Article body

At first they had tried to make a machine that could translate texts. […] One afternoon they fed it Poe’s ‘William Wilson’ and asked it to translate it. Three hours later the teletype began to print the final version. The story was stretched out and modified to such a degree that it was unrecognizable. It was now called ‘Stephen Stevenson.’ That was the first story. […] We had wanted a machine that could translate; we got a machine that transforms stories. […] It takes what is available and transforms what appears to be lost into something else. That is life.

Ricardo Piglia, The Absent City, 2000, p. 37

I. Translation as an Error Machine [La traducción como una máquina de errores]

Time and again, almost regardless of the periodical or publication in which reviews of their work appears, literary translators suffer two possible fates: they are either only mentioned in passing, as if to help impose their invisibility, or they are berated for their errors and mistakes. The implications of the former case, the ignoring of the translator, whether it be explicitly or implicitly (e.g., by praising the translation’s “fluidity” and “naturalness”) have been extensively analyzed in translation studies as of late, particularly by Lawrence Venuti in his pivotal The Translator’s Invisibility (1995). Less has been said about the berating of the labor of the literary translator, except in the occasional defense of this oft-called thankless task.[1] In any case, as different as they may seem, both kinds of critiques–the praising of a translator’s “invisibility” or the attacking of a translator’s errors and mistakes–share a crucial driving preconception: they both inevitably hold the original higher than the translation; they both repeatedly celebrate the superiority of the source over what is nearly always found to be the inferiority of the target text. Even in more scholarly contexts, critics in contemporary translation studies have tended to evaluate translations as being either domesticating or foreignizing[2]–terms that update yet reproduce the traditional metaphrase v. paraphrase debate, and which still, in either case, insist on the predominance of the source over the target. Often without questioning this insistence on the superiority of the original (or of the value of the “foreign,” in Venuti’s terms), critics in translation studies today tend to show their hand by critiquing the naturalizing or normalizing aspects of a translation, and perhaps suggesting a handful of ways that such a translation may have better “foreignized” the text. And the critic’s hand is this: that behind such valuing of the “foreign” lies the deterministic preconception that the original is a force to which the translation can never quite measure up.

However much potential and value there may be in such an approach, identifying a translation’s domesticating aspects is also another way of pointing out the translator’s insufficiencies. In other words, critiquing a translation for naturalizing or normalizing the foreign is another way to underscore the translator’s weaknesses; in this case, his/her inabilities to foreignize the text. It is another way to state that the translator has erred or been mistaken in his/her attempt to establish “equivalence” or “accuracy” or “fidelity”–all highly laden and storied and still problematic terms in the study of translation.[3]

But is an original always better than its translation? How are such evaluations made, and by whom? Who determines what is an error, what is a mistake? For that matter, are errors and mistakes the same thing? What is it about errors that make us so anxious? Paul de Man uses the following quotation from Marcel Proust as the epigraph to his book of essays Blindness and Insight: “Cette perpétuelle erreur, qui est précisément la ‘vie’….” I cite De Man quoting Proust here as a way to begin my analysis of the potential gains–and not only losses–of errors, of dislocutions and disconnections, of mis-readings and mis-translations–and especially of their importance for various literatures of the Americas.[4]

For my purposes in this article, I would maintain that “error is not mistake” (Corngold, 1982, p. 492). The latter, the mistake, can be seen as trivial or banal, without true value; a mistake is something that a translator commits from lack of information or knowledge. Mistakes, in translation, are usually semantic or lexical in nature: translators might mistake the meaning of a given word or phrase; they do not recognize an idiom or expression; etc. Mistakes, in principle, can be corrected to improve accuracy or, better yet, equivalence. Error, on the other hand, is a less stable, much more uncertain, and hence, much richer category. It is related to the “essential ambivalence” (Corngold, 1982, p. 492) of literary texts and, whether they are made consciously or unconsciously, errors open a field of potentiality in the translation. Error is closer to equivocation and digression, to betrayal and infidelity, to Derrida’s differance, to the burrows and rhizomes of “minor literatures” (in Deleuzian terms[5])–and thus to mis-readings and mis-translations.[6]

As Walter Benjamin posits in “The Task of the Translator,” since “real” poems and works of art are not intended for the receiver, neither should their translations. The “poetic” quality of a literary text, according to Benjamin, is independent of the reader, whether it be in the original or the translation (Benjamin, 1992, pp. 71-72). If error is an essential element of the literary aspect of the source, then it follows that error–ambiguity, ambivalence, abuse–must be an essential element of the target, if it is to be literary as well. There are times, after all, when–as Borges says in “About William Beckford’s Vathek”–“the original is unfaithful to the translation.”[7] If we follow Borges, we soon see that it does not really mean anything to say that a translation betrays or is unfaithful to the original. The question–the only question, really–is whether such a betrayal is fruitful.

This article opens with a quotation from The Absent City, a section dealing with the “origin,” if we can call it such, of the machine at the center of the Argentine Ricardo Piglia’s enigmatic 1992 novel. Now I quote again from The Absent City as a way to suggest a connection between the category of error and what I am calling, here and elsewhere, a theory of mis-translation:[8]

At first the machine would get it wrong. Errors are the first beginning. The machine ‘spontaneously’ breaks up the elements of Poe’s story and transforms them into potential fictional nuclei. That is how the initial plot had emerged. The myth of origin. All the stories came from there…. Reality was defined by the possible (and not by what was). The true-false opposition had to be substituted by the possible-impossible opposition.

Piglia, 2000, p. 83

Error and misreading in the fading line between the possible and the impossible: therein the fleeting potential of mistranslation.

II. A Borgesian/Joycean Mis-Encounter [Un desencuentro, una dislocución]

Thinking of Borges as a translator of Joyce is, at first glance, highly unlikely. Thinking of the ever-expansive, massive Joyce as a predecessor of the ultra-condensed Borges does not seem possible. Yet, in his work, Borges maintained a life-long dialogue with Joyce, a deeply ambivalent one in which the Argentine quotes and translates and refers to Joyce as much to disagree with him as anything else. More than an encounter with his Irish counterpart, Borges seems to strive for a mis-encounter. If a writer does indeed create his own precursors (thinking now of Borges’s own “Kafka and his Precursors”), then we may want to add that a writer re-creates and in fact mis-creates his own precursors–or so at least in the case of Borges and Joyce.

Borges’s dialogue with Joyce begins in January of 1925, just three years after the initial publication of Ulysses in English, and four years prior to its full translation into French. Borges–referring to himself as the first “Hispanic adventurer” to dare to make his way through the vast novel (1925, p. 3)–undertakes the first translation, and one of the first readings, into Spanish of James Joyce’s Ulysses. Borges’s translation consists of the last page of Joyce’s text, and is accompanied by a brief analysis of Ulysses, both of which appear in the Argentine avant-garde journal Proa. Borges’s partial translation is a brief masterpiece that has marveled a handful of writers and critics through the years. The fact that the translation is a partial one turns out to be significant in and of itself, as the fragmentary aspect of this initial translation of Ulysses contributes to a certain aesthetics of mis-translation which Borges would then develop and continue to practice throughout his life.

Borges delineates what I am calling his theory of mis-translation most clearly in his essays “The Homeric Versions” and “The Translators of The 1001 Nights”; he also puts this theory of mis-translation into play in his own ficciones. And I propose that one of Borges’s major contributions to translation theory and practice–in particular, his innovation of mis-translation based on rewriting fragments of foreign as well as domestic texts to create new representations of multiple pre-texts in an Argentine context–is its unexpected foundational role in the periphery, and hence its potential to renovate Latin American literatures. I purposefully raise the question of what it means to say that a text or a moment is “foundational” by focusing on the equivocations of such moments; in this case, on the equivocations of moments of translation–moments, which, in turn, take on foundational aspects. All of this being said, I propose that Borges’s 1925 translation of the last page of Ulysses is a foundational text that is a mis-translation, a rejoicing (if you’ll pardon the pun) in a complex play of error and dislocution. And for this I recall the etymological root of “error,” from the Latin erre, meaning: “The action of roaming or wandering; hence a devious or winding course, a roving, winding” (Oxford English Dictionary).

In fact, the innovation of Borges’s partial translation of Ulysses lies precisely in Borges’s infidelities, in his betrayals, to Joyce’s text. On the one hand, Borges’s translation is a literary gem in how it is able to re-create (to re-invent) the flow of part of Molly Bloom’s interior monologue into Spanish–although we must really say into castellano (i.e. into a Río de la Plata diction and syntax). In his version, Borges completely recontextualizes Joyce for an Argentine reader through a series of omissions and alterations, of “creative and joyful infidelities,” to use Borges’s own phrase about another translation–to which I will return momentarily. Borges introduces three fundamental changes in his partial rendition of Penelope. The first is that he eliminates nearly every proper name and place referring back to earlier moments in Ulysses, thus underlying a rupture with the source text and converting Borges’s translation into a potentially autonomous piece. The second change is that Borges has Molly speak in an Argentine voseo (i.e., the second person singular conjugation used in the Río de la Plata region); this creates a startling effect, part of which is to displace Molly and resituate her at the shores of the Río de la Plata. And the third change is this act of fragmenting and condensing Joyce’s novel, so to speak, as if the entire project of Ulysses could be metonomycally reduced to one page: to one partial and acriollada Borgesian page.

I cite just one line as an example: “Para vos brilla el sol me dijo el día que estabamos tirados en el pasto de traje gris” Borges’s Molly says at one point (1925, p. 8; emphasis added), whereas Joyce’s Molly had said “the sun shines for you he said the day we were lying among the rhododendrons on Howth head in the grey tweed suit” (Joyce, 1990, p. 782). Here we see (or hear, more accurately) Molly speaking in the Argentine voseo; we hear her flow of consciousness, the technique inaugurated by Joyce in Ulysses, but we hear it in Spanish, acriollado in castellano rioplatense. Also, this example contains the most striking deletion in Borges’s version: the omission of the reference to “Howth head,” a crucial site in Molly’s monologue and, of course, a central geographic and textual reference in the entire novel.

Borges’s deletions of most proper nouns and references to previous parts of Ulysses, the multiple omissions and alterations that rupture important links with the original, transform Borges’s version into an independent text. This, combined with the use of the voseo and the Río de la Plata idioms and colloquialisms, leads to an appropriation that is both textual and contextual. Borges perceives the relationship between the content and the context in Joyce’s version (the context being everything that has to do with the language and the setting of Dublin on June 16, 1904), and recreates it in his own image (which includes an analogous context: Borges’s Buenos Aires of the mid 1920s). Through his manipulation of language and his stylistic decisions, Borges manages to transpose the relationship between text and context from source to target.

Having Joyce’s Molly speak in the rioplatense voseo creates a strange literary effect. It is as if her voice were displaced, and yet it is not entirely transported to Argentine Spanish. It is here and there at the same time; paradoxically, it is Molly Bloom, with her beautiful fluid stream of consciousness, but it is Molly in Argentine Spanish; she is dislocuted, the text creates a strange effect of being two texts at once, undoubtedly Joyce’s Molly but unquestionably in Borges’s Spanish. Simultaneously an appropriation and a recreation; better yet: a recreation that succeeds as it appropriates. The paradox creates this strange effect, it creates a rupture with the source (through the linguistic displacement) while maintaining an unexpected continuity with the same source, even as it is recreated in a new language for a new context, on the margins of the Río de la Plata.

In 1945, twenty years after Borges’s partial version of Ulysses, the first complete translation into Spanish of the Irish novel comes out in Argentina, translated by Salas Subirat. In his 1946 review of this translation, Borges advises translators of Joyce that: “Joyce expands and reforms the English language; his translator must try to undertake congenerous liberties” (1946, p. 49). In 1925, in one page of rioplatense Spanish, Borges had already accomplished this task in a highly unexpected and irreverent fashion.

III. A Borgesian Theory of Mis-Translation [La infidelidad creadora y feliz]

A few years after he undertakes his partial translation of Ulysses, and immediately prior to when his ficciones begin to appear, Borges more explicitly articulates his ideas about translation. His 1932 essay “The Homeric Versions” (“Las versiones homéricas”) opens with the following memorable statement: “There is no problem as consubstantial to literature and its modest mysteries as that raised by a translation”. And Borges adds that, even better than so-called “original” writings: “translation seems… destined to illustrate aesthetic debates” (Obras completas 1, 1996, p. 239). For Borges, to talk about translation is to talk about aesthetics, and there is perhaps no better point of entry into literature and “its modest mysteries” than the study of translation.

Borges constantly questions the notion of fixed and stable originals. In its place, he suggests an extremely fluid conception of the relationship between any two texts, including originals and translations: “To presuppose that every recombination of elements is necessarily inferior to its original, is to presuppose that draft 9 is necessarily inferior to draft H–as there can be only drafts. The concept of a definitive text belongs only to religion or fatigue” (Obras completas 1, 1996, p. 239; emphasis in the original). This statement challenges just about all traditional theories of translation, which have always privileged the original–read: “definitive text”–and assumed that translations are the inferior of the two. If all texts are “drafts,” there can be no original, no “definitive text,” to which to measure up. “[T]he superstition of the inferiority of translations” (Obras completas 1, 1996, p. 239), as Borges calls it, is shown to be just that: a superstitious belief in something that does not exist. Borges’s challenge of the “definitive text” shifts the value from the original to the translation, and legitimizes the new version–in a new context, in Latin America–as a re-writing of one or more pre-texts.

Borges goes on to explore the potential gain that accompanies translation in his 1935 essay “The Translators of The 1001 Nights” (“Los traductores de Las 1001 Noches,Obras completas 1, 1996, pp. 397-413). Here, Borges sidesteps traditional demands for fidelity by suggesting that the merit of a translation lies in how well the translator makes use of the infidelities that inevitably occur when one transposes a text from one language and context into another.[9] In his discussion of Richard F. Burton’s version of the Arabian Nights, for example, Borges identifies the countless substitutions, re-writings, alterations, omissions, and interpolations that Burton undertakes, and praises these falsifications, which he argues improve on the original (Obras completas 1, 1996, p. 405). Of another translation of the Nights (the 1889 version into French by J. C. Mardrus), Borges states: “To celebrate Mardrus’s fidelity is to omit Mardrus’s soul, it is to not even speak of Mardrus. It is his infidelity, his creative and joyful infidelity, with which we should be concerned” (Obras completas 1, 1996, p. 410; emphasis added). “Creative and joyful infidelities” or, as De Man underscores in Proust–drawing a connection sideways, if you will–“That perpetual error, which is ‘life’ itself” (De Man, 1971, p. 1).

IV. A Tower of Betrayal: Argentina’s Errant Tradition of Mis-Translation [La irreverencia de la periferia: Tradición, traición y potencialidad]

So what does it mean to say that a translation–and a mis-translation, at that–is a foundational text? In the case at hand, Borges’s 1925 partial translation introduces Joyce’s stream of consciousness technique into Spanish, into a particular South American Spanish. This narrative technique, as is well known, would go on to enjoy a long and fortuitous lineage throughout Latin American literature.[10] In his partial version of Ulysses, Borges irreverently displaces Joyce toward the shores of the Río de la Plata through a cultural and linguistic mis-translation. Beginning what would be a complex, life-long, ambivalent position towards Joyce,[11] Borges re-invents Molly and, by doing so, paradoxically finds a feasible recreation of the Joycean stream of consciousness in Spanish, in castellano. Borges’s displacement and irreverence problematizes issues of originality and influence, while bringing to the forefront the central role of translation in the development of Latin American literatures. By translating just the last page of Ulysses, Borges metaphorically closes Joyce’s version in English, and opens a Spanish version that becomes an important originating move into twentieth-century Latin American literature.

This brief but poignant example strongly suggests that discussions of the practices and outcomes of translation in Latin America should include translation’s role as a point of rupture and continuity, of translation’s ability to both question and establish traditions, and of translation’s foundational potentiality. It is therefore important to consider not only direct lines of influence, such as the “Ulyssean” novel in Latin America, but also certain key “scenes or moments of translation”–moments that define traditions and identities as much as moments of so-called original writings. Borges’s valorization of mis-translations would be significant in and of itself even if it applied only to Borges. But this idea of a practice of fragmentary, equivocal translation–of translation as an appropriative re-writing, of translation as an “errant” movement–is actually found in a number of other key moments of Argentine literature.[12]

In fact, a quick look through this lens reveals that translation–better yet: mis-translation–is found throughout the history of Argentine literature, beginning with certain foundational moments of the nineteenth century. Although such a thorough study falls outside the scope of this article, I maintain that if one were to write a literary history of Argentina–and for that matter, of most literatures of the Americas–one would have to begin with its most important translations. As an example, I mention briefly Ricardo Piglia’s reading of the opening of Sarmiento’s Facundo (1845). The foundational, hybrid text Facundo begins with the following epigraph: “On ne tue point les idées.” The phrase comes from a sentence by Diderot. But Sarmiento misquotes it, mis-translates it, and attributes it to another thinker altogether, Fortoul. Sarmiento takes the phrase “On ne tue point les idées” (“Ideas cannot be killed”) and renders it as: “A los hombres se degüella; a las ideas no” (“Men can be beheaded, but not ideas”). By introducing the erred concept of the beheading, a creative infidelity not at all present in the original, Sarmiento takes the idea from the French (with irreverence) and re-writes it (displaces, mis-translates it) in a context that alludes to the dictator Juan Manuel Rosas’s mazorca [13].

What does it mean for one of the most intelligent and prolific intellectuals of Latin America’s nineteenth century, Piglia asks, to commit such an error? In Sarmiento’s “error,” Piglia identifies: “A line of equivocal references, false quotations, and apocryphal erudition which is a sign of Argentine culture at least up to the time of Borges” (1974, p. 132). And we could add that this line does not end with Borges at all, but that it extends all the way through Piglia himself. This is a line that would include, just to name a few of the most alluring moments of translation in Argentina’s history, the following events: other texts and declarations by the Literary Salon of 1837; the avant-garde projects of journals like Proa and Martín Fierro; Sur and Victoria Ocampo’s cultural importation machine; the moment of the translation of Ferdydurke by Gombrowicz in Buenos Aires (from Polish through French into an Argentine Spanish with multiple collaborators in 1947); perhaps the whole of Borges’s oeuvre, as well as much of Julio Cortázar’s and Manuel Puig’s; and, if we jump to contemporary fiction, the machine at the center of Ricardo Piglia’s 1992 novel The Absent City, or the importance of translation in issues of identity and sexuality in Sylvia Molloy’s latest novel, El común olvido (2002). I also note that whereas mis-translation functions as a site of national foundation in Argentina’s nineteenth century, it becomes a site of innovation and renovation–of narrative potentiality–in the avant-garde movements and in Borges, and, in Ricardo Piglia’s late twentieth century, of political cultural resistance.

Finally, one lasting effect of the fact that Ulysses is introduced in Latin America through a mis-translation, through Borges’s “creative and joyful infidelities,” is that future Ulyssean emanations in Latin American letters resonate with an equivocal foundational source. This is the case with Piglia’s The Absent City, which I have mentioned in passing a couple of times now. The end of The Absent City resonates distinctly with the final “yes” of Joyce’s Ulysses. But this resonance is mediated through Borges equivocal and fragmentary 1925 translation of the last page of Joyce’s novel. And we might say that Joyce’s entire influence in Latin American literature, or at least the Ulyssean line of this influence, is also mediated by Borges’s 1925 mis-translation. The fact that such influence is traced back to a moment of mis-translation, to an equivocal South American re-invention, complicates and problematizes not only our notions of originals and originality, but of influence and tradition as well. Borges’s treatment of Joyce’s Ulysses in 1925–much like certain key moments of Argentina’s nineteenth century, and like the machine in Ricardo Piglia’s The Absent City, as well as numerous other moments of translation throughout Argentine literature–is an originating moment, a foundational move, based on an act of mis-translation.

V. Conclusions: Moments of Potentiality [En y desde las orillas latinoamericanas]

At some level, all translations betray the source text. As Borges saw early on, however, this betrayal need not lead to despair or anxiety, it need not cause a “misery of translation” (as Ortega y Gasset frightfully put it). Each moment of translation is a site of potentiality. Even if most such moments are wasted, even if most errors and infidelities are neither “creative” nor “joyful,” translation is still rife with potential, and some of translation’s infidelities luckily turn out to be both “creative and joyful.” Borges’s translation of the last page of Joyce’s Ulysses is certainly one such moment. But the implications are broader than this kind of aesthetic evaluation. For every “moment of translation” for Latin American writers and thinkers, precisely due to Latin America’s peripherality–precisely due to the irreverent potential of the periphery–is an opportunity to re-interpret and renovate the source, in the target language and culture. The potential to reread and rewrite through error and dislocution: the kind of innovation and renovation that is mis-translation’s potential to play a foundational role in Latin American literatures.