Limite does not intend to analyse. It shows. It projects itself as a tuning fork, a pitch, a resonance of time itself.

– Mário Peixoto


Discussions about avant-garde films of the 1920s usually relate to European, Soviet or North-American productions, while South American films like Limite (1931), written and directed by Mário Peixoto (1908-1992) and, at least in Brazil, one of the most important cult movies, are hardly ever mentioned in this context, as can be confirmed in publications by Rudolf E. Kuenzli (1987), Sitney P. Adams (1990) or William C. Wees (1992). (1) Classic works on film history frequently do not offer more than very short, general and sometimes incorrect comments on Brazilian films of the 1920s. In his small paragraph on Limite, Peter B. Schumann dates it to the year 1929, two years too early, and describes Peixoto as “a man not yet 20” (2), while he was in fact 23 years-old by the time of its première. Georges Sadoul mentions in his History of Cinema on Brazilian films of the 1920s and ’30s the “remarkable talent of 20 year old Mário Peixoto” (3), without any further comments on the movie itself.

Even though it might be suspected, from a postcolonial point of view, that these obvious disregards represent the typical neglect of non-European or non-American art production through cultural hegemony, they are frequently the result not so much of a political attitude but much more of pragmatic circumstances. The reason for Sadoul’s short commentary on Limite, for instance, may simply be explained by the fact that the French critic did not get a chance to see the film at the time, a difficulty probably shared by many European or American critics to the present day. In the case of Sadoul, who had referred to the film as an “unknown masterpiece”, this is confirmed only in the special Brazilian edition of his book on the History of cinema (1963), where we are told that, in 1960, while visiting Brazil for film research, he had made a trip to Rio de Janeiro just to see the film. (4) Unfortunately, it was a fruitless journey because the deteriorated nitrate film was undergoing restoration, a process that lasted from 1959 to 1978, when Limite was finally screened again. The film has been available on VHS over the past few years, and a DVD version based on another restoration process on the 35mm film is to be released.

In general terms, it needs to be said that the quite mysterious situation regarding the “unknown masterpiece” has changed for the better over recent years. In 1996, the director and producer Walter Salles founded the Mário Peixoto Archive, located within his company, videofilmes, in Rio de Janeiro. There, Saulo Pereira de Mello, one of the restorers of Limite and a former friend of Peixoto, is taking care of an immense collection of manuscripts, scenarios, correspondence, audio tapes, photographic material and published works by and about Peixoto. But since these texts, and among them the original scenario for Limite (Peixoto, 1996), were published in Portuguese, they consequently only reach a very limited public. The present article may therefore be regarded as a small contribution towards better understanding of the film itself, as well as its significance within æsthetic discussions on cinema in Brazil.

This article is structured such that it will first show the historical context that gave rise to Limite, and then cover the reception given to it in the 1960s by Cinema Novo and particularly Glauber Rocha. Finally, the article will look into the relationship between Limite and Walter Salles, today certainly the most successful Brazilian director – Centra do Brasil (Central Station, 1998), Diarios de motocicleta (Motorcycle Diaries, 2004) – and co-producer – Cidade de Deus (City of God, Fernando Meirelles and Kátia Lund, 2002).

The historical context

Media such as photography and film arrived in Brazil quite early, imported mainly by foreign immigrants. The first Daguerreotype had already been reported in 1840 (5) and, in 1898, moving images were shot in the Bay of Guanabara, in Rio de Janeiro. Films with a plot were produced from 1908 onwards, ranging from adaptations of novels such as O Guarani by José de Alencar to police cases like The Stranglers. A more professional film structure was implemented only during the 1930s, within the context of the efforts towards industrialisation made by the Vargas government.

With regard to the artistic scene at the beginning of the 20th century, it needs to be said that the innovative styles and expansionary dynamism of the European avant-garde caused a clear rupture within Brazilian arts. The Modernist movement of the 1920s took cubist or expressionist tendencies as its driving force in the search for a genuinely national, postcolonial modern art style that would incorporate formal innovations coloured by Brazilian ethnic and pictorial qualities. With regard to cinema, many Brazilian film productions of the 1920s went in the same direction, producing the so-called regional cycles that mainly used local costumes, episodes or folkloric figures for shaping their films. This tendency towards centring on a new national culture obviously did not favour films of an avant-garde nature. The films produced frequently displayed greater interest in æsthetic innovation than in ideological references.

Within this historical context, basically four somewhat experimental movies from Brazil or made by Brazilian directors at that time can be cited. Alberto Cavalcanti (1897–1982) filmed Rien Que Les Heures (1926) in Paris, where he had worked as art director with Marcel L’Herbier for his L’Inhumaine (1923). In 1929, Rodolfo Lustig and Adalberto Kemeny tried a Brazilian remake of Walter Ruttmann’s Berlin: Die Sinfonie der Großstadt (Berlin Symphony) from 1927 called São Paulo – Sinfonia da metrópole. However, due to the lack of rigorous rhythmic structure and constant interruptions by intertitles praising the industrial growth of the city, as well as an involuntary humorous imitation of Fritz Lang’s Metropolis (1927) at the end, it was not a very convincing production. This was clearly seen in a contemporary review that considered it, despite “some original images, well photographed and pleasant to watch”, to be a pitiable copy of the original Ruttmann film. “As a documentary it does not pay off. As a rhythmic film it is even worse.” (6)

Last, we have the two films that have shown themselves to be the most relevant for Brazilian cinema history. Ganga Bruta, by Humberto Mauro (1933), is a mixture of different genres, partially talkie and partially silent, and it would later on be the overall reference for Glauber Rocha and the Cinema Novo in their attempt to redefine a national cinema history. This film was qualified by Sadoul as “one of the most important films of the universal cinema” (7). And we have Limite, by Mário Peixoto, which was filmed mainly in 1930 and first screened in 1931. This has, over the past 70 years, become a legendary cult movie, voted several times as one of the best Brazilian films of all time, and credited as one of the main inspirations for contemporary filmmaker Walter Salles.

The making of Limite

Just like many Brazilian artists of his time, Mário Peixoto received important artistic stimuli from Europe. His stay in England at the Hopedene School in Willingdon, near Eastbourne, Sussex, in 1926-7, at the age of 19, evoked his first inclination towards acting, developed his strong appreciation for cinema, especially for Russian and German movies, and probably led to his first experimentation with his homosexuality. With regard to cinema, Peixoto particularly admired the work of directors such as Fritz Lang, Georg Wilhelm Pabst, F. W. Murnau, Ernst Lubitsch, Sergei Eisenstein and Vsevolod Pudovkin. He would return to Europe in 1929 with the express intention of seeing the latest cinema productions that were unavailable in Brazil.

Photo by André Kertesz

Another source of inspiration was his contact with the local art scene in Rio de Janeiro, such as with the cameraman Edgar Brazil, the director Adhemar Gonzaga – Peixoto assisted in the shooting of one of his films, Barro Humano (Human Clay), in 1927 – and the critic and writer Octávio de Farias. The latter was a member of the Chaplin Club, a loose circle of friends founded in 1928 that, until 1930, published a magazine called The Fan, which was dedicated to debates on the æsthetics of silent cinema.


According to Peixoto, he got his final inspiration for Limite in August 1929, on his second trip to Europe. While walking through Paris, he saw a photograph by André Kertesz in the 74th issue of the French magazine, VU, a magazine that other famous photographers like Man Ray had also been working for. It was this picture that led to the writing of the scenario for Limite, which was published for the first time only in 1996. The image of a woman embraced by a man in handcuffs returned into the film in the opening and ending sequences as a prototype-image.

The scenario with its 220 listed shots shows itself to be a very explicit manual with detailed descriptions of camera positions, angles and movements for cameraman Edgar Brazil to use. The final cut of Peixoto’s film sticks very closely to the scenario.

Shot 73 might serve as an example:

fusion close up – hand of the woman who has fish and some vegetables in her basket – camera follows her and, once again, close up showing the basket and all of her purchases – woman keeps on walking – camera moves with her (8)

In comparison with the scenarios of other silent avant-garde movies of the 1920s, for instance Man Ray’s manuscript for L’étoile de mer (1928) (9), or even the script by Carl Mayer and Hans Janowitz for Das Cabinet des Dr. Caligari (Robert Wiene, 1920), it must be said that Peixoto’s text does not tell a story, nor does it give insights into any kind of psychological state of mind among the three main characters. Rather, it “thinks” in pictures, movements and angles, trying to intertwine the diverse visual fields by using certain symbolic themes and variations. From the outset, the filmic style of Limite is part of the scenario and not a result of an interpretation or transformation of the textual outline by subsequent shooting. The metaphor of the “camera brain” – a frequent term used by many avant-garde filmmakers – is also present in Peixoto’s scenario, in which the use of intertitles is avoided, with one short exception, and reliance is placed overall on the camera and its movements. Limite therefore accomplishes what Germaine Dulac had demanded in 1927: the “real” filmmaker should “divest cinema of all elements not particular to it, to seek its true essence in the consciousness of movement and visual rhythms” (10).

Taking in account the scenario as well as the actual movie, Limite must be seen as a film with a clear, elaborated and recognisable concept. This may explain Peixoto’s dislike of surrealistic movies, specifically those of Luis Buñuel, and the rejection of chance as an artistic principle, as found in Man Ray or Dada. Limite starts off with the image of a woman embraced by a man in handcuffs, a prototype image that goes on being modified throughout the film. The opening proto-image, from the photograph he saw in Paris in 1929, introduces the leitmotiv of imprisonment, of being trapped, and gives way to a long, almost hypnotic boat scene that is to transport us into the continuum of time, a rather fluid amorphous state in which the camera then moves into the past, tracing certain memory lines, episodes and associated details, objects, movements and images. These visual flashes of limitations are reflected in other images and thus escape from their fixed, limited and solid status, only to disappear or fade out without further explanation. The wrecking in the storm at the end then leads us back to the original proto-image, the initial theme, now extended and enriched by the visual and rhythmic variations that have been experienced. The scenario and film can therefore best be characterised as a visual cinematic poem that explores the medium for its poetic capacities, instead of using it for transporting non-visual conceptions and narratives.

Peixoto then offered the scenario to his director friends Gonzaga and Mauro. But both of them declined and advised him to make the film himself and to hire the cameraman Edgar Brazil, who would have the necessary experience to ensure completion of the project. Shooting began in mid-1930, using imported panchromatic film material with a high sensitivity for grey scales.

Limite had its première on 17 May 1931, in the Cinema Capitólio in Rio de Janeiro, in a session organised by the Chaplin Club. It received favourable reviews from the critics, who saw the film as an original Brazilian avant-garde production, but it was also rejected by part of the audience and never made it into commercial circuits. Over the years, it was screened only sporadically, as in 1942, when a special session was arranged for Orson Wells, who was in South America for the shooting of his unfinished It’s all True, and for Maria Falconetti, lead actress of Carl Theodor Dreyer’s La Passion de Jeanne d’Arc (The Passion of Joan of Arc, 1928).

Limite remained the only film Peixoto succeeded in finishing, even though a number of other projects were discussed or even initiated up until the 1960s. His literary ambitions, which went as far back as 1930 and included poetry, short stories, theatrical plays and a six-volume novel with strong autobiographical traits called O inútil de cada um (The uselessness of everyone) and on which he worked obsessively almost until the end of his life, also did not gain a wider response among the public. So far, only the first volume of this novel has been published (in 1984), while the remaining volumes are being prepared for publication by the Mário Peixoto Archives.

Because of financial problems, Peixoto had to sell most of the property he inherited from his wealthy family later on in his life, and he moved into a small hotel. His final years were spent in a small flat in Copacabana, where he died in 1992. He had only survived a severe illness in 1991 because of financial support from Walter Salles, who not only declared Limite to be one of the main inspirations for his career, but also included several direct references to Peixoto in his movie Abril Despedaçado (Behind the Sun, 2001), on which I will comment below.

The reception given to Limite

The reception given to Limite has been partially influenced by certain myths surrounding the movie. As mentioned earlier, due to the lengthy restoration process, it disappeared for almost 20 years, and there was speculation that the film had actually never existed. The fact is that, in 1959, the nitrate film began to deteriorate and two dedicated admirers, Plinio Süssekind and Saulo Pereira de Mello, started a frame-by-frame restoration of the last existing negative. Limite only returned to festivals and screenings in 1978. Even though hardly anybody could see the movie between 1959 and 1978 – as in the case of Sadoul and his unsuccessful trip to Rio de Janeiro in 1960 – it still served as a reference for controversial discussions and statements. In 1963, Glauber Rocha, a leading figure within the “new cinema”, the Cinema Novo, described Peixoto as “far from reality and history” (11) and the unseen movie as “unable to comprehend the contradictions of bourgeois society” (12), and a “contradiction historically overcome” (13), only to confirm his judgement of Limite as a product of the intellectually decadent bourgeoisie again in 1978, after finally having seen it. Even though Cinema Novo and Limite do share common grounds with regard to low-cost production, financed partially by the actors, directors and producers involved in the respective project, and similar concepts of camera movements and angles can even be found overall, with regard to the use of a “untied” free-moving handheld camera as an important filmic element, Rocha and his colleagues did not merely have the intention of creating an æsthetic revolution within the national film scene. In his manifesto, “Aesthetics of hunger”, from 1965, he made it clear that the rejection of colonial, exotic and primitive views about Brazil that misinterpreted the social reality and contributed towards its present-day misery was the main objective of his artistic production. Cinema Novo “intended to show the violence of hunger through appropriate aesthetics of violence” (14), thereby replacing tropical clichés by images of poverty in all its aspects: landscapes, dialogues and lightning, or showing “people eating dirt, people killing to eat, people running away to eat, and dirty ugly filthy characters” (15).

It is quite clear that Limite, with its poetic, introverted and hermetic conception, must have presented quite the opposite of the radical, political and rebellious approach adopted by Rocha and others. This becomes even clearer if we compare Rocha’s terms with Peixoto’s comments on his own film, expressed in a text called “A Movie from South America”, which was for a long time attributed to Sergei Eisenstein, as for instance in Robert Stam’s chapter on the “Brazilian Avant-Garde cinema” (16). But it was Peixoto himself who created this legendary article, while attributing its authorship to the Russian filmmaker. Peixoto at first said he had translated the article from a French version of the original English text written by Eisenstein, and later on claimed that the cameraman Edgar Brazil had translated it from German into Portuguese. However, according to Saulo Pereira de Mello, Peixoto finally admitted to having written it himself. The article was then republished by Mello (2000) as an original text from Mário Peixoto. In this article, Peixoto first emphasised the role of the “camera-brain” and the instinctive rhythmic film-structure of Limite. According to the director, his film was “as meticulously precise as the invisible wheels of a clock” (17), in which long shots are surrounded and linked by shorter ones, as in a planetary system. Peixoto characterised Limite as a desperate scream aiming for resonance instead of comprehension. “The movie does not intend to analyse. It shows. It projects itself as a tuning fork, a pitch, a resonance of time itself” (18), thereby capturing the flow between past and present, object details and contingence. This short extract may serve to illustrate the immense gap of programmatic intentions between Peixoto and Rocha, a gap that also existed on a personal level, since the reserved, English-educated Peixoto simply could not stand Rocha’s rebellious and provocative attitudes on the rare occasions when they met.

Poster for Ganga Bruta

Much more than a follower of Peixoto’s Limite, Rocha saw himself in the tradition of Humberto Mauro and his film Ganga Bruta.

In his article, “Humberto Mauro and the historical situation” (19), Rocha compares the two films:

If Limite corresponds without a doubt to the French avant-garde cinema and reveals an artist full of subjective aestheticism, Ganga Bruta not only corresponds to the poetic vein of filmmakers like [Jean] Vigo or [Robert J.] Flaherty but is also implemented within a situation that did not limit Mauro to pretentious language. […] He had before him the landscape of Minas Gerais and inside of him the vision of a filmmaker educated by sensitivity, intelligence and courage. (20)

What may have most attracted Rocha towards Mauro’s Ganga Bruta was the combination of rebellious dissonance in the montage: it starts off as an expressionist film, then shifts into a more documentary style, followed by a Western sequence, just to end on a more melodramatic note. According to Rocha, this was combined with a “comprehension of the objective values of the physical and social landscape” (21).

Differing from the more eclectic art of Peixoto, Cinema Novo was permeated by a combination of different cultural levels: popular and high culture. In his films, Glauber Rocha worked tirelessly on the blending of popular and high art in music and literature. The literature of Guimarães Rosa and Euclides da Cunha was mixed with elements of the Cordel tradition. Heitor Villa-Lobos and J. S. Bach were juxtaposed with the popular songs of the arid northeastern region. The disregard for Limite and the praising of Ganga Bruta must therefore be seen as part of the desire within Cinema Novo to find an adequate national filmic language that could, at least to some extend, represent an authentic Brazilian identity that would be outside of European or American patterns. Even though neither Peixoto nor Rocha was able to win a wider audience at the time, it is incontestable that Rocha succeeded in giving Brazil an internationally recognised “face” through his films and has marked out the country on the universal cinematic map.

Walter Hugo Khouri and actress Odete Lara: 1964

Other interesting productions from the time of Cinema Novo nowadays have a hard time emerging from its dominating shadow. Peixoto, for instance, preferred films like Noite vazia (Empty Night) from 1964, directed by Walter Hugo Khouri, whose movies might be seen as a filmic counterweight to Cinema Novo in the 1960s, showing intimate, existential conflicts in mainly urban environments, underlined by cool photography in clean geometric architecture and with jazzy music. Instead of the rural settings of Cinema Novo, Khouri’s films show urban middle-class people in search of some kind of transcendental experience, driven by an underlying erotic tension.

If Peixoto and Mauro were to divide many filmmakers of the 1960s with their highly political attitudes and concerns, it is interesting to observe how the later generation of filmmakers in the 1990s, such as Walter Salles, have had no difficulty in integrating the two tendencies: the æsthetic experiments of films like Limite and the efforts towards “Brazilianness” by Rocha and others. Salles not only frequently reaffirms his desire to pay homage to Cinema Novo directors, such as Nelson Pereira dos Santos or Glauber Rocha, but also, as I pointed out at the beginning, founded the Mário Peixoto Archives, thus preserving the memory of this unique figure in Brazilian film history. With regard to plurality as a culture principle, especially for Latin America, Salles believes,

that there is not just one Latin American cinema, just as there is no single Brazilian cinema. There are cinemas, made of sometimes-contradictory currents that often collide, yet which come together in a desire to portray our realities in an urgent and visceral manner. We make films that are, like the melting-pot that characterises our cultures, impure, imperfect and plural. [...] Unlike Europe, we are societies in which the question of identity has not yet crystallised. It is perhaps for this reason that we have such a need for cinema, so that we can see ourselves in the many conflicting mirrors that reflect us. (22)

Walter Salles: an introduction

The success of Walter Salles’ films outside of Brazil is certainly due to his unique combination of regional landscapes, atmospheres and faces with narratives of universal appeal, along with a filmic sensitivity of international standard. This, of course, is a point frequently held against him. In an allusion to Glauber Rocha’s “Aesthetics of hunger”, critics like Ivana Bentes tend to label films like Central Station as “cosmetics of hunger” (23), with the accusation that Salles and other contemporary directors are using the sertão, the arid north-eastern region of Brazil, only to produce folkloric mass spectacles:

These classic films of the 1960s created aesthetics based on the dry cut, the nervous framing, the overexposure, the handheld camera and the fragmented narrative that mirrored the cruelty of the sertão. These are the Cinema Novo aesthetics, whose purpose was to avoid turning misery into folklore. Those films proposed ethics and aesthetics for the images of pain and revolt. However, the idea rejected by those films, of expressing the intolerable through beautiful landscapes, thus glamorising poverty, emerges in some contemporary films, in which conventional language and cinematography turn the sertão into a garden or a museum of exoticism, thus ‘rescuing’ it through spectacle. (24)

Salles himself sees in this kind of criticism only another form of colonial attitude that tries to impose on filmmakers from Latin America a certain filmic style and themes they are supposed to deal with. And clear biographic reasons can be found for the apparent tension between the fascination for European or American cultural innovations and the wish for a Latin American identity that underlies his career. Born in 1956 and coming from a wealthy family – his father was a banker and diplomat in various countries – Salles lived for several years in Paris, studied economics in Rio de Janeiro, and then took up audiovisuals at a university in California. When he was watching Cuban director Tomás Gutiérrez Alea’s Memorias del subdesarrollo (Memories of Underdevelopment, 1968) for the first time, it was, Walter Salles now admits,

Like a shock to me [...] having spent part of my childhood in Europe, I had better knowledge of Italian neo-realism and the French new wave, than I did of the cinema from my own land. (25)

In many ways, his subsequent career may be understood as an attempt to correct this oversight.

Salles, recently nominated as one of the most influential people in Brazilian cultural life by the magazine Veja and, in 2004, ranked 23rd on The Guardian’s list of the 40 Best Directors In The World, is probably best known internationally for two movies: Central Station and The Motorcycle Diaries.

Central Station, from 1998, is the story of a journey undertaken by a former schoolteacher, now a professional letter-writer for illiterate people, and a young boy attempting to locate a father he never knew. It is a geographical quest that goes from the metropolis of São Paulo up to the poor rural regions of the northeast and which is, of course, a search on different levels: the quest for a personal relationship between the former teacher and the boy, a quest by the boy for his father, and also a quest for what might be called the “fatherland”. At the Berlin Film Festival in 1998, the film and the actress Fernanda Montenegro received the main awards. Central Station was elected Best Foreign Film by the British Academy Awards and the Golden Globes, not to mention numerous Latin American distinctions. Altogether, Central Station has won more than 50 international awards and has been seen by more than 7 million viewers, including 1.6 million in Brazil.

More recently, The Motorcycle Diaries, another road-movie by Salles, about the journey of a young Che Guevara (Gael García Bernal) through Latin America, accompanied by friend Alberto Granado (Rodrigo de la Serna), has been seen by more than 10 million viewers and has been awarded or nominated for 40 different prizes. But even though internationally established, the sensitivity of Salles regarding his Latin American roots became clear at the last Oscar ceremony in March 2005. Jorge Drexler, the Uruguayan composer and singer of the nominated Diaries song, “Al Otro Lado Del Rio”, was banned by the network responsible for the transmission of the ceremony. Instead, they contracted Antonio Banderas and Carlos Santana for the presentation, a disregard that resulted in a harsh reaction by Salles and which clearly shows his very determined position regarding cultural domination. A declaration signed by the entire film crew expressed their

dissatisfaction with what seems to be morally unacceptable […] and which shows a complete disinterest for the diverse cultural currencies/currents/matrixes of our continent. The decision, based on market values only, without any concern for the cultural aspects involved [...] ironically disrespects one of the artistic features of the film that is supposedly being honoured. (26)

As mentioned earlier, both road-movies, Central Station and The Motorcycle Diaries, are about a search for a double identity: by one of the protagonists as an individual, and also this person’s fate within the context of the Brazilian or Latin American geopolitical situation and history. It was certainly this aspect that was decisive for the screening of the Diaries on occasions like the World Social Forum of 2005, the alternative antipode to the meeting of world leaders and celebrities in Davos. Not only was Salles present at the Forum, to accompany the screening of the film, but also his production company, videofilmes, founded in 1985, was participating with two more documentaries: Entreatos (Between Acts, 2004), made by his brother João Moreira Salles, about the election campaign of President Luiz Inácio Lula da Silva, and Peões (Workers, 2004), by Eduardo Coutinho, about Lula’s time as a militant union leader at the end of the 1970s.

With regard to the Diaries, it is not only the movie itself, but the whole five-year-long process of making it that has become quite a significant biographical milestone for Salles. He has repeatedly stated that filming the Diaries signified an overwhelming experience of solidarity that profoundly changed his way of working, which until then had been a detailed, planned process. This gave way to intuition, improvisation and concomitant drafting of the script and accomplishment of the filming, bringing together all the participants in this project, from countries like Argentina, Chile, Peru, Mexico and Brazil. Salles therefore understands his film as an expression of an artistic-political attitude that was developed in a co-operative workstyle and not a result of outside impositions.

But all this does not mean that one can reduce Salles’ films and position to a simplified North-South conflict. The fact, for instance, that, in economic terms, Central Station and The Diaries were only possible through the collaboration with Robert Redford and the Sundance Festival, as well as through the association with Swiss producer Arthur Cohn, speaks for itself. Central Station received an award for best script from the Sundance Institute, and also a payment by NHK for Japanese television broadcast, and thus obtained the financial basis for carrying out the project. The filming of Che Guevara’s early Diaries had been a long-time project of Redford’s, and he saw in Salles an adequate partner and director.

In general terms, Salles’ activities as a director over the past twenty years have shown a wide range of intercultural interests, including videos on popular song writers from Brazil. Starting with productions for Brazilian television, his first works include detailed studies on Japan and China, as well as portraits of Brazilian artists like Tomie Ohtake or Rubens Gerchman, or interviews with famous film directors or actors such as Federico Fellini and Marcelo Mastroianni. In 1987, he made another documentary for television that would outline his further career: the story of Frans Krajcberg (Krajcberg – the poet of the remains), based on a script by his brother João Moreira Salles. In the story, Krajcberg, a Pole whose family is exterminated during the Nazi occupation, escapes to Russia where he joins the Red Army. Then, in 1940, he studies art in Leningrad, but is soon drafted into the Second Polish Army. At the end of the war, at the age of 24, without any surviving family and burned out, he first goes to Germany to study with Willi Baumeister, then to Fernand Léger in Paris and finally, in 1948, arrives in Brazil, where he becomes one of the most recognised eco-artists in the 1960s, protesting through his sculptures and installations against the destruction of natural resources. A newspaper article on the artist and his work with burned wood, giving it a second birth, finds its way to Socorro Nobre, a 36-year-old woman in prison. She then writes a letter to the sculptor telling him how moved she was by his artistic work, and that it gave her new strength and hope for her own life. This story, as well as a meeting between the two of them, is shown in a short film by Salles called Socorro Nobre from 1995. The idea that something as prosaic and simple as a letter could change a whole life stayed in Salles’ mind and formed the basic outline for Central Station. The script then won the award offered by the Sundance Festival commemorating 100 years of cinema and the project got underway, with Socorro Nobre playing a minor role in the opening sequence.

But Salles’ first cinema appearance dates back to 1991, in a Brazilian-American co-production with English dialogue. The film, A Grande Arte (High Art), is based on a novel by ex-police officer Rubem Fonseca, and is a cynical crime story with infinite cultural allusions. Even though the film offers some convincing scenes, it should be regarded as a somewhat ambiguous first attempt at adapting a work of fiction for the cinema, which does not do justice to the far richer novel. And before filming Central Station, Salles also co-directed several musical films on singers and songwriters such as Tom Jobim, Caetano Veloso and Chico Buarque. In 1995, he directed the low-budget film, Terra Estrangeira (Foreign Land, co-directed by Daniela Thomas), filmed in black and white with a handheld camera and a small crew. This re-established Salles’ pleasure in filming works of fiction after the somewhat disappointing experience of High Art. The movie, which includes filmic references to Alain Tanner’s Dans la Ville Blanche (In the White City, 1983) and Wim Wenders’ Alice in den Städten (Alice in the Cities, 1974) and Der Amerikanische Freund (The American Friend, 1977), as well as to the visual style of Cinema Novo (27), tells the story of Alex (Fernanda Torres) and Paco (Fernando Alves Pinto), who leave Brazil during the government of President Collor and its catastrophic financial policies, and try to build a new existence in Portugal, Brazil’s former colonial master, where they get involved in illegal activities.

The international success of Central Station then resulted in invitations for projects such as the series 2000 Seen By, created by German-French television channel Arte in order to collect different views concerning the end of the millennium, including contributions from Hal Hartley and Miguel Albaladejo. Salles, once again joined by Daniela Thomas, participated with the episode, “O Primeiro Dia” (“Midnight”, 1998), a kaleidoscopic narrative of different characters in Rio de Janeiro at the dawn of the 21st century.

Poster for At the edge of the earth

Salles and Peixoto

Salles’ next film was released internationally as Behind the Sun (2001) but was originally entitled Abril despedaçado or Broken April, maintaining the title of the novel by Albanian writer Ismail Kadaré that inspired the movie. If certain visual connections to Mário Peixoto’s Limite can already be found in Socorro Nobre, since both deal with bars, limitations and imprisonment, it is here that Salles included several more explicit references to Peixoto. Sergio Machado, a long-time assistant to Salles, even characterised the movie as an explicit homage to Limite, recognisable in the dialogues as well as in the montage and conception of camera plans (28). In the same year of 2001, Machado also released a prize-winning documentary on Mário Peixoto, produced by Salles’ videofilmes, called Onde a terra acaba (At the edge of the earth), which was the title of one of Peixoto’s unfinished projects.

The novel by Kadaré, presented to Salles by brother João Moreira Salles, tells the story of two rival Albanian families bound by the century-old rules of the Kanun, an ancient code based on a “blood for blood” principle that was mainly carried out in distant regions where the state was weak and institutions did not function. Walter Salles transposed this story to the setting of the arid northeast of Brazil in 1910, in which similar vendetta structures existed. This adaptation was praised by Kadaré himself as a very successful one. Literature on the theme was consulted; for instance, publications like Lutas de familia no Brasil (Luís Aguiar Costa Pinto, 1949), and The history of a family and a community in the Northeast of Brazil by Billy Jayes Chandler (1972).

But differing from his previous productions, Foreign Land, Central Station and Midnight, which showed urban violence and social contrasts in a very contemporary and up-to-date dimension, Behind the Sun, which was located in a primitive sugar cane mill in a deserted, arid, rural environment, took on a more extemporised, ancient drama-like proportion. This was mainly due to conversations between Salles and Kadaré, in which the writer emphasised the roots of his story in classical literature, arguing that blood feuds were at the root of Greek tragedy, such as in Aeschylus. The results of this research were then presented to the cast during the weeks of preparation in loco, incorporating the life of hard labour under real conditions. Visual inspiration was found in the paintings of Brazil by the German artist Hildebrandt (1818-69) and, regarding colours, it was determined that the only red-tinted cloth in the film was to be the shirt with its bloodstain. Another source of inspiration came from film tradition itself. Movies like Grigori Aleksandrov and Sergei M. Eisenstein’s Staroye I novoye (The General Line, 1929) and Peixoto’s Limite were shown and analysed in relation to their expressive force, driven by the montage as an important filmic principle. With regard to Peixoto’s relevance to Behind the Sun, firstly Salles chose Breves as the family name of Tonho (Rodrigo Santoro), which is part of Peixoto’s complete name: Mário Breves Peixoto. A second reference can be seen in a small sequence with the patriarch that was directly inspired by a scene from Limite. But the main dialogue between Behind the Sun and Limite is certainly related to the question of time and clockwork as its symbol. Both were central themes in Limite as well as in Peixoto’s literary work, above all in his six-volume novel O inútil de cada um.

Behind the Sun - The clockwork
Behind the Sun - The pendulum

As mentioned in the first part of this paper, Limite is a film that oscillates between a fluid memory stream and solid, concrete objects and episodes, which emerge as fixed points in the continuity of time. The limitations and the feeling of being trapped in time that we experience in Limite are also present in Peixoto’s novel. In a key scene of the book, the moving of a clock handle is commented on as follows: “every time the clocks counts one more, it is actually saying one less” (29), not showing a progression but rather the vanishing remains of time. The same phrases were also used by Peixoto during a first encounter in 1990 with Salles, who then inserted them into Behind the Sun. And, of course, they fit well into the story of Tonho, who is condemned to a short survival time once he has carried out the revenge. The idea of an archaic, all-imposing clock moving relentlessly and grinding down all existence is well presented in Behind the Sun by the image of the sugar mill, with its visual resemblance to a real clockwork mechanism, moved endlessly in circles by the oxen that even continue their rounds after they have been unyoked, and also the motion of a swing that imitates the movements of a clock handle.

With regard to the present, it is worth noting that Salles recently completed his first Hollywood assignment, directing the US remake of Hideo Nakata’s supernatural thriller, Honogurai mizu no soko kara (Dark Water, 2002), based on the novel by Suzuki Kôji, who also wrote Ringu (The Ring). He has also finished his episode for Paris, je t’aime, a project where twenty international filmmakers each have five minutes to portray a district in Paris (Salles’ is the 16th). These single episodes are then to be intertwined as an atmospheric story. And, once again in collaboration with Daniela Thomas, Salles is working on a Brazilian film about four brothers in a suburban area of São Paulo, in which each of them is trying to escape the social environment in a singular, peculiar way. As to the international context, among his next projects is the filming of Jack Kerouac’s novel, On the Road. The movie rights have been with Francis Ford Coppola since 1979.

Also consulted

Billy J. Chandler, The Feitosas and the Sertão Inhamuns: The History of a Family and a Community in Northeast Brazil 1700–1930 (Gainesville: University of Florida Press, 1972).

Carl Mayer and Hans Janowitz, Das Cabinet des Dr. Caligari (München: edition text und kritik, 1995).

Mário Peixoto, O inútil de cada um (Rio de Janeiro: Record, 1984).

Luís Aguiar Costa Pinto, Lutas de Família no Brasil (São Paulo: Companhia Editora Nacional, 1949).


  1. Rudolf E. Kuenzli, Dada and Surrealist Film (New York: Willis Locker & Owens, 1987); Sitney P. Adams, Modernist Montage: The Obscurity of Vision in Cinema and Literature (New York: Columbia University Press, 1990); and William C. Wees, Light Moving in Time: studies in the visual aesthetics of avant-garde film (Berkeley and Los Angeles: University of California Press, 1992).
  2. Peter B. Schumann, Handbuch des brasilianischen Films (Frankfurt- München: Vervuert, 1988), p. 17.
  3. Georges Sadoul, Historia do cinema mundial, das origens aos nossos dias (Lisboa: Livros Horizonte, 1983), p. 409.
  4. Georges Sadoul, História do cinema mundial (São Paulo: Martins, 1963).
  5. Pedro Karp Vasquez, A fotografia no Imperio (Rio de Janeiro: Jorge Zahar Editor, 2002), p. 8.
  6. Maria Inez Machado Borges Pinto, “O cinematógrapho e a ilusão espetacular da São Paulo moderna”, available at ProQuest (14 June 2002), p. 3.
  7. Sadoul, 1963, p. 502.
  8. Mário Peixoto, Limite “scenario” original (Rio de Janeiro: Sette Letras, 1996), p. 59.
  9. Kuenzli, pp. 207-19.
  10. Germaine Dulac, “The Aesthetics. The Obstacles. Integral Cinegraphie”, Framework 19, 1982, p. 9.
  11. Glauber Rocha, Revisão Crítica do Cinema Brasileiro (São Paulo: Cosac & Naify, 2001), p. 59.
  12. Rocha, p. 66.
  13. Rocha, p. 67.
  14. Ute Hermanns, Schreiben als Ausweg, Filmen als Lösung? (Frankfurt- München: Vervuert Verlag, 1993), p. 62.
  15. Glauber Rocha, “Uma Estética da Fome”, available from ProQuest (12 March 2005), p. 2.
  16. Randal Johnson and Robert Stam (Eds), “On the Margins: Brazilian Avant-Garde Cinema”, in Brazilian Cinema (East Brunswick, New Jersey: Associated University Presses, 1982), pp. 306-327.
  17. Saulo Pereira de Mello, Peixoto – Escritos sobre cinema (Rio de Janeiro: aeroplane, 2000), p. 88.
  18. Mello, p. 89.
  19. Rocha, 2001, pp. 43-55.
  20. Ibid, pp. 44-5.
  21. Ibid, p. 53.
  22. Walter Salles, “I have seen the light”, available from ProQuest (14 April 2004), p. 3.
  23. Ivana Bentes, “The sertão and the favela in contemporary Brazilian film”, in Lucia Nagib (Ed.), The New Brazilian Cinema (London and New York: Tauris, 2003), pp. 118-35.
  24. Nagib, p. 124.
  25. Salles, p. 5.
  26. Walter Salles, “Walter Salles protesta contra decisão do Oscar”, available from ProQuest (2 February 2005), p. 2.
  27. Walter Carvalho, Terra Estrangeira (Rio de Janeiro: Relume Dumará, 1997), p. 8.
  28. Sérgio Machado, “Sérgio Machado recupera memória de Mário Peixoto”, available from ProQuest (29 June 2002), p. 3.
  29. Peixoto, 1984, p. 372.

About The Author

Michael Korfmann is Professor of Comparative Studies at the Federal University of Rio Grande do Sul at Porto Alegre, Brazil. He is currently editing a book on Mário Peixoto and wishes to thank CNPq for a scholarship.