This paper was presented at the Society for Cinema Studies Conference 2000 at the Congress Plaza Hotel, Chicago on March 11, 2000, as part of the panel H6: Towards a Rhetoric of Film

* * *

My topic is identification – identification as rhetoric. Let me begin with a simple example to illustrate what I have in mind. Young lovers are shown walking in a meadow, with flowers around them, trees, a sunny sky with a few puffy white clouds, maybe a river softly flowing in the distance. This is of course a romantic cliché. The young lovers are being identified with nature. We have here a clear example of identification as rhetoric. It is the rhetoric of such a romantic scene – the way it appeals to an audience, its means of persuasion – that we should see the young lovers as natural and nature as beautiful, that we should see their romance as beautiful as nature is beautiful, that we should identify them with nature.

Proceed if you will to an ideological analysis. You could say that this is an example of naturalization, the social regarded as the natural, the construction of the couple – the heterosexual couple, the particular couple the film is portraying, the Southern white man and the Northern white woman in The Birth of a Nation (DW Griffith, 1915), for example – construed as the course of nature. Ideology is but a comprehensive persuasion, a form of rhetoric. Ideology is often treated as coercion, as a kind of police; but it is only when ideology fails that the police has to be brought in. Ideology works not by force but by appeal, by persuasion. The study of rhetoric is the best way to understand the workings of ideology. Take the concept of interpellation, which in Althusserian theory is treated as a kind of coercion. Is it not better treated as a kind of persuasion, persuasion by identification? Being interpellated, one is being identified as something – a subject position one is being put into – and one is being asked to identify oneself with something – what that subject position represents. Identify as and identify with: these are two senses of identification that we need to distinguish but that are often inextricably bound up together. I was born in Havana. If I identify myself as Cuban I am also identifying myself with Cubans. If I then want to dissociate myself from the right-wing Cubans who have been making such a deplorable fuss about that boy Elián González, this illustrates how, in a social, political and ideological context, identification functions rhetorically. If I want to dissociate myself from right-wing Cubans it is because being identified as Cuban I am identified with Cubans and Cubans in the United States tend to be identified with the right wing.

Let us return to the lovers in a meadow. The young lovers in Carl Dreyer’s Day of Wrath (1943) are consistently shown out of doors, and the camera dwells on the natural environment, the grass, the river, the trees. But in Day of Wrath the romantic rhetoric of nature is complicated by other considerations, and countered by another rhetoric that construes nature quite differently. For one thing, these young lovers are adulterous and even incestuous, a parson’s young wife having an affair with his son by a former marriage. We may still regard their affair as natural - the much older parson took advantage of his social position to impose a loveless marriage on a young woman who is naturally drawn to the younger man – but surely this is not a naturalness we can accept without qualms. Moreover, nature itself in Day of Wrath comes under suspicion. Set in seventeenth-century Denmark, the film takes us back into a Lutheran society that looked upon nature as dangerously pagan, a realm where witches roam and the devil lurks. We heirs of romanticism may admire and embrace nature, but those Lutherans would keep it at arm’s length. Set in seventeenth-century Denmark but of course aimed at us who take a different view, Day of Wrath does not make it easy for us to decide (as Arthur Miller does in The Crucible) that we are right and they were wrong. Dreyer has cunningly, unsettlingly constructed his film around the split between these two different rhetorics of nature, these two different ideologies.

The opening section of Day of Wrath depicts at some length the apprehension, trial, torture, and execution by burning of an old woman accused of being a witch. The parson plays a central part in this; his wife Anne and his son Martin are merely bystanders, their illicit affair not yet begun. In a perceptive review of the film at the time of its first American release in the late forties, Robert Warshow noted the insistent pattern of leaf shadows on the face of the old woman as she lies bound to the ladder, about to be thrust into the fire; and he further noted how

when Anne first tries her “power” in order to call Martin to her side, Dreyer repeats on her face the shifting pattern of leaves that appeared on the face of the old witch before she was burnt; when the lovers walk in the fields, the camera keeps turning upward to the trees above their heads. In general, there is an attempt to equate the outdoors, the world of nature, with evil (the pastor’s mother, who is the one firm moral pillar, is never seen outside the rigidly ordered household she controls); but the camera cannot create a religious system (1).

No, the camera cannot create a religious system; but it can express the tenets of such a system in a rhetoric of images, which here takes the form of an identification of the pattern of leaves with the old witch, then of the same pattern with Anne, so that Anne becomes identified with the old witch, and witchcraft becomes identified with nature. And so the romantic rhetoric of the lovers in the fields with the trees overhead, an identification that would usually lead us to see the lovers as good because acting in accordance with nature, is reversed by this religious rhetoric that leads us to see the lovers as evil by identifying them with nature as the realm of witches and the devil. Both rhetorics work by an identification of the lovers with nature, but the romantic rhetoric makes an identification of nature with the forces of good, whereas the religious rhetoric makes an identification of nature with the forces of evil. Warshow takes Dreyer to be espousing the religious rhetoric, but it seems to me he’s doing something more complex and sophisticated. He’s assuming that in our time we would all tend to espouse the romantic rhetoric, the identification of nature with good, and he’s countering that ideology with a religious rhetoric that he knows we will resist – who among us can readily accept the identification of nature with evil? – so as to make us aware that, whether the romantic or the religious, either way we are yielding to the sway of rhetoric, either way we are falling under the spell of ideology. Neither merely rhetorical nor merely skeptical, Day of Wrath is a reflection on the workings, the effects, the inescapability of rhetoric. Neither merely ideological nor a mere critique of ideology as false consciousness – for it makes us recognize how ideology lays the paths our consciousness takes – Dreyer’s film is a rare work of what may be called comparative ideology.

Those of you familiar with the work of Kenneth Burke will have recognized my debt to him. Identification is a key term in his Rhetoric of Motives, and I have been using the term in the way he uses it. Rhetoric is the art of persuasion, and Burke argues that persuasion rests on identification: a speaker persuades an audience by identifying his cause with their interests, by identifying himself with something that appeals to them, that has their approval. One example Burke offers is “the politician who, addressing an audience of farmers, says, ‘I was a farm boy myself.’” Identification is not the same thing as identity. “A is not identical with his colleague B,” Burke writes. “But insofar as their interests are joined, A is identified with B. Or he may identify himself with B even when their interests are not joined, if he assumes that they are, or is persuaded to believe so.”(2). An identification is not an identity but a commonality, something shared in common or believed to be so shared. (Recall the “commonplaces” of classical rhetoric.) Identity is complete: A is A. Identification is a matter of degree: A is identified with B insofar as their interests are joined. Identity is sameness and is always the same. Identification is of different kinds, depending on the interests that may be joined, the things being shared in common, the different grounds of commonality. And, as Burke teaches us, identification not only works to join but also to divide. The politician making an identification with the farmers, with the interests of the country, is at the same time effecting a division from the interests of the city. It is the task of a rhetoric of film to examine, as Burke does in his Rhetoric of Motives, the different degrees and different kinds of identification, its different ways of joining and dividing, the different effects that it can have.

What about, you may want to ask, our identification with characters? Isn’t that what we mostly mean when we talk of identification? That may be, but our identification with a character is never just with that character, never just with an individual for his or her intrinsic character. Our identification with a character always takes place in a context – the context of a story, of a setting, of a genre, the context the work establishes and the context we bring to the work. This context will involve other identifications. If we identify with the young lovers in a meadow, it is because they are identified with nature, and because in our culture (unlike the culture of Lutheran Denmark in the seventeenth century) nature is identified with good things. If the farmers identify themselves with the politician, it is because he has identified himself with them (“I was a farm boy myself”). Our identification with a character depends on how the character is identified – identified as and identified with. The same traits of character may in different contexts be differently identified. The same film seen in different contexts may provoke different identifications.

Ordinary moviegoers often talk about identification with characters, and so do film critics and theoreticians, who think they have a better understanding of such issues. In Hitchcock’s Films (1965) Robin Wood had much to say about Hitchcock’s manipulation of our identification with the characters in his films. Psychoanalytic film theory has also placed importance on character identification (as well as on identification with the camera). More recently, however, cognitivists such as Noël Carroll and Murray Smith have called the notion of identification into question. Carroll would discard the notion altogether. Smith thinks the notion conflates different things that ought to be distinguished, and one of the principal distinctions he draws is between alignment, which “describes the process by which spectators are placed in relation to characters in terms of access to their actions and to what they know and feel,” and allegiance, which “pertains to the moral and ideological evaluation of characters by the spectator.”(3) Although alignment and allegiance are often combined, Smith insists that they are not to be confused. Alignment with a character – focusing on the character, following the character around, seeing what the character sees, etc. – is not the same thing as allegiance to that character. Allegiance means approval, taking sides with the character in a moral sense, rooting for the hero against the villain. According to Smith, allegiance is what we ordinarily mean when we say we identify with a character.

In films like Taxi Driver (1976) and King of Comedy (1982) Martin Scorsese plays on a calculated split between alignment and allegiance. In both films we are aligned with the Robert De Niro character; in neither film do we approve of him. But, even though we don’t approve, even though we don’t even like him, do we not in some significant way identify with him? How else to explain our response to that scene from Taxi Driver, for example, in which Travis, having succeeded in getting Cybill Shepherd to go out with him, chooses to take her to a porno movie? We feel acute embarrassment. This may not be exactly what he feels, but surely we wouldn’t be feeling it if we weren’t putting ourselves in his place. We don’t want to be in his place, we want to get out of there, but the film leaves us no choice, and it derives its peculiar impact from the way it puts us there, where we don’t want to be, in the position of identification with a character we’d rather have nothing to do with. We have no allegiance to these Robert De Niro characters – not as Smith defines the term – but we have an uncomfortable, unmistakable identification with them.

Alignment and allegiance are two different modes of identification with a character that Smith usefully distinguishes: in the former we share the character’s path, his or her point of view in the physical sense; in the latter we share the character’s values, his or her point of view in the moral sense. But usually these two different modes are joined: alignment serves rhetorically to promote allegiance; point of view in the physical sense becomes aligned, allied, with point of view in the moral sense; sharing the character’s path helps persuade us to share the character’s values. What the usual rhetoric joins, the rhetoric of Taxi Driver and King of Comedy unsettlingly divides. That the dividing of allegiance and alignment should be unsettling to us demonstrates how used we are to their joining.

Taking identification to mean identity, the illusion that we are one with the character, that we are the character, Carroll thinks he has disposed of the notion by pointing out that we in the audience never feel exactly what the character feels:

If we feel pity at Oedipus’ recognition that he has killed his father and bedded his mother, that is not what Oedipus is feeling. He is feeling guilt, remorse, and self-recrimination. And, needless to say, we are feeling none of these. We respond to Oedipus’ plight, which includes his own feelings of repulsion about incest, an evaluation which we may share. But this gives rise to an overall feeling in us which does not match up with Oedipus’ emotional state. For example, if Aristotle is right, we come to fear that such calamities could befall us and, then, we undergo catharsis. But the time for fear has passed for Oedipus.(4)

Concerned with proving to us that we are not Oedipus, Carroll misses the way in which Aristotle’s pity and fear correspond to two classic forms of identification with a character, sympathy and empathy, feeling for and feeling with. We feel pity for Oedipus; we feel fear with Oedipus. As Michael Davis argues in his study of the Poetics, Aristotle saw that our response to tragedy is double: we respond from outside the drama, from our perspective as spectators, and we respond from inside, from the perspective of the characters. “As spectators we pity; as participants we fear,” Davis writes.

Imitations announce to us their simultaneous reality and falsity. They can only affect us insofar as they are real; we must therefore enter into the perspective of the character. On the other hand they can only affect us as images if we are aware of their unreality; this distance allows us to separate ourselves from the plight of the character. If a ferocious beast leaps at a character in a movie, we in the theater cannot properly appreciate the film unless we are at some level afraid. On the other hand, were we simply afraid, we would leave our seats and run for the exits. We can savor our fear only insofar as our experience is simultaneously real and unreal. Although it is usefully understood as the split between the perspective of the character in the drama and that of the audience, this initial dualism is not stable. Pity is not so easily separable from fear, for if we did not fear on some level we could not pity the plight of one in danger. To be worried about the danger facing a character means in some sense to feel it as danger. (5)

Pity and fear, sympathy and empathy, are both forms of identification, and not so easily separable, because they both involve our putting ourselves in the character’s place. Whether we feel for or feel with the character, in neither case do we feel exactly what the character feels, but what we feel in both cases depends on a significant commonality with the character, on some degree, some kind – degrees can become kinds – of identification.

Carroll’s dismissal of identification occurs in his study of horror. Horror is a genre that arouses fear more than pity. Where does that fear come from if not from our identification, our empathy, with the plight of a character in danger? We know we are not in danger, and watching the Frankenstein monster up there on the screen would not frighten us nearly as much if the monster were not threatening a character in the movie with whom we identify. Two famous vampire movies, Murnau’s Nosferatu (1922) and Dreyer’s Vampyr (1932), have as a figure of identification a journeying young man who finds himself in dangerous territory. For the first third of Nosferatu we follow the young husband in his trip to Transylvania; this is a clear example of alignment with a character – we travel alongside him and see what he sees, know what he knows – and of course this young man has our allegiance as well against the vampire he goes to meet. But the young husband in Nosferatu doesn’t recognize as we do the danger he faces; he is a smiling innocent who shrugs off all the signs of danger that alarm us. He’s not afraid; we are afraid for him. By contrast, the young man in Vampyr, whom we follow through the entire movie, is afraid from the start. We are afraid with him; what this young man is feeling is much closer to what we are feeling. Does this mean that we identify with him more than we identify with the young man in Nosferatu? I wouldn’t be so sure. For we are more frightened in Nosferatu than we are in Vampyr; we fear more for the young man who is not afraid than for the young man who is. Partly this is because the young man in Nosferatu is in greater danger; the young man in Vampyr is more of a bystander to danger, more an observer than a participant. Thus his situation parallels that of the audience watching the movie. Much more than the young husband in Nosferatu, the young man in Vampyr is a figure of the spectator, our double inside the movie, a wandering consciousness twinning our own. Doesn’t this make for a stronger identification with him? It depends on what kind of identification is meant. Even though we don’t feel the same thing that the young man in Nosferatu is feeling, we identify with him, identify emotionally, because we feel what we would be feeling if we were in his place. Even though our situation and our feelings match up much better with those of the young man in Vampyr, our identification with him is paradoxically less emotional and more reflective. One way of putting the difference would be to say that we identify ourselves with the young man in Nosferatu, whereas the young man in Vampyr is identified with us.

Our identification with Norman Bates as he cleans up after the shower murder in Psycho (1960) is often cited as an example of Hitchcock’s manipulative mastery. Up until her murder in the shower at the Bates Motel, we have been systematically led to identify with Marion Crane: we’ve been aligned with her from the beginning, and Hitchcock’s rhetoric of alignment powerfully overrides any moral reservations we may have about giving our allegiance to a loose woman and a thief. “Never,” wrote Robin Wood, “has identification been broken off so brutally.”(6) And yet, right afterward, we switch allegiances and become identified with her murderer. True, when first seeing the film we don’t know at this point that Norman Bates is her murderer, but it’s remarkable all the same that we should so quickly switch allegiances to the man cleaning up the bloody mess in the bathroom and dumping her dead body, wrapped in the shower curtain and shut inside her car trunk, into the depths of a swamp. I’ve been saying that we switch allegiances, but allegiance in Murray Smith’s sense of moral approval has little to do with it. “We have been carefully prepared for this shift of sympathies,” holds Robin Wood. “Norman is an intensely sympathetic character, sensitive, vulnerable, trapped by his devotion to his mother – a devotion, a self-sacrifice, which our society tends to regard as highly laudable.” (7) We’re drawn to identify with Norman, Wood is saying, because Norman is identified with things we value, devotion to his mother above all. But the moral and personal qualities we perceive in Norman – the qualities intrinsic to an individual, what people call “character” in American politics nowadays – surely don’t explain our readiness to identify with him in the aftermath of the murder. He’s cleaning up the mess his mother made, but, since we take her to be a mad killer, his devotion to her would scarcely be the reason we identify with him. Our identification comes rather from our involvement in the task he’s carrying out, the very effort and detail of cleaning up the mess. It’s not the devotion to his mother that draws us closer to Norman, but the task to which he devotes himself, and which we watch as attentively as he performs it. It’s not the intrinsic quality of character that most engages us, but the action in which we see the character engaged. Alignment, our alignment with Norman in the activity of cleaning up after his murderous mother, once again takes precedence over allegiance in bringing about identification.

Psycho is of course a joke on the notion of individual character, the core of qualities intrinsic to a person. Norman has no such core: he is split in two. Who is it we’re identifying with when we identify with Norman Bates? Psycho invites us to ask that question, and to ask what it is in ourselves that makes us disposed to such an identification. We may not be aware of it the first time we see the film, but our identification with Norman cleaning up the mess he made himself betrays our complicity in making that mess, our unacknowledged complicity in the murder. Just as Norman’s devotion to his mother is the other side of Norman’s murderousness, so our identification with Norman the dutiful son is the other side of our identification with Norman the murderer. In neither case, however, neither with the murderer nor with the dutiful son, are we identifying with an individual. Norman Bates is not an individual. The bias of our culture leads us to assume that we identify primarily with individuals, with the intrinsic qualities of a person, but, as we’ve just seen, it is primarily an action – an individual engaged in an action, if you like, but an individual seen in terms of that action rather than in his intrinsic qualities of character – that we identify with when we identify with Norman cleaning up the mess.

One of the advantages of treating identification as rhetoric is that rhetoric, being always public, always addressed to an audience, helps dispel the illusion that identification is a personal matter, that we identify one on one, the spectator as an isolate individual with the character as an isolate individual. Noël Carroll is right to object to the notion that we become one with a character. But, treating identification as something we do one on one, he makes the same mistake as those he is criticizing. The moment we recognize how we always identify in a context, a situation, and how often we identify with an action – an action that, in that context and situation, draws us in – it becomes clear that we’re not dealing with a oneness but with a commonality. Rhetoric, as Kenneth Burke insists, has to do with action, with bringing people to act together. It is a moving to action, or at least an inducement to attitude, which, as Burke says, is incipient action. The issue of “character” in American politics is a pretense that we are dealing with the intrinsic qualities of an individual when other matters and other interests are at stake, and simply to see that “character” is rhetoric, not a personal issue but a public persuasion, is to begin to deal with what is really at stake.


  1. Robert Warshow, The Immediate Experience, New York: Atheneum, 1979, p. 266.
  2. Kenneth Burke, A Rhetoric of Motives, Berkeley: University of California Press, 1969, p. 20.
  3. Murray Smith, “Altered States: Character and Emotional Response in Cinema,” Cinema Journal, Vol. 33, No. 4 (Summer 1994), p. 41.
  4. Noel Carroll, The Philosophy of Horror, New York: Routledge, 1990, pp. 90, 93.
  5. Michael Davis, Aristotle’s Poetics: The Poetry of Philosophy, Lanham, MD: Rowman & Littlefield, 1992, p. 39.
  6. Robin Wood, Hitchcock’s Films Revisited, New York: Columbia University Press, 1989, p. 146
  7. Ibid., same page