Material Cartography: João Guimarães Rosa’s Paratexts

Mapping the Novel

In 1957, the Brazilian critic Antonio Candido published a piece titled “O Sertão e o Mundo” praising João Guimarães Rosa’s major novel, Grande Sertão: Veredas (The Devil to Pay in the Backlands), which had come out the previous year.¹ In his essay, Candido develops an insightful reading of the novel’s structure based on the interplay between three anthropological categories: the land, the man, and the fight. The entwinement of these main dimensions is the foundation for Candido’s argument about the novel, in which men and action become a function of space: “An obsessive physical presence of the physical medium; a society whose rules and destiny depend on it; as a result, conflict among men.”²

The importance of space in Candido’s reading is paramount. Although Candido does not consider its implications at a theoretical level, it is through the essay’s considerations of the spatial organization of the novel that its key hermeneutic problems may be posited, and it is also in the spatial and geographic projections of the narrator’s discourse that the effects on the reader become visible for the critic. Projected visuality, Candido indirectly suggests, is central to this 600-page novel.

A detailed reading of some of the essay’s passages may show us how cartographic metaphors structure this view. In one of the essay’s key moments, Candido explicitly posits the relation of the novel to a map:

Lowlands where horses gallop, sierras where horses drag themselves; grey fields, ..., holms, corrals and villages. At every step, the tangible reality of the Northern part of Minas, up to Piauí, where the man of the South is a stranger. Bent over the map, we are able to identify the majority of toponyms and the approximate route of horse rides. Guimarães Rosa’s world seems to be limited to observation. But we need to be careful. Pressed by curiosity, the map collapses and recedes. Here, a void; there, an impossible combination of places; further on, a mysterious route, unreal names. ... Let us unfold this map thoroughly. Like a large ox-skin, the Northern territory of Minas Gerais extends itself, cut along the loin by the River São Francisco—physical accident and magical reality, water flow and river God, the axis of the Sertão.³

This much-cited passage has been discussed throughout Guimarães Rosa’s critical reception, of which mapping is an integral part. After extensive travels and geographic research, Alan Viggiano, in 1971, was proud to present a large number of identifiable toponyms to counter Candido’s indeterminacy thesis, situating the novel’s actions in accurate maps of the region.⁴ More recently, the critic Willi Bolle, reenacting the protagonist’s and Viggiano’s travels along the landscape of Minas Gerais, concluded that obsessive geographic rigor and poetic indeterminacy balance each other through the intricate and self-canceling paths of the novel’s characters.⁵ The labyrinthine space of the novel hovers between referentiality and indeterminacy, critics seem to argue. In this, Guimarães Rosa’s work appears to respond to one of the main tensions in the relations between literature and cartography.

Less noted, however, is the rhetorical shift that brings the figure of the map into Candido’s reading; not at all obvious, the map appears as a referent that slowly slides into a discourse occupied with an immersion in the novel’s landscape. What triggers the shift to the map as a possible figure of space is the recognition of the alien condition of the man from the South (the critic, the reader, Candido himself) in relation to the landscape described: “... where the man of the South is a stranger. Bent over the map, we are able to identify. ...” But if the map is initially posited as a possible double for the novel—both map and book are materially observed by the extraneous reader; the map mirrors and validates the novel, both isomorphically participating, “like a large ox-skin,” in the landscape described—it will later be revealed as a metaphor for the novel itself. This may be seen through the deictics used (“here, a void; there, an impossible combination of places”), which seem to hesitate between the book and the map (where exactly is this void?), leading the way for the establishment of a relation between the materiality of the map and the materiality of the novel, while in the same gesture denying the very possibility of a map. One might think of Melville’s famous description of Queequeg’s birthplace in Moby-Dick: “Queequeg was a native of Kokovoko, an island far away to the West and South. It is not down in any map; true places never are.”⁶ Grande Sertão: Veredas seems to call into question its own referentiality, and to suggest a poetic cartography, a mapping of true places, one might say, where the text at the same time elicits and collapses cartographic representation.

Figures 15.1 and 15.2. Front and back cover of the 6th edition of *Grande Sertão: Veredas* (Rio de Janeiro: José Olympio, 1968).

To contemporary readers, Candido’s description of the novel is entirely familiar. The two maps that make up the inside cover flaps of Grande Sertão: Veredas (figures 15.1 and 15.2) seem to match Candido’s comments, expanding the meaning of the maps of Minas Gerais at their basis through a series of symbols, references, and illustrations suggestive of a symbolic and poetic relation to space. Furthermore, the maps’ cartographic aspirations are established through the sole elements that could be superimposed on an actual map—the rivers that run through the territory. In fact, it has not escaped readers of the novel’s illustrations that each of the flaps corresponds to one of the banks of the São Francisco River. Candido’s reading seems to echo this division: “If we think of [the river’s] function in the novel, we realize that it divides the world in two parts, different in quality: the left bank and the right bank, loaded with the magical-symbolical meaning that this division represents to the primitive mentality. The right bank is auspicious; the left, inauspicious.”⁷ Indeed, if we take a closer look at the book’s materiality, we will find that this split is represented at all levels. “The São Francisco divided my life in two parts”⁸ is the central sentence of the narrator’s discourse, in the exact middle of the novel, textually dividing Grande Sertão: Veredas in two. Materially, the text enacts what the illustrations suggest; the map is embedded in the novel, as a frame and as a figure for its spatial structuring. Both a container and a contained, it seems to recall the ambiguities of the third term that Walter Moser relates to the encyclopedic project:

In their complementarity, System and Dictionary would suffice to constitute the encyclopedia. One is entitled to wonder what is the contribution of the third element, the Map, also called “world map,” “figured system,” “table.” At first sight it seems to have the status of an extra, an element that is not necessary, because it adds to a work that is already complete. It seems to be redundant information coming from a transsemiotic translation, which says the same thing in another semiotic register. The point is that as a synoptic table it belongs to another textuality. Making the transition to the pictorial, it situates itself on the margin of the verbal discourse that is common to the System and the Dictionary. The fact is also confirmed by the technical problem its inclusion in the book represents. The marginality of the map, however ... is contradicted by the fact that the ideal of a synoptic representation is an insistent fantasy throughout the history of the encyclopedic project whose center it apparently aims to occupy. Many citations could confirm this idea of Novalis: “The more difficult it is to put a book in a picture, the less good it is.”⁹

It is therefore surprising to note that the famous maps of Grande Sertão: Veredas were only added to the second edition of the novel, published in 1958—one year, that is, after Candido’s famous article. The first edition contained no illustrations aside from Poty’s drawings for the front and back cover. Whatever the correlation of these two elements—the critic’s trope and the author’s decision—the episode is significant of the cartographic vocation, one could say, of Guimarães Rosa’s novel, and of the importance of the map as a figure for its reading. As a threshold element, calling attention to a visual and spatial interpretation of the novel’s thematic threads, the map functions, I argue, as an interface where an image of the book is projected and reading indications are visually set forth.

The purpose of this chapter is to examine this relation, taking Grande Sertão: Veredas as its departure point in questioning the interconnectedness of two sets of relations: the relation between literary text, obscure referentiality, and literary landscape in the building up of Guimarães Rosa’s sertão (explored in the next section); and the relation between cartography and the questioning of the material book (discussed in the third and fourth sections of the chapter). Guimarães Rosa’s work, I argue, is an extreme case of the enactment of the implications of mapping in literature—not only, as we have begun to see, in its poetic transfiguration of the elusive map of the landscape of the sertão heartland, but also in the overlapping of the map and its metaphors with the unstable cartography of a book in permanent reconfiguration. In this questioning of the textuality of the literary map, I believe, resides its main theoretical interest for the relations between literature and cartography.

The Building Up of Guimarães Rosa’s Sertão

To better examine the building up of a literary landscape, let me turn to “Meu Tio o Iauaretê” (“The Jaguar,” in David Treece’s translation)—a short story completed by Guimarães Rosa at the time he wrote Grande Sertão: Veredas, but published only posthumously—that shares many themes and other features with his novel. At the beginning of the story, the narrator warns his visitor: “Aqui é muito lugaroso” (literally, “Here it is very lugaroso” or “This place is very lugaroso / has too many places”).¹⁰ Typical of Guimarães Rosa’s neological derivations, the coinage lugaroso (an adjective formed from lugar, place) seems to convey the idea, at the same time, of this obsessiveness of place and of the multifarous nature of this complex geography. It is also a warning: in the tense context of the short story (in a lost cabin in the woods, a strange man tells a visitor the tale of his metamorphosis from jaguar hunter to hunter of men, suggesting he will be his next prey), lugaroso means excessive in the sense of dangerously unlimited or uncontrollable.

If we take into account the construction of this short story—as in Dostoevsky’s hidden dialogue,¹¹ a supposed dialogue between two characters is rendered through the direct discourse of just one voice—the sentence may be read differently. One of the first intelligible sentences in the story is an invitation to enter an uncanny place (a cabin without walls), addressed, we come to understand, to some foreigner at the threshold: “Uh huh, you wanna come in, come on.”¹² Necessarily, the reader entering the text identifies with this invitation, thus beginning to perceive an unsettling coincidence between the space of the narrative and the text itself, or between text and territory. The story will explore this tension dramatically, representing the death of the narrator through an abrupt textual interruption. Aqui é muito lugaroso, then, could also mean the cartographic quality of this literary work: haunted by a labyrinthine structure, obsessed with spatial demarcations, and determined by its own spatial materiality. Again, the geography in the text and the geography of the text seem to overlap, suggesting a dangerous excessiveness.

If, as I began to suggest, Guimarães Rosa’s prose presents itself as the poetic cartography of a territory suspended between its referential nature and the rhetorical destabilization of such referentiality, the inescapable immensity of the territory of the sertão “as big as the world”¹³ may form the basis of this idea of fiction, where book, map, and landscape continuously collapse on one another, and where the reader is continuously confronted with the paradoxical legibility of that which has no limit (geographically) or form (textually). We are not far from the totalizing ambition (opere mondo) Franco Moretti finds in works such as Faust, Moby-Dick, or Ulysses.¹⁴ But in the consistent play on form and on the reading conditions of a blurred and yet insistent cartography, Guimarães Rosa’s fiction explicitly takes the inscribed cartography as a representational and hermeneutic problem.

The title of Grande Sertão: Veredas points in that direction. In one of the rare moments in which the text incorporates its own title (embedded reflexivity is one of the most important features of this novel), the narrator Riobaldo describes a denied perspective: “Do you know the great sertão? The ones who know it are the vultures, hawks, kites, and birds like that: they are always high up there, feeling the air with lowered feet, sizing up at a glance all joys and sorrows.”¹⁵ The veredas—small rivers, streams, marshes, or valleys with the function of routes—are pathways to this unlimited and inapprehensible space, impervious to any kind of totalizing view: this is the narrator’s (and the reader’s) position. The landscape is thus described in tension with the impossibility of legibility or representation, its inaccessible totality being construed through the immersed perspective of its characters. The title is therefore an opposition between different modes of construing the same space. Indeed, the most operative definition of the sertão that the novel offers the reader may be found on its first page: “The sertão describes itself. It is where the grazing lands have no fences.”¹⁶ In the terminology of Deleuze and Guattari, this immense, deserted space may be characterized as a smooth space: the points are subordinated to the trajectory in an unstable errancy in which direction is often blurred by the instability of the toponyms.¹⁷ Traversed by the veredas who seek their way through the immense space of the backlands, the landscape does not allow for fixity or distinctions, for borders between inside and outside, relating the impossibility of a map disconnected from its tours, in Michel de Certeau’s terms,¹⁸ with the impossibility of narrating an ever-open action: “The sertão has neither windows nor doors.”¹⁹ Describing the formless form of this world may be the challenging task of Guimarães Rosa’s fiction: a true place, one could say, is a “moving world.” I propose now to see how the material book becomes a fundamental tool in this cartography of the unlimited.

Cartography and the Material Book

For a walk and back again, said the fox. Will you come with me? I’ll take you on my back. For a walk and back again.

This epigraph from Sagarana, Guimarães Rosa’s first book, taken from “Grey Fox,” a children’s story, points to a double movement that runs through his entire oeuvre. On the threshold of the book, on the one hand, it introduces traveling as a common theme of the short stories that constitute it, reinforcing the unity of the whole; on the other hand, as a preliminary invitation, akin to the one we’ve seen in “Meu Tio o Iauaretê,” it suggests the superimposition of the description of a double movement (back and forth) and the reading experience. The suggestion, however, is not developed in Sagarana. The association of the act of reading with a movement bent on itself will be construed through the five books published by Guimarães Rosa, finding its clearest design in Tutaméia, in which rereading is clearly stated as a condition of legibility of the book.

Suzi Sperber has emphasized the presence, in Guimarães Rosa’s fiction, of characters engaged in back-and-forth movements.²⁰ It is a feature that should be considered not only from a structural point of view but also in the context of the tension between mapping and reading I underlined above. And in the two books published in 1956—Grande Sertão: Veredas and Corpo de Baile—it is important to note how the semantic field of travel begins to function as a reflexive or metaliterary problem. On one level, this move takes the form of an increasingly precise metaphorization of the theme—in which terms such as travel, traverse, and road are combined in relatively fixed structures, aimed at a representation of the relationship between experience and knowledge. In the variations dictated by the temporality of the trip, space seems to translate a permanent instability in reading the world. As Peter Brooks puts it: “If there is a knowledge provided by narrative ... it is of a particular sort: not only knowledge that comes too late, but recognition of the perpetual belatedness of cognition in relation to action.”²¹

We may think of the most important descriptions of error in Grande Sertão: Veredas:

I go through an experience, and in the very midst of it I am blind. I can see only the beginning and the end. You know how it is: a person wants to swim across a river and does, but comes out on the other side at a point lower down, not at all where he expected. Isn’t life really a dangerous business?²²

Had I but guessed then what I later came to know, after many bolts from the blue. … One is always in the dark, only at the last minute the lights come on. I mean, the truth is not in the setting out nor in the arriving: it comes to us in the middle of the journey.²³

In terms now centered on its temporal implications, this is again the problem of form in Guimarães Rosa’s fiction, represented by the invalidation of the extremes of the trip: the center is the place of blindness, and the metaphorization of movement continually exposes the late temporality of the constitution—and apprehension—of the way. The impossible bird’s-eye view of the sertão I have commented on had the birds contemplate “at a glance all joys and sorrows,” but no simultaneity is allowed on the ground. “We live on the road,”²⁴ says the narrator of “Estória do Homem do Pinguelo,” when reaffirming the fundamental unreadability of this world’s confusing revelations.

The epigraph from Sagarana allows the reader to realize that it is in the form of the book that the metaphorical dimension assumed by this topography in motion finds its performative value, temporalizing the mapping experience through the material limits—beginning and ending—of the book. Corpo de Baile and Primeiras Estórias already suggest this. Both books are organized around the two extremes of the trip. Between the final departure of Miguilim as a child at the end of the first story and his expected arrival at Buriti Bom at the close of the long final text, the book’s structure is defined by the idea of movement. The same is true of Primeiras Estórias, in which two inverted travels by plane open and close the book for the same character in its first and last story (“And life was coming toward him” is the formula that concludes both the travels and the reading experience).²⁵ The inscription of maps—and of maps of reading, I want to suggest—in the paratextual margin reflects this thematic insistence.

A careful reading of the editions prepared by Guimarães Rosa clearly reveals an engagement with the book as a form and with its physical materiality. In Johanna Drucker’s terms for describing experimental typography, what is at stake is “an investigation upon the book as an artist’s medium, rather than an editor or publisher’s domain.”²⁶ Guimarães Rosa’s control over his own books is confirmed both by his publishing house, José Olympio, and by his intense correspondence with translators and foreign publishers. The material edited by Paulo Rónai in the posthumous publication of the volume Estas Estórias clearly exemplifies this aspect. We can see a set of provisional tables of contents designed by the author as indications for the illustrator.²⁷ We now know that the books were scrupulously edited under the instructions of Guimarães Rosa, who had access to all stages of production. The promotional texts were written or suggested by the author, and anecdotes of Guimarães Rosa converting typos into neologisms in his careful review of the proofs became famous.²⁸

Ironically, the editions of Guimarães Rosa’s work published after his death have for various reasons sacrificed these elements to the point that the question of the material book has become, for more than thirty years, essentially unreadable. Only on the fiftieth anniversary of the books he published in 1956 was an effort made to recover materials sacrificed in earlier editions. This recovery is even more significant if we consider that the implications of Guimarães Rosa’s close involvement with the production process of the book go beyond a simple manifestation of authorial control. These editions now allow us to understand the complexity of what we may call his book projects and to see how paratextual materials are placed at the service of a general indication of reading and of a mapping of the reader’s path. In particular, they make visible the affinities and correlations between Guimarães Rosa’s work on the illustrated maps and the inclusion of duplicated tables of contents (reading and rereading indexes) in many of his works. In both we can see the inscription of a reflection on the effects of rereading, of repetition, and of a double assessment of the same graphic material that involves the temporal and spatial questioning of the printed form.

Paratextual Doubling in Guimarães Rosa’s Work

The material perspective I have elaborated here allows us to see a number of Guimarães Rosa’s other works in a new light. Indeed, from the gigantic enterprise that is Grande Sertão: Veredas to the minimalism of his short story collection Tutaméia, all Guimarães Rosa’s works after Sagarana have some kind of doubling in their margins that frames the text in two noncoincidental presentational elements and some kind of enactment of the index. Corpo de Baile, in its first edition of 1956, has two tables of contents, one at the beginning and one at the end of the seven stories, that literally tie the work together. The two versions offer different genre classifications for the same texts, belatedly identifying in the book’s parabasis a space for theory and reflexivity. Primeiras Estórias, published in 1962, offers an illustrated table of contents at the end of the book that visually translates the short stories that integrate it. Tutaméia, published in 1967, takes this doubling further by clearly prescribing a revision of the book’s first reading in a “Rereading Table of Contents,” as we will see in this final section. Mapping the book’s textual territory, these indexes both build up the work’s unity and subject it to a process of reconfiguration through repetition and variation.

My purpose here is to draw attention to the similarities between different forms of paratextual doubling in Guimarães Rosa’s work, of which the maps embracing Grande Sertão: Veredas and the tables of contents of Tutaméia will be key examples, and to the implications of this coincidence between geographic and bibliographical descriptions for an interrogation of the textuality of the literary map. My suggestion is that in all these cases—whether conveyed through illustration or through a creative use of indexes, tables of contents, or epigraphs—the book builds up, in its own extremes, a dissonant cartography conveying, in the same gesture, the spatialization and the temporalization I have hinted at as major formal devices in Guimarães Rosa’s work. In this, it brings to the fore the temporal implications of his inscribed visible and readable cartography. Through the usage of maps as tables and of tables as maps, Guimarães Rosa’s books seem to establish rereading as a constitutive movement of the reading. They suggest a movement upon themselves that affects the mapping of the book, defining repetition as the impossibility of fixating their form and guiding the act of reading to the paradoxical readability of a device in perpetual reconfiguration. Indeed, the infinity symbol that the reader finds at the end of Grande Sertão: Veredas could be the visual sign for this recursive movement. We saw earlier that the theme of traveling functioned as a trope for a permanent deferral in the reading of the world. The inscription at the book’s liminal extreme of reading as a tour through the book’s shifting places is another figure of this paradoxical legibility.

As we have seen, the split map on the cover flaps of Grande Sertão enacted this tension with the book’s completeness. As a graphic index of sorts, the double map indicates and constitutes, at the margin of the book, a visual and spatial whole—significantly split and dependent on the reunification of the book’s extremes to build its picture—that in the same gesture suggests and collapses the reference to the sertão and the book’s production of totality.²⁹ On the other hand, giving the reader the first unitarian image of the book while also breaking it down into its components, the table of contents seems to respond in dissonance to the hybrid function of the map as a figure that overlaps, in Paul Zumthor’s terms, description, representation, and interpretation.³⁰ In the margins of Guimarães Rosa’s books, duplicated maps and tables of contents thereby seem to share a synthetic and structural function that at the same time builds up and threatens the book, inscribing in its paratexts the cartography of a visual and temporal reconfiguration.

Let us then turn to the most extreme case of duplication in Guimarães Rosa’s work: the “Rereading Table of Contents” at the end of Tutaméia, Guimarães Rosa’s 1967 short story collection and his most sophisticated experiment with the possibilities of the material book. It will be helpful to take a closer look at its two extremes (figures 15.3a–b).

Figures 15.3a–b The two tables of contents of *Tutaméia* (Rio de Janeiro: José Olympio, 1967)

In the first table of contents, the stories are listed in the order in which they appear in the book, and one notices from the outset that this order is alphabetical, with the exception of two stories whose titles, beginning with G and R, spell the sequence JGR (the author’s initials, a signature of sorts). The second table of contents mirrors the first, but with some significant changes: four stories, whose titles were printed in italics in the first table, are now grouped together at the beginning as “prefaces” to the other forty stories. Two different images of the same book are conveyed, building up the dissonant cartography we began to see with earlier examples. In fact, the book’s identity changes from one table of contents to the other, as title and subtitle of the book are reversed from Tutaméia: Terceiras Estórias to Terceiras Estórias: Tutaméia. The same textual material projects two different structures and two different spatial and unitarian images of a scattered whole. Significantly, it is only in the second of these images—explicitly posited as a rereading table of contents, to be found after the reading—that structural positions are revealed. Throughout the reading the prefaces may have functioned as prefaces, but they are declared so only in a second moment. The purpose of this belated identification of an opening function and position for some of the stories is to call into question the border between text and paratext: the prefaces see their paratextual nature blurred and their exteriority disturbed from the moment they are positioned in a series of recognized and recognizably fictional stories. Book and margin are intermingled from the outset.

The temporality of repetition is of course what defines the originality of this play on the literary map and its effects on the cartography of the book. The idea of rereading is introduced through two epigraphs from Schopenhauer’s preface to the first edition of The World as Will and Representation. The source is indeed significant, for Schopenhauer’s is an extreme case in the tradition of the philosophical preface, given its bitter reflections on the book as a medium and on the necessity of reading “the book twice” for the book’s legibility. For Schopenhauer, the material book is an inadequate medium. From the outset, from the very margin, the preface tries to compensate for the discrepancy between a fixed form, which requires margins, and the organicity of the thought that rejects them. If the thought that the book brings forth is organic and unique, it will not endure the divisions and borders a book has to have. Schopenhauer writes: “A single thought, however comprehensive, must preserve the most perfect unity. … But a book must have a first and a last line, and to this extent will always remain very unlike an organism, however like one its contents may be. Consequently, form and matter will here be in contradiction.”³¹ This contradiction makes the preface and its demands necessary for an understanding of the book. The role of the marginal elements then becomes central: only the preface allows the book to be read, since it is the preface that states its rereading.

Both epigraphs in Guimarães Rosa’s book articulate a critique of the limited form of the material book. The first mentions the relationship between a patient first reading and a revelatory second reading: “Therefore, as I have said, the first reading demands patience, derived from the confidence that with a second reading much, or all, will appear in quite a different light.” In terms of reading protocols and expectations, the first table of contents thus has a destabilizing function: before even beginning a first reading, the reader is already projected toward a later revision, toward a subsequent teleological revelation of meaning. We are not far from one of the effects of the split in the maps of Grande Sertão I discussed earlier.

Completing this tentative first reading, the hypothetical reader then encounters a “Rereading Table of Contents” at the end of the book with a second quote from Schopenhauer: “The structure of the whole, which is organic and not like a chain, in itself made it necessary sometimes to read twice the same passage.”³² If the first epigraph brought the reader to the necessity of a suspension based on the knowledge that the second reading would change much or everything, at the end of this process the reader is projected back to her first reading, as she is told that a rereading has already occurred. Attention is called to the center, now defined as a space of difference, for the revelatory character of a second reading is subject to a temporal slip—it comes late, or belatedly, just as awareness (for instance, that Tutaméia is a book with four prefaces) comes at a later moment.³³

It is possible to see in this movement an enactment of the impossibility of distinguishing between reading and rereading that Matei Calinescu discusses in his work on the subject. He calls attention to the fact that a first reading is often “double,” intermingling linearity (typically characteristic of a first reading) with “structural attention” (usually related to second readings).³⁴ But what we find described in this interplay of anticipation and retrospection is not exclusively a matter of a more attentive or “spatial” reading, as is the case in the main discussions of rereading. In the building of a double mapping frame, what seems at stake is the interplay between difference, repetition, and deferral. In this, Guimarães Rosa’s cartography of the material book marks its originality, for the idea of a literary map subjects the cartographic gesture to the temporality of writing and reading. Embedded in the book’s own space, the map is necessarily contaminated by its paths, just as the bird’s-eye view of the sertão was accessible exclusively through the narrator’s veredas or pathways. Its relation to the book is metonymic, one could say, and not exclusively metaphorical. Inscribed geography and reflexive bibliography, margin and center, complement each other in the double movement through which the book as a space is constructed and mapped out while at the same time threatened by the instability of its form.

If the first table of contents of Tutaméia projects the reader toward the end, from the story to its map, one could say, the second version is there to make her go back to the center, to the “middle of the passage,” to the book as a transformed space. Tables of contents and maps offer a paradoxical prescription that, as in Schopenhauer’s case, deals with the impossibility of fixing the book in a settled form because of the contradiction between the spatial projection of the book and its inherent shifting movements that defy stable forms and fixed delimitations. Marginal elements, then, characterize Guimarães Rosa’s books as the space of this transitory relationship between a fixed form—the text, the map, the printed matter—and its revision enacted through repetition in its temporality. In that sense, the margin can never be interpreted as an accessory. It is rather a way of establishing the book’s unstable legibility, of taking its apparent fixed materiality as the point of departure for an organic transformation. Through the margin, Guimarães Rosa builds up the cartography of the book, in its material impossibility that depends on materiality, as a map in permanent reconfiguration.

Notes

9. My translation. The original quotation reads, “Dans leur complémentarité, Système et Dictionnaire suffiraient pour constituer l’encyclopédie. On est en droit de se demander ce que peut encore y apporter le troisième élément, la Carte, appelé aussi ‘mappemonde,’ ‘système figuré,’ ‘tableau.’ A première vue, il semble avoir le statut d’un supplément, d’un élément qui n’est pas nécessaire, parce-qu’il s’ajoute à un ouvrage qui est dejá complet. Il semble constituer une information redondante, issue d’une traduction transsémiotique, qui dirait encore une fois la même chose dans un autre registre sémiotique. C’est que, tableau synoptique, il relève d’une autre textualité. Faisant la transition au pictural, il se situe en marge du domaine discursif verbal qui est commun au Système et au Dictionnaire. Le fait est d’ailleurs confirmé par the problème technique que pose son insertion dans le livre. Cependant, la marginalité de la carte … se trouve être contredite par le fait que l’idéal d’une répresentation synoptique se maintien avec l’insistence d’un fantasme à travers toute l’histoire du projet encyclopédique dont il semble même vouloir occuper le centre. De nombreuses citations pourraient confirmer cette idée de Novalis: ‘Plus il est difficile de mettre un livre dans un tableau, moins il est bon’” (Walter Moser, “L’Encyclopédie Romantique: Le Brouillon de Novalis,” in Le Genre / Die Gattung / Genre (Strasbourg: Université de Strasbourg, 1979), 503–504).

11. Cf. the Bakhtinian description of hidden dialogicality: “Imagine a dialogue of two persons in which the statements of the second speaker are omitted, but in such a way that the general sense is not at all violated. The second speaker is present invisibly, his words are not there, but deep traces left by these words have a determining influence on all the present and visible words of the first speaker. We sense that this is a conversation, although only one person is speaking, and it is a conversation of the most intense kind, for each present, uttered word responds and reacts with its every fiber to the invisible speaker, points to something outside itself, beyond its own limits, to the unspoken words of another person” (Mikhail Bakhtin, Problems of Dostoevsky’s Poetics, ed. and trans. Caryl Emerson (Minneapolis: University of Minnesota Press, 1984), 197).

15 Material Cartography: João Guimarães Rosa’s Paratexts