1. A non-euclidean geometry of the cinematographic medium
In 1960 Roland Barthes published one of his first essays about cinema: Le probléme de la signification au cinéma. What is discussed in that work, apart from the first attempts to establish a semiotic theory for movies, is the notion of analogic relation between signified and signifier within a film:
Our art, and in particular our cinematographic art, places a very thin distance between signifier and signified; it is a strictly analogic semiology, and not a symbolic one: to signify a General, it is a uniform of a General that will be displayed, with all the details. Our filmic semiology considers the viewer a priori as untutored, and tends to provide a complete imitation of the signified (our theatre, instead, has passed through symbolist periods). The film director is obliged to produce a pseudo-physis; he is free to multiply the signifiers, not to rarefy them or to make them abstract; he has no access to the symbol, or to the sign (if not motivated), but only to the analogon.
Further on, Barthes describes the paradox of the Western film director: he is obliged to create something new, something original (unlike in Chinese theatre, where convention is very important and every action is ritualized) but always reproducing the natural: that is, the real, the logical. He has to constantly elaborate a neologism, without the chance of being abstract.
Looking at the history of cinema nowadays, is it possible to find examples of non-analogic cinema as theorized fifty years ago by Barthes?
One possible answer is provided by Peter Wollen. In his essay about Signs and meaning in the cinema (1998) he clarifies and summarizes the different attempts that have been made to study cinema as a system of signs (Bazin, Barthes, Metz, Panofsky among the others). Using Peirce’s terminology, he states that
unlike verbal language, primarily symbolic, the cinema is, as we have seen, primarily indexical and iconic. It is the symbolic which is the submerged dimension. We should therefore expect that in the “poetry” of the cinema, this aspect will be manifested more palpably.
This can be seen as an alternative way of formulating Barthes’ point. But Wollen argues that a symbolic aspect of cinema does exist.
One first form is “Primitive Symbolism” (as Panofsky calls it). In its first years, cinema was much more symbolic due to the lack of technical devices and (overall) to the lack of an audience’s familiarity with this new art. Panofsky recalls the Vamp and the Straight Girl, the Family Man and the Villain. Each of these type-characters presented a set of visual clues that, symbolically, reminded the viewer of their function in the story. In Wollen’s opinion, “primitive symbolism” survives still, in different forms, today. Plus, we can trace another story of symbolism in cinema, that is the story of cinematographic language in itself: crossing-fades, fades to black or to white, blurring, only to name a few, are used as a true symbolic language to signify things (time or space shifts, death, sleep, end of a sequence).
Secondly, we can refer to the Expressionist season in cinema as a highly symbolic period.
Bazin developed a bi-polar view of the cinema. On the one hand was Realism (‘the good, the true, the just’, as Godard was later to say of the work of Rossellini); on the other hand was Expressionism, the deforming intervention of human agency.
Bazin always supported the first view, and after all these years we have to deduce that the whole of cinema took that direction. Nonetheless, Expressionist movies have been made, and, watching The Cabinet of Dr. Caligari for example, we notice that a symbolic visual language was in operation. Probably Barthes, writing in 1960, saw that direction as closed.
In this essay we do not want to trace the genealogy of symbolic cinema. As reported above, symbolic significations in cinema are working now, as have always been working in the history of this medium. We want to broaden Barthes’ notion of analogon: as “seeking a non-analogic cinema” here we intend to seek the possibility of a cinematographic experience that can be non (entirely) signic, non-referent to the common logic or to the common experience of the world. The possibility of a cinema beyond the mere reproduction of the world. A non-euclidean geometry of the cinematographic medium, in a way.
The debate about postmodernism has as one of its favourite subjects the problem of the reference to the real. Baudrillard was the first to address this issue. He overturned the reference problem as raised by Barthes, saying: “the real is not only what can be reproduced, but that which is always already reproduced”. Much less has been talked about an art which does not try to reproduce the real (or the hyperreal, as postmodernist theorists would put it).
A paradoxical question arises: is it possible that a non-analogic art, an art that does not refer to the common (hyper)reality, can have – in the contemporary context –more possibilities to be relevant for the world we live in?
In this perspective, it is important to talk about cinema because it still is an artistic medium of mass-culture; easier – but probably less significant - would be applying this study to video-art. Two films were selected as possible examples of non-analogic cinema, because each of them provides an artistic method and a poetic direction.
The first film we want to consider as a possible non-analogic work is Stalker (1974), directed by Andrei Tarkovsky. It describes the journey of three men travelling through a post-apocalyptic wilderness called the Zone to find a room with the potential to fulfil one's innermost desires. The movie is not strictly non-analogic in the way Barthes would have probably intended the word: the objects still have an iconic (using Peirce’s terminology) reference to reality: to show a tree, Tarkovsky shows a tree, the same with a train or with a car. But we cannot say that the whole milieu where Stalker is set can be iconically referred to our reality. The first reason is that, even if the single objects are analogic, the entire reality of the movie does not seem to be. The Stimmung of the Zone is an oneiric one: it seems to have much more to do with dreams. One example: the “Meatgrinder” through which the characters have to go, is iconically described by the images as an ordinary tunnel; nonetheless, the viewer is fully convinced by the dialogues that it is a death device.
The second reason why we can consider Stalker as a non-analogic work is probably more important, and it is about the action and the motivations of the characters. The set of values and rules that dominate the moves of the main character within the Zone does not follow a usual logic. To decide the direction in which to go, for example, Stalker throws dice in an apparently random manner. Furthermore, even if the three protagonists see their destination (the Room) from the beginning of their journey, they are obliged to make a long detour to reach it, for no apparent reason. Having said that, we are tempted to state that the non-analogical way in which Tarkovsky drives his story is not much evident in what he shows (the objects are still iconic); it rather lies in what he does not show: for example the rules of the Zone and the reasons for the actions of the characters.
Using the notions of syntagm and system as elaborated by Barthes in his Elements of Semiology (1964), it is possible to state that Stalker avoids referentiality by shifting along the system axis. That is, the level of metaphor. It is in the actions (and in the objects) that the movie loses its reference to “reality”. The order of succession (the syntagm) of the story is quite normal: Tarkovsky here does not operate any huge innovation in narrative.
Someone could object that Stalker is obviously non-analogic because it is a science fiction. We can delineate here the difference between non-analogic narration and science fiction. Stalker is based on a novel by the Strugatsky brothers. In the novel (which can be described as science fiction) the action takes place in the future, and explanations for the existence of the Zones (in the book there are more than one) are provided. It is a science fiction, but our conventional logics can explain that. Tarkovsky, instead, does not explain the setting. We don’t know why the Zone exists, and we little know about what it is. The Russian director is not interested in the reasons or in the possible (logical) explanations. He just takes this situation, this setting, and just sets the story there.
To better define the reasons why we are stating that Stalker is not a science fiction, we will quote narratologist Paul Ricoeur. In his essay Narrative Identity, he asks himself a question quite similar to ours: is it possible to have a narration that is able to disregard our corporal condition of earthly human beings?
Imaginative variations of narrative fictions bear on the variable connection between selfhood and sameness, the imaginative variations of science fiction bear on a single sameness, the sameness of this thing, of this manipulable entity, the brain. An impersonal account of identity thus seems to be dependent on a technological dream in which the brain has from the start been the substitutable equivalent of the person.
The real enigma is whether we are capable of conceiving of alternative possibilities within which corporeity as we know it, or enjoy it or suffer from it, could be taken as a variable, a contingent variable, and without having to transpose our earthly experiences in the very description of the case in question.
For my part, I wonder whether we are not violating something that is more than a rule, or a law, or even a state of affairs, but the existential condition under which there exist rules, laws, facts at all. This violation may be the ultimate reason why these experiments are not only unrealizable, but, were they to be realizable, they ought to be prohibited.
Ricoeur is not positive in his last assertion, but his hypothetical non-human narration (shall we call it this?) is too far from what we are discussing here. What interests us is the section when he talks about science fiction as a narration that does not really break the conventions of “selfhood and sameness”. SF only substitutes individuals with brains, relying on a projection of our future. Technology, in this sense, becomes a device through which we logically imagine a possible future; better say: our possible future, our culture’s. Stalker’s world, instead, does not rely on this kind of logical projection. The movie is not set in our future, but in a parallel path from what is ours; a different dimension, a non-analogic world.
Johnson and Petrie state that Stalker’s Zone “obeys laws of its own that differ from those of everyday reality, which is, after all, one of the premises of both the original story and the film itself”, and further on:
Tarkovsky succeeds in this film in creating a world governed by its own dream logic. [...] Within his framework the everyday world in all his commonplace and often sordid reality is authentically transformed and made strange, so that for two hours and forty-one minutes we live inside it and accept its laws.
What can be objected in this view is that the connection between Stalker and our reality lies in an allegory. Many critics read Stalker as that. Although this view is not embraced, and Tarkovsky himself never admitted that he wanted to refer to the Socialist reality, we can accept the idea that an allegorical intention might be there. But we must be careful not to consider the movie only as an allegory. Doing that would be reducing the broad suggestive and meaningful potential of the text to one single direction. More interesting is seeing a relevance for reality (that is a fundamental element for an artwork) not within an allegorical form, but through a different way. By opening our minds to the possibility of a parallel kind of logic, of a world with different rules than ours, Tarkovsky’s effort seems even more relevant within a reality where ideology still worked as an inescapable logics (Russia in 1979).
3. Inland Empire
I never saw any whole, W-H-O-L-E, I saw plenty of holes, H-O-L-E-S. But I didn’t really worry. I would get an idea for a scene and shoot it, get another idea and shoot that. I didn’t know how they would relate.
This is how David Lynch talks about the way he created his Inland Empire (2006). What can be described, so far, as his most experimental movie from a narrational point of view, is the second example of non-analogic cinema we are going to study.
As much as we said that the non-analogic value of Stalker lies in its shifting along the system axis, we can state that that of Inland Empire lies in the way it plays with the syntagm axis, that is the level of metonymy. Roland Barthes, in his Introduction to Structural Analysis of Narratives, talks about functions as those elements that “involve metonymic relata”: “the purchase of a revolver has for correlate the moment when it will be used”. In Inland Empire, the missed reference with “reality” is not so much in the actions themselves, but rather in the way all the different narrative elements never reveal the (if any) underlying fabula. The sequences and the objects (which are very important in Lynch’s language) all work as narrative functions (in the way Barthes uses this term). In other words, it is not the conventional story anymore that drives the narration: here the narration is driven by the connections that the viewer can make between all the elements. That is why we can say that in Inland Empire there is nothing that cannot be considered as functional: everything seems to have a correlation with another moment of the movie. For example, a table lamp is emphasised in a sequence, as if it was a very important object. The viewer never knows what that lamp means (in a referential way), but the object appears again in other sequences, and grabs the attention because it makes possible a functional connection between the two moments of the movie. The reference of the Signifier (the meaning of the lamp, as well as all of the other objects in the film – and of the film itself) is left open.
But it is possible to argue that in the openness of this movie, in the way we are pushed to figure out the story (to give sense to it all) lies its particular seductive power: as if Lynch considers the “storytelling vocation” an innate feature of the human being. Emphasizing a table lamp means telling the viewer: “this one is really meaningful”. We believe it is very meaningful, but we do not know what the meaning is: that is why it has been said that lynchian movies feature the Signifier without the Signified. But there is another possible view. We can consider the massive amount of signs in Inland Empire as autoreferential: apart from their functional value, they don’t have a clear meaning outside the limits of the movie itself. But they do have a meaning inside the movie: we – as audience – experience it. Only, it is a meaning that cannot be described out of this particular logic and language – a meaning that cannot take place outside this non-euclidean geometry. Opening up the possibility of a different mode of reference is probably one of the major merits we can acknowledge to Lynch.
David Lynch can truly be defined as a postmodern author, since he plays with both narrative customs (as the functional elements described above) and with cinematographic language (as the use of music or zooming-in to underline certain details or moments) employing them in a distorted fashion. To surprise the audience, he relies on it being accustomed to this kind of language. One characteristic of Inland Empire is: the viewer does not know what is going to happen; and not in the long-term, but in a 5-second-interval. This is because the narration does not follow a usual logic, and it is impossible to predict anything. Two devices that Lynch uses to keep the tension at a high level are unexpected sudden noises and sudden zooming-in (both work due to physiological human fears).
To better explain the different modes of escaping the analogon, as employed by Tarkovsky and Lynch, we can focus on the importance of objects in their movies. In Inland Empire, an object works as a functional device because it is the nucleus of a net of potential metonymyc correlations (never openly exploited). In Stalker, instead, the objects avoid analogon because of their use. It is a metaphoric (associative) logic that the character “Stalker” challenges when using dice to get the direction.
Probably the best way to conclude this study is quoting – once again – Roland Barthes, with whom we started. We want to employ in a broader sense his words about literature, from the famous essay The Death of the Author. Today it is necessary to state that not only literature but cultural production in general,
by refusing to assign a “secret”, an ultimate meaning, to the text (and to the world as a text), liberates what may be called an anti-theological activity, an activity that is truly revolutionary since to refuse to fix meaning is, in the end, to refuse God and his hypostases – reason, science, law.
 This is the typology of the sign as distinguished by phenomenological category of its way of denoting the object (set forth in 1867 and many times in later years). This typology emphasizes the different ways in which the sign refers to its object -- the icon by a quality of its own, the index by real connection to its object, and the symbol by a habit or rule for its interpretant. See Peirce, C.S. (1867), "On a New List of Categories", Proceedings of the American Academy of Arts and Sciences 7 (1868), 287–298.
 Non-Euclidean geometry systems differ from Euclidean geometry in that they modify Euclid's fifth postulate, which is also known as the parallel postulate.
 From now on, with the term “reality” is intended the set of values shared by the audience: what we can call both “consensus” or “ideology”.
 “For Saussure, the relationships between linguistic terms can develop on two planes, each of which generates its own particular values; these two planes correspond to two forms of mental activity. [...] Jakobson, in a now famous text, has applied the opposition of the metaphor (of the systematic order) and the metonymy (of the syntagmatic order) to non-linguistic languages”. Barthes (1968).
 In the second example, Inland Empire, we will study a movie that is non-referential in the syntagm axis, on the level of metonymy: in the narrative method.
 Opposite to functions, in Barthes’ terminology, are indices, the integrational units, “not referring to a complementary and consequential act”.
 “The units which have something in common are associated in memory and thus form groups within which various relationships can be found”. Barthes (1968).
Barthes, R. (1968) Elements of Semiology, Jonathan Cape, London.
Barthes, R. (1977) Image, Music, Text. Fontana Press, London.
Barthes, R. (1995) Sul Cinema, Il Melangolo, Genoa.
Baudrillard, J. (1983) Simulations, Semiotext(e), New York.
Johnson, V. T. & Petrie, G. (1994) Andrey Tarkovsky – A visual fugue. Indiana University Press, Indianapolis.
Lim, D. (2006) David Lynch Returns: Expect Moody Conditions, With Surreal Gusts. (online). New York Times. Avaiable from: http://www.nytimes.com/2006/10/01/movies/01lim.html.
David Wood (ed.) On Paul Ricoeur (1991) Routledge, London.
Wollen, P. (1998) Signs and Meaning in the Cinema, British Film Institute, London.