Cannibals at the Carnival: Cinema Novo and Marginal Cinema

Glauber Rocha shooting "Land in Anguish", 1967.

Glauber Rocha shooting “Land in Anguish”, 1967

The Brazilian Cinema Novo is widely regarded as a staple feature in the history of avant-garde cinema. Throughout the 1960’s and early 70’s, Novo filmmakers participated in a broader movement of cultural transformation that was entwined with the politics of ‘modernising’ and ‘developing’ the nation. This period constitutes the birth of modern Brazilian cinema, the success of which “is the result of many years of struggle by filmmakers determined to create a strong film industry in that country” (Johnson 1984: 1). The struggle was bound up with the socio-political landscape in which Novo was situated. What began as a decade of popular optimism in Brazil gave way to military dictatorship and institutionalised repression. In response, Cinema Novo became increasingly radical, evolving “through a series of discernible phases, each of which corresponds to a specific sociopolitical conjuncture” (ibid). The purpose of this essay is not to examine the intricate changes of the form and content of Novo across these phases. Rather, it is to chart the centrifugal direction of it in relation to cultural traditions within Brazil. I call this direction a centrifugal one because, as well as corresponding to specific sociopolitical conjunctures, it represents a movement away from centres of power towards the marginalised elements of society. Conducive with this movement is the tendency towards cannibalism, a specific trope that falls under the broader concept of the carnivalesque. I will argue, therefore, that the movement from Cinema Novo to Marginal Cinema represents a shift from carnival to specifically cannibal strategies. The dialectical nature of this shift should be noted, as it stems from increasing political repression and is bound up with a nationalistic yet ‘tri-continental’ counter-hegemonic position. As Paul Willemen notes “a cinema that seeks to engage with the questions of national specificity from a critical or counterhegemonic position is by definition a minority and a poor cinema, dependent on the existence of a larger multinational or nationalised industrial sector” (Willemen 28: 2006).

The foundation of my analyses are the cultural theories of Mikhail Bakhtin. While Bakhtin never wrote about cinema or Latin America, his theories on the carnivalesque have been deployed by many scholars to gain an understanding of Latin American traditions and cultures. Situated under the broad umbrella of the carnivalesque, the Bakhtinian concepts of dialogism and centrifugal speech are evident in the reaction of Cinema Novo to increasing political repression on the part of the Brazilian state. The two films analysed here: Glauber Rocha’s Land in Anguish (1967) and, to a lesser extent, Rogerio Sganzerla’s Red Light Bandit (1969) represent the different strategies of Novo and Marginal during a period of transition from the former to the latter. It would be an understatement, however, to say that these directors saw faults in their respective filmic strategies. Thus it is possible to capture a dialectical moment of conflict and transition within Brazilian cultural discourse.

Robert Stam’s work on Bakhtin is central to my argument. According to Stam, “Bakhtin gives the name “carnival” to the de-centralising (centrifugal) forces that militate against official power and ideology” (Stam 1989: 122). This involves a contrast between the carnival and dominant power as well as one between “open” and “closed” cultures on a national and international level. A culture becomes permeable as a result of its “hearing” others, its hermetic rigidity is loosened and “decentering” occurs. The idea has its roots in Bakhtin’s analysis of 19th century literature, particularly the work of Dostoyevsky and Francois Rabelais. In the work of Dostoyevsky, Bakhtin finds that the authorial voice of the narrator is subsumed by the multitude of characters he creates, with each articulating a different subjective experience of the world. The text is constituted by myriad voices, and in this way it becomes dialogical. At the same time the images created by Rabelais “have an undestroyable, non-official nature; “no dogma, no authoritarianism, no narrow-minded seriousness can co-exist with [them]”” (Stam 1989: 86). The carnivalesque, then, is a dialogical mode of addressing authority, the strength of which lies in its permeability with the Other.

Carnival is rooted in the ancient traditions of Socratic dialogue and Mennipean satire, the essential characteristics of which stand in contradistinction to the traits of European Verisimilitude. In his book on Bakhtin, Stam highlights 14 of these characteristics. They include:

 ·      The constant presence of the comic element

·      A fondness for the experimental and the fantastic

·      Elements of social utopia

·      A polystylistic language and approach

·      An emphasis on moral-psychological experimentation, split personality, insanity, and abnormal psychic states showing the “unfinalisability of man” and his “noncoincidence with himself” (Stam 1989: 85)

The carnival aesthetic is, as we can see, not limited to Brazil. European Surrealists were using Mennipean tactics in film as early as Un Chien Andalou (1929), and Luis Bunuel continued to embrace aspects of the carnival until his death in 1983. As I will show later, the early articulation of the Mennipean in film was to have particular influence on the Brazilian Modernist movement of the 1920’s, and then again in the re-examination of Modernism as Novo entered its Tropicalist phase. The main difference between the European and Brazilian conceptions of carnival lies in the latter’s extension of the carnivalesque to the material world. Whereas in Europe carnival remained an abstract concept that failed to become a real social practice, in Brazil it was (and still is) a genuine lived experience. Real life carnivals, as Stam explains, are “complex crisscrossings of ideological manipulation and utopian desire” that cannot be separated from the “institutional sites in which the complex relations of discourse and power are actually negotiated” (Stam 1989:96). It is for this reason (among others) that carnival over the years has been susceptible to reclamation by dominant powers, a subject touched upon in Land in Anguish. It was not just the carnivalesque that became literalised. Cannibalism has also become more tangible in comparison to the metaphors of the European avant-garde.

 Although cannibalism was a common trope among many European avant-gardists, only in Brazil did anthropophagy [the eating of human flesh] become a key cultural trope that was to prolong itself over many decades…through Oswald de Andrade’s speculations in the 1950’s concerning anthropophagy as the “philosophy of the technicised primitive”, to the pop recyclings of the metaphor in the tropicalist movement of the late 1960’s…Thus we find in Brazil a literalisation of the metaphors of the European avant-garde (Stam 2004: 238).

This literalisation takes place in the films of Novo and Marginal Cinema as a response to urgent conditions of poverty that were unbeknown to European cultural producers. It is part of a radical film movement swept across Latin America throughout the 1960’s and which can be seen to have arisen in a dialectical relationship to conditions of political repression and economic hardship that were not present in Europe. A centrifugal movement away from dominant power took place in Latin American film industries. While having certain ties to Europe, these industries developed along different lines in accordance with the different material conditions in which they were situated. Conversely, the “Mennipean chronotope”, as a historical form of literary and filmic discourse that contrasts with European verisimilitude, has, in many cases, been “misjudged or misappreciated” because its works have “been judged by canons of “good taste” and “political correctness” rather than as prolongations of a perennial carnivalesque tradition” (Stam 1989: 115).

Oswald de Andrade, father of the Caniibalist movement

Oswald de Andrade, father of the Caniibalist movement

For Paul Willemen, this amounts to an imposition of Euro-American aesthetic theories upon non-European cultural practices, leading to what he calls “film theoretical malpractice”. In his analysis of Bakhtin and discourses surrounding the national he identifies three ways of relating to the cultural practices of the Other. The first involves a projection by the reader/viewer of his/herself onto the text, leading to to a “scornful amusement” which results from the incompatibility of two different world-views. This “projective appropriation accompanies efforts to internationalise a restrictive regime of making sense. It is concerned with conquering markets, eliminating competition and securing monopolies” (Willemen 29: 2006). The second way of relating involves a complete immersion in the culture of the Other, and a neglect of the reader/viewer’s specific cultural enrootedness. A drastic veering towards the opposite end of the spectrum, this relation involves a form of cultural “ventriloquism” that allows the coloniser to maintain their authority “while masquerading in the clothes of “the oppressed”” (ibid). Bakhtin’s third type of relating to the Other, according to Willemen, is a dynamic relation between the viewer/reader and the culture being observed. This entails a realisation that both the culture under analysis as well as the reader’s own culture exist as sociocultural formations that are specific moments within a historical construct. The observation of cultural practices transforms both the practices themselves and the observer. In becoming conscious of this transformation, the observer replaces an attitude towards cultures as a series of hermetic traditions, with one in which “the object of study is precisely the dynamic interconnections of processes that combine to form a social formation…a sociocultural constellation” (Willemen 2006: 33). This third type of relating is dialogical because

 For Bakhtin, creative understanding requires a thorough knowledge of at least two cultural spheres. It is not simply a matter of engaging in “dialogue” with some other culture’s products, but of using one’s understanding of another culture to reperceive and rethink one’s own cultural constellation at the same time (Willemen 2006: 30)

These three types of relation are broadly analogous to three major factors within the development of Brazilian film throughout the 1960’s and early 70’s. The first relation is comparable to the U.S hegemonic stance on Latin American film production. The second bears familiarity to the Novo movement in its first phase, with its desire to return to a specific Brazilian-ness. The third, dialogical relation, reflects the Novo movement in its final Tropicalist phase and its re-examination of the Modernism of the 1920’s. It represents a tendency within Marginal Cinema towards the recognition of Brazilian cultural specificity as one that is inextricable from the fact of colonisation and its consequences. With this framework in mind we now move towards an analysis of Novo’s trajectory towards Marginal Cinema.

It is widely understood that Cinema Novo began to emerge in the mid 1950’s in the wake of the chanchada genre and Vera Cruz produced films. While the chanchada was a “specifically Brazilian genre that offered valid alternatives” to U.S dominated film production, it later became “defiled by bad taste and by the most sordid commercialism, representing, finally, a form of cultural prostitution” (Diegues 1995: 65). Meanwhile, Vera Cruz, according to Carlos Diegues, was simply “a bizzare, structureless monster, without roots in our culture, nourishing itself on the dream of a European cinema in an illiterate and impoverished Brazil” (ibid). The self-adopted remit of the group of independent filmmakers who came to constitute Cinema Novo was to re-articulate foreign values into explicitly Brazilian processes. The goal was to create an independent, educational and revolutionary cinema for the Brazilian masses. While Novo did, in an anthropophagic sense, consume the auterist philosophy of the French nouvelle vague, the urgency and despair of the political situation in Brazil led theorist/filmmakers such as Rocha to articulate an “aesthetic of hunger” that turned this auterism toward political ends. In 1961 Rocha wrote:

 We do not want Eisenstein, Rosselini, Bergman, Fellini, Ford or anybody else…our cinema is new because Brazilian man is new and Brazilian problems are new, and our light is new and that is why our films are born different from European films. (Hollyman 1983: 17)

Responding to the economic failure of Vera Cruz, Novo filmmakers spoke of how a foreign controlled film industry was unsustainable due to its inability to provide sufficient returns on the expensive studio films it produced. Like most industries in Brazil at the time, the film industry was dominated by foreign interests. The centrifugal movement of Brazilian cinema commences, therefore, with the promotion by Novo filmmakers of low-budget, artisanal modes of production that actively undermined the hegemonic power of foreign interests, particularly those of the U.S. Under the influence of revolutionaries such as Che Guevara and Frantz Fanon, Novo filmmakers looked to re-conquer the Brazilian film market from Hollywood domination. Their focus on the industry would later become a subject of much debate between Novo and Marginal Cinema. The three phases of Cinema Novo have been identified by Randal Johnson as such: 1960-64 was a period in which the developmentalist policies of the previous Kubitschek government continued under Joao Goulart. Opposition from both right and left wing tendencies within Brazil led to a military coup d’etat, marking the beginning of the second phase 1964-68. This period is overshadowed by the passing of the Fifth Institutional Act, which suspended habeus corpus and declared a nationwide state of siege. Thus followed a period of repressive military rule in which filmmakers were forced to turn to allegory to express opinion. This phase ran between 1968-73. After this it is widely acknowledged that Cinema Novo as a movement ceased to exist. Increasing repression by the Brazilian state radicalised Novo to the point where it entered a final Tropicalist phase before being replaced by Marginal Cinema, a perfect example of how dialectical conflicts transform cinematic movements.

DVD cover for Glauber Rocha's "Land in Anguish"

DVD cover for Glauber Rocha’s “Land in Anguish”

While the Goulart government provoked disdain from the Brazilian left, it was his overthrow by the right wing that provided the basis for a sea-change in the concerns of the Novo movement. “The focus” Randal Johnson notes:

 shifted from rural to urban Brazil, from the lumpen proletariat to the middle class, as filmmakers turned their cameras, so to speak, on themselves in an attempt to understand the failure of the intellectual left in relation to the events of 1964 (Johnson 1984: 3)

In 1965 Glauber Rocha’s essay “An Aesthetic of Hunger” was published in the Brazilian journal Revista Civilizacao Brasileira. Two years later he would echo the ruminations of Solanas and Getino, advocating a “guerilla cinema” in his essay “The Tricontinental Filmmaker: That is Called The Dawn”. The “tricontinental” refers to Che Guevara’s grouping of Africa, Asia and Latin America, three regions that had been subjected to colonial barbarism and were now, as he saw it, primed for a revolutionary explosion. During the period 1964-68 Rocha directed two feature films: Black God White Devil (1964) and Land in Anguish (1967). The latter, a baroque dramatisation of the events of 1964, looks at the failure of the intellectual left and the faults of populist politics. Set in the mythical Eldorado, it is the story of two parallel political campaigns, that of the populist Vieira, and the right-wing Diaz. Situated in the middle is the inconsistent political activist and poet, Paolo. Throughout the film, a morosely comedic light is shed on tragic events. Populist figureheads and political activists are revealed to be weak and pathetic, while Diaz assumes an air of demonic mysticism that is a common trope in many of Rocha’s films. Indeed, the whole narrative is set in a temporally and geographically ambiguous chronotope (a Bakhtinian term that translates literally as time-space), increasing its historical resonance and “tri-continental” appeal. The multi-temporal aspect is introduced through shots of Diaz arriving on the shores of Eldorado in a leaky boat that is being rowed by an Indian and a white coloniser, both dressed in 15th century historical regalia. At the same time, Diaz is wearing a modern business suit and carrying a cross. Here he is both a product of history (specifically, the historical conjuncture of white colonisation over Indians) and its forbearer, bringing christianity and capitalism to Eldorado. The dialogical aspect is consolidated in the figure of Diaz qua Diaz, who exists as a result of the interjection of two cultures.

Paulo Autran plays Porifio Diaz in Glauber Rocha's "Land in Anguish"

Paulo Autran plays Porifio Diaz in Glauber Rocha’s “Land in Anguish”

This example of what Stam calls “chronotopic multiplicity” (Stam 2003: 32) echoes throughout the film in the ambiguity of the mise-en-scene, particularly with respect to the presidential palace. Thick jungle provides the backdrop, surrounding the foreground in which the action takes place. Here Rocha’s mysticism, enunciated by the unclear connection between foreground and background, takes the shape of a history from which the present emerges. This history is bound up in the present. Its lack of clarity is reflected in the frenetically confused baroque style of the film, which presents it as both comic and tragic at the same time. The developmentalist policies of the Goulart government are allegorised and parodied during the scene in which Vieira drafts his resignation letter (the 1964 coup). Pacing back and forth on top of the presidential palace, surrounded by socio-political figureheads, he is asked by Paolo “what’s this all about”, to which he gives the mechanical reply “it’s about economic and social development”. The inconsistency and indecisiveness of the left becomes a primary trope within a film that was “saturated with a feeling of mourning because of the realisation that a delay of radical confrontation meant a long-term defeat, making the present unbearable” (Xavier 1997: 124). Rocha places the responsibility for this defeat on bourgeois cultural elites. The protagonist, Paolo, for example, is manifest inconsistency, bouncing between the two seemingly opposing political poles of Vieira and Diaz. Sometimes he is a political activist, at others he becomes a recluse engaging in nothing but “the vain practice of poetry”. He campaigns for Vieira, the “voice of the people”, and then violently represses the people when they begin to protest. Rocha constructs a critique of populism through the portrayal of Paolo and Vieira’s relationship. In one scene they are shown with another “activist” on the roof of the presidential palace. The three of them sit around a dining table, drinking wine and laughing ecstatically as they congratulate each other on their high populist morals. This suggests their engagement with an abstract concept and concurrent neglect of material reality, a use of terms in place of a commitment to what they mean, in short, a lot of words and no action.

Taking a radical stance, the film articulates nationhood in terms of class struggle. For example, when the media mogul Fuentes wearily complains to Diaz that he is “left wing”, Diaz mocks both him and the audience by addressing the latter and asking to the camera “what is your class?” This direct address to the art-house crowd, who would have constituted the audience, forces them into the kind of self-analysis advocated in Willemen’s reading of Bakhtin. It destroys any illusions of a political difference between Vieira and Diaz, and thus between the cultural elites and the military government of Brazil. These were some of the tenets of Novo as it entered its Tropicalist phase: a radical movement away from established cultural and political power in response to the increasing repression enacted by it. Eventually, however, Novo became the target of criticism for an emerging community of Underground filmmakers. The intellectualism implied by Rocha’s recuperation of Brazilian folklore in an attempt to articulate a specific Brazilian-ness, as well as his mysticism and poeticism, led filmmakers like Rogerio Sganzerla to the opinion that Rocha only advanced the kind of cultural elitism he seemed to detest. This so-called revolutionary cinema could never reach the people because it utilised archaic cultural forms to portray a Brazil and a Brazilian-ness, which, for the majority of uneducated, poverty stricken Brazilians, no longer existed. “This audience” wrote Joaquim Pedro De Andrade in his post 64’ re-examination of Novo

 Simply does not see Novo films. When it sees them it neither accepts nor understands them…I sometimes doubt that it [Cinema Novo] intended to reach large sectors of the public…Brazilian cinema had an international impact but the problem of a Brazilian audience for that cinema remains unresolved (De Andrade 1995: 75)

In its rejection of all non-Brazilian cultural values in favour of a traditional Brazilian-ness (a method advocated by the Modernist of the 1920’s) Cinema Novo had neglected the fact of the country’s colonisation. The re-examination of 1920’s Modernism in relation to the failures of Novo led to the emergence of a new aesthetic based on the acknowledgement of Brazil’s neo-colonised position and status as a dumping ground for literal and metaphorical trash from the U.S. In a cannibalistic act of consumption and regurgitation, these filmmakers would make use of cheap and trashy popular forms, inverting the symbolic juxtaposition between “civilized” and “savage”, “rational” and “irrational”, “filth” and “cleanliness” in what became known as an aesthetics of garbage. This was seen as both “a practical necessity as well as an artistic strategy” (Stam 2003: 35). A cannibalistic text is by its very nature what Stam calls a palimpsest “the parchment on which are inscribed the layered traces of diverse moments of past writing” (Stam 2003: 34). The cannibal consumes the cultures of any historical moment or geographical location that are relevant to his/her situation and purpose. In this way both the text and its creator are multi-layered hybrids that embody the Bakhtinian concept of dialogism, as they are necessarily made up of many voices. We saw this kind of hybridity with the shots of Diaz on the boat in Terra em Transe. It can be seen to a much greater extent in Marginal Cinema films, particularly Sganzerla’s Red Light Bandit, which, from its beginning, employs a Brechtian fragmentation of sounds and images from different mediums.

The film follows the protagonist, the bandit, as he rampages through the Boca do Lixo (Mouth of Garbage, a real region in Sao Paulo) assuming various identities while constantly posing the question “Who am I”? The opening credits are displayed on a news ticker, high up on a building overlooking a busy street. Here Sganzerla is already making use of the junk surroundings present within a media dominated society. The news ticker is used throughout the film to sensationalise the narrative, and this constitutes a cannibalistic ingestion of Western media techniques. We also see frenetic jump cuts between a radio tower coupled with the beeping of Morse code and children playing in a junkyard, while the final chase between the bandit and his assailant, detective “Big Head” Sade, is permeated with a crude animation of a flying saucer chase. The music over the end credits is a distorted fusion of Jimmy Hendrix and bongos, throughout which we see wild, tribal stances appear through the flames of burning trash. This “imagistic chaos” (Xavier 1997: 101) is accentuated by the fragmented sentences of the male and female narrators, which is itself punctured with interruptions from characters within the diegetic world. The Brazil of “eloquence and drama” depicted by Cinema Novo has changed into one of “irony or self mockery…a comic province far from the civilised world” (Xavier 1997: 120). This “strategic redemption of the low, the despised, the imperfect and the “trashy” as part of a social overturning” (Stam 2003: 35) is what situates Marginal Cinema at the outer limits of the peripheral dominated by centres of cultural and political power. The antiauthoritarian aesthetic of Red Light Bandit inverts the symbolic order, and, using Mennipean satirical techniques, turns existing power relations on their head. Although Sganzerla’s film is not explicitly political, politics, like the mass population at whom the film is directed, is an ever-present undercurrent in the film. Instead of trying to reclaim a lost set of Brazilian traditions, Sganzerla “focuses on those who do not have a known tradition-the ignorant, the excluded, the underground population that, for him, offers the best image of the whole nation” (Xavier 1997: 126).

Rogerio Sganzerla plays the Bandit in his film "Red Light Bandit"

Paulo Villaca plays the Bandit in Rogerio Sganzerla’s “Red Light Bandit”

In the transition between Cinema Novo and Marginal cinema we have an example of how film movements transform dialectically, against complex backgrounds of socio-political development. Brazil is an especially interesting case, since it shows how state repression necessitated an increasingly radicalised cultural response. Further analysis may entail a more in-depth discussion of the relationship between Cinema Novo and the Brazilian state run film industry, Embrafilme. We might also turn to the contradictions inherent in Novo from the beginning, and which led some theorists to describe its aesthetic not as one of hunger, but of gluttony. In any case, it should be understood that while Novo as a movement is widely regarded as a staple feature in the history of avant-garde cinema, its successor, Marginal Cinema, offers a dynamic interpretation of cannibalism and its uses in countering ne-colonialism. This strand of the radical cinema that emerged from Latin America in the late 60’s has been maligned by both the left and right in Brazil. For this reason it is something of a hidden history, and should be paid more attention by film theorists who wish to make a contribution to radical film and its analysis.


Bakhtin, Mikhail (1984) Problems with Dostoyevsky’s Poetics, Minneapolis: University of Minnesota Press

Bakhtin, Mikhail (1984) Rabelais and his World, Bloomington: Indiana University Press

De Andrade, Joaquim Pedro (1995) “Criticism and Self Criticism” in Johnson, Randall, Stam, Robert (eds) Brazilian Cinema, p.72-75, New York: Columbia University Press

Diegues, Carlos (1995) “Cinema Novo” in Johnson, Randall, Stam, Robert (eds) Brazilian Cinema, p.64-67, New York: Columbia University Press

Fernando, Solanas, Getino, Octavio. (1968) Towards a Third Cinema. Available: [Last accessed 19/12/2012]

Hollyman, Burnes Saint Patrick (1983) Glauber Rocha and the Cinema Novo in Brazil, New York: Garland

Johnson, Randall (1984) Cinema Novo x5: Masters of Contemporary Brazilian Film, Austin: University of Texas Press

Stam, Robert (1989) Subversive Pleasure: Bakhtin, Cultural Criticism and Film, Baltimore: Johns Hopkins University Press

Stam, Robert (2003) “Beyond Third Cinema: The Aesthetics of Hybridity” in Guneratne, Anthony. R, Dissanayake, Wimal, Re-Thinkng Third Cinema, p.31-49, London: Routledge

Stam, Robert (2004) Tropical Multiculturalism: A Comparative History of Race in Brazilian Cinema and Culture, Durham: Duke University Press

Willemen, Paul, Valentina, Vitali (2006) Theorising National Cinema, London: BFI

Xavier, Ismail (1997) Allegories of Underdevelopment, Minneapolis: University of Minnesota Press


Black God White Devil (Glauber Rocha, 1964, Brazil)

Land in Anguish (Glauber Rocha, 1967, Brazil)

Un Chien Andalou (Luis Bunuel, 1929, France)

Red Light Bandit (Rogerio Sganzerla, 1969, Brazil)

Author: Anthony Killick

Share This Post On

Submit a Comment

Your email address will not be published. Required fields are marked *