[This essay appeared for the first time in the Canadian Review of Comparative Literature. 38.1 (March 2011): 40-56]
"Sans une étude génétique du Cahier qui n'a cessé de se modifier le long des ans jusqu'à sa forme définitive, on ne peut prétendre à une lecture approfondie du poème" Pestre de Almeida, "La cosmogonie césairienne, fête d'Eros"Writing roughly two decades ago on the evolution of the poem from its 1939 appearance in a Paris journal to the 1956 Présence Africaine "édition définitive," Thomas Hale acknowledged, with a finder's glee, that it had become "increasingly apparent that the Cahier d'un retour au pays natal is the product of a complex dialectic between the poet and his text, between the poet and his experience." The story of how this complexity became "increasingly apparent" serves as a paradigm for the critical attention that the history of the text has received. Furthermore, it is well worth examining as a form of history of the history of the text. Already several years after his seminal "Bio-Bibliography," Hale came across his ‘discovery’ through a fortuitous accident of the kind one perhaps hears described more often in popular accounts of scientific breakthroughs. In the course of editing an article by Gerard Pigeon, it became “increasingly apparent” that Hale, using the 1947 Brentano's edition, and Pigeon, using the 1947 Bordas edition, did not have matching references. Hale describes the episode thus:
Until recently, it was assumed that these two 1947 versions were identical. But in the course of editing [Pigeon's article] for Cahiers Césairiens, we discovered numerous discrepancies between his citations and our own text, the Brentano's edition. After a comparison of the two editions, Bordas and Brentano's, it became apparent that we had two different stages in the evolution of the poem. ("Two Decades" 190)The identity of the two 1947 versions was not the only assumption made by scholars two decades ago, according to Hale. In retrospect of another project, he confesses,
In our analysis, we assumed that there was really only one version of the poem, that published in 1956, the definitive edition, which contained a few modifications of the version published in 1947. Further, we assumed that what appeared in the Parisian review Volontés in 1939 were, as many critics have suggested, merely fragments of the later, definitive edition. (“Two Decades” 187)I emphasize here Hale's avowed dependence on assumption to highlight a trend in the study of the poem, due in part to the lack of access to the 1939 text and in part to the privileging of the "definitive" edition. Hale was not the first to notice the discrepancies, of course. In the second issue of Cahiers Césairiens in Fall, 1975, Rodney E. Harris had written a delightful piece on the permutations of the poem from 1939 to 1956, teasing out a Césaire fated to make constant changes "due to his desire to be scrupulous in his search for exact truth in details." All the same, we can understand how Hale's ‘discovery’ of the "two decades, four versions" was delayed. When the "definitive edition" came out in 1956, there were more pressing uses for the text than satisfying our genetic yearnings. Some decades later, when Hale published the account of his anagnorisis, he was riding the swell of an international Césaire reception which could now give itself the luxury of attending to structural concerns and intertextualities. As he suggested, it was the exchange of references over the Atlantic, prompted by the fourth (and last) installment of the Cahiers Césairiens, which led to the realization that the Bordas and the Brentano's texts were different texts altogether. For Hale, the textual evolution of the poem marked the movement from a personal to a collective consciousness. His history of the text astutely paralleled at the macro level what he had already pointed out elsewhere as the internal trajectory of the poem. With this in mind, at the end of his piece he calls on scholars "to follow the signs in the four versions" (“Two Decades” 195). It is perhaps sad, but very telling, that Hale's invitation did not guarantee a marriage between hermeneutical and textual analysis in the decades which followed his study; and what ultimately made his study paradigmatic was not even mere velleity for the scrupulous attention to textual transmission that the study calls for, but the fact that the study itself came in the wake of a series of assumptions. On the one hand, many subsequent critics continued to take for granted much that the work of Hale and Harris had already belied. In many of these studies, "the poem" became synonymous with a timeless 1956 text. On the other hand, many well-meaning and conscientious critics who heeded Hale's call still managed to inadvertently promote some serious misconceptions. This is the case of Lillian Pestre de Almeida, for example, who in the 1980s became the leading advocate for a genetic reading of the text. In 1984, on the occasion of Césaire's 70th birthday celebrations, two collections of essays were published with important contributions by Pestre. One of these studies, "Les versions successives du « cahier d'un retour au pays natal »" remains today the most complete and accurate published tally of textual variations in the poem. The other, "La cosmogonie césairienne, fête d'Eros," recruits this kind of close attention to textual transformation in the service of a close hermeneutical reading of a popular passage (“vienne le colibri…”). The former study is divided into two parts. In the first one, Pestre de Almeida examines variations in substantials (i.e., nouns, verbs, etc.) and accidentals (i.e., commas, dashes, etc.) regardless of changes in the stanza sequence, which as the reader will shortly see, define the evolution of the poem. In the second part she compares side-by-side, in three columns, the 1939 Volontés, the 1947 Bordas and the 1956 Présence Africaine, respectively. In order to compare the texts, Pestre de Almeida distilled the stanzas into exiguous thematic descriptions. The first stanza of the 1947 Bordas became, for example, "ouverture et proposition initiale" ("Versions successives" 79). Although she included the 1947 Brentano’s in the first part, she excludes it from the side-by-side comparison without an explanation. We must look to her "fête d'Eros" piece for the reason. She explains,
Volontés et Bordas sont, sémantiquement et structuralement, beaucoup plus proches du texte définitif que Brentano's avec de nombreux déplacements de strophes à l'intérieur du poème. De Volontés (1939) à Bordas (1947), de Bordas à Présence Africaine (1956), il y a soit des ajouts, soit des suppressions sans que l'ordre des passages subisse des modifications. Dans Brentano's (1947), par contre, en plus des additions et/ou suppressions , on observe des déplacements d'épisodes et de groupes de strophes provoquant des infléchissements de sens importants. (334)Because she took the "déplacements de strophes" to be a feature of the 1947 Brentano’s text and not, as I will soon demonstrate, a feature of the 1947 Bordas, the former became an immiscible "anomaly" to be cast out from the conforming community. The Volontés and the Bordas were thus rendered canonical because they approached the ever-central "definitive edition." As a result, sustaining the fiction of a definitive edition, Pestre de Almeida inadvertently gave rise to a few assumptions that have carried over to the present day. In 2008, Pestre de Almeida revisits the problem posed by the Brentano’s edition in her book-length study of the poem and repeats and expands the same argument, adding another factual error along the way that further explains her conclusion that Brentano’s was an anomaly. In the book, she uses the passage that formed the center of her earlier “cosmogonie” piece as the example that supposedly demonstrates the anomaly:
Avec Brentano’s, au contraire, en plus des additions et/ou suppressions, on observe de nombreux déplacements d’épisodes et de groupes de laisses provoquant des infléchissements de sens importants. Un seul exemple suffira à nous convaincre : le passage de la cosmogonie rêvée (« vienne le colibri… »), dans Volontés, Bordas et Présence Africaine, se situe après la grande « révolution » intérieure du narrateur… The problem is, of course, that this passage is simply not present in Volontés as Pestre de Almeida assumes. To illustrate how her influence has led other scholars to perpetuate this mythical history of the text, we need only open another, relatively recent book-length study dedicated to the poem, Aimé Césaire: Cahier d'un retour au pays natal, by Dominique Combe. In an otherwise able and comprehensive study of the poem—in the 1956 sense of the word "poem"— Combe publishes a series of misleading statements based on his reading of Pestre's study: a) When introducing the 1947 Bordas edition after just having introduced the 1947 Brentano's, he declares "Les deux versions comportent peu de variantes." (39) The opposite is true: A simple word count reveals that there is a difference of 574 words between the two. This is a higher figure than if we were to compare the 1947 Bordas with the 1956 Présence Africaine, for example, which yields 278; b) After pointing to Pestre de Almeida as his source, he notes,
Qu'il soit simplement noté que, dans l'ensemble, Césaire a revu son texte dans le sens d'un enrichissement : la version définitive est nettement plus longue que la version primitive de la revue Volontés, selon un mouvement d'amplification et non de restriction, qui va bien dans le sens de l'épique. (40)This statement is misleading as well. Although it is true that the 1956 edition is far larger than the 1939, this "movement of amplification" has not been continuous. In fact, the largest of all the versions is the 1947 Bordas, tallying up to 9,873 words total. The 278 word difference we pointed out above between the Bordas and the 1956 Présence Africaine comes not from additions to the poem, but from the numerous deletions, or rather expurgations, which characterize the 1956 so-called "definitive edition." In fact, these deletions contribute to the argument that the 1956 text represents not an enrichment, but an impoverishment. Such an argument remains implied in James Arnold's recent revitalization of the issues at stake for a genetic study of the poem, "Césaire's 'Notebook' as Palimpsest: The Text Before, During, and After World War II.” Coincidentally published on the occasion of another Césaire anniversary, this time his 90th birthday, this study brings historical context to bear on the four different avatars of the poem we've discussed so far. While describing each of the editions, tracing out their different contexts and influences, Professor Arnold delineates a three-stage trajectory which goes from emphasis on spiritual/biblical language to the eruption of a surrealist poetics, to the erasure of spirituality/sexuality on a Marxian platform. Each of these stages corresponds to a particular moment in the poet's development and affiliation. At the moment when the text approaches sanctification as a "definitive" edition and for reasons that Arnold suggests, "Starting in 1956, [Césaire] took a step backwards that opens the Cahier to a flat or literal reading by situating it in the context of the Cold War." (139) We must add to Professor Arnold's study by clarifying that by no means do these attributes, religious/surrealist/Marxian, constitute the central feature of these different versions, but rather their excess. In other words, these are the elements that a textual study such as ours highlights because they constitute the work's temporal difference. We should not forget that the poem's central theme and approach remain unchanged: a radical critique of colonialism, assimilation and amnesia developed through highly iconoclastic formal innovations. Having qualified it thus, Arnold's historical analysis falls at the end of my history of the history of the text and serves as the point of departure for the rest of our work here. In the next section I will try to argue for the primacy of the 1939 text as a sort of gravitational center which gives coherence to the other versions. Following this, I will delve into a brief reevaluation of the trajectory from manuscript to the 1947 Brentano’s edition, tracing what we know of the production history of the poem during this period, in order to open the possibility for a dynamic reading of the poem that belongs to no one version isolated from the rest.
Pluralitas non est ponenda sine necessitate William of Occam
¿pa’ que complicase la vida? Dominican idiom.The correspondence between Césaire and Breton, and between Breton and Yvan Goll, the exiled Franco-German poet who along with Breton was instrumental in getting the poem published in New York, confirm what many critics already adduce based on circumstantial evidence and historical hearsay, mainly that the 1947 Brentano's, translated into English by Lionel Abel under the editorial eye of Goll, was drafted before the 1947 Bordas edition. The texts themselves provide us enough internal evidence that the 1947 Bordas is indeed a later text, to make the archival evidence hold. At the outset, it seems evident that the additions belonging to the 1947 Bordas were made in excess of the additions belonging to 1947 Brentano's. Some of them were even inserted between text that already constitutes an addition to the 1939 Volontés, the best example of this being the all-too-famous "voum rooh oh" chant, which comes in between a passage already inserted in Brentano’s. (Fig. 1) If in our textual endeavors we prefer to follow the lesson of Occam’s razor, which favors the simple solution over unnecessary complexity, we can agree that it is more likely that in the case of this passage we move from absence to addition, rather than from the absence of the passage, to the addition of the passage and on to the deletion of the passage. Furthermore, several corrections were made in the 1947 Bordas of "typos" which remained consistent between the 1939 Volontés and the 1947 Brentano's. Finally, the 1947 Bordas permanently deletes material present in both the 1947 Brentano's and the 1939 Volontés, including the whole line, "Un nègre à la voix embrumée d'alcool et de misère." Therefore, when Pestre ascribes "nombreux déplacements de strophes" to the 1947 Brentano's edition she omits one crucial detail: The supposedly displaced stanzas in question are all additions, i.e., they are simply not part of the 1939 Volontés text. Once we agree on the simple line of succession between editions, the absence of the "displaced stanzas" from the 1939 implies that no part of the original 1939 text changes position in relationship to itself in the 1947 Brentano's, a fact which clearly suggests that there is no displacement of stanzas in the 1947 Brentano's as Pestre de Almeida suggests, only additions. On the contrary, it becomes clear from an analysis of the sequence of texts that the 1947 Brentano's added several stanzas in a particular order, and only during the preparation of the 1947 Bordas, were these additions then transposed around a passage in the middle of the original 1939 text. I am calling this 1939 passage around which the exchange takes place, “The Bridge.” (Fig. 1) Although we don’t have direct evidence confirming whether Césaire himself was responsible for the order of stanzas in the Brentano’s edition, we have textual and circumstantial evidence to suggest it was indeed him. The largest addition to the Brentano’s, marked 1947 C in our Fig.1, and beginning with the lines “O terre mer almée,” comes almost in its entirety from the short poem “En guise de manifeste literaire” published in the 5th issue of Tropiques, April 1942. In the process of incorporating it into the Cahier, Césaire added three blocks of text to the short poem, one at the beginning, one at the middle and one at the end. A short passage was deleted from the short poem and several regularizations took place. The amount of changes made to this small poem on its way to being incorporated to the larger poem are disproportionate in comparison with the scant changes made to the Volontés text for the same edition. A reading of the additions to the short poem suggests Césaire made these changes to adapt the short poem to its new environment. Then there is the matter of Césaire’s penchant for transpositions. As you can see from Fig.1, Césaire also shifted a small passage in the transition from the Bordas to the 1956 PA edition. Around the same time as he was working on the Brentano’s, Césaire was also working on Et les chiens se taisaient. A comparison of the manuscript of this text with the first published version, that of the 1946 edition of Les armes miraculeuses, shows no less than 64 transpositions. It seems likely then, that Césaire moved around the text when preparing the Bordas edition years after the Brentano’s text had been envisioned. Why is this important? Once we have begun to describe the history of the poem more accurately, do we learn anything about the nature of the changes? How are these migrations, additions and deletions relevant to our reading of the poem? In the chart below, the reader may notice that the structural sequence of the 1939 text remains intact in relation to itself throughout the transformation of the poem. This is even the case—excepting small, albeit important deletions—for the 1956 Présence Africaine text. In this last text, the pattern of addition and displacement continues, with purgation playing a more prominent role than ever before, but ultimately leaving the 1939 unmoved. The relative stability of the 1939 base text contrasts sharply with the highly volatile additions, which migrate and decay throughout the latter versions. We can conclude from the preceding that the 1939 text forms a veritable backbone or framework for the poem. Throughout the years, and apparently open to his editors’ suggestions, Césaire worked at the poem, trying out new sequences, adding material, etc., but never was the elder 1939 text "displaced." If there is a poem here, in the static sense of the word, it would have to be the unmovable bulk of the 1939 text—deletions and additions aside. Of course, the argument can be made that there is just no "Poem" here, just different texts sheltering under the useful myth of an umbrella title. No matter. Despite our base esemplastic tendencies to believe in the existence of a transcendental "poem" outside of its material manifestations, it remains clear that a re-evaluation of the 1939 text is in order. Once we have understood the relative instability of the additions in comparison to the first published text, even those of us who would still give primacy to the 1956 because it has been arguably the most widely read version and/or because it carried the author's final imprimatur, must agree that the 1939 text, still unmoved in the "final" version, forms what I would call a core of structural meaning. Once we shift our focus away from the 1956 and back to the 1939 “backbone” text, we are well poised to start a reevaluation of previous critical attempts to analyze the poem's structure, most of which have relied on the "definitive" edition. Apart from Lilyan Kesteloot's relatively complex structuring of the poem, the tendency has been to divide the poem in three. Peter Guberina, in his preface to the Présence Africaine edition already calls the poem “a poetic drama in three acts.” A range of other critics, from Thomas Hale (“Structural Dynamics” 166) to Maryse Condé, has also recognized in the poem a three-tiered trajectory. Notwithstanding differences in critical outlook, the formal breakdowns overlap substantially; usually, the conceptual structure of the poem ends up looking something like this: 1) Rediscovery, 2) Identification with the (African) Past, and 3) Metamorphosis and revolt. It is both this tendency to triangulate and its inherent vulnerability to condensation that has led many critics, starting with Sartre in "Orphée Noir" (1948), to invoke Hegel's dialectical movement from thesis, antithesis to synthesis as a privileged structure when discussing the poem. While tripartite readings might arguably work for each version taken in isolation, the linear/teleological narratives of the Hale and Condé variety lose their explanatory power once we look at the history of the text. This is so for two preliminary reasons: 1) The sequences expand or reconfigure with the different editions, and; 2) Linearity or linked seriality is excluded from the geometries of the poem in all of its versions. This much can be deduced from Jérôme Game and Jean Khalfa's study of the formal properties of the poem, where linearity itself is variously gainsaid or exorcised by the curvatures and involutions—spirals, pustules, volcanoes, etc.—sculpted by the poem into its own duration. Although Game and Khalfa make no distinction among the different editions, their argument holds for all of them, starting with the 1939 text. If we were to structure the poem according to its textual variations, on the other hand, we would notice that the large additions introduced in 1947 Brentano’s, and subsequently rearranged for the 1947 Bordas, effectively divide the later versions of the poem into five large segments, while retrospectively dividing the 1939 into three 'stationary' textual segments. (Fig.1) These three 'stationary' segments, which I here label 'textual' to contrast them with the familiar 'hermeneutical' divisions of others, serve as the best possible argument that indeed the text (and only the 1939 text) could be divided into three parts. The difference is that these parts are not to be read as a teleological whole, but rather as the textual blocks that remain relatively unchanged (only three small additions and no major deletions) throughout the history of the poem and around which most deletions, additions and migrations take place. From an examination of these three segments, the argument could indeed be made that they provide the foundation for the ‘hermeneutical’ divisions of the critics, further proving the centrality of the 1939 text, but it doesn’t follow that the other versions simply adjust to the conceptual model provided by the early text. Few will dispute the variations in style that came as a result of historical pressures made to bear on Césaire’s poetic practice. The large segments that surface in the 1947 Brentano’s —labeled 1947 A, B, C and D below— begin the process of transformation by opening the poem to a surrealist reading that was just not there before. As Arnold, Daniel Delas and others have already pointed out, this "poetic revolution" was linked directly to Césaire's involvement with Breton and the New York surrealists during and immediately after the war, bringing context to bear on the creative process. ("Palimpsest" 136) As a result of the additions, what critics have generally understood as the prophetic voice of the poem suffers a radical change from a more biblical to an overtly surrealist mode. By the time we reach the 1956 expurgations, a time in which Césaire was already grounded in the political climate of the Cold War, the movement to secularize the poem would be complete. Of the surrealist material, the overt sexual references bore the brunt of the attack, but many subtle lines survived censure. The strange communion of the Christian and the surrealist, and all the shared material, together under the final 'sober' shell, makes the 1956 version a veritable congeries of discourses, which perhaps accounts best for the variety of interpretational onslaughts which the text has provoked and withstood over the years: from Jungian universalism to Senghorian Negritude, from Bachelardian dreaming to post-structural dismantling, from Marxist inquest to Créoliste insurrection, et cætera. Paradoxically, it is what I called above the core of structural meaning, the bulk of the unchanged 1939 text, which provides these disparate discourses their center and their coherence. Despite these efforts to move away from both catholic and surrealist love, the later editions could not sweep off all traces of the 1939 text without damaging the integrity of the poem. After the poet and his editors had refigured or disfigured preceding published versions, the illusion that ‘it’ was still the same poem could only be sustained so long as the "original," i.e., the 1939 sequence was preserved. The end-result is clear: The "definitive" edition still bears most of the 1939 base-text in the same sequence (in relation to itself) as it was when it was originally published. In the section which follows, I will return to the source, to our "backbone" texts, starting with the surviving manuscript and its first appearance in 1939 in the Paris journal Volontés, to trace the first strand of this "strange communion" —perhaps the most curious strand once we consider it has been the least studied. From there I will move through to Brentano’s, focusing on the production history with the hope that, as our picture of the poem’s evolution becomes clearer, we become more capable of distinguishing those discursive disparities which belong to the historical changes of the text proper and thus lay the groundwork for a holistic reading of the poem that avoids the old Augustinian trap of an a-historical reconciliation of manifest dissonances which has marked previous readings of the poem.
The preceding text is a pre-print draft of a text forthcoming from the Canadian Review of Comparative Literature in 2010. All rights reserved © Alex Gil