the erotic object.¹¹ The voyeur engages the physical presence of the other only to keep him at a distance, behind the wall, beyond the door or window. The complicity of the onanists who look complacently back at the guard as he peeps into their cells defines a type of masochism intimately related to scopophilia: exhibitionism.

The fourth segment consists of two inner visions, the guard’s and the prisoner’s. The guard’s fantasy begins when the guard closes his eyes and the image fades to black. One is reminded of the association of blindness, or closing the eyes to physical reality, with the generation of inner visions in Genet’s novel, Notre-Dame-des-fleurs. In the film the guard “sees” himself in various erotic postures with anonymous nude men, including the swarthy prisoner. In the guard’s fantasy, several motifs represented early in the film recur: a garland of flowers is swung into the image from the left side of the screen, a hand from the right attempts to grab it. Two men chew on a flower uniting/separating their mouths. A third scene refers directly to the guard’s violence toward the swarthy prisoner: a symbolic rape is performed as the guard forces his gun into the man’s mouth. This and the other scenes of the fantasy are played against a black backdrop in an undefinable space. The guard’s fantasy ends and we are returned to the prison cell in which the guard is beating the swarthy prisoner to his knees.

The prisoner’s fantasy begins as an image of him looking up at the guard fades to black and cuts to a country scene. In this shot, he and his fair neighbor are free, playing chase in a woods. Attached to the fair man’s belt is a garland of flowers from which he offers a flower to the dark man. This action is a variation of the rose-plucking figure in Miracle de la rose, which I discussed in chapter 3. In the novel the narrator’s gesture symbolizes his desire to rape Harcamone. In the film the object of desire offers the lover his “flower.” The film sequence ends as the swarthy man caresses his friend, lying on the ground. No nudity or explicit sex are included in this scene. The prisoner’s fantasy, ironically, represents the ideal of “normal” love in the film, homosexual desire. This sequence ends and we are returned to the prison scene. Ashamed and defeated by the guard, the swarthy man throws himself on the cot and buries his head in the mattress.

The final segment takes up and resolves the various motifs of the film. The guard walks away from the prison as he entered the film: looking up at something as he did in the opening of the film. This time he looks behind him. Inside the prison, the swarthy man knocks on the wall, in hopes of communicating with the prisoner behind it. Outside, the arms continue to play toss and catch with the garland of flowers. The hand on screen right catches the garland.

The composition of the images themselves reiterates the theme of unresolved desire in terms of formal oppositions—between left and right of frame, left and right sides of the prison wall, inside and outside of the prison cell, and in the contrast of light and dark. This last contrast dominates the illumination of every shot. The fair prisoner wears a white shirt and his cell is well lighted; his swarthy neighbor wears a dark uniform and his cell is dim. The contrast of light and dark in the scenes of exhibitionism and voyeurism reiterates the theme of alienation and solitude associated with those perversions. The first prisoner wears a black and white-striped T-shirt; the second, a black man, wears white trousers.

The strong contrast of light and dark in the scenes depicting the guard’s inner vision reiterates, in formal terms, the perverse nature of the guard’s desire. He sees himself in various erotic postures with men, but in each one physical union is either mediated by an object, such as a flower or a gun, or defines a perversion of heterosexual intercourse, such as fellatio. The motif of the two arms swinging and attempting in vain to catch the garland recurs here, again against a black backdrop. For the guard, intercourse with other men is impossible, since his desire to dominate replaces love with sadism.

The contrasts of light and dark and left and right of frame are resolved in the swarthy prisoner’s fantasy, which represents the consummation of the erotic ideal of the film, homosexual love. The pastoral setting of the scene and the absence of pornographic detail serve to legitimize this traditionally “perverse” relation between men. Here, flowers are a natural part of the diegesis, rather than simply sexual symbols. In the film, as in the novels, the explicit representation of sex acts serves not to add realism or even sensationalism to erotic discourse, but to implicate the spectator in the perverseness of representation. The innocence of the prisoner’s fantasy denies the spectator voyeuristic participation in the characters’ desire.

It could be argued that this scene depicts a homosexual version of romantic love. For the only time in the film the contrast of light and dark is resolved in a spectrum of grey tones. The barriers separating the two men in reality (in prison) have disappeared, and a wealth of realistic detail relieves the image of the sharp formal conflicts which characterized earlier shots of flowers and prisoners.

The symbolic representation of eroticism and frustrated desire serves to expand the meaning of desire in Un Chant d’amour beyond the purely sexual, physical realm to an esthetic and psychological one. This kind of displacement permits our reflection on the eroticism of poetic discourse and the place of the spectator in that eroticism. It is no coincidence that the pornographic parts of the film are organized according to a tightly woven cross-cutting montage between the voyeur looking and the object of his look. In Genet the pornographic is inseparable from the manipulation of intersubjectivity (I/you, he/she) in discourse. Genet dares the spectator to desire the film, then perverts this desire by implicating the spectator in the sadism of the guard. It is precisely in this sense, and not simply with regard to the explicit representation of sexual perversions on screen, that Un Chant d’amour is an exercise in perversion, the perversion of the cinematic subject or “I” by means of displacements of the camera “eye”.

Christian Metz and Jean-Louis Baudry argue that in order to discuss the eroticism of spectating we need to redefine an ontology of cinematic representation in terms of the spectator’s desire to identify with the film, rather than focussing primarily on the eroticism of the filmic signified.¹² In defining the “essence” of the cinematic medium, traditional film theories privilege the reproduction of physical reality in the photographic image and the simulation of movement in the projection. This tradition stems from the criterion of mimesis shaping literary and artistic criticism from the classical period: works of art were (and still are) judged, classified, and censored with reference to their modes of and techniques for imitating reality. André Bazin goes so far as to say that the invention of the camera, and with it the ability to mechanically reproduce visual reality, relieved the other arts of their aspiration to mimesis:

The reality principle that has dominated film theory from the earliest days of the film industry focusses on the film object and the mimetic relation between cinematic representation and its object. As Baudry points out, the technical achievement of the camera to reproduce physical reality on film only partially explains the famous “illusion of reality” experienced by the spectator.¹⁴ The spectator is disposed in advance to believe or desire to believe that the projections of light and dark on the film screen represent “reality.” In other words, the illusion of reality can also be defined in terms of the disposition of the spectator. The current psychoanalytic bent of semiotic theory aims at redefining the “essence” of cinema in terms of the dialectic interaction of the spectator’s own psychological projections with the mechanical projection of recorded reality on screen. According to this approach, advanced by Baudry, Metz, and others, the illusion of reality resides not simply in the technical ability of the camera to record physical reality, but also in the play of presence and absence in the representation itself, which implicates the spectator’s desire in the film.¹⁵ This approach also shifts attention from the iconic aspect of the film image to its indexical function.

Umberto Eco refines Peirce’s classification of signs into icons, indices, and symbols (Saussure’s “sign”) by replacing the notion of “sign” with “semiotic function,” saying that an element of discourse may have a mimetic function or an indexical function without being a “sign” per se.¹⁶ This change permits explaining the semiotic function of nonlinguistic systems such as cinema, composed of both indexical and iconic aspects. Eco’s “trace,” as exemplified in hoofprints discovered in a roadway, illustrates the semiotic specificity of the cinematic signifier, which not only resembles its object more perfectly than any other type of representation, but points to the former presence of the physical reality it represents. The film image is thus an indexical icon, functioning within cinematic discourse as a trace of an absent reality.

In “The Imaginary Signifier,” Christian Metz initiated the current shift of film theory from esthetic and strictly semiological concerns to the psychoanalytic question of the subject of cinema. Metz, modelling his ontology of cinema after Jacques Lacan’s theory of the subject of the mirror phase of psychological development, focusses on the spectator’s implication in the look of the absent camera. The indexical nature of the cinematic image permits Metz to define cinematic representation as an “imaginary signifier,” a signifier of absence, which invites the spectator to invest the film with his own desire:

Metz employs the Lacanian division of psychological experience into the imaginary (the configurations of the individual’s desire for the object, based on the movement between projection and identification), the symbolic (the displacement of the imaginary onto secondary, objective representations, such as images), and the real (the social situation of the subject and his entourage).¹⁸ As a signifier or representation, the film is a symbolic apparatus, a socially determined system of semiotic codes. As a signifier of absence (a trace of physical reality), the film is an imaginary apparatus, inviting the participation of the spectator. As a physical, technical object, the film is a real apparatus. Metz points out that while all experience tends to oscillate between these three moments, the experience of cinema leans heavily to the side of the symbolic-imaginary, or the spectator’s identification with the film by means of psychological projection.

Psychological projection is a defense mechanism which displaces an unconscious conflict (in the imaginary) onto the object of desire.¹⁹ Through projection, the original relation to the mother (Melanie Klein’s “good object”), is transferred to subsequent objects in the subject’s erotic life, including displacements such as art objects (the realm of symbol formation). In Klein’s corollary to the Freudian notion of projection, the sequel to projection is identification. Once invested with the spectator’s projections, the film screen becomes a sort of mirror for the subject, a reflection of the self.²⁰

The cinema industry, Metz’s “capitalist apparatus,” strives to produce films which will exploit the spectator’s desire to merge with the film in a quasi-hallucinatory state. Though some films permit greater degrees of identification and scopic gratification than others (to this extent they are “good objects”), the industry does not promote films which remind the spectator that “it’s only a film” and thrust him back into the reality of the moment (to this extent they are “bad objects”).

The cinema is not simply a technological phenomenon, a mechanical apparatus for reproducing physical reality, nor simply a subjective phenomenon, consisting of the spectator’s projections into the film, but a “series of mirror effects [between the film screen and the spectator’s psychological screen] organized in a chain.”²¹ Furthermore, as Metz says, the institution of cinema is primarily an erotic apparatus, preserving and perpetuating the ideal of the film as “good object,” as guarantee of the subject’s closure with the Other in the imaginary realm.²²

There is a sexual meaning to this movement of the subject in the film, a meaning turning around the figure of castration. The child’s entrance into the symbolic order, the awakening of the drive to project the lost object of desire into substitutions, occurs at a stage in psychological development when the (male) child recognizes the sexual difference between the parents and between (him)self and the mother. The recognition of difference provokes castration anxiety, the fear of one’s own cleavage as a result of the real separation from the Other.²³ The Oedipal scenario is played out repeatedly in adult life as the endless fluctuation between the constitutional division of the subject from its other, and the imaginary closure of the subject with its other and Others across the imaginary/symbolic realm.

As Stephen Heath points out, in Lacanian theory the term “castration” covers two distinct regimes of meaning. It describes both the division of the subject in the symbolic order and the real division between the sexes.²⁴ Dominant discourse reduces the dual notion of “lack” to the second meaning, the lack of a specific organ: the penis. Lack, in dominant discourse, is modelled after the biological difference between male and female.

Cinematic representation perpetuates the monolithic order of castration. As substitute for the erotic object, the film performs as a fetish. In Freud the fetish is a figure for the lack, a substitute for the penis the male subject projects into the mother’s body to make up for an organ she is presumably missing.²⁵ By projecting the “same,” the phallus, into the Other (the image of woman), the male subject defends himself against the fear of his own castration. The fetishistic function of cinema demands that the spectator be confirmed as “whole” subject through projective identification with the fiction.²⁶ Masking the fragmentation of the enunciation (including the film’s articulation into shots and the disjuncture between sound and image tracks) through semiotic operations in the discourse, classical narrative films suture the subject’s own lack and confirm his recognition of (identification with) himself in the Other, the film. In this sense discursive practices in dominant cinema repress the reality of castration as the division of the subject in the imaginary/symbolic order.²⁷

Next, dominant cinema poses the solution to the original division of the subject in terms of the organization of subject-address in narrative discourse. The positions of looking and being looked at parallel the opposition of male and female in the diegesis. The man is usually the voyeur, the woman the object of the look, whose own ability to look and to shape discourse is lacking. Projecting the lack into woman’s body guards against the (man’s) recognition of difference, of bisexuality, as condition of possibility of the constitution of the subject.²⁸

To the extent, then, that films perform as fetishes, they guarantee a male order of the symbolic, an ideology of the “same” turning around the phallus (and its lack). Moreover, by representing sexual difference as the opposition between male voyeur and female object of the look, classical narrative films confirm the omnipotence of the male gaze, the voice of authority in cinematic discourse. Such cinematic practices in turn define the spectator—man or woman in reality—as “he” in the imaginary/symbolic order. For this reason, women often take pleasure in films which represent the sexual exploitation of other women.

For the semiologist, cinema represents the institutionalized practice of perversion, controlled by the technical/economic/political structure of the film industry. In Un Chant d’amour Genet underlines the fundamentally perverse nature of dominant cinema by representing a voyeur, isolated prisoners of desire, and fetishistic displacements in the image. This diegetic (iconic) representation is but an overdetermination of the more crucial manipulation of (indexical) relations of subject-address in the discourse. By displacing the “normal” practice of scopophilia from the imaginary/symbolic realm to a conscious reflection upon that practice, Genet prevents the spectator’s imaginary closure with the film, reminds him of his real separation from the erotic object, and disturbs the fetishistic function of the film. This brings lis back to the ideological threat posed by Genet’s film.

If the cinematic apparatus is essentially perverse, then what constitutes a more or less “perverse” film discourse? Un Chant d’amour is “perverse” not simply because it represents sexual perversions on screen, but because it displaces the normal practice of cinematic scopophilia from the realm of the imaginary to the realm of the real. Semiotic practices in the discourse upset the spectator’s identification with the film and transform the film into a “bad object.”

It could be argued that the critical and legal resistance to Un Chant d’amour mentioned earlier determines in advance the film’s undesirability, its stature as “bad object.” However, the film has been canonized in film d’art circles, as evidenced in its inclusion in the highly selective collections of the Anthology Film Archive and the Museum of Modern Art in New York. The film’s lack of commercial distribution indicates that “mainstream” audiences do not seek out the film in the first place; therefore, we can only discuss the privileged audience of museums and archives. For this narrow public, Un Chant d’amour constitutes à priori a “good object,” since it fulfills certain esthetic priorities of the art film crowd. An example of privately financed, independently made “counter-cinema,” Genet’s film explores the symbolic and poetic possibilities of cinematic representation, expresses a personal vision at the expense of commercial appeal, celebrates one-man authorship, and so on.

The semiotic practice of perversion in Un Chant d’amour derives from a disturbance of the classical narrative closure between primary and secondary identification in the spectator (the looks of the camera and of the characters, respectively). Primary identification can be explained by analogy with the inscription of subjectivity in linguistic discourse, along the lines of indices defined by linguistic code, the personal pronouns “I” and “you.”

As Emile Benveniste explains, subjectivity is not a metaphysical essence but a function of discourse.²⁹ “I” is uttered, therefore I am. There can be no “I” of discourse without its corollary “you,” the receiver of the message. The implication of “I” and “you” in turn permits the eventual identification of the subject with “he” and “she” in narrative discourse. For Benveniste, this convention puts into play shifts between the linguistic categories of person and non-person.

Benveniste’s observations about linguistic discourse can be used to describe the inscription of subject-address in cinema, and the shifts between primary and secondary identification in Genet’s film. The camera eye, defined by the parameters of the frame on the screen, resembles the speaker of linguistic discourse, the “I” and organizer of cinematic discourse, with one important difference: the positions of I and you do not shift between two subjects in cinema as they do in a dialogue between two speakers. The spectator’s look is implicated in film discourse according to psychological identification with the camera I/eye: cinematic discourse cannot exist without a viewer to perceive and read it. This initial inscription of the spectator in the film defines primary identification.³⁰

Primary identification is a condition of secondary identification, through which the spectator is implicated in the looks of the characters. In the cinema the shift from primary to secondary identification entails the identification or matching of the look of the camera with the look of the character, achieved primarily but not always by means of crosscuts joining shots of a character looking and the ostensible object of his look. Eye-level matching and diegetic contiguity between the character and the object of his look support the logic of the cross-cut.³¹

In classical narrative films, we are not usually disturbed by the shift from primary to secondary identification in the cross-cut, because the look of the camera transcends any particular identity in the diegesis. This shift is similar to the movement between the discourse of an omniscient, transparent narrator of the third-person novel and the reported speech of characters speaking in the story. The omniscient narrator pretends to represent characters “speaking in their own voices.” Convention dictates that we ignore the mask assumed by the narrator to produce characters with whom we can identify.

As I discussed in chapter 2, something quite different occurs in certain postwar novels, such as Pompes funèbres, in which the narrator reveals himself as “I,” an anonymous narrating presence, who occasionally trades places with his characters and speaks through their mouths. In Pompes funèbres, the narrator, named “Jean Genet,” moves in and out of the roles of Erik and Hitler at different moments of the narrative. This kind of duality draws attention to the cleavage of the narrating subject, to the difference between “I” and “he,” person and non-person of discourse, in the conventional representation of reported speech. This cleavage of the subject of modern literature is announced in Rimbaud’s famous statement, “I is an other” (“Je est un autre”).

In Un Chant d’amour, Genet transcribes the figure of the mask into cinematic terms, creating an awareness of a first-person cinematic narrator. By this means, Genet disturbs the shift between primary and secondary identification in the spectating subject.

We identify primarily with the narrator along the lines of the dialectical implication of “I” and “you.” As narrative “I” identifies with “he,” that is, with first one and then another character speaking in the novel, the reader can only identify with the characters across the intermediary of the mask. Furthermore, unveiling the mask worn by the narrator in his various impostures interferes with the reader’s identification with the characters. If the reader is implicated in the discourse along the dialectical relation between “I” and “you,” and if “I” has a dual referent (I = he), then the reader is divided in turn, as he identifies with the narration, denied the illusion of closure with the fiction. The disturbance of the shift between primary and secondary identification constitutes a “critical space” in which the reader moves between identification with and reflection upon his place in the discourse. This kind of movement defines a major source of tension, not only in the novels of Genet, but of Samuel Beckett and Marguerite Duras, for instance.³²

First-person cinema, a genre which has been inadequately explained by recent film criticism,³³ is a problematic issue which requires careful study. Metz has said, both in early and recent essays, that the difficulty of distinguishing the camera eye from the diegesis it brings into view prohibits our discussing a narrating subjectivity or self-conscious “I” of cinematic discourse, one independent of the looks of the characters viewed on screen.³⁴ If this were true, there could be no “first-person cinema” comparable to the novel from which the term derives. Metz recently suggests that “strange camera angles” might draw attention to the fact of the camera,³⁵ but he does not examine how such self-conscious techniques change the spectator’s relation to the discourse.

In Un Chant d’amour Genet creates awareness of the camera by mismatching the looks of the camera eye and the voyeur in the diegesis in a cross-cutting figure. He also addresses the looks of the characters to the fact of the camera as a presence within the diegesis. Such figures trace the intervention of a narrating I/eye into the film and disturb the shift between primary and secondary identification in the spectator.

First, there is a recurring disparity between the distance from the guard looking to the object of his look, and the distance from the camera eye to the object it observes in the subsequent shot. Moreover, the eye levels of characters looking off-screen and of the camera looking at the object of their looks are mismatched. For instance, the guard looking up in the opening shot is seen in medium long-shot. The prison exterior is not visible in the frame. The subsequent shot of prison windows is in close-up and from a different eye-level than the guard’s look. Later, inside the prison, the guard peeps in on the swarthy prisoner, who is subsequently viewed in medium shot. When the guard returns to peep in on the same prisoner masturbating, a cut from the guard looking through the key-hole to an extreme close-up of the prisoner’s body disturbs the logic of the cross-cut. We have difficulty inferring that the prisoner’s body is seen from the guard’s point of view. In the scenes of exhibitionism, the prisoners are shown in medium long-shot, looking back at the guard spying on them. Yet we see the object of their looks, the guard’s eye penetrating the peephole, in extreme close-up.

Such interferences in the normal function of the cross-cut derive from a movement between the code (the assumption that a shot following a shot of a character looking off-screen represents that person’s point of view) and a figure of lack (the implication that the matching shot in fact represents the point of view of the camera, the locus of the absent spectator). Such modifications in the code do not disturb the continuity and coherence of the diegesis as much as they draw attention to the look of the camera as an autonomous narrating presence (independent of the looks of the characters).

In the same way that the enunciation of “I” in the novel constitutes a reflection both on the production of discourse and on the complicity of the reader in that discourse, so awareness of the camera eye in Un Chant d’amour constitutes a reflection both on the cinematic apparatus and on the spectator’s place in that apparatus. The interruption of the cross-cut by the imposing “look” of a narrating I/eye makes visible the spectator’s primary implication in the discourse as an absent voyeur.

Awareness of the camera is reiterated in the shots of the onanists looking directly into the camera. As Metz says, in ordinary cinema the complicity between the voyeur and the exhibitionist, a relationship characteristic of most voyeurism, is short-circuited, since the actors are not only absent from film representation as such, but are forbidden by convention to look into the camera, to acknowledge the presence of the camera eye and the spectator’s look.³⁶ The success of the average pornographic film derives from the spectator’s sense of anonymity and denial of complicity in the events on screen, resulting from the inscription of lack and absence in the cinematic signifier. In Un Chant d’amour, recognition of the camera as an index for the absent voyeur of film discourse, the spectator, implicates the specator in the perversion of the film.

As Metz has said, spectator identification with (desire of) the film demands that the spectator both “take himself” for the character by means of psychological projection (in the realm of the imaginary), and not (mis)take himself for the character by comparing himself with the person on screen (in the realm of the real). Thus in Un Chant d’amour, reflection on the spectator’s primary identification (with the camera) perverts the normal practice of cinematic scopophilia. The cross-cutting code shaping cinematic voyeurism opens the film to the spectator’s projections, permitting him to participate in the film qua fiction (the realm of the imaginary/symbolic). The figure of lack created by the transgression of this code, which leads to the recognition of the camera eye, invites the spectator to recognize his situation in front of the film (a return to the real) and take himself for the guard, the voyeur in the film. This movement from the imaginary to the real along semiotic operations in the discourse transforms the film into a bad object and twists the scopic drive into feelings of bad conscience.

Recognition of the camera as autonomous narrating presence causes a doubling of the reference for the look, the “I” of the discourse. We recognize that I/eye refers to the autonomous narrator and the guard at the same time. This marks a cleavage in the normal closure of person and non-person in cinematic discourse, when “I” shifts into identification with “he,” a character in the story. Since “you,” the spectating subject, is constructed by dialectical implication with “I,” and since “I” is here split into a doubled reference to subject and other, the spectator is taken up in a figure of the double, making visible the real division of the self in the symbolic order. “I” is an o/Other.

Thus, the movement between the realms of the imaginary/symbolic and the real of cinema in Un Chant d’amour “defetishizes” the film object and produces displeasure. As Serge Daney puts it:

Genet not only implicates the spectator in the voyeurism of the guard in the diegesis, but reflects on the “garde,” the scopic drive as defense against the real of castration. As Metz says, “What defines the scopic realm of cinema is not as much the distance kept [“gardée”], the defense [“la garde”] itself, (primary figure of lack, common to all voyeurism) as the absence of the object seen.”³⁸ The implication of the spectator in the dual figure of the narrating subject (I = he) is inscribed with the identification of male and female, since the guard in the film desires not women but men subjected to the position of the feminine by his sadistic gaze. In other words, the fetishistic function of the film breaks down. Rather than recuperate the missing penis by projecting it into woman’s body (creating the figure, he = [imaginary] he), the guard’s sadistic looks at men in the throes of solitary pleasure effectively castrate the men in the film. These looks open up a gap in the imaginary there where the phallus is supposed to be, creating the figure he = (imaginary) she.

Genet reiterates this twist on the conventional representation of women in the sequence where the swarthy prisoner fantasizes himself in the countryside with the fair man, plucking a flower from the latter’s garland. As Ken Kelman observes, this scene approaches a parody of the Hollywood boy-pursues-girl scenario, culminating in an ironic perversion of the literary cliché of “deflowering” the maiden, a figure I discuss in chapter 3 with regard to Miracle de la rose.³⁹

Displacements of the look, from the narrating camera to the guard and to prisoners looking back, also problematize the sexual identity of the gaze as organizing principle of the discourse. The guard’s own look and sometimes his organ of vision are the objects of the gazes of camera and prisoners. Sexuality is therefore represented in the film in terms of a constant shifting of positions between voyeurs and erotic objects (masculine and feminine being constructed thereby), rather than along the biological division between men and women in the diegesis. In this way, Genet constructs the subject of discourse around a figure of bisexuality, the spectator (s)he.

Jean Cocteau once defined cinema as the act of looking at events through a key-hole, implying that the voyeur is a metaphor for the spectator.⁴⁰ Genet characteristically twists this metaphor into an ironic reflection on the spectator’s own perverseness, on the dubious sexuality of the subject of the look, and on the sadistic origins of scopic pleasure. By violating the law of the “same” as ontological support of the cinematic image, Genet, himself a former prisoner, effectively imprisons the spectator in his own dark cell of desire. Thus the spectating subject becomes the victim of Genet’s irony. In chapter 2, I discussed similar operations in the novels which implicate the reading subject in the narrator’s irony by perverting relations between narrative voice and the voices of characters speaking in the story.

In Un Chant d’amour Genet succeeded, in a rudimentary way, in exposing the terms of film performance, that is, the moment by moment identification of “I” with “eye” in film discourse, by foregrounding a first-person narrative “voice.” By disturbing the spectator’s dreamlike state during the film, by introducing a disquieting reflection on the cinema’s predisposition to voyeurism, Genet stages the spectating subject in an ironic play of (self-)reflections. In this sense, Genet anticipated the dialectical cinema of avant-garde filmmakers of the 1960s through the 1980s, including Godard, Straub-Huillet, Duras, and Ackerman, whose work problematizes the spectator’s relation to film discourse by exposing the distance between the voice of the enounced and the voice of the enunciation. Unlike those filmmakers, who take advantage of the material disjuncture of sound and image in order to stage the division of the subject in film narrative, Genet was limited by working only with the visual track. He nonetheless transcribes into cinematic terms the figure of the “double,” originating in the novels in the form of the intrication of voices in narrative discourse (see chapters 1 and 2 of the present book). He also anticipated the conditions of representation which would characterize his “theater of the double.”⁴¹

Un Chant d’amour might be viewed as an esthetic exercise for Genet, who had, by 1950, written Haute Surveillance (1947) and Les Bonnes (1948), but had not yet realized a theatrical production of either play. Theater is not primarily text but the performance of a text. Moreover, theatrical performance includes visual and spatial dimensions which can only be imagined in narrative. In the course of adapting his notion of textual performance to the needs of the spectating subject, Genet was forced to examine problems of spectator identification and transcribe his poetic vision into concrete, plastic images.

As I discuss in chapter 5, Genet created a style of theatrical performance which exposes the double at the origin of mimesis, thus destroying the illusion that a fixed reality precedes and even transcends representation. Following Artaud, Genet viewed reality as a house of illusions, as an imitation of theatrical performance.⁴² As in the film, the staging of the spectating subject of Genet’s theater constitutes a kind of recuperation of the imaginary by the real, since the spectator participates in the performance as an actor on the stage of life.

If cinematic representation exists as an imprint of absent realities, theatrical representation exists as an illusion of total presence: the presence of the spectators and actors in a single time and space; the presence of the place of production of the representation (the stage); and the visibility of the costumes, make-up and props employed by actors to create their roles.⁴³ The polemical movement between the representation as a symbolic-imaginary apparatus (as a fiction which invites our participation) and a real, technical apparatus is crucial to Genet’s ideal of “theater about theater.”

The kinds of doublings we discovered in Genet’s novels—in the dual identities of the narrator, in ambiguous indications for narrative space and in the endless mirroring of narrative discourse and reported speech, are the bases for Genet’s ideal of theatrical performance. These doublings are possible in the novel because of the conventional separation of narrator and narrating event from characters and the narration. They are possible in the theater because of the autonomy of the verbal text from miseen-scene, the distance between the fiction and the means of representation, and so on. Such doublings are somewhat more difficult to articulate in cinema, especially silent cinema, because of the conjunction between the look of the spectator, the look of the camera, and the diegesis it brings into view. While Genet might have chosen the route of filmmaker-authors such as Duras or Robbe-Grillet, who place the spectating subject on stage by means of disjunctions between sound and image, he evidently discovered that the theater was a more powerful medium in which to explore the psychological and philosophical question of performance than the cinema. Genet made only this film before devoting himself to the theater, with its vast repertory of means for staging the subject in the drama of performance.⁴⁴