Rhizomatic Translation and the Censor State: Publishing Aim é C é saire in Estado Novo Portugual

: Poet, theorist and politician, Aim é C é saire produced, over the course of his literary and political career, a singular body of work equally noteworthy for its analysis of colonialism as for his poetic innovation. Publishing his texts in Portuguese under the censorship regime of the Estado Novo required therefore a series of delicate decisions on the part of editors in order to avoid ofﬁcial prohibition. This article examines in particular the only Portuguese publication to make it to press during this period, an anthology of poems published by Snu Abecassis in 1970, and the relationship of this anthology to the broader C é sairean paratext, arguing for a “rhizomatic” understanding of the publication and translation, necessitated by the political

Translation is an act of transmission. The publication of a translation makes it possible for a new audience to enter into contact with a number of ideas or data which might otherwise be inaccessible. Doing so in a repressive, authoritarian regime poses unique challenges and requires strategical sensitivity by translators, editors and publishers. 1 The present case, that of a Portuguese translation and publication of an Aimé Césaire anthology, illustrates both the stakes and vicissitudes of such a mobilisation. Césaire, already established as both a poet and an anticolonial theorist, offered considerable potential interest to Lusophone readers, especially in Portugal, which enforced political and cultural suppression through its secret police, and its African colonies in Mozambique and Angola, then already engaged in wars of liberation.
Two Portuguese-language Césaire texts made their way to press under the state censorship regime, overseen during this period by the DGS (Direção-Geral de Segurança; General Directorate of Security). In 1970, a selection of poems, Antologia poética, was published by the legendary figure of Ebba Merete "Snu" Abecassis via her press, Publicações Dom Quixote. This was followed in 1971 by a complete translation of Césaire's Discourse on Colonialism (Césaire 1971). The latter was banned by the authorities, while in the former case, the publishers successfully evaded censorship. In light of this, the poetry anthology offers important retrospective insight into the particular censorship regime, but more broadly, the navigations of editors, translators and publishers in an authoritarian context.
In addition, the two Portuguese texts serve to add to the broader discussion of the early Césairean "paratext" (Watts 2000). Early translations and anthologies of Césaire in the Spanish-speaking world have been well documented, making it possible to glimpse how these tangible publication projects interacted with other, interlinguistic Caribbean contexts (Maguire 2013;Rodríguez 2007;Seligmann 2019). A related paratextual analysis reveals that a small, popular anthology, produced by an editorial team keenly aware of state censorship mechanisms, engages not only concerns about translation, but a wide array of texts in other languages, and depends on the awareness and tactical decisions of not only the translator, but publishers, scholars and even clandestine book couriers, which I cast in terms of a "rhizomatic" network.

Censorship and the Estado Novo: A Conspiracy of Silence
Government censorship under the Estado Novo 2 was organised under the auspices of the Secretariado de la Propaganda Nacional (renamed Secretariado Nacional de Informação, Cultura Popular e Turismo in 1945), and overseen jointly by the political police-the PVDE (Polícia de Vigilância e Defesa do Estado) 1933-1945, later the PIDE (Polícia Internacional e de Defesa do Estado), from 1945 to 1969, and finally the DGS-who wielded plenary power. Unlike the daily press, subject to regular and pre-emptive screening, censorship of books was far less systematic, especially in cases of highly literary or foreign-language books (Seruya 2010, p. 129). The political police seized books from printers, publishers, booksellers, private homes, and monitored their circulation in the postal service, posing potentially disastrous consequences for all parties with a financial stake in the process (Alvim 1992, p. 3).
The range of censored publications was vast and standards were often capricious. Recent scholarship, led by Teresa Seruya (Seruya and Moniz 2008;Seruya et al. 2009;Seruya 2010Seruya , 2013Seruya , 2016, and the online bibliography of banned books maintained by Maria Luísa Alvim reveal a number of subjects and intellectual currents targeted for censorship. 3 Likewise, the Torre do Tombo censorship archives (now digitised) 4 and archival research reveal the sometimes complicated rationales for banning or authorising individual books (Seruya 2010, pp. 132-36). Topics liable for prohibition include primarily: Marxism, sexuality (libertinism, like the Marquis de Sade-a 1966 translation of La Philosophie du boudoir was confiscated, as was a Portuguese edition of Henry Miller's Sexus, but so were the Kamasutra and books of popular medicine dealing with sexual topics), anticolonial movements (in particular those of Africa, Vietnam and Cuba; Frantz Fanon's Les Damnés de la terre), 5 pacifism (James Thurber's The Last Flower was published by Dom Quixote for Christmas and marketed as a children's book in a failed attempt to avoid censor scrutiny; see Pinto 2011, p. 129; Report n • 8974) and domestic politics. Similarly, contemporary dissident voices attest to the breadth and arbitrary functioning of the censoring authorities in all domains of cultural production. Not only were serious works scrutinised, but seemingly neutral, even trivial pieces faced censorship. Among these included lyrics to Fado songs and personal ads (Soares 1975, p. 51), popular theatre (Santos 2006, p. 12) and horoscopes, especially, it was believed, for Taurus, since that was Salazar's sign (Pires 1972, p. 24). The breadth of censorship by no means implied a corresponding depth-often the title of a book was more important than any of its contents. Portuguese publication of Norman Mailer's novels (including a Dom Quixote translation of The Armies of the Night) were threatened with bans due to a suspicion of Vietnam war opposition-despite the author's actual position and harsh critiques of the anti-war movement-solely based on the title of his novel Why are We in Vietnam? (Pinto 2011, p. 129). Another, meticulouslydocumented case, that of the banning in Portugal of a Brazilian translation of C.R. Boxer's Race Relations in the Portuguese Colonial Empire, 1415-1825, shows that just a preface could also be sufficient evidence for a book's suppression. Here, a censor closes his report by 2 The one-party Portuguese regime, described by scholars as either Catholic corporatist or fascist, which governed Portugal from 1933 until the Carnation Revolution of 1974 and is primarily associated with Prime Minister António de Oliveira Salazar, who ruled until his incapacitation in 1968. 3 In addition to sources which consider Portugal, recent scholarship on translation and censorship either generally, or spanning multiple national contexts include (Ní Chuilleanáin et al. 2009;Seruya and Moniz 2008;Billiani 2014;Sherry 2015). 4 The Relatórios de livros censurados is consultable via the following URL: https://digitarq.arquivos.pt/details?id=4331838. Individual reports will be cited by their four-digit report number. In addition to these reader reports, the archive also preserves final decisions, which sometimes overrule readers, or appeals by the readers to higher authorities. 5 See report n • 9198. stating, "Reading the Prologue [to Brazilians] is enough to conclude, in my opinion, that the sale of this book should not be authorized" (cited in Terenas 2008, p. 35). 6 The role of the censor was not only to impose silence on a certain number of "forbidden" subjects, but also to reinforce the image of a population and a state in perpetual, harmonious accord. This need became increasingly urgent, from the point of view of the regime, as Portuguese economic development lagged progressively further behind the rest of Europe. The censor thus played a central part in perpetuating the consoling mythology of "o viver habitualmente," the Salazarist slogan which emphasised a depoliticised family life, and resistance to the consideration of alternative regimes or the possibility of progress. Consequently, every decision of the censor is meant to serve the aims of a principle-never made explicit but operating in the background of every deliberative process-that Portuguese society was one in which "all conflict had theoretically come to an end" and so "nothing which attested to conflict's persistence could be allowed to be published" (Costa Pinto 1999, p. 20). 7 This logic of obscuring conflict could be applied to seemingly broad and innocuous works, such as the hardly incendiary Poemas da liberdade: De Dante a Brecht. This anthology of translated poetry was banned despite the censor's admission that "there seems to be no reason or valid motivation for prohibition" of a work which amounts to "vague and undefined aspirations and calls for freedom and against tyranny [ . . . ] which we also despise" (Report n • 8659). 8 For authors, explicit intimidation and the looming danger of imprisonment were accompanied by the insidious tendency towards self-censorship. Writing from London, the novelist José Cardoso Pires argued that this process had deeply penetrated Portuguese cultural production: "The censored pen, seized up in a routine, begins to ossify, become sclerotic, increasingly limited to safe, no-risk gestures" (Pires 1972, p. 242). One can understand then both the need for, and potential disruption posed by a writer as jarring and mordant as Césaire as an antidote to this "sclerosis." Césaire's work, in this light, could become not only a source of new and pertinent ideas, but also through his poetry, which was well suited to break through any doctrinaire routine that could shake his readers and the general public to the core of their assumptions, "like an ulcer, like a panic" showing them "images of catastrophes and freedom and fall and deliverance, devouring endlessly the liver of the world" (Césaire 1978, pp. 7-8).

Navigating Censorship: Publishing Antologia poética
By the time of the Césaire initiative, Publicações Dom Quixote had already established itself as a publishing house willing to push limits and risk censorship. Founded in 1965 by Snu Abecassis, the press took its name from the idealism of a "quixotic" undertaking, according to Carlos Araújo, who became its literary director in 1966. The enterprise reflected the ambitions of its indomitable founder. Born in 1940 to Danish journalists, Ebba Merete Seidenfaden, or "Snu," as she was called in her Portuguese years, moved to Lisbon with her husband Vasco in 1961. A singular figure of her era (and subject of a 2018 eponymous biographical film), Snu entered the popular imaginary due in part to her time as collaborator and lover to the reformer and future prime minister Francsico Sá Carneiro, and for her death in the plane crash which killed the couple less than a year after he assumed office.
The mission of Dom Quixote, as announced on the cover of its first publication, was to respond to the "lack within the Portuguese market of books both inexpensive and of high quality, able to reach a wide and diverse readership" (Pinto 2011, pp. 137-38). The press, even when challenged by initially low sales, held to this goal. They kept costs as low as possible-cheap paper, covers limited to two colours, allowing for sale prices starting at 20 escudos 9 (0.10 euros). To fulfil the other half of the mission, Publicações Dom Quixote took on both "risky" subjects (in the eyes of the regime), and prestige authors from around the world. In the former category, the press released a series on potential futures for former colonies as well as postcolonial theory (Cambodia, 1970;10 Vietnam before Peace, 1971), socialist movements (Chile: Reform or Revolution, 1970;Cuba and Socialism, 1971) 11 and democratic movements (Czechoslovakia at the Hour of Democratisation, 1968; The Revolt of Black Americans, 1971-a selection of texts by James Baldwin, Andrew Kopkind and Tom Hayden; And After Franco?, 1971) 12 . Parallel to these, Dom Quixote systematically and strategically published a translation of the Nobel laureate each year, from Miguel Ángel Asturias to Yasunari Kawabata, with a rapidity that inspired rumours-false, according to Carlos Araújo-that Snu had advanced knowledge of the deliberations of the Swedish Academy thanks to her Scandinavian connections (Pinto 2011, pp. 121-22). 13 Even if her foreign ties protected her (according to Araújo, this allowed her significant flexibility compared to a native-born Portuguese), Snu and Dom Quixote were hardly free from risk. After a first ban in 1968 (a selection of texts on May '68 by Jean-Paul Sartre, Daniel Cohn-Bendit and Henri Lefebvre), the press would see five books on the censors' index in 1969 and Snu was forced to appear before censor committees with increasing frequency. The choice, then, to publish an anticolonial author, himself a former Communist Party member, would require a certain degree of finesse. Just one year after the poetry anthology appeared, the Portuguese version of the Discourse on Colonialism was confiscated. The report by DGS censor Simão Gonçalves summarily and disdainfully illustrates the kind of reception Dom Quixote was risking by publishing Césaire: "The author is black, communist, and was for some time a French deputy. The text is nothing more than a diatribe against Western civilisation, in the form of a pseudo-defence of black, Oriental or Indian civilisation. To be prohibited" (Report n • 9253). 14

Editing and Translating the Anthology
The editorial process of the anthology was overseen by Armando Silva Carvalho, who had previously translated Kawabata's Snow Country for Dom Quixote in 1968. Carvalho, whose third volume of poetry had just appeared, had proven himself to be a "mordant, ironic, even devastating poet" (de Vasconcelos 1998, p. 12). These qualities were well suited to Césaire's own poetic orientations, and especially important given the challenges posed by the translation.
The poems chosen (see Table 1) for the anthology represent all five of Césaire's collections to appear before 1970, not counting the genre-defying Cahier d'un retour au pays natal. No structure of the sections or their order of inclusion is evident, and no 9 The cover price of the for the Césaire anthology was 30 escudos. On the economics and of both hardback and paperback books, and the corresponding relationship to censorship, see (Seruya 2010, p. 123). 10 Report n • 8975. 11 Report n • 8975. 12 Report n • 9340. 13 There is also some indication that the press was particularly adept at navigating the censorship process. On at least two occasions, individual censors were overruled when their reports recommended prohibition: Bolívia depois de Guevara (9004) and Cuba e o socialismo (8975) were both ultimately authorized despite scathing reports. In other cases affecting Dom Quixote (as in the prohibited volumes 8774 and 8785 on Spain and China, respectively), reports are handwritten rather than typed, contrary to the usual procedure). 14 This report can also be found in a non-paginated facsimile insert, reproduced by the daughter of the original editor, accompanies the re-edition of 2012, as part of the former's exposition "1975" at Modam Luxembourg. The report, the original of which is housed at the Torre do Tombo archive, was delivered 8 December, 1971-a mere two days after the book's publication.
editorial indication explains a rationale of the selection process. 15 Ferrements, the most recent collection of Césaire's to appear, is by far the best represented, accounting for slightly more than half all the anthology's poems. The translation itself faced a number of difficulties inherent to poetry, but especially to a number of specific features of Césaire's work. According to his English-language translators, Clayton Eshleman and Annette Smith, Césaire is particularly "devilish" with respect to translation. His syntax-modelled according to them on the more fluid word order of Latin-presents the reader with "a process approximating coitus interruptus" (Eshleman and Smith 1983, pp. 25-26). 16 In addition, Césaire's rarefied and technical vocabulary poses a series of lexical problems for translators, as does his penchant for neologising. His lexical choices, as Eshleman and Smith explain, "keep the translator bent over various encyclopaedias, dictionaries of several languages (including African and Creole), botanical indexes, Atlases, and history texts" (Eshleman and Smith 1983, p. 26). This in turn prompts the question of "to what extent the esoteric tone should be respected 15 The selections overlap somewhat with the poems included in Kesteloot's earlier anthology (to which the publisher had access-see further on), though differ enough to allow us to take at face value that they were indeed chosen by the translator. 16 Or, according to Charles Edgar Mombor, Césaire's syntax, "an avalanche of gleaming red [rutilant] words on the poet's lips" has the effect of destabilising the sentence itself and consequently bringing to the fore "the cry for liberty of a people murdered by colonisation." His syntax makes Césaire "the spokesperson for a new way of speaking the French language, associating sonorities, colours and the re-evaluation of words within a language which rings out as the symbol of a revolution" (Mombo 2012, p. 288).
in the English," since the terminology does not simply name obscure plants or animals, but rather names known entities in a particular lexical context, meaning an otherwise unknown word signals "nothing more than a morning glory (convolvulus) or a Paraguayan peccary ('patyura')" (Eshleman and Smith 1983, p. 26). Indeed, flora and fauna are not only challenging, they are of particular importance to Césaire's oeuvre, so "one must be wary to betray neither Césaire's use of vegetable forms as a symbol of black culture nor his concrete interest in Caribbean flora demonstrated in Tropiques" (Eshleman and Smith 1983, p. 26). 17 Césaire's "interest" in these categories forms a systemic, deliberate project. In a retrospective interview with Jacqueline Leiner on the 1941-1945 run of Tropiques, Césaire outlines how his strategic deployment of highly specific vocabulary, making use of both local and Latin terms for native plants and animals, produced a "shocking [boulversant]" effect for his readers: We did not want to produce an abstract cultural journal, but rather capture as well as possible the reality of Martinique within the context of Martinique-to situate it properly. We wanted this journal to be an instrument which could make it possible for Martinique to re-centre itself. We realised that there was nothing of the kind in the field! Absolutely nothing! So we decided to study, systematically, the flora, fauna, etc. Even a very repressive regime can't stop you from doing that! (Césaire and Leiner 1978, p. IX, authors' emphasis) Leaving aside for the moment the question of censorship (I return to this in the following section), Césaire's own orientation underscores the potential value of heavily annotated and explanatory "thick" (Appiah 1993) or "prismatic" (Anderson 2019) translation strategies. Eshleman and Smith largely reflect the former approach, while the latter, which emphasises the delicate negotiation of conveying understanding, without overtaking the prerogative of autochthonous authors, finds an echo in Césaire's instruction to Lilyan Kesteloot that his poetry be interpreted "à l'africaine" (Césaire and Kesteloot 1962, p. 182).
For his part, Armando Silva Carvalho-translating for a popular audience, with concerns for publishing cost and censorship making extensive annotation impracticable -tends to employ a "domesticating" approach for Lusophone readers, eliding a number of geographical and cultural references (Venuti 2008). In the translation of "La parole aux oricous," for example, oricou, an endemic African species, becomes in the Portugues an "abutre"-a generic term for a vulture. 18 In the translation of "Ode à la Guinée," "la mourre" (the Mediterranean odds-and-evens hand game morra, which famously appears in Apollinaire's "L'Ermite") becomes "as ascondidas," a kind of hide-and-seek. In "Afrique," the porcine terms "boutis" and "groin" lose their specificity, in exchange for more generic animal vocabulary. Regarding Césaire's idiosyncratic syntax and characteristic supersaturated verses, the translator sometimes maintains a word order which closely mirrors the original. But at times, he modifies, as in the case of the translation "À la mémoire d'un syndicaliste noir." In this case, Carvalho "smoothes" Césaire's use of jarring syntanx within the poetic line. The original reads: tu plissais les paupières tu les plisses aujourd'hui tu ne parlais guère tu ne parles guère maintenant. (Césaire 1983, p. 338) The repetition of the two verbs (plisser and parler), with modification but without separation by either punctuation is a typical example of Césaire's supersaturated syntax, creating a dramatic, claustrophobic text, an effect heightened by the relentless sequence of plosive p's. Eshleman and Smith's English preserves this structure, while also maintaining some degree of the relentless assonance: you used to squint you're still squinting 17 On the importance of the botanical terminology, see also: (Césaire and Kesteloot 1962, pp. 52-54). 18 Compare this to Eshleman and Smith's more recent English rendering, which preserves the unfamiliar term, but provides a gloss which explains the terms as a "popular name for the great black African vulture" (Césaire 1983, p. 404).
you hardly ever spoke you hardly speak any less now. (Césaire 1983, p. 339) Carvalho's Portuguese, however, adds and subtracts to Césaire's structure: tu franzias as pálpebras como agora acontece não dizias palavra como não dizes agora. (Césaire 1970, p. 94) This rendering (which I approximate in English as "you folded your eyelids like you do now/you didn't say a word like you don't say [any] now") loses some of the force of repetition, with the conjunction 'como' smoothing the halting rhythm of the original. 19 These syntactical and lexical choices represent pragmatic solutions for grappling with a poet whose technical innovations and declared intention to "keep poetry living" make his verses so difficult to translate, and as such do not necessarily diminish the overall editorial project. Rather, these decisions should be joined to another set of concerns, namely a sensitivity to the possibility of censorship, including the concerns over which poems are included in the anthology, and which are excluded. Absent, for example, is "Pour saluer le Tiers Monde/à Léopold Sedar Senghor," which celebrates the transition to sovereign African states and expresses a new "exultant" tone with respect to the continent (Munro 1999, p. 28). This poem would have been more apt to attract the attention of censors, less susceptible to recognising subtext (recall the case of Mailer's Why are We in Vietnam?), simply because the African independence leader is explicitly mentioned in the title. The titular poem of Corps perdu also addresses the theme of African self-determination, concluding in the famous verse, "I shall command the islands to exist [je commanderai les îles d'exister]" (Césaire 1983, p. 230). This poem, too, is absent, despite its privileged place in its collection. The poem which gives its name to Les Armes miraculeuses, with its vivid rendering of violence in the colonial context ("Le grand coup de machette du plaisir rouge [ . . . ]") is likewise excluded. Finally, the three poems which explicitly invoke the Civil Rights Movement ("Lynch," "Mississippi" and "...sur l'état de l'Union," with its famous evocations of the slain child Emmett Till) are left out of the anthology, removing an important concrete reference for political engagement. These limitations are regrettable, but the mission of both the specific poetry series and Dom Quixote as a whole is only to provide a first step for the discovery of authors and debates. The press characterises this mission as one of "cultivating the love of poetry among the broadest audience possible and inspiring the reader to encounter the poets represented in this collection in greater depth" (Césaire 1970, front matter).

The Sprout and the Rhizome: Translation under Surveillance
When considering this spirit of "inspiring the reader [ . . . ] to encounter in greater depth," it becomes possible to place the small translation anthology into the context of both the surveillance imposed upon its production and a series of complex ways in which it joins to other texts and writers. Indeed, a still further indication of the delicate balance the series' editors and publishers had to maintain can be found on the anthology's back cover. Here, we find praise from Lilyan Kesteloot, which reads: "We cannot cease to be impressed by the number and extreme variety of symbols in the poetry of Aimé Césaire. They manifest with such vigour that they themselves form a rich and coherent cosmogony." 20 The citation of Kesteloot is not surprising-the Congo-born Belgian scholar established herself, with Les Écrivains noirs de langue française: naissance d'une littérature, as perhaps the most important European scholar of African literature (Midiohouan 2002), and one keenly focused on Césaire in particular, whose work she had presented in an earlier anthology of her own editorial direction (Césaire and Kesteloot 1962). With the many pages Kesteloot had written on Césaire, the choice of these words, however, would not appear to be an obvious one. Cosmogony aside (which would serve to mark the book as a highly literary one, thereby lessening censor scrutiny), the blurb emphasises the aesthetic features of his poetry over any content, either implicit or explicit, of a political bent, a facet of Césaire's career she had discussed extensively.
This minimisation of politics is all the more noteworthy since Kesteloot herself had found, within the very aesthetic phenomenon she attributes to Césaire in this short passage, a "weapon" which could operate in an analogous fashion to how Césaire saw some of the subject matter of Tropiques functioning-one of the "thousand ruses" which formed a veil of obscurity between themselves and the then-Vichy-controlled censors (Kesteloot 1977, p. 229). 21 In this environment, surrealism served as a protective "citadel" (Kesteloot 1977, p. 212). Indeed, the translated passage included as the Antologia's blurb is taken from her short study and anthology of Césaire's work dealing with just this subject (Césaire and Kesteloot 1962, pp. 56-57). In the sentences which immediately follow, she argues that for Césaire, surrealism was "the boring machine which excavated deeply and brought back up original and inestimable combustible material which lay below the surface of the Martiniquais unconscious" and "fed the militant's sense of rebellion" (Césaire and Kesteloot 1962, p. 57). Even more directly, she was well aware of the Césaire's tactical use of obscurity and unusual or surreal imagery, which acted as: a code language in which the journal's editorial team could express their sense of revolt and their hope to transmit without too much fear of reprisal. This literary form thus became an indispensable vehicle for ideas of opposition, the only one adequate to the situation, since it was sufficiently allusive, obscure even, to prevent authorities and enemies from immediately seizing the publication. 22 (Kesteloot 1977, pp. 232-33) Consequently, the anthology's small blurb, which characterises Césaire as a vigorous and creative poet, de-emphasises his political context, but does so in such a way that points back, however indirectly or subtly, to the very tactics Césaire initially used to advance political goals in his own context of surveillance.
I have catalogued a number of ways in which the Dom Quixote anthology was "limited," but as the use of the carefully and tactically-selected Kesteloot citation further shows, the choices made by the translator and publisher are only limitations if the Antologia is viewed as a stand-alone text. Instead, I would propose, following Césaire's own botanical tendencies, to classify the translation as part of a larger rhizomatic system, in the terms developed by Deleuze and Guattari (1980) and Glissant (1990). According to this model, the relationship between source text and translation can be recast into an "un-centred, non-hierarchical" network of connections between various people and other books, not merely the source text, and in which "any point can connect to any other point" (Deleuze and Guattari 1980, pp. 31-32). Rather than an outright "transplantation" of the original into foreign soil, or a further branch stemming off from a single root, the humble sprout that is the Antologia maintains and builds connections to a network of other potential sources in a censorial environment where such connections needed to remain underground. The successfully navigated publication of the Césaire anthology underscores the possibility that a constrained edition might nevertheless serve as an introduction to the oeuvre of a writer whose works one will be later inspired to seek out in neighbouring languages. Through established networks, left-wing intellectuals and militants had access to original sources. For example, the publisher of the political Cadernos para o diálogo series, António Daniel Abreu, attests that he came into possession of the 1955 volume of Présence africaine (and therefore the second French edition of Césaire's Discours sur le Colonialisme) through a "clandestine" courier or "passeur" with ties to the banned Portuguese Communist Party (Céasar 2012). Each such contact would extend the rhizomatic network to which the anthology belonged.
Additionally, at the time of the Dom Quixote publication, translations of Césaire circulated more broadly in the Spanish-speaking world (thanks to publishing in countries where Césaire's political positions would not meet censorship). Lusophone readers -for whom written Spanish is highly intelligible-could perhaps access a copy of Césaire texts including two translations of the Cahier d'un retour au pays natal (one translated by Lydia Cabrera with illustrations by Wifredo Lam published in Havana in 1945 and a bilingual edition published in Mexico City in 1969, translated by exiled Catalan poet Agustí Bartra), or a poetry anthology (more extensive than its Portuguese counterpart), edited and translated by the Chilean poet Enrique Lihn. In a similar vein, readers approaching Césaire through the Antologia would not find the name of Senghor in a translation of "Pour saluer le Tiers Monde/à Léopold Sedar Senghor," and thus no mention of a man who represented, as head of a sovereign African government, a name of obvious interest to both the censors and to the many Portuguese growing increasingly tired of fighting colonial wars against such movements. The anthology does, however, give readers, the far less famous name of Louis Delgrès, the Guadeloupe resistance leader who fought the Napoleonic reintroduction of slavery, and the unnamed (but identified as the Martiniquais Albert Cretinoir) "Syndicaliste noir." 23 These figures can serve as proxies, bridging the gap between what is able to pierce the surface of censorship, and what (or whom) must remain underground.
This case study of the edition and the navigation of the censors produces only a single, limited narrow portrait of Césaire, but which also serves as a first point of contact with a vast rhizomatic network. Not only does such a publication fulfil Dom Quixote's stated and subtly subversive "initiation to discovery" mission, it also makes it possible to grasp the parts and parties of the larger textual production. This includes the translator and the savvy publisher, but also scholars such as Kestleloot, translators of Césaire into other languages, and clandestine document runners. 24 It may even include, following Piotr Kuhiwczak, the censorship apparatus itself, comprising the readers and policy makers, but also the broader standards and expectation of readers in the censorship society (Kuhiwczak 2009, p. 47). 25

Conclusions
However inclusive one considers the collaborative process of this particular translation (or any translation for that matter), the rhizomatic model offers the utility of drawing together paratextual elements and participants, with potential insights applicable to translation generally, and specifically in a censorship context. Glissant's famous reformulation of Deleuze and Guattari's earlier evocation of the rhizome's disruptive capacity for deterritorialisation and re-territorialisation (Deleuze and Guattari 1980, p. 17) casts the concept of "rootedness" in new light. While the singular taproot "takes all for itself," the concept of the rhizome "keeps the fact of rootedness (enracinement), but renounces the idea of a totalitarian root" (Glissant 1990, p. 23). In the mere fact of translation, this broadens the picture of transmission from the "root" of a source text growing in unified fashion into an idealized target-language reproduction. 26 When translation under censorship is approached in this broader and interconnected way, an understanding of the constraints imposed on translators and publishers brings to the fore the difficult decisions and creative navigations employed by these parties in the shared goal of transmission. This can be true whether censorship is considered either in the literal sense, as this article has, or is taken to 23 For further discussion of Césaire's use of these figures, see (Rumeau 2014). 24 One could draw a parallel between such a figure and another term from botany: a runner (or stolon) is, like a rhizome, a mode of plant propagation which offers an alternative to the more exclusive, hierarchical root/tree model. 25 Kuhiwczak cites a particularly dramatic formulation of Miklós Haraszti's Velvet Prison to this effect: "A new aesthetic has emerged in which censors and artists alike are entangled in a mutual embrace" (Kuhiwczak 2009, p. 55). 26 Such considerations, following Glissant, may pertain with particular force to the Caribbean. See (Forsdick 2015). include a broader set of societal and cultural standards or expectations. Such an approach resists, as Maria Tymoczko argues, "the simplistic binary notions of victims and heroes in the translation process" which constructs a reductive division of "good translators who are passive victims of evil censors [ . . . ] and on the other hand heroic translators who fight the evil empire" (Tymoczko 2009, p. 30).
Moreover, though the case of the Césaire Antologia has as its backdrop the particular censorship regime of Estado Novo Portugal, it may offer insights into others, and the gulf between explicit censorship regimes and more liberal ones may be less great than is generally assumed (Ní Chuilleanáin et al. 2009, p. 2). States without formal censorship administration may still have juridical constraints (laws governing incitation, obscenity, defamation of individuals or groups, historical memory, name suppression, reporting of certain demographic data, copyright, etc.), but also a broad set of pressures which fulfil similar functions to the censorship process, from the exigencies of algorithmic marketing to prevailing societal attitudes. All of these pressures can be, as under an explicit censorship regime, anticipated and adopted by self-censoring authors, editors and translators. Future studies adopting a rhizomatic approach to the multiple steps and participants of a published translation that incorporate new roles and processes (the recent advent of sensitivity readers is one such example), may yield promising results.