Corps de l’article

Introduction: The Beginnings of Concrete Poetry in Brazil

In the late 1950s Brazil emerged from World War II with newly gained confidence in its national future. Under the optimistic leadership of President Juscelino Kubitschek, whose motto was “50 years of progress in 5,” social and political changes ushered in a series of parallel and unprecedented transformations in many areas of the cultural arena.[1] Literature witnessed the appearance of the concrete poetry movement, arguably the first homegrown postwar avant-garde movement in Latin America, which put Brazil on the global literary map for the first time.

Brazilian concrete poetry was launched by Haroldo de Campos, Augusto de Campos, and Décio Pignatari, who at the time were barely in their twenties. These gifted and ambitious young poets issued the “Pilot Plan for Concrete Poetry” in 1958, a manifesto whose title and spirit mirrored the modernist aesthetics of the International style in architecture and celebrated the construction of Brasília.[2] Opposing a vague “poetry of expression,” which had thrived in the postwar period (the so-called Generation of 1945), concrete poetry laid particular emphasis on the material elements of language, doing away with the concept of the line of verse and substituting it by visual arrangements of words on a page that could be read in any direction. This move attempted to replace a discursive mode of language by an “ideogrammic” method of poetic composition. As the “Pilot Plan for Concrete Poetry” stated:

Concrete poetry: product of a critical evolution of forms. Assuming that the historical cycle of verse (as formal-rhythmical unit) is closed, concrete poetry begins by being aware of graphic space as structural agent. Qualified space: space-time structure instead of mere linear-temporal development.

Campos, Haroldo de et al., 2007b [1958], p. 217

Furthermore, instead of espousing a vision of language that blindly acquiesced to the Saussurean divide between signifier and signified, Brazilian concrete poetry aimed at remotivating language by integrating sound, image, and meaning to the highest degree, a “verbivocovisual” approach where no single element is privileged. The strict correspondence they sought between these elements could be termed “isomorphism,” or similarity in form. The form, in other words, aspired to reflect the content: “We call isomorphism the form-subject conflict looking for identification” (ibid., p. 218). In an article published in 1957, one year before the “Plano Piloto,” Haroldo Campos had elaborated on the notion of isomorphism as the result of an evolution from the phenomenology of composition to a more advanced “mathematics” of composition: “The passage from the phenomenology of composition to the mathematics of composition coincides with another passage: that from the organic-physiological to the geometric-isomorphic”; this advanced stage would effectively entail the “elimination of the descriptive poem: the content of the poem will always be its structure” (my translation, 1975 [1957], p. 94).

In addition to poetic experimentation, the Brazilian concrete poets also launched a program of literary translation mostly based on their model and mentor Ezra Pound, from whom they derived the title of their first little magazine, Noigandres[3] and the name of their group. This translation program involved locating examples of a similar poetics in the past (from the medieval troubadours to the avant-gardes of the interwar period, passing through key figures like Mallarmé) and translating them in a way that made them relevant for modern poetics. Their practice of translation became, thus, a “laboratory” for writing and at the same time was the source of their theoretical reflections on translation.

In this paper, I want to consider the isomorphic correspondence inherent in the conception of concrete poetry put forth by the Noigandres poets. I demonstrate how this characteristic of concrete poetry doubles itself in the translation concepts that emerge from their poetic practice. I will also survey later developments by Haroldo de Campos, including a shift towards greater independence of the translation and the agency of the translator, to culminate in what critics have termed Haroldo de Campos’s cannibalistic approach to translation. This trajectory illustrates how the concepts evolved from an early preoccupation with primarily the material/linguistic aspects of a poem to one that also involved the cultural politics of literary production. As I will show, although some of these concepts were developed at times earlier and also independently from mainstream discussions in translation theory, they may be said to coincide with certain concepts such as André Lefevere’s view of translation as “rewriting” and the so-called post-colonial turn in translation studies put forth by scholars such Susan Bassnett and Harish Trivedi, among others. This evolution also signals a fundamental change in the way de Campos viewed his own praxis as poet, translator, and critic from the late 1950s to the mid-1980s and mirrors developments in cultural politics in Latin America in those periods.

The Translatability of Concrete Poetry

In order to approach the emergence of these translation concepts, it is perhaps worthwhile to consider first the implications that the demands placed on the production of concrete poems have for their translation. In other words, how do we translate texts such as concrete poetry, where the material and conceptual aspects are so tightly wrought and interdependent? In considering this question it is important to remember that this poetry aims at diminishing the distance between signifier and signified, integrating them as much as possible and laying a significant emphasis on the material qualities of the signifier, something that, inevitably, will vary from language to language. Taking a step back, this means posing the question if concrete poetry is at all translatable.

The answer clearly depends on what we understand by translation. To illustrate the possibilities available for the translation of concrete poetry I would like to consider two examples of English translations of Brazilian concrete poetry. The first is the famous poem by Décio Pignatari, “beba coca cola,” which plays on the Portuguese version of the advertising phrase “Drink Coca-Cola”:

beba coca cola

babe  cola

beba coca

babe cola caco

caco

cola

 cloaca

1983, p. 138

The poem manipulates the commercial slogan using a series of anagrams and permutations of its words: beba (drink) and babe (drool); coca (Coke or cocaine) and caco (shard). The manipulation of these words and progressive breakdown of the original meaning evoke the negative consequences of buying into the “coca-col(on)ization” of the world. Literally, if you follow the injunction “drink coca-cola” (beba coca cola), you run the risk of “drool[ing] glue shard[s]” (babe cola caco) and ending up in the “sewer” (cloaca).[4] Interestingly, the concluding word cloaca, in a sort of dubious homage, contains all the letters in the name of this ubiquitous drink.

In its close integration of sound, semantics, and image (the original was printed in white letters against a red background, in an advertising-style layout designed by Pignatari), “beba coca cola” is a prime example of the isomorphic ideal of concrete poetry. The form is not independent from its content, especially when it comes to sound, as the puns and wordplay become an integral part of the signifying process: the detritus evoked by the final word “cloaca” is literally a result of the disintegration and rearrangement of the slogan’s words. (The downward motion of the poem, one could also argue, mirrors the descent into the underworld of the sewers.)

“Beba coca cola” also clearly exemplifies the various difficulties involved translating concrete poetry. This is reflected in the “translation” provided in the anthology Brazilian Poetry (1950-1980), which consisted of a list of the dictionary definitions of the words and a commentary by Haroldo de Campos aimed at explaining, rather than recreating, the aesthetic and political message of the poem.[5] Perhaps finding the challenge of translation potentially insurmountable, the Noigandres poets (whom we could assume were involved in the selection of the poems for this anthology) chose in this specific instance not to translate.[6] A peculiar choice indeed, particularly, in view of the translation program pursued by the concrete poets themselves. One could also see this as the case of poem still waiting for its translator, for, as Eliot Weinberger has provocatively argued, “[t]here is no text that cannot be translated; there are only texts that have not yet found their translators” (2000, p. 11). Weinberger echoes Benjamin’s claim in “The Task of the Translator” that the “translatability” of a text is not limited to the existence of an actual translation or potential translator, but rather constitutes an essential feature of the work itself.[7]

In contrast to the above refusal to translate, we may consider Haroldo de Campos’s famous concrete poem “fala prata,” another challenging text from the point of view of translation:

fala

prata

 cala

 ouro

 cara

 prata

 coroa

 ouro

 fala

 cala

 para

 prata  ouro

 cala  fala

 clara

1970a, p. 101

As Marjorie Perloff has observed, the poem “plays with the hackneyed proverb ‘Silence is golden’ (“A palavra é prata, o silêncio é ouro”) as well as the classical epithet ‘silver-tongued.’” After a detailed close reading,[8] she concludes that “[v]isual placement is central to meaning” because, although a temporal reading going from top to bottom and left to right is clearly implied, “the text is also self-reflexive, each item pointing back to its previous partner as well as forward” (Perloff, 2004, pp. 176-177). In this sense, the poem is isomorphic in the correspondence between the arrangement (form) and the meaning, as well as in the puns and rhymes the poem employs.

In Mary Ellen Solt’s anthology, Concrete Poetry: A World View, first published in 1968, Solt provided her own English version:

speech

silver

 silence

 gold

 heads

 silver

 tails

 gold

 speech

 silence

 stop

 silver  golden

 silence  speech

 clarity

Campos, Haroldo de, 1970b, p. 102

While here the original vowel- and end-rhymes in a (fala/cala/prata) clearly could not be replicated, Solt creates a similar structure with initial alliterative sibilants and visual rhymes (silver/silence). Solt’s version nicely reproduces the visual layout, and the narrative as described by Perloff is also preserved. However, the obvious constraints of the original also become evident, especially in the realm of sound and the ambiguity of meaning, something that Perloff’s otherwise insightful reading misses.[9] In assessing the challenges involved in the translation of these texts, Perloff elsewhere notes that:

[p]art of the difficulty, of course, is that Augusto’s poems, like those of his brother Haroldo, are notoriously difficult to translate into English, where their punning, homophonic play, and morphology are invariably lost. But, as in the case of the Galáxias, Haroldo’s sequence of prose poems ingeniously rendered in French by Inez Oseki-Depré, perhaps the solution, for the moment, is to leave aside English, a language markedly different in sound, syntax and morphology from Portuguese, and translate the Brazilian Concrete Poets into French, where it is perfectly possible to give the feel of the original.

2003, n.p.

She is right in pointing out the problems that the differences in the structure of languages (let alone the construction of concrete texts themselves) pose to the translator, and those may well be the reasons behind the refusal to translate “beba coca cola.” Yet, this difficulty was, for the concrete poets, the most tempting aspect of translation: crafting inventive possibilities out of apparently insuperable challenges.[10]

The Noigandres Group and Translation

The examples above visibly illustrate that the very principles that underlie the construction of concrete poetry (that is the isomorphism between form and content) are precisely those that make its translation impossible or, at least, promote its “resistance” to translation. Yet, those challenges were precisely what egged the concrete poets on in their translation endeavors. As I suggested earlier, their proposals for renewal of poetic form were accompanied by an equally innovative program of literary translation, following in the footsteps of their mentor Ezra Pound. The most active of the three in the realm of criticism, Haroldo de Campos also began his forays into translation theory based on his readings of Pound, along with considerations of other translation and literary theorists such as Albrecht Fabri, Max Bense, and eventually Walter Benjamin.

In his 1963 landmark essay “Translation as Creation and Criticism,” Haroldo de Campos begins by citing the German critic and translator Albrecht Fabri on the “tautological” character of art. In a phrase that recalls Archibald MacLeish’s famous poem “Ars Poetica,” de Campos claims, following Fabri, that art does not mean, but is. In literature what is unique to the language of a literary work, is the “absolute phrase,” the phrase “which has no content other than its own structure” and “which only exists as its own instrument” (2007c [1963], p. 312). Translation brings out the less perfect or less absolute character of the phrase. Therefore, “‘all translation is criticism’ because translation ‘is born from the deficiency of the phrase, from its insufficiency to be something by itself’” (ibid., p. 312). Following Fabri, Campos claims that one does not translate what is language in a text, but what is not language.

Using a different terminology derived from the German philosopher and critic Max Bense, de Campos distinguishes between three types of information: “documentary,” “semantic,” and “aesthetic.” “‘Documentary information,’” argues de Campos via Bense, “reproduces something observable and forms an empirical phrase [and] ‘semantic information’ goes beyond what is purely documentary, [designating] an element that is not in itself observable […] such as the concept of true and false” (ibid., p. 313). Finally, “aesthetic information,” which transcends both documentary and semantic information, belongs to the realm of the unpredictable, “the surprise and the improbability of the order of signs,” in other words, “all that is inherently fascinating in a work of art” and is therefore “fragile” because it “cannot be codified except in the form in which it was transmitted by the artist” (ibid., pp. 312-313).

For de Campos then, both Fabri’s “absolute phrase” and Bense’s “aesthetic information” are articulations of the problem of untranslatability of artistic writing, or, in Benjamin’s terms “the inaccurate transmission of an inessential content” (1968, p. 70). The problem, de Campos notes, becomes “particularly acute in the case of poetry” and other highly “creative” texts (2007c [1963], p. 315). But for de Campos, once the gates of language shut down, the gates of translation immediately open up. Acknowledging the impossibility of translation, argues de Campos, means necessarily accepting the corresponding corollary of recreation. In other words, since concrete poetry and other kinds of highly textured or isomorphic literary works are “impossible” to translate, the only possible translation is in effect a re-creation of their forms as a separate creation (cf. ibid., p. 315).

De Campos stresses this “autonomy” of translation while at the same time bringing out the isomorphism between translation and original:

The texts may exist, then, as Bense wishes, in two languages and as two bodies of autonomous aesthetic information, which […] will be linked to each other through an isomorphic relation: they will be different in language, but like isomorphic bodies, they will crystallize within the same system.

ibid., p. 315

In chemistry, “isomorphism” refers to “the existence of two or more substances (isomorphs) that have the same crystal structure, so that they are able to form solid solutions.”[11] De Campos’s metaphor suggests then that original and translation will have a similar structure, although their substance (language) may vary.

It thus becomes clear that the translation concept put forth depends on the choice of texts to be translated. It is intimately connected to that choice because the nature of those creative texts calls for this sort of “creative” translation or “transcreation,” to use the term de Campos coined. Furthermore, as he notes again in the same essay:

The more intricate the text is the more seducing it is to “re-create” it. Of course in a translation of this type, not only the signified but also the sign itself is translated, that is, the sign’s tangible self, its very materiality (sonorous properties, graphical-visual properties, all of that which forms, for Charles Morris, the iconicity of the aesthetic sign, when an iconic sign is understood as that which is “in some degree similar to its denotation”).

ibid., p. 315

In other words, certain types of texts compel us to transcreate them, to attempt to create a form that mirrors the content. Conversely, the choice of texts to translate will also ultimately depend on how “transcreationable” they are, the degree to which a creative operation can be performed in their translation. The more a text presents a challenge because of its isomorphic character, the more it will compel the translator to seek the same isomorphic qualities in the translation, i.e., to produce a translation where signifier and signified are bound, to the greatest degree possible, in a similar way in the target language as in the source.

Before going on to consider a bit further what those texts were for the concrete poets, i.e., the canon of literary translation inspired by the principles of concrete poetry, as well as other theoretical questions de Campos tackled in later years, I want to briefly zero in on another issue de Campos raises here, namely, the graphical visual properties, which are evident in the approach to translation both by Haroldo’s brother, Augusto de Campos, and the other Noigandres member, Décio Pignatari.

Augusto de Campos used the term (in)tradução, a word that puns on both prefixes “in” and “intra” and might be rendered in English as “un(in)translation,” to describe a non-translation, an internal, interior, intimate translation[12] that seeks to translate the inner structures of the poem, not an “inessential content” (Benjamin). Commenting on Augusto’s translations of poems by e.e. cummings, Haroldo de Campos sensed that these were “poems in which even ‘optical data’ had to be translated (through typographical spacing, fragmentation, and interlinear relations), and which, at times, necessitated taking into account the number of letters and physical (plastic, acoustic) coincidences in the verbal material to be employed” (ibid., p. 322).

Similarly, Décio Pignatari’s practice of what could be termed “language design” involved incorporating both words and images into an ideogrammic construct. In Pignatari’s design for the cover of Mallarmé, a collection of versions of this French poet by the three Noigandres members, the photographic portrait of Mallarmé reproduced in the book is “translated” into a digitized image made up of letters spelling out his last name (Figure 1). The letters are arranged into a pattern that suggests the outline of the face, the eyes, the nose, and the moustache.

Figure 1

Portrait of Mallarmé (left) and cover of Mallarmé(right), a collection of translations by Augusto de Campos, Décio Pignatari, and Haroldo de Campos. (Images used with permission. Courtesy of Fany Kon, Editora Perspectiva, São Paulo, Brazil.)

The Translation Canon: From Pound’s “Make it New” to Co-Authorship

-> Voir la liste des figures

While Mallarmé, a pioneering innovator and a translator of Edgar Allan Poe, was certainly a model for the Noigandres group, in the manner as well as in the choice of texts to be translated, Ezra Pound was an early and long-lasting influence.[13] Pound used translation as a kind of laboratory, a dry run for his own poetic writing, practices that, as I have been suggesting, also went hand in hand for the Brazilian concrete poets. Regarding Pound’s writings on translation, Edwin Gentzler argues that he focused “on specific images, individual words, fragments, and luminous details. His method […] emphasized juxtaposition and combination, hoping that the new configurations would react chemically, combining into a new compound, and thereby give off energy” (1993, p. 28). The chemical metaphor echoes Haroldo de Campos’s above-cited characterization of literary translation as akin to isomorphic compounds that produce similarly shaped crystals from the combination of different elements. Furthermore, the concrete poets were sympathetic to Pound’s idea that translation was “A mode of criticism. ‘Criticism by translation’” and that in that respect it constituted “‘nutriment for the creative impulse’” (Campos, Haroldo de, 2007c [1963], pp. 316 and 322).

As part of that “nutriment,” they first translated Pound himself, and then took their cue from him in translating classical Chinese poems and Japanese Noh plays; Provençal troubadours; Guido Cavalcanti; and the French Symbolists. To the Poundean paideuma they added the Russian poets Vladimir Mayakovsky and Velimir Khlebnikov, Joyce’s Finnegans Wake, the English Metaphysical Poets, Mallarmé, and Dante, among others. All of these were seen as potential models for the kind of inventive writing they themselves were trying to produce. As Haroldo observed, Pound’s “work is both criticism and pedagogy, since while it diversifies the possibilities of his poetic idiom it also offers up to new poets […] a whole repertory […] of basic poetic products now reconsidered and revivified. ‘Make it new’: give new life to the valid literary past via translation” (2007c [1963], pp. 316-317).

It was indeed Pound’s injunction to “make it new,” to translate poets from the past and make them relevant for modern poetics, which drove their translation practice. And that practice of recreating the past through translation led to a new attitude, a move from an interest in the purely linguistic aspects of translation (isomorphism), to a greater role acquired by the translator, and a greater autonomy (not only linguistic, but also political) gained by the translation. This was emphasized in particular by Haroldo de Campos, the most prolific translator[14] and theorist of the three original members of Noigandres group.

As Médici Nóbrega and Milton observe, “from the beginning of his theoretical activity, Haroldo rejected the biased view that translations are inferior products, as the translator now, far from being the author’s servant or mouthpiece, or a reproducer of meanings, becomes a recreator and a critic, choosing texts that deserve to be translated, and successfully recreating them” (2009, p. 260). In “transcreation,” the translator becomes in effect a co-author, his or her role being creatively at least equalized to that of the author. More emphasis is laid on the agency of the translator, on the role of the translator as an independent agent rather than a subservient passive force. Nowhere is this more evident than in the postscript de Campos wrote to his translation of Goethe’s Second Faust, an essay entitled “Transluciferação mefistofáustica” (“Mephistofaustian Transluciferation”), where, instead of “transcreation,” he calls his translation a “transluciferation.”

Drawing on Benjamin and Goethe, in that essay he characterizes translation as an act of transgression, inverting the traditional conception of translation as subservient to the original, bound by a “tribute of fidelity.” Instead, he argues following Benajmin’s concept of a “pure language,” it is really the original which owes to the translation, which, once it has rid itself of the task of inaccurately transmitting an inessential content, it can instead, following Lucifer’s rebellious motto non serviam (I shall not serve), devote itself to another task of fidelity, fidelity to the form (Campos, Haroldo de, 1981, p. 179). Here too we hear an echo of the early concrete concern with form, and the isomorphic relation between meaning and matter. But, going beyond that, and in a new twist, this denial to submissively serve the content entails a discussion of a power differential. In this case, translation becomes a “Satanic” enterprise, approaching a transgression of signs analogous to Satan’s semiologic sin of hubris, as in Dante’s Paradiso, (Canto XXVI, line 117). Lucifer was banished because he attempted to transgress the sign (“il trapassar del segno”), a transgression of signic limits (ibid., p. 180). What is significant here is that the agency of the translator is emphasized, in a way that evinces an attitude of rebellion against a purported power differential between original and translation, metaphorically God and Satan. This new emphasis on questions of power becomes crucial for de Campos in his later phase where the metaphor of cannibalism emerges.

Cannibalism and the Alexandrian Barbarians

Concurrently to his work in translation and translation theory, in his work as literary critic, de Campos increasingly addressed ideological questions regarding the position of the Latin American writer vis-à-vis the European tradition. This corresponded to his interests shifting from primarily formal questions of language to a more political (and politicized) view of translation and culture, particularly the vindication of Latin American cultural production in view of its perceived dependency on European models.

In 1981, the same year de Campos’s translation of Goethe was published, he wrote an essay entitled “Da Razão Antropofágica” (“Anthropophagous Reason”), where he posed the question of whether or not an avant-garde movement could emerge in conditions of economic underdevelopment and, more importantly, of cultural dependency. As de Campos notes, his inquiry was encouraged by a similar preoccupation expressed by the Mexican poet Octavio Paz in his 1967 essay “Invention, Underdevelopment, Modernity,” where Paz declared: “Aside from the fact that I am very much averse to reducing the plurality of cultures and the very destiny of man to a single model, industrial society, I have serious doubts as to whether the relationship between economic prosperity and artistic experience is one of cause and effect” (cited in Campos, Haroldo de, 2007a [1981], p. 159). Although Paz’s critique[15] is couched in the economic language of the 1960s developmentalism, his (and de Campos’s) concerns may be read today in the wider political arena of what is called Postcolonial Studies, going beyond economic dependency to scrutinize the effects of colonial powers upon their colonized subjects.

Citing as precursor Oswald de Andrade’s 1928 “Cannibal Manifesto,”[16] de Campos focused primarily on the figure of the all-knowing, all-devouring Latin American writer and intellectual, who is not only all too aware of European culture, but also exercises a strategy of “planetary devouring.” Among the practitioners of this “global cultural consumption,” he cites figures such as Octavio Paz, José Juan Tablada, Jorge Luis Borges, and José Lezama Lima. These writers, whom de Campos terms the “Alexandrian barbarians,” have been re-synthesizing European and other sources,

chemically, through an impulsive and uncontrollable metabolism of difference […] equipped with chaotic libraries and labyrinthine card catalogs […] For some time, the devouring jaw of these new barbarians has been chewing up and “ruining” a cultural heritage that is ever more global. In relationship to this heritage, the barbarian’s ex-centrifying and deconstructing attack acts with the marginal impetus of the carnivalesque de-sacralizing, profaning anti-tradition, evoked by Bakhtin in counterpoint to the main road of Lukácsian epic positivism, to monologic literature, to the closed, univocal work. In contrast, the combinatory and ludic poly-culturalism, the parodic transmutation of meanings and values, the open, multi-lingual hybridization, are the devices responsible for the constant feeding and re-feeding of this baroquizing Almagest: the carnivalized transencyclopedia of the new barbarians, where everything can coexist with everything. They are the machinery that crushes the material of tradition with the teeth of a tropical sugar-mill, transforming stalks and husks into bagasse and juicy syrup.

ibid., pp. 173-74

This exuberant description of what de Campos sees as emblematic of the cultural processes of Latin America (if not the New World) has implications for the way de Campos’s views on translation and translation theory develop in his later years, and as I will note below, can be seen through the lens of postcolonial translation. In analyzing this rich passage, it is interesting to note first the recurrence of the chemical metaphor alluded to earlier in the discussion of de Campos’s early translation concepts. Here, however, several new elements emerge: the chemical transformation is now aimed at contrast, rather than isomorphism (“a metabolism of difference”), and the operation is now one of physical consumption, devouring, indeed cannibalism. This is an important departure from his earlier discussions. Rainer Guldin perceptively notes that “the metaphor of feeding and nourishing plays only a very marginal role” in de Campos’s early theorizations on translation, such as “Translation as Creation and Criticism” (2008, p. 113).[17] Yet, the cannibalistic attitude of Haroldo and Augusto de Campos becomes paramount in their work as translators.[18] In other words, the de Campos brothers situate themselves and their translation and creative practice squarely among that of the cannibalistic Alexandrian barbarians.

Conclusion: Haroldo de Campos’s Concepts vis-à-vis Contemporary Translation Theory

In order to situate and contextualize the discussion of de Campos’s contribution to translation theory, it is important to note his visionary role, since he was first formulating these concepts in Brazil in the late 1950s and early 1960s, geographically and chronologically removed from mainstream formal theoretical discussions of translation.[19] His early concepts where informed by Pound[20] and Eliot, and more contemporaneously, the work of literary translators and philosophers associated with the Hochschule für Gestaltung in Ulm, Germany, such as Albrecht Fabri and Max Bense, whom de Campos met in his travels. However, de Campos’s concept of isomorphic “transcreation”[21] can be productively brought into dialogue with contemporary theorizations, such as André Lefevere’s concept of translation as rewriting. Bassnett and Lefevere argue that,

Translation is, of course, a rewriting of an original text. All rewritings, whatever their intention, reflect a certain ideology and a poetics and as such manipulate literature to function in a given society in a given way. Rewriting is manipulation, undertaken in the service of power, and in its positive aspect can help in the evolution of a literature and a society. Rewritings can introduce new concepts, new genres, new devices and the history of translation is the history also of literary innovation, of the shaping power of one culture upon another.

1992, p. vii

Transcreation based on isomorphism certainly reflects a specific ideology and a poetics. The form matters in this new poetics that, indeed, seeks to establish itself as a model for future poetic production, namely a concrete poetics. The commitment to a poetics is a sign as well of its affiliation with a position of power. The “Pilot Plan for Concrete Poetry,” like many avant-garde manifestoes, attempted to situate the new poets and poetics in an adversarial, if not superior, position to a conservative poetic establishment. Likewise, through their emulation of Pound and his translation method and practice, the concrete poets sought to introduce new forms and devices, to consciously and deliberately innovate through translation.

Yet, as scholars have noted, and is the case with Haroldo de Campos’s early phase, “translation was seen as purely an aesthetic act, and ideological problems were disregarded,” despite the fact that “the act of translation always involves much more than language [and t]ranslations are always embedded in cultural and political systems, and in history” (Bassnett and Trivedi, 1999, p. 6). Similarly, as I have shown, de Campos’s concern with linguistic form eventually gave way to a more politicized position that involved not only more autonomy of the text, but also, greater agency of the translator and in cultural terms a reversal of the patterns of dependency through the figure of the cannibal. De Campos’s adoption of the cannibal as metaphor for Latin American cultural processes coincides with a general tendency by 20th-century intellectuals in Latin America and elsewhere to question the primacy of the Western cultural legacy concept of originality, originally associated with Europe.[22] Consequently, this entailed also a reassessment of the concept of translation.[23]

Although de Campos didn’t specifically address translation in his landmark essay “Anthropophagous Reason,” the anthropophagic operation he describes as a pattern of Latin American cultural production is ideally suited to the practice of translation, and indeed has been theorized in that way by other scholars in his wake. As early as 1991, Susan Bassnett, in a series of lectures, acknowledged that

Haroldo and Augusto de Campos […] use Derrida to develop something like a post-modern and non-eurocentric approach to translation. In their approach to literary translation, the de Campos brothers refuse any sort of preordained original, but instead view translation as a form of transgression. They use terms not part of any European approach or science, but come up with their own terms, one of which is translation as a form of “cannibalism.”

Gentzler, 1993, p. 192

It is fascinating to note, then, that de Campos’s move predates the so-called “postcolonial turn” in Translation Studies, as characterized by Bassnett and Trivedi: “So in this post-colonial period, when, as Salman Rushdie puts it, the Empire has begun to write back, it is unsurprising to find radical concepts of translation […] that challenge established European norms about what translation is and what it signifies” (1999, p. 4).

Elaborating on the relation between anthropophagy and culture, Else Ribeiro Pires Vieira has argued how in de Campos, a view of translation based on the figure of the cannibal who devours a revered enemy in order to render homage and also acquire its forces, can be used to characterize Latin American translation of Western texts and indeed deconstruct the center/periphery dichotomy. This happens when we consider the fact that the bodies of both cannibal and cannibalized, through the sort of exchange of energies, transfusion, consumption and reincorporation, become indissolubly linked, so that one no longer can assert the independent existence of one another. The plenitude of origin is questioned when the enemy (the original) text becomes literally incorporated into the cannibal’s body (of texts).[24]

In other approaches to the cannibal metaphor, Rainer Guldin argues that de Campos “focuses mainly on the act of cannibalistic absorption itself […], the deconstruction of the foreign by chewing and swallowing, as well as the relationship arising from this interchange” (2003, p. 117), whereas Michaela Wolf suggests that, “[t]ranslation as cannibalization, on the other hand, does not conjure away the ‘original,’ but devours it in order to create a cultural attitude nourished by foreign influences and enriched by autochthonous input which helps to dismantle the traditional asymmetrical power relations between the cultures involved” (2003, pp. 126-127).

Cannibalism and the greater autonomy and agency of the translator also had implications for the formation of a literary canon that the Brazilian concrete poets sought to establish alongside their proposals for poetic renewal. Here again, we might turn to Lefevere’s argument that “the process resulting in the acceptance or rejection, canonization or non-canonization of literary works is dominated […] by very concrete factors that are relatively easy to discern […] such as power, ideology, institution, and manipulation” (1992, p. 2). It is well known that, through their publications and activity in these fields, the Brazilian concrete poets were able to garner considerable influence in the literary scene in Brazil. To this day their poetic and translation practice remain an important point of reference, including the authors they translated and promoted. But as some have also insightfully suggested, conversely, the formation of such a canon does not remain an uncontested fact.[25] The metaphor of cannibalism they helped popularize and its relation to translation continues to inspire new research in contemporary production, such as popular music.[26]

In closing, I would like to note that de Campos’s last translation was a new Brazilian Portuguese version of Homer’s Iliad, a project of epic proportions (literally) he completed just months before his death in 2003. Its seems particularly significant that de Campos would end his career in this way, with a highly transcreative and baroque-izing version of a work widely considered one of the founding texts of Western literature. He, in a sense, came full circle. From his early avant-garde poetics that rejected traditional forms and focused on isomorphic original texts and translations, he eventually cast himself as one of the cannibalistic Alexandrian barbarians who feed on the Western canon, acquiring its energies while transforming it to reveal and radicalize the difference translation is conventionally thought to erase.