Theoretical challenges for a genetics of translation

ABSTRACT

Translation archives are sources of insight into the creative processes of literary translation. This article examines the challenges of using the methodology of genetic criticism, which was devised for studying literary composition through archival evidence, for research in translation. It identifies conceptual limitations in this theory’s consideration of translated texts and their specific ontology, and it assesses the applicability of genetic criticism’s primordial categories – literary exogenesis and literary endogenesis – to translation. This paves the way for the construction of a methodology and a conceptual vocabulary for genetic process research in translation studies. It is argued that this new methodology can alleviate certain ambiguities or anxieties in current debates relating to the capacity of the translator to exercise their creativity and perform acts of writing that some would equate with authorship. Finally, adopting a translation studies perspective is suggested as a way for genetic criticism to shed its untenable Romantic heritage.

KEYWORDS:

• Genetic translation studies
• translation archives
• genetic criticism
• critique génétique
• history of translation

In recent years, archival sources have been used increasingly to gain insight into the compositional processes of literary translators. While there is no common methodology for conducting this research in translation studies, two approaches have emerged. Both require researchers to venture into public and private collections, into state archives and those held by publishers, all in search of drafts, manuscripts, page proofs, notes, contracts and correspondence, public records, or indeed any clue that helps discern translators’ working methods, strategies, and choices, their processes of composition and revision, and how they interact with editors, correctors, revisors and other collaborators. The first method emerged out of descriptive translation studies (Toury [1995] 2012, 213–240) and uses literary analysis (close reading, attention to poetics and form, comparative lexical analysis) to discern how a translation emerges with respect to the cultural and translational “norms” of the target culture. It has since been combined with historical studies approaches, notably in light of two articles by Munday that promote documentary evidence as a source of “potentially unrivalled insights into translator decision-making” (2013, 125) and micro-history as the most suitable method for exploiting such evidence (2014). Munday’s call for translation scholars to engage with archives has been answered by many,¹ some of whom continue to refine a micro-historical approach for translation (e.g. Paloposki 2017; Israel and Frenz 2019; Atefmehr and Farahzad 2022) or propose alternative historical methods like prosopography (Pickford 2021).

The second orientation applies the methodology of French critique génétique (genetic criticism) to what it terms the avant-texte (pre-text), thereby allowing “a certain reconstitution of the genetic operations that precede the text” (De Biasi 2004, 34). Until recently, international (Anglophone) translation studies was largely unaware of this Continental tradition (Cordingley and Montini 2015), and, although the two approaches are not mutually exclusive or in competition, the development of the historiography of translation history over the past two decades, with its independent metadiscourse (Rundle 2021), betokens a depth of reflection not yet matched in genetic translation studies. Just as translation history furnishes both objects of research (e.g. Bible translations in seventeenth-century England) and methodologies for apprehending historical subjects, periods, and processes (Rundle 2012), so too do translation archives yield objects and subjects of inquiry: they contain the material evidence of translation, which needs to be understood within the logic of the translational processes and discourses that structure them (Cordingley and Hersant 2021). While the historian’s methods are transferrable between document types and genres, genetic criticism was developed specifically for literary authorship. This remains its strength and, at present, its weakness: it is equipped with a sophisticated conceptualization of textuality and a descriptive vocabulary for how authorial processes and strategies are reflected in manuscripts and documentary evidence (Bellemin-Noël 1972; Hay 1993; Grésillon 1994; Van Hulle 2008, 2013; De Biasi 2011; Ferrer 2011), though it is yet to acknowledge adequately how the inherent differences between translatorship and authorship impact its methodology, notably in terms of the translation’s inexorable relationship to its source text (ST) and the role of ethics in structuring a translator’s writing performance. This article addresses these challenges by considering the ontology of the translated text from a genetic perspective. Theoretical limitations within genetic criticism are shown to complicate descriptions of translation genesis when this methodology separates “authorship” into phases of exogenesis and endogenesis. A genetic description of translation is proposed to help disambiguate some recent debates in translation studies with respect to translator creativity and to the status of translation as rewriting or authorship. Finally, a translatorly approach to the genetic criticism of authorship is suggested as a way to surpass this methodology’s Romantic heritage and gain a fresh perspective on literary creation.

The writing of translation

The very distinction between “original” and “translation” was called into question by Jorge Luis Borges in his 1932 essays on translation, “Las versiones homéricas” (“Some Versions of Homer”). Well-known to translation scholars, this argument has profound implications for a genetics of translation. Borges sees no ontological difference between authorship and translatorship, only a continuum of writing that renews itself in versions, which he equates with drafts:

Are not the many versions of the Iliad – from Chapman to Magnien – merely different perspectives on a mutable fact [un hecho móvil], a long experimental game of chance played with omissions and emphases? … To assume that every recombination of elements is necessarily inferior to its original form is to assume that draft nine is necessarily inferior to draft H – for there can only be drafts. The concept of a “definitive text” corresponds only to religion or exhaustion. (Borges 2001, 69)

Source and target texts are not two stable entities within a binary, mimetic logic; rather, they are different perspectives on “un hecho móvil” (Borges 1974, 239), a moveable fact (or event), which, for Borges, is a function of literature’s eternal narratives that form and reform with each articulation. His view that literature is writing in motion, that no text is a definitive version, and that all there can be is drafts resonates with many tenets of genetic criticism. Almuth Grésillon, for example, conceived of literature as “a doing, as an activity, as movement” (1994, 7; my translation). For her, the “ultimate goal” of genetic research “is not the text, but writing, understood as advent and event, as a process of written enunciation” (109). Similarly, Louis Hay (1993) theorized the literary work as a becoming not a finality. The text, Hay suggests, is “a necessary possibility” of its drafts; it is not the fact of writing but “one manifestation of a process which is always virtually present in the background, a kind of third dimension of the written work” (1998, 75; original emphasis). Genetic critics believe that avant-texte should be examined not with the sole intention of unlocking the secrets of the published text – even if this is of undeniable interest – but because it offers an opportunity to witness the process of textualization.

In translation studies, Lefevere (1992) argued forcefully that all literary translation is rewriting. From his descriptive perspective, translators assume authorial functions when they operate within a discursive system of literary conventions and institutions; they regularly shape the image of the authors they translate; and their texts typically function as originals in their culture of reception (109–110). Concurrent with the increasing recognition of translator agency through the sociology of translation (Wolf and Fukari 2007; Milani 2022) and with the intention of ending the much-maligned invisibility of the translator, the “creative turn” in translation studies has sought to foreground translator creativity and the translator’s authorial skillset (Bassnett and Bush 2006; Perteghella and Loffredo 2006; Buffagni, Garzelli, and Zanotti 2011; Wilson and Gerber 2012). These views emerged against the backdrop of post-structuralist theories of textuality – all texts express their fundamental intertextuality (there is no original), and all meanings are produced by the reader, who has no access to the author. Like many post-structuralists before her, Bassnett (2013) evokes Borges, to argue against the ontological distinction between original and translation. This reinforces her affirmation, drawn from her own practice, that the translator is an independent writer, author, and creative artist (Bassnett 2011). Arrojo (1994) too retraces the emergence of the idea of the translator as producer rather than reproducer, following Godard’s (1984, 15) feministic readings of the translator “as an active participant in the creation of meaning” to highlight the translator’s manipulation and subversion of the original.

Yet there is a key distinction to be made between these perspectives: if genetic criticism shares with Borges notions of text as movement and the potentiality of writing, such ideas arose in reaction to the disappearance of authorial agency within structuralist and post-structuralist conceptions of authorship.² Genetic critics resisted the notion that writing and text is the signification of difference within a syntagmatic web of variegated signs and intertextuality only. Against this flattened ontology they strove to refocus attention on the materiality of writing and the nature of its diachronic emergence. Borges’s notion of the Iliad as “un hecho movil” may resonate with Grésillon’s (1994) aforementioned conceptualization of text as a moving “advent and event”, hecho being an action or a work, a consequence, or a subject (Real Academia Española 2014). However, for genetic critics, the ontology of the text is inextricable from the individual genesis of the work, and each new text is the outcome of a distinct genetic process, inscribed in time. Borges challenges genetic critics of translation to imagine the diachronic movement of text stretching beyond the limits of an author; for him, writing is a collective process that transcends the epistemology of an individual, and texts are continually reformed and rewritten throughout the ages. In the passage quoted above from “Some Versions of Homer”, Borges suggests that designating any text as the original, definitive version is as illusory as identifying any Homeric manuscript (“draft H”), itself the representation of a pre-existing oral text, with the singular Homer, and then using this as justification for its originality and superiority. Indeed, the movement from “draft H” to “draft nine” implies a passage between semiotic systems (alphabetical to numerical) and therefore translation. By naming the imputed original a “draft”, Borges reverses the usual hierarchy of the original and translation, and by repudiating the idea that the subsequent draft is “inferior” to its source, he invites the reader to imagine the value of the H[omeric] text to be dependent upon its translations (variants H-9). Having rejected any imposed hierarchy between original and translation, Borges affirms that all writing is translation, thereby denying the possibility of a definitive text.

How does this compare with the way that genetic critics have theorized the act of translation and its product? Contributors to the first such attempt, Serge Bourjea’s (1995) collection Génétique et Traduction, were inspired above all by the French poet Paul Valéry and his reflections on translating Virgil’s Eclogues: “Faced with my Virgil, I had the sensation (well known to me) of a poet at work … with as much freedom as if it had been a poem of my own” ([1958] 1985, 303). Valéry describes a process whereby translatorship becomes authorship: he retraces the footsteps of the author back to the source inspiration, then proceeds to create the oeuvre anew, composing as a poet. Numerous contributors to Bourjea’s volume find an affinity between this idea and the work of both genetic criticism and translation. In 2014, the flagship journal of genetic criticism Genesis devoted a special issue to translation; its editor reiterated this idea from the inverse perspective, affirming that “[t]ranslation puts genetic procedures into action at every moment” (Durand-Bogaert 2014, 7). In the same special issue, Tiphaine Samoyault (2014) considered the translated text as a subsequent draft of the source’s avant-text; because translations are context-bound and situated in time they are eternally perfectible, they call to be updated; yet for this very fact they reveal the fragility of the oeuvre (its capacity, even its need, to be renewed) and its vulnerability (the erosion of its singular authority). Her proposition is close to Clive Scott’s (2006), who was one of the earliest to introduce critique génétique to Anglophone translation studies in an essay published in Bassnet and Bush’s The Translator as Writer, a volume that made a strong impact in advocating for recognition of the translator’s authorship. Scott argued that translation in fact reactivates the genesis of the oeuvre, it “‘unfinishes’ the ST, multiplies its possibilities of being” (2006, 107); crucially, it renders the original an avant-texte of its translation: “The translator transforms the text of the ST into an avant-texte (draft), transforms the text back into a process of writing, a textualization, or”, referencing Grésillon (1994, 109), “a process of writerly énonciation (rather than the transcription of an énoncé)” (Scott 2006, 107, original emphasis). Scott’s Borgesian manoeuvre maps a semiotic understanding of intertextuality over a genetic description of textualization. He later comments on two ways in which translation is intertextual. In the first, translation brings the source text to bear on the target culture, brokering a relationship between the two: translation applies formal, syntactic, lexical, and semantic pressure upon the target culture. The second form of intertextuality is the “postmodernist vision” in which “intertextuality denotes the textual traffic into and out of a text”, where the text is “unowned” and part of “the kaleidoscopic nature of the literary, of the linguistically permutable” (115). In this second sense, closest to Borges’s view of translation cited above, authority is shared between the semiotic system (texts) and the reader, who brings their own unique experience and intertextual repertoire to the text at hand (or for a translator, to the source text).

In Scott’s view, textual genesis involves a Borgesian movement of writing, within which the ontology of an original is, if not equivalent to, then indistinguishable from its translation. Whether or not this is compatible with a methodology for genetic translation studies is an important question that I will address below. Note first, however, that Scott also writes:

If, alternatively, the translator treats the ST as an avant-texte, then it is by no means evident what form the target text (TT), either as a text or another avant-texte, should take. And behind every manuscript there lurks any number of other, lost manuscripts, and any number of unrecorded hours of mental activity. A text is always what it might be or what it might have been. (Scott 2006, 108; original emphasis)

It is important to recognize that, even if one is aware of the contingent nature of the source text, its dependence upon a reader, and the instability of its signs, this does not free the translator to do whatever they wish with it. Translation is an act of rewriting within which translators exercise their individual creativity, yet this creativity is delimited by the writer’s acknowledgement of their function as either author or translator (self-translators are both, so their translations are not subject to the expectations that non-authorial translators face), as well as their publisher’s expectations and those of the target culture. A translator is constrained by a specific ethical paradigm, which may be formalized in a contractual arrangement or the code of conduct of their professional association or employer, or it may be implicit, arising from their own professional commitment. If the traditional understanding of the ethics of representation has been questioned (e.g. Chesterman 2001; Tymoczko 2000), surveys of working literary translators show that they readily acknowledge their responsibility to represent the source, even if this involves making departures to suit a new audience, modulating style, or adapting certain elements (Jansen 2019). This does not remove the translator’s creativity, rather, it imposes conditions and limits upon it (Bantinaki 2020). If these conditions and limits are by their nature impossible to define exactly, if they are not universal and change with any culture-bound definition of translation, there is then an expectation that the translator can justify every translation decision with reference to the source text and/or its function for a target reader. This does not negate instances when a translator’s ethics cause them to refuse to translate a part of a text or when the source is freely adapted for a new audience – examples are legion, from Western missionaries incorporating, appropriating, indigenous belief systems into their translations of the Bible, to Chinese translations of Dickens that mimic local literary forms, not to mention the notorious French belles infidèles, classical or oriental texts adapted to Enlightenment tastes, and feminist positions that advocate subverting patriarchal structures by “woman-handling the text and actively participating in the creating of meaning” (von Flotow 1991, 76). These are not examples of the negation or failure of translatorial ethics; rather, they render more visible the fact that translators’ decisions always negotiate different options posed by competing ethical systems, socio-political discourses, and commercial pressures.

Building on Toury’s (1995) functional approach – especially, the analysis of translation norms and text reception (in which any text received as a translation is, for all intents and purposes, a translation) – Anthony Pym introduces the notion of a “translation form”, which is the common understanding of translation between “producers, receivers and theorizers on all levels, no matter how paradoxical or unscientific those ideas may be”; the translation form necessarily “reduces complexity so that different social agents may believe they are discussing a shared object called ‘translation’” (2005, 7–8; see also Pym [2011, 2012, 61–67]). Indeed, when translators diverge from this implicit pact they often feel an obligation to disclose their act, and this maintains the integrity of the form itself. The translator’s preface is the usual venue for this disclaimer, but it can be implied by a work’s title, like Robert Lowell’s 1960 book of poems, Imitations, or made explicit with a subtitle, as in Chantal Wright’s English rendition of Yoko Tawada’s (2002) “Porträt einer Zunge”, Portrait of a Tongue: An Experimental Translation (2013). Australian poet John Kinsella offers a more baroque example with his collection, The Jaguar’s Dream: Translations, Adaptations, Versions, Extrapolations, Interpolations, Afters, Takes and Departures (2012). Kinsella’s compatriot Chris Edwards appears even to anticipate the present discussion with the title of his English version of Mallarmé’s famous roll of the dice, A Fluke. A Mistranslation of Stéphane Mallarmé’s “Un Coup De Dés … ” with Parallel French Pretext (2005). His pun on “pretext” as the justification/excuse/alibi for his own transgression echoes the standard English translation of genetic criticism’s avant-texte. Edwards, who is well versed in French literary debates, may be a parodying the notorious mistranslations and mutations of French theory in English (Cusset [2003] 2005), but the pun is funny primarily because it subverts the translator’s claim to legitimacy. Pretext is a false friend; it betrays something of an affinity between Edwards and Scott’s theoretical position given the author’s admission that he has rendered the original a draft, while also, and more firmly, declaring a departure from an ethics of translation. Advertising his error, Edwards implicitly affirms the validity of the translation form and distances himself – in his role as author – from its ethical paradigm.

If a genetic approach to allograph (non-authorial) translation is to posit a Borgesian continuum between the drafting of the “original”, its published form, the drafting of “translation”, and its printed text, it must not only reconcile such textual transmission with the acknowledged functions of each writer (acknowledged by themselves and by their target readers); it needs also to consider the potential for ontological differences between source and target texts. This requires a clarification of how exogenesis and endogenesis play out in translation. These central concepts of genetic criticism distinguish phases in the development of the avant-texte, designating elements that originate outside the writing of the work and those that emerge from within it, respectively (Debray-Genette 1979). While the limits and accuracy of these terms have been debated in genetic criticism for decades (as discussed below), they have not received any sustained treatment with respect to genetic translation studies.

The exogenetics of translation

De Biasi’s functional typology of genetic documentation for the production of literary texts, “What is a literary draft?” (1996), helped systematize the terminology and concepts of genetic criticism internationally. It appeared in English translation in Yale French Studies two years before the French original (1998), which was subsequently integrated into de Biasi’s opus Génétique des textes (2011), and its argument was then reiterated in de Biasi’s contribution to Genetic Criticism: Texts and Avant-textes (Deppman, Ferrer, and Groden 2004), an important collection of essays translated from French into English that is often the first contact English speakers have with the theory of genetic criticism. De Biasi details the process of a literary text’s evolution from avant-texte, to text and post-text through five main phases, i.e. from precomposition to composition, prepublication to publication, and post-publication. He cautions that his typology “should be seen not as the depiction of the actual genetic unfolding along the axis of time, but as the abstract diagram of the logical links that allow us to name and classify genetic documents relative to their function” (1996, 36). The actual writing of a text might advance and retract, the author formulating and reformulating, even erasing and beginning anew, although the documents that bear witness to this process can be categorized within a teleological progression towards publication. De Biasi considers the case of an author who revises their prepublication proofs so significantly that these proofs “function for the writer as a new tool for textualization, in other words, as a rough draft or a corrected fair copy” (36). In this case, the prepublication phase occasions a return to the compositional phase. Indeed, for self-translators, who write and translate their own works, it is not uncommon to change languages during the compositional phases of a text’s elaboration. This may result in a single work or two (or more) texts in multiple languages that maintain the continuum of writing. Jhumpa Lahiri, for instance, recently commented on the process of translating one of her Italian texts into English, during which her published text began literally to resemble a work in progress:

My copy of Dove mi trovo in Italian is a now dog-eared volume, underlined and marked with Post-its indicating the various corrections and clarifications to make. It has transformed from a published text to something resembling a set of bound galleys [pre-publication proofs]. (2021)

Dirk Van Hulle (2011) introduced the term epigenetics to describe the phase when an author modifies a text after its publication, a process sometimes provoked by the experience of self-translation. Lahiri’s Italian text has, she admits, “lost its published patina” and “needs to be opened up again for a few discreet procedures”; no longer primordial, it “now feels incomplete … stands in line behind its English-language counterpart” and “[l]ike an image viewed in the mirror, it has turned into the simulacrum” (2021). In this kind of re/writing, the continuum of textual genesis is bilingual and mobile, stretching across the various language versions produced by the same author, advancing, retracting, and beginning anew. However, genetic criticism is yet to account for the relationship between source and target text in relation to the theory of exogenetics and endogenetics in the case of allograph translation.

De Biasi specifies that exogenetic materials are related to the text but come from outside it – they typically include notes, research materials, vocabulary lists, letters, and diaries – and may be integrated into the endogenetic writing process, that is, when they become part of the drafts, manuscripts, typescripts, or page proofs before the “pass to press” moment, when the author effectively marks the transition of avant-texte to texte, from manuscript genetics to textual genetics. Exogenetic materials are commonly found in the precompositional phase (“a provisional, exploratory, and preparatory process whose operational functions are: orienting, exploring, decision-making, conceiving, and initial planning” [1996, 39]), while endogenetic materials are situated in the compositional phase (a process of structuration, documentation, and composition whose operational functions are: structuring, researching, and textualizing) or the prepublication phase (a postcompositional process whose operational functions are: adding finishing touches and preparing for publication) (42). Remembering that these phases are not necessarily linear – an author can revert back to an earlier phase if the project changes tack – exogenetic materials are more likely to be found in the precompositional phase, that is, before they are integrated into the writing of the work itself by the author or, to employ the more neutral genetic term, the scriptor.³

However, there are some differences in the definition of exogenesis offered by genetic critics that impact a theory of translation genetics. De Biasi specifies that “[e]xogenetics designates any writing process devoted to research, selection, and incorporation, focused on information stemming from a source exterior to the writing” (1996, 43). Crucially, it is the textualization of an external source by the author during the writing process that incorporates it into the genesis of a work. In literary translation, exogenetics will include terminological research and other investigations necessary for resituating the source in its new context and finding an appropriate style; it may also involve reading other texts by the author and/or their translation; there may be past translations of the text at hand; one might consult with the author or colleagues; and, increasingly, literary translators are experimenting with machine translation to produce a first draft, if only for inspiration or to have a proposition to work against. If avant-textes of the source text are available, studying the author’s compositional process will only enrich the translator’s understanding. The translator’s textualization of all this preliminary research is properly exogenetic.

De Biasi stresses that the potential empire of a work’s exogenesis is limited not by the life experience of the author, their reading and environment, but by the written processing of this externality. He states categorically:

exogenetics does not designate the ‘sources’ of the work (such and such a real person, place, literary work, etc.), but the locatable trace of these source-referents in terms of documents (written or transposed) present in the collection of genetic evidential material. (De Biasi 1996, 45)

Exogenetics refers therefore not to the “sources” of a work but “only to written or drawn documents, excluding the empirical objects or data to which they refer” (44). The particularity of translation, however, is the de facto presence of the source text; faced with the task of rendering it into another language, as soon as the translator begins to read the source, they are invariably already cognizing its translation, partially or intensively, “hear[ing] in the text echoes of the translation it is about to become” (Boase-Beier 2010, 32). Yet, if one applies de Biasi’s categories to translation, the source text is itself excluded from the translation’s exogenesis, only becoming part of the avant-texte when textualized by the translator.

Van Hulle offers an alternative definition of exogenesis. Discussing the work of Samuel Beckett, he writes, “[t]he term ‘exogenesis’ denotes external source texts relating to the creative process, such as the texts from which the jottings in the ‘Dream’ Notebook derive” (2011, 802). Here, the sources from which Beckett took his notes are precisely the elements denoted by the term exogenesis, in addition to those notes themselves, all of which are exogenetic to the endogenesis of the novel Beckett wrote at a later date, Dream of Fair to Middling Women. For Van Hulle, Pierre Garnier’s 1894 Onanisme seul et à deux will be an “external source text” for Beckett’s ‘Dream’ Notebook; yet, for de Biasi, the flesh and feathers of Flaubert’s stuffed parrot are not (Flaubert’s beloved bird was indeed a “text” of inspiration, resulting in the heroine of “Un coeur simple” seeing in her own parrot the movement of the Holy Ghost).

In the 2021 issue of Genesis devoted to “Intertextualité – Exogenèse”, de Biasi and Van Hulle open the journal with separate theoretical explorations that illuminate the complexity of the theme, all the while reaffirming their earlier positions with respect to source texts. For De Biasi, exogenesis is once again “any writing process devoted to the work of research, selection and integration relating to information or elements emanating from a source external to the writing” (2021, 21). And for Van Hulle, exogenesis continues to be “made up of external sources that exist independently of the work that the author is writing” (2021, 42). In the latter definition, exogenesis, which is normally considered to be a process, includes external material documents. This difference may have been overlooked to date because it appears at first not to affect the exogenesis of a literary text significantly, yet for the genesis of a translated text, whose entire existence is determined by the reality of an external source text, these questions are primordial.

Van Hulle’s exploration of manuscript genetics from the perspective of cognitive narratology and extended mind theory in Modern Manuscripts (2013) depicts the external world as shaped by and inseverable from the individual’s cognition of it. In this “post-Cartesian paradigm”, Van Hulle writes, “cognitive processes do not exclusively take place ‘in’ the head, but in constant interaction with an external environment. This two-way interaction is regarded as a cognitive system in its own right” (2013, 1). His use of this theory with respect to what genetic criticism terms auctorial (i.e. author-produced) literary manuscripts suggests that he would argue for no clear distinction to be made between external sources and their textualization in the case of allograph translation, just as he has stressed in Manuscript Genetics that “the border between exo- and endogenetics is usually blurred” (2008, 51).

So, while an external source cannot ontologically-speaking be part of the genesis of the work for di Biasi, it can for Van Hulle, who holds that an author’s cognizance of this external object already affects the exogenesis. This does not mean that Van Hulle argues, as Scott does, that translation transforms the original into a draft of the translation – “unfinishes” the source text and returns it to the status of avant-texte (Scott 2016, 107) – only that the original is incorporated into the exogenetic phase. But Scott is careful to contend not that the ontology of the physical source text is transformed, but rather that “the translator transforms the text of the ST” into an avant-texte, and this “multiplies its possibilities of being” (107; my emphasis). The text of the source text has become dematerialized and anonymized; determined by the translator’s reading of its signs, it has no ontological independence, unlike its physical fact. Yet, in the extended-mind thesis, the externality of the original text is a source of traction for the subject’s cognition, its physicality underwrites an experience of the world, and its ontological independence enables that two-way interaction necessary for it to be cognized and translated. Indeed, by recognizing the embeddedness of the source text within its environment when defining exogenesis, genetic translation studies may enter into a productive dialogue with approaches that characterize translation as coextensive with, and even a metalanguage for, ecological (Cronin 2017) and biosemiotic (Marais 2019) systems.

The endogenetics of translation

In dialectical opposition to the exo- (external) of exogenesis, the endo- of endogenesis denotes internal textualizing. “Endogenetics”, for de Biasi, “designates any writing process focusing on a reflexive or self-referential activity of elaborating pre-textual data, be it exploratory, conceptual, structuring, or textualizing work” (1996, 42). Van Hulle uses this term in a similar sense, all the while emphasizing de Biasi’s understated caveat, that there is no clear line that separates the two processes, for their nature is to merge into one another: “Logically speaking, there is no such thing as a purely exogenetic element: every exogenetic fragment bears the primitive seal of endogenetics” (De Biasi 1996, 47). De Biasi’s definition poses challenges for a theoretical description of translation: “[e]ndogenetics is the process by which the writer conceives of, elaborates, and transfigures pre-textual material, without recourse to outside documents or information, through simple reformulation or internal transformation of previous pre-textual data” (1996, 43; my emphasis). If it is not hard to imagine a creative writer whose endogenetic writing nonetheless includes an occasional return to sources (if only to check on a reference or a source of inspiration), those returns are even more frequent in the case of translation. A translation maintains an inerasable dialectic with its source, and, whatever a translator’s approach, the early stages of a translation will involve an intensive textualization of the source text. The first draft is typically a first pass at inhabiting the source, rendering its sense, feeling its rhythm; an attempt to find an appropriate voice and style. According to de Biasi’s definition, translation can only enter an endogenetic phase when the translator no longer has recourse to the source text. However, in one prototypical model of a translation’s genesis, the dialectical relationship between exogenetics and endogenetics will gradually invert with each successive draft, the translator using less and less energy to grapple with the source and more and more energy to refine their recreation of it. Any predominantly endogenetic phase, where writing occurs without recourse to outside documents or information, comes later in the piece and may in fact never arrive if the translator keeps their exogenetic sources always within sight. Indeed, if every exogenetic fragment bears the primitive seal of endogenetics, for translation the reverse is especially true; every endogenetic fragment bears the trace of its primitive exogenetics. This translatorly perspective in fact highlights the reality of genetic processes for all writing, at once confirming another tenet of genetic criticism, that no writing emerges ex nihilo. Furthermore, memory is another medium of recourse, whether previously textualized or not, and through it the endogenesis of a translation maintains – if a translation is to remain a translation – that dialectical relationship with its exogenesis (even if genetic criticism excludes the immaterial, the unverifiable).

The theory of genetic criticism therefore implicitly characterizes allograph translation as a far more exogenetic affair than auctorial writing. However, in a genetics of translation it is not true to say, as de Biasi does when discussing authorship, that, although it is not “exclusively endogenetic, the domain of the rough draft is the endogenetic domain par excellence” (1996, 43). In translation the endogenetic domain can be the place for revising, reviewing, refining the body of a text, the major part of which is textualized through mainly exogenetic processes. Teachers of translation regularly counsel students to make a first draft, walk away, take a break, then pick up the draft again and reread it without reference to the source, revising for fluency. What happens next will vary from person to person – I was taught to reread this version against the source once again, to be sure of my intuitions and to guarantee that I had not inadvertently altered the meaning or introduced errors. This back and forth can continue through multiple phases – a process discussed by translators Tim Gutteridge and Tim Parks (Gutteridge 2019) – which, from a genetic point of view, entails an ebb and flow between exogenetic and endogenetic processes, the recursion and advance necessary for the being of the translation to emerge.

But, in an alternative model of translation genetics, this complex relationship of source text to exogenesis is circumvented in order to characterize the writing of translation as endogenetic. This appears to be implied in Scott’s stimulating proposition that the translator transforms the text of the source back into avant-texte; the translator’s drafting of the source is hereby privileged, it bypasses exogenesis and enters directly into the endogenesis. Yet, given the different functions of author and translator evoked above, it is impossible to construct a seamless genetic transition between these two ontologically distinct text types. The translator who reads a source for translation has a cognitive engagement that produces “drafts” of a different order to an author who may freely adapt the material; issues of ethical and potentially legal responsibility structure a translator’s reading in a way that is absent from other forms of auctorial literary writing. Joyce’s rewriting of Ulysses was hardly constrained by his ur-text, and the same can be true for a film-maker’s interpretation of a script they first receive. Furthermore, behind this theoretical representation of the translation process looms the reality of professional obligations and pressures, financial considerations, and market forces. Recent scholarship on the collaborative dimensions to translation has highlighted the multifaceted, complex, and heterogenous negotiations that translators make with editors, authors, revisors, copy editors, spouses, peers, and authors’ estates (Jansen and Wegener 2013; Cordingley and Frigau Manning 2017; Hersant 2020a). Furthermore, the manuscript the translator submits to “pass for press” may undergo only minor editing by others. Yet, very often it is regarded by those who receive it as a provisional text (and more so than for auctorial texts); it can be significantly edited and revised as it passes through the hands of professional editors and/or translation revisers and proofreaders. In some cases, the revised manuscript is returned to the translator for changes to be approved, but not always. Solum (2018) observes that the translator’s own habitus and symbolic capital, their professional reputation, and other sociological factors affect both the degree of autonomy they exercise over their text and their willingness to defend their translation choices. Peripheral interventions may themselves be classified as exogenetic or endogenetic, following the definitions already established, in which case, the most determinant factor is whether or not the new scriptor consults the source text when introducing changes to the draft (and indeed whether or not they are capable of reading the original). Similarly, the singular translatorship assumed in the discussion thus far, whereby a single person is responsible for the production of the avant-texte, needs to be revised in cases of collaborative translatorship, team translation, and author-translator collaboration, where exogenetic and endogenetic translatorship is shared at different phases of the avant-texte.

A genetic perspective on the translator-author debate

As Paul Valéry translated Virgil’s Eclogues he imagined himself pacing towards the source of the poet’s creativity and communing with the author, like two musical instruments in concert. His metaphor implies his transcendence of linguistic difference, and indeed Valéry’s narrative continues with his descent from these spiritual, musical heights back “down”, to transform ordinary, terrestrial words into this “language of the gods” ([1958] 1998, 305–306). Here, the ideal of the translator is to become an author(-god), to experience what the author experienced through a kind of mystical transportation that allows the translator to recreate the spirit of the work in another tongue. Valéry affirms his fidelity to the meaning of the source text, sacrificing formal elements such as rhyme in his rendition of Virgil’s poems. He writes, “I did not even consider making the alexandrines rhyme, for this would undoubtedly have led me to make too free with the text, whereas I allowed myself scarcely more than a few omissions of detail” (Valéry [1958] 1998, 298). Valéry’s eye is focused resolutely on the text; his fidelity is to it and not the culture of his translation’s reception. For early genetic critics in Bourjea’s 1995 collection (many of whom were Valéry scholars), this particularly source-oriented view of translation offered a neat paradigm with which to hypothesize translator creativity as an exogenetic encounter with a source, followed by its endogenetic recreation in a phase equivalent to authorship. From such a perspective, there may seem little need to develop a theory for the specificity of translation. Yet, just as Valéry’s account is a touch too Romantic, for most translators the exogenetic-endogenetic journey is unlikely to be so teleological, or neatly divided, passing through defined stages like those of a spiritual path or passion.

A genetic perspective on the dialectic between exogenesis and endogenesis in translation can serve to demystify some debates in translation studies. The titles of publications such as The Translator as Writer (Bassnett and Bush 2006), not to be outdone by The Translator as Author (Buffagni, Garzelli, and Zanotti 2011), betoken a desire to raise the status of translators to that of authors and attribute to translators the prestigious “agency” and “creativity” of authority. Such arguments strive to promote the translator out of the secretarial pool and into the corner office (at least in terms of how the translator is perceived); they are felt also to negate the ills of closeminded nationalism and hegemonies that propagate geopolitical and ecological injustice. This trend nonetheless has its detractors. Kahn (2011) presents a more or less common-sense view that translators cannot be authors because they do not create the imaginative world of their text, they transfer it, albeit with creativity. Anthony Pym (2011, 2012, 61–67) differentiates translatorship from authorship on grounds of formal pragmatics (and the results of process research that monitors the behaviour of working translators). He argues that translators are not responsible for their words and so, unlike authors, cannot be held accountable for them. Bantinaki qualifies Pym’s argument: translations are ontologically related to their sources in that they maintain a relationship of “constrained representation” to them, but the translator is accountable only for the “quality of the representation” not what is being represented (2020, 312–313; original emphasis). But why then does she state that her “account attributes authorship to translators: authorship of representations of literary works, rather than authorship of literary works proper” (314)? She had admitted that translators were authors of the qualities of the representation not the representation itself. If Edwards, the poet discussed above, had in fact made an effort to accurately represent Mallarmé’s poem, he would be judged not on what he represented but on the quality of his representation. The Oulipo authors proved that constraints do not inhibit literary creativity, that they can in fact become creativity’s motor. Similarly, to admit the constrained or conditional nature of translatorship in no way negates its creativity; it simply acknowledges the conditions that separate the function of translatorship from that of authorship. To ignore the differences between these practices or the ontological specificity of the texts they produce is to deny the very nature of translation’s encounter with the other, its need to accommodate and surpass linguistic and cultural difference, its ethical paradigm.

But if legal, ethical, and other pragmatic questions are bracketed, and one examines translation archives for signs of how the quality of the representation emerges, there are very many cases where the translator’s more autonomous or self-reliant drafting within the endogenetic phase does not necessarily correlate with their most creative work. A genetic approach to translation need not equate creativity with authority, autonomy, or agency, or treat these terms as synonymous. Translators often resemble authors during their exogenetic grappling with the source text, when they reimage their own encounter with an original for a new audience. Here they will engage in self-reflexive, broad macro-textual and communicative framing, negotiate conventions of genre, and make stylistic choices based upon a comprehension of their own text as a whole, its cohesion, and coherence; factors which inform their inventive or “literary” uses of language at the phrasal level. This does not mean that the first draft comes out in a polished form; it can be very rough or relatively literal, perhaps made quickly not to interrupt the flow of writing, and this too may resemble authorship. Proust’s style in the early drafts of his translation of Ruskin’s The Bible of Amiens shows him composing “with the haste that betrays great discovery” (Tadié 1971, 230). Writing the early drafts of his French version of Truman Capote’s The Grass Harp, Maurice Coindreau put the dictionary aside, leaving the search for specific ornithological, botanical, and other specialized terms for a later moment, because he preferred not to interrupt the verve of creation (Hersant 2020b).

Translation is very much an exogenetic art, but not only: every translation draft bears the weight of both exogenetics and endogenetics, and the portion of each shifts from one draft to the next, often in a non-linear fashion. Accepting this fact, adopting a genetic perspective on creativity, helps overcome cultural anxieties about translation’s ontological difference, which some might overlook in their bias towards unfettered authorship or their desire to emancipate the translator from a perceived subservience. Daniel Simeoni, for instance, implies that a translator’s commitment to an ethics of representation necessarily confines them to political and aesthetic subjugation: “The only space left for creativity and innovation is in the ways chosen for achieving the goals of subservience (nothing to sneer at for sure, but clearly a substitute for higher ambitions)” (Simeoni 1998, 12). Such a view has been countered by Hanne Jansen (2019) in her survey of some 190 members of professional translation associations in Norway, Denmark, and Sweden, who avowed a commitment to both an ethics of representation and the creativity of their art. She concludes that “it could, on the contrary, be argued that not insisting on seeing translators as authors in order to recognize their creativity would be a way of heightening the status of translators and translatorship” (685–686).

In their discussion of collaborative authorship genetic critics Donin and Ferrer (2015, 41) recognize that “genetic criticism has trouble doing away with the concept of the author”, that its focus on the genesis of works is historically an offshoot of Romanticism’s cult of genius, and that “[t]he study of avant-textes participates in a fascination for the mysteries of divine creation understood as a singular emanation”; for them interactions between author and translator belong to an order of succession different in nature from that of primary (co-)authorship, as exemplified by the composer and musician who create a work together from scratch. Ferrer’s (2011, 74–78, 127–129; Bourget and Ferrer 2007) work on the genesis of cinematic texts makes it clear that, for him, the authorship of a film’s director is no less than that of its scriptwriter or editor. Yet does a translator enjoy a freedom equivalent to that of a film director? Comparing literary translation with artforms marked by succession or phases of shared authority, like film and theatre, promises to shed light on the theoretical issues evoked. Without the space to interrogate this proposition here, the evidence examined confirms that the succession or rewriting of literary translation has lacked precise description within genetic criticism, and to simply map genetic theories of literary authorship over translation processes is, at present, an inherently flawed endeavour.

Conclusion: archives in translation

I have argued that a theory of translation genesis must allow the creativity of the translator’s work to be appreciated with respect to the specific ontology of the translated text, namely its emergence within an ineluctable dialectic between exogenesis and endogenesis on the one hand, and ethical, editorial, and commercial constraints on the other, including the work’s capacity to dialogue with its target readership. While many genetic critics, including de Biasi, have examined non-literary forms of creativity, his affirmation that the “rough draft is the endogenetic domain par excellence” (1996, 43) generates a vision of literature emerging from within a writer distanced from sources, which is incompatible with literary translation. It also subordinates forms of creative iteration and variation, like Kinsella’s translations, adaptations, versions, extrapolations, interpolations, afters, takes and departures, not to mention other parodies, spoofs, hoaxes, and the like. The poet Chris Edwards, evoked above for his mistranslation of Mallarmé, revived the avant-garde technique of collage in his 2001 utensils in a landscape and the 2011 People of Earth, rearranging only the words of others into new poetic forms (with hundreds of sources listed in an appendix). Such experiments highlight that, beyond affirming a semiotic understanding of intertextuality, many writers, especially modernists, engage in a poetics of sources that resists the very categorization of writing into exogenetic and endogenetic phases. Perhaps genetic scholars of translation will find a way to surpass these terms altogether, judging them anachronistic and ill-suited to the recursive hermeneutics of literary translation. For now, I propose the following hypothesis, a genetic palliative for the translator as author debate: if one recognizes translation’s ethical paradigm as an ontological limit, one can more easily celebrate the translator’s ingenuity and creativity during the translation’s exogenesis without the anxiety of needing them to emulate or reproduce a model of authorial endogenesis.

Yet further work remains to be done. Translation’s dialogic relation to its source text has not only revealed the need for a theory of translation genetics specific to the ontology of the translated text, but it has also highlighted central ambiguities in the current theorization of exogenesis and endogenesis for all literary works. Under the test of translation – l’épreuve de l’étranger, as Antoine Berman put it – some versions of genetic criticism betray a reluctant bias towards the equation of literary creativity, including literature itself, with individual creation isolated from “sources”. Whether or not Borges is right and all writing is translation, the difficulty of applying the present categories of genetic criticism to the process of translation suggests that it would be useful for genetic scholars of literary authorship to adopt a translator’s perspective. To see how all writing is enmeshed in a textual ecology that is both textual and material is to recognize the fundamentally translatorly nature of any literary endeavour and its product.

Because it is not focused on taking sides in a debate about hierarchies of creativity (of the literary over the translatorly), precisely because it sees itself as the study of processes that are interesting and significant by virtue of their situated differences, genetic translation studies reveals how translation archives – in the indicative – how translation records a work’s transformations, divergences, and regenerations as that work is parsed by cultures who redefine themselves, and it, with each archival event (translating). Translation therefore fulfils the archival functions of being both a domicile for a work over time and a record of the work’s event or genesis, its emergence within time. That record can only ever be an imperfect reflection of the complexity of the creative processes and decision making that gave rise to it, but because genetic criticism is not wedded to a model of writing as the expression of a singular genius, the record of inscription should be understood within the network of other processes that bear upon it, records of which are found in the archives of state institutions, publishers and authors, colleagues and correspondents, friends, partners, and family. A genetic approach to translation may be combined with historical, sociological, and other approaches to the archive, which interpret these records with their specific expertise and insights. Yet, by characterizing the archive as both a domicile and record in translation, a genetic perspective dislocates the archive from its image as a vaultlike edifice of state; it allows the literary work to exist as a moving archive within the panoply of its translational afterlives, its linguistic as much as semiotic and material transformations. Genetic translation studies returns the archive to translation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie [grant number 893904].

Notes on contributors

Anthony Cordingley

Anthony Cordingley is Robinson Fellow at the University of Sydney, on secondment from the Université Paris 8, France where he is Associate Professor in English and Translation. He co-/edited Self-translation: Brokering Originality in Hybrid Culture (Bloomsbury, 2013), Collaborative Translation: from the Renaissance to the Digital Age (Bloomsbury, 2016), and special issues of Linguistica Antverpiensia, “Towards a Genetics of Translation” (2015) and Meta “Translation Archives” (2020). He edits Samuel Beckett Today/Aujourd'hui, authored Samuel Beckett's How It Is: Philosophy in Translation (Edinburgh UP, 2018), and recently completed a Marie Skłodowska-Curie Fellowship project, “Genetic Translation Studies” at KU Leuven's Centre for Translation Studies.

Notes

1 For an overview see Cordingley (2021).

2 This is less true of early theorization (e.g. Bellemin-Noël 1972) than in what the movement came to represent.

3 Grésillon (1994, 245) employed the term scripteur to designate a text’s creator – a term that could be translated into English as writer or author, except that by avoiding both écrivain and auteur, Grésillon was evidently searching for a term that was less ideologically charged, and that could also designate someone using a writing instrument who was not responsible for the content (e.g. an amanuensis).