Initial notes (1996) on Deleuze, G. ( 1992) Cinema 1: The Movement Image, London: The Athlone Press.

NB I took these notes in 1996 ( it is now 2011) , before I had read anything much else on Deleuze. Some of it makes more sense now!  One problem is that I am no longer sure about the origin of the asides in brackets. I have added doubtful ones in square brackets, not to claim credit but so as not to malign Deleuze.  One day I will go back and check. The actual works are spattered with examples from films of all kinds. I have not seen that many of them,although they are actually much more accessible now online or via LoveFilm or equivalent. I have largely omitted these examples, but indicated where in the text they can be found for anyone interested.
There is a discussion on the book and the movement image between Deleuze and two blokes from Cahiers which is quite useful -- here

NB I have collected together all the bits on Bergson from the 2 volumes here

Cinema raises the old problems and paradoxes of movement and how to analyse it—as a series of discrete ‘immobile sections’ together with some abstract quality of ‘movement’?  A series of transitions between essential ‘poses’ (common in art) or as ‘any – instant – whatevers’ (common in analytical sciences)?  Bergson’s philosophy can lead into clarifying the dilemmas.  Here, it will lead in turn to attempts to pin down the specific effects of cinema as a combination of action plus movement— ‘movement – image’ [image in the Bergsonian sense of concept, I assume, something that is both objective and subjective, material and a representation ?]. 

For Bergson, movement is a qualitative change in the whole, that is the relation between a moving body and its environment—a hungry dog wants to eat something and this changes the whole, the dog and the food.  Movement is the dynamic, the ‘vibration’ which changes elements (9).  The whole is never given but always open—durée describes the way in which human beings relate to that Open.  The Open can be closed in various ways by classifying things as belonging to sets, or by pursuing abstractions.  Nevertheless, movement cannot be understood as a closed relation between immobile sections [I think the argument is here that even apparently immobile sets still relate to some mobile section].  Images are best understood as an instant, when they are immobile, and as movement-images when they are ‘mobile sections of duration’.  There are also time-images—‘duration images, change images, relation images…  Images which are beyond movement itself’ (11).

Sets of elements are framed as in a cinematic image.  They can be saturated with elements, or ‘rarefied’ featuring a single element.  There is a correspondence with the depth focus shot and the close up.  They may be explicit and pedagogical, as in Godard, geometric (formalist, featuring lines and angles), or dynamic (changing irises and other variables).  Sometimes they shape the frame within overall frames, or feature a vocabulary of gradings and distinctions, of light, for example.  The cinema also frames  the ‘dividual’ (14) [a possible reference to the essay on social order which emphasises the replacement of real individuality with a more commercial and structural version, but also see the Glossary at the end]: some standard of measurement is applied to disparate objects , such as faces or landscapes, which produces deterritorialization.  Point of view stances also feature pragmatic rules, [such as continuity editing].  There is a clear implication of a context beyond the frame too, and this can allude to the notion of the Open, or the whole again, although this is not the intended effect of continuity editing which merely hopes to synthesise different frames.  There can also be a radical allusion to the whole, sometimes suggested by a single image.

Scenes and shots are cut in two ways as well.  There can be a succession of sections with an allusion to the whole duree, combining immobile and mobile sections.  [Camera] movement expresses the changes in relations in the wholes [for example Hitchcock’s moving camera crossing the road and going up and down the stairs in Frenzy].  In this way, cinema depicts duration—the shot ‘acts like a consciousness’ (20), but consciousness invested in the camera rather than in the spectator or the hero.  Cinema constantly reconstitutes the relations between wholes and elements, unlike any participant or viewer.  This is how the films express ideas, in movements and relations, for example in the combination of constrained, lattice- like shots and circular sweeps in the The Third Man, which is akin to Kafka’s literature (21).  Kurosawa has camera movements like brush strokes in Japanese characters, and various other spatial metaphors.  In these examples, the shot is the movement image, the mobile section.  Shots are more continuous and varied than natural perception, and can act as ‘the general equivalent’ of all the means of locomotion, offering some abstracted essential movement, intended to emancipate [from naturalistic perception?] (23).

The fundamental openness of the whole is suggested by cinematic flexible recompositions, which express ‘time itself as perspective or relief’ (24), rather than aiming at stasis as does photography.  This allusion to movement arises from sequencing scenes, as in montage, and from the mobility of the camera [Deleuze says that the early or contemporary commercial cinema offers mere sequences of the mobile sections again, with limited movements].  There can be an allusion to a depth as well [including political and theoretical depth].  There are in fact many detailed different possibilities [listed pages 26-27].  Cinema can never actually represent the Open, of course, and there are always discontinuities and ruptures [and conventional requirements to tell simple stories and impose naturalistic relations between elements]. The Open is merely testified to, for example by deliberately false continuities [montage as in Eisenstein?]; these show the arbitrary nature of normal continuity, and invoke [the context], the ‘out of field’.

Shots combine to give montages, sometimes these can be assembled within shots.  There are different relations between parts and wholes, as a general way of expressing ideas.  Options include Griffiths on the organic unity of the USA as the functional adjustment of different strands, personalised in the form of duels and so on, through to Eisenstein’s Marxist critiques, offering new notions of the whole in the form of dialectical relations between parts and wholes.  The various laws of the dialectic influenced Eisenstein, for example the shift from quantity to quality.  There also notions of synthesis and transcendence in the ‘pathetic’ shots—the connections of objects at different levels of the dialectic spiral, for example the way individuals are connected to great events, the way consciousness dawns.  [There is a very interesting discussion of Vertov and Dozhenko, with lots of detail].

French cinema took a different option, away from Griffiths’ notion into social science, offering mechanical composition, the notion of a whole as a machine (including Renoir).  Individuals acted as units, whereas the wholes were depicted as ‘passion’, a kind of zeitgeist.  Mechanical analogies included industrial machines or rivers, or the sea.  The relations between the units followed mechanical rhythms, an algebra to generate the maximum amounts of movement [part of a drive to establish what was distinctive about moving pictures—‘ photogeny’].  Alternation rather than dialectic was offered.  There are links to the Kantian notion of the sublime here, through the connections between mundane local movements and movements in nature or the universe as a whole (46).  The whole is not just the sum of the parts.  Instead, French cinema aimed at ‘simultaneousism’, seen in developments such as triple screens, polyvision, superimposition.  It was impossible to quantify the effects of these techniques, leading to a feeling of measurelessness and immensity.

German expressionism used light rather than movement to get the same effect.  Light and dark are seen as primitive givens.  Instead of the dialectic, we get stripes, contrasts or blends.  These allude to the non organic side of life which absorbs life and death, oppositions between the vital and the organic ‘as in elan vital’ (51) [as in Bergson again?].  This produced a characteristic horror of puppets, robots, golems and so on.  [Clever stuff on how the transitions from the organic to the non-organic were coded by the relations of light and dark, 52-3].  Increasing intensity produces the formless (another variant of Kant, the dynamic sublime), via a montage of contrast.

More philosophical problems emerged, for example the relations between images and movements.  Was one ideal and the other material?  However, movements produce images and images engender movements, at least in human beings.  The cinematic image also offers possibilities here.  For example it organises gestalts as forms of intentional consciousness, showing how the world becomes image. 

Classical phenomenology privileges natural perception rather than cinematic forms, and again Bergson gets closer: he sees cinematic perception as one possible way to focus and centre perceptions from a flux.  Each set equals an image.  These include movement for Bergson, necessarily, as a kind of implied context and set of relations.  The images are the primitive category, and there is an infinite set of them.  The image is identical with movement (59) and the movement – image is identical with matter, So Bergson is a materialist.  However, reality is not a mechanical system which is closed.  The fixed parts are best seen as a ‘machine assemblage of movement – images’, hence ‘the universe [is a] cinema in itself, a metacinema’ (59).  Bergson is the theorist of movement, and solid bodies are seen as merely temporary [reifications], ‘movements on a plane of immanence’.  Matter is formed from the identity of the image and the movement (59).  The mechanistic universe is only a subset, characterised by its discrete ‘block of space – time’ one of an infinite number of ‘presentations’ of the plane (59).The image alludes to becoming, to ‘everything that they have not yet become’ (60). 

Potential sites for action are immanent in the sense that light is [some constituting medium which diffuses]. There are connections here with the theory of relativity as a deliberate project in Bergson.  Light passes on unopposed, and only becomes an image when it is obstructed, just as in photography.  Matter is a fixed form of light, leading to an interesting point about objects not being revealed until they reflect light.  In a similar way, images are not recognized until they are perceived, but they do exist independently of the perceiving subject none the less. 

As with light, so with human consciousness.  Unlike the old model which sees the consciousness is shedding light, as a torch beam does, which classical phenomenology shares, although in a modified form, ‘For Bergson…  Things are luminous by themselves without anything illuminating them’ (60).  Living beings are unique images, however—they focus and organise, they frame and organise into sequences (action), organise the flux of variation, operate in the gap or interval between reception and execution.  Hence the organisation and roles of images is at the centre of this process, and many are omitted according to needs.  Human beings have different facets for receiving and transmitting images.  The reception facets close and isolate images as in perception or framing (62) [called ‘condensation’ in Bergsonism?] .  At the other end of the interval, transmission produces unpredictable open consequences, unintegrated with reception. [Really unpredictable or emergent as a complex effect of duration?]  This unpredictability produces new features of human action, indeterminism around living images.  Thus, located on the plane, images react constantly and infinitely among themselves; around living matter they are organised or closed and become centres of variation.  There is an evolutionary process here—as matter solidifies into particles, so images are subject to more and more or elaborate perceptions [this makes a bit more sense after reading Deleuze on Bergson and duration].

Perception is therefore ‘subtractive’, producing things ‘minus that which does not interest us’.  In this way, ‘an atom perceives infinitely more than we do’ (64) [this is what intentionality does?].  The cinema can deliver both sets of perceptions, from the ‘total objective perception which is indistinguishable from the thing’ to a ‘subjective perception’ (64).  The latter is a perception–image, a subset of the movement-image.

Subtraction is not the only process.  The rest of the world is recognized as a horizon, a potential for action again.  Perception is inseparable from action (64).  The distance between the core and periphery is a way of thinking about the gap between action and reaction.  There is a connection between perception–image and action–image, both of which are subsets or ‘avatars’ of the movement-image (65), an organisation (‘incurving’) of the universe rather than a simple subtraction or framing. There is another possible avatar too—affect, which ‘fills the gap [between] troubling perception [and] hesitant action’ (65).  This refers to subjectivity from the inside [normal human perspective] and to the notion of quality [Bergson’s separation of multiplicities with differences of kind and those with difference of degree?].  Affect supplies a motor tendency. It is found combined in human beings.  [So the four types -- movement -image and its three avatars—are illustrated with examples from cinema 69-70, and there is particular attention to Beckett’s work Film, page 66 F. There are implications for time- images too.]

[You can watch all 17 minutes of Film here and see if you read it as Deleuze does. It is a convenient example to match with the extraordinary levels of interpretation Deleuze offers. Deleuze says Film is:

an astonishing attempt...[to] rid ourselves of ourselves, and demolish ourselves...[like Berkley] to be is to be perceived, declares Beckett...[the bit where the character scurries along hugging a wall, filmed only from the back, is]...a perception of action, an action-image...[The bit where the character enters the room and goes round closing windows, covering mirrors and evicting pets, filmed largely from the back with the camera not intruding past 45 degree angles, is]...the perception of perception or the a double system of reference...[Then]...  the finally seen [by the camera] from the front ...[and] ...the last convention is revealed: the the double of [the character], the same face, a patch over one eye (monocular vision) with the single difference the [character] now has an anguished expression and [the camera] has an attentive expression...We are in the domain of the perception of affection: the most terrifying, that which still survives when all the others have been destroyed: it is the perception of self by self the affection-image...

So – OK this is pretty understandable, but the next bit is much more

The end [fade-out on the character rocking in a chair and a close-up eye] suggests – death immobility, blackness. But, for Beckett, immobility, death, the loss of personal movement and of vertical stature...are only a subjective finality...only a means in relation to more profound end. It is a question of attaining once more the world before man...the position where movement was...under the regime of universal variation, and where light, always propagating itself, had no need to be revealed...[Proceeding to ] the extinction of action-images, perception-images and affection images, Beckett ascends once more to the luminous plane of immanence, the plane of matter and its cosmic eddying of movement -images. He traces the three varieties of image back to the mother movement-image...Beckett’s originality is to be content to elaborate a symbolic system of simple conventions [how much the camera can reveal of the character] – according to which the three images are successively extinguished – as the condition which makes possible this general tendency of experimental cinema (66—68). 

There are two references in this section to Beckett’s own commentary on the film –in French. They offer Beckett’s own analysis of the film – according to Deleuze, Beckett says there are three ‘moments...the street, the staircase the room’. But Deleuze says his schema of action-image ‘groups the street and the staircase, while the perception image covers the room and the affection image the ‘hidden room and the dozing of the character in the rocking chair’ (n 31, 272). A couple of issues seem relevant : in the first place, the writer is highlighted as the author of the film, not the director – for the first and only time as far as I can see, since in every other example it is the director, ignoring altogether the writer, not to mention the actors, the lighting camerapersons and the technical crew. [My colleague Ian Gilhespy reminded me of this--There is a performance of Krapp's Last Tape directed by Beckett here].  In Film, was it not likely to be Keaton who added the clowning bits [blimey – I sounded like Zizek there]?  While we are here, the concept of genre is not critically discussed either throughout and is taken to be an artistic term, never a commercial one.

In Note 32, we are told Beckett has not done enough in his commentary  ‘to represent the set of all the movements’ (227). It seems it was Fanny Deleuze who completed the picture with a rather baffling diagram (see below -- I can't see a point B can you?).  O is the character, OE is the camera. No doubt this makes a nice neat diagram -- but is this the only reasons for theorising like this – some notion of completion, mastery or explanatory power? The symmetry of the diagram somehow guarantees the validity of the classification? This reminds me of the obsessive listing and classifying in Anti-Oedipus or Thousand Plateaus

 diagram on p 228

Of course these are also centred readings, with the philosophical meanings given pride of place. I think the film can be read equally well as being about identity and the ‘social mirrors’ that support it – the views of other people, photos, actual mirrors etc. The film adds the perceptions of pets as mirrors. But it could equally be seen as about stardom and anonymity, no doubt.

Peirce’s work on signs is also useful, if we see the sign is a particular type of image, one which represents (from the point of view of composition, generation or even extinction).  There is work to be done on the relations between Peirce’s classifications and Deleuze’s types of image [see Glossary].

Films usually have montages or assemblages of different images, and often one is usually dominant [in experimental film the perception image tends to be dominant].  Shots correspond too—the long shot indicates perception, the midshot action, the close up affect.  The shots can help us offer a whole readings of films (70).

The perception-image can be subjective or objective, the former belonging to a participant and the latter external. These can shift. The subjective image can become a collective one as the camera moves among the characters, taking on an anonymous generalised viewpoint rather than a single pov. This is definitely not naturalistic, and corresponds to the difference between direct and nondirect speech. Pasolini especially used this linguistic analogy, and his camera  develops a ‘free indirect discourse’ (72). [There is an aside on Bakhtin on this too, pointing out the odd tendency to be a subject in speech capable of referring to oneself as an object – I think (73). The process parallels the use of dialects in an utterance] This shows the duality of the (ordinary empirical) subject and a (necessary) Transcendental Subject (73).  However, in such discourses, there is never a complete formal split but rather ‘an oscillation of the person between 2 points of view of himself’ (74) [A necessary, ironic reflexivity?]. The camera itself often does this oscillation – it observes and comments on actors as they act, rather than obediently doing subjective and objective perceptions. An example is Pasolini’s ‘insistent’ or even ‘obsessive’ framing (74) (where the camera frames the scene before and after the actor is in it). Cameras can also use different lenses on the same image, including ‘excessive use of the zoom’ (74). This shows the ‘cinema of poetry’ [NB I noted to myself that all this is derived from Deleuze’s critical perception and it might contrast strongly with what the directors themselves believed they were doing]. The effect can be to produce some sort of reflexivity among the actors too as they watch themselves acting, and as the directors become neurotic (75)

Bergson considered the subjective as showing where images vary according to one central and privileged image rather than constant variation. Subjective variations can produce the objective after [a sort of relativist] moment, where  lots of movements of points of view allude to a notion of the Absolute or the Sublime. This is expressed in French cinema’s obsession with flowing water, as an embodiment of variation and flow [lots of examples follow pp78-9].

Vertov developed a notion of gaseous perception. He clearly aimed at demonstrating the movement-image, showing the possibilities of universal variation in a whole cinema of interaction and variation rather than a standard materialist perception. The human eye was seen as immobile, and thus inevitably subjective (the centre of variation). By contrast, the camera showed movement through montage, which objectified perception, to overcome ‘boundaries and distances’ (81 –quoting Vertov). It does this by adopting the perspective of matter [images are materialist], for example by establishing intervals and gaps that prefigure the role of the human subject. The gaps between two correlated images replace consecutive images, and when these are collected, an assemblage of matter ap[pears, a deciphering of reality. Vertov extends from suggesting images joined by a subject to variation as such [the example here is, obviously, Man With a Movie Camera]. There are even montages of elements of an image. Vertov strayed into formalist dialectics of matter rather than the humanism of Eisenstein, or the ‘spiritual ’ aims of French cinema’s depiction of natural objects.

American cinema was influenced by Vertov too, aiming to discover molecules of matter or ‘photogrammes’. Their programme corresponds with Castenada as an attempt to stop the world, see the gaps between objects etc [D really likes Castenada!]. A US experimental film is described as an illustration (86).

The affection-image. The face in close-up shows this best. A face is any image (even a clock face) which shows a ‘reflecting surface and intensive micro-movements’, which indicate normally hidden aspects [a better start than the pseudy delirium of the commentary on faces and faciality in Thousand Plateaus].The passive immobile surface signifies a surface while the micromovements show the hidden effects of desire, hence different kinds of close up [LCU and VLCU?]. [Very large close ups] offer a series of images that produces a qualitative leap, such a movement from one ‘pole’ to the other. Griffiths and Eisenstein illustrate the possibilities (91f). Eisenstein aimed to show the progression from individual to dividual ( ‘an immensely collective reflection...the unity of power and quality’ (92)).

It is possible to use light and dark on a face to signify a relation to matter generally [amazing examples 92—3]. In a close up of the face, images escape from boundaries and immediate contexts, producing ‘pure affect’ ,located in ‘any-space-whatever’(97). The same effect can be generated by CUs of other parts of the body or of things. Both the object and what it expresses can be seen as related, as in the icon [Peirce?], ‘the set of the expressed and its expression’ (97). When actualised, the qualities of affection-images become the ‘quale’ of the object, the actual emotions or impulses [associated with it] (97), and thus its action-images.

This is easier to see [!] in concrete or relational terms, rather than via Peirce’s insistence on ‘first’ qualities: ‘qualities or powers considered for themselves without reference to anything else, independently of ...their actualisation’ (98). Better to describe these as affection-images (not just action or image), preserving the idea of potentiality. These can only be expressed by ‘a face, a face-equivalent or a proposition’(99), especially in CU. In CU, faces lose their normal function, to individuate, socialise [have a social role] or relate/communicate. Instead, faces in CU offer allusions to nihilism, the void, Fear, as the face expresses nothingness [fucking weird stuff! A fancy example of the uncanny in Freud?].

Power qualities are not produced by real events since they pre-date them. They are both expressed for themselves and then realized or embodied [See DeLanda on how inorganic elements preform complex molecules without human intervention etc?]. Faces in CU also show pure (‘virtual’) relations, especially how objects and emotions are related. These relations are shown in montage (including internal montages as different bits of the same CU are shown). The ways in which faces turn away from and towards [objects] expresses relations – singularities and relations form complex unities like the dividual again – ‘that which neither increases or decreases without changing qualitatively’ (105) [seems completely opposite to the earlier definitions]. We can see how faces are linked to propositions. These are virtual relations of possibility and potentials rather than real(ized) relations, but demonstrating the latter can allude to the former (the example is the Besson film Joan of Arc which shows the realized trial but alludes to the expressed passion of Joan) [Pretty simple point after all the fuss?]. Further examples (107) lead to the  definition of a CU as a shot framed so as to eliminate depth or perspective, or to include a fragment of a field, to allude to a virtual relation.

It is possible to convey ‘any-space-whatever’ in other ways too. The Besson film shows fragmentations of space, deframings, collision-type links. The any-space-whatever is demonstrated as a concrete singular space rather than an abstraction. It has ‘lost its homogeneity’, it alludes to an infinite number of possible linkages, to potentiality (109). So the a-s-w is another way to express the affection-image, as well as a face [make your bleeding mind up, and/or stop making this up as you go along!]. The a-s-w would be a ‘qualisign’ for Peirce. Rapid montage can allude to pure power/quality  [ I am not at all sure I know what this means -- another argument with Peirce? A repetition of the earlier point?].

The a-s-w is constituted [in Besson? Always?] by the play of light and shadows, especially shadows. These represent the virtual, the infinite. Expressionism offers a ‘lyrical abstraction’ of the alternatives (112-3) according to whether light and dark alternate or are seen as opposites. There are possible ‘spiritual’ parallels, producing the classic different types of ethical choices, including no choice; moral necessity; physical necessity; psychological necessity (from desire) [Expressionist cinematic equivalents of techniques of neutralisation?]. Alternatives also appear according to whether one is aware of the choices or not – one can choose to choose or live unaware of choice. These are also illustrated in lyrical abstractions where the theme is avoiding false choices, choices that lead to no choice (Faust), or a choice to renew choice even if it ends in self sacrifice. Themes can develop into a denial that there is even a self rather than just actualisations of possible selves, a ‘moralism opposed to morality’ and ‘faith opposed to religion’ (116).

These are further examples of the relationship between philosophy and the cinema – the choice to choose is beyond specific choices; the light constitutes the whites, blacks and greys; fragments suggest the whole; spaces suggest any-space-whatever.

Whole new dimensions are added with colour. [And there is an aside about musical comedy ‘extracting an unlimited virtual world from a conventional state of things’ 118]. It is not that colour codes affect – colour is affect, it absorbs objects into relations (examples discuss Antonioni and Bergman). The asw is now a matter of absorption, an empty space. [An aside notes that post-War cities are also full of empty spaces, potentials, asw]. Experimental films offer examples (122) [one good example is provided – a slow zoom to explore a room].

The action-image.  This is an intermediate form between action and affect, between the ASW/affect pair and the ‘detached milieux/mode of behaviour’ pair.  It also belongs to the ‘originary world/elementary impulse’ pair.  The originary world can be seen as a pure set, the source of origin of actual milieu.  Impulses are seen as the initial energy required to seize fragments.  This provides a naturalist rather than realist notion of actualised forms.  [All references to Bergson’s philosophy, I assume].  Primordial origins can be alluded to as well (125) faith, understood as being immanent to the originary.  They can be represented by symptoms (‘the presence of impulses in the derived world’), and idols and fetishes (‘the representation of the fragments’) (125).  These aspects are illustrated in the cinema of Stroheim and Bunuel (and the examples here include the drawing room in Exterminating Angel, the desert of columns in Stylites).  These are films about the originary world which ‘carries the milieu along’, and represents both origin and end (as in the escape of the bourgeoisie from the drawing room, only to end in a cathedral).  The effects of the originary world on real milieu often appear as degradation or cyclical return, although both tend to represent the negative effects of time.

Impulses are always directed at fragments (including shoes or even ‘invalids’).  Impulses can carry on until they cross boundaries and exhaust other milieu.  They are deeper and more general than affects attached to objects.  They can cross social boundaries (the beggars in Viridiana) and aim at the general goal of fragmenting, gathering up the scraps and ending in the death impulse. Bunuel is  is a naturalist in this sense rather than a structuralist proper.  He alludes to faith rather than religion, however, in his particular way of discussing spirituality and choice which connects into the tradition of lyrical abstractionism discussed above.

[A commentary ensues on a number of directors who represent such a naturalism—133f].  The role of female actresses in American cinema can allude to the originary—‘originary women’ (134), esp. King Vidor in Duel in the Sun [much discussed by Mulvey as an example of a fiery tempestuous woman who disrupts normal male society, but has to be killed as a result --trailer here ].  However, staying at the level of impulse is the problem and there is a tendency to stray into realism, of actualised milieux and behaviour. Losey is discussed (137): his trick is to allude to originary violence as too great to be actualised, even in the characters.  Here too, women offer an escape into something outside the ‘hermetic’ world of men.

Action – image, the large form.  Milieux are made concrete and actualised in action – images proper—‘realism’.  This can still be linked to originary milieu.  Realism can include fictional realism, as in the dream, defined merely by ‘milieux which actualise a mode of behaviour which embodies’ (141).  The action-image is the relation between the two.  Milieux can actualise several qualities and powers which interact on a character, producing a response, and a new situation.  The process of individualisation produces a kind of duel, between man and milieu, or an hourglass structure—the broad and narrow dimensions represent the broad and narrow dimensions of the character.  This is the ‘large form—“Situation/Action/Situation”’ (141).  Each stage has its characteristic signs, including ‘synsign’ and  ‘binomial’ [using Pierce again].  The example here is the walkdown/shootout in the western.

Documentaries can be seen as a subtype of the action image. Flaherty gives an example of the duels between communities and hostile milieu, without third terms, for which he has been criticised (like the role of colonising powers and the effects on community).  For example, Nanook [clip here] represents duels between man and environment producing survival and changed situations. 

There is a psycho – social subtype in King Vidor, an ‘ethical’ realist form relating individuals and collectivities.  It is possible for individuals do not always trial for change the situation, and they can even produce worse results.  The form is still realist, sometimes hourglass, rather than the Expressionist notion of entropy discussed earlier, and there is still a focus on concrete milieux and their pathologies, or specific behavioural disorders.  This is the world of the born loser, the heroic drunk or criminal, especially in its film noir variant.  The rise and fall narratives of gangster movies also fit—here the milieu of the underworld is a ‘false community’, a jungle, while the behaviour of the hero also reveals fundamental flaws, which can sometimes be exploited by the minor characters [including the femme fatale?].  The films act as critics of society—or as a compensatory nightmare to strengthen the American dream (145).

In the western, the landscape encompasses via metaphors of ‘breath’ [sic, 146] or later, colour codes, especially the ‘sky and its pulsations’ (146).  The community and the land is mediated by a leader.  The milieu itself can ‘breathe’ especially in Ford movies, and can bring together minorities in a melting pot.  We are close to epics here, but westerns also include elements of tragedy and romance.  Ford offers change as a spiral movement rather than a circle which permits ethical commentaries, for example of the transformation of natural law to written law.  There are also ‘healthy’ dreams about a community, ‘vital illusions, or realist illusions which are more true than pure truth’ (148).

There are variations on the birth and rebirth of a nation in American cinema as well, just as in Soviet cinema, but in an ‘organic’ dialectic [there are British organicist analogies too].  This often takes place via analogy or parallels between American and other civilisations, especially classical civilisations, and develops notions of the growth of vigorous national states which have two elements of the American dream—the melting pot, and the ‘ferment which creates leaders’ (148).  These historical themes are found in all the other American genres too, including gangsters.  This is an ideological form of history of course [with references to Nietzsche on history, 149 f], for example the ‘monumental’ themes (great buildings and events for example) which are linked periods and tend towards the universal.  There are the usual problems in grasping the real movements of history with its tensions, which are often represented and individualised as duels.  ‘Antiquarian’ themes involves the reconstruction of past societies as some kind of social context for the present.  ‘Ethical’ themes depict the battle between good and evil, decadence and vigour in a ‘constant discovery of America’ (151).  These are often all work together to produce ‘a strong and coherent conception of universal history’ (151).

So there are common elements among the genres: (a) organic composition and combinatories in typical action images (sky, landscape etc.)and  ‘encompassers’ [ see Glossary], movement within images; (b) movements which contract into a duel which must be produced by convergent lines from a situation, as in the ‘large form’ like the hourglass [illustrated with reference to M]; (c) duels themselves present moments of simultaneity where parallels converge; (d) duels dovetail and interact with other duels; (e) the large gap between the milieu and behaviour can only be bridged progressively, because the power of the hero needs to be developed, for example by a group, and only after moments of doubt or impotence.  Strength can be transferred from one character to another.

There is an emphasis on behaviour as mobile from one milieu to another.  Behaviour is to be structured by the milieu.  [Then there is an odd bit about the differences between passive vegetable and active animal forms, terms which make a link with Bergson—vegetables ‘accumulate the explosive on the spot, whilst the animal undertakes the demolition’ 156].  War films are an example of these alternating forms of structuring.  Or Baby Doll, [clip]  with its slow accumulation of tensions then a violent release.  Or combinations of little episodes which accumulate [On the Waterfront is the example].  There is a tendency for greater toughness and exaggeration of this process, linked to the collapse of the American dream (157).  These films feature real behaviourism because they see behaviours as caused, although internal motivations are important too.  In fact the focus is often on the internal dimension, the whole point of fictional realism.  This inner focus is what overcomes the artificiality of realism and of acting.  These the elements must appear in the image—each action image displays an ‘emotion/object’ pair, as much as any other pairing of aspect and face, impulse and fetish.  The inner, the ‘impression’ links the situation and the explosive action.

The action image small form. This is an alternative possible form linking action/situation/image (see above) where action ‘discloses’ the situation which triggers off new actions, moving from an habitus to a partially disclosed situation, a local example.  This disclosure can be elliptical rather than spiral, constructed around events rather than organic, and comedic rather than epic.  The signs here act as indexes.  The initial actions are an index for the situations, which implies a lack of knowledge, a ellipsis [for example as in sudden shifts forward in time.  Other examples include highly condensed metonyms and metaphors—a shirt collar falling from a drawer indicates an affair].  This offers a ‘reasoning image’ (161), where the audience has to work things out.  There are also indices of equivocity [for skilled readers only?], where only slight differences appear in actions, but where these lead to two very different interpretations, hence the ellipsis.  These often linked with the need to be economical, as in the French new wave. 

Again there are different genres: (a) the comedy of manners; (b) the costume film as descriptions of an habitus.  Unlike historical epics of the large form, which are about historical developments and so on, these are about habits [with a possible pun to allude to dresses]: dress here acts as an index; (c) Grierson and free cinema instead of Flaherty, featuring concrete modes leading to social situations as sites of struggle, enabling the underdog perspective to appear; (d) the detective film, where actions become indices; (e) westerns as well, especially those produced by Howard Hawks, which replaced the organic encompassers with ‘pure functionalism’, functional groups rather than organic communities, nor to work based but arbitrary, often collections of travellers.  Here local interiors offer unexpected events in contrast to large experience, such as the odd bits where the functions of men and women interchange in Hawks movies; (f) the neowestern, which focuses on micro politics in small groups, such as The Wild Bunch, where the bigger groups fade out—there is racial indifference for example.  There are no grandiose actions, no dream, focus on local westerns rather than ‘the west’, episodes are linked by a broken line or vector rather than an organic form; (g) burlesque which is devoted to this shift, for example in Chaplin [how small differences in behaviour reflect large differences in reality of the situation—a shaking body can indicate sobbing or shaking a cocktail] (169).  Analogy merely shows complete differences between situations.  Chaplin films are vectors joining these episodes.  Another example would be Harold Lloyd on the demonstration of the ‘perception image’—he appears to be driving a large car but this is revealed as him cycling behind it.  Both comedians did gestures which could be comic or tragic/emotional.  Some commentary was also offered, as when Hitler and the little man were seen as similar—the idea being that social forces or discourses produce one rather than the other (172).

The talkies introduce discourses of this kind [commentaries, dialogues?].  These links to the larger forms in Chaplin.  It is different for Keaton, who locates himself in large milieux and borrows scenes from films of those, for example Griffith, producing large gaps between the location and the comic action.  This gap can be bridged in a number of ways: (a) the ‘trajectory gag’, a rapid sequence of actions [jumping gaps, sliding down poles, running along trains and so on]; (b) the ‘machine gag’ where weird ( ‘Dadaist’) machines are produced (175) with definite functions.  This can involve ‘minoring’, where the immensity of machines is reduced to a personal scale, alluding to a political ‘anarchistic machine’ to assert human rights to use the big machines (176).  Machines can also offer ‘recurrent gags’ with absurd causal sequences [as in Heath Robinson]. Minoring links the action to the situation, recurrent gags ‘make the hero equal to the situation’ (177).

Figures.  The figure is a sign of ‘deformations, transformations, or transmutations’ (178).  Scripts can be turned into either large or small forms according to the conceptions of the director (which can turn on a sudden realization of the signifying power of some detail, for example ) [Storm Over Asia is the example –clip here --  or Ordinary Fascismclip here -- where everyday events tell the story].  In the first example, a dialectic of quantity quality shifts tell the story, a guided small form.  Other directors saw the dialectic in terms of parts and wholes which produced a larger form [Dozhenko is the example].  In Eisenstein the issue is the reconciliation of opposites, a transforming form [the discussion turns on how big forms produced concrete embodiments.  In 1996 I read this as rather Hegelian, or buts, presumably it is much more to do with the actualization of virtual potentials in Deleuze’s own philosophy?] (181). In Eisenstein, development of the small level, the concrete or embodied becomes ‘pathetic’ [I do not know if this has a special meaning].  Such transformation is often done via special images, such as the theatrical interludes in Ivan...  [clip here] The links between these images are indirect,  via an image which prefigures, or indexes, depending on whether one is moving from action to situation or vice versa.

This leads to discussion of figures of discourse, such as tropes, where a word in the figurative sense replaces another word—metaphor, metonym, synecdoche.  There are also imperfect tropes such as allegories or personification, all cases where words are substituted for each other in a strictly sense, as in reversals.  There are also a ‘figures of thought which do not pass through any modifications of words (deliberation, concession, support, prosopaeia)’ (183).  In each case, cinema has figures which correspond. [Prosopaeia is ‘a rhetorical device in which a speaker or writer communicates to the audience by speaking as another person or object’, according to the indispensable Wikipedia.]

There are also visions.  The example is Herzog’s cinema, depicting visionaries where actors exceed the requirements of the situation, leading to sublime action (in the large form) and heroic action (which is appropriate). Herzog also demonstrates the enfeeblement and reduction of the small form [with some weird examples 185].

Transformations take place in particular domains: (a) the physio – biological milieu which acts as a transmitting fluid; (b) the mathematical domain, describing space and relationships like the whole and the part, the global and the local: elements of the local make sense by connection to a global (187); (c) the aesthetic domain, such as the landscape [and there is a weird aside about Chinese painting which apparently operates according to principles of either insisting that the elements have a unity in the One, or that elements are separated into distinct autonomous events.  This produces metaphors of breath, circle, spiral for the former, and broken line or wrinkled stroke for the latter—so some earlier allusions are explained?] Kurosawa’s cinema is the example, where the exposition of all the given factors in the situation is followed by explosive action.  This is not the same as the usual sequence from situation to action, though, since the givens are quantitative: an actual situation is depicted with allusions to a notion of ‘any quantity whatever’’ [this seems to lead to some sort of critique of positivism here, with its failure to grasp the quantitative properly by acting too hastily and literally, acting or reacting before a considered approach]. In the Seven Samurai, the question is not the pragmatic one of how to defend a village, but the more general issue of ‘what is a samurai today?’ (191).  [Deleuze’s admits that this looks ‘mundane humanist’ (192)]. 

Another approach [Mizogauchi] displays the opposite tendencies, developing from the small form (192 f).  The small scenes are linked, for example by high angle shots for individual scenes, and contiguous shots ‘which produces a sliding effect’ (193).  There is a persistence of medium shots and circular movements of the camera, and the use of a special kind of ‘rolling shot…  [which] unravels successive fragments of the space to which are…  attached vectors of a different direction’ (194) [the commentary referred to here is in French].  This leads to a lengthening, the connection of fragments and a demonstration of vectors increase, demonstrating the ‘line of the universe’’,  linking together heterogeneous elements.  The lines go through women for this director, although he is aware of the social appearance of women as prostitutes or oppressed.  Reality thus appears as disconnected, disoriented. [God knows what this actually looks like but try this clip from Ugetsu]

The crisis.  Peirce discusses ‘thirdness’, signifiers, their law or relation, as for itself, irreducible to actual examples.  There is a focus on acts rather than actions interpretations or relations, on natural and abstract relations rather and signification and the law, on natural unities and progressions  [the example here is again one of painting and how portraits led to an interest in the face and then to the circumstances of painting art].  There is less interest in abstract connections through geometrical properties.  This approach in necessarily involves mental/intellectual/cognitive aspects.

Affection-images and action-images imply some mental or intellectual element, but the latter can have its own image.  This image can explicitly signify ‘relations, symbolic acts, intellectual feelings’, and a direct relationship with thought.  [A strange discussion ensues on burlesque and the Marx Brothers—Groucho represent symbolic reasoning as comedy as in “either my watch has stopped or this man is dead”].  Hitchcock was always more interested in relations than a mere whodunit, crimes was seen as a part of the wider set of relations, of exchange, say:  this becomes a symbolic act of which the actors are unaware, although the camera explains (for example why the hero of Rear Window has hurt his leg—we see pictures of a racing car).  Even the audience can be seen as the third term, since there reactions become part of the film.

English philosophy developed an interest in relations as a key element of logic.  The relation became the starting ‘postulate’ of the film for Hitchcock [who was therefore interested in English philosophy or somehow influenced by it?].  His films develop possible variations and changes, were action is often determined by ‘experienced conjunctions; because…  although…  sense…  if…  even if’ (202).  To do this, Hitchcock develops specific signs of relationships—not the detective’s account, but various ‘marks’ (where some natural or customary series is being referred to) and ‘demarks’ (where one term leaps out of natural relations, such as birds, or the crop sprayer out of context).  There also symbols—concrete objects bear a relation, where a key can indicate a marital relation, birds become representative of the relations of nature, or objects express nodes of abstract relations [no example given here —maybe the dead stuffed birds in Psycho express the weird relation between Norman and hiss dead mother?]  Hitchcock’s work led to a crisis in traditional images of the cinema, because the emphasis on relations casts into doubt the status of the images of concrete objects themselves—they become mere terms in a relation (205).

This crisis was always there though.  There was always a tendency to deny action as such in favour of showing relations.  The milieux in sequences of situation and action are seen as no longer decisive, but merely one constitutive element of action, part of a multiplicity rather than a decisive fact.  The same goes for decisive actions—these in turn become, turn into improvisation, develop the present for their own sake rather than a part in a narrative.  The crisis has its origin in a number of external and internal (artistic) impulses.  [These look rather like a list of the factors that have produced postmodernism—the War, the end of the American dream, the consciousness of minorities, excessively imagery, and the effects of experimental narratives from literature (206)].  Conventional sequences of situation and action are no longer believable, and new signs emerge [we have here the usual claim that changes detectable by French intellectuals somehow represent real changes grasped by everybody].

In the new cinema:

1.        The image refreshes situations which are ‘dispersive’, with multiple characters and narratives, for example Altman.

2.       The lines of the universe are broken, become elliptical, featured discrete and segmented actions rather than regular transformations, or are replaced by chance—‘white events’— rather than expressions of personal interest [the example here is the actions in Taxi Driver] There is indifference, the interchangeability of the action-image and the affection-image.

3.       There is the dominance of the stroll rather than the journey or the project or initiation.  Aimless movements, repetitions, things that can happen anywhere.

4.       Cliché becomes important as a unifying principle of sets of elements rather than totality and linkage.  Unities are provided by actualities (news, interest items, songs) and the ‘eye of the camera’ [which seems to refer to some internal monologue of a third party].  These become inner psychic clichés.  The inside becomes like the outside [the examples are the clichés in Altman].  The political function of this is to make the outside tolerable, almost as a plot (209).

5.       The condemnation of this conspiracy takes the form of attacks on or demonstrations of the power of the mechanical reproduction of images and sounds.  There is a potential role here for cinema in offering a critical reflection of its own role (210).  Directors become aware of their own activities in selecting images, for example.  However it is common to direct this, at least in American cinema, both into a critique of misappropriation in two abstract repetitions of themes of [Altman again] to mere parody for—American cinema now has its own traditions to dominate it.

Critical trends in Europe include: (a) Italian cinema post-War.  The cinema had escaped fascism both, so there was no need to preserve national honour (as in France), permitting new beginnings to include the excluded. [The example here is Rossellini]. Neo-realism is the precursor of the five characteristics of the sign mentioned above, as in the fragmented scenes of  Rome Open City, and voyaging for the characters in Bicycle Thieves.  Fellini also offers a plethora of asw, but there is a reconstruction in the form of clichés of Italianness.  There is a condemnation of the mafia conspiracy too (212 – 13).(b)The French new wave features the voyage-form for the present in its own right, offering random links,  again lots of asw.  There is also ‘making false’: warping perspectives, slowing down of time and an alteration of gestures (213) as signs of the new realism. There are clumsy fights rather than stylised duels. There is an awareness that these have become clichés too. (c) New German cinema where the characters can choose to become clichés, leading to a general theme of suspicion of a general conspiracy aimed at enslavement.

There are problems with these trends too, including the danger of descent into parody. There are examples of thoughtful reflection on what an image is in Godard, extracted from clichés. There are links to Hitchcock as well as Marx in the French new wave (pursuing common interests in mental images and thirdness), but also an interest in smashing the system of perception-action – effect, a deliberate prolongation of the crisis in order to liberate a new thinking image.

NB There is also a glossary – I don’t know if it really helps though! It shows how Deleuze interprets Peirce (and Bergson) for his commentary. I have scanned it and included it below (217-8):


ACTION-IMAGE: reaction of the centre to the set [ensemble].

AFFECTION-IMAGE: that which occupies the gap between an action and a reaction, that which absorbs an external action and reacts on the inside.

IMAGE CENTRE: gap between a received movement and an executed movement, an action and a reaction (interval).

MOVEMENT IMAGE: the acentred set [ensemble] of variable elements which act and react on each other.

PERCEPTION-IMAGE: set [ensemble] of elements which act on a centre, and which vary in relation to it.


Dicisign: term created by Peirce in order to designate principally the sign of the proposition in general. It is used here in relation to the special case of the ‘free indirect proposition’ (Pasolini). It is a perception in the frame of another perception. This is the status of solid, geometric and physical perception.

Reume: not to be confused with Peirce’s ‘rheme’ (word). lt is the perception of that which crosses the frame or flows out. The liquid status of perception itself.

Gramme (engramme or photogramme): not to be confused with a photo. It is the genetic element of the perception-image, inseparable as such from certain dynamisms (immobilisation, vibration, flickering, sweep, repetition, acceleration, deceleration, etc.). The gaseous state of a molecular perception.

AFFECTION-IMAGE (quality or power):

Icon: used by Peirce in order to designate a sign which refers to its object by internal characteristics (resemblance). Used here in order to designate the affect as expressed by a face, or a facial equivalent.

Qualisign (or potisign): term used by Peirce in order to designate a quality which is a sign. Used here to designate the affect as expressed (or exposed) in an any-space-whatever. An any-space-whatever is sometimes an emptied space, sometimes a space the linking up of whose parts is not immutable or fixed.

Dividual: that which is neither indivisible nor divisible, but is divided (or brought together) by changing qualitatively. This is the state of the entity, that is to say of that which is expressed in an expression.



Symptom: designates the qualities or powers related to an originary world (defined by impulses).

Fetish: fragment torn away, by the impulse, from a real milieu, and corresponding to the originary world.

ACTION-IMAGE (the force or act):

Synsign (or encompasser): corresponds to Peirce’s ‘sinisign’. Set of qualities and powers as actualised in a state of things, thus constituting a real milieu around a centre, a situation in relation to a subject: spiral.

Impression: internal link between situation and action.

Index: used by Peirce in order to designate a sign which refers to its object by a material link. Used here in order to designate the link of an action (or of an effect of action) to a situation which is not given, but merely inferred, or which remains equivocal and reversible. We distinguish in this sense indices of lack and indices of equivocity: the two senses of the French word ellipse (ellipse and ellipsis).

Vector (or line of the universe): broken line which brings together singular points or remarkable moments at the peak of their intensity. Vectorial space is distinguished from encompassing space.


Figure: sign which, instead of referring to its object, reflects another (scenographic or plastic image); or which reflects its own object, but by inverting it (inverted image); or which directly reflects its object (discursive image).

MENTAL IMAGE (relation):

Mark: designates natural relations, that is, the aspect under which images are linked by a habit which takes [fait passer] us from one to the other. The demark designates an image tom from its natural relations.

Symbol: used by Peirce to designate a sign which refers to its object by virtue of a law. Used here in order to designate the support of abstract relations, that is to say of a comparison of terms independently of their natural relations.

Opsign and sonsign: pure optical and sound image which breaks the sensory-motor links,  overwhelms relations and no longer lets itself be expressed in terms of movement, but opens directly on to time.