Ingmar Bergman’s Persona (1966) begins, famously, with a meta-cinematic sequence. The illuminated tips of an arc lamp flare in the darkness, and then images begin with flashing clips evoking film history and genres: life, comedy, sex, religion, theater, death. And then we get a brief scene. A boy wakes up in an antiseptic room, puts on glasses, reads, and then looks out in front of him at a large screen with a woman’s face, slightly blurred, projected on it; he reaches out and caresses it, almost lovingly but also wonderingly.

The opening sequence of Persona is a broad exercise in forms of reflexivity, moving from the technical conditions of cinema to the historical forms of its appearance (as in Fängelse (1949), Bergman remakes early cinema). But it is the boy, touching the screen, who comes to stand in for the cinematic encounter itself. It is of course partly because the face is that of Elizabeth and we know she has a son—is he reaching out toward (an image of) her? It’s also because we see him again at the end of the film, again reaching out toward the face, and again touching it, this time before the arc lamps separate and blackness recurs.

The opening of Persona is a powerful gesture of modernist cinema. It is also emblematic of a deep tendency of film theory, namely the fantasy of an identification with the image and a corresponding transformation of the spectator into the position of a child—if not an infant. Consider a recent resonant passage from Mary Ann Doane: “desire for the screen is desire for a technology that simulates the pleasure of breast, the face, and the blissful state of nondifferentiation of the infant (of the Imaginary in Lacanian terms).”[1] It may as well have been a gloss on the scene in Persona.

“I am the camera eye,” Dziga Vertov proclaimed in a fantasy of human/machine identity, and it’s a fusion (and a fantasy) that runs through classical film theory. “Our eyes are in the camera,” states Balazs; Barthes says that, in the theater, he finds himself “glued” to the image. While oft-repeated, this fantasy gets formalized in the more avowedly theoretical writings in the 1970s, especially around the articulation of a logic of identification. Asking the question “with what, then, does the spectator identify during the projection of the film?,” Christian Metz answers: “The spectator can do no other than identify with the camera.”[2] We find the same argument in Laura Mulvey’s “Visual Pleasure and Narrative Cinema,” where she organizes the three “looks” of cinema, and their gendered nature, around the fundamental identification of the spectator with the camera—and, in the theater, the screen itself.[3]

Yet this psychoanalytically inflected theory carries an implication that is not within the older fantasy. For Freud, and for those writing in his wake, identification was fundamentally an infantile drama: how infants begin to relate to the world around them, how they differentiate self from others. Metz is explicit in his characterization of identification as an infantile position, one that evokes “the primitive undifferentiation of the ego and the non-ego.”[4] Faced with the screen, and identifying with the camera, we become like children—cinema makes infants of us all.

Jean Ma has recently described what she calls the “regressive thesis in the domain of cinematic reception,” in which theorists assume that going to watch a film in a theater is equivalent to traveling backward in mental development, moving toward increasingly younger stages of mental development until we reach the condition of infancy: the absence of rational thought.[5] It’s not just the reliance on Lacan’s idea of the “mirror stage”—an infantile situation—as a way to understand cinematic encounters. Metz writes that, in the movie theater, we are “like the child, in a sub-motor and hyper-perceptive state... like the child again, we are prey to the imaginary, the double.”[6] And Raymond Bellour has posited “the child who sleeps in every spectator,” connecting the childlike dimension of spectatorship to hypnotic suggestibility.[7]

The dual move here—an emphasis on identification as our fundamental relation to the image, and on the infantile posture of identification—shapes the articulation of theory and the imagination of cinematic possibilities. In an important essay, Richard Wollheim worries about the reliance on identification precisely because of what he calls its “primitive” or “regressive” tendencies.[8] His worry is that identification is such a dominant and primary developmental force that to treat it as a primary aesthetic principle will mean that it is impossible to separate our engagements with the moving image from the terrain of identification and its infantile associations. Wollheim was concerned, that is, that once you describe the relation of the spectator to screen, and thereby to diegesis, in terms of identification, all other aesthetic possibilities are organized by its force.

The dominance of identification even shaped the counterpart to the fantasy of merging with the image: a form of reflexivity, or baring of the device, that aims to distance the spectator from the cinematic image by revealing the constructedness of the image—a gesture similar to the opening of Persona. As if to replace the fundamentally infantile nature of identification with sober rationality, what the subsequent political avant-garde would think of as the necessity for an anti-illusionist cinematic practice. Yet these critical alternatives, powerful as they are, still operate on identification’s terrain: to be engaged in the image is to risk losing ourselves.

It doesn’t have to be this way. In her work on historical spectatorship, Jacqueline Stewart imagines a “kaleidoscopic” approach to “the shifting combinations of factors that structured the appeals of moviegoing” for Black audiences.[9] Elsewhere, and writing about the avant-garde, Annette Michelson imagines the need for what she calls “critical athleticism” in moving between the ways we might inhabit the moving image.[10] In my own work, I’ve argued against identification on the grounds that it limits the way we can imagine ourselves inhabiting the world of the film, arguing instead the category of imagination can be used to get at more complex mental or cognitive positionings. Call it a film theory for grown-ups. Regardless of how we define it, or imagine its possibilities, what’s clear is that we need to put away our fantasy of an infantile relation to the image—even when, perhaps especially because, it is so powerful.