Summer 2010

– Summer 2010

Contextual Essays


Events, Works, Exhibitions

Realism, not Reality: Pedro Costa’s Digital Testimonies

Volker Pantenburg

Tags: Jacques Ranciere, Pedro Costa

Pedro Costa, Casa de Lava, 1995, colour 35mm film, 110min, still. © Pedro Costa

Pedro Costa, Casa de Lava, 1995, colour 35mm film, 110min, still. © Pedro Costa


In the last ten years, the Portuguese director Pedro Costa has established himself firmly in the international film-festival circuit. His films have been shown to critical acclaim in Canada, the USA, Japan and Europe, and a recent retrospective at Tate Modern (in autumn 2009) has given his name further resonance in contemporary art contexts. Those who do not care much about cinema or contemporary art but follow Jacques Rancière's writings have had the chance to come across the director's name on more than one occasion. In Rancière's theoretical framework, Costa plays the role of an upright counterpart to the political endeavours of those artists associated with Relational Aesthetics, a movement that, according to Rancière, lacks integrity and shows how 'the attempt to overcome the inherent tension of a politics of art leads straight to its opposite'.1 In No Quarto da Vanda (In Vanda's Room, 2000), on the contrary, Rancière identifies a force 'that lies in the tensions between the settings of a miserable life and its inherent aesthetic possibilities'.2

At first glance, it seems difficult to reconcile what has become known as the Fontainhas trilogy - Ossos (Bones, 1997), No Quarto da Vanda and Juventude em Marcha (Colossal Youth, 2006), all of them set in the poor Lisbon neighbourhood of Fontainhas - with Costa's beginnings as a director. (The Criterion Collection has recently released this trilogy in a DVD box set titled 'Letters from Fontainhas'.) In terms of production, the turning point in his career comes with Vanda. After Vanda, all of his films, including the documentary on Jean-Marie Straub and Danièle Huillet (Où gît votre sourire enfoui?, or Where Does Your Hidden Smile Lie?, 2001) and his recent portrait of the actress and singer Jeanne Balibar (Ne change rien, 2009) - have been shot with digital cameras, while the preceding three feature films owe themselves to comparatively conventional modes of production. In Costa's case, the shift to digital, low-budget technology is by no means circumstantial, but accompanies and permits decisive changes in his working methods and, particularly, in his conception of realism.


Costa's first film, O Sangue (Blood, 1989), an oneiric black-and-white epic about two adolescent siblings on the run, is a stupendous debut feature. The very first two shots - a boy getting slapped in his face by an adult in the midst of a nondescript field - enact what they depict: they are themselves slaps in the face. The power of these shots results from Costa's rigorous sense of framing and from the abrupt violence that we are thrown into without warning. Yet this power is also due to Costa's strong allusions to film history. If one is familiar with D.W. Griffith's early films for the Biograph Company, one cannot help but feel reminded of the sparse field that the miserable peasant crosses to till his seeds in the anti-capitalist classic A Corner in Wheat (1909).3 And when the father lifts the suitcase and leaves immediately after hitting his son, you are not mistaken to identify this suitcase with the one that Karl Rossmann (Christian Heinisch) forgets on deck of the steamship in Straub/Huillet's Kafka adaptation Klassenverhältnisse (Class Relations) made five years prior to Costa's debut. Costa's film self-consciously inscribes itself into a strong cinematographic tradition, and in retrospect, it seems quite different from the films he made later in Fontainhas, both in aesthetic and economic terms.4

For although O Sangue was a small and independent production, it was made with a team, a script, actors, lighting technicians, an assistant director and so on. Costa felt uncomfortable with the burdens and impositions of a regular film production, but it took him until In Vanda's Room to find an alternative. The two films he made in between, Casa de Lava (1994) and Ossos, were shot in colour, but the basic parameters of production stayed the same: he still operated with a bulky production network, used 35mm stock and worked with a director of photography.

However, between Casa de Lava and Ossos, a different kind of reality began to impose itself. The artificiality that every film production entails is confronted with forces of resistance due to political and economic realities. Casa de Lava is set on the Cape Verde Islands, one of the earliest sites of Portuguese colonialism and a crucial location for transatlantic slave commerce. In Ossos the Fontainhas neighbourhood of Lisbon, a shanty town and home to a large number of immigrants from continental Africa and Cape Verde, slowly becomes the protagonist of the picture. Between the reality of Casa de Lava and the reality of Ossos, Costa himself acts as a messenger. As he has recounted numerous times, he took on the role of an amateur postman:

We were accommodated in a small village, and at the end of the shooting on the Cape Verde Islands, the people gave us letters for their sons, husbands or cousins who had emigrated to Lisbon. They handed us news, presents, tobacco and coffee for friends and family members. When I arrived in Lisbon, I looked for the neighbourhood that is situated a little north of the city - an African and particularly Cape Verdian neighbourhood - and played the postman. I told them: your son is fine, we made a movie together, and since I spoke a bit of Creole, I was accepted very quickly. I was constantly invited for dinner, to a party or wedding and I started staying there, passing the time, talking to people and making observations, drinking with them. This is how, one day, the idea came up to make a movie in this neighbourhood.5

At first sight, this sounds like nothing more than a pretty anecdote. Yet beyond the anecdotal character, Costa's recollection indicates a shift on several levels: from the film- historical references of O Sangue and Casa de Lava to the social reality of Fontainhas, from fiction to documentary,6 from working under 'professional' conditions to working in small communities akin to family contexts. The roots in film history that are so clearly sensed in O Sangue are at the time of making the trilogy complemented by a method that draws its energy from a particular location - Fontainhas - and the non-professional actors that Costa has been collaborating with ever since Casa de Lava, in Ossos, In Vanda's Room and Colossal Youth.


This shift is already perceptible in Casa de Lava. Costa initially wanted to remake Jacques Tourneur's I Walked with a Zombie (1943), an uncanny tale of vodun and obsession set in the West Indies. Casting Edith Scob for one of the female leads added a second strong reference - she had worked regularly with the French film-maker Georges Franju as a young girl, and in particular in his horror classic Les Yeux sans visage (Eyes Without a Face, 1960). Costa's notebook bears witness to his original plans for the film - they show collages directly confronting stills from Franju's film and other film-historical sources with photographs from the Cape Verde Islands.7 Yet the story as it eventually unfolded detached itself from these bonds: the comatose exiled Cape Verdian construction worker Leão (Isaach de Bankolé), who is accompanied to his native island by a young nurse (Inês de Medeiros), has clearly been marked by Portugal's colonial history and its post-colonial echoes. And when one watches the movie, one can sense that the location - the jagged, volcanic hills and exuberant colours of the Cape Verde Islands - as well as the non- professional actors from the local population offered resistance to Costa's scripted ideas and made the film change its direction. 'Casa de Lava may be the film of Costa's that poses the most constant and furious tug of war between Hollywood narrative and the non- narrative portraiture of both places and people, staging an almost epic battle between the two,' Jonathan Rosenbaum writes about the film.8


By the time of In Vanda's Room, this tug of war - to pick up Rosenbaum's metaphor - had definitely been decided in favour of the 'portraiture of both places and people'. Costa's film shows the lives of Vanda and the other inhabitants in between the noises and rubble of the demolition works. The film can be seen as both a subtraction and an expansion. A subtraction, in that Costa works without a script and without 'action' in a conventional sense. He does not use a cameraman and focuses mainly on one location - Vanda's room - which gives the film its title, and acts like centre of gravity, a shelter in the midst of the deafening works of dredges and wrecker's balls tearing down the neighbourhood.9 On the other hand, In Vanda's Room is the result of an immense luxury Costa allowed himself: the luxury of taking his time, spending months and years with the people living in Fontainhas. Both In Vanda's Room and its 'sequel' Colossal Youth only became possible under two conditions: they required that Vanda Duarte and her sister Zita, Ventura (whose surname we never get to know) and the others - the characters in his films as well as the people he lived amongst - accepted Costa not as a film-maker but as a trustworthy fellow inhabitant. And they required the technical means to give testimony to this situation of duration and togetherness without having to involve a large team. This is where digital technology and a particular form of realism come into play: when he started working on the film with Vanda and the others, Costa bought a small low-priced digital camera (a Panasonic DVX 100), and started working with light reflectors rather than additional light sources.10 What kinds of realism does this shift make possible in Costa's work? And how does it relate to traditional discussions of realism?


Traditionally, there have been two distinct (if sometimes overlapping) ways of thinking about realism in the cinema. On the one hand, realism was conceived of as a quality inherent in the photographic character of the medium. Often referred to in terminology building on C.S. Peirce's semiotics as 'indexical realism', it means that the images taken by the camera and stored on the filmstrip are inextricably tied to the objective reality, regardless of what exactly they recorded.11 Siegfried Kracauer's 'redemption of physical reality', Stanley Cavell's 'automatic world projection' or André Bazin's 'ontology of the photographic image' all point to this same quality, no matter which metaphor the process attracted (the image as 'trace,' 'imprint' or 'transsubstantiation' of the real). This model has, of course, been much criticised since the 1960s, when terms such as simulation, construction or deconstruction were given far more credit than the supposedly naïve concept of realism.

On the other hand, realism was used as a label for stylistic features attributed to specific historical film contexts or movements. French 'poetic realism' of the 1930s, Italian 'neorealism' or what some critics recently described as contemporary North American 'neo-neo-realism'.12 In these cases, the label is not meant to describe an a priori capacity inherent in the medium, but a particular way of utilising it: this might imply mobile cameras and smaller teams that make it possible to work without film studios and artificial lighting, or it might imply the prominence of the long shot and stories dealing with marginalised social groups.

I would argue that Costa's films since In Vanda's Room challenge us to replace the ontological question what realism is by an attempt to locate different forms of realism at different points of the production and reception of movies. Without any claim to completeness, I would propose to differentiate four phases of realism, or, more aptly, four phases where effects of realism are produced: (1) There is a form of realism that presents itself as the result of specific means of production. The fewer people involved in making a movie, and the fewer props and technology you need, the greater the chance to capture something 'real' without suffocating it in logistics. Historically, this form of realism has regularly been propelled by ever smaller cameras and sound equipment: 16mm stock and the Nagra tape recorder, digital cameras, etc. (2) There is a realism specific to the photographic process. However, this indexical realism, I would argue, constitutes just one aspect of the whole complex of realism, and it depends much less on the filmstrip and the chemical process of inscribing and storing than is usually understood. Rather, it is much more closely linked to the lenses and the physical transmission of light onto whatever support these are then inscribed onto. I would tentatively call this 'testimonial realism'. (3) A third aspect of realism is produced by disproportionate elements within the narrative. In his famous description of the reality effect in Gustave Flaubert's A Simple Heart (1877), Roland Barthes ascribes this effect to the numerous details in Mme Aubain's room - details that do not have any narrative or psychological purpose but, as he puts it, 'finally say nothing but this: we are the real'.13 If one applies Barthes's concept to film, one would have to identify disproportional elements on different layers of the film: I am thinking of excessively long takes that do not use time for narrative necessities, or of dialogue that does not aim at getting somewhere. A movie such as Andy Warhol's Chelsea Girls (1966) best exemplifies how the absence of an imposed storyline or script leads to a strong feeling that something 'real' is happening in front of the camera. (4) Realism is a mode of perception; it has subjective components on the spectator's side as much as it has objective ones in the apparatus and the technical process of making the film. This aspect was much discussed when Jurassic Park and Forrest Gump took computer-generated imagery to a new level in the 1990s. Yet the term 'perceptual realism', coined in this context, might also be adequate for the time modulations in James Benning's latest movie Ruhr (2009), where he conflated the 60 minutes of 'real time' showing a coke-quenching tower - a mechanism for absorbing gases released in the processing of coal - with the light changes that occurred over 90 minutes.14


Pedro Costa has given a comprehensive definition of realism that combines technical, economical and contextual aspects. 'I'm always within a kind of realism,' he says: Everything feeds into this realism: the way I work, the decision what to spend the money on, which camera you use, which microphone. All that is part of what I call 'realism'. And I know very well that every transgression - of a threshold or a border that I vaguely call realism - makes me risk losing myself and the film. This means: if I have too many people around me, too sophisticated a machine for the location where I'm filming […] I think, we have established a kind of balance between the technical means, the people and the money, all these things. We have established an equilibrium that runs into the film, I think. In any case, something runs from us into the neighbourhood, and from the neighbourhood, a lot runs back into the camera and to us.15

In the light of the aforementioned four phases of realism, I would describe Costa's digital realism as follows: (1) It is explicitly bound to an intimate and collective production process that guarantees a proximity and forms of collaboration that would not be conceivable without a small camera and practically unlimited stock. (For In Vanda's Room, Costa shot more than 150 hours of footage; for Colossal Youth, the material amounted to more than 300 hours.) Digital production thus is the primary precondition for films such as the ones Costa has been making over the last ten years. (2) The integrity of the indexical process has been much contested in digital imaging, yet, as I would argue, in Costa's Fontainhas films the ontological doubt that has infested discussions about the potential manipulations of digital images does not really carry weight; it is more than compensated for by the testimonial powers of the lens and the optical apparatus. No one will have any doubt that Ventura, Vanda, Nhurro or Bete were actually there at the moment of the shooting, in the narrow alleys, between the ruins or in the newly-built quarter that the inhabitants are relocated to.16 It makes no difference whether the images are stored digitally or on filmstrips; nobody will question the fact that In Vanda's Room depicts the present of the derelict houses and improvised sheds of Fontainhas in a specific and very concrete manner. (3) The third facet of realism results from the aesthetic and temporal structures within Costa's films. The time that they depict does not follow narrative concerns. The fixed shots tend to sink in and persist, not flow. When asked about potential models for his way of addressing reality, Costa not only mentions D.W. Griffith and the journalist and photographer Jacob Riis, who photographically documented the lives of 'the other half' of the US population in the 1880s and 90s - the poor and aggrieved - but also Andy Warhol's work as a film-maker. He states that Vanda is a close relative to Edie Sedgwick in Warhol's Beauty #2 (1965), which emphasised the long duration of shots instead of hewing to a scripted dialogue and tightly measured takes. Similarly, in the very first take in Vanda, we witness Zita and Vanda smoking heroin on Vanda's bed over a period of almost five minutes. Vanda's severe coughing and her yawns cannot be subsumed to any psychological or narrative purposes. Rather they inject a strong sense of contingency and bodily presence that rigorously insists on just being there. (4) In both Warhol's and Costa's films, realism is a temporal form of experience that needs a certain extension in time. This realism relies on duration and patient observation, on the side of the director as well as on that of the spectator. This is why Costa has been reluctant to have his work shown in a museum or gallery context and wants his films to be seen in cinemas: 'I'm not a video artist, I am a film-maker and film is a construction. Pieces are made to fit together, if they don't the whole thing will collapse, or worse, will lack movement and tension. Every shot or scene I do depends on the one that comes before and on the one that will come after.'17

Tom Gunning has stressed that in cinema we are dealing with 'realism', not 'reality': 'Theatre, for instance, makes use of real materials, actual people and things, to create a fiction world. Cinema works with images that possess an impression of reality, not its materiality.'18 This distinction has been made repeatedly in the history of film theory, yet it risks being forgotten when people mistake the world on screen for something they can dive into and identify with. Costa is unambiguous about this crucial difference when he - counterintuitively, you might say - describes the ethical task of documentary film-making as a gesture of 'closing the door' between film and spectator. 'We film life, and the more I close the doors, the more I hinder the spectator from taking pleasure in seeing himself on the screen - because I don't want that - the more I close the doors, the more I'm going to have the spectator against me, perhaps against the film, but at least he will be, I hope, uncomfortable and at war. That is, he will be in the uneasy situation of the world.'19 Costa's martial notion of the antagonism between film and spectator situates 'realism' on yet another level: the realities on screen and of the audience are not linked by any direct form of empathy or identification. What unites them is a feeling of uneasiness.

This form of realism, one could sum up, is nothing that is simply there to be taken and consumed. It results from an experience of sharing a life and a time, be it with the human beings in Fontainhas or with their portrait on screen that contains their collaborative testimony. It remains an open question where Costa's endeavour takes him now that one of the pillars of his outstanding form of digital realism, the reality of Fontainhas, no longer exists.

  1. Jacques Rancière, 'Die Politik der Kunst und ihre Paradoxien', (trans. Maria Muhle), Die Aufteilung des Sinnlichen, Berlin: B-Books 2006, p.96. English Translation the author's.

  2. Ibid., p.98.

  3. See Helmut Färber's meticulous study of D.W. Griffith's film in Helmut Färber, A Corner in Wheat von D.W. Griffith: Eine Kritik, Munich and Paris: Verlag Helmut Färber, 1992.

  4. Costa even tried to hire Stanley Cortez, the director of photography of Charles Laughton's classic The Night of the Hunter (1955) for O Sangue. He wrote him a letter without knowing that Cortez was already dead at the time. See Mark Peranson, 'Pedro Costa: An Introduction', Cinema Scope, issue 27, Summer 2006, p.9.

  5. Johanna Bedeau and Mariani Diphy, Passeur du réel: Pedro Costa, radio feature, France Culture 2008 (author's translation). Letters have been a prominent, almost allegorical element in Costa's films since Casa de Lava. In Colossal Youth a letter from the earlier film resurfaces and becomes a central relay connecting the different temporal layers. Rancière has devoted a beautiful essay to Ventura and his letter. See Jacques Rancière, 'La lettre de Ventura', Trafic, issue 61, Spring 2007, pp.7-9.

  6. Costa has talked at length about his particular (if somewhat counterintuitive) understanding of 'documentary' on the occasion of a seminar at the Tokyo Film School. The transcript has been published online by Rouge. See Pedro Costa, 'A Closed Door That Leaves Us Guessing', Rouge, issue 10, 2007, (last accessed on 11 March 2010).

  7. Excerpts from Costa's notebook, mostly collages of texts and images, are to be found in the extras of the French DVD of Casa de Lava.

  8. Jonathan Rosenbaum, 'A Few Eruptions in the House of Lava', 25 June 2008, (last accessed on 11 March 2010).

  9. The expression 'without a script' by no means implies that Costa didn't develop scenes and acting routines with Vanda and the other locals of the neighbourhood. As Costa explains in the illuminating conversation with Cyril Neyrat, the process usually began with an observation of some everyday gesture, dialogue or interaction that Costa would ask Vanda, Zita and the others to repeat several times. See Cyril Neyrat (ed.), No Quarto da Vanda. Conversation with Pedro Costa [DVD and book], Nantes: Capricci, 2008, pp.63-73 (especially the chapter 'Le texte et la répétition').

  10. Costa shot In Vanda's Room and Colossal Youth with a digital camera but transferred the footage onto 35mm film for the final print.

  11. Tom Gunning has recently made an interesting suggestion to reconsider cinematic indexicality in terms of movement rather than in terms of photographic referentiality. See Tom Gunning, 'Moving Away from the Index: Cinema and the Impression of Reality', differences, vol.18, 2007, pp.29-52. See also Philip Rosen, 'Old and New: Image, Indexicality, and Historicity in the Digital Utopia', Change Mummified: Cinema, Historicity, Theory, Minneapolis: University of Minnesota Press, 2001, pp.301-49.

  12. See A.O. Scott, 'Neo-Neo-Realism', The New York Times, 17 March 2009 and Richard Brody, 'About Neo-Neo-Realism', The New Yorker, 20 March 2009. Scott discusses Kelly Reichardt's Wendy and Lucy (2008) and Ramin Bahrani's films Man Push Cart (2005), Chop Shop (2007) and Goodbye Solo (2008); he tries to relate them each to Italian neorealism and a somewhat idiosyncratic choice of what he understands as other historical examples of neorealism.

  13. Roland Barthes, 'The Reality Effect', The Rustle of Language (trans. Richard Howard), Oxford: Blackwell, 1986. p. 148.

  14. See Benning's clarification in the comment section of the 'Rotterdam Film Festival 2010 Diary: Part III', (last accessed on 11 March 2010).

  15. J. Bedeau and M. Diphy, Passeur du réel: Pedro Costa, op. cit. Translation the author's.

  16. In the film we never get to know the surnames of most of the people/actors.

  17. 'From black box to white cube', round-table discussion with Pedro Costa, Catherine David and Chris Dercon (moderator), Jan van Eyck Video Weekend, 26 May 2007, (last accessed on 11 March 2010).

  18. T. Gunning, 'Moving Away from the Index', op. cit., p.44.

  19. P. Costa, 'A Closed Door That Leaves Us Guessing', op. cit.