Embedded Files
What do we mean by "self-translation"?
What do we mean by "self-translation"?
According to contemporary translation theory, "self-translation occurs when an author composes a text in one language and translates it into another” (A. Cordingley, Cambridge Handbook of Translation, 2022). The term "self-translation" did not exist in the early modern period, but the practice itself was common, and other terms or phrases were used to refer to it. For instance, the title page of Antonio Possevino's Apparato all'historia di tutte le nationi (Venice 1598) states that the work was "first published in the Latin language and recently made into Italian by the author himself” (see image). Similarly, Jean Bodin's De republica libri sex (Paris 1586) came with the subtitle: "made into Latin by the author, and much richer than before" (Latine ab autore redditi multo quam antea locupletiores).
Given the multilingual context of early modern culture, self-translation occurred not only in the context of written textual production, but also in oral settings (including preaching). It must also have characterized the inner thought process of authors, who may have thought in one language while writing in another. Such processes of oral or mental self-translation are fascinating and important, but also inherently elusive: thus, our census focuses less on the process of self-translation than on its tangible outcomes in the form of written books, whether manuscript or printed. This, in a nutshell, is the kind of material included in our database.
Above: Antonio Possevino's Apparato all'historia di tutte le nationi (1598), self-translated from his Latin Apparatus ad omnium gentium historiam (1597). Image: GoogleBooks.
Criteria for inclusion
Criteria for inclusion
To be included in WBC, a work must satisfy both of the following conditions:
1) It must exist, or have existed, or have been claimed or planned to exist in two or more language versions that are attributable, entirely or in part, to the same author;
2) At least one of these versions must either be written in French or Italian (linguistic criterion) or have been produced in territories where different varieties of French or Italian were a common language of expression (geographical criterion), or both. Most works in our corpus satisfy both criteria but some only satisfy one (see e.g. the case of John Leslie > Practical guidance, point B).
WBC currently includes 369 "unit works" that satisfy this definition, corresponding to 667 individual language versions.
Above: The concept of "unit work" enables us to distinguish between self-translations and looser cases of bilingual rewriting. To be included in our corpus, it is not enough for two works to share some textual content in different languages; they also need to be presented, or at least demonstrably intended, as versions of each other. For instance, Jean Bodin’s Six Livres de la République (1576, centre) and De republica libri sex (1586, right) were explicitly presented as the same work in two language versions, and thus count for us as a self-translated "unit work". On the other hand, Bodin’s Methodus ad facilem historiarum cognitionem (1566, left) was both conceived and marketed as a distinct work, even though substantial chunks of it later turned up (in French) in Bodin’s République. Following the concept of "unit work", we include the République // De republica - but not the Methodus - in our corpus. Images: GoogleBooks.
Scope and coverage
Scope and coverage
Our corpus does not aim to be all-comprehensive. To keep our scope realistic while providing access to the widest possible range of material, we have made the following preliminary decisions:
1) Prose vs verse
1) Prose vs verse
Our census focuses programmatically on prose as opposed to verse. Poetic self-translation is harder to define and identify, as the formal constraints and possibilities of this medium often result in dramatic textual alterations. Where do we draw the line between self-translation proper and looser forms of bilingual rewriting? This problem, which to some extent also applies to prose materials, is even more acute in the case of verse. Initially focusing on prose will allow us to develop a methodology that we can later apply to the more challenging case of verse. In the future, we do intend to expand our census to also include poetic materials. Note that our census does include some mixed cases, where one version is in prose and the other is in verse (e.g., Étienne Dolet’s Francisci Valesii Francorum regis fata // Les Gestes de François de Valois roy de France; Guillaume Léonard’s Geographia nova // Abrégé de la géographie). Often in these cases the authors themselves justify their choice by stating that prose allows for a more direct equivalence between the two versions.
Above: Guillaume Léonard's bilingual Geographia nova (1655), composed in French prose and facing Latin verses. Images: GoogleBooks.
2) Print vs manuscript
2) Print vs manuscript
Our primary focus is on printed materials, for several reasons. First, preliminary research suggests that printing stimulated a quantitative boom in self-translation, while book market dynamics heavily affected the way in which authors went about self-translating and why they chose to do so in the first place. It is also easier to identify self-translated "diptychs" or "unit works" with respect to printed books, especially with the development of elaborate title pages from the early 16C onwards. Unit works as identified by titles also represented the point of application for the printing privilege system that formed the legal framework of book production in the period. Finally, the advent of print created new incentives for producing self-translations or, at least, for portraying certain works as such. Focusing on printed materials enables us to better study these dynamics as well as to highlight the economic and legal dimensions of early modern self-translation.
3) Subject & genre
3) Subject & genre
Our goal is to map prose self-translations of all kinds, with no restrictions of subject or genre. The range of subjects currently represented in our catalogue shows how pervasive the practice was across all areas of learning and textual production – from theology to the liberal arts, from law and medicine to travel writing, from political controversy to engineering and other types of technical writing. We hope that this approach will allow users to reassess widespread assumptions regarding the relative spread of Latin vs vernacular in different fields, and to achieve a more nuanced understanding of early modern language dynamics.
4) Missing works (lost or projected)
4) Missing works (lost or projected)
Our corpus includes lost works, which were produced but (as far as we know) have not survived to our day [*], as well as works that were planned or projected but probably never produced. While these works do not (or no longer) exist in a physical sense, including them in our corpus allows us to form a clearer picture of historical perceptions and receptions of these materials, as well as to appreciate the incentives that authors and printers may have had in producing - or planning to produce - a self-translation.
[*] Works that are believed to be lost may, of course, turn up at some point. For instance, no copy was known to survive of Antoine Mizauld's Latin Ephemeris for 1557 (for which a French version has been preserved). But a recent search uncovered a copy at the Ludwig Maximilian University Library in Munich (see French corpus for details).
5) Pseudo-self-translations and plagiarized works
5) Pseudo-self-translations and plagiarized works
"Pseudo-self-translation" occurs when an author falsely presents their work as a self-translation. The practice was not uncommon in the early modern period.
In some cases, authors pretended to have self-translated their vernacular works from Latin originals that almost certainly never existed (cf. Pierre Boaistuau and Jean Liébault in the French corpus). In others, they stole somebody else's work and translated it into one or more languages, using (pseudo-)self-translation as a smokescreen to cover up their plagiarism (cf. Joos de Damhouder and possibly Jean Liébault in the French corpus).
Pseudo-self-translation also encompasses cases of false attribution, where a work was more or less deliberately misattributed to an author by others than the author himself, thus creating an impression of self-translation whereas in fact the various versions were the product of different authors (cf. Leonardo Bruni in the Italian corpus).
In some instances, it can be difficult to decide whether a work is missing because it went lost or because it was entirely fictitious. The debates surrounding Liébault's works (see French corpus) are illustrative of these difficulties.
Above: a probable case of "pseudo-self-translation". Pierre Boaistuau pretended to have translated his Théâtre du monde from his own Latin original, but the latter almost certainly never existed (see French corpus for details). Image: GoogleBooks.
Above: Joos de Damhouder's Enchiridion rerum criminalium (1555) presents us with a unique case of multiple (pseudo-)self-translation combined with plagiarism. Damhouder “stole” an unpublished Dutch manuscript by 15th-century jurist Philips Wielant, translated it into Latin, and published it under his own name in 1555. A French version, explicitly marketed as Damhouder's self-translation, followed shortly thereafter. Finally, Damhouder also published a Dutch version that heavily relied on Wielant's original but was once again presented as Damhouder's own work. The imperial privilege secured in 1552 explicitly covered the work in all three languages (“tam Latino, Gallico, quam Teutonico idiomate” (see French corpus for details). Images: GoogleBooks.
6) "Self-translations" and "semi-self-translations"
6) "Self-translations" and "semi-self-translations"
Our database includes both "self-translations" (produced by the authors themselves) and so-called "semi-self-translations" (Dasilva, 2016), which feature some kind of collaboration between the author and other actors.
Collaboration – whether acknowledged or invisible – was central to early modern textual production, particularly (but not exclusively) in the realm of print. For authors wanting to issue their own works in multiple languages, self-translation was not the only option available: they could choose to delegate parts of the translation process to others (often students, family members, secretaries, or close associates), while reserving the right to authorize the draft for publication.
A significant number of works in our corpus are collaborative in this broad sense. They range from translations genuinely produced in tandem by an author and his helpers (collaborative translations, e.g., Sebastian Castellio’s De l’impunité des hérétiques) to translations that were 100% allographic but published with the consent of their original authors (authorized translations, e.g., Antonio Possevino’s Moscovia, translated into Italian by his nephew Giovanni Battista; or Guillaume Rondelet’s Libri de piscibus marinis, translated into French by Rondelet’s student Laurent Joubert). Other cases sit somewhere along this spectrum of authorial agency, being largely allographic but reviewed and sometimes revised/augmented by the original authors (supervised translations): for instance, three famous works by René Descartes (Discours de la méthode, Meditationes de prima philosophia, Principia philosophiae) were translated by friends or collaborators, but Descartes carefully reviewed their work and not only made changes and improvements but in some cases added more content, including an important new preface for the French Principes. Finally, some translations were largely or entirely authorial, but were revised by somebody else prior to publication (revised by allographic translator, e.g. Pietro Bembo's Historia vinitiana or Charles Estienne's Agriculture et maison rustique)
While including such works in a database of self-translations might seem objectionable, leaving them out is even more problematic. For:
1) how can we be sure that even works marketed as ‘pure’ self-translations did not in fact result from invisible and unacknowledged collaborative processes?
2) Many of these works do after all contain important authorial elements, even though in many cases it can be difficult to distinguish the author’s contribution from that of other actors.
It should also be noted that these works were all marketed as fully authorized, and thus partook of the same authorial aura that characterized "pure" self-translations and (arguably) made them palatable to contemporary readers.
Above: Etienne de Courcelles's Latin translation (Specimina philosophiae, 1644) of René Descartes' Discours de la méthode, avec les Essais de cette méthode (1637) is a paradigmatic case of allographic supervised translation. The work contained a preliminary note by Descartes himself, which stated: “These essays, which I wrote in French and published seven years ago, were shortly afterwards translated by a friend into Latin, and this translation was handed over to me in order that I may freely change anything that displeased me in it. And so I did, in various places, but perhaps I neglected many others still; and one will be able to distinguish the latter from the former in that the faithful translator has nearly everywhere attempted to translate word for word, whereas I have often changed down to the very sentences, and I have made an effort to correct, not the translator’s words, but my own meaning everywhere.” Corinne Vermeulen has carefully studied the nature of these changes in her 2007 PhD thesis (Utrecht). Image: GoogleBooks.
Page updated
Google Sites
Report abuse