The Exorcist

This is an extract of a chapter from the book project, Freud’s Worst Nightmares, expected to be published by Cambridge University Press as part of their “Studies in Film” Series. For more information, contact William Rothman or Steven Jay Schneider.

* * *

Psychoanalytic theories of film, and of horror film in particular, have been subject to attack from various quarters. This essay addresses these criticisms-defending a psychoanalytic approach to horror cinema from objections raised by theorists such as Stephen Prince, Andrew Tudor, Jonathan Crane, Noël Carroll, and Berys Gaut. Some of these objections do little more than wheel out the well-worn objection-a common one even in Freud’s time-that psychoanalysis is “unscientific.” But even if is true that psychoanalysis is unscientific (by some often objectionable standard), this does not ipso facto show that it is false. Adolf Grünbaum’s critique of Freud’s so-called “Tally Argument” (see below) is an example of one such “objectionable standard.” This critique is basically a gussied-up version of the claim that psychoanalysis is not falsifiable. However, the falsifiability (in principle), of a scientific theory, has to be interpreted in way suitable to the theory in question. It is clear that psychoanalysis is not going to be falsifiable (in principle) in the way that the physical or biological sciences are- that is, by producing an experiment that can conclusively falsify it. Nevertheless, as I point out below, aspects of psychoanalysis certainly are falsifiable, and indeed have been falsified. (1)

It is also not difficult to produce examples of disciplines and theories that are (by certain standards) unscientific but true, or likely to be true. Many philosophical theories-whether broadly or narrowly construed-are unscientific and true, though it may be difficult to say which are the true ones. Similarly, some very general theories in social science may be true but unscientific according to the standards of the physical and biological sciences. Thus, Clifford Geertz’s enormously influential theory of “religion as a cultural system” (1973: 90) is in my view true but not experimentally falsifiable.

Other criticisms are based on a misunderstanding of fundamental aspects of psychoanalysis. (2) Although there are many interesting issues in film theory that relate specifically to the horror genre, the critique of psychoanalytic approaches to interpreting horror is usually more general. Such criticism is often directed at any kind of psychoanalytic approach to understanding film and spectatorship. My approach therefore mirrors the form of this criticism. It often discusses critiques of psychoanalytic approaches to film generally rather than horror in particular- even where horror films are the examples used.

Psychoanalysis and Horror: The Objections

Most of the objections to psychoanalytic film theory are founded on well-worn objections to psychoanalytic theory generally. There are, however, more specific objections to the application of psychoanalysis to film, especially film spectatorship, and to particular film genres such as horror. I shall discuss the general objections before considering a sample of more specific criticisms- in particular those of Gaut (1994). I argue that the general objections to psychoanalysis are unfounded, while the specific ones are often based on a misunderstanding of fundamentals of psychoanalysis. This is the case even allowing for the wide variety of disparate views amongst psychoanalytic theorists themselves.

Psychoanalytic approaches to film are often contrasted with cognitive approaches, those who support the latter typically eschewing the former. Freud, however, would have rejected this dichotomy. A psychoanalytic approach to film would be regarded by Freud and most psychoanalytic theorists as a cognitive approach. Those who favor cognitive approaches to film err in not recognizing this that is how it sees itself, and in pointing to cognition as a basis for distinguishing their own approach from that of psychoanalysis. (3) Whether psychoanalysis is a fundamentally flawed cognitive approach is another matter.

General Objections to Psychoanalysis

Until recently most, but by no means all, philosophical work in connection with Freud has been concerned with the question whether Freud’s work was scientific- whether it was testable, verifiable, or falsifiable, in accordance with accepted scientific procedure. Grünbaum (1984; 1993); Erwin (1996; 1993); Macmillan (1997); Cioffi (1988; 1985); and others (cf. Robinson, 1993) have been far less concerned with the significance and wide-ranging philosophical implications of psychoanalytic theory, or even with the truth of such theory, than with the nature and methods of science. (4) But arguably the scientific status of psychoanalysis has always been a problem for the philosophy of science in general, rather than having any intrinsic connection to psychoanalysis. What constitutes a scientific theory (cf. Lear 1990: 216ff)? What are appropriate scientific inductivist canons? (5) Furthermore, even if psychoanalysis is not a “science” given some agreed upon scientific inductivist canons, it may nevertheless be more or less true. Since a theory can be true without being scientific it is a mistake to see psychoanalysis as false if not scientific. Its support may come from various sources and considerations- both theoretical and empirical.

Grünbaum’s account of Freud’s “Tally Argument” shows that Freud himself was responsible for focusing so much attention on the scientific status of his theories. Freud says that, “After all, his [the patient’s] conflicts will only be successfully solved and his resistances overcome if the anticipatory ideas [i.e., psychoanalytic interpretations] he is given tally with what is real in him. Whatever in the doctor’s conjectures is inaccurate drops out in the course of analysis. .it has to be withdrawn and replaced by something more correct” (1917, XVI: 452). Grünbaum interprets this “as a conjunction of the following two causally necessary conditions. (p. 1) Only the psychoanalytic method of interpretation and treatment can yield or mediate for the patient the correct insight into the unconscious causes of his neuroses. (p. 2) The patient’s correct insight into the conflictual cause of his condition and into the unconscious dynamics of his character is in turn causally necessary for the durable cure of his neuroses” (1988:14). He refers to the conjunction of these claims as the Necessary Condition Thesis. (6) Such claims appear subject to inductive assessment in accordance with scientific (inductivist) canons, and Freud’s apparent adherence to the Necessary Condition Thesis suggests that he agrees. (7)

No doubt Freud believed his theories to be scientific and testable to a degree. However, it should not be overlooked that, whether or not he had sound and/or ulterior motives in claiming scientific status for psychoanalysis, it was probably a good thing for psychoanalysis that he did. It is unlikely that psychoanalysis would have received the attention it did had it not associated itself early on with the mystique of science in a scientistic age. At any rate, even a cursory reading of Freud suggests that his theories were not based solely on clinical data and “evidence,” but also, and perhaps even more so, on his observations of everyday life. He understood only too well the presence of mitigating factors, and the difficulty of achieving a psychoanalytic cure. Textual evidence notwithstanding, he never strictly adhered to the Necessary Condition Thesis. Freud carried his couch around with him. In psychoanalysis, like anything else, too selective a reading-or too close a reading-can at times be as misleading as the alternatives.

What then is the evidence for the validity of psychoanalytic theory? According to Hopkins, psychoanalytic explanation is an extension of common sense or folk psychology. (8) The relevance of thematic affinity (e.g., dreaming of drinking water when one is thirsty)-how it supports Freud’s theory-is crucial to Hopkins’ defense of psychoanalysis. He says that Grünbaum objects to the idea

that claims as to a causal connection between mental items can be cogently supported by a connection in content-a “thematic affinity”-between them. He speaks of “what might be dubbed ‘the thematic affinity fallacy’,” and appears to reject the basing of causal claims on connection in content “no matter how strong the thematic affinity.” For, he stresses, “thematic affinity alone does not vouch for etiologic linkage in the absence of further evidence” [34 ].(9) Grünbaum has made no case against the view.that much of Freud’s reasoning can be regarded as cogently extending commonsense psychology. If Grünbaum has missed something about connection between content and wish-fulfilment [i.e., the significance of thematic affinity], and if what he has missed constitutes reason to accept Freudian claims, then his conclusions systematically understates the support for Freudian theory. (1988: 59)

In addition and related to instances of thematic affinity found in dreams and neurotic symptoms, ordinary experience of the kinds Freud relied upon should not be discounted in support of psychoanalysis. Thus, the view that our strongest denials are often affirmations; that slips of the tongue are (often) meaningful; that individuals regularly deploy a variety of ego defence mechanisms at their disposal- all are Freudian insights that have now become commonplace. They can jointly be taken as part of the evidence for the truth of psychoanalytic theory. Together, they allegedly support the inference that psychoanalytic explanations of these phenomena are often the most plausible ones (see Freud 1901). When judged against alternative explanations, psychoanalytic explanations of phenomena like racism and sexism can also allegedly support fundamental psychoanalytic truths. (10) Whether or not Woody Allen’s psychoanalysis reaches a successful conclusion is-except for Woody- relatively unimportant.

Prince states that “the primary and to my mind insurmountable problem with basing general theories of spectatorship on psychoanalysis is that such theories must remain unsupported because psychoanalysis is a discipline without reliable data” (1996: 72-73). However, just as the hard science’s notion of falsifiablity is too narrow as applied to psychoanalysis), so too is Prince’s notion of “reliable data.” They are too narrow given the kind of “science” psychoanalysis is. Prince endorses Colby and Stroller, who claim that “psychoanalytic evidence is hearsay, first when the patient reports his or her version of an experience and second when the analyst reports it to an audience. Reports on clinical findings are mixtures of facts, fabulations, and fictives so intermingled that one cannot tell where one begins and the other leaves off” (1988: 3, 29). This charge of unreliable data is just another way of systematically rejecting the kinds of support for psychoanalysis that are in fact available. The criticism falsely assumes that what constitutes reliable and accurate psychoanalytic data and interpretation is up for grabs- that there are no criteria for judging what is and is not correct in psychoanalysis. Of course there is going to be disagreement amongst theorists as to what those criteria are, both generally and in specific cases. And of course those criteria are going to be very different than the criteria for theoretical adequacy in, say, the physical sciences. But the idea that psychoanalytic practice and theory necessarily employ “unreliable data” is a roundabout way of failing to recognize what does constitute psychoanalytical data and why. (11)

Far from regarding his work as final, Freud frequently reiterated that psychoanalysis has a bright future regarding “further discoveries” about the individual, civilization, and the connections between them. Freud constantly revised his theories in view of theoretical concerns and new data. Perhaps the most notorious example (recently notorious) is his revision of the seduction theory. Freud first claimed that actual childhood seductions were the cause of various neuroses. He later claimed that imagined seductions could suffice in terms of the aetiology of neuroses- given all the other conditions of susceptibility, instinctual endowment, etc. Wishful phantasies of various sorts came to be considered important, including phantasies of seductions (e.g., phantasies of parental intercourse). Both the original and altered positions were heavily and vehemently criticized, though for different reasons.

Another example is Freud’s introduction of the death instinct in 1920 to supplement the sexual (libidinous) instinct. The conflict between libido and ego-instincts becomes a conflict between eros and the death instinct. There is also his revision of the account of anxiety in 1926: first he thought that anxiety was transformed libido; then he realized that (signal) anxiety was a reaction to threat from the external world, the superego, and instinctual forces. And then there is Freud’s change in topography from Uc.-Pcs.-Cs. to id, ego and superego in 1923. Such revision-and the extent to which he revised truly is striking-goes hand in hand with the falsification of earlier theory and interpretation. So aspects of Freud’s theory were (and are) not only falsifiable, but were indeed falsified, and recognized by Freud as such. Subsequent developments in psychoanalysis build on aspects of Freud’s views without regarding those earlier views as sacrosanct. The work of Melanie Klein, for example, and David Winnicott, accept as well as revise aspects of Freud. The ongoing development of psychoanalysis is indicative of the fact that it continues to be revised, refined, and in many cases corrected. By its own light, aspects of psychoanalysis-even key aspects-are falsifiable.

The standards or criteria for falsification are different for psychoanalytic theory than they are, e.g., for physics. But even those in psychoanalysis who claim that it is science have never claimed the standards to be the same- or if they did, they were confused and mistaken. If the demand for criteria of falsification is a demand for criteria of the falsification of psychoanalysis in its entirety, then this is different kind of demand and standard. It is more akin to asking for criteria of falsification not just for a theory or aspect of physics, but for physics per se. And it would seem that there is no more readily acceptable standard of falsification for physics per se than there is for psychoanalysis as such.

Objections to Psychoanalytic Accounts of (Horror) Film Spectatorship

Psychoanalytic explanations of spectatorship in relation to horror films-for example, why some people like them and what this says about spectators-are for the most part an aspect of psychoanalytic accounts of film spectatorship more generally. Categories like perversion, voyeurism, fetishism, masochism, and sadism that are used to psychoanalytically explain film spectatorship are the same as those used to explain the (seemingly inexplicable) attractions of horror. Nevertheless, there remain some psychoanalytic concepts and theories-like repression, the “uncanny,” and “doubling”-that seem relevant primarily to horror films.

Prince (1996) couples his rejection of psychoanalytic approaches to horror with an endorsement of what he sees as a more empirically-oriented research program. He writes that

The empirical research on factors affecting levels of attention and comprehension can give us a more nuanced portrait of spectatorship than does psychoanalysis, and this, in turn, can help us to construct theories that are sensitive to the differences, as well as the similarities, among viewers. Rather than continuing to base theories on a concept of the scopically driven, fixated (or “positioned”) viewer, I suggest that we begin to derive our theories and research from the constructs of “attention” and “attentiveness.” This will enable us to make a key advance in the way we model viewing behavior. It will enable us to conceptualize, and study, viewing processes in terms of levels of information processing and emotional response. (p. 78)

His claim about a “key advance” notwithstanding, it is hard to see what the merits of Prince’s proposed research program may be in regards to a theory of film spectatorship. His is a program that is concerned with “viewing behavior” is a limited sense, for some vague purpose. It appears to focus on the nature of reception rather than interpretation. Or perhaps he is asking for an explanation of the processes by which film is able to do what it does. (Even if psychoanalysis explains content, it does not explain the remarkable fact that flickering images on the screen can evoke this content.) (12) But what Prince calls for abandons film theory’s traditional concerns with the elucidation of an account of film spectatorship- e.g., how and why certain kinds of films engage spectators in certain ways, and what these ways are.

Here is an example of the empirical research to which Prince refers:

Observations of real spectators furnish a rather different portrait of viewing behavior. A great deal of research has studied the ways young children watch television. Preschool children in a room furnished with toys and containing other adults or children do not stare with steady fixation at the screen. Instead, the child will repeatedly glance away from the screen, averaging about 150 looks toward and away from the screen per hour. Furthermore, glances at the screen are quite brief, and most last no longer than 15 seconds. Observations of adult television behavior reveal a similar pattern: looks at the screen are extremely brief and are punctuated by regular glances away from the screen, with non-looking pauses averaging as high as 22 seconds. (1996: 77)

He contrasts this with psychoanalytic film theory:

Aside from psychoanalytic film theory’s failure to model a sophisticated perceptual process, the claims it does make ill fit the available evidence on how viewers watch film and television. The problem with the “scopic drive” is that it models viewing as a driven and reactive process during which the viewer’s passion for looking is cathected by particular formal cues (for example, “fetishizing” close-ups). The scopic drive implies a unifocal fixation within the viewer maintained by a match of formal features and inner fantasy. (1996: 77)

But the empirical research Prince cites does not conflict in any obvious (or otherwise) way with a psychoanalytic account of spectatorship partly in terms of a scopic drive. (13) The fact, if it is one, that “glances at the screen are quiet brief” does not conflict with it. It has nothing at all to do with such claims! (A brief glance can be as scopically driven as a long look.) The concerns of the research, the aspects of spectatorship being investigated, are different. Prince’s perception that psychoanalytic film theory about spectatorship “ill fit[s] the available evidence on how viewers watch film and television” is not supported by the evidence he cites. Where is the connection? Why should psychoanalytic film theory seek to “model a sophisticated perceptual process?” Such theory has rightly been concerned with interpretation and understanding- not with perceptual models.

Prince goes on to say that “Psychoanalytic theory fails to grasp [that].when watching media presentations, the viewer simultaneous executes multiple levels of information processing and engages in a series of interpretive moves. Psychoanalytic theory tends to collapse the viewer’s responses into a single dimension fed by primary process energy and the unresolved childhood traumas associated with it” (1996: 79). But Prince has not shown that psychoanalytic film theory is reductive in the way he claims, and there is nothing in psychoanalysis that prevents it from grasping what Prince claims it fails to understand. What grounds are there for claiming that psychoanalysis would deny that “the viewer simultaneous executes multiple levels of information processing and engages in a series of interpretive moves?” Psychoanalysis would be foremost among the original champions of such a view, for it seeks to explain various interpretive moves in terms of its theoretical resources. If one is suffering from castration anxiety, for example, psychoanalysis seeks to explain that anxiety in terms of various interpretive moves without denying that the person sees whatever it is that they see. Repression, it could be argued, necessarily involves multiple levels of information processing.

Continuing his critique, Prince says that

By employing the constructs of “attention” and “attentiveness,” film scholars can formulate investigable problems. Indeed, we ought to stipulate as a criteria of future theory building that postulated theories be able to generate researchable questions. Current psychoanalytic theories of spectatorship do not do this, in part because psychoanalytic critics often aim only to produce interpretations of particular movies, and because many film scholars employing psychoanalytic theory seem interested in abstract and ideal, as opposed to actual, viewers. (1996: 80)

But it is unclear why film scholars, as opposed say to cognitive scientists, should be interested in the constructs of “attention” and “attentiveness” (see Prince 1996: 80-84). After all, film theorists’ concerns-and not just psychoanalytic film theorists-with spectatorship and horror has been aimed at understanding (e.g.) their appeal; how and why they affect us as they do; what such films say about us, and what they represent. (14) Additionally, the idea that psychoanalytic theories of spectatorship do not generate researchable questions should be rejected. Prince means, of course, that they are not researchable in the way that certain questions in cognitive science/psychology are. But psychoanalytic theorists are not alone in rejecting the standards applicable to those research programs as inappropriate and inapplicable to their own fields. Social scientists, economists, and some psychologists do so as well. For that matter, so do individuals working in the humanities. The success and worth of their research is not to be judged by standards appropriate to, for example, the physical sciences.

Furthermore, it is simply not the case that psychoanalytic theory is “interested in abstract and ideal, as opposed to actual, viewers.” Psychoanalysis and the approach lauded by Prince agree that understanding actual viewers requires theory. The concepts psychoanalysis sees as essential to understanding spectatorship generally (e.g., voyeurism, fetishism, sadism, masochism) are also seen as necessary to understanding why and how specific individuals view films as they do. Prince’s claim that film theory generally and psychoanalytic theory in particular has ignored real spectators is, from a psychoanalytic perspective, false. Such theory does not agree with what Prince thinks important about spectatorship, but that is another matter. In short, Prince basically ignores the question of why film theory should be concerned with the constructs he thinks are key- for example, those of “attention” and “attentiveness.”

Crane (1994) too claims that psychoanalytic approaches to spectatorship (e.g., Freudian, Lacanian, Kristevan) tend to be reductionistic. This, he claims, is partly because they are ahistorical, and (echoing Prince) because they tend to reduce group dynamics to the operations of the individual psyche. Discussing horror, Crane says that

Whatever the new terrors mean, they can only be understood in conjunction with a much larger world than that of the solitary misfit coming to grips with his or her own discomfiting sexuality. By consistently reading the audience as an individual psyche, horror films have been turned into tokens of universal, unchanging, and, ultimately, undifferentiable archetypes or psychic black holes with no historical relation to the times of the people who made and understood them. (p. 39)

The claim that psychoanalysis is ahistorical is only partly true, and Crane’s own exposition covers a variety of psychoanalytic approaches that are not, or need not be, reductionistic. Psychoanalytic theory does not deny that human beings are historically conditioned in certain ways- for example, it allows that the forms particular neuroses take are historically conditioned. Thus, the neuroses that Freud treated in Vienna at the turn of the century were indeed partly a product of that era and locale. In particular, they were a product of the treatment of women and their role in society at the time. It is basic mental processes and some contents that are relatively unconditioned by time. The often heard, but insupportable objection that psychoanalysis is only applicable, if applicable at all, to middle-aged and middle-class Viennese at the turn of the 19th century is perhaps what Crane has in mind. Crane seems to think it is a telling criticism of psychoanalytic film theory that versions of one-time frightening monsters such as Frankenstein and Dracula are now used to sell breakfast cereals. But psychoanalytic film theory offers explanations of how and why what frightens us is in some respects historically conditioned, and may indeed change over time. (15) Psychoanalysis sees an audience not as an individual psyche, though there may be group dynamics in an audience. It is sees an audience as a group of individual psyches- which is precisely what it is. (16)

Crane continues by quoting an earlier article by Prince:

In reading a culture and its attendant texts, Freud and his varied followers move rather too quickly from the individual to the social, attempting explanations of such collective phenomena as group dynamics, art or religion in relation to the operations of the psyche, writ large. By collapsing the social into the psychic life of the individual, Freud risked losing the social, and a similar problem exists with regard to our theoretical understanding of horror films. (39; Prince 1988: 20)

It is true that for Freud aspects of the social are explainable in terms of the psychic life of the individual, but he never reduced the former to the latter. Freud was well aware of, and discussed, various ways in which group dynamics affect the individual psyche, and vice-versa; and he recognized that groups were often best seen as more or other than the sum of individual psyches. At any rate, neither Prince nor Crane give any substantive examples where the problem they allege interferes with a particular psychoanalytic explanation of horror.

Tudor (1997) criticizes Grixti’s (1989) “beast within” approach as a way of explaining the attraction of horror. According to Tudor, Grixti traces the expression of this approach

in the views of such popular horror writers as James Herbert and Stephen King as well as in those academic perspectives which invoke catharsis as a key mechanism or claim that horror appeals to deep-seated, psychoanalytically intelligible repressed desires. Underlying such arguments, he says, is the belief “that human beings are rotten at the core” (Grixti: 86), whether by nature or nurture, and that horror resonates with this feature of the human condition. The genre serves as a channel releasing the bestiality concealed within its users. If the model is that of catharsis, then the process is deemed to be beneficial: a safety valve. If the model is one of articulation and legitimation, then the genre is conceived to encourage consumers in their own horrific behaviour. Either way, the attraction of horror derives from its appeal to the “beast” concealed within the superficially civilized human. (p. 445)

First of all, there is no psychoanalytic explanation that explains the attraction of horror in terms of a “beast within.” Furthermore, a “beast within” approach need not and does not equate to a belief that human beings are rotten to the core. Leaving that aside, psychoanalysis does not assume that human beings are rotten to the core, nor does it need such a view to claim that “horror appeals to deep-seated, psychoanalytically intelligible repressed desires” and that the genre serves as a channel for certain kinds of emotional release. If repression was incompatible with decency then we would all be indecent. But psychoanalysis strives to show how repression and its resulting neurotic activities are quite compatible with both “normalcy” and moral decency.

Tudor claims that if all people are supposed to have a “beast within,” then explanations of why people like horror in terms a “beast within” fail to explain why only some people like horror and not others (p. 445). He rightly points out that other factors must then be employed to explain this differentiation. But those who explain the appeal of horror in psychoanalytic terms such as the return of repressed, the reconfirmation of previously surmounted beliefs, or sadistic tendencies do not deny this. (I have already pointed out that psychoanalysis does not explain horror’s attractions in terms of a beast within or the rottenness of humanity.) Why particular individuals like horror but not others will depend upon the individual’s particular psychosexual development- a development that depends upon nurture as well as nature. Support for this is scattered throughout Freud’s writings on the stages of psychosexual development and character types (see, e.g., Freud 1915-1917). Tudor says, “even if ‘beast within’ arguments are prima facie plausible, they do not really answer ‘what is it about people who like horror?’ That requires a different kind of explanation” (p. 445). But this is not right. Psychoanalytically conceived, the explanation requires elaboration in terms of both general and particular psychological features of individuals. This is common sense, and does not entail that a “different kind” of explanation is required. It is worth noting in the context of Tudor’s discussion that, although psychoanalysis offers explanations for the attraction of horror, it also has explanations for the attractiveness of other genres in film and fiction, such as “action” and “romance.” It does not single out horror for explanation in the way that Tudor suggests. The only reason it may seem this way is because of the interest the particular question about the appeal of horror arouses. (Even noted horror film theorist Robin Wood, in his seminal 1979 essay, “An Introduction to the American Horror Film,” speculates that his “return of the repressed” theory of the horror film may be fruitfully extended to other genres, e.g., the Western.)

Tudor (p. 449) claims that “an account of horror in terms of the ‘return of the repressed’.does not directly address the question of the attraction of horror.” (17) Though he does not mention them by name, he elaborates this claim by blending together two so-called “paradoxes”: the paradox of horror and the paradox of fiction. Firstly, if fictional horror produces unpleasurable emotions such as fear, then why do people enjoy it? Secondly, if emotion requires belief that the objects of said emotion exist or are real, then how is it that fiction can induce emotion? Tudor says, “one might argue that bringing such fearful things to the surface is likely to be far from pleasurable. Freud himself recognizes that the uncanny as experienced in reality (as opposed to in fictions) is a matter for fear, and to understand it as a source of pleasure when induced through fiction requires further elaboration” (p.449).

Whatever one may think of the interest of these so-called paradoxes, they are not genuine. An explanation of the attraction of horror in terms of the return of the repressed need only deny what in the end most resolutions of the paradox of horror deny. Emotions such as fear, horror, disgust, etc. are not intrinsically unpleasant. In certain circumstancesnot just when watching films, but in everyday lifethey obviously can be enjoyed. Fear and even horror may be enjoyed in certain sporting activities, for example. And disgust may at times be a source of amusement. (In the case of the paradox of fiction, what is usually denied is that belief in the existence of the objects of emotion is required for an emotional response.) If these paradoxes can in the end be resolved-if they are not genuine-then it seems that an explanation of the attraction of horror in terms of “the return of the repressed” is far more direct than Tudor would have it. This doesn’t mean that further elaboration of such an explanation might not refer to more general features of emotion, or indeed to additional psychoanalytic mechanisms. But such elaborations do not undermine the “directness” of the psychoanalytic explanation of the attraction of horror in terms of the return of the repressed. There is nothing in the alleged paradoxes that confounds or short-circuits such explanation.

Tudor claims that whatever psychoanalytic explanations are given for the pleasures associated with horror, more is needed than just an account in terms of the “return of the repressed.” He says that “it is necessary to pose supplementary mechanisms to bridge the gap between a general account of repression and the specific explanation of pleasure, and these supplementary mechanisms lead away from the pure form of the repression model” (p.449). But such supplements are not only available, they are also part of any more complete explanation of the attraction of horror in terms of “the return of the repressed.” Of course, there are various and often-conflicting accounts of what additional elements are needed to explain the attraction of horror. Thus, Kristeva’s notion of “abjection,” taken up by Barbara Creed in her (1993) theorization of the “monstrous-feminine” in terms of a Lacanian account of the “Real” and the construction of the feminine as “Other,” seeks to give a more complete explanation of the attraction of horror. (18) Whether or not, and to what degree, these notions (and Lacanian psychoanalysis generally) is compatible with what Freud says about horror, and Freudian theory in general, is controversial. But the existence of such psychoanalytically-informed theoretical elaborations on the “return of the repressed” do not subvert but instead support this model.

Tudor appears to regard the need for recourse to additional explanation in terms of psychoanalytic theory as a kind of ad hoc maneuver, but it is not. In psychoanalysis as elsewhere, a more complete explanation requires additional theoretical detail. The “return of the repressed” may be pleasurable for a variety of different reasons depending on the nature of the repressed element being returned and also on the particular spectator involved. The pleasures of horror, dependent as they are upon the effects of the repressed, may involve temporary substitutive satisfactions- much like neurotic activity. Neurotic activity provides a replacement satisfaction for something else that did not occur. (19)

Given that psychoanalytic themes are to varying degrees consciously adopted in some films, including and especially horror films, psychoanalysis may be indispensable in interpreting them. But this is not primarily what those who claim that psychoanalysis is essential to understanding the attraction of horror films have in mind. Their claim is that psychoanalytic theory is relevant because of what it tells us about spectatorship in general and in relation to horror films specifically. If horror films are attractive because they depict and embody the “return of the repressed” in various ways, then they cannot properly be interpreted, nor can their attraction be understood, apart from such theory.

The psychoanalytic claim is that horror films-qua-horror films must be psychoanalytically interpreted. Bad horror films-ones that fail to scare or attract, or ones that are really not part of the genre-may of course require no psychoanalytic theory about horror to enable interpretation. However, Carroll claims that although psychoanalysis is necessary for interpretation in cases where psychoanalytic themes are consciously built in, it is not essential to understanding and interpreting horror films generally (1990: 168-69). (Note that Carroll is here assuming that authorial intention is relevant to interpretation.) He acknowledges that psychoanalysis may explain the attraction of particular horror films, but says “the psychoanalytic account is not comprehensive for the genre” (p.174). However, even those who claim that psychoanalysis is essential for understanding the attraction of horror and interpreting horror films need not claim that such theory is “comprehensive for the genre.”

No doubt there are some horror films-ones not central to the genre-that may be attractive to spectators, and interpretable, for reasons other than those usually given by psychoanalytic explanations of horrors. Indeed, some of these other explanations may themselves turn out to be psychoanalytic. But exceptions can be acknowledged without undermining the claim that psychoanalysis is necessary for understanding horror. To clarify: in claiming that psychoanalysis is essential for understanding the attraction of horror, I mean that horror as a genre, and most cases of being attracted by horror, cannot be understood without it. This does not mean that psychoanalysis might not need to be supplemented by other theories and explanations in some cases- or even that psychoanalysis might not be immediately necessary in all cases.

The distinction between (i) acknowledging that many, if not most, films (or at least horror films) require psychoanalytic interpretation (i.e., that they are in fact perverse in some way), but that (ii) nevertheless psychoanalytic categories are not essential to (horror) films (i.e., that perversion, etc. is not essential to film and spectatorship), so that films need not necessarily be psychoanalytically interpreted, is an important one. This view leaves open the possibility that, contrary to what may now in fact be the case, films can be non-perverse, and that spectatorship and the pleasures of the cinema need not necessarily involve some combination of voyeurism, fetishism, sadism, etc. psychoanalytically understood. (20)

Works Cited

Carroll, N. (1990). The Philosophy of Horror; or, Paradoxes of the Heart. New York, Routledge.

Chodorow, N. (1987). Feminism and Psychoanalytic Theory. New Haven, Yale University Press.

Cioffi, F. (1988). “‘Exegetical Myth-Making’ in Grunbaum’s Indictment of Popper and Exoneration of Freud.” Mind, Psychoanalysis and Science. Clark, P. and C. Wright (eds.). Oxford, Blackwell: 61-87.

Cioffi, F. (1985). “Psychoanalysis, Pseudo-Science and Testability.” Popper and the Human Sciences. Currie, G. and A. Musgrave (eds.). Dordrecht, Nijhoff: 13-44.

Clark, P. and C. Wright, Eds. (1988). Mind, Psychoanalysis and Science. Oxford, Blackwell.

Crane, J. L. (1994). Terror and Everyday Life: Singular Moments in the History of the Horror Film. London, Sage Publications.

Creed, B. (1993). The Monstrous-Feminine: Film, Feminism, Psychoanalysis. New York, Routledge.

Erwin, E. (1996). A Final Accounting: Philosophical and Empirical Issues in Freudian Psychology. Cambridge, MA., MIT Press.

Erwin, E. (1993). “Philosophers on Freudianism: An Examination of Replies to Grunbaum’s ‘Foundations’.” Philosophical Problems of the Internal and External Worlds: Essays in the Philosophy of Alfred Grunbaum. Earman, J., A.I. Janis, G.J. Massey, and N. Rescher (eds.). Pittsburgh, University of Pittsburgh Press: 409-460.

Freud, S. (1915-17). Introductory Lectures on Psychoanalysis. The Standard Edition of the The Complete Psychological Works of Sigmund Freud XV-XVI. Strachey, J. (ed.). London, The Hogarth Press.

Freud, S. (1920). Beyond the Pleasure Principle. SE XVIII.

Freud, S. (1901). The Psychopathology of Everyday Life. SE VI.

Gaut, B. (1994). “On Cinema and Perversion.” Film and Philosophy 1: 3-17.

Geertz, C. (1973). “Religion as a Cultural System.” The Interpretation of Cultures. New York, Basic Books: 87-125.

Glymour, C. (1991). “Freud’s Androids.” The Cambridge Companion to Freud. Neu, J. (ed.). Cambridge, Cambridge University Press: 44-85.

Grixti, J. (1989). Terrors of Uncertainty: The Cultural Contexts of Horror Fiction. New York, Routledge.

Grunbaum, A. (1984). The Foundations of Psychoanalysis: A Philosophical Critique. Berkeley, University of California Press.

Grunbaum, A. (1988). “Precis of The Foundations of Psychoanalysis.” Mind, Psychoanalysis and Science. Clark, P. and C. Wright (eds.). Oxford, Blackwell: 3-32.

Grunbaum, A. (1993). Validation in the Clinical Theory of Psychoanalysis: A Study in the Philosophy of Psychoanalysis. Madison, CN., International Universities Press.

Hopkins, J. (1988). “Epistemology and Depth Psychology: Critical Notes on The Foundations of Psychoanalysis.” Mind, Psychoanalysis and Science. Clark, P. and C. Wright (eds.). Oxford, Blackwell: 33-60.

Hopkins, J. (1982). “Introduction: Philosophy and Psychoanalysis.” Philosophical Essays on Freud. Hopkins, J. and R. Wollheim (eds.). New York, Cambridge University Press: vii-xlv.

Lear, J. (1990). Love and Its Place in Nature: A Philosophical Interpretation of Freudian Psychoanalysis. New York, Noonday Press.

Levine, M., Ed. (2000). The Analytic Freud: Philosophy and Psychoanalysis. New York, Routledge.

Levine, M. (2001). “Depraved Spectators and Impossible Audiences.” Film and Philosophy 5.

Macmillan, M. 1997, Freud Evaluated (1991), Cambridge, MA., MIT Press, 1997.

Sachs, D. (1982). “On Freud’s Doctrine of Emotions.” Philosophical Essays on Freud. Hopkins, J. and R. Wollheim (eds.). New York, Cambridge University Press: 92-105.

Prince, S. (1988). “Dread, Taboo, and The Thing: Toward a Social Theory of the Horror Film.” Wide Angle 10.3: 19-29.

Prince, S. (1996). “Psychoanalytic Film Theory and the Problem of the Missing Spectator.” Post-Theory: Reconstructing Film Studies. Bordwell, D. and N. Carroll. Madison, WI: University of Wisconsin Press: 71-86.

Robinson, C. L. (1993). Freud and His Critics. Berkeley, U.C. Press.

Schneider, S. (2000). “Monsters as (Uncanny) Metaphors: Freud, Lakoff, and the Representation of Monstrosity in Cinematic Horror.” Horror Film Reader. Silver, A. and J. Ursini (eds.). New York, Limelight Editions: 167-91.

Schneider, S. (1997). “Uncanny Realism and the Decline of the Modern Horror Film.” Paradoxa: Studies in World Literary Genres 3.3/4: 417-28.

Tudor, A. (1997). “Why Horror? The Peculiar Pleasures of a Popular Genre.” Cultural Studies 11.3: 443-63.

Wood, R. (1979). “An Introduction to the American Horror Film.” American Nightmare: Essays on the Horror Film. Britton, A., R. Lippe, T. Williams, and R. Wood. Toronto, Festival of Festivals.

Young-Breuhl, E. (1996). The Anatomy of Prejudices. Cambridge, Harvard University Press.


  1. Grünbaum actually claims that psychoanalysis is falsifiable, and that by failing the standards set out in Freud’s Tally Argument it has been falsified. But this is really just a needless complication, since the standards for falsifiability that Grünbaum applies-those of the Tally Argument-are inapplicable. Psychoanalysis can never be falsified by means of the Tally Argument.
  2. “Psychoanalysis” refers both to psychoanalytic theory and practice. In this paper it refers mostly to the former.
  3. See Glymour 1991. Depending of course on what one takes cognitive science and cognitive psychology to be, there is no reason why these have to be mutually exclusive.
  4. See Macmillan 1997 for an extensive bibliography. It goes largely unnoticed that much of the recent criticism is a reiteration, in substance and tone, of criticisms Freud faced throughout his career. For essays that discuss the implications and significance of psychoanalysis, see (e.g.) Levine 2000. Also see the opening paragraph of Sachs 1982: 92.
  5. Hopkins says that “Schematically, Grünbaum and many he criticizes as ‘hermeneuts’ agree that causal claims generally cannot be supported other than in accord with scientific (e.g., inductivist) canons, and that much psychoanalytic evidence about motives is non-canonical. One party sees that psychoanalytic accounts of motive have non-canonical support and so ignores the causal role of motives, while the other keeps causality in clear view but ignores non-canonical support. Neither draws the obvious conclusion from the fact of non-canonical evidence for the causal role of motives, namely that the canons leave evidence on certain causes-motives-out of account. This, I think, is because neither attends to the way commonsense understanding uses and displays causal information” (1988: 50).
  6. Grünbaum (1988:14) notes that Freud “asserts it with respect to the ‘psychoneuroses’ [those caused by repressed infantile experiences] as distinct from the so-called actual neuroses [those caused by aetiological factors in an ‘adult’s current life situation’].”
  7. Grünbaum’s account of “actual neuroses” versus psychoneuroses may not be accurate. Actual neuroses result in symptoms that lack the meaning or sense which the symptoms of psychoneuroses have. And children can have actual neuroses, too. See Hopkins 1988 for a detailed and wide-ranging discussion of Grünbaum on the Tally Argument.
  8. Clark and Wright explain Hopkins’ critique of Grünbaum as follows:

    The explanatory inferences characteristic of much of psychoanalysis, Hopkins argues, are comparable to those in commonsense psychology in which a desire or motive is postulated as a cause of action. To be sure, in the explanation of action, motives together with beliefs produce actions aimed at the satisfaction of desire; whereas, in, for example, the psychoanalytic explanation of dreams, unconscious motives produce wishes and mere representations of satisfaction, like wishful imagining. But in both cases, Hopkins argues, there is a relation of shared content between the description of motive and the resulting action, be it overt or simply an episode of imagining that which one desires. This connection of content present in the explanations both of commonsense psychology and Freudian theory, Hopkins suggests, makes for a kind of explanatory achievement which is not possible in the case of the purely causal claims typical of scientific theory. If this is overlooked, Hopkins claims, the degree of support for Freudian theory ill be systematically understated; and is, in the Hopkins’s view, by Grünbaum (ix).

  9. Hopkins notes that Grünbaum (1984; 1988) “might have argued that the ‘meaning kinship’ between thirst and drinking does not vouch for causal connection in the dream of drinking. Since such interpretation-like all interpretation-turns on the thematic affinity of presumed causes and effects, we can see that the methodology which Grünbaum here uses against Freud would, if applied, render all everyday understanding groundless” (1988: 52).
  10. See Chodorow 1987 for a discussion of how psychoanalytic theory illuminates sexism, and why feminist thought needs psychoanalytic theory. See Young-Breuhl 1996 for a discussion of psychoanalytic explanations of racism.
  11. Without undermining the significance of clinical data, it is clear that in formulating various hypotheses Freud relied on data from everyday life as well. See Freud 1901.
  12. This is Tamas Pataki’s interpretation of what Prince is after.
  13. See Tudor (1997), who asks, “So what is its [the horror film’s] appeal? Although scholars of different disciplinary persuasions have recently researched the genre at some length,.none of them.provides an entirely satisfactory answer to that question. In part the problem is empirical. These studies are unable to marshal any more than anecdotal evidence as to the composition and preferences of horror audiences and so are forced to build their arguments on what may he ill-founded speculations” (444). However, the problem is not basically empirical (even in part), but interpretive. How can further evidence as to as to “the composition and preferences of horror audiences” help answer the question about the appeal of horror- except by bolstering particular interpretive schema? Is more data needed? Is more data going to help bolster a psychoanalytic interpretation in terms of ‘the return of the repressed,” or some rival cognitive approach? Try to imagine or invent some data that would.
  14. Tudor sees these as separate questions: “‘What is it about people who like horror?’ and ‘what is it about horror that people like?'” (444). Psychoanalytically speaking, however, these are just two sides of the same coin.
  15. See Schneider 1997; 2000 for an explanation based on Freud’s 1919 essay, “The ‘Uncanny’.”
  16. See Freud 1920-1922, S.E. vol. 18.
  17. See Schneider 1997; 2000 for an explanation of how (i) the “return of the repressed,” and (ii) the reconfirmation of previously surmounted infantile beliefs, does address the question of the attraction of horror. Schneider also explains-again in terms of the above- why certain kinds of monsters and horror films lose their appeal and fail to frighten by failing to evoke the “uncanny.”
  18. Tudor (1997) cites The Exorcist (1973), Carrie (1976), Alien (1979), The Brood (1979), and The Hunger (1983) as examples of films in which the monstrous-feminine “does play an important role” (450). It is unclear how he can claim this given what appears to be his rejection of feminist film theory that employs “structural psychoanalysis.”
  19. See Hopkins’ (1982: xxi) discussion of the table cloth lady.
  20. It is the consideration of this possibility that leads to the Paradox of the Depraved Spectator. See Levine 2001.