Operationalizing the male gaze: reconstructing Laura Mulvey’s analysis of visual pleasure

Abstract

Laura Mulvey’s ‘Visual pleasure and narrative cinema’ (1975) is the most widely cited, heavily anthologized and endlessly summarized essay in Film Studies. It is a foundational text that frames mainstream narrative cinema as a type of discourse that disseminates patriarchal ideology. Rather than summarize the essay again or repeat the many critiques developed by other scholars, this essay presents a line-by-line commentary on Mulvey’s interpretation of visual features in Josef von Sternberg’s films presented in section III C2 of her essay, in order to determine the logical relation between those visual features and theoretical concepts (especially fetishism). This involves rationally reconstructing the implicit operational procedures (empirical methods and evidence) Mulvey employed to interpret patriarchal values in mainstream narrative cinema. Operationalizing demystifies Mulvey’s interpretive activity by bridging the gap between her abstract theoretical predicates and filmic images, rendering the act of interpretation replicable by students new to film analysis.

Introduction

Laura Mulvey’s ‘Visual pleasure and narrative cinema’ (1975) is the most widely cited, heavily anthologized and endlessly summarized essay in Film Studies. ¹ It is a foundational text that frames mainstream narrative cinema as a type of discourse that disseminates patriarchal ideology. Like other ‘contemporary film theorists’ of the 1970s, Mulvey combined semiotics, Marxism and psychoanalysis to develop a politically motivated type of film analysis (innovative in the 1970s) that generated radically new knowledge about popular narrative cinema by revealing that its aesthetics and narrative are structured by patriarchy. I do not summarize the essay here or repeat the many critiques developed by other scholars; instead, I present a line-by-line commentary on Mulvey’s interpretation of visual features in Josef von Sternberg’s films presented in section III C2 of her essay (pp. 14–15), in order to determine the logical relation between those textual features and theoretical concepts (especially fetishism). This involves rationally reconstructing the implicit operational procedures (empirical methods and evidence) Mulvey employed to interpret patriarchal values in mainstream narrative cinema.

The need to operationalize ‘Visual pleasure and narrative cinema’ emerges from a challenging pedagogical exercise – to introduce novice film students to the practice of feminist film analysis by training them to interpret a new narrative film in the same way Mulvey interpreted Sternberg and Hitchcock. My approach to explicating the conventions of Mulvey’s interpretive activity, her ‘way of doing’ things, is through operationalism. For Hasok Chang, ‘the operationalist dictum could be phrased as follows: increase the empirical content of theories by the use of operationally well-defined concepts’ (2004: 147). Operationalizing Mulvey’s essay involves: (i) setting up referring expressions rich in descriptive content that identify the observable features of film that Mulvey’s essay denotes; and (ii) presenting rule-bound procedures to follow in order to apply those referring expressions to a new film example. Ideally, the referring expressions need to be sufficiently precise to ensure that students can apply them unambiguously. Of course, advocating rule-following risks promoting learning by imitation rather than by understanding. But, in their well-known model of expertise, Hubert and Stuart Dreyfus (2009: 9–10) argue that following clearly defined rules is fundamental to training a novice to become proficient, competent, and eventually an expert:

The instruction process begins with the instructor decomposing the task environment into contextfree features that the beginner can recognize without benefit of experience. The beginner is then given rules for determining actions on the basis of these features, like a computer following a program. Through instruction, the novice acquires rules for drawing conclusions or for determining actions based upon facts and features of the situation that are recognizable without experience in the skill domain being learned.

Interpreting film is a complex activity whose rules can be operationalized, taught and followed. To become proficient, a beginner needs to internalize these rules and apply them to new examples, making discoveries and learning something new. In other words, a beginner’s practice improves not only via experience but also via a deeper understanding of rules. Reconstructing Mulvey’s theory as a series of operational rules develops this deeper type of understanding, for it makes her theory explanatory – it emphasizes, for example, that a fetishistic image is not a self-evident fact. Operationalism explains exactly how a specific system of interconnected codes of a filmic image can be interpreted as a fetish.

But, these rules should not be followed mechanically, without understanding; operationalizing should not be conceived as a demand for blind rule-following analogous to a computer following a program. Rather, it should be presented as a guide:

A being can be said to be following a rule only in the context of a complex practice involving actual and potential activities of justifying, noticing mistakes and correcting them by reference to the rule, criticizing deviations from the rule, and, if called upon, explaining an action as being in accordance with the rule and teaching others what counts as following a rule. (Bennett and Hacker, 2008: 256)

Rules constitute a common foundation that justifies an interpretation, identifies mistakes by defining them as deviations from the rules and ensures the resulting interpretation can be replicated or repeated by others to achieve the same results. But, as Dreyfus and Dreyfus (2009: 22) stress, ‘learning more and more rules and routines to structure practice works only as a temporary measure to shore up the safety and suspension of the novice in highly complex unfamiliar situations’ (emphasis added). Temporary rules make complex situations familiar to novices, and it is precisely when these situations become familiar that novices transcend their status as novices. My operationalizing of ‘Visual pleasure and narrative cinema’ spells out the rules necessary for a novice entering a highly complex situation unfamiliar to them – undertaking a psychoanalytic interpretation of a mainstream film following Mulvey’s feminist agenda. However, following those rules is a temporary measure that aims to assist novices to transition to a higher level of knowledge.

Although brief, Mulvey’s case studies of Sternberg and Hitchcock influenced the way a generation of film scholars and students carried out feminist film analysis in the decades since its publication in 1975. To this day, Mulvey’s essay is still used as an initiation into feminist film theory and analysis, and remains one of the most cited and downloaded articles from the journal Screen.

Mulvey’s Act of ‘Initial Baptism’

What implicit series of operations did Mulvey follow to carry out her interpretation of Sternberg (the Sternberg–Dietrich films)? How did she bridge the gap between abstract theoretical concepts (fetishism, voyeurism) and Sternberg’s images? We can answer this by examining how the referring expressions in Mulvey’s interpretation identify relevant features of those images, which requires a short detour through internal (descriptivist) and external (causal) theories of reference. For descriptivists, referring expressions do not directly denote their referents, for the act of referring is mediated through descriptions (Bertrand Russell) or sense (Gottlob Frege), and the referent is whatever fits the description/sense. In other words, a description determines reference even though it is not causally linked to a referent. This means that a modified description may designate a different referent, or an imprecise description may not function correctly as a referring expression (in the same way that a set of vague operations for measuring length may lead to inconsistent measurements). Descriptivism is contested by causal theories of reference (Kripke, 1980; Putnam, 1975), in which reference is directly fixed via an ‘initial baptism’ and is spread by the speakers of a language community via a chain of communication without any need for a description. Although both theories were initially developed in relation to proper names, the arguments have been extended to nouns (especially to mass nouns, such as water and gold). As with proper names, causal theorists argue that the referent of a noun is named via baptism rather than a description. Although Kripke (1980: 96–97) acknowledged that a description may accompany a baptism, he argued that it is not essential. Several philosophers have combined both theories into a hybrid, causal-descriptive theory of reference (e.g. Psillos, 2012), in which a descriptive apparatus is added to the causal relations. One advantage of a hybrid theory is that mismatches between referent and description can be explained. Whereas the causal theory tends to take a successful act of reference for granted (since the reference was fixed in an initial act of baptism), the descriptive theory accounts for failures in communication in terms of incomplete, inaccurate, or revised descriptions of the referent. A causal–descriptive theory, therefore, offers a balanced account of reference: the causal model explains how the initial act of reference was established and the descriptive model explains how reference is subsequently modified and revised by the language community.

Mulvey’s essay became influential precisely because it is one of the first theoretical texts to baptize stylistic and narrative features of Sternberg’s and Hitchcock’s films as patriarchal. ² In interpreting Sternberg and Hitchcock, Mulvey supplements her act of initial baptism with descriptions; however, those descriptions are fragmentary and generic. When referring to the way framing fetishizes the body of a woman, for example, Mulvey’s descriptive phrases do not sufficiently distinguish between framing in general (a universal feature of film form) and a specific style of framing with particular attributes that contribute to fetishism. Novices attempting to analyse an image as fetish following Mulvey therefore risk referencing any type of framing of a woman’s body. To replicate Mulvey’s interpretive activity, novices require a set of procedural rules and referring expressions to identify one type of framing – framing that fetishizes the woman’s body. The rules and expressions need to be exact, precise and rich in empirical detail to ensure that their act of reference is successful (that they pick out in a consistent way the same textual features). In more technical terms, the descriptions and operations need to establish identity relations that bridge the gap between high-level second-order abstract theoretical concepts, such as fetishism, and low-level first-order textual features of films, such as framing. Yet, while avoiding reference to all types of framing, etc., one should also avoid making the descriptions so specific that they refer only to the type of framing employed in Sternberg’s films. In other words, descriptive referring expressions need to be detailed but not too specific. In Husserl’s terms, descriptive referring expressions must identify the ‘determinable X’ – the abstract property that enables a referent to be identified and re-identified as the same even though it may undergo changes in appearance. ³

An interpretation entails theoretical concepts coming up against the empirical reality of film. The following line-by-line commentary on Mulvey’s interpretation of Sternberg’s images examines the descriptive empirical vocabulary in her referring expressions, which determines whether they express meaningful observations or empty metaphysical generalizations.

Mulvey on Josef Von Sternberg

Mulvey’s referring expressions aim to pick out textual features in the Sternberg–Dietrich films that satisfy the conditions expressed in the psychoanalytic definition of fetishism. Mulvey’s interpretation of Sternberg is carried out in one long paragraph (425 words) towards the end of the essay. These interpretations present a series of statements that identify a specific type of film style and narrative structure formed by patriarchal logic. What evidence does Mulvey provide for attributing hidden patriarchal causes to the formation of the style of Sternberg’s films? Patriarchy is an external cause that structures films into a specific pattern of signification. This specific pattern and its cause are Mulvey’s primary objects of investigation. We have no direct access to the cause (patriarchy), for it is a structuring absence in which only the effects are observable – effects that become symptoms of patriarchy embedded in both film style and narrative structure. Analysis of these patterns of patriarchal symptoms begins with the fundamental features of images that have already been established in formalist film, which Mulvey then reinterpreted using psychoanalytic concepts of fetishism and voyeurism (and sadism, when discussing narrative in Hitchcock). These features constitute a new type of evidence: film style and narrative structure no longer attest to the aesthetic qualities of individual shots or scenes or to a film’s overall unity; instead, they serve as evidence of an external force –patriarchal logic – shaping these features. For example, reinterpreting a unified film style using the concept of fetishism explains the gendering effects of this style on male spectators: the unified images fetishize the woman’s body, rendering it ‘whole’ and ‘safe’ for the male spectator’s unconscious fear of the ‘castrated’ woman (more precisely, the fetishistic images disavowal a perceived or imagined lack). A unified film style is therefore predicated an additional property: it is fetishistic. More generally, Mulvey established new objects of study – the fetishistic image, the male gaze (voyeurism) and sadism in narrative structure – in order to demonstrate the influence of patriarchy upon mainstream cinema.

Mulvey not only retains the premises of formalist film theory, she also preserves many of the principles of traditional auteur theory. In the 1960s, auteur critics identified key features of Sternberg’s and Hitchcock’s films. In terms of Sternberg’s visual style, Raymond Durgnat (1976[1965]) analysed six of his films; Andrew Sarris (1966) placed Sternberg in his pantheon of film directors (Sarris, 1968: 74–77) and wrote a short book on him (Sarris, 1966); Sternberg’s name appears numerous times in Cahiers du Cinéma in the 1960s; writing under the pseudonym ‘Lee Russell’, Peter Wollen profiled him in the New Left Review (Russell, 1966), and later (using his real name) expanded and refined his comments in his seminal essay ‘The Auteur Theory’ (Wollen, 1969: ch. 2); while Carole Zucker (1988) continued this tradition in the 1980s in The Idea of the Image. The statements of these critics are aesthetic reflections that appear to have a high degree of observability. However, in terms of empirical confirmation, auteur theory is more problematic than formalist film theory, for auteur critics refer to a director’s typical stylistic traits – that is, they create an idealized model that denotes general features of a director’s style. In order to pick out the relevant features of a new filmic images that satisfy the theoretical postulates presented in ‘Visual pleasure and narrative cinema’, and to ensure the interpretations can be repeated by others, Mulvey’s general and idealized auteur model of Sternberg needs to be enriched with specific and detailed descriptions of their films’ visual features.

For Mulvey, Sternberg and Hitchcock represent two ideal types of film: those whose symptoms of patriarchy are located in the image (Sternberg) and those whose symptoms are located in the narrative (Hitchcock). Mulvey introduces Sternberg by paraphrasing his well-known anecdote that the content (story and characters) of his films is insignificant, for his films could just as well be projected upside down in order to appreciate the pure play of light and shadow (Mulvey, 1975: 14). Taken literally, Sternberg reduces his own films down to their low-level raw, concrete sense data –patterns of light and shade projected onto a flat screen. Mulvey modifies Sternberg’s anecdote, recouping the reality represented in these images – especially the figure of ‘the woman’. Once reinstated, this universal figure of the woman frames the remainder of Mulvey’s analysis of Sternberg’s films.

Frame and Space

After citing Sternberg’s anecdote, which privileges image over narrative, Mulvey begins with Sternberg’s distinctive style. The first statement Mulvey (1975: 14) makes is the descriptive phrase:

• the pictorial space enclosed by the frame is paramount rather than narrative or identification processes.

The words the, pictorial space, enclosed and frame in the first part of this phrase require comment. Frame and space constitute the first and second stylistic features of film presented in Mulvey’s interpretation of Sternberg, and both are promoted to the status of primary (‘paramount’) features. The frame is a pre-given formal feature of cinema, one of the film image’s conditions of possibility. What this frame encloses is another fundamental pre-given formal feature – space. Using the formal features frame and space to identify the specificity of Sternberg’s film style may seem odd, for style designates a specific use or manipulation of form. However, Mulvey qualifies space with the adjective pictorial (from the Latin pictorius, ‘of painters’). The phrase pictorial space confers upon Sternberg’s images the quality of having the aesthetic (rather than physical) features of a painting, thereby connotating the meaning that Sternberg’s films are artistic. Sternberg’s framing (and lighting) eschews symbolism and turns film space into pictorial space. At this early stage, Mulvey’s interpretation of Sternberg remains at the level of stylistic and auteur analysis. ⁴ Nonetheless, her analysis eventually feeds into her ‘image of woman’ (woman as spectacle) argument, which depends on the formation of the image of woman enclosed – that is, positioned and fixed – in film space. In other words, Mulvey’s analysis of the frame enclosing space develops into an argument that female characters are confined within a self-contained, framed space. Describing the space as enclosed further suggests that female characters tend to look ‘inward’ rather than off screen. This means that the character does not motivate a cut to the next shot, but instead becomes an object to be looked at rather than the subject controlling the look of the camera and motivating the cuts. In contrast, male characters, while also framed by the camera, look outwards beyond the frame, bringing into play off-screen space, thereby controlling the discursive movement of the film from shot to shot. This sense of control is heightened when the male character is an unseen voyeur.

Mulvey’s comments on frame and space do not refer to a specific frame enclosing a specific pictorial space; instead, she refers to ‘the pictorial space enclosed by the frame’ (emphases added). In other words, Mulvey discusses two features of Sternberg’s images using the generic form of the definite article the, which refers to an abstract class or category of things rather than to particular instances. Using a generic phrase weakens the referential function of the description. Mulvey’s initial empirical statement already raises an issue of the focus of her investigations, which appears to be a combination of film form (cinematic constants, signified through the generic form of the definite article the), film style (a director’s – in this instance, Sternberg’s – specific manipulation of film form) and symptomatic images (images interpreted as symptoms of an external cause – patriarchy –manipulating film form). No conflict exists between the auteurist and the symptomatic interpretations, for patriarchy can be read as mediated through the (male) director; the main conflict is between the focus on form (as a universal feature of film) and style (a specific manipulation of film form): are patriarchal values inherent in film form, or are they evident only in film style (a specific manipulation of film form)? Blurring the boundary between form (the universal) and style has created confusion in carrying out feminist interpretations as well as determining whether an alternative (non-patriarchal) style – an alternative use of film form – is possible.

Mulvey ends the phrase just analysed with the words ‘rather than narrative or identification processes’, which sets up a conceptual opposition between ‘pictorial space enclosed by the frame’, on the one hand, and ‘narrative or identification processes’, on the other. And, like all binary oppositions operating within an ideological (in this instance, patriarchal) framework, the terms are not equal but are organized into a hierarchy: in Sternberg’s films, Mulvey argues, ‘framed pictorial space’ is given prominence over ‘narrative or identification processes’. ⁵ These ‘narrative or identification processes’ cannot be observed directly, in two senses: firstly, Mulvey is here describing the marginality of narrative and identification processes in Sternberg’s films. Secondly, narrative is a structure that is not given directly in experience because it structures experience; it is not immediately perceivable and is not, therefore, directly reducible to observation. Identification processes can be reduced to observable terms, but Mulvey again claims that they are absent in Sternberg’s films.

Fetishism

• Sternberg produces the ultimate fetish, taking it to the point where the powerful look of the male protagonist (characteristic of traditional narrative film) is broken in favour of the image in direct erotic rapport with the spectator. (Mulvey, 1975: 14)

The statement ‘the image in direct erotic rapport with the spectator’ asserts an immediate (‘direct’) relation between the spectator and the image. Direct in what sense? Mulvey does not mean the spectator’s relation to the image is an unmediated direct perception, for her essay is premised on theorizing this relation as phantasmatic, mediated via fetishism, voyeurism and sadism. Instead, Mulvey refers to the absence of one contingent aspect of mediation – the look of a male character. The narrative investigative look embodied in a male character is downplayed in Sternberg’s films in favour of the direct relation between the spectator’s look and the image.

This direct relation is further attributed the quality of ‘eroticism’. How can eroticism be identified and confirmed? Fetishism is a key property of eroticism. In its literal sense, fetish means artificial (to make by art). In religion, the fetish is a sacred inanimate object that is venerated because it embodies a divine spirit. In psychoanalysis, the fetish becomes an artificial secular object of worship: the fetishist overinvests in the value of an object that generates sexual arousal by functioning to contain lack. In Freudian psychoanalysis, the fetish is interpreted as supplementing the male subject’s traumatic perception of lack in the female body: the fetish acts as a substitute object for the ‘missing’ phallic object; it tries to protect the male subject from castration anxiety that the woman’s body represents to him (a body which patriarchal ideology codifies as lacking). A fetish is an object that negotiates lack by standing in for the ‘missing’ phallus. Mulvey identifies features analogous to the fetish in Sternberg’s images of women: specific film images of female characters are similar to and function like a fetish object. The analogy aims to establish a correspondence between abstract features of the psychoanalytic concept of fetishism and empirical features of certain film images. A fetishistic image is not a self-evident fact; it is not given in advance. For an image to function like a fetish, as an object that stands in for the ‘missing’ phallus, it needs to meet certain empirical conditions. How do specific parts of the image function as a fetish? Mulvey answers this question in the next statement.

Close-ups, editing and stillness

• The beauty of the woman as object and screen space coalesce; she is no longer the bearer of guilt but a perfect product, whose body, stylised and fragmented by close-ups, is the content of the film and the direct recipient of the spectator’s look. (Mulvey, 1975: 14)

Here Mulvey reinforces and expands her claim in the previous statement that women are depicted from a patriarchal perspective in mainstream films, and Sternberg’s films in particular. The woman’s body is ‘the direct recipient of the spectator’s look’ – that is, the male spectator’s look, unmediated through a character (but still mediated through the codes of cinema). Mulvey repeats the opposition between image and narrative by claiming that the woman ‘is no longer the bearer of guilt but a perfect product, whose body, stylised and fragmented by close-ups . . .’ Here, Mulvey develops her discussion of the image of woman via a technical element of film style – the close-up – and its multiple effects: stylization, fragmentation, perfect product. The close-up is the third stylistic element of film presented in Mulvey’s interpretation of Sternberg, after frame and enclosed space. The close-up, in fact, is a particular type of frame, a stylistic choice that restricts the spectator’s field of vision by enclosing a small area of space: it extracts a limited but significant part of the scene, thereby fragmenting and isolating it from the rest of the scene. Moreover, Mulvey does not discuss a single close-up but a series of close-ups combined via editing. These close-ups create a stylized (pictorial) image that fragments both space and the woman’s body, breaks them up into pieces but then recombines them via editing. The editing needs to arrange the series of spatial fragments into a unity that conceals absence and exaggerates the female body, thereby turning it into a fetish. Stephen Heath (1981: 40) expressed this idea in theoretical terms: ‘The need is to cut up and then join together in a kind of spatial Aufhebung that decides a superior unity . . . fragmentation is the condition of a fundamental continuity.’ Heath recognized an apparent paradox of narrative film: its ‘superior unity’, its continuity, can only be created through an initial fragmentation. This follows the process of reification, where the space and time recorded in each shot are abstracted from their original everyday context – they are broken up and recombined to create a new, artificial (filmic) space and time.

‘A perfect product’. Perfect suggests a complete aesthetic stylization of the image of woman, while product suggests that this perfect image of the woman is manufactured – it is made into a perfect product. In this respect, we can note in passing that perfect derives in part from the same word as fetish: facere, ‘to make’ (fetish: to make by art; perfect: to make complete). Mainstream cinema creates from fragments the impression of a perfect or unified image of woman, and this perfect image is manufactured via the logic of fetishism.

In stating that ‘The beauty of the woman as object and screen space coalesce’, Mulvey sets up three separate entities – woman, object and screen space – and argues that they are subsequently combined to form a whole. The Oxford English Dictionary gives the original meaning of object as ‘something placed before or presented to the eyes or other senses’ and cites a technical meaning in which the object is ‘the thing or body observed with an optical instrument’. In this technical sense, it is an optical instrument that places the object before the observer’s eyes. The film camera is an optical instrument whose close-up framing places the woman before the spectator’s eyes by producing an enclosed screen space that reduces her to her physical appearance. The body of the woman is therefore defined as an object positioned and located in filmic space, that is, an object seen and represented from the perspective of film, whose codes (close-ups, the frame, camera distance) are manipulated by an external force, patriarchal logic, to create an erotic image of the woman’s body.

Within a fetishistic sequence, editing functions descriptively, portraying a woman through a series of close-ups that imply no temporal progression from shot to shot. The woman is transformed into an autonomous spectacle, an attraction (in Eisenstein’s sense of the term), a frozen or still representation that is both part of the film’s diegesis but also separate from it. The stillness of the fetishistic image is the fourth stylistic element of film presented in Mulvey’s interpretation of Sternberg. A few pages before introducing Sternberg, Mulvey (1975: 12) developed these ideas of stillness and attraction by referring to ‘the device of the show-girl’, whose performance momentarily freezes time, ‘takes the film into a no-man’s-land outside its own time and space’. The stillness of the fetishistic image momentarily pauses the film’s narrative progression driven by voyeurism and sadism.

Flat images

• Sternberg plays down the illusion of screen depth; his screen tends to be one-dimensional, as light and shade, lace, steam, foliage, net, streamers, etc. reduce the visual field. (Mulvey, 1975: 14)

The flat image is the fifth stylistic element of film presented in Mulvey’s interpretation of Sternberg, after frame, enclosed space, close-up and stillness. It is defined both negatively (it downplays the illusion of spatial depth, reduction of the visual field) and positively, in terms of a list of specific empirical features: chiaroscuro (light and shade), lace, steam, foliage, net, streamers. How do these elements create a reduction of screen depth? We can see from observing these features in Sternberg’s films that they are placed in the foreground of the image, which reduces or blocks visual access to the background. Such shots are therefore the opposite of the deep focus images in Renoir, Wyler and Welles praised by André Bazin (1967: 35), who argued that such shots ‘bring the spectator into a relation with the image closer to that which he enjoys with reality’. In creating the opposite of the deep focus shot, Sternberg presents a more abstract, non-realistic, artificially manufactured image.

Mulvey also writes about Sternberg’s screen rather than image, a word choice that establishes a similarity between image and screen, suggesting that Sternberg’s flat images are reflexive to the extent that they imitate the flat screen on which they are projected: Sternberg does not try to conceal the flatness of the image with the spatial illusion of depth. Earlier, Mulvey (1975: 12) noted that ‘one part of a fragmented body destroys the Renaissance space, the illusion of depth demanded by the narrative, it gives flatness, the quality of a cut-out or icon rather than verisimilitude to the screen’. Flatness therefore contributes to the formation of Sternberg’s stylized pictorial space and, crucially, it also facilitates the coalescing of woman and screen space mentioned in the previous statement.

Non-mediated look

• There is little or no mediation of the look through the eyes of the main male protagonist. On the contrary, shadowy presences like La Bessière in Morocco act as surrogates for the director, detached as they are from audience identification. (Mulvey, 1975: 14)

Here, Mulvey restates the lack of mediation between spectator and image in Sternberg’s films – a lack of mediation in the sense of the male character’s look. ‘Little or no mediation’ is an observable and quantifiable statement, and it is the first time that Mulvey cites a specific film as evidence, although no definite moment in the film is mentioned. Instead, she refers to the male character La Bessière (Adolphe Menjou) in Morocco as one instance of a shadowy figure who stands in for the director.

Sternberg’s narrative

• Despite Sternberg’s insistence that his stories are irrelevant, it is significant that they are concerned with situation, not suspense, and cyclical rather than linear time, while plot complications revolve around misunderstanding rather than conflict. (Mulvey, 1975: 14–15)

Mulvey argues that narrative still has a place in Sternberg’s films, a point she makes by setting up (without developing further) a series of oppositions between concepts central to narrative (suspense, linear time, conflict) to those that are secondary (situation, cyclical time, misunderstanding). Conflict ties in with suspense, for spectators receive information about a conflict and anticipate its unfolding and resolution. Mulvey implies that Sternberg’s narratives consist of nondramatic sequences of actions and events that are organized into a cyclical structure and that no fundamental disagreement exists between characters, merely misapprehension.

• The most important absence is that of the controlling male gaze within the screen scene. The high point of emotional drama in the most typical Dietrich films, her supreme moments of erotic meaning, take place in the absence of the man she loves in the fiction. (Mulvey, 1975: 15)

By now, Mulvey’s empirical statement about the absence of the controlling male gaze within Sternberg’s films is clearly established. Mulvey repeats it to emphasize the way its absence functions in Sternberg’s narratives (especially the six films starring Marlene Dietrich): the high point of the female character’s emotional drama is witnessed by the spectator but not by the man she loves. These moments are, however, quite rare, for in Sternberg’s films (and classical Hollywood cinema more generally), female characters rarely appear in scenes in the absence of male characters. Mulvey ends her interpretation of Sternberg with two examples:

• At the end of Morocco, Tom Brown has already disappeared into the desert when Amy Jolly kicks off her gold sandals and walks after him. At the end of Dishonoured, Kranau is indifferent to the fate of Magda. In both cases, the erotic impact, sanctified by death, is displayed as a spectacle for the audience. The male hero misunderstands and, above all, does not see. (Mulvey, 1975: 15)

Mulvey names two instances in Sternberg’s films (Morocco and Dishonoured) where the female characters’ emotional drama is shared with the spectator but takes place beyond the male hero’s sight and knowledge. The male character’s gaze is absent while the spectator’s gaze is prominent, and that gaze is directed at a moment of drama rather than a spectacle. In terms of narrative knowledge, the male hero knows less than the spectator and female characters, who are positioned on the same level.

‘Image of Woman’ Defined Operationally

Mulvey’s referring expressions are only partly successful in identifying the features of film images relevant to a fetishistic interpretation because their descriptions are brief and elliptical, consisting of atomistic propositions, each of which refers to an individual feature of film. This gives the impression that concepts like fetishism are embedded in these features taken individually. To operationalize Mulvey’s analysis requires more detailed descriptive expressions. However, a detailed description of each feature in isolation will simply perpetuate the atomistic view of features; instead, we need to examine specific combinations of stylistic features, for it is only a cluster or interconnected set of features that can be labelled fetishistic. This is because fetishism is of a higher logical type than individual empirical features such as the film frame or shot scale. As a theoretical concept, fetishism is not, therefore, located in the individual features of films, and it is not located in just any aggregate of multiple features; instead, it is a supervenient quality that emerges from a specific structure of multiple empirical features of film that are not in themselves fetishistic. ⁶ In practice, due to Mulvey’s atomistic perspective, brief descriptive expressions and a lack of operational procedures, many novices who apply her ideas to new film examples all too often decide immediately and conclusively that any series of images of a woman are fetishistic. But this does not constitute an interpretation; instead, it amounts to confirmation bias – a top-down activity that imposes an interpretation on a film and then selectively picks data that supposedly confirms that interpretation. To avoid confirmation bias, an interpretation needs to develop from the bottom up: begin with a close and detailed analysis that identifies the combination of multiple visual features that tentatively appear to be fetishistic. Only then determine if they are fetishistic or not based on interpretive rules and on descriptions of an interconnected set of features, ensuring that the initial tentative hypothesis is tested and is confirmed or discarded.

The following comments present a first attempt to operationalize Mulvey’s ‘Image of Woman’ concept via a series of descriptions and rules. Firstly, here is a recap of the specific empirical features that Mulvey identified:

• The frame (a visual boundary).

• Enclosed space (defined by the frame which isolates the space within it, making it autonomous).

• Close-up (measured in terms of frame, space and camera distance), which has the effect of fragmenting the woman’s body.

• Editing: a series of fragmented shots reified into a unity of space with no strong temporal progression from shot to shot (time is frozen).

• Flat image (absence of screen depth) caused by:

• ○ absence of the male character’s look at the female character (and, implicitly, the woman’s look is inward, contained within the frame).

• The story is organized around secondary narrative elements such as situation, cyclical time and misunderstanding, rather than primary elements such as suspense, linear time and conflict.

To meet the conditions of fetishism, these visual features of film need to be combined and arranged into a specific composition that creates unity by concealing absence. These conditions can be written out as a set of explicitly defined rules and procedures formulated as a series of questions:

1. Begin with the image’s fixed border, the frame. But not just any frame; instead ask: does the frame function in a specific way—does it shut in and isolate the space within it? In other words, does the framing make the image autonomous and complete, or does it create an incomplete image that draws attention to off-screen space (which would motivate a cut to the next image)? Only autonomous enclosed framing can contribute to fetishism, for it objectifies the female character’s body by confining it within this autonomous space.

2. Does the lighting (e.g. a spotlight) create a chiaroscuro effect, drawing attention to the woman’s body and, at the same time, isolating it from the surrounding space?

3. Does this enclosed type of framing move in close to the female character, abstracting her face (and/or other body parts) from the surrounding space and décor?

4. Does the foreground contain objects (such as lace, steam, smoke, foliage, nets and streamers), and is the woman’s body positioned immediately behind those objects? This type of composition flattens out the image, reducing the body’s three-dimensionality.

5. When edited together, do the flat, autonomous, close-up images of the female character’s spot-lit body create the impression of a unified space with no strong temporal progression from image to image?

6. Are the images of the female character’s body in direct rapport with the spectator, or are they mediated through a male character’s vision?

If we identify these empirical features in a series of images, then – hypothetically at least – we can map the abstract theoretical predicate is fetishistic onto them.

We can illustrate the operational procedures just outlined by briefly analysing the wedding scene from Sternberg’s The Scarlet Empress (1934). The scene comprises 29 shots over 5 minutes and 20 seconds, including 9 shots isolating Catherine (Marlene Dietrich) in her own space (8 of her face, 1 of her hand): 1. The frame initially shuts in and isolates the space within it, of Catherine during her wedding ceremony. However, the image is not completely autonomous, for an incense burner swings back and forth into and out of some of the images, and the Priest’s hands enter the frame during Holy Communion and the Ring Ceremony; 2. The lighting (including the candle Catherine holds) highlights Catherine in her white dress against a dark background; 3. As the scene progresses, the camera focuses on Catherine’s face; 4. The incense burner billows smoke into the foreground, and Catherine’s wedding veil covers her face, flattening the images of her; 5. As the scene unfolds, the camera cuts in closer, creating a series of shots fragmenting and flattening the space and Catherine’s face; 6. At first, Catherine is contemplative, looking inwards. But, as the scene progresses, she looks off-screen left, which motivates a cut to Count Alexi, who is looking off-screen right back at her. The images of Catherine are no longer autonomous and in direct rapport with the spectator but are mediated through a male character (not, however, his optical point of view; nonetheless, the shots of Catherine now represent Count Alexi’s awareness – he claims these images, or at least competes with the spectator).

How can we justify an interpretation of a series of film images as fetishistic? If fetishism is non-demonstrative (that is, not deductively entailed from the empirical features of a series of filmic images), then the ‘determinable X’ (in this instance, the property of fetishism) cannot be defined unequivocally. Additional support for an interpretation of a series of images in The Scarlet Empress (or any other film) as fetishistic needs to overcome at least two issues: ambiguity in measuring the individual observable features and the quantity of observable features that are needed to constitute a fetishistic sequence of images. In regard to ambiguity, can an enclosed, autonomous space be unequivocally identified? Does the incense burner swinging in and out of the image in The Scarlet Empress negate its autonomy (by drawing attention to off-screen space) or does it only weaken it? When does an image cease to manifest depth and become a flat image? No matter how detailed they are, the descriptive phrases of referring expressions can never be 100 percent definitive. In regard to quantity, is it possible to identify additional features that Mulvey did not consider (the duration of a shot, for example)? Furthermore, features 2, 3 and 4 are dominant in the scene from The Scarlet Empress, with 1 and 6 initially dominant at the beginning but diminishing as the scene unfolds, while 5 does not apply at all. When does a series of images become fetishistic? That is, what is the threshold/borderline for the minimum number of features for a visual system to become fetishistic? Are some features necessary, or are several features merely sufficient? And can fetishism be manifest via several different interconnected combinations of features, or is it manifest by one unique combination of features? No absolute criteria exist that can answer these questions. Operationalism would need to formulate a series of detailed descriptive phrases and rules to address each issue.

The more studies that test a theory, the stronger the case becomes for confirming or refuting that theory. Mulvey only presented a small number of carefully selected examples; however, the operational rules presented here not only enable novices to carry out an accurate analysis of visual pleasure; they also enable the empirical claims made in ‘Visual pleasure and narrative cinema’ to be rigorously and more comprehensively tested, supported, modified, expanded and transformed.