Kuleshov Effect

The Kuleshov Effect

From ‘Facial Expression, Montage, and the Kuleshov Fallacy’, Film, Art, and the Third Culture, 139 (on facial expression in Hitchcock’s Psycho)
Is the notion of the Kuleshov effect no more than a museum piece? Hardly. Arguments and assumptions of this type concerning the power of editing over other aspects of film technique, and particularly performance, are still very much alive. Consider, as a parallel example, Howard Goodall’s analysis of Bernard Herrmann’s score for Hitchcock’s Psycho (1960). Goodall claims that the extended sequence of shots showing Marion Crane (Janet Leigh) driving away from Phoenix would take on an entirely different emotional character if they were juxtaposed with upbeat music, rather than the unsettling harmonies and jagged rhythms of Herrmann’s score. That much is certainly true; there can be no doubting the ‘added value’[1] that sound and music bring to the film image in general, nor of the importance of Herrmann’s score in shaping our experience of the story of Marion Crane and Norman Bates in particular. Goodall cannot resist implying, however, that Leigh’s facial expression is entirely inert, and that her performance plays no significant role in conjuring up the mood of the scene. ‘What’s remarkable about [Herrmann’s cue for the scene] is how much impact it has on the pictures. Without it, what we’re looking at is someone driving along in a car, and there’s nothing dramatic, tense, or worrying about that’.[2] But what Goodall implies by this claim is manifestly false: across the sequence Leigh’s face betrays many symptoms of anxiety and agitation (nervous sideways glances, lip biting, strained squinting into the headlights of oncoming traffic as she drives long into the night). In this respect the example differs from the classic Mosjoukine case, for Leigh’s face is not expressively neutral. To be sure, a different score would interact with these facial expressions to obtain a different effect to the one actually achieved with Herrmann’s score (as Goodall demonstrates); but that does not secure the claim that Leigh’s facial expressions are playing no significant role. We are back to a version of the Kuleshov fallacy: only context – here provided by the score, rather than editing – counts in determining the meaning or emotional character of a given facial expression. In this respect, Goodall’s otherwise very compelling analysis of the musical character of Herrmann’s score and its contribution to the film strikes a false, if very familiar, note.
  1. On the notion of ‘added value’, see Michel Chion, Audiovision: Sound on Screen, trans. Claudia Gorbman (New York: Columbia University Press, 1994), chapter 1.
  2. Howard Goodall’s 20th Century Greats: Bernard Herrmann (Howard Goodall and Francis Hanly, 2004).
 

About Film, Art and The Third Culture

More Sequences 

startle effect

The Startle Effect in Kurosawa’s Ran

From ‘Startling Sights and Sounds,’ Film, Art, and the Third Culture, 95-7 (on the startle effect in Kurosawa’s Ran)

Now, it is obviously the case that our appreciation of this scene and the film as a whole depends upon a whole array of mental capacities that go far beyond the brief, involuntary, reflex-like reaction that is the startle response. In order to make even basic sense of the action, viewers need to understand the dialogue, to interpret facial and vocal expressions, and to figure out the spatial relations among the agents and features of the setting. If their understanding is to go beyond rudimentary comprehension, spectators will in addition need to bring to bear a great deal of culturally-specific knowledge – about Marvel Comics, Robert Downey Jnr, and the ‘war on terror’, for example – on the unfolding action. None of this knowledge, however, makes startle cues like the explosion in Iron Man, and the primitive response that it triggers, any less significant, any less a designed feature of the film that exploits a real and important feature of human psychology. Here the startle cue performs a number of functions: it marks a major narrative development (indeed, a plot point); it symbolizes in concrete form the danger posed by the Afghan insurgents; and it delivers a thrilling ‘whomp’ – one of those powerful bodily sensations so characteristic of action and horror movies, very much akin to the visceral, stomach-in-mouth pleasures of fairground rides.

With such disreputable associations in mind, Baird notes that the startle effect has been ‘[m]aligned as mindless and a hallmark of B-movies and exploitation fare’.[1] Cultural prejudice of this sort, however, should not stop us from seeking a deeper understanding of such genres, nor of the dispositions and capacities upon which they depend. Tallis and other traditionalists might be inclined to agree that the startle response is, precisely, ‘mindless’ – a reflex response of the body bereft of flexibility and variability. But such a stance is untenable. Unless we are prepared to embrace Cartesian dualism, human psychology cannot be regarded as dividing conveniently into a wholly mechanical, bodily dimension and an entirely voluntaristic mental dimension. Tallis’ own adage regarding ‘affirmative action for the body’ is ironically apt here. Tallis emphasizes that the brain cannot be understood properly in isolation from the body.[2] But it is equally true that cognition itself cannot be understood properly when detached from its neural and bodily grounding, since the cognitive capacities an organism possesses depend, at least in some ways and to some degree, on that organism’s (neuro-)physiology. As we saw Darwin putting it (in the epigraph to chapter 2), ‘the mind is [a] function of the body’.

In any event, the startle effect is not uniquely associated with the horror film or with other popular genres even though it has certainly been exploited most frequently and routinely in these types of film. Ran (1985), Akira Kurosawa’s free adaptation of Shakespeare’s King Lear, shows the effect at work in a much more ‘exalted’ cultural context.[3] In the film, the Lear-figure Hidetora (Tatsuya Nakadai) has ceded effective leadership of the Ichimonji clan to his first son, Taro (Akira Terao), while Hidetora retains his role as titular leader of the clan. Initially in residence at Taro’s castle, Hidetora leaves after arguing with him, eventually moving into the castle of his third son, Saburo (Daisuke Ryu). Taro and Jiro (Hidetora’s second son, played by Jinpachi Nezu) together mount a sustained and bloody attack on Saburo’s castle.

Across a lengthy sequence, we witness the extraordinary carnage of battle (figure 3.2), and the gradual progress of the attacking armies towards the capture of Saburo’s castle. All diegetic sound from the battle – the clash of armour, the galloping of horses, the cries of injury and death – is suspended, however; in its stead, we hear only Toru Takemitsu’s mournful, Mahler-inspired score. Until, that is, we see Taro, leading the offensive army and entering Saburo’s castle on horseback – at which point he is shot in the back, a single, loud gunshot terminating the musical cue abruptly and marking the restoration of diegetic sound.

As in the Iron Man sequence, a striking change in the overall rhythm of the scene, initiated by the occurrence of an unanticipated event incarnated by a sudden loud sound, prompts the startle response. And here, the force of the startle is both to underline a dramatic turning point, and to remind us of the physical brutality of war. The startle effect caps a process of defamiliarization, whereby our appreciation of what is being represented is heightened by the shift in style. First, we witness slaughter on a mass scale, through the stylized, ‘operatic’ phase of the sequence, when only the score is audible; then, with the re-entry of diegetic sound, the realistic potential of the visual imagery is drawn out. The startle generated by the gunshot marks the dividing line between the two styles; the effect of defamiliarization therefore turns on this moment. (We return again to the notion of defamiliarization in chapter 4.)

Thus far in this discussion of the startle response, I have assumed that the response itself is an invariant feature of human physiology; what variation of function we see arises from the contexts in which it is spontaneously triggered or exploited by design, as in Iron Man and Ran. In fact, to some extent, context may affect the form as well as the function of the startle response. ‘Unlike the knee-jerk, the sneeze, or the flinch’, writes Baird, ‘startle reflexes are modified by emotional and cognitive states’.[4] According to Peter Lang and his collaborators, ‘the vigor of the startle reflex varies systematically with an organism’s emotional state…the startle response (an aversive reflex) is enhanced during a fear state and is diminished in a pleasant emotional context’.[5] Baird points to the ‘threat scene’ as evidence of horror filmmakers’ understanding, whether fully conscious or more intuitive, of this fact. In such a scene, a character is presented as fearful of a hidden, often indeterminate, off-screen threat; after depicting the character’s anxiety for a period, the immediate space around the character is subjected to a sudden intrusion, typically from an unpredictable direction or in an unpredictable manner. The threat phase of the sequence primes us for the startle response; in an anxious and fearful state, we are more likely to be startled. (Baird suggests that this is why we find the startle effect much more frequently in horror and thriller films than in any other filmic context.) Threat scenes thus perform a balancing act between priming and misdirection: an unsettling, fearful mood is created, but within that broad affective context, our attention is taken away at the critical moment from the precise spatial location of the threat. (On the relationship between moods and emotions, and the priming of the latter by the former, see the discussion of Heimat in chapter 6.)

Baird analyses the bus scene from Cat People as an important historical exemplar of the threat scene. Alice (Jane Randolph) is seen making her way through the streets at night. She senses that she is being followed by someone or something, and glances several times to screen left. After this pattern has been sustained over some seconds, a bus abruptly and noisily enters from screen right. Note here the role and force of the deictic gaze, that is, our impulse to track the gaze of others; Alice’s glances to the left pull our attention in that direction as well. The sequence from Iron Man is closer to this antecedent than the sequence from Ran. In the former case, general knowledge of war zones and particular knowledge about attacks on military convoys in Afghanistan puts us on alert. The relaxed social atmosphere in the Humvee – Stark’s cocktail, AC/DC hammering away, the badinage and business around the taking of the photo – are designed to distract us from the possibility of an attack, however; our visual attention in particular is (mis-)directed to the right of the frame (where the photographer lines up the camera) and away from the centre of the frame (where the Humvee in front will explode) (figure 3.1d). In the Ran sequence, the sense of threat is somewhat dampened and flattened by the suspended diegetic sound, and the fatal bullet is less visibly intrusive than the bus (in Cat People) or the missile (in Iron Man); even so, the basic constituents (character, implied threat, and sudden intrusion) are in place here as well.

Endnotes

  1. Baird, ‘The Startle Effect’, p. 15.

  2. See in particular Tallis, The Kingdom.

  3. In describing Ran as a ‘free’ adaptation, I mean to indicate that the film borrows many elements from King Lear and seems designed to bring Shakespeare’s play to mind, but it does not seem intended as a ‘faithful’ adaptation, in which overall fidelity to the play is a primary goal. See Paisley Livingston, ‘On the Appreciation of Cinematic Adaptations’, Projections 4, no. 2 (Winter 2010): pp. 104–127, doi:10.3167/proj.2010.040207. Livingston notes the use of the phrase ‘free adaptation’ in the credits for Fellini Satyricon (Federico Fellini, 1969).

  4. Robert Baird, ‘Startle and the Film Threat Scene’, http://www.imagesjournal.com/issue03/features/startle1.htm.

  5. Peter J. Lang, Margaret M. Bradley, and Bruce N. Cuthbert, ‘Emotion, Attention, and the Startle Reflex’, Psychological Review 97, no. 3 (July 1990): p. 377, doi:10.1037/0033-295X.97.3.377; quoted in R. Baird, ‘The Startle Effect’, p. 20.

About Film, Art and The Third Culture 
More Sequences