Mirror

14/04/2015

The Rhythm of Consciousness

What is a cinematic self-portrait? In one scene during Andrei Tarkovsky’s Mirror a young boy flicks through a book and comes across a a self-portrait of Leonardo da Vinci. We have no problem naming it as such, and self-portraiture in art has a long tradition, but what of self-portraiture in film? Some might think of Derek Jarman’s Blue, the personal work of Alain Cavalier, Godard’s JLG versus JLG. But few films have more ambitiously found a form for personal reflection than Mirror, and it is entirely apt that the film goes by such a name.

The film is both astonishingly personal and fascinatingly general. At one moment the camera passes through a room and we see a poster of Tarkovsky’s second film, Andrei Rublev on the wall. The poetry we hear in  voice-over is Tarkovsky’s father, Arseny, reading his own work. Tarkovsky’s wife, Larissa Tarkovskya and his mother make an appearance, and the film is an account of both Arseny and Andrei’s childhood with the same actress, Margarita Terekhova, playing both Arseny and Andrei’s mother. Yet there is also the impersonal history: stock footage from various different time periods, including the Spanish Civil War and WWII. These moments are invoked neither arbitrarily nor concretely. The Spanish Civil war footage comes during a moment when some Spanish friends of the family are around the house; the WWII footage after the father (Alexei) speaks to his son Ignat on the phone and mentions a girl he was fascinated by during those years as we notice the son is also playing the role of the father as a boy. We can notice the same mole above his lip, just as we notice that the mother and wife have three moles on the neck. Obviously these are attributes of the actor and actress; but they’re also distinguishing features that will leave us in no doubt the same person is playing both roles.

Thus we have personal reflection with impersonal archive, yet we also have black and white footage intermingling with the colour image without clear distinctions made between the use of one over that of the other. Many films of course use black and white to signify one thing and colour photography to illustrate something else. The Wizard of OzA Matter of Life and DeathRaging BullPleasantville and Rumblefish allow us to locate the distinct use of each. In The Wizard of Oz Kansas is in monochrome, Oz in colour; in A Matter of Life and Death earth is in colour and the heavenly way station in black and white. Raging Bull uses colour for home movie footage, while Rumblefish utilises colour symbolically evident in the colourful fish of the title. In Pleasantville colour is steadily more evident as the townspeople become more radiant and happy. Tarkovsky himself used colour consistent with these approaches in Andrei Rublev: most of the film is in black and white, but the conclusion shows various Russian icons painted by the artist of the title in colour.

But as in his own Stalker, as in Nagisa Oshima’s The Diary of a Shinjuku Thief and Boy, Lindsay Anderson’s If… colour and black and white are shown indeterminately. We can’t seek a logic in the image but must accept that the associations become vague and troubling. There seems to be a general use of black and white or sepia in the archival footage and in the scenes that suggest a dream; while the scenes in the thirties and forties and the present are in colour. But this logic doesn’t hold: there are dreamlike images in colour; moments in the present in black and white. There is deliberate vagueness too in casting Terekhova as both the wife and the mother. Like Ulysses’ Gaze which has the same actress playing four roles, or That Obscure Object of Desire, where two actresses play the one part, Mirror calls into question the gap between the character and the actor. If we have the choice between colour and monochrome utilised in such a way that the image becomes indeterminate by the use of both without a clear rationale, then equally film has the opportunity to distinguish between the role and the actor. In most fiction films the purpose is to close the gap between the two to create a seamless impression of a subject, of an indivisible self in the role, no matter how divisible they actually happen to be. Yet by casting the same actress in two parts, Tarkovsky creates a crisis in the image equal to his play with black and white and colour. In each instance the crisis happens to be epistemological: what can we safely say about the film we are watching? How can we be sure that at certain moments we are watching the wife and not watching the mother, and vice versa, just as we cannot be sure what the shift from black and white to colour means. Certainly we can make certain assumptions about where we are in time, but Tarkovsky doesn’t make things easy for us to do so partly by casting the same person in two roles.

This is where a couple of epistemological options make themselves available to us. One is to assume that the film doesn’t ask for sense but just washes over the viewer, so that we let our minds go and respond affectively to these images that we can’t quite shape into narrative sense. The other is to insist that they must add up to a coherent whole and we thus shape them into a determinate meaning that allows us to eradicate the crisis the image generates. But surely for Tarkovsky and other masters of the period (like Herzog, Pasolini, Wenders and Oshima), the very crisis is what interests them. Tarkovsky insisted that “as I began working on Mirror I found myself reflecting more and more that if you are serious about your work, then a film is…an action which will affect the whole of your life. For I had made up my mind that in this film, for the first time, I would use the means of cinema to talk of all that was most precious to me…” (Sculpting in Time). Talking about literature and free indirect discourse, but at the same time understanding its possible use in cinema, Pasolini said “…entire novels are no more than entire Free Indirects in that either there is a total identification of the author with a character, or the characters are a pseudo-objectification of the author, or the characters are devices for expressing the thesis of the author in a substantially unified language, or finally – unconsciously – the characters perfectly inhabit in the same way the social and ideological world of the author…” (Heretical Empiricism) Both Tarkovsky and Pasolini are talking about crisis however as an indirect expression, evident when Tarkovsky adds “Mirror was not an attempt to talk about myself, not at all. It was about my feelings towards people dear to me; about my relationship with them; my perpetual pity for them and my own inadequacy – my feelings of duty left unfulfilled.” (Sculpting in Time)

This suggests that rather than assuming the film should wash over us (a version of the affective fallacy where our feelings are what is most important in the art work), or believing that the film has a clear meaning if we only look hard enough to find it (the intentional fallacy where we credit the importance to the rigour of the artist), we instead accept a tentative and speculative relationship that invokes our own responses into what we might assume are the director’s intuitions. We neither assume a grand plan or settle for its absence, but feel instead unsettled by the fact it is neither one thing nor the other, and try and find a position between these two antinomies. Serge Daney, reviewing Stalker, said “A film can be interpreted. This one in particular lends itself to it (even if in the end it hides its secrets.) But we are not obliged to interpret it. A film can be watched too. One can watch for the appearance of things which one has never seen before in a film. The watcher-viewer sees things which the interpreter-viewer can no longer make out. The watcher stays at the surface because he doesn’t believe in depth.” (Liberation)

To be a watcher-viewer however means making sense of our senses. It doesn’t mean surrendering to them, and it doesn’t mean denying them for symbolic interpretation. It means remaining within the troubling realm. If we think about how the shots are constructed intuitively, if we feel the film is resisting our faculties of expectation, we are remaining on the surface of the film, but we are not refusing its Tarkovskian originality: the fact that it wants to be a work of cinema within the world of singular sensibility.

To help us understand the Tarkovskian we will have more to say about the freedom with which Tarkovsky creates his images. But it would also be useful to say a few things first about the Soviet montage cinema against which Tarkovsky was rebelling. We might think especially here of Lev Kuleshov’s experiments and both creative geography and the manipulation of emotions through film form. In the former example Kuleshov would show someone walking along the street and then the filmmaker would cut to the Eiffel Tower, Red Square or Capitol Hill and the viewer would make the assumption that this  is where they are going. Again this indicated the importance of the cut. The actor could be walking along any street, perhaps many miles from the place in the next shot, but the film language makes the viewer assume that they are closely linked and this is where the character intends to go. In the latter, an actor would retain exactly the same expression on his face, and the director would cut to an attractive woman lying on a settee, to a bowl of soup, to a young girl in a coffin, and the viewer would read different expressions on the face each time. This showed that it wasn’t acting in cinema that necessarily counted, but the juxtaposition of shots. This was the Kuleshov effect: “The Kuleshov effect is an editing principle in which an interpretive effect is created by a sequence of shots that by themselves are not explanatory. The imagined connection between the shots establishes the meaning of both shots.” (Slavische Studies)

Though there were many points of disagreement between Kuleshov, Eisenstein and Pudovkin, they all agreed in the importance of montage. Tarkovsky saw this as a false priority, and thus instead talked up the rhythm of the shot: the integrity of the particular sequence and the integrity of performance and location. Yet this wasn’t a straightforward approach to what was in front of the camera. His films weren’t interested especially in creating a copy of the world, but instead in creating out of the camera a vision that would be integral and not overly edited, or rather, if edited, edited with an awareness of the time that each shot contains. “Editing exists in every art form, since material always has to be selected and joined. What is different about cinema editing is that it brings together time, imprinted in the segments of film.” (Sculpting in Time)

This is a variation perhaps of Pasolini’s notion of the written language of reality. If literature could be reduced to recognizable smallest units, phonemes or letters, then the shot had no such reducibility. What was the smallest unit in film? If we say it is a shot, then we have to acknowledge that most shots are made of many items within that unit. If the smallest unit in literature could be a or e, the smallest in film would be closer to: tree, earth, sun, cloud. It would be made up of a series of things, not merely the equivalent of one letter. Here is where Pasolini asks what cinema does. “What does the grammar of the language of cinema fish from reality? It fishes the smallest units, the units of the second articulation: the objects, the forms, the acts of reality which we have called “kinemes”. After having fished them, it keeps them in itself, encapsulating them in its units of first articulation, the monemes – that is the shots.”

One of the key differences between film and literature is that the word is without time, it belongs not to the temporal order of the world, but a symbolic order on the page. It is removed from reality, while the film image is contained by reality. The is why Pasolini can talk of film as the written language of reality and why Tarkovsky can insist that “editing entails assembling smaller and larger pieces, each of which carries a different time”, adding “their assembly creates a new awareness of the existence of that time.” What Eisenstein and other montage filmmakers lacked was what Tarkovsky called “time-truth”: that, in many films, by playing up the editing whilst playing down the temporal it is as though “in terms of the editing they are quite indistinguishable.” “On the other hand “you will always recognise the editing of Bergman, Bresson, Kurosawa or Antonioni; none of them could ever be confused with anyone else, because each one’s perception of time, as expressed in the rhythm of their films, is always the same.” (Sculpting in Time)

Taking into account both Pasolini and Tarkovsky’s remarks, what we see is a need to acknowledge the language of reality in the accumulation of the shots. This is of course what Kuleshov and others claimed was unnecessary. What mattered in Kuleshov’s experiments, for example, was not the language of reality but the language of film, and this is why the viewer would automatically assume that if someone walked along the street and you cut to the Eiffel tower, then this would be where they are going. This is why if you cut from a bowl of soup to the man looking on, we will assume he is hungry. It is as if the montage filmmakers, and indeed many a Hollywood filmmaker thereafter, wanted the film’s grammar to carry the film’s meaning more than the language of reality that allows for its existence.

However this doesn’t leave Pasolini and Tarkovsky as naive realists; if anything it is their respect for the written language of reality that leads to the complexity of their work. Thus if Mirror remains such an elusive film, it resides in the cut losing many of its grammatical coordinates whilst still holding on to elements of its temporal milieu. When for example the film shows us a burning dacha, the grammar indicates a sequence in one temporal zone, but the features of the sequence spatially makes it difficult to organize the scene perceptually with confidence. As we see the two children (Alexei and his sister) looking out by the front door in reflection, a boy comes from the rear of the house and passes by the mirror that is showing the kids reflected. The two children seem to be watching the very burning dacha that we will see the boy looking at, but the play with space here is slightly baffling. As the boy looks out from the house, he sees in front of him an old man nearer to the fire, and a woman looking on. The film then cuts to what we might at first assume is a closer shot of the woman, but instead to the Alexei’s mother, Maria, coming towards the burning building, as the man tries to put it out. The grammar of the sequence indicates that they are part of the same temporal moment, but the logic of the shots doesn’t easily match our grammatical expectations.

Tarkovsky refuses the usual assumptions. He could have shown the children looking out, cut back to a reaction shot from the children’s faces, and then shown the mother in front of them going towards the fire. Instead it is as though the camera must float through and round the characters, leaving the story and the figures within it secondary to a narrational reality based on memory and reflection. Of course numerous other film makers were refusing this focus on story and character, but just because other filmmakers were doing likewise (Antonioni, Angelopoulos, Garrel), this shouldn’t quite pass for an explanation, otherwise the very remarks about Bergman et al would become redundant. We would notice that while Tarkovsky wouldn’t be editing like a Hollywood filmmaker, he would be editing like an arthouse one. It would also indicate a weak dramaturgy next to the strong dramatic reasoning of Hollywood, where the filmmaker could easily justify their editing decisions on the basis of dramatic logic. The arthouse filmmaker would be positioned as merely reactive. We need to find a logic in Tarkovsky’s images equal to those in more commercial cinema, and for the added purpose of suggesting a singular response: the Tarkovskian invoked by Daney. Yet this isn’t the same as crediting Tarkovsky with a given meaning either. It is closer to a resistant (quite different from reactive) aesthetics; one that tries to find a meaning not out of assertive assumption (as we find in many a mainstream work), but out of tentative exploration: “the true cinema image is built upon the destruction of genre, upon conflict with it.” (Sculpting in Time)

How is this achieved? Let us think of a couple of scenes in the film that play up anxiety. There is the pre-credit sequence where Ignat turns on the television, and the film then cuts to a young man who has a stutter corrected by a speech therapist. As the opening shot shows Ignat looking at the TV before the film cuts to show the scene between the stutterer and the coach, there is no tracking shot into the television to suggest the boy is watching a particular programme, and the TV we see has a blank blueish screen. The film then cuts to this black and white sequence showing the therapist working with the young man. These links are what we might call cinegrammatically weak: shouldn’t the TV show Ignat watches be in colour, and shouldn’t we have shots back and forth between the programme and Ignat to show this is what he is watching? Instead Tarkovsky allows the scene to point up the anxious without feeling obliged to dramatise the scene generically. The theme is there, the problem explored, but not quite contained, confined, boxed in.

Many a film, and numerous great ones in the Hollywood idiom, offer the cinegramatically strong partly by containing the anxiety within generic demand. Whether it is Hitchcock’s The Birds or Spielberg’s Jaws, the high anxiety generated contains a categorical notion of terror. Yet Tarkovsky dilutes the terror in a twofold manner. First he removes genre: there is no horror tension available as there happens to be in Hitchcock’s film and Spielberg’s movie, and secondly by diluting the connections between the very shots themselves. This sequence is ‘about’ anxiety, but it is as though Tarkovsky was interested in Kierkegaard’s distinction between existential anxiety and experiential fear: the difference between an intangible dread and a horror we see in front of our eyes. Both the birds and the shark are tangible fears; but how to locate anxiety somewhere else, in the intangible, and find an image structure to capture such indirectness? Tarkovsky does so by removing the generic framework and then also the grammatical expectation. This scene has nothing directly to do with the rest of the film, and isn’t even very concretely located within the context of Ignat watching the television.

The second scene of anxiety comes when the mother believes she might have allowed a typo to go through at the newspaper where she works. The scene follows Maria through the building in a series of long shots that captures well her feeling of urgency met by apprehension. By the end of the sequence Maria realizes that the error hasn’t been made and we are aware that the sequence is basically an anxiety dream. Now in a different type of film Tarkovsky could be accused of cheating: that here he is creating a scene possessing a twofold inconsequentiality. First of all the scene would seem to be a nightmare; secondly there is no mistake in the newspaper. This could be the very definition of empty suspense at its most ruthless, and how often do we see horror films that either show a character getting into what seems like a dangerous situation only for that strange noise to be no more than a broom falling, or getting horribly killed only for it to be an anxiety dream out of which the character wakes? But the very reason why we ‘forgive’ Tarkovsky and not the horror filmmaker is because Tarkovsky eschews the twofold generic framework and the grammatical expectation. The dream and the mistake aren’t there to manipulate the audience into a world of concrete fears with the occasional trick, but to work from the very basis of abstract fears and to offer a series of scenes that capture something of Kierkegaard’s angst over terror.

In the horror genre this takes place in a tangible world. Of course we are watching a fiction film, but within that fiction it is only in the trick dream sequence where we usually feel that what we have been watching has been unreal. In Mirror, though, the whole film takes on this unreal quality, and this is where it segues into a technique not unlike Pasolini’s interest in free indirect discourse; into stream of consciousness. If Pasolini drew on free indirect discourse to indicate a filmmaker’s freedom to deviate from the story he tells, to allow the camera to show us images that need not only indicate a character’s point of view, but also the perspective of the filmmaker, then Tarkovsky finds a variation of it and one consistent with his interest in the flow of time. A term created by philosopher William James, and practised by Joyce, Woolf and Faulkner, stream of consciousness is described by James thus: “Consciousness, then, does not appear to itself chopped up in bits. Such words as ‘chain’ or ‘train’ do not describe it fitly as it presents itself in the first instance. It is nothing jointed; it flows. A ‘river’ or a ‘stream’ are the metaphors by which it is most naturally described. In talking of it hereafter, let us call it the stream of thought, of consciousness, or of subjective life.” (The Stream of Consciousness)

One of the definitions of stream of consciousness is that it lacks standard punctuation to give greater freedom to thought, and if Pasolini’s theory showed deviations in the grammar of film as Antonioni and Godard would sometimes leave behind their characters for an interest in their surroundings, Tarkovsky pushes still further into the grammatically incoherent by a juxtaposition of scenes that don’t so much deviate from, but eradicate the basis of, the story. By way of comparison, and to bring out the difference between cinematic free indirect discourse and stream of consciousness, let us look at ostensibly similar moments in Mirror and Antonioni’s The Passenger. In the latter, the penultimate shot is justly famous. On both a technical and formal level it is distinctive. Here we have central character David Locke lying on the bed as the camera drifts off through the bars and focuses on various actions taking place outside the hotel room. This type of shot is quite common in cinema, however brilliant Antonioni’s example happens to be. We find variations of it in Le Mepris, where the camera leaves the characters behind and focuses on the bushes, in Taxi Driver when a three hundred and sixty degree shot leaves Bickle behind in the carpool and then picks him up a few seconds later, and in Battle in Heaven, where during a lovemaking session the camera exits the room and circles the view from outside the window. These would all be shots consistent with free indirect discourse: the centre of the narrative holds, but the digression is available.

In Mirror we follow the camera round a room as the phone rings and off screen Alexei answers it while the camera continues roaming through the space, passing through a corridor and into a large sitting room. The film then cuts to the anxiety dream we’ve already invoked, with Maria working at the printing house. The place had just been mentioned in the conversation between Alexei and his mother, when Maria says that a colleague of hers from many years earlier had just died. Now we can locate this scene with the roaming camera and the telephone call presumably in the relative present (we see the Andrei Rublev poster on the wall), and see the printing house sequence as a recollection to earlier times. But there are several elements that would stop us doing so. Firstly we are not cued to assume this is a flashback; we are merely asked to accept its weak linkage through the phone conversation. Instead of cutting when Maria talks about Liza to the scene where Maria goes to the printing press to see if she has made a mistake, the camera continues moving through the house and then abruptly cuts to Maria running along the street, with the film moving into black and white. There is no sense that she is relaying this information to her son which would have given it flashback assertion; instead it suggests much more a stream of consciousness from a position beyond the diegesis, where the linkages aren’t causally connected nor causally diluted (as we find in the free indirect approach), but causally disconnected and instead associative. It is as if the rain in the scene (which begins shortly after she starts running to the printing works) is matched by water from the shower at the end of the sequence, just as the orb of light we see early on as she runs along the street is matched by the lights we see in the corridor.

This is an associational cinema that insists the connections are weak because of our understandable assumptions about strong connections. If for example somebody changes the direction of the conversation on the basis of a pun rather than on a point being made, then we feel the weakness of the link, as if the person was looking for the flimsiest of premises to change the topic. If the person responds to the previous point made, then a strong link is assumed. Most films work off these strong linkages, and Pasolini’s remarks on Antonioni, Godard and Bertolucci simply indicate their modest weakening rather than their radical exhaustion. This modest tweaking was not quite there in Mizoguchi and early Bergman, Pasolini says, but became evident in Godard and others. In Mizoguchi and Bergman the films “were not, in other words, shot according to the canons of the “language of poetry”. Their poetry was elsewhere than in language as technique of language.” (Heretical Empiricism) Yet what Pasolini says of Mizoguchi and Bergman could still be said of sixties Godard and most of Antonioni. They were still making films with a high degree of linearity, while allowing digressions within this linear narrative, and indeed most of Tarkovsky’s films also are closer to a digressive free indirect discourse than stream of consciousness. They digress from character but still hold to character; where in Mirror the latter technique obliterates character and replaces it with this still freer form.

This is partly why we have here a narrative consciousness rather than a character. We know that Alexei is around forty, know that he has split up from his wife and has a son, but he remains a voice and a mind, not a character within the diegesis except as a boy and in memory. It isn’t so much that the film doesn’t make sense; more that it doesn’t make the sense we expect. There are moments throughout the film, for example, that help us locate the film in time. At one moment the wife talks about the fact that the Alexei’s mother was in a state for days after Liza’s death but Alexei hadn’t been in contact. At another moment there is a mention of recalling the burning dacha. Yet the film remains enigmatic in its form. This is why we feel Daney is right to talk of a watcher-viewer over the interpreter-viewer: much of the mystery lies in the very approach the film takes stylistically.

To understand the radical nature of the project, we can mention remarks made by Tarkovsky’s regular cinematographer Vadim Yusov and by chairman of the centralized government agency Filip Yermash. Yusov refused to shoot Mirror, saying, Tarkovsky conveys, “he found the frankly autobiographical nature of the work distasteful from an ethical point of view; he was embarrassed and irritated by the unduly personal, lyrical tone of the whole narrative, and by the author’s desire to talk exclusively about himself.” (Sculpting in Time). When shown the finished film, Yermash rejected it. “He found it all incomprehensible.” (Time within Time) Now there are certain films Mirror ostensibly resembles, like W.R, Mysteries of an OrganismIllumination and My American Uncle, where the film is both argumentative in its form and magpie in its relationship to footage. When “an explicitly argued film does come along”, according to Seymour Chatman, “the general public and critics are likely to be puzzled and angry.” (Coming to Terms) Now Mirror isn’t an explicitly argued film, as Chatman believes My American Uncle to be, but it is, like the others, a work of diegetic vulnerability – it is willing to weaken its story by contaminating it with extra-diegetic features like archival footage, and making scenes follow from each other in states of semi non sequitur. They don’t possess a powerful, Aristotelian through line. They all utterly and deliberately fail to conform to Aristotle’s notion of the three unities: that there is a unity of time, of place and of action. “A whole is that which has a beginning, a middle and an end.” (Poetics)

We see this for example In the scene in Mirror, where we have the son talking on the phone with his father, and here the film does allow the phone conversation to invoke directly the return to time past but then quickly accepts a stream of consciousness approach to take over. As the father tells his son about his time in the army as a boy and the presence of a beautiful redhead, the film cuts to a girl with red hair before showing the son and then a boy’s experiences in an army shooting range where a grenade almost goes off accidentally. The film then cuts to archival footage of WWII. The comment about the redhead isn’t an opportunity to relate a story from the past, but to move into recollection, and these aren’t quite one and the same thing. Again we can think of actual analogies. When we’re in conversation and someone mentions the time they worked in a bakers or wherever on leaving school, after they’ve finished their anecdote we might match it with one of our experiences working in a job when we were young, but the idea is that we will do so anecdotally – that we will give it small scale unity as we tell our tale: Aristotle in miniature. However, we might instead allow our thoughts for a few moments to recall our own past experiences but leave them unformed, fleeting, unconnected pieces of past events, and choose not to divulge them. It is the latter that stream of consciousness often works with in literature, and it is something of this freedom Tarkovsky seeks in Mirror. As the films drifts on to the moment with the grenade, the redhead gets ignored and then the archival footage is shown. We have some indication that the scene with the grenade and the documented footage take place on the same temporal plane, but that the former is fictional the latter factual, that the former is in colour, the latter black and white, dilutes any dramatic continuity or ready sense.

Perhaps Tarkovsky asked himself what is intimacy in cinematic form; how can one generate a global intimacy just as Communism hoped to create a global fraternity, and in the process created individual excavation? Soviet montage cinema created a people, instead of a person. “I reject the principles of ‘montage cinema’ because they do not allow the film to continue beyond the edges of the screen: they do not allow the audience to bring personal experience to bear on what is in front of them on film.” (Sculpting in Time) This however isn’t to make personal cinema, especially, but instead cinema that is personal: a cinema of intimate contact with the world. If Pasolini wanted to find in free indirect discourse a means by which to free the filmmaker’s aesthetic by tying it to a character’s consciousness and then showing it indirectly, then this is not quite what Tarkovsky desired in Mirror no matter if both of them indicate deviations. When we think of the penultimate shot in The Passenger, or the scene in Battle in Heaven, we can say in Pasolini’s terms, that the film isn’t showing what the character sees (a direct point of view), but what he might be thinking at one remove. It is as though the shot in The Passenger asks, through Locke, what would it be like to leave this world after earlier trying to leave his ‘body’ by changing identity  – by utilising the passport of the dead Robertson? In Battle in Heaven, the “thought’ as the camera floats around the courtyard could be central character Marcos’s, wondering what other life he could have if he could somehow free himself from the body he is trapped in. But in Mirror Tarkovsky instead wants a different form of intimate discourse: he wants not to deviate from his narrative and create a hypothesis around the character, but instead to allow the ‘mind’ – the film – to wander associationally as it seeks to discover cinematically in a literary technique a freedom antithetical to the rigorous external discourse of the Soviet state.

Indeed, and in conclusion, when James talks of ‘substantive’ and ‘transitive’ states of mind, it resembles Tarkovsky’s remarks on the difference he sees between his work and the Soviet montage filmmakers. “Let us call the resting-places the substantive parts’ and the places of flight the ‘transitive parts’ of the stream of thought. It then appears that our thinking tends at all times towards some other substantive part than the one from which it has just been dislodged. And we may say that the main use of the transitive parts is to lead us from one substantive conclusion to another.” (The Stream of Consciousness) Tarkovsky believes, “we can only talk about the components rather arbitrarily, dividing it up artificially for the sake of theoretical discussion” (Sculpting in Time) and that while “every art form involves editing…adjusting parts and pieces”, the cinema image comes into being during shooting, and exists within the frame. It is both substantive and transitive. The Soviet montage filmmakers concentrated in contrast too much on the substantive and failed to capture the rhythm of the shot, the flow of time within the very image that comes from lived reality. This flow of time within the image is like the flow of time we have to engage in private thoughts, to have the room for our stream of consciousness to meet the film’s and arrive at a properly intimate encounter. As the director said of editing: “it is not a question of mastering the technique like a virtuoso, but of a vital need of your own, distinct, individual expression.” (Sculpting in Time)

Tarkovsky’s film might be about his life, his mother’s and his father’s, and it might have in small roles his wife and his mother as well as running through it his father’s poetry, but it is also, and perhaps especially, personal, in its creation of images that indicate what very few films have managed to do and that is invoke a stream of consciousness. There have been hints at it in some of the films we have mentioned, and some critics could namecheck Last Year at Marienbad and L’Immortelle, but Mirror remains the film that takes the free indirect discourse theorised by Pasolini and that can be aptly applied to the narratively digressive works of Antonioni, sixties Godard and a number of other films, into the realm of a properly cinematic stream of consciousness, into a mirror of the mind. This is where a character is not set, where an actress can play both wife and mother, where colour can intermingle with black and white, where a scene can drift from what would appear to be its narrative centre, and where time is conditionally contained by much more than the givens of dramatic expectation. It remains of course a masterpiece, but that word still doesn’t quite do justice to its singularity. The film is a mirror that manages to reflect the most intimate of feelings without at all ignoring history both cinematic and social.

 

© Tony McKibbin

Tony McKibbin Tony McKibbin

Mirror

The Rhythm of Consciousness

What is a cinematic self-portrait? In one scene during Andrei Tarkovsky's Mirror a young boy flicks through a book and comes across a a self-portrait of Leonardo da Vinci. We have no problem naming it as such, and self-portraiture in art has a long tradition, but what of self-portraiture in film? Some might think of Derek Jarman's Blue, the personal work of Alain Cavalier, Godard's JLG versus JLG. But few films have more ambitiously found a form for personal reflection than Mirror, and it is entirely apt that the film goes by such a name.

The film is both astonishingly personal and fascinatingly general. At one moment the camera passes through a room and we see a poster of Tarkovsky's second film, Andrei Rublev on the wall. The poetry we hear in voice-over is Tarkovsky's father, Arseny, reading his own work. Tarkovsky's wife, Larissa Tarkovskya and his mother make an appearance, and the film is an account of both Arseny and Andrei's childhood with the same actress, Margarita Terekhova, playing both Arseny and Andrei's mother. Yet there is also the impersonal history: stock footage from various different time periods, including the Spanish Civil War and WWII. These moments are invoked neither arbitrarily nor concretely. The Spanish Civil war footage comes during a moment when some Spanish friends of the family are around the house; the WWII footage after the father (Alexei) speaks to his son Ignat on the phone and mentions a girl he was fascinated by during those years as we notice the son is also playing the role of the father as a boy. We can notice the same mole above his lip, just as we notice that the mother and wife have three moles on the neck. Obviously these are attributes of the actor and actress; but they're also distinguishing features that will leave us in no doubt the same person is playing both roles.

Thus we have personal reflection with impersonal archive, yet we also have black and white footage intermingling with the colour image without clear distinctions made between the use of one over that of the other. Many films of course use black and white to signify one thing and colour photography to illustrate something else. The Wizard of Oz, A Matter of Life and Death, Raging Bull, Pleasantville and Rumblefish allow us to locate the distinct use of each. In The Wizard of Oz Kansas is in monochrome, Oz in colour; in A Matter of Life and Death earth is in colour and the heavenly way station in black and white. Raging Bull uses colour for home movie footage, while Rumblefish utilises colour symbolically evident in the colourful fish of the title. In Pleasantville colour is steadily more evident as the townspeople become more radiant and happy. Tarkovsky himself used colour consistent with these approaches in Andrei Rublev: most of the film is in black and white, but the conclusion shows various Russian icons painted by the artist of the title in colour.

But as in his own Stalker, as in Nagisa Oshima's The Diary of a Shinjuku Thief and Boy, Lindsay Anderson's If... colour and black and white are shown indeterminately. We can't seek a logic in the image but must accept that the associations become vague and troubling. There seems to be a general use of black and white or sepia in the archival footage and in the scenes that suggest a dream; while the scenes in the thirties and forties and the present are in colour. But this logic doesn't hold: there are dreamlike images in colour; moments in the present in black and white. There is deliberate vagueness too in casting Terekhova as both the wife and the mother. Like Ulysses' Gaze which has the same actress playing four roles, or That Obscure Object of Desire, where two actresses play the one part, Mirror calls into question the gap between the character and the actor. If we have the choice between colour and monochrome utilised in such a way that the image becomes indeterminate by the use of both without a clear rationale, then equally film has the opportunity to distinguish between the role and the actor. In most fiction films the purpose is to close the gap between the two to create a seamless impression of a subject, of an indivisible self in the role, no matter how divisible they actually happen to be. Yet by casting the same actress in two parts, Tarkovsky creates a crisis in the image equal to his play with black and white and colour. In each instance the crisis happens to be epistemological: what can we safely say about the film we are watching? How can we be sure that at certain moments we are watching the wife and not watching the mother, and vice versa, just as we cannot be sure what the shift from black and white to colour means. Certainly we can make certain assumptions about where we are in time, but Tarkovsky doesn't make things easy for us to do so partly by casting the same person in two roles.

This is where a couple of epistemological options make themselves available to us. One is to assume that the film doesn't ask for sense but just washes over the viewer, so that we let our minds go and respond affectively to these images that we can't quite shape into narrative sense. The other is to insist that they must add up to a coherent whole and we thus shape them into a determinate meaning that allows us to eradicate the crisis the image generates. But surely for Tarkovsky and other masters of the period (like Herzog, Pasolini, Wenders and Oshima), the very crisis is what interests them. Tarkovsky insisted that "as I began working on Mirror I found myself reflecting more and more that if you are serious about your work, then a film is...an action which will affect the whole of your life. For I had made up my mind that in this film, for the first time, I would use the means of cinema to talk of all that was most precious to me..." (Sculpting in Time). Talking about literature and free indirect discourse, but at the same time understanding its possible use in cinema, Pasolini said "...entire novels are no more than entire Free Indirects in that either there is a total identification of the author with a character, or the characters are a pseudo-objectification of the author, or the characters are devices for expressing the thesis of the author in a substantially unified language, or finally - unconsciously - the characters perfectly inhabit in the same way the social and ideological world of the author..." (Heretical Empiricism) Both Tarkovsky and Pasolini are talking about crisis however as an indirect expression, evident when Tarkovsky adds "Mirror was not an attempt to talk about myself, not at all. It was about my feelings towards people dear to me; about my relationship with them; my perpetual pity for them and my own inadequacy - my feelings of duty left unfulfilled." (Sculpting in Time)

This suggests that rather than assuming the film should wash over us (a version of the affective fallacy where our feelings are what is most important in the art work), or believing that the film has a clear meaning if we only look hard enough to find it (the intentional fallacy where we credit the importance to the rigour of the artist), we instead accept a tentative and speculative relationship that invokes our own responses into what we might assume are the director's intuitions. We neither assume a grand plan or settle for its absence, but feel instead unsettled by the fact it is neither one thing nor the other, and try and find a position between these two antinomies. Serge Daney, reviewing Stalker, said "A film can be interpreted. This one in particular lends itself to it (even if in the end it hides its secrets.) But we are not obliged to interpret it. A film can be watched too. One can watch for the appearance of things which one has never seen before in a film. The watcher-viewer sees things which the interpreter-viewer can no longer make out. The watcher stays at the surface because he doesn't believe in depth." (Liberation)

To be a watcher-viewer however means making sense of our senses. It doesn't mean surrendering to them, and it doesn't mean denying them for symbolic interpretation. It means remaining within the troubling realm. If we think about how the shots are constructed intuitively, if we feel the film is resisting our faculties of expectation, we are remaining on the surface of the film, but we are not refusing its Tarkovskian originality: the fact that it wants to be a work of cinema within the world of singular sensibility.

To help us understand the Tarkovskian we will have more to say about the freedom with which Tarkovsky creates his images. But it would also be useful to say a few things first about the Soviet montage cinema against which Tarkovsky was rebelling. We might think especially here of Lev Kuleshov's experiments and both creative geography and the manipulation of emotions through film form. In the former example Kuleshov would show someone walking along the street and then the filmmaker would cut to the Eiffel Tower, Red Square or Capitol Hill and the viewer would make the assumption that this is where they are going. Again this indicated the importance of the cut. The actor could be walking along any street, perhaps many miles from the place in the next shot, but the film language makes the viewer assume that they are closely linked and this is where the character intends to go. In the latter, an actor would retain exactly the same expression on his face, and the director would cut to an attractive woman lying on a settee, to a bowl of soup, to a young girl in a coffin, and the viewer would read different expressions on the face each time. This showed that it wasn't acting in cinema that necessarily counted, but the juxtaposition of shots. This was the Kuleshov effect: "The Kuleshov effect is an editing principle in which an interpretive effect is created by a sequence of shots that by themselves are not explanatory. The imagined connection between the shots establishes the meaning of both shots." (Slavische Studies)

Though there were many points of disagreement between Kuleshov, Eisenstein and Pudovkin, they all agreed in the importance of montage. Tarkovsky saw this as a false priority, and thus instead talked up the rhythm of the shot: the integrity of the particular sequence and the integrity of performance and location. Yet this wasn't a straightforward approach to what was in front of the camera. His films weren't interested especially in creating a copy of the world, but instead in creating out of the camera a vision that would be integral and not overly edited, or rather, if edited, edited with an awareness of the time that each shot contains. "Editing exists in every art form, since material always has to be selected and joined. What is different about cinema editing is that it brings together time, imprinted in the segments of film." (Sculpting in Time)

This is a variation perhaps of Pasolini's notion of the written language of reality. If literature could be reduced to recognizable smallest units, phonemes or letters, then the shot had no such reducibility. What was the smallest unit in film? If we say it is a shot, then we have to acknowledge that most shots are made of many items within that unit. If the smallest unit in literature could be a or e, the smallest in film would be closer to: tree, earth, sun, cloud. It would be made up of a series of things, not merely the equivalent of one letter. Here is where Pasolini asks what cinema does. "What does the grammar of the language of cinema fish from reality? It fishes the smallest units, the units of the second articulation: the objects, the forms, the acts of reality which we have called "kinemes". After having fished them, it keeps them in itself, encapsulating them in its units of first articulation, the monemes - that is the shots."

One of the key differences between film and literature is that the word is without time, it belongs not to the temporal order of the world, but a symbolic order on the page. It is removed from reality, while the film image is contained by reality. The is why Pasolini can talk of film as the written language of reality and why Tarkovsky can insist that "editing entails assembling smaller and larger pieces, each of which carries a different time", adding "their assembly creates a new awareness of the existence of that time." What Eisenstein and other montage filmmakers lacked was what Tarkovsky called "time-truth": that, in many films, by playing up the editing whilst playing down the temporal it is as though "in terms of the editing they are quite indistinguishable." "On the other hand "you will always recognise the editing of Bergman, Bresson, Kurosawa or Antonioni; none of them could ever be confused with anyone else, because each one's perception of time, as expressed in the rhythm of their films, is always the same." (Sculpting in Time)

Taking into account both Pasolini and Tarkovsky's remarks, what we see is a need to acknowledge the language of reality in the accumulation of the shots. This is of course what Kuleshov and others claimed was unnecessary. What mattered in Kuleshov's experiments, for example, was not the language of reality but the language of film, and this is why the viewer would automatically assume that if someone walked along the street and you cut to the Eiffel tower, then this would be where they are going. This is why if you cut from a bowl of soup to the man looking on, we will assume he is hungry. It is as if the montage filmmakers, and indeed many a Hollywood filmmaker thereafter, wanted the film's grammar to carry the film's meaning more than the language of reality that allows for its existence.

However this doesn't leave Pasolini and Tarkovsky as naive realists; if anything it is their respect for the written language of reality that leads to the complexity of their work. Thus if Mirror remains such an elusive film, it resides in the cut losing many of its grammatical coordinates whilst still holding on to elements of its temporal milieu. When for example the film shows us a burning dacha, the grammar indicates a sequence in one temporal zone, but the features of the sequence spatially makes it difficult to organize the scene perceptually with confidence. As we see the two children (Alexei and his sister) looking out by the front door in reflection, a boy comes from the rear of the house and passes by the mirror that is showing the kids reflected. The two children seem to be watching the very burning dacha that we will see the boy looking at, but the play with space here is slightly baffling. As the boy looks out from the house, he sees in front of him an old man nearer to the fire, and a woman looking on. The film then cuts to what we might at first assume is a closer shot of the woman, but instead to the Alexei's mother, Maria, coming towards the burning building, as the man tries to put it out. The grammar of the sequence indicates that they are part of the same temporal moment, but the logic of the shots doesn't easily match our grammatical expectations.

Tarkovsky refuses the usual assumptions. He could have shown the children looking out, cut back to a reaction shot from the children's faces, and then shown the mother in front of them going towards the fire. Instead it is as though the camera must float through and round the characters, leaving the story and the figures within it secondary to a narrational reality based on memory and reflection. Of course numerous other film makers were refusing this focus on story and character, but just because other filmmakers were doing likewise (Antonioni, Angelopoulos, Garrel), this shouldn't quite pass for an explanation, otherwise the very remarks about Bergman et al would become redundant. We would notice that while Tarkovsky wouldn't be editing like a Hollywood filmmaker, he would be editing like an arthouse one. It would also indicate a weak dramaturgy next to the strong dramatic reasoning of Hollywood, where the filmmaker could easily justify their editing decisions on the basis of dramatic logic. The arthouse filmmaker would be positioned as merely reactive. We need to find a logic in Tarkovsky's images equal to those in more commercial cinema, and for the added purpose of suggesting a singular response: the Tarkovskian invoked by Daney. Yet this isn't the same as crediting Tarkovsky with a given meaning either. It is closer to a resistant (quite different from reactive) aesthetics; one that tries to find a meaning not out of assertive assumption (as we find in many a mainstream work), but out of tentative exploration: "the true cinema image is built upon the destruction of genre, upon conflict with it." (Sculpting in Time)

How is this achieved? Let us think of a couple of scenes in the film that play up anxiety. There is the pre-credit sequence where Ignat turns on the television, and the film then cuts to a young man who has a stutter corrected by a speech therapist. As the opening shot shows Ignat looking at the TV before the film cuts to show the scene between the stutterer and the coach, there is no tracking shot into the television to suggest the boy is watching a particular programme, and the TV we see has a blank blueish screen. The film then cuts to this black and white sequence showing the therapist working with the young man. These links are what we might call cinegrammatically weak: shouldn't the TV show Ignat watches be in colour, and shouldn't we have shots back and forth between the programme and Ignat to show this is what he is watching? Instead Tarkovsky allows the scene to point up the anxious without feeling obliged to dramatise the scene generically. The theme is there, the problem explored, but not quite contained, confined, boxed in.

Many a film, and numerous great ones in the Hollywood idiom, offer the cinegramatically strong partly by containing the anxiety within generic demand. Whether it is Hitchcock's The Birds or Spielberg's Jaws, the high anxiety generated contains a categorical notion of terror. Yet Tarkovsky dilutes the terror in a twofold manner. First he removes genre: there is no horror tension available as there happens to be in Hitchcock's film and Spielberg's movie, and secondly by diluting the connections between the very shots themselves. This sequence is 'about' anxiety, but it is as though Tarkovsky was interested in Kierkegaard's distinction between existential anxiety and experiential fear: the difference between an intangible dread and a horror we see in front of our eyes. Both the birds and the shark are tangible fears; but how to locate anxiety somewhere else, in the intangible, and find an image structure to capture such indirectness? Tarkovsky does so by removing the generic framework and then also the grammatical expectation. This scene has nothing directly to do with the rest of the film, and isn't even very concretely located within the context of Ignat watching the television.

The second scene of anxiety comes when the mother believes she might have allowed a typo to go through at the newspaper where she works. The scene follows Maria through the building in a series of long shots that captures well her feeling of urgency met by apprehension. By the end of the sequence Maria realizes that the error hasn't been made and we are aware that the sequence is basically an anxiety dream. Now in a different type of film Tarkovsky could be accused of cheating: that here he is creating a scene possessing a twofold inconsequentiality. First of all the scene would seem to be a nightmare; secondly there is no mistake in the newspaper. This could be the very definition of empty suspense at its most ruthless, and how often do we see horror films that either show a character getting into what seems like a dangerous situation only for that strange noise to be no more than a broom falling, or getting horribly killed only for it to be an anxiety dream out of which the character wakes? But the very reason why we 'forgive' Tarkovsky and not the horror filmmaker is because Tarkovsky eschews the twofold generic framework and the grammatical expectation. The dream and the mistake aren't there to manipulate the audience into a world of concrete fears with the occasional trick, but to work from the very basis of abstract fears and to offer a series of scenes that capture something of Kierkegaard's angst over terror.

In the horror genre this takes place in a tangible world. Of course we are watching a fiction film, but within that fiction it is only in the trick dream sequence where we usually feel that what we have been watching has been unreal. In Mirror, though, the whole film takes on this unreal quality, and this is where it segues into a technique not unlike Pasolini's interest in free indirect discourse; into stream of consciousness. If Pasolini drew on free indirect discourse to indicate a filmmaker's freedom to deviate from the story he tells, to allow the camera to show us images that need not only indicate a character's point of view, but also the perspective of the filmmaker, then Tarkovsky finds a variation of it and one consistent with his interest in the flow of time. A term created by philosopher William James, and practised by Joyce, Woolf and Faulkner, stream of consciousness is described by James thus: "Consciousness, then, does not appear to itself chopped up in bits. Such words as 'chain' or 'train' do not describe it fitly as it presents itself in the first instance. It is nothing jointed; it flows. A 'river' or a 'stream' are the metaphors by which it is most naturally described. In talking of it hereafter, let us call it the stream of thought, of consciousness, or of subjective life." (The Stream of Consciousness)

One of the definitions of stream of consciousness is that it lacks standard punctuation to give greater freedom to thought, and if Pasolini's theory showed deviations in the grammar of film as Antonioni and Godard would sometimes leave behind their characters for an interest in their surroundings, Tarkovsky pushes still further into the grammatically incoherent by a juxtaposition of scenes that don't so much deviate from, but eradicate the basis of, the story. By way of comparison, and to bring out the difference between cinematic free indirect discourse and stream of consciousness, let us look at ostensibly similar moments in Mirror and Antonioni's The Passenger. In the latter, the penultimate shot is justly famous. On both a technical and formal level it is distinctive. Here we have central character David Locke lying on the bed as the camera drifts off through the bars and focuses on various actions taking place outside the hotel room. This type of shot is quite common in cinema, however brilliant Antonioni's example happens to be. We find variations of it in Le Mepris, where the camera leaves the characters behind and focuses on the bushes, in Taxi Driver when a three hundred and sixty degree shot leaves Bickle behind in the carpool and then picks him up a few seconds later, and in Battle in Heaven, where during a lovemaking session the camera exits the room and circles the view from outside the window. These would all be shots consistent with free indirect discourse: the centre of the narrative holds, but the digression is available.

In Mirror we follow the camera round a room as the phone rings and off screen Alexei answers it while the camera continues roaming through the space, passing through a corridor and into a large sitting room. The film then cuts to the anxiety dream we've already invoked, with Maria working at the printing house. The place had just been mentioned in the conversation between Alexei and his mother, when Maria says that a colleague of hers from many years earlier had just died. Now we can locate this scene with the roaming camera and the telephone call presumably in the relative present (we see the Andrei Rublev poster on the wall), and see the printing house sequence as a recollection to earlier times. But there are several elements that would stop us doing so. Firstly we are not cued to assume this is a flashback; we are merely asked to accept its weak linkage through the phone conversation. Instead of cutting when Maria talks about Liza to the scene where Maria goes to the printing press to see if she has made a mistake, the camera continues moving through the house and then abruptly cuts to Maria running along the street, with the film moving into black and white. There is no sense that she is relaying this information to her son which would have given it flashback assertion; instead it suggests much more a stream of consciousness from a position beyond the diegesis, where the linkages aren't causally connected nor causally diluted (as we find in the free indirect approach), but causally disconnected and instead associative. It is as if the rain in the scene (which begins shortly after she starts running to the printing works) is matched by water from the shower at the end of the sequence, just as the orb of light we see early on as she runs along the street is matched by the lights we see in the corridor.

This is an associational cinema that insists the connections are weak because of our understandable assumptions about strong connections. If for example somebody changes the direction of the conversation on the basis of a pun rather than on a point being made, then we feel the weakness of the link, as if the person was looking for the flimsiest of premises to change the topic. If the person responds to the previous point made, then a strong link is assumed. Most films work off these strong linkages, and Pasolini's remarks on Antonioni, Godard and Bertolucci simply indicate their modest weakening rather than their radical exhaustion. This modest tweaking was not quite there in Mizoguchi and early Bergman, Pasolini says, but became evident in Godard and others. In Mizoguchi and Bergman the films "were not, in other words, shot according to the canons of the "language of poetry". Their poetry was elsewhere than in language as technique of language." (Heretical Empiricism) Yet what Pasolini says of Mizoguchi and Bergman could still be said of sixties Godard and most of Antonioni. They were still making films with a high degree of linearity, while allowing digressions within this linear narrative, and indeed most of Tarkovsky's films also are closer to a digressive free indirect discourse than stream of consciousness. They digress from character but still hold to character; where in Mirror the latter technique obliterates character and replaces it with this still freer form.

This is partly why we have here a narrative consciousness rather than a character. We know that Alexei is around forty, know that he has split up from his wife and has a son, but he remains a voice and a mind, not a character within the diegesis except as a boy and in memory. It isn't so much that the film doesn't make sense; more that it doesn't make the sense we expect. There are moments throughout the film, for example, that help us locate the film in time. At one moment the wife talks about the fact that the Alexei's mother was in a state for days after Liza's death but Alexei hadn't been in contact. At another moment there is a mention of recalling the burning dacha. Yet the film remains enigmatic in its form. This is why we feel Daney is right to talk of a watcher-viewer over the interpreter-viewer: much of the mystery lies in the very approach the film takes stylistically.

To understand the radical nature of the project, we can mention remarks made by Tarkovsky's regular cinematographer Vadim Yusov and by chairman of the centralized government agency Filip Yermash. Yusov refused to shoot Mirror, saying, Tarkovsky conveys, "he found the frankly autobiographical nature of the work distasteful from an ethical point of view; he was embarrassed and irritated by the unduly personal, lyrical tone of the whole narrative, and by the author's desire to talk exclusively about himself." (Sculpting in Time). When shown the finished film, Yermash rejected it. "He found it all incomprehensible." (Time within Time) Now there are certain films Mirror ostensibly resembles, like W.R, Mysteries of an Organism, Illumination and My American Uncle, where the film is both argumentative in its form and magpie in its relationship to footage. When "an explicitly argued film does come along", according to Seymour Chatman, "the general public and critics are likely to be puzzled and angry." (Coming to Terms) Now Mirror isn't an explicitly argued film, as Chatman believes My American Uncle to be, but it is, like the others, a work of diegetic vulnerability - it is willing to weaken its story by contaminating it with extra-diegetic features like archival footage, and making scenes follow from each other in states of semi non sequitur. They don't possess a powerful, Aristotelian through line. They all utterly and deliberately fail to conform to Aristotle's notion of the three unities: that there is a unity of time, of place and of action. "A whole is that which has a beginning, a middle and an end." (Poetics)

We see this for example In the scene in Mirror, where we have the son talking on the phone with his father, and here the film does allow the phone conversation to invoke directly the return to time past but then quickly accepts a stream of consciousness approach to take over. As the father tells his son about his time in the army as a boy and the presence of a beautiful redhead, the film cuts to a girl with red hair before showing the son and then a boy's experiences in an army shooting range where a grenade almost goes off accidentally. The film then cuts to archival footage of WWII. The comment about the redhead isn't an opportunity to relate a story from the past, but to move into recollection, and these aren't quite one and the same thing. Again we can think of actual analogies. When we're in conversation and someone mentions the time they worked in a bakers or wherever on leaving school, after they've finished their anecdote we might match it with one of our experiences working in a job when we were young, but the idea is that we will do so anecdotally - that we will give it small scale unity as we tell our tale: Aristotle in miniature. However, we might instead allow our thoughts for a few moments to recall our own past experiences but leave them unformed, fleeting, unconnected pieces of past events, and choose not to divulge them. It is the latter that stream of consciousness often works with in literature, and it is something of this freedom Tarkovsky seeks in Mirror. As the films drifts on to the moment with the grenade, the redhead gets ignored and then the archival footage is shown. We have some indication that the scene with the grenade and the documented footage take place on the same temporal plane, but that the former is fictional the latter factual, that the former is in colour, the latter black and white, dilutes any dramatic continuity or ready sense.

Perhaps Tarkovsky asked himself what is intimacy in cinematic form; how can one generate a global intimacy just as Communism hoped to create a global fraternity, and in the process created individual excavation? Soviet montage cinema created a people, instead of a person. "I reject the principles of 'montage cinema' because they do not allow the film to continue beyond the edges of the screen: they do not allow the audience to bring personal experience to bear on what is in front of them on film." (Sculpting in Time) This however isn't to make personal cinema, especially, but instead cinema that is personal: a cinema of intimate contact with the world. If Pasolini wanted to find in free indirect discourse a means by which to free the filmmaker's aesthetic by tying it to a character's consciousness and then showing it indirectly, then this is not quite what Tarkovsky desired in Mirror no matter if both of them indicate deviations. When we think of the penultimate shot in The Passenger, or the scene in Battle in Heaven, we can say in Pasolini's terms, that the film isn't showing what the character sees (a direct point of view), but what he might be thinking at one remove. It is as though the shot in The Passenger asks, through Locke, what would it be like to leave this world after earlier trying to leave his 'body' by changing identity - by utilising the passport of the dead Robertson? In Battle in Heaven, the "thought' as the camera floats around the courtyard could be central character Marcos's, wondering what other life he could have if he could somehow free himself from the body he is trapped in. But in Mirror Tarkovsky instead wants a different form of intimate discourse: he wants not to deviate from his narrative and create a hypothesis around the character, but instead to allow the 'mind' - the film - to wander associationally as it seeks to discover cinematically in a literary technique a freedom antithetical to the rigorous external discourse of the Soviet state.

Indeed, and in conclusion, when James talks of 'substantive' and 'transitive' states of mind, it resembles Tarkovsky's remarks on the difference he sees between his work and the Soviet montage filmmakers. "Let us call the resting-places the substantive parts' and the places of flight the 'transitive parts' of the stream of thought. It then appears that our thinking tends at all times towards some other substantive part than the one from which it has just been dislodged. And we may say that the main use of the transitive parts is to lead us from one substantive conclusion to another." (The Stream of Consciousness) Tarkovsky believes, "we can only talk about the components rather arbitrarily, dividing it up artificially for the sake of theoretical discussion" (Sculpting in Time) and that while "every art form involves editing...adjusting parts and pieces", the cinema image comes into being during shooting, and exists within the frame. It is both substantive and transitive. The Soviet montage filmmakers concentrated in contrast too much on the substantive and failed to capture the rhythm of the shot, the flow of time within the very image that comes from lived reality. This flow of time within the image is like the flow of time we have to engage in private thoughts, to have the room for our stream of consciousness to meet the film's and arrive at a properly intimate encounter. As the director said of editing: "it is not a question of mastering the technique like a virtuoso, but of a vital need of your own, distinct, individual expression." (Sculpting in Time)

Tarkovsky's film might be about his life, his mother's and his father's, and it might have in small roles his wife and his mother as well as running through it his father's poetry, but it is also, and perhaps especially, personal, in its creation of images that indicate what very few films have managed to do and that is invoke a stream of consciousness. There have been hints at it in some of the films we have mentioned, and some critics could namecheck Last Year at Marienbad and L'Immortelle, but Mirror remains the film that takes the free indirect discourse theorised by Pasolini and that can be aptly applied to the narratively digressive works of Antonioni, sixties Godard and a number of other films, into the realm of a properly cinematic stream of consciousness, into a mirror of the mind. This is where a character is not set, where an actress can play both wife and mother, where colour can intermingle with black and white, where a scene can drift from what would appear to be its narrative centre, and where time is conditionally contained by much more than the givens of dramatic expectation. It remains of course a masterpiece, but that word still doesn't quite do justice to its singularity. The film is a mirror that manages to reflect the most intimate of feelings without at all ignoring history both cinematic and social.


© Tony McKibbin